f/EPA
United States
Environmental Protection
Agency
Environmental Sciences Research |
Laboratory
Research Triangle Park NC 27711
Research and Development
EPA-600/S2-81-070 July 1981
Project Summary
Development of Measurement
Methodology for Evaluating
Fugitive Particulate Emissions
Edward E. Uthe, John M. Livingston, Clyde L. Witham, and Norman B. Nielsen
An experimental study was con-
ducted to demonstrate a measure-
ment methodology for evaluating
fugitive particulate emissions. The
program focused on the application of
the lidar (laser radar) technique under
field conditions but in circumstances
that simplified and controlled the
variables of the general problem.
The lidar was used to make elevation
scans perpendicular to an aerosol
plume generated by controlled release
of particulate material into the atmos-
phere. The lidar backscatter signa-
tures were processed in terms of
cross-plume integrated backscatter,
and these values were related to inde-
pendently measured particulate
emission rates. A very high correlation
was obtained between time-averaged
lidar observations and emission rates
(correlation coefficients of 0.9 or
better in most runs), with substantially
less correlation for individual lidar
observations. Relatively high correla-
tions were also obtained between
smoke-reader data on downwind
plume opacity and smoke emission
rate as well as lidar backscatter. For
dense smoke, attenuation of the lidar
energy was shown to be of importance
in interpreting data in terms of smoke
concentration.
Finally, the lidar was used at the site
of an actual fugitive particulate source
to demonstrate that appropriate data
can be collected for measuring source
emission rates.
This Project Summary was develop-
ed by EPA's Environmental Sciences
Research Laboratory, Research Tri-
angle Park, NC, to announce key find-
ings of the research project that is fully
documented in a separate report of the
same title (see Project Report ordering
information at back).
Introduction
The total source strength of pollution
emitted by industrial plants is the aggre-
gate of all diffuse and minor specific
emissions as well as major identifiable
point sources. Therefore, for many
plants, measurement of individual
emissions from a multiplicity of sources
is neither economical nor practical. The
only feasible approach is to measure, as
accurately as possible, the concentra-
tion throughout a cross section of the
downwind plume of the combined fugi-
tive emissions, to integrate these, and,
from a measurement of the integrated
wind velocity through the plume of the
cross-section, to calculate the pollutant
mass flow.
The problems of accomplishing such
measurements with existing technol-
ogy are many. Specifically, with in-situ
samples it is virtually impossible to
characterize adequately the concentra-
tion of particles throughout the total
cross section of the plume, to relate any
measurements made to the plume's
total envelope, or to determine its
extent. This is especially the case above
-------
the surface, because of the extended
and variable nature of the multiple
sources of fugitive emissions.
Lidar observation fulfills, as no other
method does, the requirement for
delineating the spatial distribution of
elevated particulate pollution plumes
and for readily distinguishing between
pollution background and pollution from
the source being studied. Although
there are limitations and difficulties in
using lidar backscatter measurements
for determining absolute particulate
concentrations, it is possible to
evaluate, with useful accuracy, the
near-instantaneous distribution of
particulate material within a selected
cross section or envelope. It is thus
possible to obtain a series of such cross
sections in time and, from a measure-
ment of mean wind velocity and infor-
mation on the backscatter-to-mass-
concentration ratio, to derive an
estimate of source emission rate.
In the present study, three experf-
mental field programs were conducted
to demonstrate a lidar methodology for
measuring fugitive particulate emis-
sions. The first two field programs used
controlled release of various types of
particulate material to simulate
emissions from fugitive sources. A lidar
system was used to make elevation
scans perpendicular to the transport
direction of the aerosol plume about
500 m downwind from the source. The
lidar backscatter signatures for each
elevation scan were processed for
values of cross-plume integrated back-
scatter, and these values were related
to independently measured particulate
emission rates. In the second field test,
a trained smoke inspector made down-
wind plume-opacity readings in addition
to the lidar observations. The third field
program was conducted to demonstrate
use of the lidar technique at the site of
an actual fugitive emission source.
Results of this study showed that
downwind measurements of lidar
cross-plume integrated backscatter and
smoke-reader plume opacity generally
increase linearly with particulate
emission rate. Relatively high correla-
tion coefficients between these
measured quantities demonstrate that
lidar and smoke reader provide two
possible methods for evaluating fugitive
particulate emissions.
Procedure
A series of field experiments was
designed to demonstrate the method-
ology by making lidar measurements of
an aerosol plume generated by continu-
ous release of particulate material of
known properties into the atmosphere
at known rates. The lidar system used
was SRI International's Mark IX, which
is van-mounted complete with data-
processing and power-generating
capabilities. Figure 1 is an example of an
intensity-modulated video display
depicting cross-plume aerosol structure
observed by scanning the lidar in eleva-
tion. Computer-generated vertical
concentration profiles are plotted on the
cross section for locations indicated by
the cursor marks drawn above the
plume return. Similarly, the backscat-
tered data can be spatially integrated to
determine a relative cross-plume
density.
The Mark IX lidar was used to make
cross-plume observations downwind
from controlled emission sources with
known particulate properties, as shown
by the example presented in Figure 1.
The lidar typically observed the plume
from a distance of about 300 to 500 m,
about 200 to 500 m downwind from the
source. On some experimental data
runs, the trained smoke inspector made
plume-opacity readings near the emis-
sion source; on other runs he made
readings at downwind distances corre-
sponding to the lidar observations.
Three methods of aerosol generation
were used. An aerosol generator was
constructed for releasing fine silica
powder in the atmosphere at 1-m and
10-m heights. The powder emission
rate was controlled by a grooved-disk
feeder.
A second method of aerosol genera-
tion used a smoke generator operated
by the State of California Air Resources
Board for certifying smoke inspectors.
Both white smoke (produced by vapor-
ization and condensation of diesel fuel)
and black smoke (produced by incom-
plete combustion of toluene) could be
emitted through a 10-m modified stack.
The white smoke was found to evapor-
ate downwind from the source; there-
fore, only black smoke was used in the
experiments. The emission rate was
controlled by the fuel combustion rate,
and smoke quantity was evaluated with
an in-stack white-light transmissom-
eter calibrated in terms of mass
emission.
The third method of smoke generation
consisted of igniting zinc chloride
smoke pots and candles. The mass
emission rate was determined by
experimentally evaluating the emission
from a single pot and candle and multi-
plying by the number of pots or candles
(1, 2, 4, or 8) ignited simultaneously.
Results and Discussion (
The experimental results of this study
show linear relationships of relatively
high correlation among the quantities of
downwind lidar cross-plume backscat-
ter, smoke-reader plume opacity, and
particulate emission rates. This is illus-
I
-j
2
o
-Q
I
Horizontal Distance From Lidar
Figure 1. Example of computer-generated profiles of vertical plume density.
Lidar is located at lower left corner. The height and distance scale is
75 m/div. Vertical concentrations of the plume (relative to clear air, with ,
a scale of JO dB/div) are plotted at the lower left and the horizontal
position associated with each profile is plotted in the upper right.
-------
rated by the example presented in
Figure 2 for data collected during a
smoke candle experiment. This example
and other data presented in the final
report demonstrate the feasibility of
using either lidar or smoke inspectors
for estimating participate emission from
fugitive sources.
Relatively large scatter of data occur-
red, probably caused by downwind
aerosol density fluctuations introduced
by turbulent transport. Linear correla-
tion coefficients squared for all data
points (R,2) collected during an experi-
mental run typically ranged from 0.6 to
0.9. By averaging data from three or four
lidar scans at each smoke concentration
(requiring a 5-to 10-minute period), the
correlation coefficients squared (Rm2)
typically were greater than 0.9. There-
fore, the data clearly show that time-
averaged measurements of the down-
wind plume are required.
For high-density smoke generated by
igniting smoke pots, significant attenu-
ation of the laser energy was evident. A
factor-of-two increase in the emission
rate resulted in substantially less than a
factor-of-two increase in the cross-
plume integrated lidar backscatter.
_ Experiments with lower-density smoke
provided the expected one-to-one
jorrespondence between lidar
response and smoke emission rate.
Plume opacities derived from the lidar
data by analyzing the clear-air returns
on the far side of the plume returns
were substantially less tha'n those esti-
mated by the smoke reader. Correction
applied because of .the longer wave-
length lidar and submicrometer smoke
particles explained only a part of the
difference between lidar and smoke-
reader observations.
A field program was conducted to
demonstrate the use of the lidar system
at an actual fugitive emissions source.
The lidar was used successfully to make
downwind vertical scans across the
particulate plume generated by the
Permanente Cement Plant located in
Cupertino, California. The data were
collected in the same way as the test
smokes and therefore could have been
processed in the same way as the test
data to estimate plant particulate
emission rates.
Conclusions
The major conclusions of this study are:
• Cross-plume integrated back-
scatter evaluated from data
collected by scanning a lidar
system across a downwind smoke
plume behaved predictably with
variations in particulate emission
rate.
• Correlations between downwind
lidar observations and source
emission rates were greatly
increased by averaging data from
multiple lidar observations made
at each emission rate (three or
four lidar scans requiring 5 to 10
minutes).
• For low-density plumes, the lidar
responded linearly in a one-to-
one ratio with changes in particu-
late emission rate. Linear correla-
tion coefficients greater than 0.9
were obtained in most cases.
• For higher-density plumes, the
lidar signal increased with
increasing particulate emission
rate, but at less than a one-to-one
ratio because of extinction
processes.
• Correlations between downwind
lidar observations and particulate
emission rates were only slightly
improved when corrections for
wind speed variations (measured
at the lidar site) were applied.
• Testing of lidar methodology was
best accomplished using
commercially available smoke
candles. The quantity of smoke
was controlled by the number of
units ignited.
• Downwind plume opacities evalu-
ated by a trained Method-9 ob-
server were highly correlated with
particulate emission rate and with
the lidar cross-plume integrated
backscatter.
• A mobile lidar system can suc-
cessfully make appropriate cross-
plume observations at actual
fugitive particulate emission
sources.
Recommendations
This study demonstrated a methodol-
ogy, using the lidar technique for
measuring fugitive particulate emis-
sions. Several additional studies
suggested to further develop and
demonstrate the lidar technique for this
purpose are discussed below:
• In-Situ Measurement of Back-
scatter-to-Mass-Concentration Ratio
—The study showed that the
cross-plume integrated backscat-
ter evaluated by scanning a lidar
system across an aerosol plume"
responds predictably with particu-
late emission rate. However,
measurement of the absolute
emission rate requires calibration
of the lidar backscatter in terms of
aerosol concentration. An in-situ
measurement of the backscatter-
to-mass-concentration ratio
would provide the needed calibra-
tion. In addition, the measure-
ment of the variability of this ratio
for different types of particles
would provide an estimate of the
accuracy of lidar measurement for
the case of particle characteristics
changing in the vertical.
Therefore, an instrument that
measures absolute values of
backscatter-to-mass-concentration
should be developed.
• Airborne Lidar Measurement of
Particulate Emissions from Large-
Area Sources—Many fugitive
emission sources, including coal
mining, oil refining, and cement
plant operations, generate a
particulate plume with a large
horizontal and vertical extent.
From these sources, the pollution
plume is more readily observed
with an airborne lidar than with a
surface-based system. A
downward-pointing lidar could be
flown along a path that encom-
passes the plant site to observe
upwind and downwind particulate
flow. Observations made in com-
plex terrains would be greatly
simplified as compared with those
using a surface-based lidar. A
two-wavelength airborne lidar
system has recently been demon-
strated.* Backscatter data at two
wavelengths may provide the
necessary information to estimate
absolute mass concentration of
observed aerosols. An airborne
lidar system should be considered
for measurement of particulate
fugitive emissions.
*Uthe, E E , N B Nielsen, and W Jimison, "Air-
borne Lidar Plume and Haze Analyzer (ALPHA-1),"
Bull. Am Mel Soc, 61:1035-1043, 1980
-------
200
.§ /50
c
3
0)
'^5
•2
*
^
i wo
fc
t/J
J<
G
k.
3 50
.
1
.
_
^
_ <
t
0
A"
•
, . , .
4
/
#
s
0
t
t
*
* •
*
«
*
«
/' L
j X
/'
•
* —
/
•* -
.•
•i
^
R? = 0.83
• • m —
1^1^
0.98
1
L
2000
| 1500
;
3
gj
<
"i
i
k
!
i
WOO
0
c
i
-4
c
0
0
j
0
J
3
OQ
I
CO
T
°0 20 40 60 80 100
Emission Rate — $r m/'n 1
fa) Cross-plume integrated backscatter
as a function of emission rate.
inn j-
i i/i/
90
80
1
se 70
!
8 60
'S
I
I*.
1
«c
1 40
-------
Choice of Lidar Wavelength or
Wavelengths— Successful use of
the lidar technique for remote
measurement of particulate con-
centrations require knowledge of
the relation between aerosol
optical and physical parameters.
Because fugitive emissions
frequently are comprised of emis-
sions from several source types
and the percentage of particulate
from each type may vary, the
particle characteristics (size,
shape, and composition) may also
vary in both space and time. The
backscatter-to-concehtration ratio
is dependent on these particle
characteristics and, therefore,
a fugitive-emissions lidar system
m.ust be designed to be insensitive
to changes in particle characteris-
tics. Experiments should be
conducted to establish the proper
wavelengths for minimizing the
effect of particle characteristics
on the backscatter-to-concentra-
tion ratio.
Edward E. Uthe. John M. Livingston, Clyde L Witham, and Norman B. Nielsen
are with SRI International. Menlo Park, CA 94025.
William D. Conner is the EPA Project Officer (see below).
The complete report, entitled "Development of Measurement Methodology for
Evaluating Fugitive Particulate Emissions." (Order No. PB 81-196 594; Cost:
$8.00, subject to change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield. VA22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Environmental Sciences Research Laboratory
U. S. Environmental Protection Agency
Research Triangle Park. NC 27711
I US GOVERNMENT PRINTING OFFICE 1861-757-012/7164
-------
United States Center for Environmental Research Fees Paid
Environmental Protection Information Environmental
Agency Cincinnati OH 45268 Protection
Agency
EPA 335
Official Business
Penalty for Private Use 5300
?SS ISVIR^RO^C-IIOH RGKNCi
------- |