Final Report
STUDY OF LASER 6ACKSCATTER BY
PARTICIPATES IN STACK EMISSIONS
By: EDWARD E. UTHE and CHARLES E. LAPPLE
Prepared for:
ENVIRONMENTAL PROTECTION AGENCY
NATIONAL ENVIRONMENTAL RESEARCH CENTER
DIVISION OF CHEMISTRY AND PHYSICS
RESEARCH TRIANGLE PARK
NORTH CAROLINA 27711
CONTRACT CPA 70-173
-------
xJ\ v 7 /jt
jr Zgfifz
F/V?a/ Report January 1972
STUDY OF LASER BACKSCATTER BY
PARTICULATES IN STACK EMISSIONS
By: EDWARD E. UTHE and CHARLES E. LAPPLE
Prepared for:
ENVIRONMENTAL PROTECTION AGENCY
NATIONAL ENVIRONMENTAL RESEARCH CENTER
DIVISION OF CHEMISTRY AND PHYSICS
RESEARCH TRIANGLE PARK
NORTH CAROLINA 27711
CONTRACT CPA 70-173
SRI Project 8730
Approved by:
R. T. H. COLLIS, Director
Atmospheric Sciences Laboratory
RAY L. LEADABRAND, Executive Director
Electronics and Radio Sciences Division
-------
ABSTRACT
This experimental study investigates the validity of determining
smoke plume opacity or particulate content from measurements of the
backscatter of laser radiation from plume particulates. The backscatter
experiments were conducted with the use of a specially designed aerosol
chamber that allowed the experimental geometry to simulate that assoc-
iated with actual remote plume probing.
The backscatter experiments used fly ash collected from a bituminous
coal-burning power plant. The raw fly ash was classified into various
size fractions and pneumatically injected into the chamber at controlled
concentrations. The aerosol was continuously monitored by a white light
transmissometer along the axis of the chamber. Lidar measurements were
made to determine the backscatter and transmission of the generated
aerosols.
At a wavelength of 0.7u, the backscatter-to-extinction ratio was
essentially independent of particle size, and the backscatter-to-mass
concentration ratio was inversely proportional to the volume-to-surface
mean diameter. At a wavelength of 1.06j_L, the backscatter-to-extinction
ratio was more dependent on particle size and the backscatter-to-mass
concentration ratio was less dependent on particle size than at the
0.7(j, wavelength. Transmission data at both lidar wavelengths agreed
with the white light transmission data. Mie theory computations of
backscatter and extinction values agreed reasonably well with observed
data.
These results for fly ash indicate that plume opacity may be deter-
mined from lidar backscatter measurements at 0.7^ wavelength, but that
mass concentration is better inferred from 1.06^ lidar backscatter
measurements. At either wavelength, lidar measurements of transmission
may be used to determine plume opacity.
iii
-------
ACKNOWLEDGMENTS
The authors are grateful to Mr. W. D. Conner of the Environmental
Protection Agency for his many valuable suggestions at various phases
of the study.
We also acknowledge the assistance of the following Stanford Re-
search Institute personnel: John Oblanas, for his design of the aerosol
chamber optical sensors and directing a portion of the data collection;
William Dyer, Conrad Schadt, Earl Scribner, William Evans, Robert Allen,
Ilsabe Niemeyer and August Pijma, all of whom participated in the data
collection; David Jackson and Raymond Cummings for their discussions on
CW laser techniques; and Ronald Collis, Director of the Atmospheric
Sciences Laboratory, for his helpful discussions during the course of
this project.
-------
CONTENTS
ABSTRACT iii
ACKNOWLEDGMENTS v
LIST OF ILLUSTRATIONS ix
LIST OF TABLES xiii
I INTRODUCTION 1
II SUMMARY AND CONCLUSIONS 3
A. Experimental Facility and Procedures 3
B. Data Analysis 4
C. Results and Conclusions 5
III DISCUSSION OF EXPERIMENTAL PROGRAM 9
A. Experimental Facility 9
B. Experimental Procedures 9
C. Experimental Results 16
D. Comparison of Experimental Results and Mie Theory . . 25
IV TECHNIQUES OF PLUME EVALUATION 37
A. Comparison of Opacity and Backscatter
Measurement Techniques 37
B. Laser Pulse Length Considerations 41
C. Polarization Measurements 43
D. CW Laser Techniques 44
1. Unmodulated CW Laser System 45
2. Pulse Modulation CW Laser System 46
3. FM CW Laser System 48
VII
-------
CONTENTS (Continued)
V RECOMMENDATIONS FOR FURTHER RESEARCH 51
A. Development of a Low Cost, Remote Plume Sensor .... 52
B. Experimental Evaluation with Other Particulates ... 52
C. Aerosol Chamber Improvement for Aerosol Studies ... 53
D. Extending the Accuracy of Backscatter Measurements
Made with Pulsed Lidars 54
E. Computations from Mie Theory 55
F. Development of Multiple Wavelength Lidar
Techniques 55
REFERENCES 57
Appendix A Details of the Sighting Tunnel System A-l
Appendix B Fly Ash Preparation and Properties ..... B-l
Appendix C Dust Concentration Control and Measurement C-l
Appendix D Aerosol Chamber Optical Components D-l
Appendix E Transmissometer Data E-l
Appendix F Lidar Instrumentation F-l
Appendix G Lidar Signature Analysis G-l
Appendix H Lidar Data Summary H-l
Vlll
-------
ILLUSTRATIONS
Figure 1 Diagram of Experimental System
Figure 2 Photographs of Laser Beam Size
at Chamber Location
10
12
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7
Figure 8
Backscatter Quantities Observed
at 0.7 (j, Wavelength as a Function
of the Extinction Coefficient Observed
by the Transmissometer
Volume Backscatter Coefficients ( |3) Oberved
at a Lidar Wavelength of 0.7 p Related
to Particulate Mass Concentration (M)
and White-Light Volume Extinction
Coefficients (o~)
Volume Extinction Coefficients (a) Observed
at a Lidar Wavelength of 0.7 p Related
to Particulate Mass Concentrations (M)
and White-Light Volume Extinction
Coefficients
Volume Backscatter Coefficients (c) Observed
at a Lidar Wavelength of 1.06 p Related
to Particulate Mass Cencentrations (M) and
and White-Light Volume Extinction Coefficients
Coefficients ( CT)
Volume Extinction Coefficients (a) Observed
at a Lidar Wavelength of 1.06 u Related to Par-
ticulate Mass Concentration (M) and White-Light
Volume Extinction Coefficients
Spherical Particle (Mie Theory) Backscatter Ef-
ficiency Factors (Backscatter Cross Section)
Geometrical Cross Section) as a Function
of Particle Size at 0.7 p, and 1.06 p Wavelengths
for Several Values at the Complex Refractive
Index
17
21
22
24
26
27
IX
-------
ILLUSTRATIONS (Continued)
Figure 9 Observed and Computed Extinction and Backscatter
Efficiency Factors as a Function of Size Param-
eter Computed from the Average Sauter
Diameters 34
Figure 10 Plume Opacity as a Function of Particulate Volume
Extinction Coefficient and Observed Plume
Length 39
Figure A-l Aerosol Chamber Used in the Backscatter
Experiment A- 8
Figure A-2 Aerosol Chamber Details A- 9
Figure A-3 Dust Dispenser Details A-ll
Figure B-l Flowsheet for Classification Processes B- 8
Figure B-2 MSA-Whitby Particle Size Analyses
of Blackdog Fly Ash B- 9
Figure B-3 Coulter Counter Particle Size Analyses
of Blackdog Fly Ash B-10
Figure B-4 Photomicographs of Blackdog Fly Ash B-ll
Figure B-5 Additional Photomicrographs of Blackdog Fly Ash
at Lower Magnification B-12
Figure B-6 Scanning Electron Micrographs
of Blackdog Fly Ash B-13
Figure B-7 Scanning Electron Micrographs of Fly Ash Deposits
on Sampling Filter Papers B-14
Figure C-l Sampling and Filter Arrangement C-16
Figure C-2 Delren Filter Holder C-17
Figure C-3 Bag Sampler System C-18
Figure C-4 Sampling Locations C-19
x
-------
Figure C-5
ILLUSTRATIONS (Concluded)
Typical Transmissometer Strip Chart Records
of Various Dust Sizes and Concentrations
. C-20
Figure D-l Optical Diagram of the Transmissometer
and Angular Scattering Sensors D- 6
Figure D-2 Data Flow Diagram of the Optical Scattering
Sensors D- 7
Figure D-3 Data Format of the Optical Scattering Sensors . . . D- 8
Figure E-l Effect of Fly Ash Concentration and Size
on Optical Transmission E-10
Figure E-2 Generalized Correlation for
Optical Transmission E-ll
Figure F-l SRI Mark V Lidar F- 5
Figure F-2 Diagram of Mark V Lidar System and Data Recording
for Backscatter Experiments F- 6
Figure G-l Examples of Lidar Signatures (reproduced
from Polaroid prints) for Various Mass Concen-
tration of the 0 to 2.5 ^ Diameter Fly Ash
Fraction (0.7 p, Wavelength Lidar) G- 7
XI
-------
TABLES
Table 1 Observed and Computed Aerosol Parameters 30
Table A-l Dust Feeder Calibration Data A-12
Table A-2 Effect of Compressed Air Pressure
on Dust Dispersion A-13
Table A-3 Summary of Dust Collected Inside Aerosol Duct-
work and Tunnel A-14
Table B-l Summary of Size Analysis Data
on Blackdog Fly Ash B-15
Table B-2 Summary of Mean Particle Diameters
of Blackdog Fly Ash B-16
Table B-3 Log-Probability Approximations
to Size Distribution of Each Fraction
of Blackdog Fly Ash B-17
Table B-4 Recommended "Average" Size Distribution
for Each Fraction Used in Study B-18
Table B-5 Typical Fly Ash Analyses B-19
Table C-l Summary of Initial Series of Sampling Runs .... C-21
Table C-2 Summary of Second Series of Sampling Runs C-26
Table C-3 Summary of Concentration Ratio Data for Sampling
and Filter Arrangement "T" C-32
Table C-4 Summary of Concentration Ratio Data
for Second Series of Sampling Runs C-33
Table C-5 Summary of Weather Measure Sampling Data C-34
Xlll
-------
TABLES (Concluded)
Table E-l
Table E-2
Table F-l
Table H-l
Table H-2
Summary of Transmissometer Data E-12
Comparison of Measured
and Calculated Average Efficiency Factors . .
Characteristics of SRI/EPA Mark VIII
and SRI Mark V Lidar
E-13
F- 7
Lidar Data Summary H- 3
Digital Data from Individual Runs H- 5
xiv
-------
I INTRODUCTION
Enforcement of increasingly stringent air pollution laws governing
emissions from pollution sources requires the development of objective
direct- and remote-sensing techniques. One major problem is the remote
evaluation of aerosol density emitted from stationary sources in terms
of legally acceptable parameters. Subjective plume opacity reading by
a human observer (the Ringelmann procedure) is the method used by most
pollution control agencies. The present study is concerned with certain
aspects of the development of an instrument for the objective measurement
of plume opacity or density.
Remote measurement of stack effluent opacity using a pulsed laser
technique has been described and demonstrated by Evans (1967), Conner
and Hodkinson (1967), and Cook et al. (1971). Single-ended measurements
of laser energy transmission through stack plumes are made by comparing
"clear air' backscatter returns from the near and far side of the plume
return. However, to obtain the necessary signal levels, a relatively
expensive pulsed lidar with high transmitter energy and a highly sensi-
tive low noise receiver is required. In addition as Evans reported, sensi-
tive light-receiving photomultiplier tubes may become "saturated" from the
plume return and may not recover sufficiently to make accurate measure-
ments of the clear air backscatter from the far side of the plume.
An alternative approach, the remote inference of plume density by
measurement of elastic backscatter of laser radiation from plume partic-
ulates, would place fewer requirements on the lidar components. How-
ever, the inference of plume particulate content from the laser radiation
backscattered from these particulates is susceptible to errors originating
-------
from inherent uncertainties of particle backscatter cross sections
(Twomey and Howell, 1965).
The objective of this experimental study was to investigate laser
backscatter from plume particulates in terms of plume opacity, trans-
mission at laser wavelengths, and mass concentration. The collected
data are presented and analyzed in terms of developing a low cost, plume-
measuring instrument.
-------
II SUMMARY AND CONCLUSIONS
The experimental facility and lidar techniques used during this
study and the results obtained are described briefly below. Conclusions
of significance to the development of a low cost plume-monitoring instru-
ment are presented.
A. Experimental Facility and Procedures
A major effort of the present study was to design and assemble an
experimental facility to be used for the evaluation of laser backscatter
techniques from aerosols consisting of particulates of known character-
istics and in known concentrations. The following experimental work
was performed:
(l) A special aerosol chamber was designed and constructed.
The chamber dimensions (20 in. square and 30 ft long) are
sufficient to enclose completely a Q-switched laser pulse
and allow the pulse to propagate through the test aerosol
without interaction with any chamber surfaces.
(2) Raw fly ash was collected from a coal-burning power plant
and was divided into various size fractions in the 0 to
10 u diameter range. The classified fractions were analyzed
by the Fisher Sub-Sieve Sizer, MSA-Whitby Sedimentation
Analyzer, and Coulter Counter techniques to determine their
particle size distribution.
(3) A white-light transmissometer and a 90° side-scattering
photometer were constructed and mounted in the aerosol
chamber for continuous aerosol monitoring. The optical
instrumentation was specially designed to reduce multiple
scattering effects and not to interfere with propagation
of the laser pulse through the test aerosol. (The 90 side-
scatter photometer was not extensively used because of low-
signal-to-noise problems during daylight conditions.)
-------
(4) The fly ash fractions were pneumatically injected into the
chamber at various concentrations by a rotating-disk par-
ticle feeder. The particle feeder was calibrated in terms
of mass concentration for each fly ash fraction by separately
measuring dust rates from the feeder at various feeder speeds.
(5) Extensive particle-sampling was conducted at various locations
within the chamber to assess the degree of uniformity of par-
ticulate concentration.
(6) The white-light transmissometer data were calibrated in terms
of particulate mass concentration for each fly ash fraction
by relating time-averaged transmissions to the calibrated
particle feed rates.
(?) Lidar observations of the test aerosol were made 500 ft from
the aerosol chamber. The lidar backscatter signatures were
recorded on Polaroid film using several oscilloscope displays
to provide a linear reading of return signals that can vary
over several orders of magnitude. Concurrent lidar and
chamber measurements of aerosol concentration were conducted
as a function of particulate concentration, particulate size
(using the various fly ash size fractions), and lidar wave-
length (0.7 and 1.06 u,).
B. Data Analysis
Each lidar observation (firing) produced relative measurements of:
transmitted peak power, peak power return (backscatter) from the test
aerosol, and peak power return from a black passive reflector placed
beyond the aerosol chamber. These recorded quantities were interpreted
in terms of aerosol transmission at the lidar wavelength, relative back-
scatter from the test aerosol, and absolute backscatter (volume back-
scatter coefficient) by assuming values for the laser pulse length and
the reflectivity of the passive reflector. Then these lidar-derived
quantities were related to aerosol mass concentrations and white-light
extinction coefficients derived at the chamber site for each fly ash
fraction. All data are presented as mass concentrations or optical co-
efficients that are independent of experimental geometry.
4
-------
C. Results and Conclusions
Data plots of lidar-derived aerosol backscatter showed a linear
relationship with respect to aerosol mass concentration and volume
extinction coefficient at low concentrations, becoming nonlinear at high
concentrations. Although detector saturation or limited instrumentation
bandpass could explain this result, it was concluded that within-pulse
attenuation by the test aerosol caused the observed nonlinearity. A
model expression that includes this nonlinear effect as a correction
factor to the expected linear relationship was fitted to the observed
data using a nonlinear least-squares procedure.
Data collected at a lidar wavelength of 0.7 u showed that the rela-
tionship between backscatter and the volume extinction coefficient for
white-light is nearly independent of the particle size distribution.
Since extinction is related to the second moment of the particle size
distribution this result indicates that backscatter is similarly related,
Data collected at a lidar wavelength of 1.06 p. showed that backscatter
is less related to extinction but more nearly related to the third moment
of the size distribution than at a lidar wavelength of 0.7 u. This sug-
gests that interpretation of lidar backscatter data in terms of plume
opacity is best achieved at a lidar wavelength of 0.7 LL and that inter-
pretation in terms of mass concentration is best achieved at a lidar
wavelength of 1.06 |i. The degree of experimental data scatter between
aerosol backscatter and mass concentration was similar for low and high
density plumes.
Extinction coefficients observed at the two lidar wavelengths
agreed with those observed with white light from the transmissometer
records. Hence, valid plume opacity measurements may be made by ob-
serving the near and far side clear air returns at either lidar wave-
length. The extinction at the two wavelengths is sufficiently- similar
-------
to discourage any attempt to determine mass concentration by comparing
observations of extinction at these two wavelengths. However, an im-
proved inference of mass concentrations may be possible by comparing
backscatter from the plume particulates at these two wavelengths.
Observed relationships among aerosol mass concentrations, volume
extinction coefficients and volume backscatter coefficients were compared
with those computed by use of the Mie scattering theory for spherical
particles assuming the results of the particle size analysis. In general,
the theoretical and experimental results agreed for each fly ash fraction.
The close agreement between theory and experiment encourages confidence
in the experimental techniques used, which can now be extended to study
backscatter from particulates for which theory is not complete (such as
backscatter from volumes of irregularly shaped particles).
In general, the main conclusion of the study is that the experi-
mental techniques developed are valid and useful for evaluation of laser
plume measurement instrumentation and for investigating the scattering
properties of particulate volumes. In particular, the experiments show
to a substantial extent (but not conclusively) for fly ash from coal-
burning power plants that:
(l) A lidar measurement of plume opacity by observing the clear
air returns from the near and far side of the plume return
is valid; i.e., plume transmission at lidar wavelengths of
0.7 \i and 1.06 p. and white light are equal within experi-
mental error.
(2) A lidar measurement of plume particulate backscatter at
0.7 p. wavelength may be interpreted in terms of opacity
regardless of particle size distribution. The relationship
between plume backscatter and opacity is not so well defined
at 1.06 p, wavelength.
(3) The relationship of backscatter to particulate mass concen-
tration is less dependent on particle size at 1.06 |i than
it is at 0.7 u..
-------
(4) The relationship between plume backscatter and plume opacity
at 0.7 u. wavelength is less dependent upon particle size
than is the relationship between plume backscatter and par-
ticulate mass concentration at 1.06 p, wavelength.
(5) Comparative measurements of backscatter at both 0.7 u and
1.06 u. wavelength are likely to give better indications of
particulate concentration than are single wavelength measure-
ments. Dual wavelength measurements of extinction, however,
do not appear to offer a similar advantage, due to the rela-
tivity small dependence of extinction upon wavelength.
-------
Ill DISCUSSION OF EXPERIMENTAL PROGRAM
A. Experimental Facility
An especially designed experimental facility was developed to study
laser backscatter from a volume of particulates. The facility allowed
the investigation to be conducted in a manner that nearly duplicates the
geometry of actual remote plume evaluation. Figure 1 shows the experi-
mental set up, including lidar, aerosol chamber, and particle feeder.
The chamber dimensions (20 in. x 20 in. x 30 ft) are sufficient to enclose
the effective scattering portion of a Q-switched laser pulse (typically
20 to 30 ns). Particle-laden air curtains at both ends of the chamber
effectively isolate the test aerosol from the surrounding environment
without affecting transmission of the laser pulse as it traverses a path
through the chamber. Additional design details of the aerosol chamber
are included in Appendix A.
B. Experimental Procedures
Particulates of known size distribution, shape, and refractive prop-
erties were pneumatically injected into the chamber at controlled con-
centrations by means of a rotating disk particle feeder (see Appendix A).
The generated aerosol was continuously monitored by a white light trans-
missometer along the axis of the chamber and, during periods of low back-
ground light levels, by a photomultiplier viewing 90° sidescatter from
a small portion of the transmissometer beam. (Further details of these
optical components are included as Appendix D.)
-------
BLACK TARGET
LIDAR
Air intake -
FAN
TA-653583-17
FIGURE 1 DIAGRAM OF EXPERIMENTAL SYSTEM
The lidar (laser radar) was placed 500 ft from the chamber. The
lidar consists basically of a laser transmitter that emits a very brief,
high intensity pulse of coherent light and a receiver that detects the
energy at that wavelength backscattered from the atmospheric aerosol as
a function of range. The resulting signal was displayed on an oscilloscope
and photographed. The lidar data records consist of Polaroid prints that
display relative transmitted peak power, peak power return from the aero-
sol within the chamber, and peak power reflected from a black passive re-
flector placed on the far side of the chamber. Further discussion of the
lidar characteristics and data recording procedures are found in Appen-
dix F.
Fly ash from a large coal-burning power plant was used for the partic-
ulate matter in the backscatter experiments. Raw fly ash was collected
and classified into various size fractions and these were then subjected
to various particle size analysis. Appendix B discusses the results of
particle size classification and analysis. Basically, the fly ash used
10
-------
in the backscatter experiments consisted of four size fractions (0-2.5,
2.5-5, 5-10, and 0-10 (i diameter).
Valid measurements of aerosol backscatter required that the laser
pulse pass through the aerosol chamber with negligible energy incident
on the chamber surfaces. Since light reflected from the chamber could
not be distinguished from light backscattered from the chamber aerosol
during an experimental run, extensive efforts were made to eliminate this
problem area. The lidar van was isolated from its shock mountings so
that lidar pointing was independent of weight loads or movement within
the van. The receiver field of view was coaligned with transmitted laser
pulses. The lidar was then aligned with the aerosol chamber. Initial
lidar observations made with a clean air chamber showed laser returns
from the chamber so that further modifications were necessary.
Infrared Polaroid photographs of the laser pulse size at the chamber
location (Figure 2) showed that the laser energy could be detected over
a diameter of 18 inches (21-inch skirt). Photographs made by inserting
neutral density filters in front of the camera optics reduced the spot
size to 16 inches for a 10-dB filter and to 4 inches for a 20-dB filter.
An upper limit to the spot size at the energy halfpower points (3 dfi)
is estimated as 6 inches. However, the solid surfaces of the aerosol
chamber caused significant energy return from the beam skirts (from
energy at least 10 dB down from that at the beam center). It was de-
sirable to reduce the beam size sufficiently so that negligible returns
were received from the clean air chamber. This was accomplished by
inserting an optical stop within the laser transmitter. Figures 2(b)
and 2(c) illustrate the laser pulse size before and after inserting an
optical stop. Negligible energy was intercepted and reflected back to
the lidar by the chamber surfaces with the optical stop in place.
11
-------
(a) PHOTOGRAPH OF LASER
BEAM FOR VARIOUS DEGREES OF
ATTENUATION BY NEUTRAL
DENSITY FILTERS
(bl PHOTOGRAPH OF LASER
BEAM FOR 10 LASER FIRINGS
BEFORE PLACEMENT OF
TRANSMITTER OPTICAL STOP
TAPE LENGTH hUUALS 4.3 inches
(c) PHOTOGRAPH OF LASER
BEAM FOR 10 LASER FIRINGS
AFTER PLACEMENT OF
TRANSMITTER OPTICAL STOP
TA-8730-23
FIGURE 2 PHOTOGRAPHS OF LASER BEAM SIZE AT CHAMBER LOCATION
11:
-------
The photomultiplier tube was tested for its range of output linearity.
At high light levels the tube became saturated and behaved in a nonlinear
fashion. A neutral density filter was placed within the receiver optical
path (see Appendix F) so that the return pulse from the passive reflector
was within the linear operating region of the photoraultiplier. Unfortu-
nately, the filter also decreased the returns from the particulate scatters
along the path traversed by the lidar pulse. The passive reflector w:as
painted with a black nonglossy paint to minimize the value of the re-
ceiver optical filter required for operation within the limits of detector
linearity.
A pair of communication lines was installed between the lidar and
aerosol chamber sites. One line was used for voice communication and
the other line provided electrical impulses to the data recorders located
near the chamber site at times of the lidar firing (see Appendix D).
The backscatter experiments normally consisted of making a series
of lidar firings with a clean air chamber to establish the scattering
properties of the background air and to check the lidar/chamber alignment.
This was followed with alternating series of lidar firings for a chamber
of clean air and various aerosol concentrations.
The data records collected at the lidar site for each laser firing
included relative measures of:
(l) Transmitted pulse peak power, P^
(2) Aerosol return, Pa
(3) Passive reflector return, P .
The data records collected at the chamber site included:
(l) A continuous strip chart record of the light transmission, T,
for a path along the length of the chamber.
(2) A continuous record of event marks indicating particle feeder
disk rotation speed.
13
-------
(3) A record of event marks indicating time of lidar firing relative
to the above data records.
(4) A paper tape printout of transmission at times of the event
marks of items (2) and (3) above.
(5) A continous record of relative 90° light scattering for times
of low background light levels.
(e) A "Nuclepore" filter sample of aerosol mass concentration, M,
within various parts of the chamber for some lidar data collec-
tion periods.
Examples of the raw data records are shown in Appendix C, Figure
C-5, and in Appendix G, Figure G-l. A detailed listing of the lidar and
transmissometer data collected for each laser firing is given in Appen-
dix H.
The lidar data records were used to evaluate the following quantities:
(l) Relative aerosol backscatter uncorrected for variations of light
energy transmission between the lidar and aerosol chamber.
(2) Relative reflected energy from the passive reflector normalized
by the transmitted peak power.
(3) Relative aerosol backscatter corrected for variations of light
energy transmission between the lidar and aerosol chamber by
using the target reflected return.
(4) Absolute aerosol backscatter from knowledge of the passive
target reflectivity and lidar pulse length.
(5) Absolute transmission of the laser energy across the aerosol
chamber by using data from item (2) above and assuming a clean
air transmission of 1.0.
The computational procedures used to derive these quantities are discussed
in Appendix F.
The lidar-derived quantities of aerosol backscatter and extinction
at the lidar wavelength were to be related to the white-light transmission
14
-------
and aerosol mass concentration determined from the filter samples. How-
ever, as discussed in Appendix C, difficulties were encountered with de-
riving valid aerosol samples from the aerosol chamber. It was concluded
that mass concentration evaluated from the particle feeder disk rotation
speed was most representative of the aerosol within the chamber over an
extended time period. Short term variations of the aerosol concentration
were observed for the smaller sized fly ash fractions and they probably
resulted from fly ash adhering to particle feeder components. Accordingly,
the transmissometer output T was calibrated in terms of mass concentration
M for each fly ash fraction by the expression:
T = «p-S ML , (1)
where the volume extinction coefficient
-------
The transmissometer reading at the. time of each lidar firing was used
to derive the instantaneous aerosol volume extinction coefficient and mass
concentration. These values are tabulated with the lidar quantities in
Appendix H.
C. Experimental Results
Data collected from the backscatter experiments are presented and
discussed in this section. Experiments were conducted at two laser wave-
lengths using the SRI/EPA* Mark VIII (0.7 p,) lidar and the SRI Mark V
(1.06 |_L ) lidar. The characteristics of these lidars are given in Appen-
dix F.
Figure 3 shows aerosol backscatter quantities observed at the 0.7-u,
wavelength plotted against the optical volume extinction coefficients de-
rived from the transmissometer data. The extinction coefficients may be
related to plume opacity by the graphs in Figure 10 (shown later) for a
plume length of 10 m.
Although a general increase of the volume backscatter coefficient
accompanies increases of the optical extinction coefficient, an unexpected
nonlinear relation exists at the higher concentration level. Lower values
of backscatter than expected were observed at high aerosol concentrations.
This can be caused by saturation (nonlinear response) of the optical de-
tector and degradation of large peak signals by limited electronic band-
pass, A third possible cause could be increased within pulse attenuation
by the plume particulates at high particulate concentrations. Experiments
conducted under laboratory conditions, although different from those used
in the field experiment, indicate that detector saturation or electronic
bandpass problems would not account for the degree of nonlinearity observed.
* Use of the Mark VIII lidar on this program was granted by the Division
of Meteorology, Air Pollution Control Office, EPA.
16
-------
10
CO
CJ
m
0.5
0-2
0.1
0.05
1 ''!'' ' ' I
DUST FRACTIONS
x 0-2.5 M
+ 2.5-5 At
D 0-10 At
O 5-10AI
1
10
20
50
100 200
EXTINCTION COEFFICIENT, TRANSMISSOMETER km
0 -iU
V.
1 10
I
h- 5
LU
CJ
COEFFI
to
DC 1
111
H
Y~
cj 0.5
CO
CJ
CO
ui 0.2
5
o 0.1
>
| , { , , , , | , : } , : : ,
T INFERRED FROM LIDAR
_ a x _
; (Independent of Return Signal t>z'~.
~_ from the Passive Reflector) rfJ-«''
' ^ "--n* "
»«?'x
x J*
s\v'*
ffi ~_
s
y/
s
] 2 5 10 20 50 100 200
T INFERRED FROM TRANSMISSOMETER
a
; (Dependent on the Return Signal from x* ;
" the Passive Reflector) x x_^
- x jf^:'''-:
& '')
*%i3'*
oj*
_ )*6 _
"
/
s
Xx
0 ,*a ° (c)
/°"
/ I ,,!,,,,! I , , I i , , , 1
12 5 10 20 50 100 2C
EXTINCTION COEFFICIENT, TRANSMISSOMETER km~1 TA-8730-3
FIGURE 3 BACKSCATTER QUANTITIES OBSERVED AT 0.6943 M WAVELENGTH AS A FUNCTION
OF THE EXTINCTION COEFFICIENT OBSERVED BY THE TRANSMISSOMETER
17
-------
The fraction of signal decrease as a result of attenuation within
the laser pulse by a homogeneous aerosol can be expressed as:
2 ,, T/2 -2ax 1-e / x
f = - e dx = , (2;
T Jo ar
where T is the pulse length and cr is the volume extinction coefficient.
This expression reduces to a value of one for a = 0 or for an infinites-
imal pulse length. For a value of a - 100 km and T = 0.01 km, f = I - e
« -.63, which is approximately the amount of nonlinearity observed at a =
100 km" . The dashed lines in Figure 3 represent nonlinear least squares
fit of the function:
(3)
where
Y = quantity plotted along the ordinate
X = quantity plotted along the abscissa
C , C = parameters to be determined. C., will relate to the
linear portion of the curve and C will relate to ef-
fective pulse length.
The parameters C and C were determined by the nonlinear least-
squares procedure of Marquardt (1963, 1964) using the supplied convergence
criterion. The best fit function was determined by minimizing percent
N . A . 2
residuals; i.e., finding C^ and C2 so that § = .£ [(Y. - Y )/Y.] is a mini-
mum, 6, where Y is determined from the model for each of the N data points.
The degree of fit of the model to the data points is objectively given as
~ 1 /2
the standard error of estimate defined as S = [$/($ - 2)]
18
-------
The amount of data scatter varies among the backscatter quantities
presented in Figure 3. This scatter can be caused by nonvalid assumptions
made for the data analysis or by experimental errors. As more pieces of
experimental information are included in the analysis, data point scatter
may be reduced by eliminating certain assumptions or increased by the
experimental errors associated with the additional information. The
relative backscatter is simply the ratio of aerosol return to transmitted
peak power; hence it ignores any variation during the experimental data
collection of the atmospheric transmission characteristics along the 500-ft
path between the lidar and aerosol chamber. However, both signals are
relatively large and can be read from the primary data records (Polaroid
prints) with little reading error. Mark VIII (0.7 " wavelength) data were
collected where background conditions were similar so that variations of
attenuation of the laser pulse along the 500-ft path were probably small.
Most of the data point scatter in Figure 3(a) (s = 0.16) is thought to
result from experimental limitations in deriving a valid measure of trans-
mitted peak power or from variations in the peak-power-to-energy relation-
ship; i.e., from variations in transmitter pulse shape (see Section IV B).
Volume backscatter coefficient, j3, values are shown in Figure 3(b)
and (c). The derivation of these data is discussed in Appendix G. As-
sumed in their derivation is knowledge of target reflectivity, laser
pulse length, and transmission of the laser pulse by the test aerosol,
T . As discussed in Appendix G, derivation of T from the lidar data
a a
results in an expression for p that is independent of the target reflected
pulse P . In addition to information on relative backscatter, there is
r
information in the form of lidar-observed quantities for conditions of
a clear air aerosol chamber. The amount of data scatter in Figure 3(b)
(s = 0.20) is slightly greater than that in Figure 3(a), which indicates
e
that experimental errors associated with the additional information used
19
-------
are greater than variations of clear air transmission during the experi-
ments. Figure 3(c) shows volume backscatter coefficients derived by using
T from the transmissometer data. Since wavelength dependence of trans-
0.
mission is relatively minor, the more accurate transmissometer values may
have decreased the data scatter. However, in this analysis the target
reflected energy pulse is required and is subject to large reading errors
at high aerosol concentrations. This explains the large data point scat-
ter (S = 0.23) at high aerosol concentrations in Figure 3(c).
e
In Figure 4(a) the volume backscatter coefficient is plotted against
the particulate mass concentration as determined from the particle feed
rates and in Figure 4(b) the volume backscatter coefficient is plotted
against the extinction coefficient as determined from the transmissometer
records [same plot as Figure 3(b)]. A greater nonlinear distribution of
the 2.5-5 |JL backscatter data is explained because these data, the first
collected, produced lidar returns sufficiently large to be in the nonlinear
response region of the lidar detector, and an additional neutral density
filter was placed in the lidar receiver after this data run (see Appen-
dix H), The other three fly ash fractions produced greater nonlinearity
at higher concentrations in proportion to their backscattering efficiency
at equal mass concentrations.
The data in Figure 4 indicate that the backscatter coefficient is
nearly proportional to the optical extinction coefficient independent
of particle size, but the relationship to particulate mass is dependent
on particle size. This implies that at this lidar wavelength (0.7 (j,)
backscatter is proportional to second-moment quantities of the particle
size distribution.
Figure 5(a) shows the volume extinction coefficient at a wavelength
of 0.7 jj, derived by relating the passive reflector return to the relative
transmitted peak power (see Appendix G) as a function of particulate mass
20
-------
zw
j 10
1 5
]
3
- 2
j
3
C ,
j 1
t
j
0
j 0.5
f
a
u
>
5
D 0.2
>
n 1
| 1 1 | i i i i | 1 1 i | i i i i | 1
DUST
FRACTIONS ,-"
CCS x ^
- * 0-2.5 0.100 0.007 0.18 ^s I
+ 2.5-5 0.061 0.011 0.16 $f*& * ^jf""
~ Q 0-10 0.036 0.006 0.15 y' +^^i
O 5-10 0.016 0.005 0.05 'V* n ,,-^+t ~/-
/ * Jr ^ v
_ ^ ^. ^"yt / /'
x .^ ^x *fli /
x^ ^ XX° /
x JK^ XH /x /
* / * '' ax ₯*
X w^ V^ * ' t fl. JW
Xv / +^ >W D X
-* x -*1 x
*$ ^ ^ / A1^
- /& /"$' $ ~~
/ ft. s r*
*,*,?>/
/' /' s /
- /* */ °/D% /'' M C2M / _
/ / / s
/ / ' (a)
' / /
/ \ Xlll
-------
*"
'E
* 100
a:
o
0 50
LU
1
LL
II i
cc
^
U-
2 20
_J
*
H
u 10
o
u_
u_
to uj
10 8 5
2
O
I
f)
z
1- 2
X ^
LU
1
1 1 | 1 1 1 l| | 1 1 | 1 I 1 l| },
/
/
**
DUST Y
- FRACTIONS *' /j
(/I) RELATION / 4. ,&''.
x 0-2.5 a = 0.82 M $o5< X 4/ttP* ^
_ + 2.5-5 a=0.39M *,' ,'/?aS-
0 0-10 a = 0.33 M * #* / / ,A
O 5-10 a = 0.21 M x/x*cv(^// '' 8
x' /a x
O X ^^ AQ >
+ / / / ^j .
X XX*X XD XX4± fl ft ~
^^ ' Q / ^*4 V /
X XX _ * X+ y
xx + v a°^ x ' n
x v + XOB ' / /*n
~ x x/x tx^' x7 ~
X )* XX^ ^/ ^xx x
X / X x O
xX x x'"xX x °
XX / ' j S ^
.M '/ / D
x xx x x. a
x xxx xx+
/ / / /
^ s ^ /
/x /xx/ xx
~ / '''/ / * (a]
/' ''/' '' *
1 '*' , D| X, , , , 1 1 , , 1 , I , , 1 1
1 . . 1 . 1 1 1 1 1 I 1 1 1 1 1 1 1 ,
/
s
l£
X
t '
J^x '-
vp D/ :
/ 9
/6
. o t x'W x
X n
x D x>§9x a^P ~
X n XX *
+ Q ^< v D
O * n x X^ £ O
"^^ttfX «. X D
T x X x ~
x * xx xV o x
x * D a
X «X X
X X / Of, _
/ * °
/ . D X
/
/
/
/
/
- ' o (b)
/\ °
/ 45°
/ a | \ , ,!,,,,! 1 , ,!,,,,!
5 10 20 50 100 200
MASS CONCENTRATION, FEEDER mg/m3
10 20
50 100 200
-1
EXTINCTION COEFFICIENT, TRANSMISSOMETER km
TA-8730-28
FIGURE 5 VOLUME EXTINCTION COEFFICIENTS (a) OBSERVED AT A LIDAR WAVELENGTH OF 0.7 n RELATED TO PARTICULATE
MASS CONCENTRATIONS (M) AND WHITE-LIGHT VOLUME EXTINCTION COEFFICIENTS. The dashed lines represent the
best fit linear relationship to the observed data.
-------
concentration and Figure 5(b) shows the lidar derived volume extinction
coefficient as a function of the optical volume extinction coefficient
derived from the transmissometer records. The relationship of the ex-
tinction coefficient at 0.7-Li lidar wavelength to mass concentration is
about equally dependent on particle size as the relationship of backscatter
to mass concentration [Figure 4(a)]. However, a greater overall scatter
of the data points occurs with decreasing scatter for increasing particle
concentrations. Since the largest experimental error associated with
reading of the reflector return from the Polaroid records occurs at higher
particle concentrations, the data scatter trend must be associated with
the experimental technique and probably results from small variations of
the clear air transmission that become negligible at large aerosol con-
centrations. The lidar-derived extinction coefficients (0.7 n) are only
slightly less than optical (white light) extinction coefficients [Fig-
ure 5(b)].
Figure 6 presents volume backscatter coefficients derived with the
1.06-u wavelength lidar. Unfortunately, data were not collected for the
5-10 11 fly ash fraction. Data on the relationship of volume backscatter
to mass concentration were nearly half as dependent on particle size as
data collected for the 0.7-|i lidar wavelength TFigure 4(a)]. Correspond-
ingly, the scatter of data points was greater in the plot of backscatter
coefficient against optical extinction coefficient than the scatter of
0.7 u, wavelength data (a standard error of 0.26 for the 1.06-n data is
comparable to a standard error of 0.20 for the 0.7-n data, which included
data from all four fly ash fractions). From Figure 6 it may be concluded
that, for a lidar "wavelength of 1.06 p,, the volume backscatter coefficient
is more nearly related to particulate mass concentration independent of
the particle size distribution than at the 0.7-u wavelength. Therefore,
interpretation of single wavelength lidar data in terms of mass concentra-
tion is better achieved at 1.06-|i wavelength than at 0.7-u wavelength.
23
-------
£~\J
1
1 10
r-
'E
-^
I 5
2
m
o
u.
u. 2
LU
O
o
DC
Ul i
t
<
o
OT
^
0 0.5
<
CD
UJ
5
3
J
0 0.2
>
0.1
| i i | i i M| | . i ! n.r - -f
DUST
FRACTIONS
~ (M) C1 °2 Se * -
x 0-2.5 0.115 0.016 0.17 i**'-^---
+ 2.5-5 0.090 0.009 0.16 + «f-<- " ~^^n -
p- D 0-10 0.070 0.010 0.22 2* ^^ -
r x }ffc%'0
s s'k / 9^
J&,'' s'
J*- '*^ '
j&ts /
~~~ X rj S
^ ^* i^ "^
*< yXM^Er ££
rfx-c xffl?
xx * s* *^
- ' » D
~ xX,^Ex/' ~
/ / s
S f S
- / ' / -
/ . f
' / '
/ ,' rf 1 \
/ ' Of n - M | 1-e"C2 \
_x/6/ ' \ C M /
/' ,f /' 2 /
x"x'' (^
/I i i 1 j i i i 1 1 i i 1 i i i 1 1 1
2 5 10 20
MASS CONCENTRATION, FEEDER mg/m
50 100 200
*. ,,-
->'*
o ^
D X
= 0.180 I
S = 0.26
1-e
-0.0260\
0.0260
Ib)
i ' II I M
i illi l i
10
20
50
100 200
-1
EXTINCTION COEFFICIENT, TRANSMISSOMETER km
TA-8730-29
FIGURE 6
VOLUME BACKSCATTER COEFFICIENTS (0) OBSERVED AT A LIDAR WAVELENGTH OF 1.06 n RELATED TO
PARTICULATE MASS CONCENTRATIONS (M) AND WHITE-LIGHT VOLUME EXTINCTION COEFFICIENTS (a). The
dashed line represents the best fit to the data of the nonlinear equation indicated.
-------
However, the best universal inference would be optical extinction (trans-
mission or opacity) from data collected at a wavelength of 0.7 a.
Figure 7 shows the volume extinction coefficient derived at 1.06-j.
wavelength as a function of particulate mass concentration and the optical
volume extinction coefficient. The empirical relationships between ex-
tinction and mass concentration are not substantially different from those
derived at a lidar wavelength of 0.7 u,. This indicates that extinction
measurements taken at these two wavelengths with the same experimental
accuracy as the measurements presented here could not be interpreted in
terms of particulate mass concentration, as suggested by Twomey and Howell
(1967). However, plume transmission measurements (by the technique of
observing the clear air returns in front of and behind the plume) at
either lidar wavelength may be related to white-light transmission or
opacity.
D. Comparison of Experimental Results and Mie Theory
Individual particle backscatter cross sections and integrated or
volume backscatter coefficients can be investigated theoretically using
the exact Mie scattering theory for homogeneous or stratified spheres.
Figure 8 shows results of Mie theory computation in terms of the back-
scattering efficiency, which is defined as the ratio of the computed back-
scatter cross section to the geometrical cross section. The particle size
parameter is the ratio of the particle circumference to the wavelength
of the scattered light. These results are for homogeneous particles with
a real refractive index of 1.5 and imaginary refractive indices of 0.0
(nonabsorbing), 0.01, 0.05, and 0.10.
The sharp resonances of the nonabsorbing particles have been explained
by surface waves, i.e., waves that travel along the interface between the
sphere and the medium (Bryant and Cox, 1966; van de Hulst, 1957). It is
25
-------
to
01
^uu
| 100
-it
'
CC
O 50
1-
O
LU
LU
LU
DC
1 20
Q
-1
z 10
LU
u
LU
LU
LU c
0 5
O
z
Q
p
O
S 2
X
LU
1
I 1 1 I 1 I MI 1 1 I 1 I 1 1 1, 1
X
DUST J&' /
FRACTION x ' S_
A< RELATION *x /W~
/ / /
x 0-2.5 a = 0.83 M / / W -
+ 2.5-5 a = 0.51 M x'|- x'X ^xB
a 0-10 a = 0.27 M x +/ V /X ''
* «t r, /
/ * / + a /
/ X + + -~S
X Xv ^tf'
X X *XC1
' / / _
x>< x x x'
x. x x
/ * X X
£^ x^a oxx
rf, y' a i*r
: a / xxx*g/B :
0 X X « X *
x x 'x
_ / X X 0 _
a ' x x
xX xX xX 0. D
x x Dx " a
XXX
XX X
/ X X +
xX x /x °
~ / X X ""
XXX D i.
xxx (a)
/ x x
XX X
IX X i i 1 i | i
, X| 1 1 1 1 i 1 1 1 1 1 1 1 1 1 1 1
-1 1III I I
-r
*
*
,
?
(b)
1
1 1
2 5 10 20
MASS CONCENTRATION, FEEDER mg/m
50 100 200
-1
50 100 200 1 2 5 10 20
3 EXTINCTION COEFFICIENT, TRANSMISSOMETER km"
TA-8730-30
FIGURE 7
VOLUME EXTINCTION COEFFICIENTS (a) OBSERVED AT A LIDAR WAVELENGTH OF 1.06 \i RELATED TO PARTICULATE
MASS CONCENTRATIONS (M) AND WHITE-LIGHT VOLUME EXTINCTION COEFFICIENTS. The dashed lines represent the best
fit linear relationship to the observed data.
-------
10
PARTICLE SIZE PARAMETER
20 30
40
50
10"
o
z
uu
icr
to
o
to
o
<
CD
10
,-3
10-4 t
m = 1.5 - 0.0!
f \ ' '< '; |" i
1.5 - 0.01/ I" l' ''
m = 1.5 - 0.1/
m = 1.5 - 0.05;
m = complex index of refraction
1234
PARTICLE RADIUS FOR AN INCIDENT LIGHT WAVELENGTH OF 0.7 {l
5.5
_L
2 3 4 5 G 7
PARTICLE RADIUS FOR AN INCIDENT LIGHT WAVELENGTH OF 1.06 p
TA-653581-24R1
FIGURE 8 SPHERICAL PARTICLE (MIE THEORY) BACKSCATTER EFFICIENCY FACTORS (BACKSCATTER
CROSS SECTION/GEOMETRICAL CROSS SECTION) AS A FUNCTION OF PARTICLE SIZE AT
0.7 AND 1.06 v WAVELENGTHS FOR SEVERAL VALUES OF THE COMPLEX REFRACTIVE INDEX
-------
seen that, as particle absorption increases (increase of the imaginary
part of the refractive index), the magnitude of the resonances is de-
creased, as are the overall backscatter cross sections. However, with
increasing absorption or increasing size, the backscatter approaches the
Fresnel reflection limit at normal incidence (McDonald, 1962) and increases
with increasing absorption by the particle. The important conclusion is
that the absence of surface wave effects in the backscattered light from
absorbing spheres could increase the correlation between volume backscatter
and particulate concentration for small changes in particle size distri-
butions. In addition, these surface waves are probably dependent on the
smooth surface of the sphere for their propagation and thus may be absent
within backscattered light from irregularly shaped particles regardless
of their orientation. This conclusion is supported by the scattering
measurements of Holland and Gagne (1970) on randomly oriented flat plates
that gave less backscatter than equivalent size spherical particles.
The fly ash particulates used in the backscatter experiments were
mostly all spherically shaped (see Appendix B), and hence theoretical
results may relate favorably to empirical results. The ratios of volume
backscatter coefficient to mass concentration and of volume extinction
coefficient to mass concentration were experimentally determined for each
fly ash fraction at each of the two lidar wavelengths (0.7 and 1.06 p,).
Theoretically, these parameters, which are independent of particulate
number densities, may be expressed as:
I 2
rra Q (x,m)n( a)da
CT _ £ e
M ~~ co
f rra pn( a)da
'J O
O
28
-------
and
f 2
J ira Q (x,m)n( a)da
j| _ £ b
M °°
f rra pn( a)da
J 3
o
where
a = particle radius
x = particle size parameter = 2rra/\
A. = incident light wavelength
m = complex refractive index
Q = extinction cross section/geometrical cross section
e
Q = backscatter cross section/geometrical cross section
b
p = particulate density
n = number of particles per unit volume per radius interval.
Values of
-------
Table 1
OBSERVED AND COMPUTED AEROSOL PARAMETERS
Size
Fraction
0-2.5
2.5-5
5-10
0-10
"Average" Sauter
Diameter
5
32
1.7
3.1
6.5
4.0
1/5
32
0.59
0.32
0.15
0.25
(1/D '.*
32'
1.00
0.55
0.26
0.43
Observed V/hite-
Light Parameters
a/H
0.943
.450
.214
.336
(cj/M1*
1.00
0.48
0.23
0.37
Q
e
av
2.73
2.46
2.12
2.23
Size
Fraction
0-2.5
2. 5-5
5-10
0-10
Size
Fraction
0-2.5
2.5-5
5-10
0-10
Mie Theory Computations (0.7-L: wavelength'1
a/U
0.990
.467
.210
.388
(CT/U)*
1.00
.47
.21
.39
3/11
0.0820
.0410
.0151
.0295
( s/ii ; *
1.00
.50
.18
.36
K
e
0.411
0.870
1.939
1.048
K
b
4.96
9.92
26.86
13.78
k
12.07
11.39
13.90
13.15
(k)*
1.00
0,94
1,15
1,09
Q
2.76
2.37
2.24
2.54
^
0.228
.208
.161
.193
X
7.70
14.01
29.44
18.09
Mie Theory Computations (1.06-^ wavelength^
C-/U
1.157
0.506
0.217
0.401
(a/in*
1.00
0.44
0.19
0.35
S/U
0.0520
.0479
.0160
.0281
(5/11)*
1.00
0.92
0.31
0.58
K
e
0.351
0.804
1.877
1.014
K
b
7.82
8.48
25.40
13.56
k
22.25
10.56
13.56
13.41
'ki*
1.00
0.47
0.61
0.60
Q
e
3.22
2.57
2.31
2.63
^
0.145
.244
.171
.184
X
5.03
9.17
19.26
11.84
Size
0-2.5
2.5-5
5-10
0-10
Lidar-Observed Aerosol Parameters
0.7-u Wavelength
C/M
0.817
.395
.207
.331
(cj/11)*
1.00
0.48
0.25
0.40
3/M
0.100
.0605
.0156
.0358
f S/1I)*
1.00
0.61
0.16
0.36
k
8.2
6.52
13.27
9.23
Q
2.28
2.01
2.20
2.17
Qb
0.279
.308
.166
.235
1.06-j, Wavelength
a/l!
0.833
.509
.278
( S/M}*
1.00
0.61
0.33
= /M
0.1145
.0896
.0696
(3/111*
1.00
0.78
0.61
k
7.28
5.68
3.99
Q
2.32
2.56
1.82
Qb
0.319
.456
.456
039
- Sauter diameter ( jj. ]
- = volume extinction coefficient (km"1^
5 - volume backscatter coefficient (km" ster"1;
M = mass concentration ( mg/m^ )
k = cj/5 'sterv,
x - D32/X - average size parameter
\ = wavelength (microns]
av
= average extinction efficiency factor
= average backscatter efficiency factor
av ( ster * 1
K = K factor of Ensor and Pilat ' 1971";
e
K - backscatter factor defined in the manner
b
of K
(P'* = parameter P normalized to a value of 1.0
for the 0.2- to 5-j. fraction.
30
-------
is an experimentally determined quantity and hence subject to experimental
error, and that an accurate evaluation of the complex refractive index
was not available.
Light extinction is not strongly wavelength dependent, and the cr/M
values observed with white light may be related to the monochromatic com-
putations. Especially good agreement exists between the white-light and
0.7-(j, values for all four fly ash fractions. Hence, single wavelength
lidar measurements of plume transmittance at 0.7-u, wavelength can be
interpreted adequately in terms of plume opacity (l - transmission of
white light).
Parameters of Table 1 that have been normalized to a value of 1.0
for the 0-2.5 p, fly ash fraction are enclosed within parentheses. Quan-
tities that are independent of size distribution are constant for each
fraction; quantities that vary as the particle surface to volume ratio
are proportional to the reciprocal of the Sauter diameter D . As ex-
o 2i
pected , the white-light and monochromatic extinction-coefficient-to-
mass-concentration ratios vary nearly as the reciprocal of the Sauter
diameter, 1/D . The computed ratios of backscatter coefficient to mass
O^
concentration at 0.7-(_i wavelength show greater dependence on size distri-
bution than the 1/D variations and at 1.06-u wavelength are nearly in-
O^
dependent of size distribution for the 0-2.5 ij, and 2.5-5 - fly ash frac-
tions. The 5-10 [j, fraction is approximately half as dependent on size
distribution as at the 0.7-p, wavelength. The observed ratios are in
agreement with the computed ratios as far as establishing dependence on
size distribution. Therefore, for the 1.06-u wavelength both observations
and theory predict less dependence on the backscatter-to-mass ratio on
particle size than for the extinction-to-mass ratio (proportional to
1/D32).
31
-------
The other quantities presented in Table 1 are for comparison with
the work of other authors. The value K , which is the ratio of particulate
e
volume concentration to extinction coefficient, corresponds to the value
K given by Ensor and Pilat (1971) and Pilat and Ensor (1970). The rela-
tionship between a/M and K is given by
-1 2
CT/km m\ ,32, ,,3. . .
-I - ) = l/K(cm /m )p(g/cm ) , (6)
\mg/m
3
and the computation of K has assumed a fly ash density of p = 2.46 g/cm
for each fly ash fraction (see Appendix B). The K quantities are the
b
corresponding values for the ratios of particulate volume concentration
to volume backscatter coefficient. Effective (average) efficiency factors
for volume extinction and backscatter coefficients computed from Mie Theory
are given by Q and Q , where:
e b
a Q (x,m)n(a)da
Q
e
r 2 f
J a n(
and
I 2
a Q (x,m)n(a)da
\ "^
r 2
J a nf a)da
32
-------
Observed average values of Q and Q are determined from the following
e b
expressions (Appendix E):
Q = I- DD
e \3 / 32 M
av
Q = I - | pD
b \ 3 / H 32 \ M
av
where o~/M and 9/M are experimentally determined values.
Figure 9 presents observed and computed backscatter and extinction
efficiency factors as a function of particle size parameters computed
from the Sauter diameters (x = rrD /X). The lidar-observed efficiency
O t~i
factors on the average are slightly less than the monochromatically com-
puted factors, however they agree well with the observed white-light values,
The lidar-observed backscatter efficiency factors show a general max-
imum at a size parameter of approximately 10.0, which agrees with the
maximum observed in the theoretical curves for single particles (Fig-
ure 8). The observed average values are greater than the computed average
values and exhibit more of a peak at x = 10, which indicates that the fly
ash size distributions may have been more uniform (narrower) than assumed
in the computations. However, the use of too large a target (passive re-
flector) reflectivity or too short a laser pulse length will produce over-
estimates of volume backscatter coefficients and backscatter efficiency
factors. In any future work to derive these absolute backscatter quan-
tities using the passive reflector technique, more emphasis must be placed
on deriving knowledge of the target reflectivity and laser pulse length
than was possible in the present study.
33
-------
10.0
IO
o
LU
O
LU
LU
LU
O
O
X
LU
10
o
LU
5
LL.
LU
o:
LU
U
(O
^
CJ
5.0
2.0
1.0
0.5
0.2
0.1
I I
D
D
i i i I i i r
O Observed White Light (A = 0.6^ Assumed
in Q Calculations)
A Observed 0.7 p.
Observed 1.06 H
£± Calculated 0.7 /I
n Calculated 1.06 M
m
S3
A
i i
A ° *
D
A
I
5 10 20
AVERAGE PARTICLE SIZE PARAMETER. X
50
TA-8730-31
FIGURE 9 OBSERVED AND COMPUTED EXTINCTION AND BACKSCATTER
EFFICIENCY FACTORS AS A FUNCTION OF SIZE PARAMETER
COMPUTED FROM THE AVERAGE SAUTER DIAMETERS
34
-------
The relatively good agreement between theoretical and experimental
results obtained gives support to the experimental procedures used during
this study. Hence, further experimentation into areas for which existing
theory is incomplete (such as scattering by irregularly shaped particles!
may now proceed with some degree of confidence.
35
-------
IV TECHNIQUES OF PLUME EVALUATION
A. Comparison of Opacity and Backscatter
Measurement Techniques
Plume evaluation from a lidar backscatter signature consisting of
the plume return, P , and near and far side clear air returns, P, and P ,
p ^ F
are the subjects of the following discussion.
An inference of plume transmission, T, or opacity, O , from clear
air returns is given by the expression:
= 1-T = 1
where R is the range to the appropriate return signal. This equation
assumes that backscatter coefficients for the clear air regions are equal
and that the plume transmission at the lidar wavelength adequately rep-
resents the transmission of white light. However, this technique does
not require absolute lidar calibration, and the inferred quantity is not
subject to variations of clear air transmissions over the path from lidar
to plume.
A plume opacity measurement is related to the volume extinction
coefficient a and the observed plume length L by Bougure's law of
attenuation:
Jta(l - O ) = -aL (8)
P
37
-------
Figure 10 is a graphic presentation of this expression. [Aerosol opacity
observed during the backscatter experiments may be determined from Fig-
ure 10 using the observed extinction coefficients (Section III) and a
plume (chamber length) of 10 m.] From Eq. 8 opacity is related to the
particulate mass concentration, M, of a vertically rising circular and
homogeneous plume of radius a observed at ground level with a lidar ele-
vation angle 9 by the expression:
JLn (1 - 0 ) = -(-} Ma sec 6 (9)
p UI/
The quantity a/M is independent of particle concentration; it relates the
aerosol optical density (the extinction coefficient a) to its physical
density M and may be considered a constant for particulates of a given
type and relative size distribution. The relationship of this ratio to
various particle refractive indices and log-normal size distribution
parameters has been studied theoretically for spherical particles by
Pilat and Ensor (1970), Ensor and Pilat (l97l), and Conner and Hodkinson
(1967).
The total particulate emission M from the stack per unit of time
may be expressed in terms of observed opacity as:
M = 77a Mv = -rr (-) a cos 9v In (l - 0 ) , (10)
t \ a/ P
where v is the air speed of the stack emission. The interpretation of a
lidar measurement of plume opacity in terms of particulate concentration,
M, or total emission, M , requires knowledge of the optical-physical
parameter cr/M and hence is dependent on the particle size distribution.
38
-------
500
En (1 - O ) = -al_
2 5
OBSERVED PLUME LENGTH, L meters
10
TA-8730-33
FIGURE 10 PLUME OPACITY AS A FUNCTION OF PARTICULATE VOLUME
EXTINCTION COEFFICIENT AND OBSERVED PLUME LENGTH
39
-------
The lidar opacity measurement described above is independent of the
lidar signal from the plume particulates, P (However, the large sig-
P
nals received from the plume may saturate typical lidar detectors and,
without their complete recovery, accurate measurements of P may not be
possible. ) The plume return P may be related to plume mass concentration
P
by the expression:
-2 /B \ 2 , .
P = K -R [- MT (ll)
p p \M/ c V '
where
K = a lidar constant
T = clear air transmission to the plume
c
j3 = volume backscatter coefficient.
The above statements applied to the ratio a/M also pertain to the quan-
tity 3/M- Equation 11 assumes that: particles are uniformally dis-
persed within the laser pulse; a constant energy density exists across
the beam; attenuation by the plume particulates within the pulse may be
neglected; and P is the maximum peak signal return, i.e., R is the
P P
range at which the pulse becomes completely embedded within the plume.
As shown both experimentally and theoretically in Section III, the ratio
3/M is nearly independent of particle size at long wavelengths. There-
fore, less requirement is placed on knowledge of particle size for a mass
concentration estimate based on a backscatter measurement than for one
based on an opacity (transmission) measurement.
Other advantages of the plume particulate backscatter technique
over the opacity technique are its adaptability to low density plumes
and its ability to take a direct reading of plume mass concentration.
In addition, because of the relatively large plume returns, less
40
-------
expensive lidar instrumentation is required. However, disadvantages of
the particulate backscatter technique are the requirement of an absolute
lidar calibration, dependence of the plume return on the transmission
of the laser pulse along the clear air path between the lidar and the
plume return, dependence of the plume return on the emitted energy from
the lidar, and the greater sensitivity of the volume backscatter coeffi-
cient than that of the volume extinction coefficient to particle shape
and particle refractive properties. The above operational aspects of
remote plume reading by observing backscatter of laser energy by the plume
particulates requires further investigation.
B. Laser Pulse Length Considerations
The lidar data analyzed in this report are the peak power returns
from both aerosol volumes and solid targets. The data presented give in-
formation on the scattering properties of aerosol volumes that is needed
for the remote evaluation of plumes. These data have been presented in a
manner that allows inferences about the accuracy with which remote plumes
may be evaluated without knowledge of the particle size distribution.
However, the experiments have been idealized in that the ratio of plume
length to laser pulse length has remained nearly constant. Derivation of
plume density (mass concentration) from opacity measurements requires
information on the parameter o/M and hence depends on particle size dis-
tributions. Backscatter has been shown to be less dependent on particle
size distribution (for long wavelengths) and a more measurable quantity
for low density plumes than either extinction or opacity.
The lidar equation for peak power returns may be expressed in general
form as:
P(R ) = f Y(R)P (R - R)dR , (12
o - to
o
41
-------
where
P(R ) = peak power returned from range R
o o
2 2
Y(R) = KR P(R)T (R)
P = power density
T*
K = a lidar constant.
When the pulse length is much less than the plume length (L « L ), then
P(R ) sa Y(R ) f P (R - R)dR. The instantaneous power return at a
o o J t o p m -,
given range (R ) is dependent on the transmitted energy P P (R - R)dR
o [> * ° J
and the shape of the lidar backscatter signature is nearly that of the
transformation function Y(R). A feature in Y(R), such as increased scat-
tering by a stack plume imbedded within cleaner air, is reflected in the
function P(R). This case (L < L ) relates to the backscatter experiments
where it was assumed in the data analysis that:
f P (R - R)dR = K*P
J tv o ' t
o
where P is the observed peak transmitted power and K* is a constant,
The total area under the return signal (energy return) is given as
^
f P(R )dR = K'P f Y(R )dR , (13
«J o o t «J o o
or the product of the pulse area and the area under the lidar backscatter
signature (received signal as a function of range). Assuming Y is directly
related to aerosol concentration, the return signal from a plume of greater
density than its background may easily be distinguished from the back-
ground, and the range integration of the return signal from the plume
42
-------
would relate to total concentration of the plume after applying a correc-
tion for attenuation by the plume particulates.
The pulse length may be larger than the plume it penetrates (L » L
In such a case Eq. (12) may be expressed as
P(R ) % P (R - R) P Y(R)dR
o to -J
and the lidar signature is more responsive to pulse shape than Y(R) vari-
ations. Hence, for long pulse lengths it is more difficult to distinguish
the plume return superimposed on the background (clear air) return. In
the extreme case of CW (continuous wave) lasers a single numerical value
represents the return integrated over the observed path.
C. Polarization Measurements
Important inferences on the nature of scattering particulates can
be made from polarization measurements of scattered light (Eiden, 1966;
Gilbert, 1970; Dave, 1969; Harris, 1970). Since laser energy can be
emitted as polarized light, it is questioned if polarization measurements
of backscattered laser light can be used to derive particulate concentra-
tions. The polarization of the incident light is conserved in the back-
scattered radiation from spherical particulates (except for circular
polarized light that is conserved except for reversal of the rotation
vector) and consequently is independent of the nature of the spherical
particulates. Hence, any depolarization would be attributed to multiple
scattering and would be relatable to particulate concentration. However,
the polarization of the incident light tends to be destroyed in the back-
scattered light from irregularly shaped particles. Schotland etal(l97l)
has used this property to distinguish between water and ice clouds. Since
43
-------
the primary backscattered light from smoke plume particulates would
dominate over any multiple scattered light and since the particulates
exhibit varying degrees of sphericity, it is unlikely that backscatter
polarization measurements could be related to plume density. A limited
number of polarization measurements made during this study showed no
relationship between particulate concentration and depolarization of the
energy returns. However, some experimental error was introduced because
of the difficulty of making polarization measurements with coaxial lidar.
D. CW Laser Techniques
The present study indicates that optical backscatter techniques,
along with existing lidar technology, may yield useful information on
mass concentration within a stack plume. However, several characteristics
of current lidar technology will tend to limit lidar's immediate useful-
ness as a practical, low cost, mobile monitoring instrument. The major
limitations are: equipment complexity and cost, lack of a simple real-
time data processor within the receiver, and high peak power optical
radiation that could be hazardous to eye safety.
The cost of Q-switch crystal lasers (ruby, neodymium-glass, and
neodymium-yag) currently used in lidar sensors is typically in the
315,000 to $20,000 bracket. Although these prices have declined some-
what in recent years as the result of improved technology and increased
competition, it is unlikely that the price of Q-switched crystal lasers
will decrease significantly in the near future. Accordingly, most
Q-switched crystal lasers are expected to remain unattractive for this
particular application in the foreseeable future.
In the present study extensive use was made of digital computer
techniques to process the raw lidar data. For a truly practical monitor-
ing instrument it is desirable that the lidar data be processed
44
-------
automatically and analyzed in real time and that the result be displayed
with little or no intervention by the operator. There seems to be no
adequate device for this purpose, although recent advances in integrated
circuits and memories may make feasible the use of hybrid analog and
digital computation techniques at a comparatively low cost.
If these previously mentioned limitations can be overcome, the re-
sulting device should be a relatively simple, low cost, mobile instrument
that is capable of yielding accurate, repeatable measurements even when
used by relatively unskilled personnel.
Application of CW lasers to the problem of remote evaluation of
plume density may overcome the high cost and hazard to eye safety prob-
lems of Q-switched crystal lasers. Several CW laser techniques are
briefly discussed below.
1. Unmodulated CW Laser System
The lowest cost and least sophisticated laser device for remote
evaluation of plume densities would use an unmodulated CW laser as its
energy source. The received backscattered signal would represent the
backscattered light integrated over the viewing path. This is equivalent
to path integration of the lidar backscatter signature from R = 0 to co.
Because of the inverse range-squared dependence of the lidar signature,
the nearby clear air return in a relatively polluted atmosphere may equal
the energy return by the remote plume. In addition, the range to the
plume must be evaluated independently to correct for the range dependence
of plume returns. Because of these problems an unmodulated CW laser back-
scatter device looks unattractive. However, spatial separation of the
transmitter and receiver optics so that the receiver views only a remote
portion of the complete scattering volume may eliminate the clear air
scattering problem and give an estimate of plume range. A separation
45
-------
distance of 8 to 10 ft between transmitter and receiver would allow the
system to be mounted on a mobile van or stationwagon. Some effort may
be expended, however, to align the system for a particular remote air
volume, and the system optical stability may limit the accuracy of the
plume reading. Another limitation is that the optical receiver would be
required to detect weak backscatter signals in the presence of background
sources such as the sun, street lights, and the like. Signal discrimina-
tion may be possible with an extremely narrow bandwidth optical filter
or by modulating the CW laser with a constant frequency and including in
the receiver electronics a narrow bandwidth electrical filter to reject
the unmodulated background noise.
2. Pulse Modulation CW Laser System
An acousto-optic modulator suitable for helium-neon and argon
lasers has been described by Maydan (1970). The modulator in effect
allows a continuous wave helium-neon or argon laser to be operated in
a pulsed mode with minimum pulse widths of 20 nsec and maximum pulse
repetition rates of 2 MHz. In addition, approximately 50 percent of the
power within the laser cavity is available at the output, thus increasing
the peak power of the pulse by a factor of 50 to 100 over the correspond-
ing CW power output of the same laser. The availability of a very high
repetition rate laser transmitter makes it possible for the first time
to use various signal-processing techniques within the receiver. Certain
well-known techniques from microwave radar technology, such as signal-to-
noise improvement by means of pulse-to-pulse averaging could be easily
employed, along with hybrid analog and digital real-time processing of
the resulting data.
The use of the acousto-optic modulator could, in theory, provide
a number of significant advantages to lidar system performance. These
46
-------
advantages could make possible low cost, remote monitoring of smokestack
plumes. The acousto-optic modulator offers these major advantages:
(l) It significantly increases the peak power output of a
gas laser.
(2) It provides the ability to pulse the gas laser very
rapidly (typically, 106 pulses/sec).
(3) It provides the ability to vary the width of the trans-
mitted laser pulse.
(4) The rapid pulsing of the laser transmitter produces a
received signal with a rapid data rate (e.g., a million
measurements of plume backscatter per second). This
high data rate allows the use of range gating and video
integration to increase receiver sensitivity and measure-
ment accuracy over that derived using a single pulse.
When incorporated into a system, this technique may permit pulse (range)
measurements to be made without any danger to eyes.
A simple system could incorporate the technique of automatically
sending out a transmit pulse shortly after detecting a receive pulse
from the plume. The rate at which the laser transmits pulses would be
a function of the range to the plume. The strength of the receive pulses
could then be corrected automatically for the range to the plume and
recorded; the operator would be needed only for pointing and turning on
the power switch. Other range discrimination techniques may be applicable
and should be considered.
Although detailed system performance calculations have not yet
been completed, preliminary calculations seem to indicate that this con-
cept is worthy of further investigation.
47
-------
3. FM CW Laser System
The FM CW technique (Jackson, 1971) provides the most complete
set of capabilities and, with appropriate development, at little or no
additional cost or complexity over the other CW laser systems. Ease of
operation and eye safety features of CW systems are advantages of the FM
technique. In addition, this technique has the lowest peak-power-to-
average-power ratio and is not vulnerable to the receiver overload and
recovery problems associated with pulsed systems. The sensitivity (hence,
transmitted power required) of a radar depends on the average transmitted
power and, for any given average power required from the laser, the FM
technique results in the lowest peak power.
Calculations indicate that it will be possible to: modulate a
low power CW laser beam; detect scattered laser light; and measure, in
real time, the scattering as a function of range, thus deducing plume
reflectance and plume opacity. The unique characteristic of the technique
is that the laser beam is amplitude modulated by a frequency-modulated
subcarrier. A telescope collects the light scattered by clear air and
the plume, and a photodetector recovers the FM/CW subcarrier. The FM
permits the measurement of range, signal strength, and range resolution
by standard FM radar techniques. Real-time processing can be obtained
using frequency filters and heterodyne techniques instead of the time-
gating techniques applicable to pulse systems.
What is required at this time are experiments to obtain engi-
neering data that will be needed in future FM/CW lidar development
programs. These data (which are also required for other types of CW sys-
tems) will permit engineering choices to be made of system parameters
such as subcarrier frequency, subcarrier FM rate, size and amplitude
weighting of receiver aperture, frequency of sawtooth FM, laser type and
power, laser stability specifications, and photodetector specification.
48
-------
Of primary interest in the experiments will be optical Doppler spreading
of the backscatter signal, conversion of laser FM noise into AM noise
owing to the depth of the scattering region, and receiver aperture
smoothing effects.
To summarize, the unmodulated CW laser system requires the
lease amount of electronics but provides the least amount of information
required for plume evaluation. However, the optical alignment problems
could outweigh the electronic simplicity for actual portable field systems.
The pulse-modulated CW laser system could obtain,without an eye safety
problem, the same data as a single-shot pulsed lidar. The FM/CW technique
would require more development but could actually cost less and would pro-
vide a system capable of measuring both the backscatter and opacity of
smoke plumes.
49
-------
V RECOMMENDATIONS FOR FURTHER RESEARCH
The research conducted under this study has demonstrated that
measurement of laser energy backscattered from plume particulates is
useful for remote evaluation of smoke plumes and that evaluation of
backscatter techniques is possible by using an especially designed aero-
sol chamber to generate an artificial plume consisting of known particu-
lates at known concentrations. However, a low cost laser plume-reading
instrument must still be developed and evaluated. In addition, the
techniques used and the results obtained under this study suggest other
research areas. Accordingly, it is suggested that research be undertaken
to (l) develop a low cost plume reading instrument, (2) extend the experi-
mental results of this study, and (3) further develop the aerosol facility
for use in a variety of aerosol studies and instrument evaluations. In
particular, the suggested areas of further research are:
Development of a low cost remote plume sensor
Experimental evaluation with other particulates
Aerosol chamber improvement for aerosol studies
Extend the accuracy of backscatter measurements made with
pulsed lidars
Computations from Mie theory
Development of multiple wavelength lidar techniques.
Each of these recommended research areas is discussed in the following
paragraphs.
51
-------
A. Development of a Low Cost, Remote Plume Sensor
Although high power lasers such as ruby (0.7 p,) and neodymium-glass
(l.06 p.) were used to investigate backscatter from fly ash particulates,
it is unlikely that these lasers will find extensive application for
remote plume monitoring because of their high cost and complexity and
their potential hazard to eye safety. Accordingly, further research
should be directed toward applying the results and techniques of the
present study to develop a low cost laser instrument for reading coal-
burning power plant plumes. Experimental development of CW techniques
based on bistatic, amplitude modulated and frequency modulated systems
for remote plume reading is recommended. Preliminary evaluation of
these laser systems could be accomplished using the experimental aerosol
facility developed under this study.
B. Experimental Evaluation with Other Particulates
The present study dealt exclusively with fly ash particles. The
results showed that optical characteristics could be characterized by a
specific average particle size. Fly ash particles are mostly spherical
in shape and heterogeneous in composition. It would be desirable to
obtain optical data on other synthetic and real particulates to establish
the effects of shape and composition as a guide to developing a generalized
means for assessing the applicability of optical measurements. Examples
of suggested synthetic powders, each of which has specific attributes,
are: glass spheres, aluminum spheres, flake aluminum, Cab-0-Sil, iron
oxide, ammonium chloride, dioctyl phthalate. Suggested actual industrial
dusts are: cement, open hearth fume, paper mill recovery furnace dust,
limekiln dust, sulfuric acid, and incomplete hydrocarbon combustion
products.
52
-------
Where feasible, it is proposed to grade the dusts by size, as was
done in the present study, and to evaluate both closely sized fractions
and composites.
C. Aerosol Chamber Improvement for Aerosol Studies
The present aerosol chamber provides a means of assessing laser
backscatter measurements in terms of aerosol properties. In addition,
a facility to generate large volumes of easily accessible aerosols con-
sisting of known particles at known concentrations would be useful for
a variety of other studies including:
Evaluation of techniques of deriving a valid aerosol sample
Evaluation of direct aerosol samplers
Comparison of various sampling instruments
Development of optical aerosol sensors such as multiwavelength
transmissometers and multiangle photometers
Investigation of multiple scattering effects on instrument re-
sponse functions.
It is recommended that the aerosol chamber be further developed to
provide a unique facility for both remote and direct aerosol studies.
Some improvements include:
The dust feeder system should be modified to produce a more uni-
form particulate feed rate.
Distributors should be placed at the chamber inlet to straighten
air flow within the chamber.
The problem of deriving representative aerosol mass concentration
should be resolved.
The present optical systems should be modified to eliminate drift
in transmissometer readings and permit angular scattering measure-
ments during daylight hours.
53
-------
Instrumentation should be added to provide the aerosol size
distribution measurements.
D. Extending the Accuracy of Backscatter Measurements
Made with Pulsed Lidars
High powered Q-switched ruby and neodymium-glass lasers were used
to collect aerosol backscatter data for this study. The accuracy of such
data could be significantly improved by better monitoring of the trans-
mitted peak power and the total energy of the laser pulses, and by better
knowledge of the target reflectivity at lidar wavelengths used.
Most Q-switched ruby and neodymium-glass lasers exhibit significant
pulse-to-pulse variations in output power and energy. In a well-designed
laser cavity, the pulse shape of the laser output is approximately Gaus-
sian and does not exhibit a significant pulse-to-pulse variation. The
major pulse-to-pulse variations occur in the peak power output of the
laser. For the present study, the peak power was monitored by means of
a beam splitter that directed a small portion of the output pulse onto
a diffusing surface that was viewed by a fiber optic light pickup coupled
to the receiver optics. The output energy was assumed proportional to
the peak power; i.e., it was assumed that the pulse shape remained con-
stant. The use of improved calibration techniques for assessing trans-
mitted energy should result in greater accuracy of the backscatter
measurements.
The laser energy reflected from a solid target of known reflectivity
was used in this study to derive an absolute value of the volume back-
scatter coefficient. A black target was used rather than a standard white
target in order to reduce the target return to a level of the detector
response that is linear. In future backscatter experiments using this
technique, more emphasis must be placed on evaluating the target reflec-
tivity at each lidar wavelength used.
54
-------
E. Computations from Mie Theory
Actual aerosols consist of various sized particles rather than
particles of a single size. However, it is possible to characterize every
phenomenon by some effective average particle size. The effective size
will be different for different phenomena. It is desirable to establish
what average size will characterize each of the common optical phenomena
that may be used to determine physical aerosol characteristics: trans-
mission, forward scattering, backscattering, and 90° scattering. This
can be done from the Mie theory for spherical particles. Although it is
recognized that such calculations may not apply to the irregularly shaped
particles encountered in practice, such theoretical evolutions would be
of tremendous use as a guide to the type of engineering approximation that
might be useful for assessing the optical properties of such actual systems.
In the present study, for example, experimental results indicated that the
Sauter diameter, D , was a good characteristic measure of the transmission
o *j
effects. This was also confirmed by a few spot theoretical calculations
based on Mie theory. For backscatter, Mie theory predicts multitudinous
and abrupt changes of intensity with particle size; however, for typical
particle size distributions, such variations tend to be evened out rapidly
so that the aerosol backscatter may be related to an average particle size.
It is suggested, therefore, that calculations for spherical particles
be carried out for various optical phenomena at various wavelengths,
particle refractive indexes, and particle size distributions. The aerosol
optical parameters should then be related to aerosol physical properties.
F. Development of Multiple Wavelength Lidar Techniques
The results of this study suggest that multiple wavelength lidar
measurements may provide more valid estimates of particulate concentra-
tion than single wavelength lidar measurements. It is suggested that the
two-wavelength lidar data presented in this report be used to investigate
55
-------
techniques of inferring higher moments of the size distribution from such
measurements. Also, since the Mie theory computations relate well to
experimental data for fly ash particulates , it is suggested that the Mie
theory be used to investigate various aspects of multiple wavelength
lidar probing of plume particulates. The number and position of wave-
lengths used and the manner of combining the measurements for an inference
of particulate concentration could be investigated by this means at rela-
tively low cost.
56
-------
REFERENCES
Bryant, H. C., and A. J. Cox, 1966: Mie theory and the glory. J. Opt.
Soc. Am., 56, 1529-1532.
Conner, W. D., and J. R. Hodkinson, 1967: A study of the optical prop-
erties and visual effects of smoke-stack plumes. Cooperative Study
Project, Edison Electric Institute and U.S. Public Health Service.
Cook, C. S., G. W. Bethke, and W. D. Conner, 1971: Development of a
mobile lidar system to measure smoke plume opacity, paper presented
at the Joint Conference on Sensing of Environmental Pollutants,
Palo Alto, California, November 8-10, 1971.
Dave, J. V., 1969: Scattering of visible light by large water spheres.
Applied Optics, _8, 155-164.
Eiden, R. , 1966: The elliptical polarization of light scattered by a
volume of atmospheric air. Applied Optics, £, 569-576.
Ensor, D. S., and M. J. Pilat, 1971: Calculation of smoke plume opacity
from particulate air pollutant properties. J. Air Pollution Control
Association, 21, 496-501.
Evans, W. E. , 1967: Development of lidar stack effluent opacity measur-
ing system. Final Report, Edison Electric Institute, SRI Project
6529, Stanford Research Institute, Menlo Park, California.
Gilbert, G. D., 1970: The effects of particle size on contrast improve-
ment by polarization discrimination for underwater targets. Applied
Optics, 9, 421-428.
Harris, F. S., 1970: Water-and-ice cloud discrimination by laser-beam
scattering. Paper presented at the 1970 Spring Meeting of the
Optical Society of America.
Holland, A. C. and G. Gagne, 1970: The scattering of polarized light by
polydisperse systems of irregular particles. Applied Optics, £),
1113-1121.
57
-------
Jackson, D. VI., 1971: New development of a CW laser radar. Paper pre-
sented at Joint Conference on Sensing of Environmental Pollutants,
Palo Alto, California, November 8-10.
Marquardt, D. W. , 1963: An algorithm for least-squares estimation of
nonlinear parameters. J. Soc. Ind. Appl. Math., 11, 431-441.
Marquardt, D. W. , 1964: Least squares estimation of nonlinear parameters.
IBM Computer Library SHARE program 3094.
Maydan, D., 1970: Fast modulator for extraction of internal laser power.
J. of Applied Physics, 41, 1552-1559.
McDonald, J. E., 1962: Large-sphere limit of the radar back-scattering
coefficient, Quart. J. Roy. Met. Soc., 88, 183-186.
Pilat, M. J., and D. S. Ensor, 1970: Plume opacity and particulate mass
concentration. Atmos. Environment, 4, 163-173.
Schotland, R. M., K. Sassen and R. Stone, 1971: Observations by lidar
of linear depolarization ratios for hydrometeors. J. Applied
Meteorology, 10, 1011-1017.
Twomey, S., and H. B. Howell, 1965: The relative merit of white and
monochromatic light for the determination of visibility by back-
scattering measurements. Applied Optics, 4, 501-506.
Twomey, S., and H. B. Howell, 1967: Some aspects of the optical esti-
mation of microstructure in fog and cloud. Applied Optics, 6^,
2125-2131.
van de Hulst, H. C., 1957: Light scattering by small particles. John
Wiley and Sons, Inc., New York, New York,
58
-------
Appendix A
DETAILS OF THE SIGHTING TUNNEL SYSTEM
-------
Appendix A
DETAILS OF THE SIGHTING TUNNEL SYSTEM
Details of the sighting tunnel system, diagrammed in Figure 1 of the
main text, are discussed in this appendix. A photograph of the system
is shown in Figure A-l.
Air was supplied by means of a Buffalo Forge Type BL, Model No. 365,
Arrangement 9, Class II centrifugal blower, belt driven at 2,080 rpm by
means of a 5-hp motor. This blower discharged into the horizontal duct
system shown in Figure A-2. The pressure difference across the 12-in.-
diameter nozzle located after the blower was used to measure fan air flow
rate.
Dust Feeder
Dust was metered out by a proprietary grooved-disk feeder. The dust
feed rate was determined by the size of the groove and the disk rotational
speed. The dust was exhausted pneumatically from the groove and fed into
the sonic velocity stream as shown in Figure A-3. This sonic stream also
provided the suction for pulling conveying air through the feeder. The
pressure upstream of the sonic nozzle was 50 psig during a normal run.
The compressed air flow was calculated as 33 CFM and the induced feeder
conveying air was estimated to be on the order of 2 CFM.
Separate feeder calibrations were carried out to establish the feed
rate as a function of disk speed for each dust. These were conducted by
pressurizing the feeder with a metered amount of compressed air (1.3 CFM)
that carried the dust from the groove into a separate collecting bag. The
A-3
-------
collecting bag was weighed over timed intervals. The results of these
calibrations are shown in Table A-l. The dust feed rate was usually
linear with disk speed; in these cases the average value of the grams/
revolution measured at all dust speeds was used in all subsequent calcu-
lations of feeder rates. In a few cases a significant variation in grams/
revolution was observed; thus no average value is shown for them in Ta-
ble A-l, and the calibration factor established for each disk speed was
used in subsequent calculations.
Special tests were conducted to observe the effect of the compressed
air pressure on dust dispersion. In the tests the transmissometer reading
was recorded on a strip chart for a timed interval as the compressed air
pressure was changed in steps for a given feeder disk speed setting. Typ-
ical results are shown in Table A-2. The test results would indicate that
a reproducible state of dust dispersion is obtained for pressures over
30 psig. At lower pressures the apparent dispersion is progressively
poorer, as reflected by the greater transmissometer readings. The sharp
drop in apparent dispersion at pressures below 30 psig for the speed con-
trol setting of 100 was partially due to a failure of the feeder groove
to be unloaded completely at that speed because of the lower induced air
at the lower air pressures. At a speed control setting of 40 or less,
the groove was essentially unloaded at all air pressures. Even at a set-
ting of 100, no more than 20% of the groove was not unloaded. Thus, any
increase in transmissometer reading must reflect primarily the effect of
dust dispersion and not any actual reduction in dust feed rate.
Aerosol Chamber
The dust-laden air stream is gradually decelerated in the expanding
ductwork and allowed to enter the plenum chamber beneath the sighting
tunnel. In the original equipment Arrangement A, the top of the plenum
A-4
-------
chamber was equipped with three rectangular nozzles pointing upward. Two
were at either end as shown in Figure A-2. These two nozzles each had a
throat dimension of 3 in. x 20 in. The third nozzle was located in the
center and had a throat dimension of 2 in. x 20 in. The sighting tunnel
was filled with a controlled aerosol by the dust-laden air coining into
it through the central nozzle. The two end nozzles served as an aerosol
curtain to carry away any outside air blown into the tunnel and hence to
establish a fixed boundary to the dust-laden air through which the lidar
would sight.
Because sampling problems had become obvious in the first few runs,
the central nozzle was replaced after Sampling Run 8 with the multihole
distributor shown in Figure A-2 to minimize any effects on sampling that
the high velocity from the central jet might have. This arrangement
(Arrangement B) was used throughout the rest of the study.
Air Rates
With Arrangement A the air flow could be measured by either the pres-
sure differences across the 12-in. nozzle after the fan or by the drop
across the rectangular nozzles. The total flow measured by these two
methods agreed within 2%. The pressure drop across the plenum chamber
nozzles was slightly less with Arrangement B than with Arrangement A and
the system air flow increased slightly (approx. 2rc) for Arrangement B.
The total flow with Arrangement B was taken as 5800 CFM, including com-
pressed air introduced at the feeder. It is believed that this value
should be correct within ±3%; the value also agrees well with the flow
to be expected from the fan performance curve provided by the manufacturer.
Although the air flow through the system could have been varied by pro-
viding a throttle at the fan inlet, this was not done. All data were taken
at this same flow rate. The flow rate also changed slightly with air
A-5
-------
temperature. Since this variation should have been less than ±3%, it was
not allowed for. Because air rate variations with temperature were also
not allowed for in the sampling system, the effect of temperature is
largely self-compensating in so far as overall concentration comparisons
are concerned.
Flow Pattern
The air flow pattern in the sighting tunnel was established qualita-
tively by probing with short streamers on a rod. Apparently a marked
rotary circulation in the tunnel is induced because the entrance to the
plenum chamber is located on one side. The central 55-hole distributor
alone is apparently not sufficient to eliminate the side velocity com-
ponent. Viewed from the inlet end (east end) of the tunnel, the rotary
flow was clockwise in both halves of the tunnel.
Dust Deposition
During the course of the test two inspections were made of the inside
of the aerosol chamber system by disconnecting the 4-ft X 18-in. transition
piece. In both cases it was observed that there was an obvious deposit
on the floor of the plenum chamber, the sighting tunnel, and the transition
piece. The sidewalls and roof had only a slight coating that was negligible
compared with that on the floor. The ductwork upstream of the transition
piece and including the 3-ft upstream portion of the transition piece was
essentially totally void of any dust deposit.
In the inspection on February 2, 1971 (after Sampling Run 8), a por-
tion of the deposit on the floor of the transition piece was collected by
brushing and was weighed. The weight indicated a deposit density of ap-
proximately (l/3) gram/sq ft. Because this was the region of heaviest
deposit, it was estimated that the deposit in the entire chamber system
A-6
-------
totaled less than 50 g. Since this represented the entire accumulation
after an estimated 2 to 3 Ib of dust had been fed to the system, it may
be concluded that wall deposit within the chamber amounted to less than
5% of the dust feed. These measurements were made rather crudely, however,
and it was possible that the (l/3) gram/sq ft value was low because of
windage losses.
On October 4, 1971, the chamber was opened for the second time and
was cleaned out completely with a small hand-vacuum cleaner. The dust
from the various parts of the system was collected separately in a bag
filter and weighed. The results are shown in Table A-3. These measure-
ments also indicated that less than 5% of the dust feed was retained on
the chamber system walls.
The material collected on both of the above occasions had a Fisher
diameter (corrected for slip flow) of 6.7 and 6.0 LL , respectively. Exam-
ination under the microscope showed it to be substantially all in the 5-
to 10-|j, range.
Because the chamber was subject to random vibrations and to deliberate
rapping during the experimental program, it is not possible to be certain
that a greater wall deposit had not accumulated but was dislodged because
of this mechanical action. Considering the low velocities (4 to 8 ft/sec)
in the various parts of the chamber, it is not likely that significant
reentrainment of deposited dust occurred.
Transmissometer
A transmissometer was mounted in the sighting tunnel. The collimated
beam generator or light source was located at the upper right of the inlet
end of the tunnel and the detector at the lower left end of the outlet end.
Both light source and detector were located outside the aerosol zone in
ambient air. The transmissometer details are given in Appendix D.
A-7
-------
(a) FRONT VIEW OF AEROSOL CHAMBER SHOWING
LASER PULSE ENTRANCE
(b) SIDE VIEW OF AEROSOL CHAMBER WITH
PARTICLE FEEDER AND INSTRUMENTATION
FOR THE OPTICAL SENSORS
FIGURE A-1 AEROSOL CHAMBER USED IN THE BACKSCATTER EXPERIMENTS
A-8
-------
EXHAUST AIR
TO ATMOSPHERE
6'
20"
SOUTHEAST
SIDE '
20"
-NORTHWEST SIDE
'OPEN ENDS (20" x 20")
2" HALF COUPLING
,DUST FROM FEEDER
COMPRESSED
AIR
PRESSURE TAPS
(1/8" Hole,
1/8" Nipple)
FLEXIBLE
COUPLING
V.
SAMPLE HOLE NO. 8
WOODEN STAND
(a) ELEVATION
12" o.d.
18" o.d.
FAN
DRIVE
FAN OUTLET
20" x 14 3/8"
7 l_ 1
/ AA
' IN
(b) TOP VIEW
18"
-6" R
32'-
(c) INSIDE ELEVATION OF TUNNEL (Section A-A, Northwest Side)
COLLIMATED
LIGHT BEAM
EXHAUST
18"
6'
20"i
AEROSOL
CURTAIN
SIGHTING TUNNEi.
ATMOSPHERIC
AIR
12"
u^v. i [-34-- | 1(r
36" 55 1 1/4" HOLES /\ } jl k
| ~"" ~ " ~ PLENUM CHAMBER
00'
J\.
^^^
DETECTOR
(d) SECTIONAL VIEW OF CONTROLLED AEROSOL CHAMBER (Section B-B)
TB-8730-19
FIGURE A-2 AEROSOL CHAMBER DETAILS
-------
AIR TO
PLENUM
STAINLESS STEEL TUBE
5/16" o.d., 1/4" i.d. '
RUBBER STOPPER
12" o.d. DUCT
1" PLUG WITH
3/8" HOLE
1
AIR AND DUST
FROM FEEDER
2 CFM
1/4" i.d. RUBBER HOSE
1" PIPE CAP
-1" PIPE
1" ELL
50 psig
COMPRESSED AIR
33 CFM
AIR FROM FAN
5750 CFM
FIGURE A-3 DUST DISPERSER DETAILS
TA-8730-20
A-ll
-------
Table A-l
DUST FEEDER CALIBRATION DATA
Data Taken
Groove Dimensions
Centerline Fly Speed
Diameter Depth Width Ash Control
(in.) (in.) (in.) Fraction Setting
3.50 1/16 1/8 0-10 t 15
30
40
(Groove volume = 50
1.408 cc/rev.) 60
80
100
200
5-10 u. 20
40
100
2.5-5 14 20
40
100
0-2.5 (i 20
40
100
3.50 1/8 1/2 0-10 p 20
20
(Groove volume = 40
11.27 cc/rev.) 100
5-10 p 20
40
100
2.5-5 |i 20
40
100
0-2.5 u. 20
40
100
Number
Revolutions
of
Disk
3
18
3
18
3
18
18
18
18
18
18
18
1
3
3
3
3
3
3
3
3
3
3
3
3
Powder
Collected
( Grams )
32.5
42.9
40.9
54.5
89.2
90.2
86.1
70.9
5.1
29.4
5.2
29.3
4.8
28.4
26.9
26.3
25.2
18.1
16.1
15.3
12.3
36.9
36.1
35.8
39.3
36.1
33.1
34.7
34.5
34.1
22.7
23.2
22.5
Time
(Sec.)
4620
1980
1200
1200
1800
1200
900
600
432
2882
152
911
57
327
2844
913
317
2840
943
320
157
486
154
54
490
155
57
492
156
54
480
160
53
Disk
Speed
( RPM )
0.255*
o.sost
1.224t
1.643t
2.05 *
2.81 *
3.58 t
4.20 t
Av,
0.417
0.374
1.183
1.183
3.16
3.30
Av,
0.379
1.183
3.40
0.380
1.145
3.37
0.382
0.370
1.17
3.34
0.367
1.16
3.16
0.366
1.15
3.33
Av.
0.375
1.13
3.40
Av.
Feed
Rate*
(g/rev)
1.66
1.61
1.67
1.66
1.45
1.61
1.60
1.69
. 1.64
1.70
1.63
1.73
1.63
1.60
1.58
. 1.63
1.50
1.46
1.40
1.005
0.895
0.850
12.3
12.3
12.0
11.9
13.1
12.0
11.0
11.6
11.5
11.4
11.5
7.58
7.73
7.50
7.60
Date
11/17/70
9/23/71
9/23/71
9/23/71
2/17/71
2/26/71
2/18/71
2/26/71
Determined by collecting dust for timed period (10 to 80 min) and timing 1 rev. of disk for tests conducted
on 11/17/70; all other determinations made by timing and weighing dust collected in a precise number of
disk revolutions. For sp. gr. of 2.46 g/cc the following are calculated volume voidages, e:
Groove Dust g/rev. s
1/16 x 1/8 0-10 n 1.64 0.527
0-2.5 u. 0.850 0.755
1/8 x 1/2 OrlO p, 12.0 0.567
0-2.5 ti 7.60 0.726
.Measured by duplicate determination of time for 1 revolution.
A-12
-------
Table A-2
EFFECT OF COMPRESSED AIR PRESSURE ON DUST DISPERSION*
Transmissometer
Speed Compressed Air Reading at
Control Pressure Equilibrium
Setting (psig) (percent)
100 50 38
40 38
30 41
20 55
10 63
40 50 72
40 71
30 71
20 77
10 90
20 10 92
* General conditions: Fraction dust fed: 2.5 to 5
Disk groove: 1/8 in. x 1/2 in.
A-13
-------
Table A-3
SUMMARY OF DUST COLLECTED INSIDE AEROSOL DUCTWORK AND TUNNEL*
Weight Collected
g [g/sq ft]
Transition Section 41 [2.5]
Plenum Chamber 311 [5.8]
Central Section (6 ft long) 74 [7.4]
3 ft NE and adjacent to
Central Section 30 [6.0]
3 ft SW and adjacent to
Central Section 46 [9.2]
10 ft balance of NE section 89 [5.3]
10 ft balance of SW section 72 [4.3]
Wind Tunnel 155 [3.1]
Central distributor (3 ft long) (20)t [4]
10 1/2 ft NE and adjacent to
Central distributor 67 [3.8]
10 1/2 ft SW and adjacent to
Central distributor 51 [2.9]
3 ft inside of NE aerosol curtain 6 [1.2]
3 ft inside of SW aerosol curtain n [2.2]
Total Collected* 507 = 1.1 Ib
*
Substantially no deposit in duct upstream of transition section;
upstream 3 ft of transition section also almost free of deposit.
Weights given above were swept up from bottom with a vacuum
cleaner. The sides contained a coating of dust but the quantity
of dust was negligible (< 20%) compared with deposit on bottom.
All weights in grams. Values in brackets are approximately
g/sq ft.
t Not measured or collected but estimated by noting depth and nature
of deposit relative to other measured sections.
* Total dust fed into system prior to these measurements was
approximately as follows:
Amount Fed
Dust Fraction (Ib)
0-2.5 ^ 9.0
2.5-5 |j, 7.7
5-10 u. 15.1
0-10 u 9.2
Total 41.0
A-14
-------
Appendix B
FLY ASH PREPARATION AND PROPERTIES
-------
Appendix B
FLY ASH PREPARATION AND PROPERTIES
Source
The raw fly ash used in this project was supplied by Diamond Aggregate
and Fly Ash Company (501 Eleventh Avenue South, Minneapolis, Minn. 55415)
to the Donaldson Company, Inc. (1400 West 9th Street, Minneapolis, Minn.
55431) for further treatment. Diamond Aggregate had this material in
storage, but had originally obtained it from Northern States Power
Company's Blackdog Power Plant in Minneapolis. It represents combustion
of Southern Illinois bituminous coal during January 1969. The Blackdog
power plant, built in 1954, has 4 boilers and 4 stacks and is equipped
with electrostatic precipitators. It has an output of 475 megavolts when
all four boilers are in use. Loading in the winter time is about 65-^.
It is the 4th largest power plant in the area and the only one from which
fly ash is readily available. The plant generates approximately 50,000 tons
of fly ash per year.
Treatment
The raw fly ash was delivered to the Donaldson Company, Inc. (later
designated DCl) for size classification in their proprietary centrifugal
air classifier. This classifier was originally developed for Donaldson
by SRI. Its unique features are the provisions for achieving dust dis-
persion, which is normally a limiting factor in achieving sharp size
classification in the range below 10 |j,. It was also developed to be
capable of production capacities corresponding to powder feed rates up
to 50 Ib/hr.
B-3
-------
Prior to classification, approximately 1000 Ib of the fly ash were
screened through a Tyler 8-mesh sieve to remove traces of paper, gravel,
and other superfluous debris. The screened material was then passed
successively through the classifier at settings designed to give various
splits. The complete flowsheets for all classification runs are given
in Figure B-l together with resultant material balances.
The first lot was successively classified at cut sizes of 10, 2.5,
and 5u, to yield the fractions indicated. A second lot was then classi-
fied at 10 u, to yield a 0 to 10-u, fraction. The 0 to 10-|i fraction ob-
tained for the first lot has been designated as 0 to 10 u, (F) to dis-
tinguish it from the corresponding fraction obtained from the second lot.
All the 0 to 10-p, (F) material was used up for the successive classifica-
tions. The 0 to 10-|j, and the 0 to 10-p, (F) fraction should have been
nominally identical. There was, however, a slight difference in the
operating conditions at which each was obtained with a corresponding
slight difference in results, the 0 to 10-u, fraction being cut at a
slightly finer size than the 0 to 10-|j, (F) fraction.
In the classification system, the fines are collected in a cyclone.
Any fines escaping the cyclone are collected in a filter. Because of the
small amount of material involved , all filter fractions from all the
classification runs were combined into a single fraction called "Super-
fines," as indicated in Figure B-l. Although a total of 9.0 Ib were col-
lected in the filter as determined by direct weighing, only approximately
5 Ib were physically recovered , the remainder being either left in the
pores of the filter medium or lost during shaking of the filter to remove
the dust deposit.
B-4
-------
Properties
Particle size analyses were conducted by DCI using the Fisher Sub-
Sieve Sizer, the Coulter Counter, and the MSA-Miitby sedimentation
techniques. The Fisher analysis was checked by Stanford Research Insti-
tute (SRl). The results of all analyses are summarized in Table B-l and
plotted in Figures B-2 and B-3. The mass media diameters given in
Table B-l were obtained from the 50% intercept of the curves shown in
Figures B-2 and B-3. The sp gr of material in all fractions was assumed
to be 2.46 for purposes of any calculations for interpretation of
results of size analyses. It is likely that the sp gr will vary with
the fractions although such measurements were not made during the study.
A summary of the mean sizes derived from each method is given in Table
B-2.
The data shown in Figures B-2 and B-3 can be approximated by a
log-probability relationship. This was done by visually fitting the best
straight line to the data of Figures B-2 and B-3 giving emphasis to the
data in the 10- to 90-cumulative percent range and ignoring those outside
of that range. These straight lines can then be described in terms of
two parameters, a mass media diameter and a standard geometric deviation.
These parameters are summarized for each dust fraction in Table B-3
>r=
together with corresponding calculated values of other mean sizes
(D and D ). Since the Fisher Sub-Sieve Sizer supposedly measures
32 mO
D values, the Fisher data have also been included in Table B-3 for
32
comparison.
As might be expected all methods yield different particle size
results since different particle properties are assessed by each method.
* M
The calculation methods are described by C. E. Lapple in Particle-
Size Analysis and Analyzers," Chem. Eng., 7_5 (11), 149-156, (Way 20,
1968)
B-5
-------
Table B-4 lists suggested effective values in terms of optical properties
for each of the dusts employed in this study. These were derived as a
matter of personal judgment assuming that: (1) the optical effects
are most nearly related to D values; (2) the Fisher Sub-Sieve Sizer
o 2t
value comes closest to measuring D ; and (3) the distribution (in terms
O &
of standard geometric deviation) can be taken from the MSA and Coulter
Counter data. Because of these brash assumptions, these effective
values cannot be considered as having any great precision but they are a
as good as can be obtained in the present state of the art. The values
of D and D given in Table B-4 have been calculated from the quoted
m3 mO
*
value of D and a by standard methods.
32
The fractions were examined at SRI both in the optical microscope
and in the scanning electron microscope. Photomicrographs of the various
fractions are shown in Figures B-4, B-5, B-6, and B-7.
No chemical analyses were made on the fly ash used in the project.
However, Diamond Aggregate and Fly Ash Company in 1966 supplied the
data given in Table B-5 on fly ash samples from two other Northern
States Power generating plants in the Minneapolis/St. Paul area.
These analyses are averages from samples taken over more than twelve
months of operation.
The superfine, 0-2.5^1, 2.5-5|i, and 5-lOp, fractions were prepared
by classifying the 0-10(1 fraction. The size distribution of the 0-10(1
fraction can be calculated from the mass balance data of Figure B-l and
the Fisher diameter of each of the composite fractions, assuming that the
Fisher diameter is a reasonably good measure of the mass median diameter
of each of the composite fractions. Since the standard geometric devia-
tion of these composite fractions is of the order of 1.5, it can be shown
Ibid
B-6
-------
from the methods described by Lapple* that the mass median diameter will
be only about 9% greater than the D diameter (which is presumably the
O iL,
Fisher diameter). On this basis the 0-10- fraction would have the follow-
ing size distribution:
Particle Diameter Cum. tt't. To
microns Finer Than
6.5 x 1.09 = 1.7 67.8
3.1 x 1.09 = 3.4 22.4
1.7x1.09 = 1.85 5.5
0.77 x 1.09 = 0.84 0.8
A plot of these data yields a mass median diameter of 5.4 microns.
It should also be noted that the Fisher size of the 0-10,. fraction
calculated from the measured Fisher sizes of the component fractions is
4.0^. This checks exactly with the measured Fisher size for the 0-10^-
fraction. In all the above calculations the SRI measurements of Fisher
size shown in Table B-l have been used.
*Ibid
B-7
-------
LOT
NUMBER
FIRST
RAW FEED
INTER-
MEDIATE
FRACTIONS
AND FEEDS
LOSS'
CLASSIFI-
CATION
PROCESS f
PRODUCT
FRACTIONS
SECOND
* Loss obtained by difference between feed weight and total weight of recovered fractions.
t Mechanical failure occurred early in run and both coarse and fine fractions obtained prior
to failure were discarded.
t Rates given are powder feed rates to classifier.
FILTER
MATERIAL
TA-8730-11
FIGURE B-1 FLOWSHEET FOR CLASSIFICATION PROCESSES
B-8
-------
CUMULATIVE WEIGHT PERCENT OVERSIZE
99 98 95 90 80 70 60 50 40 30 20 10 5 21 0.5 0.20.1
O Raw fly ash
O 0-10/J (F)
0-10M
O 5-10M
A 2.5-5M
D 0-2.5/J
V Superfines
0.3
0.1 0.2 0.5 1 2 5 10 20 30 40 50 60 70 80 90 95 98 99 99.8 99.9
CUMULATIVE WEIGHT PERCENT UNDERSIZE o,_ ,
TA8730-1 2
FIGURE B-2 MSA-WHITBY PARTICLE SIZE ANALYSES OF BLACKDOG FLY ASH
B-9
-------
99 98
CUMULATIVE WEIGHT PERCENT OVERSIZE
95 90 80 70 60 50 40 30 20 10
1 0.5 0.2 0.1
O Raw fly ash
(F)
0-
O 5-10H
A 2.5-5M
O 0-2.5M
V Superfines
0.10.2 0.5 1 2 5 10 20 30 40 50 60 70 80 90 95 9698 99 99.899.9
CUMULATIVE WEIGHT PERCENT UNDERSIZE T o,on
T A-873Q- i 3
FIGURE 8-3 COULTER COUNTER PARTICLE SIZE ANALYSES OF BLACKDOG FLY ASH
B-10
-------
,~<
-°
>-v
O-OQ
iQ M
O
(a) RAW FLY ASH
(b) 0-1 Op FRACTION
*
«f b
,
P *
*^
|c) 5-1 Op FRACTION
,
* . -
." "
IF
-
^
-
0
o a ° ° ,0
'
«#
^4n? 5
- -. ' o p '
o
« O
>
Uj-v-mol
(d) 2.5-5.0P FRACTION
r J>J ^- " ,W***^
,
-
^>
^r
'
"
' - ' - .'TI-I'F r^.rri\ftl
1 . - -, '*. - *
(e) 0-2.5P FRACTION If) SUPERFINE
SMALLEST DIVISION ON SCALE = 6.7/J
FIGURE B-4 PHOTOMICROGRAPHS OF BLACKDOG FLY ASH
TA-8730-14
B-ll
-------
29 .D
r
I
i
£r- ""-. ?* ^V-'*V>:. &- ,
.0 \ 0*.?? #&f.
.* » - . * 4 ' -./''
. ^ q - .
MP
V-i
(a) RAW FLY ASH
. -
.
- '
;gr o'f
o
> {.A|
/
,
k V* -
. '
. *
^ 'i *
*-
ft
, .^. v ,_v %-, ^ ;vv- . /
*.'*' '«.^ * *A'~ '**,' ' * +
* " -' '_»». ' " f » s
-< . »
v, ..- . 1 .
(c) 5-Wp. FRACTION
NUMBER ABOVE EACH COLUMN GIVES MAGNITUDE OF SMALLEST
SCALE DIVISION IN EACH PHOTOGRAPH IN THAT COLUMN
TA-8730-15
FIGURE B-5 ADDITIONAL PHOTOMICROGRAPHS OF BLACKDOG FLY ASH AT LOWER
MAGNIFICATION
B-12
-------
(a( 0-10M FRACTION
(b) 5-10JU FRACTION
-*
(c) 2.5-5M FRACTION
(d) 0-2.5M FRACTION
o O
0
2.5 H 5jU 10/J
FIGURE B-6 SCANNING ELECTRON MICROGRAPHS OF BLACKDOG FLY ASH
8-13
-------
! '
a-
(a) RUN 74A (0-2.5-fI Fly Ash)
(b) RUN 80A (2.5-5-M Fly Ash)
3 |U
NOTE: Filter paper: G.E. "Nuclepore", '\~ll pore size.
TA-8730-21
FIGURE B-7 SCANNING ELECTRON MICROGRAPHS OF FLY ASH DEPOSITS ON SAMPLING
FILTER PAPERS
-------
Table B-l
SUMMARY OF SIZE ANALYSIS DATA ON BLACKDOG FLY ASH
W
l
0 - !U[i (F)
9.0 u
7.2 .>.
5.69 >
9.8 (}i!. 1 HI . 1
9 . -I 17.1 y 7 .
,1.61
3.U7
99.5
SH.3
93.9
.1. fi 10 . 0 5.3 3 .
] . 2 43 1.0 1 .
2, 46
0. 51
-------
Table B-2
SUMMARY OF MEAN PARTICLE DIAMETERS OF BLACKDOG FLY ASH
Size Analysis Method
Fisher Subsieve Sizer, diameter
corrected for slip-flow , u,
DCI
SRI
Coulter Counter, mass median
diameter , |_i
DCI
MS A- Whit by Sedimentation, mass
median diameter, u,
DCI
Fraction
Raw
Fly Ash
8.4
8.11
13.7
5.0
o-io u, (F)
4.5
4.00*
6.2
2.75
0-10 u,
4.4
3.95
5.0
2.70
5-10 u,
7.7
6.47
8.8
5.7
2. 5-5 |j,
3.9
3.10
4.1
2.97
0-2.5 p,
2.1
1.70
1.93
1.85
Superfine
0.64
0.77
1.43
1.77
w
i
M
Calculated from analyses of component fractions.
-------
Table B-3
LOG-PROBABILITY APPROXIMATIONS TO
SIZE DISTRIBUTION OF EACH FRACTION OF
BLACKDOG FLY ASH
Diameter, micron
Fraction
Raw Fly
Ash
0-10i_i(F)
0-lOu
5-10u
2.5-5[a
0-2. 5u
Superfine
*
Size
Analysis
Method
CC
MSA
F
CC
MSA
F
CC
MSA
F
CC
MSA
F
CC
MSA
F
CC
MSA
F
CC
MSA
F
Standard
Geometric
Deviation
2
2
2
1
2
1
1
1
1
1
1
1
1
1
.74
.50
-
.41
.78
-
.48
.93
-
.51
.70
-
.50
.62
-
.52
.46
-
.64
.61
-
+
Mass '
Median
D
13
5
5
2
4
2
8
5
4
2
1
1
1
1
m3
. 5
.0
-
.80
.75
-
.75
.70
-
.60
.40
-
.00
.90
-
.85
.80
-
.46
.80
-
j.
Sauter "
D
8
3
8
3
2
4
3
2
4
7
4
6
3
2
3
1
1
1
1
1
0
32
.12
.28
.1
.94
.33
.0
.13
.17
.0
.90
.74
. 5
.68
.58
.1
.70
.67
.70
.29
.61
.77
j.
Number "
Median
D
mO
0
0
0
1
0
0
5
2
2
1
1
1
0
0
.64
.40
-
.57
.01
-
.40
.74
-
.17
.32
-
.44
.44
-
.09
.17
-
.70
.91
-
CC = Coulter Counter; MSA = MSA-\Vhitby Sedimentation; F - Fisher
Sub-Sieve Sizer
'Obtained from best straight line through data in log-probability paper
in 10 to 90 cumulative percent range
*.
Calculated from 3 and D
m3
B-17
-------
Table B-4
RECOMMENDED "AVERAGE" SIZE DISTRIBUTION
FOR EACH FRACTION USED IN STUDY
Fraction
0-lOp.
5-10)0.
2.5-5p,
0-2.5)1
Standard
Geometric
Deviation
2.2
1.5
1.5
1.5
Diameter, microns
Mass
Median
ra3
5.45
7.06
3.36
1.85
Sauter
D
32
4.0
6.5
3.1
1.7'
Number
Median
mO
0.85
4.31
2.06
1.13
B-18
-------
Table B-5
TYPICAL FLY ASH ANALYSES
Generating Plant (Station)
Location
Physical Analyses
Specific gravity
Sieve analysis
"3 passing 200 mesh
73 passing 325 mesh
Blaine, surface area, sq cm/g
equiv diam, |_L
Hygroscopic moisture, /&
Chemical Analysis, $
A12°3
Fe2°3
CaO
MgO
S°3
Ignition Loss, 79
High Bridge
St. Paul
2.36
92
81
3110
8.15
0.02
49.6
Riverside
Minneapolis
2.50
97
91
3491
6.87
0.11
47.6
18.0 | 16.9
18.5 21.9
4.8 4.3
0.8 0.8
1.3 1.0
3.4 3.3
B-19
-------
Appendix C
DUST CONCENTRATION CONTROL AND MEASUREMENT
-------
Appendix C
DUST CONCENTRATION CONTROL AND .MEASUREMENT
Dust Concentration Control
The dust concentration in the system was controlled by establishing
known rates at which both air and dust were admitted to the chamber sys-
tem. The dust feed rate was controlled by the size of groove in and the
rotational speed of the disk used in the feeder described in Appendix A.
The total air flow rate was essentially constant throughout all the tests,
a given fan being operated at a fixed speed on a system of fixed geometry.
The air rate was measured and monitored as described in Appendix A.
Calculation of Dust Concentration
Since both dust and air were admitted to the aerosol chamber at fixed
and known rates, the dust concentration existing in the sighting tunnel
can be calculated if it is assumed that there is no deposition or accumu-
lation of dust within the chamber. This calculated concentration (termed
"feeder concentration") is obtained by dividing the dust feed rate by the
total air rate. The dust feed rate is established from the feeder calibra-
tion data given in Appendix A and the feeder disk rotational speed. Al-
though the disk speed was determined by the speed control setting, some
small speed differences were noted at a given setting presumably because
of differences in frictional resistance in the drive. For this reason
the disk speed was measured in each run by a microswitch that was acti-
vated momentarily once every revolution of the disk. This generated a
mark on the transmissometer strip-chart recorder from which the actual
disk speed could be calculated for each run.
C-3
-------
Recognizing at the start of the program that dust deposition in the
aerosol chamber could be a problem, the dust sizes and air rates were
chosen so as to minimize this difficulty. Although it was not feasible
to eliminate entirely deposition with any reasonable system, it was ex-
pected that such problems should be minor, reflecting a change in dust
concentration of less than 10%. To achieve this, however, it was also
necessary to disperse completely the dusts used. The compressed-air
disperser described in Appendix A was designed to provide as good a dis-
persion as possible with known technology.
Concentration Measurements Techniques
To provide a separate check on concentration, separate measurements
of concentration were made by sampling the dust-laden stream from the
sighting tunnel. During the course of the program three different sam-
pling devices were used.
In the beginning of the program a simple probe inserted into the
aerosol chamber at right angles to the nominal flow was used as shown in
Figure C-l. This arrangement was designated as the "T" arrangement. Al-
though it is customary to use isokinetic sampling when dealing with par-
ticulates to avoid sampling bias caused by particle inertia, this would
have been very difficult to achieve because of the low air velocities in
the chamber. It was also felt that with most of the dusts the inertial
sampling bias should not be too severe and hence the simple form of sam-
pling probe was used.
Air was pulled through the sampling train at a constant and metered
rate by connecting the sampling system to a high vacuum line through a
critical flow orifice (approximately 0.025-in. diameter). This orifice
was calibrated against a gas prover. Sampling rate was held constant at
a predetermined value by maintaining a given upstream pressure on the
C-4
-------
orifice by means of a valve located upstream of the orifice. The pres-
sure downstream of the orifice was always recorded to be certain that
critical flow was being realized. A sampling rate of 2.55 x/min was used.
The filter medium was a 1-in.-diameter General Electric ''Xuclepore" me-
dium of 1-u. pore size. One side of this medium is smooth and the other
is matte. The dust was allowed to deposit on the matte side. This filter
medium was used because of its light weight (5 to 6 mg) and low moisture
retention (less than 0.01 mg, as established in another study). Before
and after sampling, the filter disks were stored in an aluminum cup that
in turn was placed in a closed plastic box. The aluminum cup was used
to mitigate the possible repulsive forces arising from electrostatic
charges on the filter medium.
The second sampling technique used the Delrin filter holders shown
in Figure C-2. These filter holders were placed over the 1-in. sampling
ports or located directly in the aerosol stream. Two such units were al-
ways operated at the same time (although not at the same sampling loca-
tion) and air was drawn through each at a rate of 2.55 x/min by a separate
critical-flow orifice similar to that used for the first sampling tech-
nique. The same "Nuclepore" filter medium was also used. The filter
holders were used in four orientations: (l) pressed over the 1-in.
sampling ports with the filter medium parallel to the chamber walls;
(2) vertical,facing into the nominal direction of air flow, so that the
filter face is normal to the nominal air flow; (3) horizontal with the
filter medium facing upward and parallel to the air flow; and (4) hori-
zontal with the filter medium facing downward and parallel to the nominal
air flow. As will be noted from Figure C-2, there were two types of
holders: (l) an "open" filter in which essentially the entire filter
medium (13/16-in. diameter free area) was exposed directly to the air stream;
and (2) a "closed" filter in which the air must pass through a nominal
C-5
-------
5/16-in.-diameter hole before arriving at the filter medium. It should
be noted that the air velocity in the sighting tunnel was nominally
4.4 ft/sec (133 cm/sec) whereas, at a sampling rate of 2.55 £/min, it
was almost the same [2.8 ft/sec (86 cm/sec)] in the inlet opening of the
closed filter. In the "open" filter with an inlet opening of 13/16 in.,
the velocity was 0.42 ft/sec (13 cm/sec).
A few runs were made with the sampling arrangement shown in Fig-
ure C-3. This entire unit was inserted into the sighting tunnel with
the nozzle mouth facing into the air stream. Large volumes of air were
thus sampled by a vacuum cleaner, and dust was collected in a paper vac-
uum cleaner bag in relatively large quantity. The air flow was measured
by the inlet nozzle to the system. The velocity at the mouth of the
nozzle (2-in. diameter) was of the order of 15 ft/sec (460 cm/sec). The
air flow in this case was determined largely by the resistance of the
bag. Although the air flow was reasonably constant during a run, it
varied markedly with the way in which the bag was inserted in the holder
because of the close clearance between wall and bag.
With all the sampling arrangements, samples were taken at various
locations, indentified in Figure C-4.
Sampling Results
The details of data taken during the series of samplings with sam-
pling arrangement "T" are given in Table C-l. In these runs the sampling
arrangement was held constant and samples were taken at a variety of lo-
cations with three of the dust fractions. It was during the course of
these measurements that the Mark V lidar shots were made.
C-6
-------
Since it was obvious that there were problems in obtaining a correct
sample, at a later date a second series of runs were made in which the
feed rate was held essentially constant (at the same speed control setting),
and the type of sampler and the sampler orientation and location were var-
ied with each of the dusts. These results are given in Table C-2.
To give a simple indication of the sampling runs, the data were
reported as a ratio of the concentration calculated from the feeder speed
to that measured on the sampling filter. These ratios are summarized in
Table C-3 for the data of Table C-l and in Table C-4 for the data of
Table C-2.
With the "T" sampling arrangement, the data of Table C-3 shows a
distinct trend of the concentration ratio with dust fraction. Although
there is some trend with sample location, this is overshadowed by the
effect of dust fraction. In the second series of runs (Table C-4), the
concentration ratio also shows a distinct trend with dust fraction although
some effect of sampling arrangement is demonstrated. In both cases this
is best illustrated by the following arithmetic averages of all concentra-
tion ratios measured for each dust fraction, leaving out concentration
ratio values that are obviously grossly in error.
AVERAGE CONCENTRATION RATIO (Feeder/Filter)
Dust
Fraction
0 -.2.5 u,
2.5 - 5 u.
5 - 10 u.
0 - 10 p.
For "T" Sampling
Arrangement
(Table C-3)
1.28
2.08
3.61
For Second Series of Runs
(Table C-4)
All
Data
0.98
1.53
1.75
1.83
Data for CHU
Arrangement Only
1.02
1.65
2. 39
2.10
Data for OHU
Arrangement Only
0.94
1.42
1.48
1.66
C-7
-------
It is apparent that the concentration ratio is essentially greater
than unity and increases with increasing dust particle size. A concen-
tration ratio greater than unity means that the concentration measured
with the filter is less than would be calculated from the dust feed rate.
Uniformity of Aerosol Concentration
The best indication of the overall degree of uniformity of the aero-
sol concentration in the chamber is through the transmissometer strip-
chart readout. Figure C-5 shows typical recordings obtained for the
various dusts at a variety of dust concentration levels. The run numbers
correspond to those cited in Tables C-l and C-2. Figure C-5 gives the
nominal transmissometer readout as a function of time. The range in the
transmissometer scale is nominally 20% to 100%, with 100% being at the
left. The chart was calibrated during each group of runs since there
was some drift in calibration and the readout was not quite linear. It
is for this reason that the transmissometer scale is termed nominal. The
100% level is almost correct on an absolute basis but the indicated 20%
reading may have been as high as 25% on an absolute basis. The strip
chart was operated at its lowest speed of 1 mm/sec in all cases, with the
smallest time-scale divisions being 5 mm (hence, 5 seconds) apart.
From Beer's law, the aerosol mass concentration is related to trans-
mission by
w = K
P
where
3
w = mass concentration, mg/m
P
T = fractional light transmission, dimensionless
3
K = proportionality factor, mg/m
C-8
-------
By differentiation
Aw
P
w
P
P -,
AT
T £n(l/T)
Here Aw /w is the fractional change in aerosol concentration with a change
in transmission, AT. For values of T greater than 70<>, a close approxima-
tion is:
Aw
P
w
P
AT
1 - T
-
~A(1 - T)"
(1 - T)
This says that for transmissions greater than 7CTc, the fractional change
in aerosol concentrations should be proportional to the fractional change
in opacity, (l - T). In the region of T between 10 and 70% a reasonable
approximation is
= 3 AT
From the above relationships, the variations in concentration with
time can be estimated from the variation in transmission with time shown
in Figure C-5. It will be noted that for the coarsest dust fraction, 5 -
10 M., which is relatively free flowing, the concentration variation dur-
ing a run is less than ±5%. For the finest and least free-flowing dust,
0 - 2.5 (J., the concentration variation is only some ±15%.
The traces in Figure C-5 indicate occasional spurts in concentration,
especially with the finest dust. These were demonstrated to result from tap-
ping or vibration of the dust feed line to the aerosol disperser; either
would momentarily dislodge a buildup of dust in this line. This could also
C-9
-------
be duduced from the fact that such a spurt is always in the direction of
increased concentration and the fact that a time of about 6 seconds
(twice the nominal aerosol chamber retention time) is required for this
perturbation in concentration essentially to disappear.
It may, therefore, be concluded that the aerosol concentration in
the tunnel remained uniform to better than ±15% throughout a run except
for an occasional short-lived increase.
Sources of Sampling Error
The following are possible sources of sampling error:
(l) Error in concentration of coarse particles because of failure
to achieve isokinetic sampling conditions. With sampling ve-
locities higher than those of the stream being sampled, a low
apparent concentration of coarse particles will be measured;
the reverse will be true if sampling velocities are too low.
The following is a summary of approximate nominal velocities
under the test conditions employed:
In sighting tunnel 4.4 ft/sec (133 cm/sec)
In plenum chamber 7.3 ft/sec (223 cm/sec)
Entering open "Delrin" filter 0.42 ft/sec (13 cm/sec)
Entering closed "Delrin" filter 2.8 ft/sec (86 cm/sec)
Entering "T" "arrangement 2.8 ft/sec (86 cm/sec)
Entering bag sampler nozzle 15 ft/sec (460 cm/sec)
The above statements assume that the sampling port points into the gas
stream. If the gas stream flow direction is other than directly into
the sampling port, there will be a tendency toward measuring low ap-
parent concentrations.
(2) Deposition in sampling lead lines. This was probably a major
factor in the "T" arrangement in addition to any sampling bias
and would result in the radically lower measured concentrations
observed with that arrangement.
C-10
-------
(3) Electrostatic repulsion at the probe intake. Particles sub-
jected to the violent impact and shearing in the compressed-
air disperser can acquire electrostatic charges, the sign of
which can be asymmetric. When such particles deposit on an
insulated wall, an electrostatic field is set up that will tend
to repel those particles bearing the predominant sign of charge.
Such electrostatic effects are likely to be more pronounced with
the coarser particles.
(4) Failure to retain particles on the filter medium. This can
happen because of electrostatic charges or because of mechani-
cal action. The "Nuclepore" medium can acquire very large
charges that could repel particles of similar charge and keep
them from landing. Mechanical vibration or flexure in handling
the filter papers could result in a loss of deposited particles.
Indeed, in a few cases it was observed that particles were pro-
pelled into space when the filter paper was removed from the
holder, although the magnitude of this action was not assessed
quantitatively. This behavior could be caused by either elec-
trostatic or vibrational effects. In either case, the action
would be expected to be most pronounced with the coarsest par-
ticles because of the much greater tendency of fine particles
to adhere to the surface. It was established that such losses
did not occur during weighing. A few filters were weighed by
successively removing them from and replacing them in their
storage containers between weighings. The loss in dust deposit
after several such manipulations was negligible (less than 5%
of total dust deposit) when dealing with the 5 - 10 u dust frac-
tion. Such weighings, however, were made a relatively long time
(hours) after sampling. Thus, strong adhesive forces may have
developed and electrostatic charges may have leaked off.
(5) Error in sample weighings. The precision of the weighing tech-
niques was checked on numerous occasions and weighings were
always reproduced within ±0.020 rag. Such an error would be less
than lOTc with all but the smallest filter deposits. Although
human error in weighing the clean filter papers could result,
such error should be very infrequent as judged by the reproduc-
ibility achieved in reweighing of dust-laden filter papers.
(6) Deposition of particles on the walls, roof, or floor of the
tunnel. Under the test conditions such wall deposition would
be most pronounced the coarser the particles. The tunnel and
dust conditions were selected to minimize such effects although
C-ll
-------
such effects could not be completely eliminated. As discussed
in Appendix A, measurements of wall deposits in the entire
tunnel after many sampling runs indicated that such deposition
was probably less than 5% of the dust feed. This conclusion,
however, is not incontrovertible because of the possibility of
dislodgement of wall deposits by vibration during periods be-
tween samplings. It was observed that what little dust deposit
occurred on the tunnel walls was almost entirely on the hori-
zontal surfaces. However, theoretical calculations confirm
that gravity settling could account for only a small amount of
deposition on the tunnel floor.
Discussion of Sampling Results
The following is a summary of the results indicated by the sampling
data summarized in Tables C-l to C-4:
(l) The concentration ratio (feeder/filter) is close to unity for
all filter arrangements and sampling locations when dealing
with the finest dust fraction, 0 - 2.5 p,.
(2) For the coarser dust fractions the concentration ratio becomes
progressively larger. In the case of filter arrangement "T"
the concentration ratio increases steadily with increasing
particle size and becomes quite large (3.6) for the coarsest
size. Although dust deposits were observed in the "T" arrange-
ment lead lines, their magnitudes could not be established quan-
titatively. Wi-th the other filter arrangements the trend is
the same, but there appears to be a major increase when going
from the finest (0 - 2.5 u) to the next finest (2.5 - 5 (j, dust
fraction.. The concentration ratio then increased only slightly
more for the coarsest fractions.
(3) Although there was some trend of concentration ratio with some
filter locations, the trends were somewhat erratic with the
"T" arrangement. In the other arrangements the concentration
ratio was more dependent on the sampling technique than on the
sampling location. The open "Delrin" filter collected more
dust (gave lower concentration ratio) with horizontal filter
medium (OH , OH ) than did the closed filter (CH , CH ). For
u d u d
the vertical medium facing into the stream, there was little
difference in concentration ratio between open and closed filter
(OS, CS).
C-12
-------
(4) Runs in which the filters contained a heavy deposit (86, 91-94)
tended to give a lower concentration ratio but the trend was
not consistent.
(5) Runs in which a double filter mat was used (87 - 90) with the
deposit between the mats gave a lower concentration ratio
(about 1.3 for the 5 - 10 p. fraction) than did most of the
other arrangements.
(6) With all dust fractions a significant number of samples indi-
cated a concentration ratio of close to unity for most sampling
arrangements, and sampling locations.
It is felt that the concentration determined by sampling is subject
to the greatest error and that actual concentrations are close to those
calculated from dust feed rates, although the actual concentrations must
be somewhat lower than those so calculated. The following tabulation
represents an "educated guess" on the relationship between actual and
calculated concentrations.
Concentration Ratio
Dust Fractions (Feeder/Actual)
0 - 2.5 p. 1.00
2.5 - 2.5 U 1.05
5 - 10 U 1.10
0 - 10 u 1.10
These values cannot be derived rigorously but are based on a considered
judgment of all the above considerations.
Although particle inertia undoubtedly accounts for some of the error
in the sampling measurements, it is felt that a major factor is associated
with loss of dust from the filter paper due to either electrostatic effects
during deposition or dust loss while removing the filter paper from the
holder. The fact that the particle was reasonably spherical could be a
significant factor in such accidental loss from the filter medium. This
C-13
-------
conjecture has not been proved but is suggested by the poor reproduc-
ibility of the sampling measurements. A major consideration in this
conjecture is the fact that the transmissometer readings showed excel-
lent agreement with feeder speed settings and did not correspond to the
wide variations in apparent concentration indicated by the sampling mea-
surements for a given feeder speed setting.
Sampling with Weathermeasure Instrument
A few high-spot evaluations of the dust concentrations in the tunnel
system were made by sampling through an instrument produced by Weather-
measure Corporation (Division of KMS Industries), P.O. Box 41257, Sacra-
mento, California 95841. This instrument measures light scattered at
90 from an incident light beam. The light detector current is allowed
to accumulate in a condenser that pulses whenever a given charge accumula-
tion is reached. The count rate, which is proportional to aerosol con-
centration, is obtained by separately timing the number of pulses. Each
such discharge is registered as a count. The instrument will also read
count rate directly for rates over 100 counts/min.
In all these runs air was sampled by means of the "l" arrangement
shown in Figure C-l. However, the Weathermeasure instrument was inter-
posed between the sampling line and the filter. All the measurements
were made in the course of the sampling series recorded in Table C-l.
The data so obtained are summarized in Table C-5.
The Weathermeasure instrument requires calibration for each dust;
hence, the results can be considered only relative. The instrument has
been calibrated with 0.32-^1 stearic acid aerosol by the manufacturer;
for this aerosol 1 count/min is reported to correspond to a particle
3
concentration of 0.01 mg/m . From the data of Table C-5, 1 count/min
for 0 - 10 (I fly ash would appear to correspond more nearly to a dust
C-14
-------
3
concentration of 1 mg/m . With the 2.5 - 5 |i fly ash, 1 count/min appears
3
to correspond to about 0.5 mg/m . Since these results are presumably also
subject to the sampling errors previously discussed for sampling with the
"T" arrangement, the actual instrument equivalents are probably about 1/2
. 3
those cited above (i.e., 0.5 and 0.25 mg/m for the 0 - 10 p and the
2.5 - 5 n fractions, respectively).
In general, resonable reproducibility was obtained although there
was some problem with a drift in the instrument calibration setting. The
manner in which the instrument was held also seemed to affect results.
It was held horizontally in all the runs reported in Table C-5. The trend
with sample location corresponds reasonably well to those observed in the
filter sampling runs with the "T" sampling arrangement.
C-15
-------
TUNNEL WALL.
1" HOLE.
1/4" i.d. TYGON TUBE
AEROSOL IN
AIR OUT
POROUS METAL, PALL CORP
GRADE F POROUS
STAINLESS STEEL,
1/16" THICK
/RUBBER STOPPER
COPPER TUBE
3/8" o.d., 5/16" i.d.
1/4" i.d. x 5/16" o.d.
FRICTION FIT
1/16" 'O' RING
3/32" 'O' RING
1/8"
I
SHT., ,rj$ys:i/i6"T i 1-1/2-
>
QUICK OPENING LATCHES
.TO CLAMP TOP AND BOTTOM
PIECES TOGETHER
BOTTOM VIEW OF FILTER
TA-8290-14R
FIGURE C-1 SAMPLING AND FILTER ARRANGEMENT
C-16
-------
4
4
\
\
"---
^.^
O"~-
-, ~-
;<;
>^H«
f(\V
?-:\
1 T/ft"
- 7/8" *~V>*.
I I'*.
^
^ v
;
i
1/8
i
1 1
3/4"
1/4"
_L
9/16"
3/8"
3/16"
-9/16-
- 7/8"-
'
3/32"
(a) OPEN GELMAN FILTER HOLDER (DELREN)
40 MESH SCREEN
(b) CLOSED GELMAN FILTER HOLDER (DELREN)
TA-8730-22
FIGURE C-2 DELREN FILTER HOLDER
C-17
-------
O
K-1
00
1" i.d.
1 1/8" o.d. TUBE
\
TO
. VACUUM x RUBBER STOPPER
CLEANER
12 1/2"-
PAPER VACUUM CLEANER BAG
FOR HOOVER MODEL 2900
PART No. C-15631
PACKAGE No. 16446
(Approximately 1.0 sq. ft. Filter Area)
3 1/2" o.d. 16 ga Al TUBE
PRESSURE
1/8" HALF COUPLING
RUBBER STOPPER
.4" ^
-1/2"
TA-8730-26
FIGURE C-3 BAG SAMPLER SYSTEM
-------
n
i
10"
18"
- 2'-
4'-
\
3A
38
3AA
30" »-
5A
12"
4"
10"
ELEVATION
PLAN VIEW
FIGURE C-4 SAMPLING LOCATIONS
IT
J_
r
4"
1 3A2 22 |
1A t ^ t f
S3Al "Cl2l
2 \
V
1
/ ,2"
"P
TA-8730-25
-------
0 - 2.5-
2.5 - 5.0L
5.0 - 1CU
0 - 10L
Nominal Transmission, percent
DUST
FRACTION
0-2. 5t
2.5-5M
5-1 DM
0-lOfi
CHART
NUMBER
1
2
3
4
5
6
7
8
9
10
1 1
12
BUST COf/CENmriON
(Based en Feeder)
jig m^
16
48
146
27
74
222
32
73
206
35
84
241
SAMPLING RUN
NUMBER
75
51
48
79
34
33
91
100
101
65
27
29
JVEHAGE TBJNSHISSOXETEfl
READING CORING RON
85
66
29
84
74
42
95
S3
67
91
76
43
FIGURE C-5 TYPICAL TRANSMISSOMETER STRIP CHART RECORDS OF VARIOUS DUST
SIZES AND CONCENTRATIONS
C-20
-------
Table C-l
SUMMARY OP INITIAL SERIES OF SAMPLING RUNS*
Run Number
Date 1/25/71 1/2G/71 1/2G/71 1/2C
Equipment
Arrangement number A A A A
Pressure differentials, mm
xylenc
Across nozzle 00 92 92 92
Across plenum (2^G) /19 .19
Atmospheric condition
Temp, '' C
Barometer, mm II;; 770 772 772 772
Oust type (fly ash) 0-10[i 0-lOu 0-10,1 0-W O-lOjo,
Feeder
G roove .s i ze
^P Va r i at: setting
to RPM
h-1
Filter
Sampler type T T T T
Initial filter wei|;lil, mi; 5.707 5.591 5.7H7 5.80S
Dual collected, nn; 0.2:19 0.205 0.100 0.117
Time, see 600 (Hid 000 600
Sampl in;; position 2266
Samp 1 i nj; rai e , ,0,/m i n
Aerosol eoncent ra 1 i on, m;1;,'m'
From ti-ed rate
I''rom filler samp 1 e
Rat 10: feeder/I' i HIM'
Average Iransmissometer
read i n;;
5
1/26/71
A
92
49
772
6
1/26/71
A
92
49
772
7
1/26/71
A
92
49
1H.5
770
8
1/2C/71
A
94
51
19.0
770
9
2/4/71
n
94
46
M.5
772
10
2/4/71
li
94
46
15.0
772
-------
Table C-l (Continued)
O
I
to
Run Number
Date
Equipment
Arrangement number
Pressure differentials, mm
xylene
Across nozzle
Across plenum (2-6)
Atmospheric condition
Temp, °C
Barometer, mm Ilg
Dust type (fly ash)
Feeder
Groove size
Variac setting
RPM
Filter
Sampler type
Initial filter weight, rng
Dust collected, mg
Time, sec
Sampling position
Sampling rate, 4/min
Aerosol concent rat ion, mg/m^
From feed rate
From filter sample
Ratio: feeder/filter
Average transmissomcter
reading
Notes
11
2/4/71
B
94
46
15.5
772
0-10(1
1/16 x 1/8
100
3.27
T
5.887
0.081
GOO
3A
2.55
32.6
3.2
9.82
0.008
3
12
2/4/71
D
94
46
16.0
772
O-10H
1/16 x 1/8
100
3.21
T
5.586
0.261
600
3A
2.55
32.0
10.2
3.14
0.905
13
2/4/71
B
94
46
15.2
772
0-10(1
1/16 X 1/8
100
3.27
T
5.626
0.195
GOO
2
2.55
32.6
7.65
4.26
0.902
14
2/4/71
B
94
46
15.0
772
0-10(1
1/36 X 1/8
100
3.21
T
5.836
0.383
600
8
2.55
32.3
15.0
2.15
0.906
15
2/4/71
B
94
46
15.2
772
0-10(1
1/16 X 1/8
100
3.27
T
6.294
0.280
600
3
2.55
32. C
11.0
2.96
0.906
1
16
2/4/71
B
94
46
14.8
772
0-10(1
1/16 X 1/8
100
3.24
T
6.070
0.276
600
5
2.55
32.3
10.6
3.05
0.911
17
2/4/71
B
94
46
14.2
772
0-10(1
1/16 x 1/8
100
3.2-1
T
6.314
0.2G7
600
6
2.55
32.3
10.5
3.08
0,912
IS
2/4/71
B
94
46
14.0
772
0-10(1
1/16 x 1/8
100
3.21
T
6.255
0.221
600
6
2.55
32.3
8.7
3.71
0.912
19
2/4/71
B
94
46
13.5
772
0-10|l
1/16 x 1/8
100
3.27
T
6.157
0.323
600
5
2.55
32.6
12.7
2.57
0.914
20
2/4/71
B
94
46
13.0
772
0-10(1
1/16 xl/8
100
3.24
T
5.451
0.315
600
3
2.55
32.6
12.3
2.65
0.909
-------
Table C-l (Continued)
0
1
to
w
Run Number
Date
Equipment
Arrangement number
Pressure differentials, mm
xylcnc
Across nozzle
Aeross plenum (2-G)
Atmospheric condition
Temp , ° C
Barometer, mm IU;
Dust typo (fly ash)
Feeder
Groove Si7.o
Vari ac sot t intf
RPM
Filter
Sampler type
Initial filter weif.ht, mi,e.
Dust colli'd eel , mi;
Time, sec
Sainpl i UK pos i t i on
Sampling rate, ^/iniu
Aerosol concent rat ion , mj.;/m-5
From feed rate
F rom i i 1 1 er samp 1 e
llat io: feeder/r i 1 1 er
21
2/5/71
I!
04
-10
10.0
70S
0-lOu,
1/10 x 1/8
20
0. 292
T
0.335
0. 171
3G02
:;A
2.55
2. '.11
1.11
2 . 55
22
2/5/71
1!
0-1
'10
17.0
7GB
0- 1 On
1/1G x 1/8
20
T
5. -1-12
o. Ir>o
:)Goo
3A
2.55
2. ill
0.9S
2. 97
23
2/5/71
1!
0-1
4G
1 5 . 0
7(iH
n-lOu,
1/16 x 1/8
1C)
1 . 07
1'
G.2:!1
0.240
1KOO
3 A
2 . 5 5
1 0 . 7
3. 1 1
3.40
24
2/5/71
B
01
40
11 .8
7GH
0-1 0|i
1/1G x 1/8
40
T
G.:n4
o.2:iK
1800
J,\
2.55
10.7
3.11
3.4 I
25
2/11/71
li
93
40
20.5
771
0-1 0|j,
1/8 x 1/2
20
0.321
T
(i.lHl
0.01G
830
3A
2.r.r>
21.2
0. 15
53.H
20
2/11/71
13
93
40
25.5
771
0-1 0|J,
1/8 x 1/2
20
0.332
T
G.284
0.217
liOO
3A
2.55
21.8
8 . 5
2 . 92
27
2/11/71
I)
93
10
') r >
771
0-10(1
1/8 x 1/2
40
]. 15
T
G.025
0. 751
GOO
3A
2 . 55
X 1.0
20.4
L'.Kl
28
2/11/71
D
93
4G
21.8
771
0-lOp.
1/8 x 1/2
40
1.15
T
li.190
O.G58
GOO
3A
2.55
81.0
25.8
3.2G
29
2/11/71
li
93
40
25.0
771
()-10|i
1/H x 1/2
100
3 . 33
T
G.311
1 .GOb
GOO
3A
2.55
241.
G3.0
3. 82
30
2/11/71
U
93
10
24.8
771
0-lOu.
1/8 x 1/2
100
3.37
T
G.334
1 . 537
GOO
3 A
2.55
214.
GO. 2
1 .05
O.ilGG 0.9M9 0.918 O.921 0.7GI 0.770 0.182
-------
Table C-l (Continued)
n
I
to
Run Number
Date
Equipment
Arrangement number
Pressure differentials, mm
xylene
Across Nozzle
Across plenum (2-G)
Atmospheric condition
Temp, °C
Barometer, mm Hg
Dust type (fly ash)
Feeder
Groove size
Variac setting
HPM
Filter
Sampler type
Initial filter weight, me
Dust collected, mg
Time , sec
Sampling position
Sampling rate
Aerosol concentration, mg/m^
From food rate
From filter sample
Ratio: feeder/filter
Average transmissometcr
read ing
31
2/18/71
B
93
46
15.8
770
2.5-5(1
1/8 x 1/2
40
1.05
T
6.200
1.302
600
8
2.55
73.5
51.0
1.44
0.727
32
2/18/71
B
93
46
15.8
770
2.5-5(1
1/8 x 1/2
40
1.05
T
6.176
1.301
eoo
8
2.55
73.5
51.0
1.44
0.734
33
2/18/71
B
93
46
15.4
770
2.5-5|l
1/8 x 1/2
40
1.05
T
5.969
0.731
6OO
2
2.55
73.5
28.6
2.57
0.740
34
2/18/71
B
93
46
15.4
770
2.5-5(1
1/8 x 1/2
40
1.05
T
6.171
0.770
000
2
2.55
73.5
30.2
2.43
0.738
35
2/18/71
B
93
46
16.2
770
2.5-5(1
1/8 x 1/2
40
1.07
T
6.195
0.843
600
3A
2.55
74.9
33.0
2.27
0.734
36
2/18/71
B
93
46
16.2
770
2.5-5|l
1/8 x 1/2
40
1.08
T
6.164
0.996
600
3A
2.55
75.6
39.0
1.94
0.737
37
2/18/71
B
93
46
16.0
770
2.5-5(1
1/8 x 1/2
20
0.297
T
6.130
0.281
600
3A
2.55
21.0
11.0
1.91
0.917
38
2/18/71
B
93
46
15.7
770
2.5-5(1
1/8 x 1/2
20
0.278
T
6.222
0.287
600
3A
2.55
39.6
11.3
1.74
0.920
39
2/18/71
B
93
46
15.2
770
2.5-5(1
1/8 X 1/2
100
3.20
T
5.980
2.733
600
3A
2.55
222.
107
2.08
0.417
40
2/18/71
B
93
40
15.0
770
2.5-5^
1/8 x 1/2
100
3.23
T
6.167
1.808
600
3A
2.55
224 .
74.4
3.01
0.387
Notes
-------
Table c-1 (Concluded)
o
1
to
Ol
Run Number
Date
Kqu ipment
Arrangement number
Pressure, differential mm
xylene
Across nozzle
Across Plenum (2-6)
Atmospheric condition
Temp , ° C
Barometer, mm 1[^
Dust t ype ( fly ash)
Feeder
G roove s i ze
Vari ac sett 'nix
KI'M
Filter
Sampl el1 t ype
Initial filter weip,lit, mr;
Dust collected, mr;
T i me , .sec
Samp 1 i n)', pos i t i oa
Sampling rale, £/min
Aerosol conren 1 ra ! i on, i;i;;-/!a:}
From food rate
F rum filter samp 1 e
Rat 10: feoiler/l i 1 1 or
A ve ra[;e t ransmi ssor.u 1 e i'
road i IH:
No! os
41
2/25/71
B
93
4G
9.8
773
0-2.5(1
1/8 x 1/2
20
0. 255
T
G.085
0.292
GOO
3 A
2 . 5 5
11.8
11.1
1 .03
0.902
12
2/25/71
B
93
4G
10.8
773
0-2.5',!
1/8 x 1/2
20
0.2G3
T
G. 153
0.301
GOO
3A
2 . 55
12.2
11.8
1.03
0.883
43 44
2/25/71 2/25/71
B B
93 93
46 46
11.4 12.0
773 773
0-2.5(1 0-2.511
1/8x1/2 1/8x1/2
40 100
0.983 2.92
T T
G.115 6.180
0.927 3.029
600 GOO
3A 3A
2.55 2.55
45.5 135.2
3G.3 118.8
1.25 1.11
0.652 0.311
45 40 17 18 49 51 51
2/25/71 2/25/71 2/25/71 2/25/71 2/20/71 2/25/71 2/25/71
n B B B B B B
93 93 93 93 93 93 93
46 46 4G 4G 46 4G 46
12.5 13.0 13.3 12.8 12.3 12.6
773 773 773 773 773 773 773
0-2.511 0-2.511 0-2.5(1 0-2.5(1 0-2.5(1 O-2.5ji 0-2.5(1
1/8 x 1/2 1/8 x 1/2 1/8 x 1/2 1/8 x 1/2 1/8 x 1/2 1/8 x 1/2 1/8 x 1/2
100 40 40 100 10 -10 40
2.92 1.05 1.02 3.16 1.04 1.03 1.03
T T T T T T T
G.lll 6.213 6.087 6.179 6.161 0.201 G.22G
2.712 1.210 1.001 2.877 1.008 0.785 0.780
600 liOO (500 GOO GOO 605 GOO
3A 8 3A 3A 8 2 2
2.55 2.55 2.55 2.55 2.55 2.55 2.55
133.2 18. 5 17.2 11G.3 18.2 17.7 17.7
10G.3 18. G 39.1 112.8 39. G 30.5 30. G
1.27 1.00 1.20 1.30 1.22 1 . 5G 1.56
0.315 0.586 O.G53 0.291 O.G31 o.ii:')] O.GG1
Not.(.>H : 1. 1'i 11 cr \ips i (It- (in\ui i n pl;u;t i r box be Con1 \vi> i r,-h i nj;.
2. Some rlust ;uJlK>rini; to in.si/Jt.- 1 i(J of nili.-r holder box, Hriu-e dust collected rni;ht be recorded as tno low.
.'t. Under m i crosi-opr 1 h i s filter eont a ined nboul 1 /! ;u; much null or i :il as filter from Hun 1^ by v i r-uia 1 rst iniat e .
t. I'eed rat o est iniat ed , not measured .
-------
Tnble C-2
SUMMARY OK SECOND SERIES OF SAMPLING RUNS*
Run Number
Date
Equipment
Arrnngement number
Pressure differentials, mm
xylene
Across nozzle
Across plenum*"" '
Atmospheric conditions
Temp, °C
Barometer, mmllg
Dust type (fly ash)
Feeder
Groove size
_ Variac setting
1 HPM
to
C* Filter
Sampler type
Initial filter weight, nig
Dust collected, nig
Time, sec
Sampling position
Sampling rate, £/min
Aerosol concentration mg/m
From feed rate
From filter sample
Ratio: feeder/filter
Average transmissometor
reading
Notes
52A 52B
8/31/71
B
89
41
25.2
--
0-lOu
1/10 x 1/8
100
3.31
T OS
6.315 0.400
0.155 0.525
707 707
2 2 1
2.55 2.55
33 . 3
5.10 17.5
6.45 1.90
0.913
53 A 53B
9/2/71
B
89
13
27.8
702
0-lOp.
1/10 x 1/8
100
3.28
OF OS
0.235 G.OG8
1 .051 0.838
832 832
2 2 1
2.55 2.55
32.9
29.7 23.7
1.11 1.39
0.908
54 A 54 B
9/2/71
B
89
43
27.8
702
0-lOu
1/10 x 1/8
100
3.33
OF OS
0.891 0.877
0.691 0.710
GO 3 003
2 2,
1
2.55 2.55
33 .2
27.0 28.9
1.23 1.15
0 . 903
55A 551!
9/3/71
B
88
13
28.5
701
0-10u
1/16 x 1/8
100
3.10
OF OS
0.215 G.095
0.158 0.3G5
059 059
3 A 3Aj
2.55 2.55
31 .0
1 0 . 4 13.0
2.07 2.02
0.905
56 A 56 B
9/3/71
B
87
42
761
0-10u
1/10 x 1/8
100
3.44
OF OS
6. 020 G.200
0.080 0.505
016 016
3 A 3A2
2.55 2.55
34.3
25.9 21.6
1.32 1.59
0 . 900
57A 57B
9/3/71
B
80
41
3 1 . 9
761
0-lOu.
1/10 x 1/8
100
3.43
OF OS
6.203 0.099
1.058 0.520
017 017
2 22
2.55 2.55
34 . 2
40.3 20.0
0.85 1.71
0 . 900
58A 58B
9/9/71
B
91
45
21.1
7G2
0-lOp,
1/16 x 1/8
100
3.33
OF OS
C.I 66 6.000
0.010 0.170
600 000
2 2y
2.55 2.55
33.2
25.1 18.1
1.32 1.80
0.905
59A flnil 60A GOB
9/9/71 9/9/71
B B
91 90
1 5 1 5
22.2 23.2
762 702
0-lOu, 0-lOu,
1/10 x 1/8 1/16 x 1/8
100 100
3.33 3.33
OF OS OF OF
0.135 0.129 G.9CO G.2.14
0.895 0.199 O.OG3 0.221
600 GOO GOO 600
2 2C 8 7
2.55 2.55 2.55 2.55
33.2 33.2
3 5.1 19. 5 2.5 8.8
0.95 1.70 13.3 3.77
0.908 0.905
1
* Common conditions: tunnel air flow 5800 CFM including compressed air; compressed air: 50 psi.
T = copper tube with 90° bend
0 = open "Delrin"
Hu = filter horizontal facing up
S = filter facing into stream
F = filter flush with wall
C - closed "Delrin"
11^ = filter horizontal facing down
V = vacuum cleaner bag sampler.
-------
n
Table C-2 (Continued)
Run Number 61A 6in 62A 621? G.'iA G3li G4 G5A G5I1 (iGA G6H 67A 67I1 (iSA 6HJ1 G9A G9W
Date
Kcju i pmen t
Arrangement number
Pressure differentials, min
xylene
Across noxzle
Across plenum^""*' '
Atmospher ic concl i t. ions
Temp, C
Haromcl er , mmlli;
Dust typo (fly as]))
Feeder
Groove' s i 7,e
Var i ac .sot t i nf^
HPM
Fi 1 tor
Sampler type**
Initial I'i 1 ter \voiu lit . niu'
9/9/71 9/9/71
1) 1!
89 89
1 3 1 3
28.1
702 702
o-iou o-jou
I/Hi x 1/K I/Hi x 1/H
100 100
3 .33 3.11
OK ())! OF Ol,,
(i. 109 7(i] 7(il 701
0-lOn O-10(J, 0-lOu O-IOU 0-JOu O-K'U
I/Hi x 1/8 I/Hi x 1/8 1/10 x 1/8 1/10 x 1/8 I/Hi x 1/K I/Hi x 1/K
100 100 100 10O 100 IOO
3.11 3.T)7 3.T>1 'i.37 3.37 3.37
0,S OH,, ("11,, OH,, CHU OH,, HI,, OH,, <'!lu Oil,, Til,,
G.2M ri.riio r>.7iiH 5.009 ri.nni r..2in <;.o79 0.2. 1:1 (;.I33 fi 775 r>.70<;
Dust eolleeted, tn^ O.17T) O.(i91 O.19K O . ;>:r/ O . :!2O 1I.3H1 O.O11 O.fiiO O.fiKi O . Ti 1 8 0.112 O.3S7 O.:':V O.171 O.IS(i O.liMd (l..r)39
T i me, sec (iOO (iOO 000 (iOO (ion (>oo Hi 730 73(i (ioo lion (ion (id!) (ioo
Sampling position 3 A 3A, 3A .'iA, 7 8 3A( 3 \c 3A(. 3A(. 3A(. '>c
Aei-osol eoneen I r a t i on ms1,/nr
From feed rate 33.2 31.0 31.0 31.0 3."i.0 3:",.2 33.(i 33.0 33.0
From filter sample 18.(i 27.1 19 . li 20.7 K.ii IT..(I 2H..'i 1 7 . r, 2O.3 10.2 1 '.'< . 2 9.9 1 S , r, 1 ) . I 27.3 21.1
liatio: feeder Ti Her 1.79 1.'.!:! 1.71 I.01 3.91, 2.20 -- 1.71 2.01 1.7'! 2.1H 2.21 3.39 1 . HL! 1.70 1.23 l.:',9
Not r
-------
Table C-2 (Continued)
O
to
00
Run Number
Date
Equipment
Arrangement number
Pressure differentials, mm
xylene
Across nozzle
Across plcnum(2-6)
Atmospheric conditions
Temp, °C
Barometer, mmllK
Dust typo (fly ash)
Feeder
Groove size
Variac setting
HPM
Filter
Sampler type**
Initial filter weight, mg
Dust collected, m(?
Time, sec
Sampling position
Snmplinp rate, ,^/min
Aerosol concentration mg/m
From feed rate
From filter sample
Ilatio: feeder/filter
Average transmissometer
70A 70B
9/20/71
D
87
43
23.0
763
5-lOp,
1/16 x 1/8
100
3.26
°"u C1IU
6.170 5.708
0.438 0.413
600
3AC
31.3
17.2 16.2
1.82 1.93
0.955
71A 71B
9/20/71
n
87
43
23.0
763
5-lOp,
1/16 x 1/8
100
3.26
5.560 0.1 «3
0.492 0.32G
600
3 Ac
31.3
19.3 12.8
1.G2 2.44
0.945
72A 72D
9/20/71
n
87
43
23.0
7G3
G-lOp,
1/16 x 1/8
100
3.26
°»u C"u
5 . 70 1 6 . 070
0.761 0.299
600
2c
31.3
29.9 11.7
1.05 2.08
0.950
73A 73B
9/20/71
B
86
44
22 5
763
G-lOp.
1/16 x 1/8
100
3.27
°»u C"u
5.567 5.953
0.705 0.345
600
2
c
31.4
27.6 13.9
1.14 2.20
0.945
74A 74B
9/27/71
B
90
44
21.6
7GG
0-2. GH
1/16 x 1/8
100
3.22
Oil CII
5.930 C.045
0.425 0.413
596
3Ac
10. 7
16.8 16.3
1,00 1.02
0.863
75A 75B
9/27/71
B
21.9
766
0-2. Dp,
1/1G x 1/8
100
.3.17
OH CII
u u
5.539 0.0 16
0.532 0.449
638
o
16.4
19.6 16.5
0.84 1.00
0.860
76A 7611
9/27/71
n
89
44
21.9
76G
0-2. 5u,
1/16 x 1/8
100
3 . 27
0.094 5.509
0.4G2 0.436
653
:IAc
16.9
16.6 15.7
1.02 1.08
0.863
77A 77B
9/27/71
n
21.9
766
0-2. 5p,
1/16 x 1/8
100
3 . 23
OH CII
u u
5.951 5.678
0.494 0.451
611
2C
16.7
19.0 17.3
0.88 0.97
0.855
7SA 78B
9/28/71
B
89
44
18.8
764
2.5-5U
1/16 x 1/8
100
,"! . 20
Oil CII,.
u »
0.124 5.058
0.400 0.440
638
:!Ac
27.2
17.2 16.2
1.58 1.G8
0.840
reading
Notes
-------
Table C-2 (Continued)
0
1
to
to
Hun Number
Date
Kqu ipment
Arrangement number
Pressure d 1 1' ferent ials , mm
xylene
Across nozzle
Aeross plenum '
Almospher ic coiuli t ions
Temp, °C
Barometer, mmlln
Dust typo (rly ash)
Feeder
(Iroove s i ye
Var ine set t in^
11PM
F i U e r
Sampler type**
Initial f i Her \iein lit. , mi;
Oust ro 1 lee t ed, in);
'I'ime, see
Samp! i n;p, pns i t ion
Sampling; I ' a t f ' , f / Iti i 1 1
Aerosol ronrentral ion nir,/'"'
From t'eeil rale
From niter sample
Hat i<>: IVeder/l i 1 t < r
79A 7011 80A 8015
9/28/71 9/28/71
n B
89 89
'1 5 4 f>
18.:!
7G4 7(11
2.5-510, 2.S-5U.
I/Hi x 1/K 1/1G x 1/8
100 10(1
.'!.!!) 3.28
on en on en
u u " »
5.750 5.(i21 (i.051 5.711
0.5:i(i 0.4G1 0.447 0.423
(115 (100
:'-,. :IA(.
27.2 27.!)
19.5 1(1. '1 17.5 1(1.11
1 .:!'.! 1 .(11 1 .59 1 .(18
81A SIB
9/29/71
B
89
43
22.5
7(i2
5-lOp.
1/10 x 1/8
100
:i . 21
Oil C\\
U I!
5. (11 2 5.5(19
0.153 0.:iGl
(100
3A
c
:i2. i
17.7 11 . 1
1.81 2 .28
82A 82D 83A 83B 84A 8111
9/29/71 9/29/71 9/29/71
11 B H
89
44
23. 0
5-lOu 5-lOp. 5-lOu
1/1G x 1/8 1/1G x 1/8 1/1G x 1/8
100 100 100
3.25 3.28 3.2(1
Oil HI Oil Cll Oil Cll
11 U (1 (1 ll tl
5.G23 5.1115 G.OR2 5.1185 5. (151 5.58G
0.173 0.315 (1.521 O.322 0.112 0.211
(127 (100 (115
3A _ 3A(. 3 A
32.2 32.5 32.3
17.7 11.X 20.5 12.11 1(1.9 9.3
1 .82 2 .73 1 .5S 2 .5H 1.91 .'! . 1 7
85A 8511 8BA 8611 87A 87B
9/29/71 9/29/71 10/8/71
11 B B
89 00
11 43
21.0 20.5 21.4
7GT>
5-lOp, S-ldu f)-I()n
1 /Ifi x 1/8 1 /Id x 1 '8 1 'Hi x 1 '8
100 100 100
3.30 3.30 3.25
OS CS OS CM OS OS
5.974 5. (115 (1.037 5,925 5.73(1 12.913
(1.884 0.899 1.190 1.2(10 O.(i59 0.519
1 200 12OO (i 1 L'
3A 3A 3A
'' rr . '' .,-''.
32.7 32.7 32.3
17.3 1 '! .11 23 .3 21.7 35.3 21.1
1 . H9 1 .8(1 1 . 1(1 1 .32 1 .28 1 , 53
O.K75
-------
Table C-2 (Continued)
Run Number
03
a
Equipment
Arrangement number
Pressure d ifferentials, mm
xylcne
Across nozzle
Across plenum'^-6)
Atmospheric conditions
Temp, °C
Barometer, mmHg
Dust type (fly ash)
Feeder
Groove size
Variac setting
RPM
Filter
Sampler type
Initial filter weight, mg
Dust collected, mg
Time, sec
Sampling position
Sampling rate, J>/min
Aerosol concentration mg/m
From feed rate
From filter sample
Ratio: feeder/filter
Average transmissometer
reading
Notes
10/8/71
13
90
43
23.3
765
5-10|J,
1/16 x 1/8
100
3.31
OS OS
5.936 12.282
0.608 0.648
606
3AC
2.55 2.55
32.8
23.6 25.1
1.39 1.31
0.946
5
10/8/71
n
90
43
21.9
765
5-lOp,
1/16 x 1/8
100
3.27
6.029 17.813
0.587 0.626
600
3Ac
2 . 55 2 . 55
32.4
23.0 24.5
1.41 1.32
0.948
5
10/8/71
n
90
43
21.2
765
5-10,
1/16 x 1/8
100
3.27
5.558 12.506
0.597 0.829
600
3Ac
2.55 2.55
32.4
23.4 32.4
1.38 1.00
0.948
5
10/12/71
D
91
44
18.0
763
5-lOp,
1/16 x 1/8
1OO
3.23
OS OH
5.724 5.960
1.891 1.360
1800
3AC
2.55 2.55
32.0
24.7 17.8
1.30 1.80
0 . 94 9
10/12/71
n
91
44
20.9
763
5-10,
1/16 x 1/8
1OO
3.27
OS OI1U
5.908 5.914
2.692 1.557
1800
3 A
2.55 2.55
32.4
35.2 20.3
0.92 1.60
0.942
10/12/71
n
87
42
32.9
704
2.5-5y,
1/16 x 1/8
100
3.37
OS Oil
5.924 5.593
2.021 1.868
1800
3AC
2.55 2.55
28.7
26.4 24.4
1.09 1.17
0.860
10/12/71
D
88
42
32.7
764
2.D-5U,
1/16 x 1/8
100
3.43
OS OIIU
6.105 5.546
1.280 1.639
1800
3AC
2.55 2.55
29.2
16.7 21.4
1.75 1.36
0.850
10/18/71
n
90
44
18.8
764
2.5-Su,
1/16 x 1/8
]00
3.17
V
13710
240
1800
3AC
490
27.0
16.3
1.66
0.90
6,7
10/18/71
n
90
44
18.8
764
2.5-Su,
1/16 x 1/8
100
3 .23
V
13680
270
1800
3AC
613
27.5
14 .7
1.87
0.88
6
-------
Table C-2 (Concludes)
o
Hun Number
97
Date 10/19/71
F.qu i pmcn t
Arrangement number li
Pressure d i f ferent in IK , mm
xylcno
Across nozzle
Across plenum (2-(>)
Atmospheric conditions
Temp, "t:
Barometer, nmillt;-
Dust type (fly ash)
Feeder
Ciroove s i /e
Var iae se 1 1 i n^
ItPM
91
M
18.8
70 n
5-lOu,
I/Hi x 1/8
100
3.30
98
10/19/71
H
91
-11
21.0
7Gn
fj-lOn.
1/1G x 1/8
100
3 . 33
99
10/26/71
13
92
1-1
1-1.0
703
fl-lOp,
1/8 x 1/2
20
0.227
101
10/26/71 10/26/71 10/26/71
7G3
5-lOu,
7G3
5-lOn.
100
3 .08
Fi1ter
Sampler type
Initial r i 1 ler v.'oit;ht., mp,
Dus t eo 1 lee ted, mp;
Time, see
Sampl i nf;- pos i t ion
Samp L i n^ rate, 1/m in
Aerosol eoncenlrat ion IIIK/ni'
From teed rate
From f i 1 t or san.p 1 e
Hat io: (Yodel- Ti I tor
imotor
13790
320
18(10
3-V
107
18.1
().S7fi
1. Filter mount.ed over side of tunm-l (not over a hole).
2 , Dummy run in wh i rh filler was moved into t unne t faring st ream with no air f I ov. through filter. 1
3, Manomet er across filler 1)1 ew out when run was 1 /2 to li/.'i over.
1. Major drift exper i enced in t ransmi ssion met or 7ero .
r>. In the "ll" runs , a spec i a 1 composi te filter was used : 1) Lop of Nuclepore cont a i ni nsc four 1 /H" 1
Nuclepore with mat te s i de up ; 3} the two spot Allied with Duco cement and separat ed at cent er will
G , Sain pi e t a ken with vacuum c 1 eaner with 1 rounded no/./ ! c po i nt ed into air st ream 1 ead i n£ d i reel 1 y
( appro x. 1 sq . f t. . area) . Pressure d rop across inlet no// 1 e used to me t e r air f 1 ow rat e .
7. Feed rate dropped erratically during last half of run, apparently due to brid^in..- in feeder hopp<
nles V, Mil
small pi i
to paper A
sh i ny side up
re of Nurlepo
aeuiun el eane r
-------
Table C-3
SUMMARY OF CONCENTRATION RATIO DATA FOR SAMPLING AND FILTER ARRANGEMENT "l"
Concentration 1
Feeder Conditions At Locatic
Dust
Fraction
0-2.5 p
2.5-5 p,
0-10
Groove
Size
1/8 x 1/2
1/8 x 1/2
1/16 x 1/8
1/8 x 1/2
Speed
Control
Setting
20
40
100
20
40
100
20
40
100
20
40
100
2
1.56
1.56
2.57
2.43
6.45
3.35
3.94
4.38
4.26
3
2.18
2.09
2.96
2.65
3A
1.03
1.03
1.25
1.20
1.14
1.27
1.30
1.91
1.74
2.27
1.94
2.08
3.01
2.55
2.97
3.40
3.44
9.82
3.14
53.8
2.92
2.84
3.26
3.82
4.05
/ Feeder \
>n Number
5
3.43
2.91
3.05
2.57
6
5.00
5.53
3.08
3.71
Sampling Run
Number
8
1.00
1.22
1.44
1.44
2.35
2.15
41
42
50, 43, 46
51, 47, 49
44
45
48
37
38
33, 35, 31
34, 36, 32
39
40
52A, 21
22
23
24
1, 5, 11, 7, 3, 9
2, 6, 12, 8, 4, 14
10, 15, 16, 17
13, 20, 19, 18
25
26
27
28
29
30
See Figure C-4 for position corresponding to each location number.
C-32
-------
Table C-l
SU-'SIARY OF CONCENTRATION RATIO DATA FOR SECOND SERIES OF SA.VPLI.NC RL'.NS
Condit ions Cormon to all RunsGroove Size: 1, 16 x 1/8
Speed Control Setting: 100
uoncentra
Sampling
Dust Location For Sampler
Fraction Number OF CF OS
0-2.5- 2
c
3A
c
2.5-5^. 2
c
3A 1.09
c
1.75
5-10- 2
c
3A 1.89
c
1.40
1.28
1.53
1.39
1.31
1.30
0.92
0-10- 2 1.11
1.23
0.85
1.32
0.95
2 1.90
1.39
1. 15
2 1.71
2
1.80
2 1.70
c
3A 2.07
1.32
1.79
1.74
3A 2.62
3A 1.53
2
3A
C
7 3.77 3.95
8 13. 3t 2.26
tion Katio I iecacr,, i liter;
Tvpe and Arrangement NLirr.ber*
CS OKU CH,, OH.-J CHd
0.84 1.00
0.88 0.97
1.00 1.02
1.02 1.08
1.39 1.61
1.53 1.68 1.65
1.59 1.63 1.87
1.17
1.36
1.05 2.68
1.14 2.26
1.86 1.82 1.93 1.58 2.58 1.66
1.32 1.62 2.14 1.91 3.47 1.53
1.81 2.28
1.82 2.73
1.41
1.32
1.38
1.00
1.80
1.60
1.23 1.59
1.22 1.64
1 . 74 2.04
1.73 2.18
2.21 3.39
1.S2 1.76
Sa-.plini; P.Ljn Number
-51 7"*
77A, 77B
74A, 7 IB
7GA, 76B
79A, 79B
93A , 7SA , 73B . 95
91A, 80A, SOB, 96
933
94B
7"\ 7"S
73A, 73B
85A, 85B, 70A, 70B , 83A, 83B, 97
86A, SJ6E, 71A, 71B, 84A, 84B. 98
87A, 81A. SIB
87B, 82A, 82E
83A. S9A
883, S9B
D1A, 90A
92A, 903
91B
92B
53A
51A
57A
53A
59A
52B
533
5 IB
573
583
59B, 69A, 69B
55A
56A
61A
62A
553, 61B, 62B
56B
65A.65B
66A, 66B
67A, 67B
68A, 68B
603, 63A
60A, 633
* Symbols used as follows:
Sampler Orientation: F = flush with wall
S - facing into nominal air st rean (medium vert i cal}
H, = filter medium horizontal facing down
H^ = filter medium, horizontal facing up.
Type sampler: 0 - open ''Delrin"
C = closed "Delrin"
V - vacuum cleaner bags (vert ical int ake}.
t Que s t i o nab1e run.
C-33
-------
T«ble C-5
SUMMARY OF WEATHER MEASURE SAMPLING DATA
O
I
co
Date
Equipment Arrangement Number
Dust Fraction (Fly Ash)
Feeder, Groove Size
Speed Control Settirg
Disperser Air Pressure, psig
Sampling Rate, liters/min
Nominal Dust Concentration, mg/m3
Weather Measure Headings, counts/min
At Sampling location
Equipment
Sec t ton
Supply Line
PlenuQ
Sighting Tunnel
Number
9
10
7
8
12
13
6
5
5A
16
4
11
1
14
2
15
3B
3A
3AA
3
Atmospheric Air
1/28/71
A
0-!0u
1/16 X 1/8
100 100
50 50
40** 2.55
32 32
42
41,42 57,47,48,50
37,38 40,40
43,41
42 49
44
21,25,29
48
3,2
2/3/71
II
0-10H
1/16 X 1/8
100
50 10 20 30 50
2.55 2.55 2.55 2.55 2.55
32 32 32 32 32
0
35
46,55,43 14 49 49 47,51
45
40
42
48 48
46
44,48
36
25,27
20
28 29
28
35,28,26
30,27,21 9 34,21,40
37,39,32
42,29,39
46,38,37
2/11/71
E
0-10n
J/8 X 1/2
20 40 100
50 50 50
2.55 2.55 2.55
24 84 242
34,32 97(106) 322(336)
102(96) 294(276)
31,29 92 272(286)
90(86) 294(256)
2/18/71
B
2.5-5|i
1/8 X 1/2
20 100
50 50
2.55 2.55
20 222
32 376(376)
342(326)
52,47 474(420)
502(470)
51 508(516)
52
31,34 322(246)
390(316)
44,45 448(446)
* Except where otherwise noted, Weather Measure Instrument exhaust was connected to the same metcred vacuum system as was used for the filter sampling runs.
** In this case instrument was exhausted to the atmosphere with the plenum pressure acting to cause flow through the instrument; the flow rate given is estimated.
-------
Appendix D
AEROSOL CHAMBER OPTICAL COMPONENTS
-------
Appendix D
AEROSOL CHAMBER OPTICAL COMPONENTS
The fly ash aerosol generated with the especially designed and as-
sembled equipment described in Appendix A was continuously monitored by a
white-light source transmissometer and, during periods of low background
conditions, by 9CT light scatter from a small portion of the transmissome-
ter light beam (Figure D-l). The transmissometer was designed to minimize
the forward-scattering problem associated with light extinction measure-
ments (Zuev, 1966). Light transmission T is defined in terms of Bougure's
(Beer's) Law as:
T = =
o
where
I = light intensity in absence of an attenuating medium
o
I = measured light intensity
L = path length along the interacting medium
o = volume extinction coefficient.
However, the light intensity I may contain some light that has been scat-
tered by the attenuating medium, resulting in an apparent extinction co-
efficient (G" ) smaller than the true coefficient (cr). The ratio of these
a
* ZUEV, V. E., 1966: Atmospheric transparency in the visible and infrared,
Translated from Russian, available from the Clearinghouse for Federal
Scientific and Technical Information (TT69-55102).
D-3
-------
coefficients depends on the transmissometer geometry (primarily path length
and receiver aperture) and the angular scattering function of the aerosol.
Since larger particles (with respect to the light wavelength) are more ef-
ficient forward scatters,
-------
Event marks on the opposite side of the chart indicate both lidar firings
and pulses signifying one revolution of the particle feeder disk. For
each of these event marks, a value of transmission (voltage) was printed
on the paper tape record.
The 90 side-scatter data were recorded in relative logarithmic units
on a second channel of the strip chart recorder. However, because of high
background light levels during daytime, these data were of only limited
use. Since the transmissometer views a much larger aerosol volume than
the side-scatter photometer, the relative variations between these outputs
are indicators of aerosol homogenity. Larger variations were normally
observed in the transmissometer output indicating the aerosol was well
dispersed within the chamber.
D-5
-------
o
O'l
LIGHT TRAPS
TUNGSTEN
'FILAMENT
LAMP
IRIS
COLLIMATED LIGHT SOURCE
FIELD OF VIEW
Adjustable from
6 to 20 mrad
PROVISION
FOR FILTERS ~~--fc-
FIELD STOP^"
PHOTOMULTIPLIER--
-------
110
POWER
SUPPLY
90 LIGHT
SCATTERING
RECEIVER
TRANSMISSOMETER
RECEIVER
AEROSOL TUNNEL
HIGH-
VOLTAGE
POWER SUPPLY
LIDAR FIRE PULSE
OPERATIONAL
AMPLIFIER
DIGITAL
VOLTMETER
PRINT PULSE
PRINTER
EVENT
MARK
STRIP
CHART
RECORDER
DIRECT
DISPLAY OF
TRANSMISSION
PRINTOUT OF
TRANSMISSION
GOAL = 1% ACCURACY
TA-8730-5
FIGURE D-2 DATA FLOW DIAGRAM OF THE OPTICAL SCATTERING SENSORS
-------
STRIP CHART OUTPUT
LIDAR FIRING
AND DISK
ROTATION PULSES
0.0--
TRANSMISSION
PRINTOUT
1 mm/s
r- O CM i-
m ir) in cn
o 6 ci d
TA-873O-6
FIGURE D-3 DATA FORMAT OF THE OPTICAL SCATTERING SENSORS
D-8
-------
Appendix E
TRANSMISSOMETER DATA
-------
Appendix E
TRAXSMISSOMETER DATA
Transmissometer readings were taken at all times when the sighting
tunnel was in operation. This appendix presents and analyzes all data
showing the relationship between the optical transmission measurements
made in the tunnel and the aerosol concentration for the various aerosols
Details of the transmissometer construction and operation are given in
Appendix D while details of simultaneous concentration and transmission
measurements are given in Appendix C.
Background Theory
Basic Optical Transmission Equations
Optical performance is usually measured in terms of a scattering
coefficient defined by the Beer's relationship:
dl = -GldL . (E-l}~
The gross geometry of an aerosol system can be allowed for by introducing
an efficiency factor, Q , defined as
e
n =n
a = PS P Q (iD"/4)an . (E-2)
e p p
n=o
* A table of nomenclature is given at the end of this appendix.
E-3
-------
The factor, Q , is a function of the ratio of particle size to wavelength
of light used (D A) and the refractive index of the particle (m). For
P
spherical particles this functional relationship is given by the well-
known Mie theory.
Thus, although Q is a function of particle size, for many prac-
e
tical purposes it is convenient to write Equation E-2 in terms of an
average efficiency factor
n=n
a = Q
av n=o
S (TTD /4)An
P P
(E-3)
where, by definition,
av
S Q D 'An
e p p
2
Z D An
P P
(E-4)
The advantage of so defining Q0 is that, as will be shown later, <3
eav eav
will be dependent primarily on some mean particle diameter when dealing
with a mixture of particle sizes.
Mean Particle Diameters
A mixture of particle sizes can be represented by any one of an
infinite number of mean sizes. Mean size can generally be defined in
the following format:
D
DqAn '
P P
L P P-
[E-5)
Certain specific mean sizes will be of special significance in ex-
pressing transmission performance. The "surface mean diameter (sometimes
E-4
-------
called the surface-to-number mean) corresponds to values of 2 and 0
for q and j, respectively,
D
20
L, D
2
an
P P
1/2
;E-G)
The Sauter diameter (sometimes called the "volume-to-surface" mean"
corresponds to values of 3 and 2 for q and j, respectively,
£ D /in
P P
'32 2
£ D An
P P
D
E-7,
But, since, by definition,
w = Z(rrp /6)D An
P P P P
(E-S'
Equation E-7 may also be written as
D
(6w /up )
P P
32 2,
£ D an
P P
(E-9)
if all particles have the same density. If the particles'density varies,
p may be defined as an "average" particle density.
P
Applied Optical Equations
By combining Equations E-l and E-3 with Equation E-6 and E-9, re-
spectively, optical performance can be written in two alternate formats:
E-5
-------
a) In terms of particle number concentration:
dl = -(n/4)q n D" IdL (E-lo)
e p 20
av
or
J2n(l/x) = j£n(l /I) = (rr/4)Q n D L
o e p 20
av
In terms of particle mass concentration:
dl = -(3/2)Q (w /p D n)ldL (E-12)
e p p 32
av
or
= Jin(l /I) = (3/2)(Q w /p D )L . (E-13;
o e p p 32
av
It will be observed that when concentrations are measured in terms
of particle number, the optical performance is dependent on the square
of the mean diameter D ; when measured in terms of mass, it is dependent
&(J
only on the first power of the mean diameter D . In this study mass
concentration of aerosol was determined as well as the Sauter diameter,
D , for each fraction. Consequently, these values in conjunction with
O*j
the measured values of light transmission can be used to evaluate Q
e
av
Experimental Results
The runs in which both aerosol concentration and transmission values
were established are discussed in Appendix C. These results are summarized
in Table E-l. Average values of aerosol concentration and optical trans-
mission are given in Table E-l for all runs made at nominally the same
conditions and with the same aerosol.
E-6
-------
As discussed in Appendix C, aerosol concentrations calculated from
the measured feeder disk speed are believed to be the most reliable mea-
sure of aerosol concentration. Consequently, for this comparison with
optical transmission, only the aerosol concentration values established
from feeder disk speeds will be used.
The data of Table E-l are shown plotted in alternate forms in
Figures E-l and E-2. In Figure E-l the transmission (on a log scale)
is plotted against the mean dust concentration (on an arithmetic scale)
for each dust fraction. As expected from Equation E-13, straight lines
are obtained for each dust, with the finest dust being optically the
most active. The upper scale expresses concentration in grains/cu ft
to permit each comparison with units used in practice.
In order to permit a more quantitative evaluation of the effect of
particle size, the data are plotted in Figure E-2 as transmission (on a
log scale) against the grouping 3Lw /2p D (termed effective opacity)
p p 32
on an arithmetic scale. The slope of such a plot is the average effi-
ciency factor. Lines for values of Q of 1, 2, and 3 are shown as
av
dashed. The data can all be correlated within the actual precision by
the single solid line. This line corresponds to an average efficiency
factor of approximately 2.4. Although there appears to be a trend in
the data for the various fractions toward higher Q values with reduced
av
particle size (as expected from theory for this particle size range),
these trends are believed to be outside of the overall intrinsic precision.
The average efficiency factor calculated for each data point is
shown in the last column of Table E-l. Because of errors in the indivi-
dual values that go into the calculation of Qe , these values for the
av
individual points will tend to exaggerate discrepancies. In general,
however, these values agree well with the value of 2.4 deduced from
Figure E-2.
E-7
-------
Values of Qe have been calculated for each of the dust fractions.
av
To do this, the size distributions given in Appendix B, Table B-4 were
used in conjunction with Mie theory to define Q and the operation in-
e
dicated by Equation E-4 was made on a CDC-6400 computer. Such calcula-
tions were made for two assumed wavelengths of light, 0.7 and 1.06^.
The results of the calculations, shown in Table E-2, show excellent agree-
ment with the experimental results for the transmissometer.
E-8
-------
NOMENCLATURE
D = particle diameter, cm
P
D - any mean particle diameter, cm
qj
D = surface-mean particle diameter, cm
D = Sauter diameter, cm
o*^
I = incident light intensity, lumens/sq. cm
I = initial incident light intensity, lumens/sq. cm
o
L = optical path length, cm
m = refractive index, dimensionless
3
n = particle number concentration, number/cm
P
q,j dimensionless integers
Q = cross section efficiency factor, dimensionless
e
Q - average efficiency factor, dimensionless
e
av
T = fractional light transmission = I/I
o
/ *?
w = particle mass concentration grams/cm
P
3
p = true particle density, grams/cm
P
-1
0" = scattering coefficient, cm
0" = standard geometric deviation, dimensionless
' g
X = wavelength of light, cm
E-9
-------
w
I
0.01
0.02
T
FLY ASH CONCENTRATION mg/m (from feeder rate)
0.03 0.04 0.05 0.06 0.07 0.08
0.1
A = 0-2.5 p.
m = 2.5-5M
V = 5-10/V
= 0-10U
20
40
80 100 120 140
FLY ASH CONCENTRATION
160
grams/ft
180
3
200
220
240 260
TA-8730-36
FIGURE E-1 EFFECT OF FLY ASH CONCENTRATION AND SIZE ON OPTICAL TRANSMISSION
-------
M
I
A = 0-2.5 /i
- 2.5-5/LI
T = 5-
= 0-10M
0.1
EFFECTIVE OPACITY = 3L wp/2 Pp D32 dimensionloss
TA-8730-37
FIGURE E-2 GENERALIZED CORRELATION FOR OPTICAL TRANSMISSION
-------
Table E-l
SUAIMAHV OF TKA.VSMISSOMETER DATA
Fly Ash
Fraction
Used1'
0-2.5 |i.
2.5-5 |L
5-10 (i
0-10 n
Mean
Diameter'
532
Microns
1.7
3.1
6.5
1.0
Feeder Conditions
Groove Speed
Size Control
Sotting
1/16 x 1/8
1/8 x 1/2
1/1G x 1/8
1/8 x 1/2
1/16 x 1/8
1/8 x 1/2
1/16 x 1/8
1/8 x 1/2
100
20
40
100
100
20
40
100
100
20
40
100
20
100
20
40
100
Feeder
Disk Speed
RPM
3.22
0.259
l.OO
3.00
3.27
0.288
1.06
3.22
3.27
0,227
1.02
3.08
0.292
3.33
0.328
1.15
3.35
AvcraRC Results*
Dust-
Concentration
3
mg/m
16.7
12.0
47.0
138.9
27.8
20.3
74.1
223.
32.3
18.1
74.5
206.
2.91
33.2
24.5
84.0
243.
Optical
Transmission
%
86.0
89.2
65.5
30.8
86.4
91.8
73.5
40.2
94.6
96.9
87.3
66.6
98.5
90.8
92.1
76.7
47.7
Data from
Run Numbers
74-77
41-42
43, 46, 47, 49-51
44, 45, 48
78-80, 93-96
37, 38
31-36
39, 40
7O-73, 81-92, 97, 98
99
100, 102
101
21, 22
9-20, 52-69
25, 26
27, 28
29, 30
Effective
Opacity
(3Lw /2p T5 )
p p 32
Dimcnsioaless
0.0550
0.0395
0.1515
0.456
0.0501
0.0366
0.1336
0.401
0.0278
0.0156
0.0641
0.1773
0.0040
0.0463
0.0342
0.1172
0.339
Average
El f iciency
Factor *
Q
e
av
Dimcnsionlcss
2.73
2.89
2.73
2.58
2.91
2.35
2.30
2.27
1.99
2.08
2.11
2.29
3.75
2.08
2.39
2.26
2.18
M
1
* Details of runs given in Appendix C.
t Details of fly ash properties given in Appendix IS.
* Calculated_from measured feeder disk speed and feeder disk dust rate calibration data.
5 Qn = 2p D in (l/T)/3Lw , using p = 2.-16 g/cm , 1, -= 30 ft.
av P 32 P I1
-------
Table E-2
COMPARISON OF MEASURED AND CALCULATED AVERAGE EFFICIENCY FACTORS
Average Efficiency Factor
Fly Ash
Fraction
0-2.51^
2.5-51-1
5-lOlJ.
0- 10M.
Measured
Q
av
( Transmissometer )
2.73
2.46
2.12
2.53
(2.23)T
Calculated^"
Q
e
A. = 0.711
2.76
2.37
2.24
2.54
A. = 1.061-L
3.22
2.57
2.31
2.63
Particle Size
Distribution
Parameters
Geometric
Standard
Deviation
r~
g
1.5
1.5
1.5
2 2
Sauter
Diameter
D
32
Microns
1.7
3.1
6.5
4.0
* Average value from last column of Table E-l.
t Calculated from Q =
6
Q D2An
- -
Z D An
P P
for an assumed log-probability
distribution with parameters taken from the last two columns of
Table B-4, as summarized in the last two columns above.
Omitting one value in Table E-l that was abnormally high and of
questionable precision.
E-13
-------
Appendix F
LIBAR INSTRUMENTATION
-------
Appendix F
LIDAR INSTRUMENTATION
The SRI Mark V lidar system (Figure F-l) was used to collect back-
scatter data at a wavelength of 1.06P- (neodymium laser), and the SRI/EPA
Mark VIII lidar was used to collect backscatter data at a wavelength of
0.69M. (ruby laser). Table F-l gives the lidar characteristics and
Figure F-2 is a diagram of the transmitter/receiver optics and of the
data recording method used with the Mark V lidar. The transmitter and
receiver optics are coaxial so that complete beam convergence occurs at
a short range from the lidar (approximately 50 m). A small portion of
the beam energy is reflected by means of a glass plate beam splitter onto
a diffusing surface that is viewed by a light pipe with a wide acceptance
angle. The output from the light pipe is reflected into one of the photo-
multiplier tubes, and the resulting pulse amplitude is used to normalize
the pulse amplitude returns from both the chamber aerosol and the passive
reflector. Since the peak return from the aerosol is proportional to
transmitted energy and the return from the reflector is proportional to
transmitted peak power, it is assumed that the ratio between these quan-
tities remains constant from pulse to pulse, i.e., the transmitter pulse
shape does not vary. Because of the high bandpass of the lidar signals,
a photographic technique is employed to record the data records. Linear
signal processing and displaying was used because of normal uncertainties
of the transfer characteristics of the video logarithmic amplifier.
However, because the lidar return signals may vary over several orders
of magnitude, a multistage display scheme was used. The output from
one detector is input to three separate oscilloscope displays, each set
F-3
-------
at a different voltage gain. The output from the other detector was
logarithmically amplified and displayed on an oscilloscope to provide
a single trace monitor of the energy return along the beam path. Data
recorded with the Mark VIII lidar, which has only one detector, was
recorded using a single higher bandpass Tektronix 556, dual-beam oscil-
loscope (30 MHz) with two linear displays.
F-4
-------
TA-728&-7
FIGURE F-1 SRI MARK V LIDAR
F-5
-------
D1
02
D3
D4
FILTERS
»-^
NARROW BAND
LOG
AMPLIFIER
LIGHT PIPE
NEUTRAL
DENSITY
TEKTRONIX 555 DUAL-BEAM
OSCILLOSCOPES
TA-8730-24
FIGURE F-2 DIAGRAM OF MARK V LIDAR SYSTEM AND DATA RECORDING FOR BACKSCATTER EXPERIMENTS
-------
Table F-l
CHARACTERISTICS OF SRI/EPA MARK VIII AND SRI MARK V LIDAR
Charac ter1st ics
SRI/EPA Mark VIII Lidar
Transmitter
Laser Rod
Wavelength (A)
Beamwidth (mrad)
Optics
Pulse Energy (joules)
Pulse Length (nsec)
Q-Switch
Cavity Cooling
Receiver
Optics
Field of View (mrad)
Predetection Filter Passband Width (A)
Neutral Density Filter
Detector
Post Detection Filter Bandwidth
Ruby
6943.0
Measured as 2 nrad and reduced to 1 mrad
during this study
2-inch refractor coaxial with receiver
telescope
0.9
Measured as 50 ns during this study
Pockels cell
Refrigerated water
Newtonian reflector (6-in.,
2 and 3 mrad used during this study
10
16 dB
RCA-7265 photomultiplier (S-20 cathode)
30 MHz
SRI Mark V Lidar
Transmitter
Laser Rod
Wavelength (A;
Spectral Line Width (A)
Beamwidth (mrad)
Optics
Pulse Energy (joules)
Pulse Length (nsec)
Q-Switched
Receiver
Optics
Field of View (mrad)
Predetection Filter Passband Width (-
Neutral Density Filter
Detector
Post Detection Filter Bandwidth
Neodynium glass
10,600
90
0.4
2-inch refractor, coaxial with receiver
telescope
1.0
Measured as 23 nsec during this study
Rotating prism
Newtonian reflector (6 in.;
1.0
100
30 dB
RCA 7102 photor.ultiplier ', S-l cathode)
30 MHz
F-7
-------
Appendix G
LIBAR SIGNATURE ANALYSIS
-------
Appendix G
LIDAR SIGNATURE ANALYSIS
Figure G-l is an example of the lidar signatures recorded for various
fly ash concentrations (Mark VIII lidar, 0-2.5 pm diameter fly ash). The
amplitude of the pulse at zero range is a measure of the peak power
emitted by the laser. As the aerosol concentration is increased, the
return from the passive reflector decreases because of the round-trip
attenuation of the laser pulse by the aerosol. The amplitude of the
aerosol return signal (normalized by the transmitted energy) increases
with increases of aerosol concentration. The primary data records con-
sist of three pulses P , P , and P , that are proportional to the peak
t a r
transmitted power (P')> the peak aerosol return (P')> and the peak passive
t a
reflector return (P'). The relationship among these quantities and the
derivation of their information content are best approached by use of
lidar equations for solid and volume targets that may be expressed in
the form:
Solid target: P'
r
-2 r 22
KP'R - T T
t r 'i c a
G-l
Volume target: P'
a
-2 T 2
KP'R B - T
t a 2 c
G-2
where
K = lidar constant
P'
t
= peak power transmitted
G-3
-------
R = range to the target
r = target reflectivity
|3 = volume backscatter coefficient
T = pulse length
T = transmission along the path of clear air
c
T = transmission along the path of the test (chamber) aerosol.
cl
The above expressions assume single scattering, no geometrical effects
over the beam path (coaxial lidar), negligible attenuation over one
laser pulse length, uniformly distributed scatters within the pulse, and
a Lambert (cosine) reflector. For a given lidar, the quantities K, R ,
R , r, and T are assumed to remain constant. The constant T implies that
a
the ratio of transmitted energy to peak power (or pulse shape) remains
constant from pulse to pulse.
Relative backscatter of the chamber aerosol is defined as a. quantity
normalized by transmitted energy, not corrected, however, for variations
of the clear air transmission, i.e.,
P
a
6 = relative backscatter = . G-3
R P
This quantity is dependent on the lidar optical and electrical efficiencies
and hence its numerical values would vary from lidar to lidar. An abso-
lute value of backscatter, the volume backscatter coefficient, requires
calibration of the lidar, which may be accomplished by use of a solid
target (passive reflector) of known reflectivity. The ratio of returns
from a solid target to a volume target leads to the expression:
G-4
-------
P R
a a r 2
n 9 ~ T G"4
P 2 n a
r R
r
The backscatter experiments required a black passive reflector so that
received signals were not sufficiently large to fall in a nonlinear
operating region of the detector (detector saturation). The target re-
flectivity (0.07) was determined from the ratio of reflected target re-
turns from the black target and a standard white target (0.98 reflectivity)
with a neutral density filter placed in the receiver to prevent detector
saturation. Pulse lengths were measured before the backscatter experi-
ments using a photodiode and a high bandpass oscilloscope (Tektronix 556).
Pulse shapes were not monitored during the experiments.
Transmission of the laser energy pulse through the aerosol chamber
T , may be derived from the transmissometer data or inferred from the
a
lidar data. Using the data recorded from a lidar observation made with
a clear air aerosol chamber c, the transmission for a test aerosol may
be evaluated from the expression:
a P P
re t
G-5
where the transmission through a "clear" chamber, T , was assumed to be
ac
unity. Using the lidar derived value of T , the expression for 0 becomes:
cl
2
P P R
a tc a r 2
R = G-6
P P P 2 TT T
t rc R
r
P
a
= const G-7
G-5
-------
and 3 is independent of the target return for lidar firings with a test
aerosol within the chamber. This is important for the backscatter experi-
ments since high aerosol concentrations may result in weak target returns
that cannot accurately be read from the primary data records. A relative
backscatter quantity corrected for variations of the clear air transmis-
sion over the 500-ft path, T , can be derived (Eq. G-6) from the expression
c
P P P
a tc tc
B ' = = 8 . G-8
PR P P PR P
t re re
The value of T in Eq. (G-4) can be replaced by the transmission
a
determined with the white light transmissometer. This transmission mea-
surement is probably a better indicator of the true laser energy trans-
mission (because of less experimental error); however, Eq. (G-4) then
requires a reading of target return energy, P , at times of high aerosol
concentration. Hence, enhanced experimental errors could be expected at
large aerosol concentrations.
The quantities P , P , and P were read from the Polaroid prints
t a. r
as voltages above the receiver noise level that, because of the black
target, was electronic noise limited rather than atmospheric background
limited. The data were placed on computer input cards and subjected to
the above analysis and then plotted in various ways by a computer micro-
film plotting system. The data are tabulated in Appendix H and shown as
data plots in Section III C of the main text.
G-6
-------
UJ
oc
Q
_l
LIDAR
REFLECTOR
T = 0.99
CLEAR AIR
T = 0.88
M = 11 mg/m'
T = 0.60
M = 18 mg/m3
AEROSOL RETURN
REFLECTOR
RETURN
T = 0.32
M = 110 mg/m
MEASURE OF
TRANSMITTED
ENERGY
MEASURE OF
RETURN ENERGY J
FROM AEROSOL
MEASURE
OF RETURN
ENERGY
FROM
REFLECTOR
RANGE
TA-8730-35
FIGURE G-1
EXAMPLES OF LIDAR SIGNATURES (REPRODUCED FROM POLAROID PRINTS)
FOR VARIOUS MASS CONCENTRATIONS OF THE 0- TO 2.5-fi DIAMETER FLY
ASH FRACTION (0.6943 ju WAVELENGTH LIDAR)
G-7
-------
Appendix H
LIDAR DATA SUMMARY
-------
Appendix H
LIDAR DATA SUMMARY
The lidar data collected and analyzed under the present study are
summarized in Table H-l in digital form. Absolute power levels were not
measured because the photomultiplier gain and/or receiver attenuation
for various experimental phases had to be varied. System studies applied
to other lidar techniques are best approached using the absolute scatter-
ing and extinction coefficients presented in Section III-C of the main
text. The relative returns (voltages) shown in Table H-l may be applied
to any lidar by calibrating the lidar in terms of received signal from
a black diffuse passive reflector.
Table H-l
LIDAR DATA SUMMARY
Run No.
1
2
3
4
6
7
8
9
10
11
12
13
14
Lidar
(wavelength )
Mk V (1.06 u.)
Mk V (1.06 u)
Mk V (1.06 p.)
Mk V (1.06 p.)
Mk VIII (0.69 u.)
Mk VIII (0.69 u.)
Mk VIII (0.69 p,)
Mk VIII (0.69 p,)
Mk VIII (0.69 |o,)
Mk VIII (0.69 (o,)
Mk VIII (0.69 |i)
Mk VIII (0.69 jo,)
Mk VIII (0.69 u.)
Fly Ash
Fraction
0-10
0-10
0-2.5
2.5-5
2.5-5
0-2.5
0-2.5
0-10
0-10
Feeder
Groove
Size
1/16 X 1/8
1/8 x 1/2
1/8 x 1/2
1/8 X 1/2
1/8 x 1/2
1/8 X 1/2
1/8 x 1/2
1/8 x 1/2
1/8 X 1/2
Screen Experiment
5-10
0-2.5
0-10
1/8 X 1/2
1/16 X 1/8
1/16 X 1/8
H-3
-------
Table H-2 contains printouts of computer runs. A glossary of terms
used in the column headings precedes the table.
Glossary for Terms in Table H-2
D = particle feeder speed control setting
P = relative transmitted peak power (mV)
P = relative aerosol peak power return (raV)
3,
P = relative passive reflector (black) return (mV)
r
P = relative passive reflector (white) return (mV)
P = relative screen return (mV)
s
T = transmission (white light)
a = volume extinction coefficient (km )
. 3,
M = aerosol mass concentration (mg/m )
H-4
-------
LTDAB FIELD DATA, si
D
0
n
n
0
P
0
n
0
AVG
20
20
20
20
20
AVG
40
40
*n
40
40
40
40
40
40
40
40
40
AVG
100
ion
100
100
AVG
P-.
t
1500
1550
1450
1310
1030
1300
1650
1450
1414
1550
1400
1500
1580
1650
1536
1520
1700
1S50
1810
1450
1650
1350
1380
1550
1520
1?PO
1290
1512
1530
1480
1530
1710
1567
P.
a
-0
-S
n
-1
-n
-ft
ft
-0
^
18
la
2?
22
24
21
75
9i
an
80
12i
135
in
105
ITS
170
Un
160
119
22n
23ft
240
27ft
24ft
p
r
390
350
300
320
250
300
420
300
327
320
300
350
310
400
336
300
400
360
360
340
390
295
300
340
360
260
280
332
300
300
300
350
313
p /P
a t
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
.012
.013
.015
.014
.015
.014
.046
.053
.048
.044
.083
.082
.076
.076
.110
.112
.109
.124
.080
.144
.155
.155
.158
.153
J* NO.
P /P
r t
.253
.226
.207
.246
.243
.231
.255
.207
.233
.206
.214
.233
.196
.242
.219
.197
.?15
.218
.199
.214
.236
.219
.217
.219
.237
.213
.217
.219
.196
.203
.194
.205
.199
1
T
,997
.997
,997
.997
.997
.997
.997
,997
.997
.992
.990
,988
,9P6
.986
,9fl8
.967
.966
.967
,965
.948
.945
.945
.947
,929
.933
.912
.935
.948
.9(19
.699
.901
.899
.902
.33
.33
.33
.33
.33
.33
.33
.33
.33
.93
1.10
1.32
1.54
1.54
1 .27
3.66
3.77
3.66
3.89
5.82
6.17
6.17
5.94
8.03
7,56
7,68
7.33
5. PI
10.40
11.61
11.37
11.61
11.25
.,
11
.11
.11
.11
.11
.1;
11
.11
.11
.29
.37
.44
.52
.52
.43
1.23
1.27
1.23
1.31
1.96
2.07
2.07
2.00
2.70
2.54
2.58
2.46
1.95
3. 5Q
3.90
3.R2
3.90
3.78
LTD'R FIELD D*TAi BUM MO.
D
n
rt
0
0
n
0
0
AVG
20
20
20
2n
20
20
20
20
20
20
AVO
P
t
800
880
12*0
800
1250
1150
1250
1047
1050
uno
sno
1150
1280
1380
1320
500
910
1000
1078
P
a
-n
-n
5
n
*9
-0
^
5
25i
350
lei
3Po
38j
363
340
lOfl
24g
240
274
p
r
540
750
900
660
1100
900
1100
850
700
1140
650
890
1000
1200
1120
340
660
800
850
P /P
a t
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0,000
.238
,250
.225
.261
.297
.261
.258
.200
.267
.240
.250
p /p
r t
.675
.852
.750
,825
.880
.783
.880
.806
.667
.814
.813
.774
.781
.870
.848
.680
.733
.800
.778
T
.995
.995
.995
.995
.995
.955
.995
.995
,9?8
.922
.'523
.923
.924
,928
.920
.924
.912
.916
.922
-
.55
.55
.55
.55
.55
.55
.55
.55
8.15
fl,86
8.74
8.74
8.62
8.15
9.09
8.62
10.05
9.57
8.86
M
,1 8
.18
.18
.18
.1 8
.13
.18
.18
2.74
2.98
2.94
2.94
2.95
2.74
3.06
2.90
3.38
3.21
2.98
H-5
-------
Table H-2 Continued '
LIDAR FIELD DATA, RUN' NO. 2 (CONTINUED)
D
0
0
0
0
AVG
40
40
40
40
40
AVG
100
100
100
100
100
AVG
0
0
AVG
100
100
100
100
100
AVG
D
0
0
1
AV3
20
2(1
20
2(1
20
20
20
20
20
2"
2o
*VG
r>
n
0
0
AVG
100
100
100
AVQ
p
t
1200
1400
2000
1200
1450
800
850
1450
1250
1300
1130
1000
1000
880
700
8SO
886
300
1100
700
600
1350
1610
1350
1640
1310
P
t
1620
1400
1450
1490
1600
1250
1310
1500
1580
1380
1240
1550
1400
1480
1800
1462
1200
1150
1060
900
1H77
1500
1490
1800
1597
P
a
5
-0
2
0
5po
510
85?
75i
686
119n
loon
880
80S
85S
944
-J)
n
S
6on
1150
155*
1305
1500
1220
LTCl'
p
a
1
I
-n
P
600
400
535
46o
Son
son
395
40n
4on
4o5
535
46?
n
-n
-n
-A
"
16fln
1635
16oo
161S
p
r
950
1000
1620
810
1095
310
400
610
610
600
506
200
150
140
120
100
142
150
650
400
100
160
260
200
300
294
»B FIELD
P
r
1480
1300
1300
1360
1200
910
860
1100
1150
910
800
1250
900
1100
1400
1053
1050
920
930
740
910
200
250
300
250
p /P
a t
0.000
0.000
0.000
0,000
0.000
.625
.600
.566
.600
.631
,6H8
1.190
i.ono
1.000
1.143
1.000
1.067
0.000
0,000
0.000
I. 000
.852
.963
,963
.915
.938
DATA,
P /P
a t
0.000
o.ooo
O'OOO
0.000
.375
.3?n
.408
.307
.316
.362
.319
,258
.286
.270
.294
.320
0.000
0.000
0.000
0.000
o.ooo
1.067
1.094
,809
1.017
P /P
r t
.792
.714
.810
.675
.748
.387
.471
.421
.488
.462
.446
.200
.150
.159
.171
.118
.160
.500
.591
.545
.167
.119
.161
.148
.183
.156
RUKi NO.
P /P
r t
.914
.929
.897
.913
.750
.728
.662
.733
.728
,659
.645
.806
.643
.743
.778
.716
.875
.800
,877
.822
.844
.133
.168
.167
.156
T
-0.000
.0.000
.0.000
-0.000
n.ooo
.758
.765
.767
.760
.751
.76f>
.494
.488
.491
,489
.491
.491
-0.000
.0.000
0.000
.496
.497
.498
.496
.493
.496
3
T
.998
.998
.998
.998
.9(12
.913
«9r7
.900
,881
,896
,892
.867
,881
.892
.861
.890
.998
.998
.998
.998
.998
.430
.420
.422
,4?4
.'
-0.00
.0.00
-0,00
-0.00
0.00
30.22
29.21
28.93
29,93
31.23
29.90
76.91
78. ?4
77,57
78.01
77.57
77.66
-0.00
-0.00
0,00
76.46
76,24
76.03
76.46
77.13
76.47
-
.22
22
22
.22
11.25
9.93
10.64
11.49
13.82
11.98
12.46
15.56
13. «2
12.46
16.32
12.70
.22
.22
.22
.22
.22
92.04
94.60
94.fl8
93.57
SI
-0.00
-o.oo
-O.OQ
-0,00
0.00
10.15
9.82
9.77
10.06
10.49
10.05
25. e*
26.29
26.16
86.21
26.06
26. ng
-0.00
.0.00
0.00
25.69
25.62
25.54
25.69
25.91
25.69
M
21
21
21
21
10.61
9.36
10*04
1".83
13.03
11.29
11.75
14.68
13.03
11.75
15.39
11.98
21
21
21
21
21
86.79
89.21
88.72
88. ?4
H-6
-------
p /p
p /p
T
40
4o
41
»n
40
40
40
40
40
4n
AVG
o
n
0
0
0
AVG
100
100
100
100
100
100
100
AVQ
7?0
1100
1350
1300
UOO
1650
1600
1910
1300
1200
13T5
400
600
450
B50
1 000
660
651
1150
300
100
100
160
ooo
1066
son
son
n on
lloo
125o
13flfi
12no
1500
85o
900
1059
-0
o
*2
o
n
n
900
145o
1700
I4no
1400
145o
1300
1371
250
460
600
500
600
700
750
850
7QO
450
586
300
450
350
700
900
540
50
100
80
90
SO
SO
BO
80
.667
.727
.815
.846
.781
.836
.750
.789
.654
.750
.762
0.000
0.000
a. ooo
0.000
0.000
0.000
1.385
1.261
1.308
1.273
1«273
1«250
1«3PO
1-293
.333
.418
.444
.385
.375
.474
.469
.447
.538
.375
.4?!
.750
.750
.778
.874
.900
.800
.077
.087
.062
.082
.073
069
080
076
.630
.680
.630
.631
.628
.537
.628
.6?5
.625
.731
.634
.998
.998
.998
.998
.998
.998
.259
.3nl
.266
.326
.3p4
293
.296
.292
50.39
42.06
SO. 39
50.21
50.^3
67. 8n
50.73
51.25
51.25
34.17
49.90
.22
.22
.22
.22
.22
.22
147.32
13Q.93
144.4]
122.23
129.85
1 33«87
132*76
134.48
47.51
39.66
47.51
47.35
47.84
63.94
47.84
48.33
48.31
3?. 22
47.05
.21
21
21
21
.21
21
138,92
123.47
136.18
115.26
122.45
126«24
125.19
126-82
LTOAB FIELD DATA, RUM NO. 4
p /p
p /p
0
n
0
n
AVG
40
40
40
40
40
4n
40
40
+ 0
40
40
40
40
*0
AVQ
n
0
0
AVG
1900
1660
1600
1900
1765
1160
1550
1200
1300
1650
1300
1730
1600
1550
900
1100
1400
1300
1680
1389
BOO
1400
1600
1Z67
-0
n
-0
-n
n
750
lf)5n
80o
900
1200
9fn
lion
1000
lion
60o
65o
9oii
1000
118o
938
-0
o
-o
n
1400
1100
1100
1490
1272
300
500
400
400
6QO
430
600
500
400
250
300
400
450
600
418
400
750
1000
717
0.000
1.000
0.000
0.000
n.OOO
.647
.677
.667
.692
.727
.692
.629
.625
.710
.667
.591
.643
.769
.702
.674
0.000
0.000
0.000
0.000
.737
.663
.688
.764
.71 8
.259
.3?3
.333
.308
.364
.331
.343
.313
.258
.278
.273
.286
.346
.357
.312
.500
.536
.635
.554
.995
.995
,995
.995
.995
.735
.7*8
.747
.7^4
.752
.732
.752
.817
.740
.740
.740
.740
.740
.740
.748
.995
.995
.995
.995
.55
.55
.55
.55
.55
33.58
31.66
31.81
30.79
31.08
34.02
31.08
22.04
32-84
32.04
32.84
32.84
32.84
32.84
31.65
.55
.55
.55
.55
.?5
.25
.25
.25
25
15.11
14.25
14,31
13.86
13.99
15.31
13.99
9.9?
14.78
14.78
14.78
14.7R
14.78
14.78
14.24
.25
25
25
.25
H-7
-------
Table H-2 .Cent inucd ^
LIDAR FIELD DATA, RUN NO. -1 (CONTINUED)
D
40
40
40
40
*0
40
AVG
t
o
AVG
40
4n
*0
40
AVG
0
0
AVG
20
20
20
20
20
2"
AVG
0
n
0
AVG
20
20
20
20
AVG
0
0
0
AVG
100
100
100
100
100
100
100
100
100
AVG
0
0
n
AVQ
P
t
1400
1400
1550
1450
1350
1700
1475
1300
1500
1400
1400
1700
1600
1SOO
1550
1300
loco
1150
1380
1200
1*10
1600
1060
920
1295
1210
11?0
1300
1220
1000
1430
lino
13(50
1207
1250
950
1200
1133
1150
1400
1151
10*0
lino
1200
1410
1550
1500
12*2
1080
mo
1108
1113
P
a
75*
100P
iioii
980
loon
lion
988
0
o
fl
loon
I25r
lino
llor
111?
-r
-0
P
400
36o
5Pn
500
256
2fto
36»
-0
-p
«A
n
29S
30ii
35n
33o
317
-n
-0
"2
n
142n
175«
1600
Uftfl
16no
168n
172?
1800
2100
16P1
-0
0
-n
n
p
r
350
500
450
450
450
570
462
700
900
800
400
500
560
500
490
TOO
600
650
700
610
810
900
540
500
677
ft90
700
900
763
480
800
600
650
632
700
600
800
700
100
150
120
120
130
150
150
160
200
142
680
710
760
717
P /P
& t
.536
.714
.710
.676
.741
.647
.671
0.000
0.000
0.000
.714
.735
.688
.733
.718
0.000
0.000
0.000
.290
.300
.311
.313
.236
.217
.278
0.000
0.000
0.000
0.000
.290
.210
.318
.254
.268
0,000
0.000
0.000
0.000
.235
.250
.391
.352
.455
.400
.220
.161
1.400
1.318
0.000
0.000
0.000
0.000
p /p
r t
.250
.357
.290
.310
.333
.335
.313
.538
.600
.569
.286
.294
.350
.3^3
.316
.538
,600
.569
.507
.508
.503
.563
.509
.54.1
.522
.570
,609
.692
.624
.480
.559
.545
.500
.521
.560
.632
.667
.619
.087
.107
.104
.111
.118
.125
.106
.103
.133
.111
.630
.612
.691
.644
T
.738
.731
.724
.742
.751
.715
.733
.995
.995
.995
.740
.728
.72?
.726
.730
.995
.995
.995
.890
.927
.922
,9?4
.928
.91P
.918
.995
.995
.995
.995
.918
.927
.910
,911
,916
.995
.995
.995
.995
.424
.418
.428
,4n6
.418
.417
.413
.415
.417
.417
.995
.995
.995
.995
-
33.13
34.17
35.22
32.54
31.23
36.58
33.81
.55
.55
.55
32.94
34.62
34.62
34.92
34.25
.55
.55
.55
12.71
8.27
R.86
8.6?
8.15
9,33
9.32
.55
.55
.55
.55
9.33
a. 27
10.28
10.16
9.51
.55
.55
.55
.55
93.57
95.12
92.54
98.30
95.12
95.38
96.43
95.91
95.38
95.31
.55
.55
.55
.55
M
14.91
15.38
15.85
14.64
14. ns
16.46
15.22
.25
.25
.25
14.78
15.58
15.58
15.71
15.41
.25
.25
.25
5.72
3.72
3.99
3.89
3.67
4,2?
4.19
.25
.25
.?5
.25
4.20
3.72
4,63
4.57
4.28
.?5
.25
.25
.25
42.11
42.81
41 .64
44.23
42.81
42.92
43.40
43.16
4?. 92
42.89
.25
.?5
i?5
25
H-8
-------
Table i[-2
LTHAR FIELD DATA. RUN NO. <
P /P
P /P
0
0
0
0
AVG
2n
20
20
20
2n
20
2r
2o
AVG
0
r,
0
AVG
40
40
40
40
40
40
40
40
40
40
AVQ
0
n
0
0
AVG
100
100
100
100
100
100
»VG
315
3*8
360
420
3*6
352
375
377
427
375
4?0
375
450
393
315
344
3«2
347
315
372
339
3<57
420
420
427
398
360
337
378
24n
3"7
360
375
320
342
300
270
285
3*0
340
319
-^
«*
-0
"5
*
152
28fl
23?
232
18S
2P?.
22*
2ls
226
-n
-n
n
n
597
630
54*
566
617
69ft
607
6i«;
49«;
55i
59,
-f<
«
-o
* n
n
77?
727
756
65?
78n
96»
77*
3«2
407
330
502
405
255
352
330
360
322
372
390
465
356
372
360
360
364
234
282
225
225
279
315
330
288
210
255
264
307
330
375
450
365
140
105
8o
75
169
142
llfl
0.000
o.ooo
0.000
0.000
0.000
.432
.768
.624
.543
.480
.671
600
.484
.575
0.000
0.000
0.000
0.000
1.895
1 .694
1 .593
1 .426
1 .469
1.657
1.422
1.545
1.375
1.635
1.571
0.000
0-000
0.000
0.000
0.000
2.257
2-423
2.800
2-268
2.167
2.689
2.437
1 .213
1.116
.917
1 .195
1.108
.7?4
.Q39
.RS7
.S43
.859
.886
1.040
1.033
.901
1.191
1 .047
.942
1.057
.743
7S8
.664
.567
.664
.750
.773
.7?4
.583
.757
.698
1.279
1.075
1.042
1.200
1.149
.409
.350
.296
.263
.469
.39*
.364
.995
.995
,994
.994
.994
.925
.921
.9?3
.916
.9?!
.915
.9?4
.9!9
.920
.993
.991
.995
.993
.764
.767
.763
.771
.779
.782
.796
.770
.776
.780
.775
.992
.991
.992
.991
.991
.474
,478
.498
.534
.509
.545
.507
.55
.55
.66
.66
.60
8.50
8.97
fl.74
9.57
a. 97
9.69
8.62
-------
LTD*" FIELD DATA» RUNi NO, 7
p /P
P /P
0
(1
0
n
AVG
2i
20
20
20
20
20
20
2o
AVG
0
0
0
c
AVG
40
40
40
+ 0
40
4n
40
4n
AVG
0
0
n
AVG
100
100
10P
100
100
100
100
100
100
100
100
AVG
150
195
175
US
171
140
130
150
150
183
185
inn
135
1*7
135
flO
140
145
125
1*0
164
155
135
165
160
115
138
149
95
125
13«
117
93
1"5
128
124
128
125
1*2
125
115
140
128
119
2
«n
-n
-0
P
12-5
7?
9^
Us
155
I7n
*5
lln
121
n
fl
-n
6
6
34*
38«
309
3on
40?
40?
266
31?
340
V A
»«
0
"
278
294
as?
36ft
357
377
3l«
3SJ,
369
37*
37n
337
130
215
160
160
166
95
88
125
140
135
US
75
105
113
135
55
1*5
105
110
65
65
55
42
65
58
45
65
58
95
115
UO
117
25
35
60
39
41
35
28
40
35
40
40
38
0.000
0.000
0.000
0.000
c.ooc
.893
.554
.633
.967
.838
.919
.950
.815
.821
0.000
0.000
0.000
n.ooo
0.000
2.125
2.366
1.994
2.222
2.455
2.500
2.313
2.261
2.279
0.000
0.000
0.000
n.ooo
2.989
2.800
I .992
2.9"52
2.789
3.01*
3.088
2.848
3.209
2.671
2.891
2.840
.867
1.103
.914
.970
.963
,679
.677
,S33
.933
.730
.784
.750
.778
.770
1.000
.688
1.036
.724
.862
.406
.396
.355
.311
.394
.362
.391
.471
.3*6
1.000
.920
1.077
.999
,?69
.333
.469
.315
.320
.280
.275
.3?0
.304
.286
t313
.317
.993
.993
.993
.993
.993
.865
.935
,9?4
.898
.864
.871
.865
.877
.886
.991
.991
,991
.990
.991
.628
.634
.625
.««*
.60!
.623
.631
.641
.624
.999
.991
.995
.995
.432
.565
,7"8
.547
.514
.515
.514
.513
.482
,5P8
.561
.534
.77
.77
.77
.77
.77
15. «2
7.33
8.62
12.95
15.94
15.06
15. «2
14.31
13.23
.99
.99
.99
1.10
1.01
50.73
49.70
51. ?5
54.26
55.52
51,6.1
50.21
48.50
51 .47
.11
.99
.55
.55
91.53
62.26
37.66
65,79
72.58
72.37
72.58
72.79
79.59
69,65
63.04
69.17
.72
.7?
.72
.72
.72
14.91
6.91
8.13
12.22
15.03
14.20
14,91
13.50
12.48
.93
.93
.93
1.03
.96
47.84
46.86
48,33
51,17
52,36
48.66
47.35
45.73
48.54
.10
,93
.5?
,52
86.31
58.71
35,51
62. n*
68.44
68, 24
68,44
68.64
75,^5
65.68
59.44
65.14
H-10
-------
LTHAR FIELD oiia, c
D
2n
20
20
20
20
20
AVG
40
40
40
40
40
AVG
100
100
100
100
100
100
100
AVG
P
t
73
114
75
118
113
95
98
95
120
134
129
120
120
95
100
95
124
108
125
130
111
P_
5s
6P
73
6f
9c
9s
7?
19?
32?
384
37?
3nS
31?
37?
37?
487
49?
46?
52?
54fl
466
P
r
75
93
75
100
105
85
89
60
44
62
55
50
54
8
13
8
13
8
15
10
11
p /P
a t
.753
.596
.973
.559
.841
1 .000
.787
2.053
2.708
2.866
2.884
2.500
2.602
3.947
3,750
5.126
3.992
4.306
4,176
4.154
4.207
'UK NO,
P /P
r t
1.027
,816
1 .000
.847
.929
.895
.919
.632
.367
.463
.426
.417
.461
.0*4
.130
.OS*
.105
,074
,1?0
.077
,096
Q
f
.911
.872
.875
.883
.875
,870
,R8l
.630
,6P4
.612
.592
,606
.609
.231
.357
.270
,3?0
.276
.310
.257
.289
10.16
14,94
14,56
13.57
14.56
15.19
13.83
50,39
54.98
53.55
57.17
54.62
54.14
159.80
1 12,32
142.78
124.26
140.39
127.72
148,17
136.49
9.59
14. rg
13.73
12.80
13.73
14.32
13,04
47,51
51 .85
50, 49
53.91
51.51
51.05
150.69
105.92
134,65
117.17
132.39
120.44
139,72
128.71
LTD'R FIELD DATA. HUN NO, 9
P /P
0
0
0
0
0
AVG
20
20
20
2n
20
20
20
20
AVG
0
0
0
r
AVr,
40
40
40
40
40
40
40
40
AVG
95
125
125
114
128
117
108
155
132
110
158
133
160
139
137
120
120
115
115
117
115
120
138
140
120
132
145
118
128
-n
n
-n
^
-0
0
85
7a
loo
6 =
8P
9o
7(1
8?
84
»i
A
n
rt
*
20=;
2o«
2li
IB?
16-»
19?
19?
Ifl?
191
100
125
93
78
110
lo.l
95
1P9
110
70
1*5
108
120
115
108
1*5
100
92
110
112
85
SO
90
80
62
100
88
85
84
0.000
0.000
0.000
0.000
0.000
0.000
.787
.503
.758
.591
.557
.677
.487
.612
.621
0 ,000
0.000
0.000
0.000
0.000
1.783
1.667
1.522
1.300
1.358
1,477
1.324
1.568
1.500
1 .053
1.000
.744
.684
.859
.868
,7S7
.703
.833
.636
.919
.812
.750
.827
.783
1 .208
.833
.800
.957
.950
.739
.667
.652
.571
.517
.758
.607
.720
.654
.909
.999
.999
.997
.999
.999
.900
.913
.9no
.907
.915
,894
.896
.892
.902
.993
.993
.992
.993
.993
.749
.750
.754
.761
.759
.759
.767
.755
.757
.11
11
11
.33
.11
.15
11 .49
9.93
11 .49
10.64
9.69
12.2?.
11 .98
12.46
11.24
.77
.77
,88
.77
. 79
31.52
31.37
30.79
29.78
30.07
30«07
28.93
30.65
30,40
.04
.04
.04
11
.04
.05
3.H6
3.34
3.86
3.S8
3.25
4.1]
4.02
4.19
3,78
?6
.26
.29
.26
.27
10.59
10.54
10.35
10.01
10.10
10. 10
9.7J
10.30
10.21
H-ll
-------
Table U-2 Continued
LIDAR FIELD DATA, RUX NO. 9 (CONTI.NTED)
D
0
n
0
AVG
100
100
100
100
100
100
100
100
AVG
D
n
0
0
0
AVG
20
20
20
20
A«G
40
41
*0
»0
AVG
100
100
100
100
100
AVG
p
t
8!
131
128
115
120
135
109
125
175
115
130
119
122
P
t
99
1(15
140
120
116
100
125
115
115
114
98
1 18
115
135
116
98
1?4
128
105
130
117
p
a
ft
ft
ft
A
36ft
405
338
34=;
36"
345
375
318
35*
LTPAB
p
a
* *
ft
ft
ft
ft
45
55
45
5ft
4Q
u;
125
14ft
16*
147
300
3l«
314
311
308
31?
P
r
Io5
140
122
122
31
45
35
30
35
35
39
29
35
FIELD
P
r
99
120
145
132
124
94
110
100
98
100
70
70
70
ln2
78
42
38
46
30
35
38
P /P
a t
0.000
o.ooo
0.000
0.000
3.000
3.000
3.101
2.760
2.880
3.000
2.885
2.672
2.912
DATA,
P /P
a t
0.000
0.000
o.ooo
0.000
0.000
.450
.440
.391
.435
.429
1.429
1.059
1.217
1.222
1.232
3.061
2.565
2.453
2.962
2.369
2.682
p /P
r t
1.235
1.069
.953
1.086
.258
.333
.321
.240
.280
.304
.300
.244
.285
"UNI NO. i
p /p
r t
1.000
1.143
1.036
1.100
1 .070
.940
.880
.870
.852
.885
.714
.593
.609
.756
.668
.429
.306
.359
.286
.269
.33(1
T
.993
.994
.994
.994
.494
.497
.490
.491
.487
,*91
.495
.489
.492
0
T
.993
.994
.993
.993
.993
.922
.937
.931
.920
.927
.774
.798
.781
.788
,7P5
.542
.543
.542
.504
.514
.579
j
.77
66
.66
69
76.91
76,74
77.79
77.57
78.46
77.57
76,68
78.01
77.41
.77
.66
.77
.77
.74
8,86
7.10
7.8(1
9.09
8.21
27.94
24.61
26.96
25.98
26.37
66.79
66.59
66.79
74.72
72.58
69.49
M
.26
22
22
23
25.84
25.62
26.14
26.06
26.36
26.06
25.77
26.21
26.01
\(
.26
72
7.6
76
25
?.98
2.38
2«62
3.06
2.76
9.39
8.27
9.06
8.73
8.86
22.44
22.37
22.44
25.11
24.39
23.35
LTOAR FIELD DATA, RUN NO. 11
D
0
0
0
0
0
0
0
0
0
0
0
AVG
p
t
139
130
148
144
138
128
145
135
140
132
139
138
P
s
n
5
-1
ft
-n
-0
ft
n
n
ft
n
-0
p
r
155
145
152
165
175
130
142
138
150
148
140
149
P /P
s t
0.000
0.000
o.ooo
1.000
0.000
0.000
0.000
0.000
0.000
o.OOO
0.000
0.000
p /p
r t
1.115
1.115
1 .027
1.146
1.268
1 .OlA
.979
1.022
1.071
1.121
1.007
1.081
H-12
-------
D
0
0
1
n
0
i
n
0
0
AVG
D
(1
0
0
n
0
0
n
n
0
0
0
0
AVG
P
t
lAfl
135
12?
124
140
125
131
1"5
125
127
P
t
130
125
125
12n
130
108
119
135
12n
135
136
130
126
P
s
-0
n
n
-n
-n
n
n
« n
-0
-0
p
5
434
44)
44f
481
42*
33T
36S
494
479
47
-------
Table H-2 Continued
LIDAR FIELD DATA, RUN NO. 12 (CONTINUED)
D
*0
40
40
4n
40
*0
*0
AVG
0
(1
0
0
AVG
100
100
100
ion
100
100
100
100
AV4
17.01
.55
.77
55
.55
.60
49.87
-------
D
100
100
ion
101
100
:oo
ion
100
AVO
40
41
4C
4,1
40
41
40
AVC
D
>
n
0
1
AVG
20
2'J
2fl
20
2')
AVG
40
41
4n
40
41
40
40
40
AVG
0
0
n
0
AVC,
100
100
100
ion
100
100
100
100
AVG
n
0
0
0
AVG
P
t
no
104
110
110
130
105
98
104
119
102
115
11"
139
110
122
110
115
P
t
89
1"5
95
qo
95
85
»5
inn
119
]n(j
99
sn
72
89
lift
90
119
115
113
99
114
90
115
112
113
100
115
<=9
no
94
110
109
li
1
.;
I?
s
q
li
9
9
2"
2i
2l
2*
2ft
24
2P
3*
26
« r>
-0
-0
n
n
65
5?
5?
61
40
65
67
60
59
n
.0
.n
-ft
rt
P
r
85
88
85
98
112
82
80
100
91
118
U)5
If 9
142
l'':5
149
11 5
119
FIELD
p
r
90
115
82
100
97
106
85
115
135
130
112
95
qn
88
122
95
125
129
126
l-<9
115
100
130
128
118
115
ion
95
110
78
102
U:9
110
102
130
110
159
132
133
p /p
a t
.955
.9?3
.909
.955
.985
.848
.827
1.231
.954
.333
.270
.264
.281
.264
.32"
.236
.281
DATA.
P /P
a t
0.000
0.000
0.000
0.000
0.000
.118
.094
.090
.OPS
.083
.094
.222
.292
.236
.245
.222
.202
.243
.265
.241
0.000
0.000
0.000
0.000
O.OPO
.620
.452
.584
.555
.511
.591
.569
.639
.565
0.000
0.000
0.000
0.000
0.000
p /p
r t
.773
.846
.773
.891
.862
.781
.81 6
.962
.833
1 .157
.913
.98?
I.0?2
.955
!.??!
.955
I ,0?9
SUN NO. 14
p /r
r t
l.hi i
1.005
."63
l.ll 1
1 .021
l.?47
1 .000
1.050
1 .144
1.193
1.1?7
1.056
1 .250
.989
1.151
1 .056
1 .050
1.1?2
1.115
1 .099
1.106
1.111
1 ,?33
1.143
1 .149
1.1 = 0
.S70
1.0*7
1 .010
.830
.9?7
I .000
1 .019
.983
1.193
1,058
1.233
1,200
1.171
T
.851
.792
,8»2
.840
.795
.874
.869
.842
.839
.951
,9*1
.959
.949
.953
.954
,958
.954
T
.993
.992
.993
,995
.993
.97fl
.986
.981
.9°0
.983
.982
.962
.959
.961
.962
.9*2
.965
.967
.969
.963
.997
.997
.997
.997
,997
.9i6
.906
,916
,9"8
.904
.916
.908
.917
,916
.9*1
,994
,993
.992
,990
17.59
25.43
18.75
19.01
25.02
14.69
15.31
18.75
19.32
5.4?
5.48
4.57
5.71
5. '5
5.14
4.6B
5.19
-
.77
.88
.77
.55
.74
?.43
1 ,C4
2.19
2.21
1 .07
2.13
4.22
4.57
4 ,34
4.22
4.22
3.«9
3.66
3.43
4.17
.33
.33
.33
.33
.33
10.77
10.77
10,77
10.5?
11.01
1C. 77
10. =2
10.64
10.72
2.09
.66
.77
.88
1 .10
16.59
23.98
17.69
17.93
23.59
13.85
14,44
17.69
18.22
5.17
5.17
4.31
5.39
4.95
4.84
4.4\
4.S9
.26
.'9
.26
.18
.'5
.8?
.52
.71
.74
.63
.69
1.42
1.53
I .46
1.4?
1 .4?
1.31
1.23
1.15
1.37
.11
.11
.11
.11
.11
3.62
3.62
3.62
3.54
3.75
3.62
3.5»
3.58
3.60
.7n
.'2
.26
.29
.37
H-15
------- |