Final Report
LIDAR STUDY OF STACK PLUMES
By: WARREN B. JOHNSON, JR. and EDWARD E. UTHE
Prepared for:
DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE
NATIONAL AIR POLLUTION CONTROL ADMINISTRATION
DIVISION OF METEOROLOGY
411 WEST CHAPEL HILL STREET
DURHAM, NORTH CAROLINA 27701
CONTRACT PH 22-68-33
STANFORD RESEARCH INSTITUTE
Menlo Park, California 94025 • U.S.A.
-------
STANFORD RESEARCH INSTITUTE
Menlo Park, California 94025 • U.S.A.
Final Report
LIDAR STUDY OF STACK PLUMES
By: WARREN B. JOHNSON, JR. and EDWARD E. UTHE
Prepared for:
DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE
NATIONAL AIR POLLUTION CONTROL ADMINISTRATION
DIVISION OF METEOROLOGY
411 WEST CHAPEL HILL STREET
DURHAM, NORTH CAROLINA 27701
CONTRACT PH 22-68-33
SRI Project 7289
Approved:
R. T. H. COLLIS
Director
Aerophysics Laboratory
RAY L. LEADABRAND
Executive Director
Electronics and Radio Sciences Division
Copy No. -'::.
-------
ABSTRACT
The feasibility of lidar (laser radar) for stack plume studies is
established from the results of an experimental investigation of plume
behavior from a 245-m power plant stack in western Pennsylvania. During
this study a total of 175 vertical plume cross sections containing about
3800 separate lidar observations were obtained, of which 64 cross sec-
tions representative of various types of plume behavior were selected
for detailed analysis. Each vertical cross section was built up from
15 to 30 lidar shots at 5 to 8 second intervals and at elevation angle
increments of 1/3° to 10°. The selected cross sections are grouped into
series which show the spatial (downwind) and temporal variations in
plume geometry and relative particulate concentration distributions.
The factors involved in interpreting the lidar data in terms of
plume rise and diffusion are discussed and exemplified. Although cal-
culated plume-rise values agree reasonably well with the observations,
it is clear from inspection of the cross sections that the important
effects of vertical wind direction shear (plume tilting and fanning)
and vertical changes in stability (plume trapping) should be taken into
account when predicting plume rise and diffusion. Close correspondence
between plume tops and levels of increased atmospheric stability was
found. Several cross sections are shown of fumigating plumes, which
occurred frequently. In a tilted plume, different portions apparently
fumigate at different times. Optimum use of lidar for diffusion studies
requires provision for obtaining 30-minute or hourly plume concentration
distributions, as well as allowances for the effect of the lidar noise
level upon plume size.
To investigate the potential of lidar for making quantitative
measurements, a sample cross section of absolute mass concentration was
computed on the basis of Mie scattering theory and independent particle
size measurements. The cross-axial integrated mass concentration repre-
sented by this cross section is 680 g/m. The corresponding value calcu-
lated from the gross power-plant data and wind speed is 875 g/m.
iii
-------
The advantages of the mobile lidar technique stem from its ability
to obtain measurements remotely and at a high density in space and time.
The quantitative application of the technique for obtaining absolute
particulate mass concentrations is limited mainly by the accuracy with
which the optical characteristics of the aerosol are known. An improved
lidar system for plume studies would incorporate better signal-to-noise
characteristics and a higher pulse repetition rate than present equip-
ment provides. in addition, an expanded data processing capability with
real-time" output is necessary.
IV
-------
CONTENTS
ABSTRACT iii
LIST OF ILLUSTRATIONS vii
LIST OF TABLES xi
LIST OF SYMBOLS xiii
FOREWORD xv
I INTRODUCTION 1
II SUMMARY AND CONCLUSIONS 3
A. Plume Characteristics 3
B. Lidar Data Analysis Considerations 4
C. Instrumentation Considerations 5
III DESCRIPTION OF THE EXPERIMENTAL PROGRAM 7
A. Instrumentation 7
B. Experimental Procedures 9
C. Summary of Lidar Data 12
IV BASIC ANALYSIS OF LIDAR BACKSCATTER SIGNATURES 15
A. General Discussion 15
B. Lidar S-Function Analysis 17
V RELATIVE CONCENTRATION CROSS SECTIONS OF KEYSTONE
STACK PLUME 25
A. May 1968 Observation Period 25
B. October 1968 Observation Period 33
VI INTERPRETATION OF CROSS SECTIONS IN TERMS
OF PLUME RISE AND DIFFUSION 53
VII EVALUATION OF PARTICULATE MASS CONCENTRATIONS
FROM LIDAR NORMALIZED SIGNAL RETURN (S) VALUES 59
VIII ANALYSIS OF ATTENUATION EFFECT
USING PARTICLE-SIZE CHARACTERISTICS 65
-------
IX RECOMMENDATIONS FOR FURTHER RESEARCH 73
A. Lidar Instrumentation 73
B. Data Handling and Reduction Capabilities 74
C. Experimental and Analytical Techniques 76
D. Advanced Lidar Concepts 7*>
REFERENCES 79
Appendix A—DESCRIPTION OF MARK V LIDAR 81
Appendix B--DETAILED LIDAR DATA SUMMARIES 89
Appendix C—LIDAR DATA DIGITIZING AND CONDITIONING DETAILS. . 97
Appendix D--PLAN VIEWS OF HORIZONTAL PLUME POSITIONS
FROM LIDAR OBSERVATIONS 101
Appendix E--LIDAR EYE-SAFETY CONSIDERATIONS 113
vi
-------
ILLUSTRATIONS
Frontispiece
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7
Figure 8
Figure 9
Figure 10
Figure 11
Figure 12
Figure 13
Figure 14
Figure 15
Mark V Lidar in Operation during Keystone
Stack Plume Study near Shelocta, Pennsylvania
SRI Mark V Lidar
Lidar Sites Used during Field Experiments .
Basic Lidar Observation of Smoke Plume. . .
Example of Multiple-Trace Format Used
for Display of Magnetic Disk Records. . . .
Lidar Data Analysis Process ,
Effect of Range Correction upon Plume Signal
Return and Receiver Noise Level ,
Comparison of Plume Cross Sections Contoured
in Linear Units (a) and Logarithmic
Units (b)
Relative Height Considerations Associated
with Lidar Plume Observations ,
Spatial Series of Lidar-Observed Vertical
Cross Sections under Stable Conditions,
25 May 1968, from Site 1, 3.5 km from Stack
Spatial Series of Lidar-Observed Vertical
Cross Sections under Moderate Diffusion
Conditions, 25 May 1968, from Site 1,
3.5 km Southwest of Stack
Spatial Series of Lidar-Observed Vertical
Cross Sections under Stable Conditions,
26 May 1968, from Site 2, 3.5 km North of
Stack
Spatial Series of Lidar-Observed Vertical
Cross Sections under Unstable (Looping)
Conditions, 26 May 1968, from Site 2,
3.5 km North of Stack
Vertical Temperature Profiles Associated
with Lidar Observations, 25-26 May 1968 . . ,
Comparison of Manual and Machine Analyses .
Time Series of Lidar-Observed Vertical
Cross Sections from Site 6, 3.8 km Northeast
of Stack, Plume Overhead, 15 October 1968 .
8
10
15
16
18
21
22
26
27
28
29
30
31
34
35
VII
-------
Figure 16 Time Series of Lidar-Observed Vertical
Cross Sections from Site 6, 3.8 km
Northeast of Stack, Plume Overhead,
16 October 1968
Figure 17 Time Series of Lidar-Observed Vertical
Cross Sections from Site 6, 3.8 km
Northeast of Stack, Plume Overhead,
17 October 1968
Figure 18 Time Series of Lidar-Observed Vertical
Cross Sections from Site 17, 13.7 km
North-Northwest of Stack, Plume Overhead,
18 October 1968
Figure 19 Time Series of Lidar-Observed Vertical
Cross Sections from Site 21, 21 km
East-Southeast of Stack, Plume Overhead,
20 October 1968
Figure 20 Spatial Series of Lidar-Observed Vertical
Cross Sections from Site 22, 4.9 km
Southeast of Stack, 21 October 1968,
Immediately Prior to Full Precipitator
Operation
Figure 21 Spatial Series of Lidar-Observed Vertical
Cross Sections from Site 22, 4.9 km
Southeast of Stack, 21 October 1968,
Precipitators in Full Operation
Figure 22 Spatial Series of Lidar-Observed Vertical
Cross Sections from Site 22, 4.9 km
Southeast of Stack, 21 October 1968 . . .
Figure 23 Spatial Series of Lidar-Observed Vertical
Cross Sections from Site 6, 3.8 km
Northeast of Stack, 22 October 1968 . . .
Figure 24 Spatial Series of Lidar-Observed Vertical
Cross Sections from Site 6, 3.8 km
Northeast of Stack, 22 October 1968,
Clean Plume
Figure 25 Spatial Series of Lidar-Observed Vertical
Cross Sections from Site 6, 3.8 km
Northeast of Stack, 22 October 1968 . . .
Figure 26 Spatial Series of Lidar-Observed Vertical
Cross Sections from Site 6, 3.8 km
Northeast of Stack, 22 October 1968 . . .
Figure 27 Spatial Series of Lidar-Observed Vertical
Cross Sections from Site 6, 3.8 km
Northeast of Stack, 23 October 1968 . . .
37
38
39
40
41
42
44
45
46
47
48
49
viii
-------
Figure 28 Time Series of Lidar-Observed Vertical
Cross Sections from Site 11, 4 km
Northwest of Stack, 24 October 1968
Figure 29 Time Series of Lidar-Observed Vertical
Cross Sections from Site 11, 4 km
Northwest of Stack, 24 October 1968
Figure 30 Time Series of Lidar-Observed Vertical
Cross Sections from Site 11, 4 km
Northwest of Stack, 24 October 1968
Figure 31 Dependence upon Particle Size of Total
Scattering (Extinction) Optical Cross
Section per Unit Mass (§e) ., and
Backscattering Optical Cross Section per
Unit Mass per Steradian (^b)
Figure 32 Sample Cumulative Particle Size Distribution
and Junge-Type Analytical Fit ,
Figure 33 Effect and Magnitude of Correction of Plume
Mass Concentration Distribution for
Attenuation
Figure 34 Visualization of Intensity-Modulated
Range/Height Oscilloscope Display of
Plume Cross Section Observed by Lidar . . . ,
Figure A-l Optical System of Mark V Lidar
Figure A-2 Electronics System of Mark V Lidar
Figure D-l Plan View of Horizontal Plume Positions,
15 October 1968
Figure D-2 Plan View of Horizontal Plume Positions,
16 October 1968
Figure D-3 Plan View of Horizontal Plume Positions,
17 October 1968
Figure D-4 Plan View of Horizontal Plume Positions,
18 October 1968
Figure D-5 Plan View of Horizontal Plume Positions,
20 October 1968
Figure D-6 Plan View of Horizontal Plume Positions,
21 October 1968
Figure D-7 Plan View of Horizontal Plume Positions,
22 October 1968
Figure D-8 Plan View of Horizontal Plume Positions,
23 October 1968
Figure D-9 Plan View of Horizontal Plume Positions,
24 October 1968
50
51
52
62
68
70
75
84
86
103
104
105
106
107
108
109
110
111
IX
-------
TABLES
Table I Keystone Precipitator Conditions during
Lidar Observations of Plume
(from Single Stack) ,
Table II Condensed Data Summary, Lidar Observations ,
Table III Values of Input Parameters Used and Results
from Plume Rise Calculations Using ASME
Formula for Stable Conditions ,
Table IV Summary of MRI Particle-Size Distribution
Data ,
Table V Values of Optical Parameters Computed
from Relative Particle-Size Data ,
Table A-I Characteristics of SRI Mark V Lidar. ...
Table B-I Data Summary, 25 May-1 June 1968 ,
Table B-II Data Summary, 15-24 October 1968
11
13
55
66
67
85
91
93
XI
-------
SYMBOLS
a = particle radius
A = effective receiver area
c = speed of light
C = particle number concentration; also parameter in formula for
Junge fit to particle-size data
C = optical cross section averaged over particle size interval
E = lidar site elevation
F = buoyancy flux
g = acceleration of gravity (= 9.8 m/s)
G = stability parameter
H(R) = magnitude of magnetic disk-recorded signal return from atmosphere
as a function of range
H = height of plume axis above lidar
H = height of top of stack above lidar
S
H, = height of plume top above lidar after maximum rise
AH = plume rise (height of axis of plume above stack top after reaching
equilibrium)
H = distance along the laser pulse propagation path
i = complex operator (y~-l)
k = slope parameter in equation for Junge fit to particle-size data
log* = logarithmic amplifier transfer function
m = refractive index of particles
M = plume mass concentration
M = background mass concentration in "clean" air
n = number of particles per unit volume per unit radius interval
n = relative density of particle-size distribution (fraction of
y*
number of particles per unit radius interval)
N = cumulative particle number concentration (number of particles
larger than a given radius)
P = detector output component due to background and internal noise
P = detector output component due to backscattered signal return
o
P = peak power emitted into the atmosphere
Xlll
-------
Q, = backscatter efficiency factor
b
Q = extinction efficiency factor
e
r = radius of stack orifice (= 4.25 m for Keystone)
R = range
R = reference range
AR = range increments of digitized lidar data
S = normalized, range-corrected signal return function
T = atmospheric transmission factor
T = ambient atmospheric temperature at stack top
a
T = beam convergence factor
T = stack gas exit temperature
T = air temperature at height of plume top
U = mean wind speed over height interval where plume rise occurs
U = wind speed at height of stack top (H )
8 S
U - wind speed at height of plume top (H ) after maximum rise
L ^
V = stack gas exit velocity
S
x = particle size parameter (= 2rr a/X)
P = volume backscatter coefficient (per steradian)
X = wavelength of the laser energy
M. = 10 m; also used generally as 10 ( )
^ = backscattering optical cross section per unit mass per steradian
(=
= total scattering (extinction) optical cross section per unit
mass (= a/M)
rr = 3.1416
p = density of the particulate matter
a = volume extinction coefficient
T = pulse duration
xiv
-------
FOREWORD
This work was carried out as part of the Large Power Plant Effluent
Study (LAPPES) conducted by the National Air Pollution Control
Administration, and was sponsored by the Division of Meteorology of
that agency under Contract PH 22-68-33.
We are grateful for the logistical and meteorological support and
experimental coordination provided by Frank Schiermeier, LAPPES Field
Manager, and by Larry Niemeyer and Charles Hosier of the Division of
Meteorology, National Air Pollution Control Administration.
We also acknowledge the able assistance of the following Stanford
Research Institute personnel: Hisao Shigeishi, Charles Brabant, and
Stephanie Briggs in the computer data processing and analysis phase;
Al Smith and John Alder in the manual analysis phase; and John Oblanas,
Norman Nielsen, and William Dyer in the experimental program. Thanks
are also due to Ron Collis, Director of the Aerophysics Laboratory,, for
his helpful suggestions in the writing of this report, and to William
Evans for the contribution of his ideas on data display and analysis
techniques .
xv
-------
FRONTISPIECE MARK V LIDAR IN OPERATION DURING
KEYSTONE STACK PLUME STUDY NEAR
SHELOCTA, PENNSYLVANIA
-------
I INTRODUCTION
The lidar, or laser radar, which was developed for meteorological
use by Stanford Research Institute in 1963, has important applications
in air pollution research and control activities (see Johnson, 1969;
*
Hamilton, 1966; Barrett and Bcn-Dov, 1967). This report covers a
recent experimental program, carried out for the National Air Pollution
Control Administration, involving mobile lidar observations of the diffu-
sion of the smoke plume from a 245-m stack at Keystone Generating Station
in western Pennsylvania. Such tall stacks are becoming increasingly
necessary in order to lower the ground concentrations of pollutants from
large plants. However, these stacks cost in excess of $3000 per meter
to build, and it is important to determine their efficiency, in practice,
in preventing high air-pollution levels. In addition, in order to refine
diffusion models used to predict pollutant concentrations from high
stacks, more data are needed. Most existing diffusion models were de-
veloped before the advent of the tall stack, and the accuracy of the
theoretical predictions for this application has not yet been adequately
verified due to the shortage of suitable data (Smith, 1968). The new
lidar technique promises to fill this need and facilitate advances in
the theory of high-stack diffusion.
Using a mobile, truck-mounted lidar system, a field team from
Stanford Research Institute participated in the LAPPES experimental
program during two periods: 25 May-1 June and 15-24 October 1968.
Lidar observations of the smoke plume from the stack of coal-burning
Keystone Generating Station were obtained on 17 clays; a total of 175
vertical plume cross sections, consisting of 3806 individual lidar ob-
servations, were collected. The quality of about one-third of these
data were degraded by poor weather conditions. Of the remainder,
*
References are listed at the end of the report.
Large Power Plant Effluent Study.
-------
approximately 10 percent were selected for detailed manual analysis (first
field program) and 46 percent were analyzed objectively using computer
techniques. Representative examples of these derived plume cross sec-
tions will be presented here to illustrate the significant information
which lidar observations can furnish regarding plume geometry and be-
havior. Although particulate emission is not the major air pollution
problem associated with power plants because of the efficacy of modern
ash removal devices, the smoke serves as a convenient tracer for the
study of diffusion processes. Since the particles are small, their
diffusion should be essentially identical with that of a gas such as SO .
-------
II SUMMARY AND CONCLUSIONS
The results of the Keystone plume study with regard to plume be-
havior in particular, and lidar capabilities and limitations in general,
are briefly itemized below, with conclusions included as appropriate.
A. Plume Characteristics
In general, the lidar observations confirmed and underscored the
known and suspected complexities of plume behavior. Complicating fac-
tors--such as those mentioned below--thnt have been largely ignored in
present theories should now be intensively studied for possible incor-
poration, if the theory is to stay abreast of the experimental state of
the art .
Plume tilting and fanning caused by vertical wind direction shear
probably affects plume rise, and definitely requires reconsideration of
the definition of plume rise.
Fanning is a very common feature of plumes from high stacks ob-
served under stable conditions and must be included in any realistic
diffusion theory.
The marked effects of elevated inversions and other levels of in-
creasing stability with height were observed frequently and need thorough
consideration. Close correspondence was found between such levels and
plume tops, as well as with haze tops when no plume was observed.
Plume fumigation occurs often, probably more often than previously
suspected. Typically, however, only a portion of the plume fumigates at
a given time, with the remainder tilted and elevated.
The meteorological regime best suited for model validation, namely
nonstable conditions, also presents the most difficult observational
situation for lidar. This is because of the diffuse, rapidly changing
nature of plumes under these conditions. (See the instrumentation dis-
cussion later in this section.)
-------
B. Lidar Data Analysis Considerations
The value of lidar for observing plumes was clearly established.
For many applications, such as the use of particulates in a plume as
tracers for the study of diffusion processes, lidar-observed relative
mass concentrations, uncorrected for attenuation, should be sufficient
to furnish the required information.
However, quantitative results in terms of absolute mass concentra-
tion were also obtained. A computed sample cross section corrected for
attenuation gave a cross-axial integrated mass concentration of 680 g/m
of plume length, compared with 875 g/m as calculated from the power
plant data and wind measurements.
Our method of calculation of absolute particulate mass concentra-
tions from the lidar data is based on the following:
(1) The same particulate relative size distribution and
optical properties are assumed to hold in the clean
air and plume. (The MRI data appear to confirm
this.) This is a convenient but not essential
assumption for obtaining a solution.
(2) The relative size distribution and optical properties
are assumed to be known (measured or estimated) .
(3) Mie scattering theory is assumed to be valid (this is
reasonable, since the fly ash particles are spherical).
Thus 5 , the extinction optical cross section per unit
mass, can be computed.
(4) The clean air outside the plume is assumed to be homo-
geneous and to have a known mass concentration M
o
(measured or estimated).
The integration procedure used in this method is fairly sensitive
to errors in the values used for the solution parameters P and M .
H ^e o
The known errors and limitations associated with the use of
collecting-type particle samplers tends to make such data incompatible
with lidar data. Optical particle samplers might give more comparable
-------
results, although there could be difficulties with any direct-sampling
instrument that disturbs the medium.
The best prospects for obtaining routine quantitative data seem to
lie in the direction of deriving additional information on the solution
parameters by means of advanced lidar techniques, such as multiple-
wavelength instruments, which may be able to eliminate the need for in-
dependent particle measurements.
C . Instrumentation Considerations
For plume cross sections built up from many individual lidar obser-
vations, an accurate shot-to-shot referencing system is essential to
correct the data for pulse-to-pulse variations in transmitted power.
Higher pulse repetition rates, on the order of one to ten shots per
second, are needed to observe rapidly changing plumes under unstable
conditions adequately and faithfully. A water-cooled system is suggested
by this requirement .
Better data handling, processing, and display capabilities are
necessary to assimilate larger quantities of data properly. New tech-
niques, ideally of the real-time analog type, are needed to provide a
capability of averaging many cross sections to obtain an hourly mean
concentration distribution with which to check diffusion theories.
The usefulness of lidar data would be considerably enhanced if in-
strumentation with higher power output and lower noise susceptibility,
and thus greater range, were used. A lidar with an effective above-
noise range (for the clear-air return) of about 8 km would be ideal for
plume studies. The neodymium system used in this study has an effective
*
range for clear-air return of about one to two km, and thus could ob-
itt
serve only the denser parts of "clean" plumes at ranges greater than
*
This range is variable, depending upon the level of background
illumination.
When all electrostatic precipitators were operating.
-------
this. The Mark V ruby lidar would give much better results in this
respect due to its shorter wavelength, narrower spectral region, and
greater receiver sensitivity, but the pulse repetition rate needs to
be increased.
In this connection, it is encouraging to note that neither the ruby
nor neodymium versions of the SRI Mark V lidar approach the limits of
current technology, and systems with the recommended performance can
readily be developed.
-------
Ill DESCRIPTION OF THE EXPERIMENTAL PROGRAM
A . Instrument at ion
The SRI Mark V lidar (Fig. 1) was used during both field programs
in a mobile, truck-mounted configuration. The characteristics of this
instrument are detailed in Appendix A. As are all lidars, the Mark V
is basically composed of a laser transmitter, which emits a very brief,
high-intensity pulse of coherent monochromatic light, and a receiver,
which detects the energy at that wavelength backscattered from the at-
mospheric aerosol as a function of range.
The transmitter of this instrument employs either a neodymium-doped
glass laser rod or a ruby rod, at wavelengths of 1.06 and 0.694 p. re-
spectively. The neodymium laser was used for the plume study because
of its faster pulse repetition rate (approximately 10 pulses/min vs
3 pulses/min for the ruby). However, due to the much lower quantum
efficiencies and higher dark currents of the best available photomulti-
plier tubes at 1.06 u, and the wider filter bandpass needed for this
system, overall receiver sensitivity, signal-to-noise ratio, and thus
range is significantly reduced over that obtainable with the ruby system.
*
The lidar data were recorded on both a video magnetic disk recorder
and on an oscilloscope equipped with a 35-mm recording camera. A second
oscilloscope used in conjunction with the disk recorder permitted an
"instant replay" of each lidar shot, thereby providing real-time moni-
toring of the data. A 3-MHz low-pass filter is used with the disk re-
t
corder to minimize noise; the range resolution is about 40 m, while that
for the 35-mm film record is less than 20 m. The 35-mm film records were
used for the manual analyses, while the machine analyses employed the
magnetic disk records.
MRV Corp. Model 100M.
+
The range resolution as used here is defined as the minimum separation
distance at which two individual targets can be resolved.
-------
'--:-- >- r
FIGURE 1 SRI MARK V LIDAR
-------
B. Experimental Procedures
For each series of observations, the lidar was situated at any one
of a number of preselected points, depending upon the location of the
plume. The sites used during the two observation periods are depicted
in Fig. 2. Where possible, the lidar was positioned so that it could be
aimed in a direction approximately perpendicular to the plume axis. At
each site, vertical cross sections through the plume were normally ob-
*
tained at several downwind distances from the stack. The technique
employed consisted of scanning vertically with elevation angle incre-
ments of 0.3° to 10°, depending upon the vertical extent of the plume
as determined from preliminary observations. Such scans were then re-
peated at other azimuth angles. The neodymium lidar permitted observa-
tions at intervals as short as five seconds. A complete vertical scan
of 20 to 30 lidar shots required about two to five minutes to complete,
and a set of three cross sections could be obtained within 10 to 15
minutes.
When the lidar was located at a considerable distance downwind from
the stack, such as at Sites 16 (9 km), 17 (14 km), and 21 (21 km), the
vertical cross sections were limited to only one azimuth angle (i.e.,
one vertical plane), because of the extended width of the plumes and
the frequently directly overhead plumes.
In order to make the stack plume visible for direct observations
with aircraft associated with other experiments, 50 percent of the
electrostatic precipitators at Keystone Generating Station were normally
shut down during most of the observations. However, to test the capa-
bility of the neodymium lidar to observe the "clean" plume, all of the
precipitators were put into operation during certain periods on 21 and
22 October. As shown in Table I, six lidar plume cross sections were
obtained under these conditions, and the results are presented in Sec. V.
*
Although the Keystone Generating Station has two stacks, each 245 m
high, only one stack was in operation during each of the observation
periods.
-------
543210
TA-72B9-8
FIGURE 2 LIDAR SITES USED DURING FIELD EXPERIMENTS
-------
Table I
KEYSTONE PRECIPITATOR CONDITIONS DURING LIDAR
OBSERVATIONS OF PLUME (FROM SINGLE STACK)
(Precipitators on indicated by X)
Date
(1968)
25 May
26 May
15 October
16 October
17 October
18 October
20 October
21 October
22 October
23 October
24 October
Lidar
Cross Section
Numbers
All
All
All
All
All
All
All
j 46, 47, 48
(All Others
j 60, 61, 62
' All Others
All
All
Bank 1
Inlet
1
X
X
X
X
X
X
X
X
X
X
X
X
X
2
X
X
X
X
X
X
X
X
X
X
X
X
X
3
X
X
X
X
X
X
X
X
X
X
X
X
X
4
X
X
X
X
X
X
X
X
X
X
X
X
X
Outlet
1
X
X
X
X
X
X
X
X
X
X
X
X
X
2 3
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
4
X
X
X
X
X
X
X
X
X
X
X
X
X
Bank 2
Inlet
1
X
X
X
2
X
X
3
X
X
4
X
X
Outlet
1
X
X
X
2
X
X
3
X
X
4
X
X
-------
The lidar observation periods were limited in length by prior
arrangement in order to forestall any possible interference between the
lidar and other experiments utilizing aircraft. One of the considera-
tions here was lidar eye-safety, which is discussed in Appendix E.
C. Summary of Lidar Data
The lidar observations obtained during the two experimental periods,
25 May-1 June 1968 and 15-24 October 1968, are summarized in detail in
Appendix B, and in condensed form in Table II, along with the distribu-
tion of the analytical effort among the data. Data obtained during
several days during the May period when there was rain, low ceilings or
fog have been omitted, since no analysis was attempted. Data obtained
on occasions when the cooling tower plume merged with the smoke plume
also were not analyzed .
12
-------
Table II
CONDENSED DATA SUMMARY, LIDAR OBSERVATIONS
(For detailed summaries, see Appendix B)
to
Date
(1968)
25 May
26 May
15 October
16 October
17 October
18 October
20 October
21 October
22 October
23 October
24 October
Time
(EOT)
0912-0925
1034-1048
1423-1453
0854-0926
1145-1208
0932-1007
1032-1150
0852-0946
1149-1206
0903-0942
1151-1216
0917-0944
0909-0941
1018-1040
0854-0935
1023-1036
0900-0945
1134-1201
1227-1245
0850-0918
0941-0949
1155-1227
0930-1204
Site
Number
1
2
6
16
16
17
21
22
6
6
11
Elevation
(m)
329
387
394
418
418
448
448
448
394
394
418
Location
from Stack
3.5 km SW
(Az. 227°)
3.5 km N
(Az. 002°)
3.75 km NE
(Az. 055°)
8.50 km N
(Az. 355°)
8.50 km N
(Az. 355°)
13.7 km NNW
(Az. 344°)
21 .1 km ESE
(Az. 104°)
4 . 90 km SE
(Az. 120°)
3.75 km NE
(Az. 055°)
3.75 km NE
(Az. 055°)
3.95 km NW
(Az. 306°)
Number of
Cross Sections
Observed
3
3
4
6
5
6
8
6
5
5
3
4
4
2
11
3
9
6
6
7
2
5
19
Analyzed
3
3
0
3
2
3
5
2
0
2
0
2
2
1
6
3
9
0
3
3
0
0
12
Analysis
Method*
M, C
M
—
M
M
C
C
C
—
over
C
—
C
C
C
C
C
C
—
C
C
--
—
C
Remarks
Spatial
series
Spatial
series
Time and
spatial
series
Plume
almost
overhead
Plume
overhead
Plume
overhead
Plume
overhead
Dirty vs
clean
plume
Dirty vs
clean
plume
Time and
spatial
series
Time
series
M—Manual Analysis, C—Computer Analysis.
-------
IV BASIC ANALYSIS OF LIDAR BACKSCATTER SIGNATURES
A. General Discussion
An example of an oscilloscope photograph of a typical lidar shot
recorded on the magnetic disk is given in Fig. 3. Here the magnitude
of the backscattered signal return from the airborne aerosol is log-
amplified and displayed on a logarithmic scale (in order to maintain a
wide dynamic range of recording) versus time (or range) on a linear
scale.
RANGE — km
0 0.3 0.6 0.9 1.2 1.5 1.8 2.1 2 2 2.7
I I I I I 1 I I I
O
o
1 1 1
TRANSMITTER
POWER
REFERENCE
PULSE
H
^x^
^\
SIGNAL
RETURN
FROM
PLUME
^"^ c« i /^ m
h
\ \ n IT T i
NEAR FIELD
CONVERGENCE
POINT OF
TRANSMITTER
AND RECEIVER
SIGNAL RETURN
FROM BACKGROUND
AEROSOL IN "CLEAN"AIR
TA-7289-9
FIGURE 3 BASIC LIDAR OBSERVATION OF SMOKE PLUME
For inspection, inventory, and editing purposes, ten lidar observa-
tions are displayed simultaneously by the magnetic disk recorder, as
shown in Fig. 4. One complete vertical "slice," or cross section,
through the plume is represented by two to four of these ten-trace
15
-------
0 0.3 0.6 0.9 1.2 1.5 1.8 2.1 2.4 2.7 3.0
Elevation
Angle
TA-7289-10
FIGURE 4 EXAMPLE OF MULTIPLE-TRACE FORMAT USED
FOR EXAMINATION OF MAGNETIC
DISK RECORDS
photographs. The "artificial" peak in the received signal at close range
*
in Figs. 3 and 4 is typical of a bistatic system, and is due to the
oppositely acting effects of signal increase with increasing degree of
intersection of the transmitter beam and receiver field of view, and
signal drop-off due to increasing range (a function of the square of the
range) .
These traces represent the basic raw lidar data, which can be used
directly to obtain plume boundaries easily and accurately. However,
meaningful quantitative information on the concentration distribution
within the plume requires further reduction and analysis. For each plume
cross section consisting of 20 to 40 lidar observations, the data must
be adjusted and corrected for:
(1) Range effect,
(2) Log-amplifier transfer function, and
(3) Pulse-to-pulse variations in transmitted power.
Current lidars have separate, noncoaxial transmitter and receiver optics
16
-------
Allowance for these factors is fairly straightf orward, and results in a
cross section of normalized signal return (S) values. These S-values
are directly proportional to particulate mass concentrations (expressed
in logarithmic units) , if the particle size distribution and optical
properties are spatially uniform, as is often a reasonable assumption
(McCormick and Kurfis, 1966), and if attenuation is negligible. Correc-
tion for the attenuation effect requires considerable effort, and is
detailed in Sec. VII.
Manual analysis was initially employed for the May data. Using a
specially prepared gridded template, range-corrected values of signal
return (in dB, relative to the background signal level from the ambient
aerosol) were extracted from the 35-mm film record. These values were
then corrected for the shot-to-shot variations in transmitted power,
plotted, and the resulting cross sections analyzed by hand at a contour
interval of 6 dB (representing an increase by a factor of four) . In
line with the preceding discussion, these S-contours represent (in
logarithmic units) approximate relative mass concentrations within the
plume .
These manual analyses are quite tedious, and an objective analysis
procedure was developed to process the October data and to check the
hand analyses of the May observations. This procedure is documented in
Fig. 5 (and Appendix C) and is discussed below.
B . Lidar S-Function Analysis
The signal received on the face of the Mark V detector, originating
from backscattered energy at range R, may be expressed as
P (R) = P^ ^ -^ T (R) p(R) T2(R) (1)
where
^
s T 2 R2 c
P = peak power emitted into the atmosphere
c - speed of light
T = pulse duration
17
-------
STEP 1—ANALOG-TO-DIGITAL CONVERSION
I
Field
Instrumentation
Electronics
Oscilloscope
Camera
Manual Curve
Tracing
Calma Model
300 Digitizer
ONE SET
OF LIDAR
OBSERVATIONS
r
i
VIDEO
MAGNETIC
DISK RECORD
—
r
i
35 mm
FILM
1
1
DIGITIZED
DATA
(AR = 3.8 m)
r
i
-»-
DATA IN
ENGINEERING
UNITS
CDC6400
Computer
STEP 2—CONDITIONING, EDITING, AND BASIC ANALYSIS
DATA 1 N
ENGINEERING
UNITS
T
1
CONDITIONED
AND EDITED DATA
(AR = 15.2m)
r
i
NORMALIZED SIGNAL RETURN (S)
VALUES Data Corrected for Range,
Pulse-to-Pulse Power Variations,
and Log Amplifier Characteristics
1 1
1 1
CDC6400 CDC6400
STEP 3—CONTOUR ANALYSIS AND ATTENUATION STUDY
S-VALUES
PARTICLE
SIZE DATA
.Selecte
d Cases
1 "
* '
i
CDC6400
MASS
CONCENTRATION
VALUES
I'
1
1
1
) „
p
i
CONTOUR PLOTS OF PLUME
CROSS SECTIONS
S Units
CONTOUR PLOTS OF PLUME
CROSS SECTIONS
Mass Concentration Units
CDC3300, 6400
CDC280 TV Graphs
Calcomp Plotter
SRI Contouring Program in
Polar Coordinates
TA-7289-11
FIGURE 5 LIDAR DATA ANALYSIS PROCESS.
on Steps 1 and 2.
See Appendix C for additional details
18
-------
A = effective receiver area
T = beam convergence factor
3 = volume backscatter coefficient (per steradian)
T = atmospheric transmission factor.
As previously mentioned, the resulting detector signal was routed through
a logarithmic (log) amplifier and recorded on a video magnetic disk, as
well as on a backup 35 -mm recording camera system.
All signal strengths were referenced to the detector level in the
absence of the applied backscatter signal P , which is denoted by P .
The atmospheric return level recorded on the disk may then be expressed
(in decibel notation) as
P (R) + P
H(R) = 10 log* - - - - (2)
N
where log* is the log amplifier transfer function, which differs slightly
from the true log function. As indicated in Fig. 5 and detailed in
Appendix C, the analysis phase was initiated by digitizing recorded H(R)
traces, using a CALMA Model 300 digitizer. Intensity values at the
digitizer standard increment (equivalent to 4 m) were recorded on mag-
netic tape. A CDC6400 computer was then employed to obtain a range-
corrected, normalized signal return (S) function defined as
P (R) R2
S(R) = 10 log -5- — - - - (3)
P R R
s \ o I o
where R is a reference range [normally chosen as the first data point
o
of the H(R) digitization], and P » P . This S function also incor-
S JN
porates a correction for the transfer function of the log amplifier.
From Eqs. (1) and (3), the normalized signal return function S(R)
may be related to the atmospheric optical parameters by
6(10 = 10 10. , (4,
e (Ro) T W
19
-------
where T (R > R ) = 1 is assumed; i.e.. the H(R) digitization process
c o
was initiated at a point on the trace for which the transmitted beam is
fully encompassed by the diverging receiver field of view. The S func-
tion then represents a relative backscatter function, which is uncorrected
for atmospheric transmission losses.
S-function contours were generated from a computer program (developed
at Stanford Research Institute) that uses linear interpolation of data
gridded in polar coordinates. The polar grid of data points need not be
equal-incremented in range or angle, thus permitting processing of lidar
data collected at unequal increments of elevation angle. The output from
the contouring program was subsequently machine plotted on a Benson-
Lehner or Calcomp plotter, or electronically graphed on the CDC280 CRT/
film system. These contour plots and their significance are presented
in Sec. V.
The lidar S function is dependent only on the atmospheric optical
properties when P (R) » P,T. However, when P (R) ^ ?„ (i.e., when the
s N ' s N _
atmospheric return is at or below the noise level), then S(R) °c 10 log R .
Invalid data of this type were omitted from the contouring program. This
omission, illustrated in Fig. 6, explains the open contours that may be
expected when plume return occurs at a range for which the clear-air re-
turn would be below the noise level. While the S-function contour
analyses give a surprisingly good picture of plume relative concentration
distributions from laser energy returns (see Fig. 23), they are biased
to varying degrees by the fact that some energy is lost from the laser
pulse as a function of distance along the path, i.e., the atmospheric
transmission of the laser energy varies along the pulse path. Accounting
for this attenuation effect requires additional information on the atmo-
spheric scattering particles and/or an explicit relationship between the
atmospheric transmission and backscatter properties at the laser wave-
length. This aspect of the analysis is discussed further in Sec. VII.
Since plume mass concentrations—and thus lidar echoes—cover such
a wide dynamic range, contouring a plume cross section in linear units
presents a problem. As illustrated in Fig. 7, the linear display (a)
20
-------
RAW LIDAR DATA
BEAM CONVERGENCE
AREA
CO
I-
_l
O
<\i o
I
0_
\
o:
a:
o
o
g
CO
LIDAR
SIGNAL
PLUME
RETURN
BACKGROUND
REFERENCE
RANGE (Ro)
V
SIGNAL RETURN IN THE
ABSENCE OF NOISE
RANGE
UNKNOWN SCATTERING
REGION
CORRECTED FOR I/R2 SIGNAL DECREASE WITH RANGE
AND LOGARITHMIC AMPLIFIER TRANSFER FUNCTION
FLAT UNTIL
NOISE LEVEL
IS REACHED
\
CLOSED
CONTOURS
OPEN
CONTOURS'
BACKGROUND
NOISE LEVEL
RANGE
REFERENCE
RANGE (Ro)
TA-7289-12
FIGURE 6 EFFECT OF RANGE CORRECTION UPON PLUME SIGNAL RETURN AND RECEIVER
NOISE LEVEL. For simplicity, a perfectly transmitting homogeneous medium has been
assumed for the clear air.
21
-------
600
DC
<
Q
o
CD
400
200
Cross section No. 64
0940:35-0942:00 22 Oct 68
Azimuth 300 deg mag
J.
_L
_L
_L
_L
1.2
600
1.4 1.6 1.8
HORIZONTAL RANGE FROM LIDAR — km
(a) Linear display—M/MQ—contour interval 25
—I 1 1 1 1 1
2.2
Cross section No. 64
0940:35-0942:00 22 Oct 68
Azimuth 300 deg mag
400
IT
<
Q
O
CO
200
_L
J_
_L
_L
1.2
1.4 1.6 1.8
HORIZONTAL RANGE FROM LIDAR
2.0
km
2.2
(b) Logarithmic display—10 Iog10 (M/M0)— contour interval 3 db
FIGURE 7 COMPARISON OF PLUME CROSS SECTIONS CONTOURED
IN LINEAR UNITS (a) AND LOGARITHMIC UNITS (b)
TA-7289-13
22
-------
loses definition at the boundary of the plume, compared with a logarithmic
display (b) with contours in decibels. The latter documents the plume
edge more accurately, and permits reasonable contour spacing.
In the linear display in the figure, the contour units are in mul-
tiples of the clean-air background mass concentration (M ). or M/M ,
o ' o'
where M is the plume mass concentration. [For this case, M was taken
3 °
to be 100 ng/m , on the basis of the visibility estimate and the work of
Noll, et al. (1968).] It may be seen that the contour spacing, and thus
the outer contour value, must be set at no lower than about 25 in order
to prevent contour crowding. On the other hand, the logarithmic display
with units of dB = 10 log (M/M ) has a contour spacing of 3 dB, which
gives an outer boundary of M/M ~ 2 without excessive contour crowding.
o
(Each contour increment of 3 dB corresponds to a change by a factor of
approximately two from the adjacent contours.) Because of these ad-
vantages, logarithmic contour intervals have been used exclusively in
the lidar cross sections.
Because of the effect of the finite above-noise range of the lidar
as explained in Fig. 6, many contours in the cross sections to follow
later do not close around the plume; with increasing range they begin
contouring the background noise and become meaningless. In addition,
contours frequently appear in the clear air beneath the plume, which
are usually probably real, but have no particular value. Both of these
effects are shown in Fig. 7(b). To add clarity to the plume cross sec-
tions, these extraneous portions of the contours have been deleted from
the illustrations. Figure 33 in Sec. VIII shows the nature of this
editing process. Considerable care was taken to ensure that this process
did not remove any portion of the plume cross section or any other in-
formation of value.
23
-------
V RELATIVE CONCENTRATION CROSS SECTIONS OF KEYSTONE STACK PLUME
A. May 1968 Observation Period
The emphasis during the May observation period was on lidar plume
observations within five miles (8 km) of the stack. The analyses for
the May data were performed manually. All lidar observations during
this period were taken when one of the two precipitator banks for the
active stack was shut down, allowing 50 percent of the effluent to pass
out through the stack untreated. The resulting plume color was a light
gray-brown.
In the plume cross section illustrations to follow, both for the
May and October data, it should be noted that the lidar is at the point
(0,0), and the horizontal range scale is usually discontinuous. Also,
heights are taken above the lidar; Fig. 8 illustrates the height rela-
tionships involved. To facilitate height transformations, the height of
the top of the stack is indicated on the ordinate scale for each cross
section by means of a horizontal arrow. (Lidar site elevations are
given in Table II, Sec. III.) Calculated plume rise heights (AH) are
also shown for those cross sections for which reasonably concurrent
temperature profiles were available, and will be discussed in Sec. VI.
Figures 9-12 show four sets of sequential vertical cross sections
through the smoke plume during the mornings of 25 May (Figs. 9 and 10) and
26 May (Figs. 11 and 12). The contours in Figs. 9 through 12 represent
range-corrected signal return in decibels relative to that from the ambient
background aerosol, including background noise, and can be taken to repre-
sent (approximately) the relative particulate concentrations. The effects
of attenuation are neglected. The inset in each figure indicates the
geometry of the observations. Poor weather conditions prevailed during
the 43 cross sections obtained during the remainder of the May period.
The cross sections in Fig. 9 were obtained about 1-1/2 hours earlier than
those in Fig. 10, while the observations in Fig. 11 precede those in Fig.
12 by almost three hours. Vertical temperature profiles closest in time
25
-------
KEYSTONE
STACK
245 m
Y////////////////.
H_ 570-E
LIDAR SITE t ELEVATION
570 m
325 m
SEA LEVEL
TA-7289-14
FIGURE 8 RELATIVE HEIGHT CONSIDERATIONS ASSOCIATED WITH LIDAR PLUME
OBSERVATIONS. Lidar site elevations, E, are given in Table II; values of HS
are indicated on each plume cross section.
-------
bO
1000
800
tr
Q 600
Ij
LU
O 400
I-
X
2 200
LU
I
LIDaR SITE NO I
PLAN VIEW OF HORIZONTAL PLUME POSITION
25 MM I968
0912-0926 EOT
CROSS SECTION MO
AZIMUTH 000° M
AH (calc)_
= 206 m
CROSS secrtON NO
0912 30 - 0916 00 EOT
0.8
1.0
1.2
1.4
1.6
1.8
2.0
-! f-
2.2 ' 0.8 1.0
RANGE km
1.2
1.4
1.6
2.0
2.2
1000
800
600
400
200
2.4 2.6
0.2
0.4
0.6
0.8
1.0 1.17 0 0.2 0.4
PLUME CROSS-WIND DISTANCE km
0.6 0.7
0.1
0.2 0.23
TA-7289-1 5
FIGURE 9 SPATIAL SERIES OF LIDAR-OBSERVED VERTICAL CROSS SECTIONS UNDER STABLE CONDITIONS, 25 MAY 1968,
FROM SITE 1, 3.5km FROM STACK. The contours represent range-corrected signal return in decibels relative to that from
the ambient background aerosol, including background noise, and can be taken to approximately represent relative particulate
concentrations. Attenuation has been neglected. See inset for horizontal positions of cross sections.
-------
1000
to
00
CROSS SECTION NO 6
1045.00-1048:00 EOT
SIS* M
CROSS SECTION NO 4
1034.00-103945 EOT
000 • M
4
0 1/2 I km
LIDAfi SITE NO I
PLAN VIEW OF HORIZONTAL PLUME POSITION
CONVERSION TftBL£
dB
0
6
12
18
24
30
MULTIPLE
BACKGROUND
|
4
16
63
251
IOOO
OF
RETURN
25 MAY 1968
1034-1048 EOT
AH (calc)
= 251 m
2.8 3.0 3.2 3.4 3.6 1.0 1.2 1.4
2.0 2.2 22 2.4 2.6 2.8 3.0 3.2
1.8 2.0 2.2 2.4
1.2 1.4 1.5 0 0.2 0.4 0.6 0.8 1.0 1.16 0
PLUME CROSS-WIND DISTANCE — km
0.2
1000
80O
600
400
200
0.4 0.48
TA-7289-16
FIGURE 10 SPATIAL SERIES OF LIDAR-OBSERVED VERTICAL CROSS SECTIONS UNDER MODERATE DIFFUSION CONDITIONS,
25 MAY 1968, FROM SITE 1, 3.5km SOUTHWEST OF STACK
-------
800
tr
< 600
Q
400
o
CO
200
X
(D
STACK \X^-/
\ ^ Xx CROSS SECTION NO 12
INFERRED N{\ X 0858.00-0902 40 EOT
/ \ \
HORIZONTAL
PROJECTION
OF PLUME
BOUNDARIES-.^ \
226' M
/
CROSS SECTION MO 14
0908 43-0913-00 EOT
253-M
LIOAR SITE NO 2
PLAN VIEW OF HORIZONTAL PLUME POSITION
CONVERSION TABLE
-------
00
o
1800
1600
1400
CE
<
Q
1200
1000
o
m
I
(3
UJ
X
800
600
400
200
INFERRED HORIZONTAL
PROJECTION OF PLUME
BOUNDARIES
NFERRED AXIS OF
BISECTED PLUME
/Z26' M
PLAN VIEW OF HORIZONTAL PLUME POSITION
LIDAR SITE N02
CROSS-SECTION NO 17
II45'30-II48'30 EOT
AZIMUTH 200' M
-TOP OF
STAXK
3.4
3.6
3.8
4.0
26 MAY I968
II4S-II5S EOT
4.2 •> 0 " 2.8
RANGE — km
0.2
0.4
3.0
3.2
3.4
_|
3.6
3.8
I
4.0
8OO
1600
1400
1200
1000
8OO
600
400
200
4.2
0.57 0 0.2 0.4
PLUME CROSS-WIND DISTANCE — km
0.6
0.8
1.0
1.2 1.27
TA-7289-18
FIGURE 12 SPATIAL SERIES OF LIDAR-OBSERVED VERTICAL CROSS SECTIONS UNSTABLE (LOOPING) CONDITIONS,
26 MAY 1968, FROM SITE 2, 3.5km NORTH OF STACK
-------
to the lidar cross sections are given in Fig. 13. On both mornings, the
sky was essentially clear.
1000
E 800 -
o
I-
10
,. 600
UJ
en
<
CD
UJ
> 400
m
I 200
26 MAY
I 40 EOT
TOP OF
STACK
— 25 MAY-*
0946 EOT
10
FIGURE 13
12 14 16 18
TEMPERATURE —°C
20
22
TA-7289-19
VERTICAL TEMPERATURE
PROFILES ASSOCIATED
WITH LIDAR OBSERVATIONS,
25-25 MAY 1968
Several general features of these cross sections are apparent. The
contours, which are in logarithmic units (decibels), represent approxi-
mate relative particulate mass concentrations as previously discussed.
Maximum concentrations are usually observed near the top of the plume,
apparently associated with the greater buoyancy there. This effect was
also found by Hamilton (1966), who took ruby-lidar observations of the
plume from an oil-fired power plant. There is considerably consistency
in the observed shape and structure of the Keystone plume from cross
section to cross section in each set, and individual features may be
followed for several kilometers downwind from the stack. This is par-
ticularly pronounced for the early-morning, nondiffusive cases. For
31
-------
example, in Fig. 9, the central clear cavity, or vault, on the bottom
of the plume, evidently associated with the bifurcation process, is
readily apparent in all three cross sections.
For the stable cases (Figs. 9 and 11), the characteristic horizontal
spreading (fanning) and tilting of the plume due to vertical wind direc-
tion shear (wind veering with height) is evident.
In Fig. 11, there is some evidence that terrain channeling of the
air flow acted to horizontally spread the effluent, in addition to the
effects of wind direction shear. The northeastern portion of the plume
(nearest the lidar) appears to follow a valley that is about 100 m lower
than the hills which are traversed by the southwestern portion of the
plume (farthest from the lidar), possibly accounting for part of the
*
rapid increase in the cross-wind (or cross-axis) dimension of the plume
from 340 m at 0.8 km downwind to 1450 m at a downwind distance of 2.2 km.
Lifting by the hills could explain the 160-m rise of the southwestern
portion of the plume (at the extreme right in Fig. 11) from Cross Section
12 to Cross Section 14.
The cross sections pictured in Fig. 10 probably represent a case of
fumigation, in that downward diffusion is fairly strong, while upward
diffusion of the plume apparently is limited. This situation could be
caused by a temperature inversion at 900-1000 m, above the maximum height
of the observed temperature profile. Note that relatively high concen-
trations are brought to very low levels within 3 km from the stack.
The rapid upward diffusion of a looping plume is illustrated in
Fig. 12. For the most part, the relative concentrations in Cross Section
19 are less than those in Cross Section 17 by about 6 dB, which repre-
sents a decrease to about one quarter of the previous levels . The tem-
perature sounding for this time (Fig. 13) shows an approximately dry-
ad iabatic lapse rate up to the maximum height sampled, 1000 m above the
base of the stack.
*
This dimension is indicated on the secondary abscissa scale for each
cross section during the May period.
32
-------
Since these cross sections were built up from lidar observations
spanning a period of two to five minutes, some distortion is unavoidable
when the plume structure is changing with time, as during unstable con-
ditions (Figs. 10 and 12). In the future, lidar systems capable of one
or two pulses per second and with automatic positioning should eliminate
such problems by furnishing essentially instantaneous plume "slices."
B. October 1968 Observation Period
Due to the large quantity of lidar data obtained during this period,
an objective, computer-oriented data reduction and analysis process was
developed, as described in Sec. IV. This process permitted the analysis
of about half of the cross sections observed. A number of illustrative
examples have been selected for presentation here. There are various
minor flaws in these analyses that would not have occurred if the work
had been carried out by hand, but the time saving is quite significant.
For example, occasional small areas of the cross sections (indicated
in the figures) were not contoured due to a minor program error. In
addition, the cross sections illustrated do not always represent the
total plume width as recorded on the original data records (photographs),
although the portion shown is correct. This truncation occurred because
of a problem in digitizing and processing those photographs that were
obtained with an oscilloscope sweep speed of 5 (as/cm, corresponding to
a full-scale range of 7.5 km.
Because of this difficulty, which could not be overcome within the
time and effort limitations of the project, the photographs taken at
2 |_is/cm were used exclusively, which limits the ranges of the analyzed
data to 3-km maximum. In order to retain the information on total hori-
zontal plume extent, maximum distances of plume return were scaled off
the original 35-mm film and are indicated on the figures by vertical
arrows (horizontal arrows when off scale) labeled with the range to the
farthest observed edge of the plume.
Several of the May (hand-prepared) cross sections were reanalyzed
by machine for comparison purposes. An example is given in Fig. 14.
The agreement is good, despite the fact that two different data records
33
-------
800
600
Q
_J
UJ
O
m
400
111
I
200
—I 1 1 |
MANUAL DATA REDUCTION
AND CONTOURING
FROM 35-mm FILM
T/T
MACHINE DATA REDUCTION
AND CONTOURING
FROM MAGNETIC DISK
Cross Section No. 1
0912:30-0916:00 25 May 68
Azimuth 045 deg mag
2.0
2.2
2.4 2.6 "2.0 2.2
HORIZONTAL RANGE FROM LIDAR — km
2.4
2.6
TB-7289-20
FIGURE 14 COMPARISON OF MANUAL AND MACHINE ANALYSES. Note that
two different records are used, with the 35-mm film having considerably
better spatial resolution than the magnetic disk. One lidar observation
is missing at the top of the machine cross section.
were used: the 35-mm film (spatial resolution about 20 m), and the mag-
netic disk (spatial resolution about 40 m). In addition, it was later
discovered that some data (the uppermost lidar observation) was missing
from the disk record; this caused the shape of the machine cross sec-
tion to differ at the top of the plume from that in the manual analysis.
The cross sections in Fig. 15 are in the same (approximately cross-
wind) vertical plane at the same azimuth angle from the lidar, and are
spaced about one hour apart; thus, they document the changing crosswind
structure of the plume with time during the morning of 15 October. The
direction to the stack is into the paper. (See Appendix D for plan views
of the horizontal plume positions, to clarify visualization of the lidar/
stack/plume orientation.) The typical tilting and fanning of the plume
in stable conditions, due to veering (clockwise rotation) of the wind
direction with height, is apparent. By 1147 EDT, the height of the
mixed layer had apparently increased to the level of the lower part of
the plume, causing fumigation of that portion. Unfortunately, no tem-
perature profiles are available for this day.
34
-------
E 800
600
> 400
00
I 200
(3
Cross Section No. 2
0939:00-0943:11 15 Oct 68
Azimuth 300 deg mag
Top of Stack
0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4
HORIZONTAL RANGE FROM LIDAR — km
2.6
2.8
3.0
E 800
600
400
O
m
200
• Top of Stack
Cross Section No. 8
1036:30-1041:00 15 Oct 68
Azimuth 300 deg mag
0.4 0.6 0.8
1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4
HORIZONTAL RANGE FROM LIDAR — km
2.6
2.8
3.0
E 800
Cross Section No. 14
1147:00-1150:00 15 Oct 68
Azimuth 300 deg mag
1.2 1.4 1.6 1.8 2.0 2.2
HORIZONTAL RANGE FROM LIDAR — km
2.6
2.8 3.0
TB-7289-52
FIGURE 15
TIME SERIES OF LIDAR-OBSERVED VERTICAL CROSS SECTIONS FROM SITE 6, 3.8km
NORTHEAST OF STACK, PLUME OVERHEAD, 15 OCTOBER 1968. All times EOT. Lidar
is located at Point (0,0). Horizontal arrows represent height of stack; vertical arrows represent
horizontal extent of plume as indicated by original record. See also Figure D-1.
35
-------
Figures 16-18 illustrate the large plume widths at 8.5 and 14 km down-
wind caused by stratification and fanning of the plume during the early
morning hours of 16-18 October, respectively. The "lumps" appearing in
these cross sections (and in those of other stratified plumes) are caused
*
by a bias effect in the computer contouring program due to the particu-
lar gridding technique used. This bias effect only occurs for very shal-
low plumes. The vertical plane covered here is directly over the lidar.
The plume cross sections in Fig. 19 are the farthest downwind (21 km)
that were observed during the experimental program. These cross sections
are in a vertical plane directly over the lidar. Of special interest
here is Cross Section 42, which shows no plume, but does show the haze
top associated with the temperature inversion. Evidently the tenuous
plume either meandered away or lost its identity by mixing downward.
Figures 20 and 21 illustrate the effect on plume density of all
precipitators being in operation (Fig. 21) compared with 50-percent
operation (Fig. 20 and others). In each figure, the cross sections are
at different azimuth angles and thus different downwind distances from
the plume. Cross Sections A44 and 46 are about 3.5 km downwind, Cross
Sections B44 and 47, 4.5 km downwind (and approximately crosswind), and
Cross Sections 45 and 48 are 6 km downwind. (It should be kept in mind
that as illustrated in Appendix D, these cross sections are sometimes
elongated and do not represent true plume width, except when oriented
in the crosswind direction. Plume height, however, is always truly
represented.)
Examining Cross Sections B44 (Fig. 20) and 47 (Fig. 21), the dirty
and clean plume, respectively, we see that the maximum normalized signal
return above the reference for the clean plume is 12 dB, which is 6 dB,
or four times, lower than that for the dirty plume. The apparent size
of the plume, as observed by the lidar, also shrinks, due to less of the
plume return being above the receiver noise level (see Fig. 6, Sec. IV).
*
The contouring program, developed recently by S. Briggs of SRI, was
used in its available form for this work. This bias effect, as well as
the cause for the contouring skips, could be eliminated from the program
in future studies.
36
-------
800
600
400
o
CD
200
Cross Section No. 16
0916:00-0918:15 16 Oct 68
Azimuth 100 deg mag
Program
Skip
AH (calc) = 169 r
0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8
HORIZONTAL RANGE FROM LIDAR — km
8 12 16 20
TEMPERATURE — °C
800
Cross Section No. 20
0943:15-0946:20 16 Oct 68
Azimuth 100 deg mag
0.2
0.4 0.6 0.8 1.0 1.2 1.4
HORIZONTAL RANGE FROM LIDAR
8 12 16 20
TEMPERATURE — "C
TB-7289-21
FIGURE 16 TIME SERIES OF LIDAR-OBSERVED VERTICAL CROSS SECTIONS FROM SITE 6, 3.8km
NORTHEAST OF STACK, PLUME OVERHEAD, 16 OCTOBER 1968. See also Figure D-2.
Calculated plume-rise heights (AH) are shown as dashed lines.
37
-------
HEIGHT ABOVE LIDAR -
*• £
o 8 S
. — -y , •••-] |
Cross Section No. 27
091 0:00-091 4:25 17 Oct 68
-, , , _ v Azimuth 090/270 deg mag
/Cjl 1 //^^x^ ••— -i$\
Top of Stack 0911
u — . 1640ml _
I , , , i A I I
u -
\v
- V
, I , , \
-1.6 -1.2 -0.8 -0.4 0 0.4 0.8 |Q 20
HORIZONTAL RANGE FROM LIDAR — km TEMP — °C
800
Q
LU
X
400
Cross Section No. 30
0936:00-0942:20 17 Oct 68
Azimuth 090/270 deg mag
Top of Stack
Program
'Skip
/ Convergence \
/ Zone \
J 1 L
-1.6 -1.2 -0.8 -0.4 0 0.4 0.8
HORIZONTAL RANGE FROM LIDAR — km
1.2
2315ml
1.6 2.0
TB-7289-22
FIGURE 17 TIME SERIES OF LIDAR-OBSERVED VERTICAL CROSS SECTIONS FROM SITE 6, 3.8km
NORTHEAST OF STACK, PLUME OVERHEAD, 17 OCTOBER 1968. See also Figure D-3.
38
-------
1600
0 10 20
TEMPERATURE — °C
-0.4 0 0.4 0.8 1.2
HORIZONTAL RANGE FROM LIDAR
1600
1200
LU
>
CO
<
LU
I
800
400
Cross Section No. 37
0941:00-0944:20 18 Oct 68
Azimuth 080/260 deg mag
p.s. 4-18— 'ie^S^^-^^F^^Sis^; f _
I I J I I 1 I
-0.8 -0.4 0 0.4 0.8 1.2
HORIZONTAL RANGE FROM LIDAR — km
A .
]/ . | . . v • i • | _
y\
7a\
\
\
1017 EOT \
\
Calculated \
Plume \ ~
Rise \ \
\ \ -
.1, AH(calc) \ \
«^ ^ = 241 m \ \ -
\ V —
i A i t i I i i i f1 i i \
1-6 ^0 10 20
FIGURE 18
TB-7289-23
TIME SERIES OF LIDAR-OBSERVED VERTICAL CROSS SECTIONS FROM SITE 17, 13.7km
NORTH-NORTHWEST OF STACK, PLUME OVERHEAD, 18 OCTOBER 1968. See also Figure D-4.
MRI particle sampling (P.S.) altitudes and run numbers are also indicated.
39
-------
1600
E
I
I
cc 1200
Q
£ 800
o
CO
*£
X 400
O
LLJ
X
0
A
']/ i i i i v ' I
\
Cross Section No. 40 »
0928:00-0932:00 20 Oct 68 *
Azimuth 200/020 deg mag 'a >
\
\
\
. \
\
\
Program \
Skip
#*&^*;mi**&e^ /i^f-^^^^^^^fi, d^1 °833 EDT
An 1 \ 1(=lfi ^^il^gP^^-T&<^* fr^—^ii -^-^ — <^Of\^t~^ i
~~~ , /
-«— Top I 2850 m' — *- )
of Stack | , , , , , A , i I , , < I
L ' ' ' ' ' ]/
1 1 1
"
~
\
\ -
\
\ —
\
\ -
\ -
r
i i i >
r*tf
-0.8
HORIZONTAL RANGE FROM LIDAR — km
TEMPERATURE — °C
1600
. 1200
C
3
J
j 800
>
3
3
C
- 400
I
j
0
Cross Section No. 41
0937:00-0941:30 20 Oct 68
Azimuth 200/020 deg mag
_
,_--,. ^ -*, „ ... _<=--*r=«cfiiJ's''&^'i;?' --=,--- • f '' ( * ^UtT /T®^
TOP '"^- •''•^^^^^i^S^^^^t^''^'1'-^^^^'^-^^
of Stack
~t , , : , 1,
-0.8 -0.4 0 0.4 0.8 1.2 1.6
HORIZONTAL RANGE FROM LIDAR — km
2.0
2.4
2.8
IT
<
Q
O
m
H-
I
O
LU
I
1600
1200
800
400
Cross Section No. 42
1018:00-1024:20 20 Oct 68
Azimuth 200/020 deg mag
•*— Top of Stack
iii.
-1.2 -0.8 -0.4 0
A
v ' ' I ' X ' I ' ' '
\ _
3 \
\
\
\ -
^ V \
Q \ \
\\-
^^
-------
800
600
400
200
LLJ
I
Cross Section No. A44
0854:00-0855:40 21 Oct 68
Azimuth 270 deg mag
- Top of Stack
_J L_
0.4 0.6 0.8
1.0 1.2 1.4 1.6 1.8 2.0 2.2
HORIZONTAL RANGE FROM LIDAR — km
2.4
2.6
2.8
3.0
800
600
£ 400
O
in
<
t- 200
I
Cross Section No. B44
0856:40-0858:35 21 Oct 68
Azimuth 225 deg mag
-Top of Stack
-A H (calc) = 203 m
0
0857 EOT
0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
HORIZONTAL RANGE FROM LIDAR — km
2.2 "4 8 12
TEMPERATURE — °C
800
-------
E 800
600
400
o
CD
200
LLJ
* 0
Cross Section No. 46
0903:00-0904:40 21 Oct 68
Azimuth 270 deg mag
- Top of Stack
0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2
HORIZONTAL RANGE FROM LIDAR — km
2.4
2.6
2.8
3.0
E 800
\ 600
3
J
i 400
0
£
- 200
L
0
J
C 0
A
V ' | ' y |
Cross Section No. 47 \ \
0906:30-0908:20 21 Oct 68 \ \
\ \
Azimuth 225 deg mag ( 7 \
/— -., j '^f^T/^iST^ \ \
PS 1-21 -•- ^-r^ ^^§^ L^ y^J \
^l^ 5T? ^Zf^-^/ A \
g^-^r -^^^,.—^.~^—, ^ H (calc) = 206 m \
0857 EOT
- — Top of Stack /
$
f i i i i i i f] i A . \ . . . \
0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 "4 8
-
\
\
\
\
\
\
\ —
\
\
V
II I \
12
HORIZONTAL RANGE FROM LIDAR — km TEMPERATURE — °C
800
cc 600
<
Q
uj 400
O
CO
<
H 200
I
0
LU
x 0
Cross Section No. 48
0909:30-0910:50 21 Oct 68
Azimuth 180 deg mag
?~
Y
Top of Stack
FIGURE 21
0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2
HORIZONTAL RANGE FROM LIDAR — km
2.4 2.6 2.8 3.0
TB-7289-26
SPATIAL SERIES OF LIDAR-OBSERVED VERTICAL CROSS SECTIONS FROM SITE 22, 4.9km
SOUTHEAST OF STACK, 21 OCTOBER 1968, PRECIPITATORS IN FULL OPERATION.
See also Figure D-6.
42
-------
Later during the same morning, the cross sections in Fig. 22 were
observed on the same azimuth angles. Here the lower portion of the
plume is fumigating, while the upper portion remains relatively intact,
probably due to the temperature inversion at 470 m.
Figures 23-26 illustrate various stages in the structure of the
plume during the morning and midday of 22 October. In these figures the
stack location is into the paper. Considering the cross sections located
nearly crosswind (Cross Sections 58, 64, and 73), we see that at 0904 EOT
the plume is compact and level (Cross Section 58), at 0941 EDT wind shear
has set in and the plume is tilted, and at 1232 EDT fumigation of the
lower portion is occurring. For the clean plume (Fig. 24), essentially
no plume is apparent in the cross sections due to the weakness of the
return signal in relation to the background noise level. Possibly be-
cause of the 3-MHz bandpass of the disk recorder, the data for clean
plumes do not show up as well on these cross sections as on the 35-mm
film, from which the plan views in Appendix D were obtained.
In Fig. 27, the plume apparently cannot penetrate the 1° C inversion
at 500 m above the lidar, and is spreading out beneath with fumigation
of the lowest portion. Cross Section 79 shows an interesting bubble at
the top, which could be a thermal which has risen through the plume.
A series of approximately crosswind-oriented cross sections at about
15-minute intervals is shown in Figs. 28-30. The rapidly changing struc-
ture of the plume is apparent, as well as the trapping of the effluent
by the stable layer at 400-500 m above the lidar. On this date, 24
October, the particulate output of the stack was reduced by 25 percent
from previous days, since two additional precipitator sections were
operating. As shown in Cross Section 104 (Fig. 29), there is some down-
ward diffusion of the plume, but the stable near-ground layer apparently
prevents the plume material from mixing to the ground.
43
-------
800
600
400
O
CO
200
LLI
X
Cross Section No. 54
1023:00-1025:55 21 Oct 68
Azimuth 270 deg mag
AH (calc)
0 0.2 0.4 0.6 0.8
HORIZONTAL RANGE FROM LIDAR — km
-T !_, /L
1.0 1.2 V
4 8 12
TEMPERATURE — °C
800
600
LU 400
O
CO
<
I- 200
I
O
LU
I
Cross Section No. 55
1027:30-1030:30 21 Oct 68
Azimuth 225 deg mag
0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8
HORIZONTAL RANGE FROM LIDAR — km
800
DC 600
<
Q
400
O
CO
X 200
Cross Section No. 56
1033:20-1036:50 21 Oct 68
Azimuth 180 deg mag
-Top of Stack
2030 m L_»
0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8
HORIZONTAL RANGE FROM LIDAR — km
TB-7289-27
FIGURE 22 SPATIAL SERIES OF LIDAR-OBSERVED VERTICAL CROSS SECTIONS FROM SITE 22, 4.9km
SOUTHEAST OF STACK, 21 OCTOBER 1968. See also Figure D-6.
44
-------
800
600
> 400
o
DO
X 200
LU
x
0
Cross Section No. 57
0900:00-0901:20 22 Oct 68
Azimuth 270 deg mag
-Top of Stack
1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6
HORIZONTAL RANGE FROM LIDAR — km
2.8
800
Cross Section No. 58
0903:00-0904:30 22 Oct 68
Azimuth 300 deg mag
T.O 1.2 1.4 1.6 1.8 2.0 2.2
HORIZONTAL RANGE FROM LIDAR — km
4.0 8.0 12.0
TEMPERATURE — °C
cc
<
Q
800
600
400
X 200
Cross Section No. 59
0905:20-0907:30 22 Oct 68
Azimuth 330 deg mag
••—Top of Stack
1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4
HORIZONTAL RANGE FROM LIDAR — km
2.6
2.8
TB-7289-28
FIGURE 23 SPATIAL SERIES OF LIDAR-OBSERVED VERTICAL CROSS SECTIONS FROM SITE 6, 3.8km
NORTHEAST OF STACK, 22 OCTOBER 1968. See also Figure D-7.
45
-------
800
= 600
Q
>
O
CO
<
400
g 200
Cross Section No. 60
0921:45-0923:10 22 Oct 68
Azimuth 270 deg mag
|.
1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4
HORIZONTAL RANGE FROM LIDAR — km
2.6
2.8
800
600
400
200
cc
o
HI
O
CO
<
I-
CJ
Cross Section No. 61
0923:50-0925:10 22 Oct 68
Azimuth 300 deg mag
PS. 4-22^
1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4
HORIZONTAL RANGE FROM LIDAR — km
2.6
2.8
cc
<
o
o
CO
<
I-
I
o
HI
X
800
600
400
200 T°P
1.4 1.6 1.8 2.0 2.2 2.4
HORIZONTAL RANGE FROM LIDAR — km
2.6 2.8
TB-7289-29
FIGURE 24 SPATIAL SERIES OF LIDAR-OBSERVED VERTICAL CROSS SECTIONS FROM SITE 6, 3.8km
NORTHEAST OF STACK, 22 OCTOBER 1968, CLEAN PLUME. See also Figure D-7.
46
-------
= 600
3
J
> 400
3
Q
I
E 200
1
J
c
0
1
800
c 600
3
J
U> 400
Q
(
E 200
p
u
c
0
1
800
: 600
C
i
j
i 400
D
1
E 200
3
J
C
0
1
Cross Section No. 63
0938:30-0939:45 22 Oct 68
Azimuth 270 deg mag
^^f - ffij
' ^iK^ri^
" — Top of Stack
J , I
0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6
HORIZONTAL RANGE FROM LIDAR — km
2.8
Cross Section No. 64
0940:35-0942:00 22 Oct 68
Azimuth 300 deg mag
P.S. 8-22 — «-
_. '•,^i£^^=^-.~^ - -
P.S.7-22— itf^ff^^}^
•« — Top of Stack ~" N^^^z2^r^—
1 .
0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6
HORIZONTAL RANGE FROM LIDAR — km
Cross Section No. 65
943:45-0945:15 22 Oct 68
Azimuth 330 deg mag
^^^7"";^^~^ 5:"-^
•* — Top of Stack
0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6
HORIZONTAL RANGE FROM LIDAR — km
2.8
2.8
TB-7289-30
FIGURE 25 SPATIAL SERIES OF LIDAR-OBSERVED VERTICAL CROSS SECTIONS FROM SITE 6, 3.8km
NORTHEAST OF STACK, 22 OCTOBER 1968. See also Figure D-7.
47
-------
800
cc 600
Q
_1
tH 400
200
-Top of Stack
Cross Section No. 72
1227:00-1229:15 22 Oct 68
Azimuth 270 deg mag
1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0
HORIZONTAL RANGE FROM LIDAR — km
800
Cross Section No. 73
1231:00-1233:00 22 Oct 68
Azimuth 300 deg mag
0
0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 ' |6 20
HORIZONTAL RANGE FROM LIDAR — km TEMPERATURE
800
DC
<
Q
600
O
m
400
200
LU
X
-•— Top of Stack
Cross Section No. 74
1234:45-1236:55 22 Oct 68
Azimuth 330 deg mag
3500 ml
0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4
HORIZONTAL RANGE FROM LIDAR — km
2.6
2.8
TA-7289-31
FIGURE 26 SPATIAL SERIES OF LIDAR-OBSERVED VERTICAL CROSS SECTIONS FROM SITE 6, 3.8km
NORTHEAST OF STACK, 22 OCTOBER 1968. See also Figure D-7.
48
-------
800
600
O
CO
400
200
Cross Section No. 79
0856:45-0858:30 23 Oct 68
Azimuth 220 deg mag
- Top of Stack
0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4
HORIZONTAL RANGE FROM LIDAR — km
2.6 2.8 3.0
800
Cross Section No. 80
0859:45-0901:45 23 Oct
Azimuth 180 deg mag
0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2
HORIZONTAL RANGE FROM LIDAR — km
4 8 12
TEMPERATURE — "C
800
cc 600
Q
O
m
400
200
I
C3
LLJ
0
Cross Section No. 81
0902:40-0904:45 23 Oct 68
Azimuth 140 deg mag
0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0
HORIZONTAL RANGE FROM LIDAR — km TA-7289-32
FIGURE 27 SPATIAL SERIES OF LIDAR-OBSERVED VERTICAL CROSS SECTIONS FROM SITE 6, 3.8km
NORTHEAST OF STACK, 23 OCTOBER 1968. See also Figure D-8.
49
-------
800
Cross Section No. 93
0933:15-0934:40 24 Oct 68
0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8
HORIZONTAL RANGE FROM LIDAR — km
4 8 12 16
TEMPERATURE — UC
800
600
u 400
O
CQ
<
H- 200
I
O
Cross Section No. 99
1004:40-1006:00 24 Oct 68
Azimuth 070 deg mag
Top of Stack
0 0.2 0.4
0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
HORIZONTAL RANGE FROM LIDAR — km
2.2
2.4 2.6
E 800
600
400
O
CO
I- 200
x
u
LU
I
Cross Section No. 101
1020:00-1021:45 24 Oct 68
Azimuth 070 deg mag
-Top of Stack
0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
HORIZONTAL RANGE FROM LIDAR — km
2.2
2.4
2.6
TB-7289-33
FIGURE 28 TIME SERIES OF LIDAR-OBSERVED VERTICAL CROSS SECTIONS FROM SITE 11, 4km
NORTHWEST OF STACK, 24 OCTOBER 1968. See also Figure D-9.
50
-------
800
= 600
Q
_l
£ 400
O
CO
200
I
Hi
X
Cross Section No. 103
1035:15-1037:50 24 Oct 68
Azimuth 070 deg mag
-Top of Stack
0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8
HORIZONTAL RANGE FROM LIDAR — km
2.0 2.2 2.4 2.6
800
oc 600
400
Cross Section No. 104
1047:30-1050:30 24 Oct 68
Azimuth 070 deg mag
O
CD
200
I
CD
-Top of Stack
AH(calc)
0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8
HORIZONTAL RANGE FROM LIDAR — km
8 12 16
TEMPERATURE — °C
800
cc 600
Q
_J
LU 400
O
CQ
200
I
CJ
LU
Cross Section No. 106
1103:30-1105:45 24 Oct 68
Azimuth 070 deg mag
-Top of Stack
0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6
HORIZONTAL RANGE FROM LIDAR — km
TB-7289-34
FIGURE 29 TIME SERIES OF LIDAR-OBSERVED VERTICAL CROSS SECTIONS FROM SITE 11, 4km
NORTHWEST OF STACK, 24 OCTOBER 1968. See also Figure D-9.
51
-------
800
600
400
200
I
(D
LU
I
Cross Section No. 107
1117:30-1119:50 24 Oct 68
Azimuth 070 deg mag
-Top of Stack
0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8
HORIZONTAL RANGE FROM LIDAR — km
2.0
2.2
2.4
2.6
800
600
>
o
m
<
LU
I
400
200
Cross Section No. 108
1134:00-1136:30 24 Oct 68
Azimuth 070 deg mag
-Top of Stack
0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8
HORIZONTAL RANGE FROM LIDAR — km
2.0
2.2
2.4
2.6
800
600
40°
Cross Section No. 109
1145:45-1151:35 24 Oct 68
Azimuth 070 deg mag
O
CQ
200
••—Top of Stack
AH (calc) = 429 m
0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6
HORIZONTAL RANGE FROM LIDAR — km
L
1.8
8 12 16
TEMPERATURE — °C
TB-7289-35
FIGURE 30 TIME SERIES OF LIDAR-OBSERVED VERTICAL CROSS SECTIONS FROM SITE 11, 4km
NORTHWEST OF STACK, 24 OCTOBER 1968. See also Figure D-9.
52
-------
VI INTERPRETATION OF CROSS SECTIONS IN TERMS
OF PLUME RISE AND DIFFUSION
The new capability to "see" the detailed structure and geometry of
plumes, which lidar makes possible, gives the instrument decided poten-
tial in plume rise and diffusion studies. However, as often happens
when a new technique enables us to obtain detailed observations where
only sparse data were available before, the lldar data have underscored
the complexities of plume behavior and made apparent the weaknesses of
certain of the present concepts and theoretical assumptions.
After examination of the plume cross sections in Sec. V, it is
apparent that plume tilting and fanning caused by wind direction shear
is a characteristic feature of plume behavior in stable conditions. It
then becomes difficult to identify any one level of the plume for use
in computing plume rise. The concept of plume rise also tends to lose
meaning under fumigation conditions, which were frequently observed,
and under unstable (looping) conditions. Equally important as compli-
cating factors are the temperature inversions and stable layers, which
regularly occur aloft and tend to trap the plumes and prevent further
rise. The lidar data for October show several instances of this. Thus
the data emphasize that there typically are significant vertical varia-
tions in stability and wind structure that have profound effects upon
plume behavior, and that are not taken into account by the plume rise
theories .
Plume-rise values (^H) for the stable cases were calculated from
formula rec
(Smith, 1968):
*
the formula recommended by the ASME Committee on Air Pollution Controls
AH = 2(F/GU)1/3 (5)
American Society of Mechanical Engineers.
-------
with
and
F = g r2 Vg (Tg - Ta)/Ta
G = (g/T UAT/AZ + o.oi)
\ a/
U = (Ut + UJ/2
AT = T. - T
X 3
AZ = H - H (after Proudfit, 1969)
t s
where
F = buoyancy flux
g = acceleration of gravity (= 9.8 m/s)
r = radius of stack orifice (= 4.25 m for Keystone)
V = stack gas exit velocity
S
T = stack gas exit temperature
S
T = ambient atmospheric temperature at stack top
a
G = stability parameter
U = mean wind speed over height interval where rise occurs
U = wind speed at height of plume top (H,) after maximum rise
L C
U = wind speed at height of stack top (H )
3 S
T = air temperature at height of plume top.
The computed values, as well as the values of the input parameters used,
are listed in Table III.
In view of the complications discussed previously with regard to
defining plume rise in a manner that permits objective determination
from lidar plume cross sections, the calculated plume rise values have
been indicated directly on the cross sections, Figs. 9-12 and 15-30.
54
-------
Table III
VALUES OF INPUT PARAMETERS USED AND RESULTS FROM PLUME RISE CALCULATIONS
USING ASME FORMULA FOR STABLE CONDITIONS
Date
(1968)
5/25
5/26
10/16
10/17
10/18
10/20
10/21
10/22
10/23
10/24
Time
(EOT)
0919
1036
0900
0917
0945
0912
0920
0942
0928
0857
0907
1024
0903
1232
0900
0934
1049
1148
Cross
Section
Number
2
4
12
16
20
27
34
37
40
B44
47
54
58
73
80
93
104
109
V
s
(ft/s)
73.0
73.0
68.5
68.6
68.6
68.7
65.9
65.9
67.1
67.0
69.0
69.0
59.2
59.2
73.1
61.3
61.3
61 .3
T
s
(°F)
275.0
275.0
270.0
292.0
292 .0
292.0
289.0
289.0
267.0
278.0
278.0
278.0
292.0
292.0
285.0
300.0
300.0
300.0
T
a
(°C)
12.8
12.7
15.4
17.3
17.3
19.7
18.7
20.0
9.6
7.5
7.5
8.6
11.5
15.7
7.3
9.6
9.8
11.0
Tt
(°C)
11.6
9.7
13.4
17.5
17.6
18.3
17.1
16.5
11.0
5.5
5 .8
5 .9
12.1
13.2
4.7
8.3
8.5
7.7
Ht
(m)
590
850
540
470
430
530
550
580
460
620
570
580
430
580
550
560
470
510
H
s
(m)
240
240
160
152
152
152
122
122
122
162
162
162
176
176
176
152
152
152
Ut
(m/s)
7.5 at 84°
5.3 at 89°
11.0 at 157°
9.0 at 186°
10.5 at 187°
12.4 at 176°
10.9 at 168°
14.3 at 165°
6.8 at 248°
9.1 at 326°
7.8 at 328°
8.1 at 325°
7.9 at 193°
14.9 at 218°
9.0*
4.9 at 158°
6.4 at 156°
6.2 at 150°
U
a
(m/s)
6.2 at 70°
4.5 at 49°
3.0 at 111°
5 .9 at 153°
5.2 at 150°
7.2 at 150°
7.2 at 136°
7.2 at 136°
6.4 at 235°
7.5 at 305°
7.5 at 305°
7.4 at 314°
5.5 at 153°
6.0 at 196°
7.5 at 241°
5.7 at 127°
6.9 at 137°
5 .8 at 127°
AH
(m)
206
251
220
169
164
183
185
241
156
203
206
242
161
203
257
218
212
429
AH + H
(m)
446
491
380
321
316
235
307
363
278
365
368
404
337
379
433
370
364
581
Estimated.
-------
In this way, detailed comparisons with the observed plume heights and
shapes are possible.
In general, the ASME formula appears to predict plume rise reasonably
well, although of course the frequent presence of temperature inversions
at plume levels diminishes the relevancy of the theory in many instances.
For the few cases where the vertical temperature gradient at plumes
levels is approximately uniform with height, and vertical wind shear—
and thus plume tilting—is weak (e.g., Cross Section 58, Fig. 23), the
agreement between the theory and the observations is excellent. On the
whole, there appears to be a weak indication that the ASME formula may
predict slightly lower plume rises than observed.
The capabilities of lidar in computing diffusion parameters are
fairly obvious, at first glance, from inspection of the horizontal plan
views of plume positions in Figs. 9-12 (insets) and in Appendix D, which
show the horizontal spread of the effluent with downwind distance.
Again, however, wind direction shear and the attendant spreading of the
plume by shearing are not adequately covered by present theories. The
best application of current theory is probably for the neutral and un-
stable cases, when vertical wind shear is not usually so pronounced.
Here another problem is encountered, in that most theories are only
valid for mean concentrations over a period of 30 minutes to an hour
(Gifford, 1961). To be most useful here, then, a lidar should be
capable of taking fast, quasi-instantaneous cross sections through a
plume at frequent intervals for extended periods on the order of an hour.
The results of these cross sections would then be averaged to obtain a
mean concentration distribution to compare with theory. The present
equipment requires about two to three minutes for a complete cross sec-
tion. Under unstable conditions when the plume position is rapidly
changing, some distortion is introduced into the cross sections due to
the lack of instantaneity of the individual lidar observations.
Another important consideration for studying plume diffusion with
a lidar is the receiver signal-to-noise ratio at the range where the
plume is observed. As seen in Fig. 6, a portion of the plume return is
56
-------
truncated when the plume is observed at a range beyond that at which
clear-air return is detected above the noise. This decreases the ob-
served size of the plume, and hence affects the diffusion calculations
unless suitable allowances are made. A prime example of this effect is
shown by Figs. 20 and 21 and the corresponding plume plan views in
*
Appendix D. Here a series of six cross sections were obtained within a
time period of 15 minutes. The precipitators were all on when the last
three cross sections were observed, furnishing a clean plume. Due to
background noise considerations, this clean plume is "seen" by the lidar
to be considerably smaller than the dirty plume. However, the diffusion
regime was the same.
The Mark V lidar has an effective above-noise range for clear-air
return on a typical bright day of about one to two km in the neodymium
t
version, but this could be readily improved by using a ruby or YAG
laser and/or additional signal-to-noise enhancement equipment and
techniques .
2k
The effect is not as pronounced in the plan views due to the greater
recording resolution of the 35-mm film system.
Ytrrium aluminum garnet, wavelength 1.06 |_i.
57
-------
VII EVALUATION OF PARTICULATE MASS CONCENTRATIONS
FROM LIDAR NORMALIZED SIGNAL RETURN (S) VALUES
As discussed in Sec. IV. the conversion of S values to absolute
mass concentrations requires allowance for the effect of attenuation of
the laser beam as it passes through the atmosphere and the plume. This
problem is best approached by solving the differential equation derived
from Eq. (4) (Sec . IV) :
dR ~ ' |3 dR ' a
where Bouquer's law of attenuation
T(R) = exp - / rrU) d S, (7)
_ ° _
has been assumed and a(£) is the volume extinction coefficient at a dis-
tance i along the path. The solution of Eq. (6) for particulate mass
concentration along the laser beam path requires relating the optical
parameters a and g to the particulate mass concentration. For this
purpose the ratio of each optical parameter to the particulate mass con-
centration M(mass/volume) is defined as
-
M
M
Substitution of Eq . (8) into the lidar differential equation [Eq . (6)]
results in an expression relating mass concentration to the lidar nor-
malized signal return:
dR
4.34
,0
(9)
where
has been considered independent of range. The validity of this
assumption will be discussed later. This nonlinear equation may be
solved by first employing the linearization transform 7] = M with the
59
-------
result
-1
M(R) = exp C
]
M
- 2 F
=
/ exp C^SU)
JR
(10)
where C = 1/4.34. This solution principally assumes that
(1)
and ^ are invariant with range and the value of
G D
can be computed, and
(2) a boundary or clear-air particulate mass concentration,
M
is known or can be estimated.
Microscopic analysis of the Keystone fly ash clearly shows that the
particles are glass spheres, apparently formed from the mineral content
of the coal by the intense heat of the combustion process; hence, the
use of Mie theory in deriving estimates of ^ and £, is justified. To
a high degree of accuracy, the volume extinction coefficient at the
lidar wavelength of 1.06 |_i may be considered to be essentially dependent
only on the atmospheric particulate matter, and the ^ values may be ex-
pressed
as :
n a Q (x,m) n(a) da
G
4 3
— TT a p n(a) da
(11)
TT a Q (x,m) n(a) da
4 3
— TT a p n(a) da
(12)
To be precise, the parameters ^e and ^b defined by Eqs. (11) and (12)
are exactly equivalent with those defined by Eq. (7) only when multiple
scattering effects are absent. This is a good assumption for this study
because of the narrow field of view of lidar systems (<1 mrad) .
60
-------
where
a = particle radius
Q = extinction efficiency factor
G
Q = backscatter efficiency factor
x = size parameter = 2n a/X
m = refractive index of particles
X = wavelength of the laser energy
n = number of particles per unit volume per unit radius
interval, and
p = density of the particulate matter.
The efficiency factors may be computed from Mie theory for a given
particulate matter refractive index, the real part of which in this case
was estimated to have a value of approximately 1.5, based upon the known
mineral composition of coal. The requirement that ^ and ? be inde-
6 u
pendent of range is satisfied when the relative particle-size distribu-
tion is invariant with range; i.e.,
n(R,a) = C(R) n (a) (13)
' r '
where C(R) is the number concentration and n (a) is the relative size
distribution. Also, with this assumption only the relative particle
size distribution is required to evaluate the ^ values. However, n (a)
is rarely completely independent of range, and the question may be asked
as to how small changes in the relative size distribution are reflected
in the ^ values. Since the £ values are expressed in terms of integrals
over the range of particle sizes, monodispersed (single-sized) particles
would cause largest ^ variations with slight changes in the particle size.
Mie computations of ^ and ^ are presented in Fig. 31 for the case of
monodispersed particles consisting of material with a density of 2.1 g/cm
(as measured at the Institute from a Keystone fly ash sample) and a re-
fractive index of m = 1.5-iO for incident radiation of wavelength 1.06 u.
61
-------
4 6
PARTICLE RADIUS — microns
8 10
TA-7289-36
FIGURE 31
DEPENDENCE UPON PARTICLE SIZE OF TOTAL SCATTERING
EXTINCTION—OPTICAL CROSS SECTION PER UNIT MASS (£ }. AND BACK
BACK-SCATTERING OPTICAL CROSS OPTICAL CROSS SECTION PER
UNIT MASS PER STERADIAN (^). For mono-disperse particles with
refractive index of 1.5-iO and density of 2.1 g/cm3
62
-------
As shown, slight variations in particle size can cause large varia-
tions in F and smaller but significant variations in F . Because of
bb ^e
this, and the fact that P is more sensitive to variations in p than a
s
[Eq. (1)], the requirement that £ remains constant with range also
satisfies the less stringent requirement that ^ remains invariant.
6
A real aerosol contains particles of many sizes, so the variations
in the ^ curves tend to be smoothed out when a weighted integral for all
sizes present is computed. The next section will explain the methods
developed to handle the supplementary particle-size data obtained during
the Keystone study for use with the lidar observations.
63
-------
VIII ANALYSIS OF ATTENUATION EFFECT
USING PARTICLE-SIZE CHARACTERISTICS
The Keystone stack plume particulates were sampled with an airplane-
mounted Goetz moving slide impactor (Goetz, 1969) by Meteorology Research,
Inc. (MRI), at times nearly concurrent with the SRI lidar observations.
Tentative relative particle size distribution data for specified sampling
periods were furnished SRI (see Table IV) along with information re-
quired to evaluate the number of particles per unit volume per unit
radius interval for increments of 0.25 p. in radius. The relative size
distribution is sufficient to evaluate the optical parameters | and F
G D
for use with the lidar data.
Ideally, however, one would like to compare mass concentrations
measured by the MRI sampler with those derived from the lidar data.
This was not possible, since our conversions of the relative particle-
size data to absolute size distributions (using the furnished sample-
volume values) and then to mass concentrations give unrealistically
3
large values of concentration for the clear air (M > 1000 ng/m ) upwind
of the stack. Hence it has been necessary to use the empirical relation-
ship derived by Noll, et al. (1968) to determine the clear-air background
mass concentration (M ) from visual estimates of atmospheric visibility.
The solution parameter ^ and the parameter ^ were evaluated from
the furnished relative size distributions for each data sample listed in
Table IV; these values are tabulated in Table V. In these computations,
integrations over particle radii were performed by means of rectangular
integration, with the size distribution assumed constant within each
sampling interval. The parameters £ and I* were evaluated from sampling-
G D
interval-averaged optical cross sections defined as:
C =
Q = (\ /4rr) fx Q(x)dx/fdx (14)
\ / J
-------
Table IV
SUMMARY OF MRI PARTICLE-SIZE DISTRIBUTION DATA
(Compiled by Stanford Research Institute)
Count
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Date
(October
1968)
16
17
17
17
17
17
17
17
17
17
17
18
18
18
18
18
18
18
21
22
22
22
22
22
Time
(EDT)
0953
0853
0957
1219
1219
1223
1223
1230
1230
1230
1252
0951
1009
1009
1009
1009
1009
1142
0912
0930
0942
0944
0945
1233
Slide
No.
5-16
2-17
16-17
22-17
22-17
23-17
23-17
24-17
24-17
24-17
26-17
1-18
4-18
4-18
4-18
4-18
4-18
6-18
1-21
4-22
6-22
7-22
8-22
12-22
Slide
Position
2
4
10
17
17
2
12
5
8
17
5
3
2
2
2
7
7
5
10
18
13
13
2
7
Photo
Exposure
Time
(s)
0.5
0.5
0.2
0.2
0.2
0.1
0.5
0.1
0.2
0.5
0.5
0.5
0.5
0.2
0.1
0.5
0.2
0.5
0.5
0.5
0.5
0.2
0.5
0.5
Slide
Exposure
Time
(s)
29
29
109
106
106
106
106
226
226
226
17
116
117
117
117
117
117
244
59
29
27
28
28
7
Aircraft
Radial Position
Relative
to Stack
(km)
*
3 DW
3 DW
16 DW
8 DW
8 DW
8 DW
8 DW
8 DW
8 DW
8 DW
7
14 DW
17 DW
17 DW
17 DW
17 DW
17 DW
5 UW
5 DW
4 DW
4 DW
4 DW
4 DW
4 DW
Relative
to Lidar
(km)
*
6 UW
6 UW
7 DW
Over
Over
Over
Over
Over
Over
Over
7
Over
3 DW
3 DW
3 DW
3 DW
3 DW
20 UW
Over
Over
Over
Over
Over
Over
Aircraft
Altitude
(Ft
MSL)
2320
2750
2650
3000
3000
1800
1800
3000
3000
3000
7
2500
3100
3100
3100
3100
3100
1250
to
3750
2800
2700
2100
2400
2700
2400
(m above
Lidar)
289
420
390
497
497
131
131
497
497
497
7
314
497
497
497
497
497
Below
692
446
429
246
338
429
338
Portion
of Plume
Sampled '
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
7
Bottom
Top
Top
Top
Top
Top
None"
Middle
Above(?)
Bottom
Middle
Above (?)
Middle
Precipitator
Capacity On
(percent)
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
100
100
50
50
50
62
Associated
Lidar
Cross
Section
No.
—
—
—
33
33
33
33
33
33
33
7
37
37
37
37
37
37
—
47
61
64
64
64
73
Ol
Notes: DW = downwind, UW = upwind.
Judging from MRI aircraft data and lidar cross sections.
Clear" air background upwind of stack.
-------
Table V
VALUES OF OPTICAL PARAMETERS COMPUTED
FROM RELATIVE PARTICLE-SIZE DATA
Count Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
?e
(mV1)
0.63211
0.64544
0.70527
0.56964
0.85449
0.36938
0.93607
0.97728
0.55544
0.51767
0.59043
0.32160
0.37274
0.47949
0.45573
0.54664
0.84976
0.39697
0.35974
0.21364
0.51050
0.25768
0.38903
0.68789
?b
(m2g~1ster~1)
0.03647
0.05744
0.06434
0.06138
0.05519
0.03603
0.05103
0.04638
0.05882
0.05197
0.02630
0.02109
0.02622
0.03176
0.02401
0.02836
0.03571
0.03233
0.02950
0.01382
0.04480
0.02137
0.03518
0.05680
It was indicated by MRI that the particle counting technique is in-
sensitive to particles with radii less than 0.50 to 0.25 |~i. The influence
of this on the evaluated quantities, as well as the validity of the rec-
tangular integration, was investigated by fitting a continuous Junge dis-
tribution function to the observed particle counts. Figure 32 illustrates
the method used to evaluate the Junge parameters from the cumulative
particle-size distribution of Count No. 21, which has been reduced to
absolute values by the tentative MRI information. This particular sample
has considerably fewer small particles (a < 1.0 |a) and also fewer large
particles (a > 3 |_i) than the indicated Junge model.
67
-------
1000
o
•a
8.c.
100
Q
UJ
o
o
2
Z
I
t-
cc
LJ
CC
_J
UJ
o
t-
cc
o.
u.
o
cc
LJ
m
10
1.0
O.I
0 O
METHODOLOGY:
n = C Q-k
Slope l-k
Intercept =
1 >~
COUNT No. 21
SLIDE 6-22
0942 EOT
22 OCT 1968
n = 201 o
-4.346
O.I 1.0 10
PARTICLE RADIUS, a microns
100
TA-7289-37
FIGURE 32 SAMPLE CUMULATIVE PARTICLE SIZE DISTRIBUTION
AND JUNGE-TYPE ANALYTICAL FIT
Assuming that the Junge model is a satisfactory representation of
the real aerosol, it can then be employed with a more exact trapezoidal
integration of ^ (maintaining 0.1 increments in x) . Integration with
o
limits of a =1.0 and 10.0 |_L results in g; = 0.37 m /g, slightly less
than the rectangular integrated value (0.51), which is expected because
of the increased number of larger particles in the Junge model. Inte-
gration with limits of a =0.1 and 10.0 (j. results in F = 0 87 the in-
=e ' '
crease occurring because of the larger number of small particles in the
Junge distribution.
68
-------
The lidar and particle sampling of the Keystone plume were best
coordinated for Cross Section 64} which is oriented approximately cross-
wind, and this case was chosen to apply the mass concentration solution
discussed in Sec. VII. The boundary parameter M was taken from the
visibility estimate (7 to 10 miles),, which gives M = 100 ug/m , using
the relationship of Noll, et al. (1968). The value of £ was chosen in
Q
accordance with the computation (Table V) for Particle Count 22 as
2 , *
0.2 m /g . The results of these absolute mass concentration computa-
tions are presented in Fig. 33, which also illustrates the manner of the
removal of extraneous clear-air and background noise contours from the
plume cross sections, which is normally employed in order to better de-
fine the plume boundaries and shape. Figure 33(c) represents the re-
sults of the graphical subtraction of Fig. 33(a) from 33(b), and thus
shows contours of mass concentration in decibels added to the back side
of the plume by the attenuation correction. It may be noted that the
maximum correction is 9 dB, which means that the mass concentration at
that point was almost 10 times too low in the uncorrected cross section,
Fig. 33(a) . It may be seen that although the attenuation correction
does change concentrations significantly in one part of the cross sec-
tion, the general shape and structure of the plume remain the same.
A planimeter was used to integrate the total mass per unit plume
length (cross-axial integrated mass concentration) represented by the
cross sections in Fig. 33(a) and (b) . The results were as follows:
uncorrected cross section (a), 350 g/m of plume length; corrected (b),
A value of 0.3 m /g was tried also,, but gave no solution due to the
integration becoming unstable. Although solutions were not attempted
for any other cross sections,, many of the high values of £,e shown in
Table IV would probably cause similar unsuccessful results. These high
values are a consequence of the surprisingly small modal particle sizes
reported by MRI. The cause of this apparent discrepancy between the
lidar data and the particle-size data is unknown, but there are a number
of possibilities, such as (1) the free-air particles could have a thin
coating of water, which increases their optical size, (2) particle
agglomerates may break up when introduced into the sampler, (3) some of
the assumptions used in the lidar data analysis may be faulty, or (4)
there may be an error in the particle-size measurements or analyses.
69
-------
600
< 400
LU
o
CD
<
I- 200
I
<£
LU
I
(a) Without attenuation correction
1.2
1.4
1.6
1.8
2.0
2.2
600
< 400
Q
O
m
200
o
LU
I
(b) With attenuation correction
1.2
1.4
1.6
1.8
2.0
2.2
Q
_l
LU
O
CD
LU
I
600
400
200
(c) (a)-(b)
CROSS SECTION NO. 64
0940.35-0942.00 22 OCT
AZIMUTH 300 DEC MAG
1.2
1.4 1.6 1.8 2.0
HORIZONTAL RANGE FROM LIDAR — km
2.2
FIGURE 33 EFFECT AND MAGNITUDE OF CORRECTION OF PLUME MASS
CONCENTRATION DISTRIBUTIONS FOR ATTENUATION. Contours
represent 10 Iog10 (M/M0) and thus are labelled in decibels. Solution
parameters for (b) are clear-air concentration M0 = 100 x 10~6 g/m3;
£e = 0.2 m2/g; dashed line faired in by hand separates clear-air and
background-noise contours from plume contours.
70
-------
680 g/m, which shows that the mass correction was significant. These
values were compared with an estimate of the particulate emission rate
from the stack, obtained from the following parameters existing at the
time the cross section was obtained:
3
Coal burning rate = 651 x 10 Ib/hr
*
Fly ash content of coal = 17 percent
Precipitator collection efficiency = 50 percent
(half in operation)
Wind speed at plume level = 8 m/s.
Using these values gives a stack source strength of 7000 g/s, and
a plume cross-axial integrated mass concentration of 875 g/m, compared
with the 680 g/m measured by the lidar, after allowance for attenuation
Although further evidence will be necessary from other comparisons, the
close agreement of this initial example does indicate strongly that the
lidar technique is capable of useful quantitative measurements.
*
Full conversion to gas-borne fly ash particles is assumed
71
-------
IX RECOMMENDATIONS FOR FURTHER RESEARCH
This first feasibility study,, although limited in scope and scale
of effort, has demonstrated that lidar is of significant value in plume
studies. The instrument can furnish information about plume rise and
dispersion from tall stacks that is unobtainable in such detail by any
other means. To take advantage of the full potential of the technique,
additional work is needed in a number of areas. These fall for the most
part within the general categories of
(1) Lidar instrumentation improvements,
(2) Additional data-handling and reduction capabilities,
(3) Improved experimental and analytical techniques, and
(4) Testing of advanced lidar concepts.
A number of specific suggestions for improving lidar usefulness in plume
studies are itemized below.
A. Lidar Instrumentation
• The observations described in this report were made with SRI's
general-purpose experimental Mark V lidar operating in its neodymium
configuration. Although this provided a higher pulse repetition frequency
(10-12 pulses/min) than the ruby version, the latter system is more sen-
sitive and could thus have acquired data at longer ranges, or better
information on "clean" plumes (i.e., with full precipitator operation).
The lower pulse rate of the ruby system (3-4 pulses/min) would have
seriously reduced the data rate.
These considerations well illustrate the problem confronting
the designer of an optimum system for use in plume observations. It is
important to recognize, however, that the SRI Mark V lidar is a rela-
tively modest unit, and that with current technology considerably higher
performance is readily attainable. The results of the present investi-
gation, even allowing for the limitations of the equipment, are so
73
-------
encouraging that the development of a more advanced lidar system appears
to be fully justified.
Such a system could be readily developed with state-of-the-art
technology based upon the experience of the present study. In selecting
an optimum system, it would be necessary to investigate carefully the
relative advantages of lasers of different wavelengths. Prime candidates
are 0.694 [a (ruby), 0.9 |_i (gallium arsenide), near infrared at 1.06 n
(neodymium glass and YAG—ytrrium aluminum garnet), or 0.53 [i (the second
harmonic of the latter), all of which are immediately available. In
making the choice, the factors of pulse repetition rate, availability of
adequately sensitive detectors, and, of course, cost would all have to
be carefully weighed.
Although at the moment indications are that a well-cooled ruby
laser, system with a PRF of 1 to 10 pulses per second would provide the
best compromise, a detailed systematic study would be a most valuable
preliminary to building an advanced lidar system for plume studies.
• A study should be made of the feasibility of calibrating the
lidar on an absolute basis so that the energy output (in joules) of each
pulse were known. A completely calibrated lidar could furnish objective
estimates of one of the attenuation solution parameters, M , the abso-
lute particulate mass concentration in the clear air.
B. Data Handling and Reduction Capabilities
• Even at the present lidar firing rates, the large quantity of
data collected is on the verge of becoming unmanageable. Each lidar
observation is not just a point measurement, but a complete profile.
Making full use of this information, particularly when using higher-PRF
instruments, requires new and better ways of handling and processing
lidar data.
*
Recent tests with the SRI Mark V ruby system indicate that relatively
modest improvements could permit it to be operated at a pulse rate
comparable to that of the neodymium version, which could make this
system an excellent interim solution.
74
-------
• A big step in the right direction would be the development of
an intensity-modulated range-height oscilloscope display of plume cross
sections. This would be similar to the familiar radar RHI display, as
simulated in Fig. 34, and would give an immediate, operationally useful
photograph showing the plume shape and boundaries, as well as the in-
ternal relative concentration distribution in terms of varying degrees
of brightness. For a more quantitative record, the video magnetic disk
recorder would be used to take photos of the same cross section with
(Each line represents one lidar shot;
trace intensity is a function of plume
paniculate concentration)
HORIZONTAL DISTANCE FROM LIDAR
TA-7289-39
FIGURE 34
VISUALIZATION OF INTENSITY-MODULATED RANGE/HEIGHT
OSCILLOSCOPE DISPLAY OF PLUME CROSS SECTION
OBSERVED BY LIDAR
different oscilloscope intensity settings, giving a contour representa-
tion. Such a display would require that each lidar observation be cor-
rected electronically for range and for equipment nonlinearities. These
real-time analog techniques would greatly simplify the data-processing
problem and reduce the number of digital computer analyses required.
• Since some of the lidar data will always need to be retained
in quantitative form, an on-line analog-to-digital conversion and re-
cording capability in the field would vastly facilitate later digital
analyses.
75
-------
C . Experimental and Analytical Techniques
• For better time continuity of the lidar plume observations,
the limitations on duration of the observation periods, which were in
effect during the experimental phase of this study, are undesirable and
future arrangements should permit uninterrupted operations, if at all
possible.
• Observations should also be obtained during the night hours,
when lidar performance is optimum due to the low background illumination
(noise) levels.
• Most of the effort in this study went into lidar data collec-
tion, processing, and analysis. More work is needed on the interpreta-
tion and use of plume cross sections in checking plume rise formulas,
testing diffusion models, studying terrain effects, etc.
• In particular, a rationale is needed to deal with the effect
of the background noise threshold in plume cross sections upon the
testing of diffusion models using lidar data. This methodology could be
similar to that developed by Gifford (1959) for use in diffusion esti-
mates from smoke plume photography.
• Some provision for time averaging of the lidar data is required
for use in testing diffusion models, which generally are designed to
furnish one-half- to one-hour averages. This capability is especially
needed for checking models for neutral and unstable conditions. It is
possible that an analog means to do this can be devised using the pre-
viously discussed intensity-modulated range-height display.
• In future absolute concentration analyses, an attempt should
be made to treat £ , the ratio of the optical volume extinction coeffi-
cient to particle mass, as a variable with range.
D. Advanced Lidar Concepts
• There is great potential for exploiting the wavelength charac-
teristics of laser energy in lidar applications for plume studies. Two
possibilities warrant early investigation:
76
-------
(1) The use of multiwavelength systems to determine
absolute mass concentrations without independent
information of the particle size distribution,
and
(2) The extension of the multiwavelength differential
absorption technique that has been applied to the
determination of water-vapor (Schotland et al.,
1962) to the evaluation of plume constituents.
In this connection the development of new types of lasers, such as the
tunable dye laser, holds great promise.
77
-------
REFERENCES
Barrett, E. W., and O. Ben-Dov. 1967: Applications of the lidar to
air pollution measurements. J. Appl. Meteor., 6, 500-515.
Gifford, F. A., Jr., 1959: Smoke plumes as quantitative air pollution
indices. Intern. J. Air Poll., 2, 42-50.
Gifford, F. A., Jr., 1961: Use of routine meteorological observations
for estimating atmospheric dispersion. Nuclear Safety, 2, 48.
Goetz, A., 1969: A new instrument for the evaluation of environmental
aerocolloids. Envir. Science and Tech., 3, 154-160.
Hamilton, P. M., 1966: The use of lidar in air pollution studies.
Intern. J. Air and Water Poll., 10, 427-434.
Johnson, W. B., Jr., 1969: Lidar applications in air pollution research
and control. J. Air Poll. Contr. Assoc., 19, March 1969.
McCormick, R. A., and K. R. Kurfis, 1966: Vertical diffusion of aerosols
over a city. Quart. J. Roy. Meteor. Soc., 92, 392-396.
Noll, K. E., P. E. Mueller, and M. Imada, 1968: Visibility and aerosol
concentration in urban air. Atmos . Envir., 2, 465-475.
Proundfit, B. W., 1969: Plume rise from Keystone Plant. Final Report,
Contract PH 86-68-94, Sign X Laboratories, Inc., Essex, Connecticut.
Schotland, R. M., A. M. Nathan, E. A. Chermack, and E. E. Uthe, 1962:
Optical sounding. Tech. Rpt., Contract DA 36-039 SC-87299, U.S.
Army Electronics Research and Development Laboratory.
Smith, M. E. (Editor), 1968: Recommended Guide for the Prediction of
the Dispersion of Airborne Effluents. American Society of Mechanical
Engineers, 345 E. 47th Street, New York.
79
-------
Appendix A
DESCRIPTION OF MARK V LIDAR
81
-------
Appendix A
DESCRIPTION OF MARK V LIDAR
The optical system of the Mark V lidar is illustrated in Fig. A-l.
The transmitter consists of a Q-switched, air-cooled, pulsed ruby or
neodymium-doped glass laser with wavelengths at 6943 A (deep red) or
at 10,600 A (near infrared), respectively. Since the angular resolution
of the lidar is determined by the transmitted beam divergence, 6-inch-
diameter collimating optics are used to reduce the laser beam divergence
and to produce an output beamwidth of 0.35 mrad. The corresponding
spatial resolution of this beam is 0.50 m at a range of 1 km in the
cross-beam direction, and about 2.3 m in range. The pulse repetition
rate of the Mark V lidar is primarily limited by the cooling rate of the
laser cavity. At present, the maximum firing rate is 3 to 4 pulses per
minute for the ruby laser, and 10 to 12 pulses per minute for the neo-
dymium. The characteristics of the Mark V lidar are summarized in Table
A-I.
The lidar receiver consists of a 6-inch-diameter Newtonian tele-
scope, identical to the transmitter optics. An adjustable field stop
at the focal plane limits the receiver field of view to a maximum of 6
mrad. A multilayered narrow-band filter with a wavelength interval
(bandwidth) of 13 A (ruby) or 100 A (neodymium) is inserted in the re-
ceiver optical path to reduce the output noise level produced by solar
radiation scattered into the receiver field of view. The detector con-
sists of an RCA 7265/S-20 photomultiplier tube for the ruby laser, and
an RCA 7102/S-l for the neodymium system.
The major electronic components of the lidar and the data-recording
system are illustrated in the block diagram shown in Fig. A-2. A
compressed-air-driven turbine rotates the laser Q-switching prism at
* 4 o
10 A = 1 l-i.
83
-------
COLLIMATING LENS
NARROW-BAND FILTER
RECEIVER SIGNAL
PRISM
FIELD STOP
(ADJ USTABLE)
PRIMARY
MIRRORS
oo
PHOTOMULTIPLIER
LIGHT GUIDE TO
OPTICAL ATTENUATOR
DIVERGING LENS
FIBER OPTICLIGHT
GUIDES
LASER ROD
_ ;f^_ ^_ ^
FLASH LAMPS
(2)
PARTIALLY REFLECTING MIRROR
(FABRY-PEROT ETALON)
REFERENCE PATH TO
COLLIMATING LENS
CALIBRATED
OPTICAL ATTENUATOR
0-45dB
RIGGER PULSE TO OSCILLOSCOPE
SILICON PHOTO DIODE
TA-7289-40
FIGURE A-1 OPTICAL SYSTEM OF MARK V LIDAR
-------
Table A-I
CHARACTERISTICS OF SRI MARK V LIDAR
Characteristics
Transmitter
Laser Material
Beamwidth (mrad)
Optics
Peak Power Output (MW)
Pulse Length (ns)
Q Switch
Maximum PRF (pulses/min)
Receiver
Optics
Field of View (mrad)
Predetection Filter
o
Wavelength Interval (A)
Detector
Postdetection
Filter Bandwidth (MHz)
Wavelength
6943 A
Ruby
0.35
10,600 A
Neodymium-Glass
0.2
6-inch f/4 Newtonian Reflector
15
15
Rotating Prism
3 to 4
50
12
Rotating Prism
10 to 12
6-inch Newtonian Reflector
1.5
13
RCA 7265
(S-20 Cathode)
30
3.0
100
RCA 7102
(S-l Cathode)
30
500 revolutions per second. Upon receipt of a fire signal, the synchro-
nizing generator triggers the flash lamp in step with a signal from the
rotating prism. A capacitor bank charged to 3 kV supplies energy for
the laser flash lamps. A photodiode senses the occurrence of the laser
pulse and produces a trigger pulse to start the oscilloscope sweep. The
output of the photomultiplier in the lidar receiver is fed to a pulse
amplifier having a logarithmic transfer function, and then to an oscillo-
scope. A Polaroid recording camera mounted on the oscilloscope photo-
graphs the lidar return signal. Other data-recording options available
include the use of an automatic recording 35-mm camera and a magnetic
disk video recorder. The latter device furnishes a steady display on
the face of the oscilloscope immediately after each shot, as well as a
permanent record, and is very useful for real-time monitoring of
observations .
85
-------
FLASH LAMP
ROTATING PRISM v \
.LASER ROD
,BEAM SPLITTER
J-fl— -X---X
FIRE
PULSE
^OUTPUT RADIATION
•PHOTO DIODE
00
a
BACKSCATTERED
SIGNAL RETURN"
PHOTOMULTIPLIER
i
HIGH VOLTAGE
POWER
SUPPLY
LOGARITHMIC
AMPL IFIER
VIDEO DISC
RECORDER
1
SWEEP TRIGGER
I
OSCILLOSCOPE
»
MONITOR
OSCILLOSCOPE
35 MM
RECORDI NG
CAMERA
POLAROID
* CAMERA
TA-7289-41
FIGURE A-2 ELECTRONICS SYSTEM OF MARK V LIDAR
-------
Although the pulse-to-pulse variations of laser power output are
not unduly severe, the presence of this random variation introduces
a corresponding uncertainty in the amplitude of the received signal un-
less provision for monitoring the transmitter power is made. The Mark V
lidar incorporates an optical feedback arrangement, using fiber-optic
light guides, that samples the transmitted energy and injects it into
the receiver optical path ahead of the narrow-band predetection filter
(see Fig. A-l). This arrangement produces a pulse occurring at zero
range on the receiver output. The pulse height is proportional to trans-
mitted power. A variable neutral-density filter is inserted in the
fiber-optic path to allow adjustment of the pulse height.
The use of a logarithmic video amplifier in the receiver is almost
essential in order to compress the wide dynamic range of the detector
(typically four or more orders of magnitude) to enable the received
signal to be displayed on a single oscilloscope trace without loss of
detailed information.
87
-------
Appendix B
DETAILED LIDAR DATA SUMMARIES
89
-------
Table B-I
DATA SUMMARY, 25 MAY-1 JUNE 1968
Date
(1968)
5/25
5/26
5/27
5/28
Site
No.
1
2
3
4
Observational
Period
No.
1
2
3
4
5
6
7
8
Cross
Section
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Azimuth
Angle
045
000
315
000
045
315
000
045
315
045
200
226
252
253
226
200
200
226
226
226
226
065
032
000
065
032
000
065
032
000
065
032
000
020
Elevation
Angle
(First)
13
25
21
30
16
23
89
89
89
60
9
16
20
15
15
9
13
30
30
30
30
12
25
24
15
25
24
15
25
20
15
25
24
89
Time
(Start)
(EOT)
0912 :30
0917:30
0922 :00
1034:00
1040:15
1045:00
1423:00
1431:00
1437:00
1448:00
0854:00
0858:00
0904:30
0908:45
0915:45
0921 :00
1145:30
1151 :30
1154:30
1200:30
1205:00
0908:10
0913:15
0918:00
0923:00
0928:00
0932 :20
1013:45
1019:45
1024:00
1035:30
1040:30
1045:45
0909:45
Elevation
Angle
(Last)
6
11
7
4
4
4
7
5
5
3
1
3
5
5
6
1
5
22
4
4
4
3
1
1
2
1
2
2
2
0
2
2
3
Time
(End)
(EOT)
0916:00
0921:00
0925:25
1039:45
1043:30
1048:00
1428:00
1435:00
1442:15
1453:00
0856:15
0902 :40
0906:30
0913:00
0918:00
0926:00
1148:30
1153:00
1158:30
1204:15
1208:30
0912:00
0916:45
0922:00
0927:00
0931 :30
0935:20
1018:15
1022 :45
1026:45
1039:30
1043:45
1049:00
0913:00
Elevation
Angle
Increment
1/2
1
1
1
1
1
3
3
3
3,1
1/2
1/2
1
1/2
1/2
1/2
1/2
1
1
1
1
1/2
1
1
1/2
1
1
1/2
1
1
1/2
1
1
3
No. of
Lidar
Shots
15
15
15
27
13
20
28
29
29
28
17
26
16
20
19
29
29
9
27
27
27
19
25
24
27
25
24
28
24
21
30
24
23
30
Subtotal
No. of
Shots
219
465
759
Weather, Remarks
0800 : O 1/4 GF C
0900 : O 7 L/V
Plume Hdg ~ 245°
1400: /© 15 140/5
Plume Looping
0800: 100©/dDlO 090/5
Plume Hdg ~ 315°
1000: 1000 vd)/© 7
090/10
Plume Looping
0800 :E25 0 4R-090/15G25
Plume Hdg 270-290°
0800 :E35®8 090/5
-------
Table B-I (Continued)
Date
(1968)
5/28
5/29
5/30
5/31
6/1
Site
No.
3
5
2
6
6
Observational
Period
No.
9
10
11
12
13
14
15
Cross
Section
No.
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
Azimuth
Angle
065
032
000
333
065
032
000
333
350
030
070
030
350
350
020
117
117
117
228
199
170
199
228
199
170
170
350
215
305
260
Elevation
Angle
(First)
20
25
25
25
20
25
25
25
15
15
12
12
20
20
20
28
25
28
10
15
15
15
15
20
20
90
85
90
90
90
Time
(Start)
(EOT)
1113:30
1116:30
1121:00
1126:40
1135:45
1139:45
1143:30
1148:20
0855:30
0900:00
0905:00
0912:15
1208:15
1210:10
1213:30
0844:00
0945:15
1018:15
0820:30
0824:45
0828:15
0831:15
0935:15
0940:15
0944:45
1018:00
1025:00
1030 :00
1034:00
1041 :00
Elevation
Angle
(Last)
3
3
3
4
3
2
2
4
2-1/2
2
4
2
2
2
2
4
5
4
1-2/3
2
3-1/2
1-1/2
3-1/3
3
3-1/2
15
6
12
6
5
Time
(End)
(EDT)
1116:00
1119:45
1123:30
1128:00
1138:00
1142:20
1147:00
1150:00
0859:00
0903:45
0908:15
0915:15
1209:30
1212:30
1215:45
0848:00
0948:30
1021:15
0823:30
0827:15
0830:40
0824:25
0938:45
0943:45
0948:15
1020:45
1027:30
1032:15
1037:00
1043:30
Elevation
Angle
Increment
1
1
1
1
1
1
1
1
1/2
1/2
1/2
1/2
2
1
1
1
1
1
1/3
1/2
1/2
1/2
1/2,1/3
1/2
1/2
5
5,3,2,1
5,3
5,3,1
5,3,2,1
No. of
Lidar
Shots
19
23
23
22
18
24
24
22
26
28
17
22
10
19
19
25
23
25
27
27
24
28
34
29
18
23
19
24
25
Subtotal
No. of
Shots
964
1105
1178
1378
Weather, Remarks
(Plume overhead,
shifting EWD, so moved
to Site 3)
Plume turning into water
cloud
1200:E12©3R — F 090/10
0830 :E50 © 5HGFC
Plume moving to ESE
0845; -X10 © 1/4R-FC
0945: E10 © 2R — FC
1015: E8 © 50 © 3R—
F 200/4
0800: E10 © 5H 220/3
Plume becoming water
cloud
0900: A 13 (D 25 © 5H
(Chopper report)
Ground fog till 1000
Plume overhead
Plume rising almost
vertically and then
spreading horizontally —
wind light/variable
Total No. of Shots 1487
CD
CO
-------
Table B-II
DATA SUMMARY, 15-24 OCTOBER 1968
Date
(1968)
10/15
10/16
10/17
Site
No.
6
16
16
Observational
Period
No.
1
2
3
4
5
Cross
Section
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Azimuth
Angle
270
300
330
270
300
330
270
300
330
300
300
300
300
300
145
100
55
100
100
100
255
230
230
230
230
090
090/270
090/270
090/270
090/270
090/270
090/270
090/270
Elevation
Angle
(First)
45
50
50
40
40
36
30
50
36
40
44
70
70
70
90
90
90
90
90
90
30
30
36
15
20
15
15
15
15
15
3
5
15
Time
(Start)
(EOT)
0932:00
0939:00
0945:30
0955:00
0959:00
1004:40
1032:30
1036:30
1044:00
1059:20
1116:00
1138:00
1143:00
1147:00
0852 :20
0916:00
0920:20
0934:00
0938:30
0943:15
1149:30
1153:20
1157:30
1200:10
1204:00
0903:30
0910:00
0916:30
0928:00
0936:00
1151:15
1201:45
1212:00
Elevation
Angle
(Last)
3
4
2
2
2
2
Time
(End)
(EOT)
0936:04
0943:11
0949:40
0957:50
1002:25
1007:25
2 1034:40
2 1041:00
1
2
2
2
1
1
1
40
5
45
5
5
6
2
2
1
1
1
90
6
6
5
5
5
5
5
1047:00
1102 :20
1118:50
1141 :30
1145:45
1150 :00
0855 :00
0918:15
0923:00
0937:00
0941 :50
0946:20
1151 :30
1155 :45
1159:00
1203:00
1206:25
Msg.
0914:25
0922 :20
0934:35
0942 ;20
1157:15
1207:10
1216:25
Elevation
Angle
Increment
2
2
2
2
2
2
2
2
2
2
2
3,2
3
3,2
5
5
5
5
5,2
5,2
2
2
2,1
1
1
5
5,2
5,2
5,2,1
5,3,2
5,3,1
5,4,3,2
5,4,3,2
No . of
Lidar
Shots
22
26
26
20
20
18
15
25
18
20
22
25
24
25
11
18
12
18
24
26
15
15
19
15
20
16
33
34
35
33
33
29
27
Subtotal
No. of
Shots
306
499
739
Weather, Remarks
0915 :Q 3 GF
Plume Hdg 022 ;
Strong vertical wind
shear in direction
1156: O 5-7H
light /variable
0945; O 7 HK 360/5
1145: O 8H 135/8
Plume Hdg 360=
0927: E60(D/dD8 090/5
Plume overhead
1216: 12008 060/8
CO
CO
-------
Table B-II (Continued)
Date
(1968)
10/18
10/20
10/21
10/22
Site
No.
17
21
22
6
Observational
Period
No.
6
7
8
9
10
11
Cross
Section
No.
34
35
36
37
38
39
40
41
42
43
44A
44B
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
Azimuth
Angle
080/260
080/260
080/260
080/260
200/020
200/020
200/020
200/020
200/020
200/020
270
225
180
270
225
180
225
180
225
270
225
270
225
180
270
300
330
270
300
330
270
300
330
Elevation
Angle
(First)
3
3
3
3
3
45
14
14
2
20
24
26
16
24
26
16
26
16
26
24
26
45
60
60
15
16
15
15
16
15
15
15
15
Time
(Start)
(EDT)
0917:20
0928:20
0935:20
0941:00
0909:00
0916:15
0928:00
0937:00
1018:00
1038:30
0854:00
0856:40
0859:30
0903:00
0906:30
0909:30
0921:30
0925:00
0928:00
0930:30
0934:00
1023:00
1027:30
1033:20
0900:00
0903:00
0905:20
0921:45
0923:50
0925:45
0938:30
0940:35
0943:45
Elevation
Angle
(Last)
3
3
3
3
2
2
2
2
1
10
5
8
3
5
6
4
9
5
9
7
9
2
3
2
2
3
2
3
3
3
2
2
2
Time
(End)
(EDT)
0922:15
0932:30
0939:45
0944:20
0914:45
0919:00
0932 :00
0941:30
1024:20
1040:50
0855:40
0858:35
0901:20
0904:40
0908:20
0910:50
0923:40
0926:00
0929:40
0932:15
0935:40
1025:55
1030:30
1036:50
0901:20
0904:30
0907:30
0923:10
0925:10
0927:00
0939:45
0942 :00
0945:15
Elevation
Angle
Increment
5,4,3
5,4,3,2
5,4,3
5,4,3
5,4,3,2
5,3,2
5,2
5,2
5,2
10
2,1
2,1
1
2,1
2,1
1
2,1
2,1
2,1
2,1
2,1
3,2,1
5,3,2,1
5,3,2,1
1
1
1
1
1
1
1
1
1
No. of
Lidar
Shots
34
33
33
33
35
24
34
34
42
16
17
18
14
16
18
13
17
10
18
14
15
24
26
27
14
14
14
13
14
13
14
14
14
Subtotal
No. of
Shots
872
1057
1304
Weather, Remarks
0900: SOQD/dDs 110/8
Plume overhead
1100: 15® 80 CD/CDS
130/10
0900 : Ql5
0945: O I5 160/7
1015: O 5H
1100 : O 5H 220/10
0830: 50Q6H 270/3
0945: 400V(H)6H 135/5
1100: 25(D50(|)8 290/10
0840: /0 7 GF
Plume Hdg 020°
1015: /(D 7 080/5
1015: undulations on
top of plume
CD
-------
Table B-II (Continued)
Date
(1968)
10/22
10/23
10/24
Site
No.
6
11
Observational
Period
No.
12
13
14
15
16
Cross
Section
No.
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
Azimuth
Angle
270
300
330
270
300
330
270
300
330
300
300
300
220
220
180
140
220
180
140
220
180
220
180
140
190
180
110
070
030
110
070
030
070
070
070
070
Elevation
Angle
(First)
20
20
20
20
20
30
26
30
40
34
30
30
16
18
26
24
20
20
26
22
26
24
30
45
50
90
30
34
26
22
30
34
30
26
34
34
Time
(Start)
(EDT)
1134:00
1137:00
1139:15
1151:50
1154:15
1158:45
1227:00
1231:00
1234:45
1238:30
1241:15
1243:45
0850:15
0856:45
0859:45
0902:40
0909:00
0912:45
0916:00
0941:20
0946:30
1155:00
1158:45
1206:00
1218:00
1223:45
0930:00
0933:15
0936:00
0938:30
0941:00
0943:30
1002:00
1004:40
1017:45
1020:00
Elevation
Angle
(Last)
2
3
3
1
0
3
1
1
2
2
2
2
1-1/2
2
2
5
2
1
5
2
2
2
1
5
1
2
3
2
5
5
4
2
5
5
9
6
Time
(End)
(EDT)
1136:10
1138:30
1140:40
1153:25
1157:30
1201:50
1229:15
1233:00
1236:55
1240:30
1242:50
1245:05
0853:15
0858:30
0901:45
0904:45
0911:20
0915:50
0918:00
0943:45
0949:40
1157:40
1201:20
1209:30
1220:30
1227:10
0932:30
0934:40
0937:05
0940 :00
0942:30
0945:20
1003:40
1006:00
1019:30
1021:45
Elevation
Angle
Increment
2,1
2,1
2,1
2,1
1
2,1
2,1
2,1
3,2,1
2
2
2
1,1/2
1
2,1
2,1
1
1
2,1
2,1
2,1
2,1
2,1
5,3,2,1
3,2,1
5,3,2
2,1
2
2,1
2,1
2
2
2,1
2,1
2,1
2,1
No. of
Lidar
Shots
17
15
15
18
25
27
19
20
20
17
15
15
28
17
17
18
19
20
18
20
21
19
23
17
22
27
21
17
16
15
14
17
17
15
17
20
Subtotal
No. of
Shots
1665
1951
Weather, Remarks
1140: /(J)7 090/10
1050: Plume looping
1205: /
-------
Table B-II (Continued)
Date
(1968)
10/24
Site
No.
11
Observational
Period
No.
16
Cross
Section
No.
102
103
104
105
106
107
108
109
110
Azimuth
Angle
070
070
070
070
070
070
070
070
070
Elevation
Angle
(First)
30
30
38
36
46
46
90
90
90
Time
(Start)
(EDT)
1032:30
1035:15
1047:30
1051:30
1103:30
1117:30
1134:00
1145:45
1202:00
Elevation
Angle
(Last)
9
8
5
6
4
2
8
4
4
Time
(End)
(EDT)
1034:10
1037:50
1050:30
1053:25
1105:45
1119:50
1136:30
1151:35
1204:45
Elevation
Angle
Increment
2,1
2,1
2,1
2
2
2
10,5
3,2
10,5
3 2
°,"
10 5
3,2
No. of
Lidar
Shots
15
22
28
16
22
23
23
25
25
Subtotal
No. of
Shots
2319
Weather, Remarks
1150: 80(J)l60©5H
180/10
co
en
-------
Appendix C
LIDAR DATA DIGITIZING AND CONDITIONING DETAILS
97
-------
Appendix C
LIDAR DATA DIGITIZING AND CONDITIONING DETAILS
(by C. E. Brabant)
1• Photographic Processing
The lidar returns were recorded on Polaroid photographs and on
magnetic disk tracks in analog form. These signals were then repro-
duced from the magnetic recordings and displayed repetitively on the
screen of a standard laboratory oscilloscope (Tektronix Model 555). A
photographic transparency on 35-mm film was made of each oscilloscope
trace with an illuminated graticule superimposed on the background to
provide reference axes for scaling the photograph. The time scales
used in the photographic frames were 2 )j,s/cm (3-km range full scale),
and when appropriate, 5 p,s/cm (7.5-km range full scale). The vertical
amplitude scale was set at 100 mV/cm.
2. Graphical-to-Digital Conversion
A Model 300 CALMA Co. digitizer was used to convert the optically
enlarged photographic trace to numerical increments recorded on digital
magnetic tape. The data on the photographs were automatically digitized
when the cursor was manually traced over the image of the lidar return
signal amplitude. Scaling was performed at the beginning of each
digitizing-recording session and immediately preceding a change in
scale whenever this occurred.
Engineering values in millivolts and microseconds were entered via
the digitizer keyboard, as were all other special control characters.
These were recorded on magnetic tape along with the records of the in-
cremental digital values describing the X and Y Cartesian coordinates
of the analog signal. These values were recorded for equal increments
in time at a resolution of approximately 25 ns, corresponding to 3.8 m
in range. The vertical scale resolution was 0.75 mV. General accuracy
obtained in curve tracing is estimated to be no worse than one-half of
99
-------
the trace width, which varied from approximately 10 to 20 mV, although
it is possible that larger deviations from the curve occurred infrequently.
3. Transformation to Engineering Units and Conditioning
A standard CDC 6400 computer program developed by the CALMA Co. was
used to recover the data, scaled in engineering units. Much of the data
required special handling because of operator errors in entering control
characters via the keyboard, foot-treadle, and console pushbuttons of the
digitizer. This step produced the first intermediate digital magnetic
tape and a control listing of the decimated data points. This listing
also provided data for verifying the scaling factors used to convert
incremental values into scaled engineering values.
A pair of special CDC 6400 computer programs were developed to
select data between predetermined starting and ending ranges (travel
time values for lidar return signals) at equally spaced intervals.
Adequate spatial resolution was obtained by selecting every fifth point
in the data record, yielding an increment of approximately 100 ns (15-m
range) between selected sample points. The second program of the pair
listed the entire output tape, point-by-point, as a check for subsequent
processing.
100
-------
Appendix D
PLAN VIEWS OF HORIZONTAL PLUME POSITIONS
FROM LIDAR OBSERVATIONS
101
-------
Appendix D
PLAN VIEWS OF HORIZONTAL PLUME POSITIONS
FROM LIDAR OBSERVATIONS
To help clarify the plume cross sections from the October period
that are presented in Sec. V, plan views of horizontal plume positions
derived from these cross sections are shown on the following pages
(Figs. D-l to D-9) . By referring to these plan views, the elongation
in the cross sections caused by the lidar azimuth angle being nonper-
pendicular to the plume axis may be determined.
r-
0"
KEYSTONE
STACKS
FIGURE D-1 PLAN VIEW OF HORIZONTAL PLUME POSITIONS, 15 OCTOBER 1968
Cross section numbers and times in EOT are shown beside each horizontal
plume position.
103
-------
\
LIDAR SITE
No. 16
20 (0945)
\
\
\l
\l
0
KEYSTONE
STACKS
i
I
I
I
I
I
I
I
I
16 OCTOBER 1968
I l I
1
SCALE—km
TA-7289-43
FIGURE D-2 PLAN VIEW OF HORIZONTAL PLUME POSITIONS, 16 OCTOBER 1968
104
-------
30 (0939)
27 (0912)
17 OCTOBER 1968
i i i l
I I
0 1 2
SCALE—km
TA-7289-44
FIGURE D-3 PLAN VIEW OF HORIZONTAL PLUME POSITIONS, 17 OCTOBER 1968
105
-------
37 (0943)
34 (0920)
i I l i
18 OCTOBER 1968
0 1 2
SCALE—km
TA-7289-45
FIGURE D-4 PLAN VIEW OF HORIZONTAL PLUME POSITIONS, 18 OCTOBER 1968
106
-------
KEYSTONE
STACKS
40
(0930) 42
(1021)
20 OCTOBER 1968
SCALE—km
10
TA-7289-46
FIGURE D-5 PLAN VIEW OF HORIZONTAL PLUME POSITIONS, 20 OCTOBER 1968
For those observation series where the azimuth angle was held con-
stant and the plume structure observed as a function of time, the first
cross section in the series is shown at the proper position, and subse-
quent cross sections are displaced slightly for graphic purposes to
avoid superposition.
The horizontal plume boundaries interpolated between cross sections
are indicated by dashed lines in the figures. It should be remembered
that these illustrations show the total horizontal projection of the
plume. At any given height, the plume would normally be narrower than
this.
The cross section number and time (EDT) are shown beside each
horizontal plume position.
107
-------
o
oo
LIDAR SITE
No. 22
48
(0910)
TB-7289-47
FIGURE D-6 PLAN VIEWS OF HORIZONTAL PLUME POSITIONS, 21 OCTOBER 1968
-------
LIDAR SITE
No. 6
63
(0939)|- /* LIDAR SITE
; / No. 6
//
//
// ''
(*( KEYSTONE Q
STACKS
1 IMll 1 1
72
(1228)
/
/
/
KEYSTONE
STACKS
/
/
/
LIDAR SITE
No. 6
22 OCTOBER 1968
1
TA-7289-48
FIGURE D-7 PLAN VIEWS OF HORIZONTAL PLUME POSITIONS, 22 OCTOBER 1968
109
-------
23 OCTOBER 1968
LIDAR SITE
No. 6
\-
KEYSTONE S' '
STACKS^ ' — -
0^---
III! 1
^
— 79
(0857)
1
J
80
(0901)
|
SCALE—km
TA-7289-49
FIGURE D-8 PLAN VIEW OF HORIZONTAL PLUME POSITIONS, 23 OCTOBER 1968
110
-------
109 (1149)
108 (1135)
107 (1119)
106 (1105)
104 (1049)
103 (1036)
101 (1021)
-99 (1005)
93 (0934)
LIDAR SITE
No. 11
24 OCTOBER 1968
0 1
SCALE—km
©
KEYSTONE
STACKS
TA-7289-50
FIGURE D-9 PLAN VIEW OF HORIZONTAL PLUME POSITIONS, 24 OCTOBER 1968
111
-------
Appendix E
LIDAR EYE-SAFETY CONSIDERATIONS
113
-------
Appendix E
LIDAR EYE-SAFETY CONSIDERATIONS
Although often exaggerated, the question of safety in the use of
experimental lidar systems should be carefully considered.
The series of observations described in this report were carried
out with careful adherence to the safety procedures that have been de-
veloped and practiced by SRI teams over the last six years of field
experiments. Some of the more important aspects of the safety question
are described below.
The Mark V lidar pulse is harmless to the skin, but could cause
eye damage in the form of a minute retinal lesion if the pulse were
viewed directly. The lidar pulse takes the form of a cylinder of light,
initially 15 cm in diameter and 6 m long, traveling through the atmo-
sphere at the speed of light. Each pulse is only 0.00000002 s (0.02 |_LS)
in duration, and the pulses occur approximately 5 s apart. The lidar
is designed to totally prevent any emission of stray laser light from
any direction other than that in which the unit is pointed.
Thus an eye hazard exists only at the exact instant of firing and
only in the precise direction the lidar is aimed. The lidar is highly
directional: the increase in diameter of the pulse with range (travel
distance) is very slight—from 15 cm at the transmitter to 1.3 m at two
miles range—so that the spatial extent of the hazard is confined to a
very narrow cone. In this sense, the hazard from a lidar is directly
analogous to that from a rifle or other firearm, except that the magni-
tude of the danger is considerably less for the lidar. This is because
the pulse can only cause eye damage, and even then, only a partial (lo-
calized) loss of vision could result. Generally this loss would take the
form of a blind spot ten times smaller than the natural one already present
(and unnoticed) in every human eye at the location of the optic nerve.
Our standard general safety procedures for field experiments (which
have proven worthy by use in many projects) are as follows:
115
-------
(1) Mechanical stops are placed on the equipment to con-
fine the pointing angles to a specific field of view.
(2) Aiming and firing is accomplished manually by an
operator who uses an aiming telescope to monitor
the field of view, as well as a large surrounding
area. This operator is the only person who can
fire the lidar, and he does not trigger the unit
until he is sure the area is clear.
(3) A separate observer equipped with binoculars also
keeps the entire firing area under surveillance.
This observer is equipped with an override switch
to prevent firing of the lidar when any human
activity (airplane, ship, etc.) is sighted in the
firing area.
(4) When not being fired, the lidar is always rendered
completely safe by a switch which discharges the
capacitor bank furnishing high voltage to the laser
flash lamps. Without this high voltage supply, the
lidar cannot be fired.
(5) As an additional precaution, the stowage position
of the lidar is such that it is never pointed at
eye level.
116
------- |