EPA-600/2-77-009
January 1977
Environmental Protection Technology Series
CARBON DIOXIDE LASER SYSTEM TO MEASURE
GASEOUS POLLUTANTS
Environmental Sciences Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into five series. These five broad
categories were established to facilitate further development and application of
environmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
This report has been assigned to the ENVIRONMENTAL PROTECTION
TECHNOLOGY series. This series describes research performed to develop and
demonstrate instrumentation, equipment, and methodology to repair or prevent
environmental degradation from point and non-point sources of pollution. This
work provides the new or improved technology required for the control and
treatment of pollution sources to meet environmental quality standards.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/7-77-009
January 1977
CARBON DIOXIDE LASER SYSTEM
TO MEASURE GASEOUS POLLUTANTS
by
R.J. Gilltneister and L.R. Snowman
General Electric
Pittsfield, Massachusetts 01201
Contract No. 68-02-1290
Project Officer
William A. McClenny
Atmospheric Instrumentation Branch
Environmental Sciences Research Laboratory
Research Triangle Park, NC 27711
ENVIRONMENTAL SCIENCES RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NC 27711
-------
DISCLAIMER
This report has been reviewed by the Environmental Sciences Research
Laboratory, U.S. Environmental Protection Agency, and approved for
publication. Approval does not signify that the contents necessarily
reflect the views and policies of the U.S. Environmental Protection
Agency, nor does mention of trade names or commercial products con-
stitute endorsement or recommendation for use.
-------
ABSTRACT
Reported here Is the continuation of work in the development of a
gas laser system for air pollution monitoring over long paths, a kilometer or
more, using Infrared absorption. Discussed are modifications to a bread-
board system for simultaneous detection of ozone, ammonia and ethylene and
the addition of beam steering optics to give the system area monitoring
capability. Operation for a two month period in St. Louis in conjunction with
the RAPS program is also discussed. Data comparing system performance
with that of conventional monitors Is presented along with the results of prob-
lem investigations. While reasonable correlation with point monitor results
is Indicated, it Is concluded that system performance can be Improved.
Recommendations are given for further activity to achieve better system
performance.
111'
-------
IV
-------
CONTENTS
Abstract 11J
List of Figures vi
List of Tables viii
Acknowledgements : ix
Sections
I Conclusions 1
II Recommendations 3
HI Introduction 5
IV System Modifications 7 .
V Field Measurements 25
VI Discussion 49
VII References 79
Vni Appendices 81
A ILAMS System Description 81
B Pollutant Concentrations Detectable with Gas Laser 91
Long Path Sensors
C Laser Long Path Monitoring Applications 97
-------
LIST OF FIGURES
Figure Page
1 Resultant Ethylene Line Selection From 25 Iteration LWSP Run 13
2 Resultant Ammonia Line Selection From 25 Iteration LWSP Run 14
3 Resultant Ozone Line Selection From 25 Iteration LWSP Run 15
4 CMFIL Output Listing 16
5 Spectrally Scanning Laser Design Concept 21
6 Layout of Monitoring Path at RAMS Site 103 26
7 View of Monitoring Path from Laser to Retro 28
8 View of Monitoring Path from Retro to Laser 28
9 Ozone Count Variation With Time - 6 Sep 74 31
10 Ozone Count Variation With Time - 3 Dec 73 ,32
11 Comparative Path/Moving Point Monitor Ozone Data - 3 Oct 74 33
12 Comparative Path/Moving Point Monitor Ozone Data - 3 Dec 73 34
13 Path and Point Monitor Ozone Concentrations - 9 Oct 74 37
14 ILAMS Ozone and Ammonia Concentrations - 9 Oct 74 39
15 Path and Point Monitor Ozone Concentrations - 10 Oct 74 41
16 ILAMS Ozone and Ammonia Concentrations - 10 Oct 74 43
17 St. Louis Six Wavelength Ozone 57
18 Two Closely Spaced Wavelengths 58
19 Two Wavelength Ammonia 59
20 Two and Three Wavelength Ozone 60
21 Two Widely Spaced Wavelengths 61
22 Improved Six Wavelength Ozone 62
23 Spatial Filter Effect on One Dimensional Intensity Distribution 66
24 Optical Configuration for Spatial Filter Experiment 67
25 Horizontal Scan with 8. 7 mm Mode Stop and Beamsplitter in Place 70
VI
-------
Figure
26 Horizontal Scan with 0.9 mm Cleanup Aperture, 8. 7 mm
Mode Stop and Beamsplitter In Place 71
27 Horizontal Scan with 6.4 mm Mode Stop and Beamsplitter In Place 72
28 Horizontal Scan with 0.9 mm Cleanup Aperture, 6.4 mm 73
Mode Stop and Beamsplitter In Place
29 Horizontal Scan with 6.4 mm Mode Stop and Beamsplitter Removed ^^
30 Horizontal Scan with 0.9 mm Cleanup Aperture, 6.4
Mode Stop and Beamsplitter Removed 75
31 Experimental Beam Propagation Optics Configuration "^
32 ILAMS Block Diagram 82
33 "V" Laser Optical Layout 84
34 Data Collection and Reduction System 87
35 Breadboard System 89
36 Beam Steering Optics 89
37 System Concept for Multipollutant Urban Area Monitoring 98
38 Area Monitor Configuration 98
39 Sensitive Sector Monitor 102
40 Mobile Monitor 102
41 Perimeter Monitor 102
42 Vertical Monitor 1°2
43 Folded Path for Atmospheric Studies 102
vii
-------
LIST OF TABLES
Table Page
1 Water Vapor - CO Laser Line Interference Data 11
£t
2 Absorption Coefficients and CL Variance of Atmospheric Species 17
3 Linear Weights and SNR's for Atmospheric Species 18
4 Cross Response of Linear Weights 19
5 3 Oct 74 Comparative Path Moving Point Monitor Ozone 35
Data Summary
6 Weights Corresponding to Each of the Curves Calculated
and Plotted 55
7 Response to Each of the Weights Shown in Table 6 56
8 Experimental Conditions for One-Dimensional Plots 69
9 Indicated ILAMS Pollutant Detectability for a Laser to 92
Retroreflector Range of One Kilometer
10 Absorption Coefficient Data for Indicated ILAMS Pollutant
Detectability 94
11 Comparison of Indicated ILAMS Pollutant Detectability with 100
Pollution Monitoring Criteria
viti
-------
ACKNOWLEDGEMENTS
The authors wish to gratefully acknowledge the support of many
people in obtaining the results reported here. In particular we thank
W. S. Bee man, S. E. Craig, J. B. Haberl, A. J. Nestl, J. Ozolins and
D. L. Roberts.
ix
-------
-------
Section I
CONCLUSIONS
Contract 68-02-1290 specified modifications of ILAMS (Infrared
Laser Atmospheric Monitoring System) and its operation in St. Louis in
conjunction with the RAPS program. The modifications, the change to a six
wavelength configuration was the largest of these, were intended principally
to extend the system's pollutant monitoring capability to three target gases,
ozone, ammonia and ethylene. Beam steering capability, to permit position-
Ing of the beam, 360° In azimuth and ±10° In elevation was also added. Analy-
sis and measurement of causes contributing to system errors noted in earlier
field tests were begun and useful information obtained. Progress was made
In Improving system performance, but resource constraints limited what
could be done In this area.
In St. Louis, useful field experience was gained operating In a
rather severe environment. The system was operated over several paths,
two of them about 500 meters in range (one way distance). The last one
exceeded 900 meters range, with much of It over a two lane concrete highway.
A fertilizer plant, emitting large amounts of ammonia and partlculates, was
located along the path.
While the data show reasonable correlation between the path and
point monitor Instruments, there Is room for Improvement In the performance
of the system. Data review and analysis Indicates:
1) Optical effects on the six wavelerigth beam pattern are a principal
source of error.
-------
2) Changes In signal processing can significantly reduce short
term excursions in ozone concentration measurements„
3) Spectral effects not factored into the wavelength selection
process were a source of error occurring under certain
conditions.
4) Zeroing of the system, independent of other instruments, is
an important feature which must be incorporated to eliminate
errors.
-------
Section H
RECOMMENDATIONS
A concerted effort to address system problems with the objective
of achieving Its full performance potential is an essential next step in the
development of ILAMS.
A large amount of measurement data has been accumulated.
Because this data exists and because the system can be precisely modeled
both as to the electronic and optical signal processing a quantitative system
error analysis can now be performed. It should be conducted to give direction
to error reduction activities. It will point out the nature and magnitude of the
sources of system error and provide the basis for specific design recommen-
dations. These design changes should be Incorporated In the system and Its
performance evaluated In field measurements to determine results achieved.
-------
Section in
INTRODUCTION
A gas laser system to measure trace gases by long path absorp-
tion techniques has been under development for some time. Called ILAMS,
it is described in Appendix A. Briefly, a flowing gas CO laser is uniquely
&
configured to produce multiple wavelengths which are transmitted in a rapid
sequence to a distant retroreflector. The retro returns incident energy to an
off-axis parabola which serves as both the transmit and receive optical ele-
ment. In the signal processor, the natural logarithm of the return to trans-
mitted signal ratio is taken for each wavelength. These logs are weighted and
summed to produce an output proportional to pollutant concentration In accord-
ance with Beer's Law principles. Transmitted wavelengths are selected with
the aid of computer programs using absorption coefficient data for target gases
and expected spectral interferences, to maximize return signal to noise ratios.
Linear weights are computer-calculated to discriminate pollutant effects from
Interferences.
Under this contract the system was modified for simultaneous
detection of O , NH and C H . This required changes to the laser dictated
o 3 24
by the wavelength selection process; a six wavelength configuration was
selected for the detection problem. New linear weights were determined.
Certain other laser modifications were also made. Beam steering optics were
added to the system so that It could be operated over multiple paths In a fairly
rapid sequence. The beam can be steered through 360° In azimuth and approxi-
mately ±10° In elevation. The system was transported to St. Louis and
operated there for a two month period, mid-August thru early October 1974, in
conjunction with the RAPS program. The drift problem, first Identified under
-------
Contract 68-02-0757, Is evident In the data. Time and dollar constraints of
the program permitted only limited study of the problem, but useful Information
was obtained. The work Is described In the sections which follow. The last
section, Discussion, reviews the field measurement results and presents same
analysis of the data.
-------
Section IV
SYSTEM MODIFICATIONS
SIGNAL PROCESSING
ILAMS measures directly the transmission (T , T , T , ... T )
of the sample region between the transmitter and the retroreflector at n
selected wavelengths. Assuming that the line width of the resonance absorp-
tion Is broad compared to the transmitted laser line width, and assuming also
that there Is no saturation In the absorbing media, then, for a uniform con-
centration of the absorber over the path, the transmission at each discrete
wavelength Is of the form, T = exp (-A C . L); where A Is the absorption
m m A m
coefficient of absorber A at wavelength m, C Is the concentration of absorber
A over the total optical path, and L Is the total optical path through the sample
region. If the concentration Is non-uniform over the path, as is the more
usual case, then C. L can be replaced by the Integrated concentration over
A,
the path.
Typically, C has units of grams/liter or atmospheres of partial
pressure, and L Is In centimeters. A Is In units to make A C . L dimen-
^ » m m A
sionless.
If a second absorber B with absorption coefficients B is intro-
m
duced Into the region, the net transmission will be the product of the trans-
mission due to each absorber.
-(A C.L+B CL)
_ m A m B
T = e
m
-------
If the natural log of the transmission at each wavelength Is taken
electronically, then
A
S - InT = -(A C . L + B C L) (2)
m m m A m B ' v
and the resulting signals have two convenient properties:
1) System response to any absorber Is a linear function of the
quantity (concentration x path) present
2) System response to several absorbers is the sum of the
responses to the individual absorbers.
Therefore, if the system can be designed to give a zero response to spectrally
Interfering absorbers, the system will respond only to the pollutant to be
measured, and the response will be proportional to the quantity present.
Speaking more generally, the above properties define a linear n-
dimenslonal vector space. Each gas Is represented by a vector in this
space whose length is proportional to the concentration. This formalization
permits the application of known mathematical and statistical techniques.
Using decision theory and multivariate statistical analysis, it can
be shown that the optimum signal processing involves the use of single or
multiple linear weights. Application of a single linear weight, W, means
taking a linear sum of the signals S S ... S to give a new signal,
1, z, n
S = W S + W S + ... W S . The quantity of absorbers present can be
X 1 2 2 n n
accurately determined by examining the magnitude of such linear sums.
Techniques may also be applied for choosing linear weights to
accurately measure the quantity of a given pollutant in the presence of known
interfering spectral absorbers, random spectral absorbers, scintillation,
-------
and other "noises". A definitive study on spectral absorption pattern
detection and estimation techniques using linear weights appears In
Keference 2. A summary of the applicable results Is given In Appendices A
and B of Reference 1.
Gas lasers offer a large number of lines from which wavelengths
can be selected to detect practically all Important pollutants (See Appendix B
and Reference 3.) On the basis of both analytical and experimental work,
several basic conclusions about wavelength selection can be drawn.
1) The relative success of a group of wavelengths depends
directly on the measurement problem. The pollutants to
be measured, pollutants and absorbers to be Ignored,
expected quantities of absorbers and the system noise levels,
all affect the choice of wavelengths.
2) For a given problem, there will be an optimum set of
wavelengths. Increasing the number of pollutants to be
estimated or Ignored will tend to Increase the optimum
number of wavelengths to be used; I.e., the more complex
the environment, the more wavelengths are necessary.
3) The finer the spectral structure of an absorber, the fewer
number of wavelengths are needed to best measure the
quantities of It present. The laser system is not limited to
detecting pollutants with fine structure such as ammonia and
ethylene. In fact, It does remarkably well In detecting or
rejecting absorbers with rather smooth spectral
characteristics.
-------
The primary target gases In this program were ozone, ammonia
and ethylene. Carbon dioxide, water vapor and neutral attenuation were the
Interferences considered for these targets,, For a single target gas, past
experience has shown that four wavelengths have done an excellent job In
handling spectral recognition problems In environments representative of the
real world. However, for this program, a six wavelength configuration was
selected because of the relative complexity of the detection problem. In
Reference 1, the general topics of Interferences and Optimum Linear Weights
are discussed, respectively, in Appendices A and B. A systematical
approach to the wavelength selection problem using mathematical and
computer techniques Is presented in Appendix C.
Wavelength Selection
In Appendix C of Reference 1 the philosophy and methodology of
wavelength selection is delineated as a two step computer-aided process. In
order to reduce the possibility of problems associated with proximity of water
vapor absorption lines, a number of CO laser lines were eliminated from
2
consideration before proceeding with the computerized wavelength selection
process. Table 1 lists the potentially troublesome water vapor lines and
nearby laser lines together with calculated absorption coefficients at the laser
line center. Data on water vapor was taken from Stanford Research Institute
computer printouts. A preliminary wavelength selection was then made to
reduce the number of potential CO laser lines from 59 to 20 by using a com-
z
puter program called LWSP. This program eliminates wavelengths of low
Information by an Iteration process of adjusting the power allocation. The
second step In the selection process Involves a combinatorial evaluation of the
reduced set by which all combinations are ranked in accordance with their
performance in measuring ozone, ethylene and ammonia. An existing program
(MFIL) was modified to perform tills operation and is called CMFIL.
10
-------
Table 1
WATER VAPOR - CO_ LASER LINE INTERFERENCE DATA
HO Line
_i
(cm )
924.988
929.071
970. 644
976.012
973.486
977.420
1029.490
1039.470
1055.500
1066.200
1073.315
1074.430
1091.240
Width
-1
(cm )
.050
.048
.040
.040
.034
.064
.041
.052
.061
.047
.027
.050
.056
Of
— 1 — 1
(atm cm )
-3
1.59 X10
-4
1.98X10
-4
1.89 X10
-3
1.57X10
-4
8.20X10
-4
4. 80 X 10
-3
1.02X10
-4
3. 90 X 10
-4
5.55 X 10
-3
7. 00 X 10
6.10X10~5
-3
2.21 X 10
-3
2.99X10
C°2
Laser
Line
-1
(cm )
924.975
929*018
970.548
975.931
973.289
977.215
1029.442
1039.369
1055.625
1066.037
1073.278
1074. 646
1091.031
a
@ Laser Line
-1 -1
(atm cm )
-3
1.25X10
-5
3.37X10
-6
7.84X10
-5
9.02X10
-6
6.06X10
-5
1.17X10
-4
1.57X 10
-5
2. 42 X 10
-5
3.12X10
-4
8.82 X10
7.17 X 15~6
c
2.92 X10
-5
5.27X10
R<
J In
Value t
- -
P-40
P-36
R-12
R-20
R-16
R-22
P-38
P-28
P-10
R-02
R-12
R-14
R-42
Jlati
ipoi
anci
1
6
11
4
13
10
3
9
7
2
12
8
5
11
-------
A 25 iteration LWSP run resulted in the preliminary line selections
for each target illustrated in Figures 1 through 30 The solid lines in these
figures represent the linear weights applied to each wavelength and their
length is Indicative of the relative importance of each line. The X's designate
normalized ozone absorption coefficients and the +'s designate the average
interferent noise level.
The top 10 wavelengths from the LWSP output were selected to
form the basis set for CMFIL. Combinations of this set were then evaluated
and ranked using CMFIL. The output listing for the three targets is shown in
Figure 4. The combination 1, 2, 3, 7, 9 & 10 (circled In the Figure) was
selected as the one that gave the best compromise between the highest signal-
to-nolse-ratio (SNR) and ease of Implementation In the laser. It can be seen
from Figure 4 that the SNR penalty In the selected combination is small.
Linear Weight Computation
After the six wavelengths were selected, the optimum linear
weights for measuring each of the target and Interferent species were com-
puted using the MFIL program. The absorption coefficients for the selected
lines are shown In Table 2. For the present detection problem the mutual
Interference of O0, HO, CO0, C H and NH was modeled by the covarlance
O L* £t £t 4 O
matrix approach as implemented In MFIL. Scintillation and detector noise
were treated as a single, uncorrelated noise source. Since neutral attenuation
Is a dominant Interference and Is not very statistically well-behaved or
modeled, the weights were orthogonally constrained to reject this interference,
i.e.,
W, + W0 + W0 + W^ + We + W,, = 0
123456
for each target. This was indirectly accomplished by assigning a large
variance to neutral attenuation (as well as to targets and Interferences),
see Table 2. The linear weights obtained are shown in Table 3. Their
cross responses are given in Table 4.
12
-------
a
ce
cr
i-
<£
o
x
o
UJ
o
00
o.
I
4-HWf -W-
4+ 4f
\/
tu
ct:
UJ
u_
8.40 8.80 9.20 9-60 10-00 !0.40 . 10-80 11. .20
-- . WnVE.LENGTH . - :., . ""
Figure 1. RESyLT^NT ETHYLENE LINE SELECTION FROM 25 ITERATION LWSP RUN
11 .60
-------
o
Csl
o
OD
O~
Z0
C"~) ^-
i — i • -
CL-
CS:
o
coo
CD°
CLo
i
0
l—o
So.
E-. -i |
UJ
3
o
i ~
o
CM
.—4
TflRGET 2.
'
y ^
TIME 08-35
DflTE 062174
/
X
X
X
X
, X
X
T
4. mm ij
i Tttnl T
r- v W ^s
r IT
X .1
H- + +f -H- -H4
\
V
H ' N
/
i—
UJ
^
a:
9
cc
o
-z.
1 —
X
0
UJ
1 —
UJ
UJ
U-
u^
UJ
^_
o
-------
cn
o
(M
a
03
o
5
a_
o
cog
do.
i
cc:
o
l—o
LU
o
on
o.
i
o
CvJ
TQPPFT
1 H K LJL !
«u^£
Tp
X
+ miu Lii _i
TTtm TTT r
TIME °8'35
DRTE 062174
-^ >y WWK
it
rT
it i-i 1 14
~Tr M T IT
3-80 9.2.0 9..6.0 10=00 10-40 10=30 11.20
.' . . - NRyELENGTH • .. ;. . •
Figure 3. RESULTANT OZONE LINE SELECTION FROM 25 ITERATION LWSP RUN
A
LU
O
a;
-------
COMBINATIONS OF 10 WAVELENGTHS, 6 AT A TIME, USING 1 THROUGH 10
123456
10
9.249900
1031.092773
SUM
333.03301
315.46305
313.30420
313.27819
305.20800
301.87397
300.45103
299.98512
299.11207
290.09062
298.81707
298.00067
297.94690
297.68326
297.27602
297.21180
296.71113
296.25343
296.25154
295.95622
295.68015
9.290800
1076.
,980145
9.505700 10.
1052.129615 967
333700 10.476000 10.494500
10
.532100 10.591000
10.
,632100
.707603 954.562813 952.880074 949.478264 944.197906 940.547966
10.
,696400
934.893982
WAVELENGTHS
2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
3
2
3
2
2
2
2
2
2
2
2
2
2
2
2
3
2
3
3
3
3
4
3
4
3
3
3
3
3
3
3
3
3
3
3
3
4
3
4
4
4
4
5
7
7
7
5
4
8
5
4
4
4
4
6
4
5
5
5
7
5
6
5
6
9
8
8
8
8
9
9
5
6
9
8
7
6
7
8
7
9
8
8
9
0
10
10
10
10
9
10
10
6
8
10
10
10
7
8
10
9
10
9
10
10
283.34100
282.89109
263.65334
263.31756
256.07702
256.00804
254.13359
253.83157
252.87314
251.46286
223.50657
217.59459
213.58671
209.35777
208.34166
207.09418
206.29518
202.81506
201.75308
190.20189
1
1
1
1
1
1
1
1
1
1
1
2
1
3
2
2
3
2
2
2
3
3
3
3
3
3
3
3
3
3
3
3
3
4
3
3
4
3
3
3
4
4
5
5
5
5
7
6
5
5
6
7
6
7
6
4
6
4
5
4
6
5
7
8
7
.6
8
7
6
6
7
8
8
8
7
7
7
7
7
6
7
6
8
9
9
9
9
9
7
8
8
9
9
9
9
8
9
9
8
7
8
7
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
8
117.84003
116.27672
116.17893
113.13588
106.88219
105.65019
84.86986
82.03304
79.31270
75.13602
48.03009
43.03641
33.66413
0.
°-
0.
0.
0.
0.
0.
2
2
3
3
2
2
3
1
3
1
1
1
1
1
1
1
1
1
1
1
3
3
4
4
1
3
5
3
5
3
3
3
3
4
2
4
4
2
5
2
5
4
5
5
1
4
6
4
6
5
5
5
6
5
4
5
5
4
6
4
6
5
6
6
6
8
7
7
7
6
6
6
7
6
5
8
6
8
8
6
7
6
7
7
9
9
8
8
8
7
8
7
8
7
66
9
7
9
9
8
9
9
9
10
10
10
10
9
9
9
9
8
9
10
9
10
8
10
10
10
*
This is a partial listing of the original listing.
This modification has been made to facilitate
reproduction of this manuscript.
Figure 4. CMFIL OUTPUT LISTING
-------
ABSORPTION COEFFICIENT
ATM'1 CM"1
Laser Wavelengths (Microns)
Variance
Species
°3
C°2
°2H4
"V
H2°
Neutral
Atten.
9.2499
0.
0.0035
0.1800
0.0640
0.0001
1.0000
9.2938
0.
0. 0040
0.2800
12. .7000
0.0001
1,0000
9.5039
12.7000
0.0035
0.1400
0.3100
0.0001
1. 0000
10.5321
0.
0.0023
29. 1000
0.8100
0.0002
1.0000
10.6321
0.
0.0021
2.1000
0.1300
0.0002
1.0000
10. 6964
0.
0.0013
1.6300
0.8600
0.0002
1.0000
(CL)
100
100
100
100
100
100
Table 2 ABSORPTION COEFFICIENTS AND CL VARIANCE OF ATMOSPHERIC SPECIES
-------
LINEAR WEIGHTS
Laser Wavelengths (Microns)
oo
Species
°3
C02
C2H4
NH3
H2°
Neutral
Atten.
9.2499
-0.5655
1.0000
-0. 3006
-1.0000
-0.4566
-0.6624
9.2938
-0.0527
0.0482
-0.0351
0.8897
0.0308
-0.0329
9.5039
1.0000
0.0000
-0.0000
-0.0000
-0.0000
-0.0000
10.5321
0.0388
0.0605
1.0000
-0.0044
-0.0397
-0. 0943
10.6321
-0.9946
-0.3133
-0.3092
-0.2748
1.0000
1.0000
10.6964
0.5741
-0. 7954
-0.3551
0.3895
-0.5345
0, 4734
Relative SN
324.18
0.047
801. 62
265.96
0.0005
16.86
Table 3 LINEAR WEIGHTS AND SNR'S FOR ATMOSPHERIC SPECIES
-------
TARGET GAS - INTERFERENCE CROSS RESPONSES
Species O CO C H NH HO Neutral
3 2 24 3 2
O 12.7 5.7X10"5 3.4 XIO"4 3.7 X 10"4 -3.8 X 10" 0.0
3
CO 0,0 21.4X10~4 3.9X10"4 3.7X10~4 -1.0X10~ 0.0
L*
CH 0.0 -3.4X10"6 27.8 -5.9X10"4 3.4X10~5 0.0
NH 0.0 -2.2X10'5 -llo2Xlo"4 11.53 101X10~5 0.0
O
HO 000 1.6X10"4 -6.9X10~5 1.1X10"4 4.3X10'5 0.0
£t
Neutral 0.0 4.9X10"5 903X10~4 5.2X10"4 2 X 10~ 0.0
Table 4 CROSS RESPONSE OF LINEAR WEIGHTS
-------
LASER MODIFICATIONS
Adopting six wavelengths for the pollutant detection problem was
the principal Impetus for change to the laser configuration shown In Figure 33
Appendix A. Changes in optical elements external to the laser cavity (beam
propagation optics) were made to improve system performance. In the spec-
trally scanning laser, rapid sequential transmission of several wavelengths
is achieved using the concept diagrammed in Figure 5. The grating disperses
the laser beam into Its wavelength components,, The mirrors between the
grating and the chopper wheel are positioned to intercept the wavelengths
selected and direct them through the chopper wheel to the end mirrors. The
chopper wheel, rotating at about 40 Hz (in this Instance) serves to "chop" the
laser beam for detection purposes and also permits but one of the selected
wavelengths to lase at a time. Originally in a four wavelength configuration,
the change to six wavelengths required a new end mirror mount with two addi-
tional end mirrors, a new chopper wheel, repositioning of the intermediate
mirrors and re-alignment of the mirror system. Co Incidentally, a solid
aluminum grating was installed, replacing the resin base grating that was show-
ing signs of waxing under the high density laser beam«
Changes In beam propagation optics were made concurrently with
laser modifications and later, during St. Louis field operations. Experiments
(see Section VI) had shown that the beam splitter was seriously degrading the
beam configuration produced by the spatial filter. It was replaced. However,
the variation in concentration readout as observed on the teletype printout
still exceeded that obtained the previous fall in a four wavelength configuration.
Satisfactory readout stability was achieved with the introduction of a second
pinhole at the focal point of the Ge focussing lens, between the lens and the
collimator (see fig. 33). This performance could not be duplicated in
St. Louis, however, and ultimately the second pinhole was removed, the
20
-------
COUPLING MIRROR
PLASMA TUBE
Figure 5; SPECTRALLY SCANNING LASER DESIGN CONCEPT
21
-------
beamsplitter was removed, and another, significantly different, beamsplitter
was installed to achieve acceptable performance. This beamsplitter was
made from .020 Ge with a 42' (minute) wedge. It was oriented in the system
so that the anti-reflection coated wedge surface was presented to the return
signal. Backscatter of the transmitted beam from beam propagation
optical elements was a continual problem. The focussing lens was operated
off-axis to minimize this effect.
A new, digital processor-compatible preamp was designed for the
thermistor bolometer detectors used in the system. The design incorporates
variable signal attenuation capability, 60 db in nine steps. This feature is
particularly useful when operating over the range of return signal experienced.,
with varying weather conditions. Two preamps, one each for the (return)
signal and reference detectors were built and Installed in the system. Other
modifications Included relocation of the return signal bolometer-preamp to
the position shown In Appendix A, Figure 33^ an easily removable, three
section plastic laser cover and screw jacks for leveling the laser mount.
BEAM STEERING OPTICS
Beam steering optics were incorporated In the system to permit
rotation of the laser beam through 360° In azimuth and approximately ±10° In
elevation. They consist of two 4 x 14 x 24 cm. flat mirrors, positioning
fixtures and structure to support the fixtures on which the mirrors are
mounted. A picture Is shown In Appendix A, Figure 36. The lower mirror,
at about 45° with respect to the horizontal laser beam, directs It up through
the roof of the trailer housing the system to the other mirror, also about 45°
and thence to the retroreflector and back. The mirrors are mounted on
Identical fixtures, and are adjustable about the 45° axis. Both are also
adjustable about an axis perpendicular to the 45° axis, though this capability
has been disabled on the upper unit. The upper mirror and fixture are
22
-------
mounted on a 360° turntable with coarse and fine adjustment as well as slew
capability.
Micrometer adjustments on the lower mirror are positioned to
facilitate observing return signal level on either a nearby oscilloscope or the
teletype printout. The upper mirror Is shielded from the weather by a cupola
atop the trailer whose design is such that manual positioning of the beam exit
window is required. Sighting optics have been Installed on the laser channel
for visual location of distant retros. A mirror drops Into the path of the laser
beam near the focussing lens focal point for sighting purposes.
23
-------
SECTION V
FIELD MEASUREMENTS
SITE DESCRIPTION
Under past EPA contracts ILAMS has been operated in rural and
urban atmospheres at ranges, laser to retro reflector distances, approaching
5 kilometers (three miles). More recently It was employed In a four wave-
length configuration to monitor ozone over a 0.67 kilometer range. Under
this contract the system was changed to a six wavelength configuration using
wavelengths selected and weights calculated for simultaneous monitoring of
ozone, ammonia and ethylene. The system and trailer In which it has been
Installed were moved to the St. Louis area and operated there for a period
of about two months. The system was located at Site 103 of the RAMS
(Regional Air Monitoring System) network, across the Mississippi River In
Illinois but actually only about 3 or 4 kilometers as the crow files from the
center of St. Louis.
The system, housed in a 2.4 x 4. 8 meter (8 x 16 foot) mobile
office-type trailer, was located in a fenced area, adjacent to the RAMS
station and about 4 meters from Big Bend Road, a two lane concrete highway.
Located just outside a trailer window was a portable ozone monitor with
several feet of Inlet tubing extending well above the trailer roof. The system
was operated over several monitoring paths at this site. The first two were
about 500 meters long. The last one, which was about the longest we could
get at this site without considerable preparation, exceeded 900 meters. See
Figure 6. Much of it was over the two'lane concrete highway. The
25
-------
HORSESHOE LAKE
—r—^C RETRO
Figure 6. LAYOUT OF MONITORING PATH AT RAMS SITE 103
-------
height of the laser beam (monitoring path) above ground ranged from about
1.0 to 3.5 meters. The highway proved to be a significant scintillation
source. There was visible shimmer from It when looking down the path with
binoculars on sunny days.
Across the highway and paralleling It toward the other (retro-
reflector) end of the monitoring path was a long galvanized sheet steel
structure 50 to 60 feet high housing a fertilizer plant. In the complex of
buildings surrounding it was a large ammonia storage tank which appeared to
serve as a distribution point to bulk ammonia users. The fertilizer plant was
thought to be an ammonia source too, as well as the nearby stockyards. In
addition, the plant was the source of dense plnklsh-tan partlculate emissions
coming from a stack about as high as the main structure. After a period of
such emissions, dust was everywhere,, Settling on the long sloping roof of the
main building, it gave a pinkish cast to the structure. Figures 7 and 8 are
views from laser and retro ends of the monitoring path.
OPERATIONAL EXPERIENCE
Because early system performance was not In accordance with
expectations, an extended period of experimentation and system adjustment
followed setup and initial operations in St. Louis. As discussed in the Laser
Modifications section, the beamsplitter was changed and the second plnhole
removed, giving Improved results based on cut and try experiments per-
formed at the site. In the signal processor, the quantization level of the
signal and reference channels was lowered to reduce quantization noise In
low level signals. A processor function which set negative input samples
to zero was disabled because it was Introducing an Intermittent positive
bias to some signals. Such a condition might possibly occur when operating
with a quarter second time constant under high scintillation but was not
likely with the 32 second time constant used In St. Louis operations. The
function substituted a zero when a negative return signal level occurred
27
-------
Figure 7. VIEW OF MONITORING PATH FROM LASER TO RETRO
Figure 8. VIEW OF MONITORING PATH FROM RETRO TO LASER
28
-------
and this was thought to be biasing the data.
A "cat's eye" retroreflector (telescope with a plane mirror at
Its focal point) was used In the 1974 field tests,, Prior to departing for
St. Louis, the effect on system operation of a cube corner retro (three
mutually orthogonal reflecting surfaces) was compared with that of a
cat's eye. No significant difference In system performance was observed.
Consequently, a 5"' effective aperture cube corner was used Initially In
St. Louis (because of handling convenience) at ranges up to 800 meters.
However, the reflecting surfaces corroded rapidly due to insufficient
Si O overcoating and it became necessary to return to cat's eye operation.
Focussed vs defocussed beam experiments conducted during
the course of the work showed a more stable return signal was achieved with
the latter configuration on both transmit and receive beams. The reason for
this was not clear. Defocussed operation was adopted for data taking.
Operation with spatial filter plnhole sizes ranging from 0.25 to 0.91 milli-
meters (0.010 to 0.036 Inches) Indicated a 0. 71 millimeter aperture as
best for operation at the final operational range. Experience with beam
steering optics produced estimates that once located, moving the laser beam
to another retro would take less than five minutes and with some minor
improvements 30 to 60 seconds. Time for Initial retro location would also be
reduced'with these improvements.
There was evidence of trailer ambient temperature changes
affecting system performance but attempts to Isolate the cause were unsuccess-
ful. An effort was made to keep ambient temperature as constant as possible
with a system of thermometers and judicious operation of the trailer s air
conditioners. Elements of the laser proved to be Insensitive to blasts of hot air
29
-------
from a hair dryer. The effects of changing temperature on the laser
cavity are expected to be compensated for by the spatial filter. It has been
suggested that the reference beam path could be the source of the problem,
but this has not been investigated.
Line voltage variations and transients were a frequent problem
at the monitoring site. A separate line was run from the power distribution
point for the laser high voltage power supply to isolate it from other
apparatus in the trailer. Transients were the cause of occasional computer
stoppages. Some of the transients were associated with air conditioner on-off
cycles. (Use of a ferro-resonant isolation transformer between the line and
the computer is expected to solve this problem).
DATA
Early St. Louis data reflected the difficulties discussed above.
Some indication of the nature of the problem can be seen in a comparison
with similar data taken last fall. Figure 9 is a plot of O count variation
O
with time for 6 Sept 74 data. For the time period shown, each point on the
plot is the average count for a one minute interval, 12 lines on the teletype
printout. Count, or the number on the printout, is equivalent to a aCL (See
Appendix B) and ozone concentration, 2.0 ppb/count for an 800 meter
operating range. For comparison, Figure 10 shows 3 Dec 73 data plotted the
same way. Its relative "smoothness" is evident.
Most of the path-point monitor comparison data taken in St. Louis
related ILAMS output to the stationary point monitor outside the trailer
window. However, on 3 Oct 74 some moving point monitor data was taken,
i. e. the portable ozone monitor was walked along the 938 meter path. This
data is listed in Table 5 and plotted in Figure 11. Figure 12 shows 3 Dec 73
30
-------
40 H
30.
o
o
.Q
a
O
o
20-
10-
1333
1343
1353
1403
1413
1423
r—
1433
TIME OF DAY
Figure 9 OZONE COUNT VARIATION WITH TIME - 6 Sep 1974
-------
20-
5 10-
co
o
O
0-
-10-
1310
1320
1330
1340
1350
- "T" *
1400
1410
TIME OF DAY
Figure 10 OZONE COUNT VARIATION WITH TIME - 3 Dec 1973
-------
co •
CO :
50-
40-
30-i
ppb
3
20-
£ Path Monitor
Moving Point
Monitor
10-
0900
1000
1100
1200
1300
1400
I
1500
1600
TIME OF DAY
Figure 11 COMPARATIVE PATH/MOVING POINT MONITOR OZONE DATA - 3 Oct 1974
-------
40 1
30 -
20 -
ppb
10 -
PATH MONITOR
MOVING POINT
MONITOR
ESTIMATED FROM
LATER READING
900
1000
1100 1200 1300
TIME OF DAY
1600
Figure 12. COMPARATIVE PATH/MOVING POINT MONITOR OZONE DATA - 3 Dec 1973
-------
Table 5
3 Oct 1974
COMPARATIVE PATH/MOVING POINT MONITOR OZONE DATA SUMMARY
(ILAMS from Brush 2-Channel Recorder Chart)
Time
0934
0945
0952
1000
1015-1029
1039-1055
1130
1145
1150
1200-1214
1217-1237
1303
ILAMS
27*
23*
33*
31*
47
50
48*
48*
43*
40
40
46
AID
27**
27*
30**
29**
34
43
44**
43**
44**
40
44
50**
* one minute average
** Stationary
35
-------
path/moving point monitor data for comparison,, As Table 5 Indicates, some
stationary point monitor data Is Included In Figure 11. By this time In
St. Louis what problems could be had been resolved and the comparison Is
favorable. Unfortunately, what looked like a good day of data-taking was inter-
rupted by an area-wide power failure.
Other examples of data taken In the usual format are shown on the
following pages, Figures 13 through 16. They were taken over the 938 meter
path. The data, in the form of analog chart recorder traces, Is for the two
days 9 and 10 October 1974C For each day there is a comparison of ILAMS
data and that of a portable chemiluminescent monitor, Analytical Instrument
Development Inc. (AID) Model No. 560. Also, for each day there Is a two
channel chart recorder trace showing ILAMS ozone and ammonia concentration
data side by side. Time on the path and point monitor traces runs from right
to left. On the O /NH traces, time runs from right to left too, beginning on
«J O
the top, and continuing on the bottom chart.
The zero baselines on Figures 13 and 15 are at the bottom of the
traces. On O /NH data there is a separate scale grid for each trace, so the
3 3
baseline or zero for the upper one is a little above the midpoint of the charts
reproduced In Figures 14 and 16. Zero for the lower trace is a short distance
above the bottom edge of the chart reproduction. The ammonia trace was near
zero much of the time. Recorded operation typically began around 9:00 AM and
continued into the evening, usually around 7:00 or 8:00 PM0 Such was the case
on these days. Both were bright sunny days with early morning haze.
36
-------
Figure 13 shows comparative path-point monitor data for 9 Octo-
ber. Both traces reflect the diurnal rise and fall of ozone concentration. The
practice was to "zero" ILAMS by adjusting Its chart recorder trace to
correspond In concentration value to the point monitor trace. This was done
by Introducing a fixed bias or offset In the fluctuating ILAMS output and then
waiting to see If It produced the desired correspondence. The procedure was
subject to error, particularly when the effect of a rising or falling ozone
concentration was added to the short term fluctuations In me AID output.
Consequently there Is early morning evidence of system output adjustment in
the ILAMS chart records and any error in duplicating the point monitor trace
location remains in the path monitor trace as a constant offset throughout the
day.
The midday discontinuity In the ILAMS trace of 9 October results,
as best we can tell, from beam blockage when the system was left unattended
during lunch time. An earlier beam blockage produced by a truck leaving the
fertilizer plant produced the swing to full scale visible on the trace earlier in
the day. The midday discontinuity required reloading of computer programs
and system rezerolng. A software change to prevent this from happening has
been designed and Incorporated In a new program which awaits final debugging.
Later In the day, the ILAMS trace exhibits some variability which is the result
of deliberate changes in beam position on the retro to assess their effect on
system performance. Also evident In the stationary AID trace are some
local, short term dips In ozone concentration not seen by the path monitor.
These dips generally correlated with the passing of cars and trucks. They
were attributed to NO from vehicle engine exhaust combining with O to
o
reduce the local concentration. In Figure 14, the characteristics of the
ILAMS ozone trace discussed above are repeated at a little higher chart speed.
Also apparent is the lack of ammonia, consistent with the wind data we have
37
-------
showing that It was blowing away from the point monitor and the system
monitoring path to the fertilizer plant.
Figure 15 again shows the typical diurnal rise and fall of ozone on
10 October 1974. Because the concentration level approached 120 ppb, a 5x
scale change for both monitors to keep the trace on the chart followed by a
later return to the original scale is evident. The ILAMS trace again shows
the early morning adjustment to the AID concentration level. Also a period
in the morning is apparent (straight line) where the processor was turned off
while the effect of some adjustments to the retroreflector secondary was
assessed. As on the 9th there are local concentration dips in the point
monitor trace which have been associated with NO from passing cars. The
dips are greater in number and more pronounced, consistant with the shift
in wind direction; 1. e. from the road toward the monitor.
Of particular interest is system performance late in the day
relative to that of the point monitor. Up to the point of scale change, path and
point monitors were reading almost exactly the same. Not long after the
scale change, which occurred around 6:10 PM, the point monitor ozone
reading fell to essentially zero; whereas ILAMS did not. It held steady for
awhile, then trended upward for reasons unknown. Such performance can
possibly be attributed to unknown spectral interferences, Inherent or
atmosphere-Induced optical drift in the system or a combination of both. 'A
combination is the most likely, but in the absence of quantitative data it is
impossible to determine the cause. However, conditions at the time invite
speculation that a major contributor to ILAMS performance was an unknown
spectral interference, which once identified, the system can discriminate
against. First, the wind was blowing from the fertilizer plant, across the
system's path. Effluent from the plant was visible in the beams of the lights
38
-------
around it; It was dark by this time. Large ammonia concentrations, which
tended to correlate with high partlculate emissions, were evident in the
system's HN channel (see Figure 16). On the 12th of October, when the v'
O
fertilizer plant was operating and the wind blowing from the monitoring
path to the plant, ILAMS, operating as it was on the 10th, went down to
zero like the AID monitor.
Figure 16 essentially duplicates the ILAMS ozone performance of
Figure 15, but without the scale change, hence the periods of full scale satura-
tion. As mentioned this was a day of high ammonia concentrations in the path.
Early in the morning, when this was apparent from the trace, a 4x scale
change was made to keep the trace on scale. The independence of O and NH
3 3
response is evident in the Figure. While ILAMS performance was not com-
pared with an ammonia monitoring instrument as it was for ozone, there is
evidence to show that the system was Indeed responding to ammonia. For
example, plots of ammonia concentration signal level as Indicated by different
wavelength pairs correlate very well in time, and in amplitude, in relation to
the ratio of their absorption coefficients. System performance and odor
thresholds give a rough idea of concentration levels experienced. The indi-
cated concentration change over the path was freo^iently 500 ppb or more when
the system responded to ammonia. An ammonia smell was often evident at the
fertilizer plant end of the monitoring path when the system was indicating NH
O
concentrations. A reported ammonia odor threshold of ,037 mg/llter (about
50 ppm) suggests ILAMS was responding to a nine meter plume of 50 ppm
ammonia when the indicated average concentration change over the path was
500 ppb.
39
-------
STATIONARY POINT MONJLQR (100 PPB FULL SCALE - AID MODEL #560)
LASER LONG-PATH MONITOR (100 PPB FULL SCALE - ILAMS)
Figure 13. PATH & POINT MONITOR OZONE CONCENTRATIONS - 9 OCTOBER 74
40
-------
vw*v«<^^
'.-.- - 41
-------
ZO.NE 000 PPB FULL_SCALE)_
Figure 14. ILAMS OZONE AND AMMONIA CONCENTRATIONS - 9 OCTOBER '74
42
-------
AMMONIA (UNCALIBRATED)
43
-------
STATIONARY POINT MONITOR (100 PPB FULL SCALE - AID MODEL #560)
-Tr
LASER LONG-PATH MONITOR (100 PPB FULL SCALE - ILAMS)
Figure 15. PATH & POINT MONITOR OZONE CONCENTRATIONS - 10 OCTOBER 74
44
-------
TTT-
45
-------
OZONE (100 PPB FULL SCALE)
Figure 16. ILAMS OZONE AND AMMONIA CONCENTRATIONS - 10 OCTOBER '74
46
-------
AMMONIA (UNCALIBRATED)
. . . i . .
_/^/\
. L
+
BTT
1
.»...
!
f
II
! .
t
M»n
•
]
t
|
1
I1
]
tr
ii
1
KtK^H^
iiji:!: ji^hi
iiliil;'!--!1:-::
! i ! ' : !--^L>v
»^r~~~~~>
•1 : • : • : •
3
y^ ^
-^^-^T
i •
£
' '
^^x-^^..
1 '
1
_^v_>^
===;»
•
K:
47
-------
Section VI
DISCUSSION
Initial experimentation, analysis of the data obtained In St. Louis
and reflection on the experience gained there has produced Information which
relates to the evaluation of the field measurement results and the direction of
future activity to Improve system performance.
SYSTEM ERROR SOURCES
With regard to optical problems In the system, the conclusions
indicated by St. Louis data review are:
1. Non-uniformity in the power density of the laser beam over Its
cross-section (zoning) produces optical attenuation at the retro-
reflector aperture and at the receiving mirror aperture. This
optical attenuation Is different at different laser wavelengths and
hence upsets the absolute calibration of the system. The optical
attenuation is changed by atmospheric turbulence and atmo-
spheric looming causing drift errors in the measurement of
gas concentrations. To achieve absolute calibration and bring
the drift down to workable levels, this zoning must be elimi-
nated.
2. Analysis of the data shows that the magnitude (but not the
pattern) of the drift error between two wavelengths Is roughly
proportional to their difference In wavelength. Although this
zoning-Induced drift might be caused by spectrally dispersive
optical elements In the transmitting and receiving optics, it
49
-------
is more likely due to diffraction effects. According to
Camplllo et al* the truncation of the laser beam at an
aperture produces Fresnel zones at points remote from
the aperture (depending on beam divergence). Moderate
truncation results in large percentage amplitude variations.
Broadly speaking, there are a number of things, optical in nature,
which could contribute to the spectral zoning that we feel is the major source of
error in the system. (In discussing these effects, we visualize a multiwave-
length transmitted laser beam, really a composite beam pattern, made up of
the beams at each wavelength superimposed on each other at the collecting
apertures in the system.) The conditions of concern are:
• Lack of far field concentricity of the beams at each wavelength.
• Beam size variation with wavelength.
• Non-uniformity in the power distribution across each
wavelength's beame
• Focussed vs. defocussed beam, return as well as transmit.
• Fresnel/Fraunhofer diffraction. • Retro obscuration.
• Dispersion (refractive effects). • Collecting aperture sizes.
There are many things which can be done to compensate or correct
for these effects. In the beam propagation optics a spatial filter has been intro-
duced to cleanup the laser beam, with the result that it made each wavelength's
beam more uniform in power distribution across it and more concentric with
respect to the others. However high resolution beam mapping has shown that the
mirrors, lens and beamsplitter which follow it degrade the beam configuration
produced by the spatial filter. Some of these elements can be eliminated or their
degrading effects reduced. For example, an alternative to the beamsplitter was
* A. J. Campillo et al - "Fresnel Diffraction Effects"
Appl. Phys. Letters, Vol. 23, No. 2, 15 July 1973
50
-------
tested early In the program as discussed later In this section in connection with
beam mapping experiments. Backscatter from the focussing lens could be
eliminated by replacing it with a reflecting element.
Diffraction effects are associated with the mode of operation, a
focussed or defocussed beam. In Syracuse during the fall of 1973 we operated
essentially focussed on both transmit and return. Fraunhofer diffraction
effects occur In this mode. About the only relief from them, given good quality
optics, Is to operate In the central Airy disk of the diffraction pattern, over-
filling optical elements. In St. Louis, focussed operation was too sensitive to
misalignment. Defocussed operation on transmit proved to be more stable,
and the return beam was slightly defocussed also0 Fresnel diffraction effects
are associated with defocussed operation and sharply defined "hard" apertures,
like the exit of the colllmator. To reduce Fresnel effects, the technique is to,
use smaller beam-to-aperture size ratios and soften or apodize limiting
apertures.
Obscuration by the secondary mirror in the cat's eye retroreflector
could be a significant effect in non-focussed operation but it has not been
evaluated. If significant, a non-obscuring configuration like a cube corner
(a cube corner and the cat s eye were found to be equivalent in focussed
operation) or an off-axis parabola could be used. The use of larger collect-
Ing apertures to reduce optical effects such as atmosphere Induced beam
motion are a possibility, but have not seriously been considered because of
cost.
Still another consideration Is the reference beam path and the
position of the spatial filter relative to it. Past practice has been to simulate,
in the reference path, the return signal (far field) aperture-detector relation-
ship. In current, defocussed operation, we still have a focussed operation
simulation In the reference path. Also, changing the position of the spatial
51
-------
filter appears to the reference detector as motion of the reference beam
across the aperture in front of the reference detector,, These effects need to
be analyzed further and appropriate changes made. It has been suggested that
the reference path is the source of at least some of the temperature sensitivity
experienced.
It is clear that there are a number of optical configuration ques-
tions to be answered and that substantial effort is needed in this area. Treat-
ment of the problems has so far been limited. It is expected that a thorough
study of beam propagation optics will indicate changes, which, when imple-
mented, will produce significant improvement in the performance of the
system.
System zeroing was a source of error as noted earlier. In previous
programs (References 1 and 4) the linearity of system response to ozone,
ammonia and ethylene has been established for the concentration range of
interest. For accurate concentration measurement with ILAMS it remains only
to locate zero or some known concentration point on the response line, ia e.,
the linear plot of system response versus concentration. In St. Louis, system
response for a known concentration (as indicated by the portable ozone monitor)
was determined and the "zero" response extrapolated from that point knowing
the ppb equivalent for one count on the printout,,
An appealing alternative, which eliminates the need for another
instrument to determine a point on the response line, is based on operating the
system over a very short or negligible path length. Such a situation is easy
to create by simply putting a retro, with appropriate attenuation, at or near
the transmit optics. Over such a short path, pollutant gas concentrations would
be essentially zero, and the system response over the abbreviated path would
be equivalent to zero concentration. However, there is an important condition
52
-------
to this approach. The multi-wavelength laser beam pattern and Its relation to
the retro (overfill, underfill, etc.) must be the same over the short and long
path. The resolution of system optical problems should make It possible to at
least satisfactorily approximate this condition.
Though not as extensively treated here or In the course of this
work, there are other sources of error, potentially significant contributors,
which also need to be examined and their error contribution determined.
Among them are:
1. Electronic (thermal) noise
2. Quantization noise and sampling error of the signal and
reference channels which are enhanced by the large range
in laser power levels
3. Quantization error In the numerical weighting functions which
results in poorer rejection of spectral Interferences
4. Cross talk error caused by the high and low frequency
response of the preamplifiers and laser signal power
amplitude fluctuations
5. Scintillation noise falling within the Information band width of
the system
6. Spectral absorption and scattered sources in the atmosphere
about which we have incomplete or incorrect information, or
that are so large that they exceed the dynamic range of the
system
ALTERNATIVE WEIGHTING STRATEGIES
It has been stated that St. Louis data review has revealed the
presence of drift error associated with the spacing between wavelengths.
While It appears that this error is caused by optics-related effects, It has
also been determined that a change in weighting strategy is useful In reducing
the wavelength separation dependent component of the drift error. Several
53
-------
strategies have been tested on some of the data from St0 Louis with these
results:
a) Choosing two wavelengths close together to measure a
particular target gas produces a significant Improvement
in drift
b) Using six wavelengths with weights selected to distinguish
between correlated and uncorrelated optical noise works
measurably better.
However, neither of these two approaches does a totally satisfactory
job. It should be the objective of further analysis to gather additional evidence
of alternative strategy effects. These alternative strategies can then be con-
sidered in the light of optical problem-solving results,, It is a relatively simple
matter to introduce and evaluate promising strategies in the data-taking during
subsequent field measurements „
These observations about wavelength strategies have been derived
from data taken on 11 Oct 74, a day when there was significant drift or random
fluctuation in the ozone output record that was unrelated to the ozone levels as
recorded by the AID point monitor. On that day some effluent was observed to
be coming from the fertilizer plant, The weights indicated in Table 6 were
applied to the teletype printout of o;CL (the Beer's Law exponent - see
Appendix B) at each wavelength using a hand calculator,, By plotting these
weighted outputs one can isolate the effects of both the effluent from the
fertilizer plant and the drift error related to wavelength spacing. Curves
generated in this way using the weights of Table 6 follow as Figures 17 through
22. The vertical scale of these plots is in arbitrary units; the horizontal is
time of day. Based on the absorption coefficients listed in Table 2 the
responses to each of those weights is given in Table 7.
54
-------
TABLE 6
WEIGHTS COKRESPONDING TO EACH OF THE
CURVES CALCULATED AND PLOTTED
WAVELENGTHS
CURVE
NUMBER
1
2
3
4a
4b
5
6
1
9.25
-.5655
0
-1
-1
-.9805
1
-.95
2
9.29
-.0527
0
+1
0
-.0195
0
-.05
3
9.50
+1
0
0
+1
+1
0
1
4
10.53
+.0388
0
0
0
0
0
+.0107
5
10.63
-.9946
-1
0
0
0
-1
-.5292
6
10.70
+.5741
+1
0
0
0
0
+.5185
55
-------
T7
WUGffiKES SfflffGfiOT
(B
1
©
-M
5
IS, 7
us, ®a
iS
-4
io
~3
©
6
0
4038 X10
2.05 X10
12.7
-4
-4
1.9X10
"5
-------
11 OCTOBER 1974
9,25, 9,29, 9,50, 10,53, 10,63, 10,70 MICRONS
en
-q _
I
H
AM
12
T
5
Figure 17. ST. LOUIS SIX WAVELENGTH OZONE
-------
Ol
00
11 OCTOBER 1971
10,63, 10,70 MICRONS
1
9
1 1
IO II
AM
I 1
12 /
1
1
2
fM
I I
3 *
l I
5 6
1
7
Figure 18. TWO CLOSELY SPACED WAVELENGTHS
-------
en
CO
11 "OCTOBER 197<*
9.25, 9,29 MICRONS
to ii
A.M.
/ 2. 3 V
AM.
Figure 19. TWO WAVELENGTH AMMONIA
/a.
-------
11 OCTOBER 1974
9,25, 9,50 MICRONS
O5
O
1
9
i
1
/o
1
1
// A
A
\
\ \
Z ' P» *
1 1
1
3
1
1 1 i
* 6 b
\ \ \
I
7
1
1
S
1
9,25, 9,29, 9,50 MICRONS
Figure 20. TWO AND THREE WAVELENGTH OZONE
-------
11 OCTOBER 1974
9,25, 10,63 MICRONS
-
1
T
3
AM
-------
11 OCTOBER 1974
9,25, 9.29, 9,50, 10,53, 10,63, 10,70 MICRONS
SUPERIMPOSED DATA POINTS *
to
I
10
II It
I
3
§
*AiD MOP6L 560
Figure 224 IMPROVED SIX WAVELENGTH OZONE
-------
Figure 17 is a plot of the output weighted for ozone as a target and
ethylene, ammonia, carbon dlozlde, water vapor and neutral attenuation as
interferences. This is the same curve as was recorded on the chart recorder
at St. Louis as the ozone output. The cross responses to this weight have a
non-zero value because the covariance method was used to calculate the
weights and system noise was considered along with the known spectral inter-
ferences.
Figure 18 was the first attempt to see if two closely-spaced wave-
lengths would exhibit very little drift. This plot showed larger fluctuation than
expected, so the presence of ammonia effects was examined.
Figure 19 compares two closely-spaced wavelengths one of which
is a strong ammonia line. This plot showed widely fluctuating ammonia con-
centrations with a. maximum average concentration of 800 ppb. A comparison
of Figures 18 and 19 showed that the two plots are an excellent match except
for scale and that the ratio of scales is the same as the ratio of the expected
responses to ammonia (see Table 7 ). This Is good evidence that ammonia was
the major spectral interferent in the effluent from the fertilizer plant.
Figure 20a Is a two-wavelength system which responds strongly to
ozone and only slightly to ammonia. This plot is considerably smoother than
the plot of Figure 17 and tracks the AID data as closely as could be expected,
except at the beginning and end of the day. This discrepancy was as bad as
40 ppb at one point after 7:00 pm,,
Figure 20b Is a closely spaced three-wavelength system designed
to respond to ozone, and not respond at all to ammonia. This plot shows a
further reduction in the ammonia response and no noticeable change In the
errors at the beginning and the end of the day. There are three possible
explanations for the presence of an ammonia pattern on this trace: 1) The
63
-------
ammonia absorption coefficients used to calculate the weights are wrong,
2) Another spectrally Interfering gas was mixed with the ammonia, or 3) The
very large concentrations recorded for ammonia using a 32-second system
response time are really the average of much larger, short-time-average
concentrations moving in and out of the path producing at times complete
extinction of the signal at* . The system in attempting to quantize the
2
residual noise produced some error,,
Figure 21 compares two wavelengths that are widely spaced yet do
not respond strongly to either ammonia or ozone in order to further assess the
effects of wavelength spacing. The curve shows large drift compared to either
Figure 18 or Figure 20, that is, large fluctuations in the amplitude that may or
may not be due to fluctuations in the concentration of some gas or spectrally
interfering material in the path. At times, at the beginning and end of the day
when the drift on Figures 18 and 20 is obvious, the fluctuation on Figure 21 is
correspondingly larger.
Figure 22 is the result of applying a weight that Is designed to
discriminate against the effects of wavelength spacing as well as produce low
responses to all the spectral Interferences. The cross responses are shown
in Table 7. Besides having the Insensltivlty to the spectral Interferences,
this plot appears to track the AID data somewhat better than Figures 20a and
20b. However, the morning and evening drift Is still apparent. Note that any
weight that discriminates perfectly against both water vapor and neutral attenu-
ation also discriminates against any random drift of the ten micron group of
wavelengths with respect to the nine micron group. The weights used for
Figure 18 had an adequate cross response for water vapor, but not for the
random drift in return signal at widely spaced wavelengths.
The covarlance technique for computing the optical weights did
consider optical noise as an Interfering source, but did not consider the fact
64
-------
that the noise would be highly correlated between wavelengths that are closely
spaced. A more judicious selection of weights would have resulted in a large
Improvement in the chart data taken In St. Louis. It appears that the weights
used for Figure 22 are the best set in terms of what Is known now.
HIGH RESOLUTION BEAM MAPPING AND OTHER EXPERIMENTS
To examine the effects of elements in the beam propagation optical
system a number of one-dimensional plots were made of the laser beam inten-
sity measured across the transmitting aperture of the system. In each case
a horizontal scan was made across the center of the 4.8-inch circular aperture
of the collimator with, a 0.3 mm by 0.3 mm thermistor bolometer detector.
The measured detector output voltage was read directly from an oscilloscope
presentation of the bolometer output. The vertical scale is relative intensity.
The significant feature of these curves is the variation in intensity with hori-
zontal position which is a measure of the partitioning of power distribution
across the axis. It Is this partitioning or zoning that the spatial filtering Is
designed to remove.
Figure 23 Illustrates the effect on a single wavelength of a pinhole
type of spatial filter. The curves show the output beam from the laser with
the beamsplitter removed (fig. 33) and with no Intra-cavlty mode stop (iris
stop, fig. 33) In the laser In order to emphasize the benefits of spatial
filtering. The optical configuration for this experiment Is shown In Figure 24.
Because the output beam goes more directly Into the beam expander, the beam
Is smaller at the focussing lens and at the output than is the case when
operating the laser In the ILAMS system.
The filter in this case was Inserted at the focal point of the beam
expanding collimator so that the only optical elements following it were the
45° mirror and the off-axis parabola of the beam expander,, The spatial filter
65
-------
Spatial Filter
OS
os
No Spatial Filter
r
-2
-1
Figure 23 SPATIAL FILTER EFFECT ON ONE DIMENSIONAL INTENSITY DISTRIBUTION
-------
Reference
Detector
ling
Mirror
Spatial Filter
pinhole diameter 0.25 mm
Laser
Output
Coupling
Mirror
Beam diameter at 1/e
Intensity points = 4 mm
Focussing
Lens
Focal Point of
Beam Expander
Figure 24 OPTICAL CONFIGURATION FOR SPATIAL FILTER EXPERIMENT
67
-------
was a 0.25 mm aperture providing a spatial cutoff frequency of 08 4 cycles/centi-
meter at the beam expander output. As the figure shows, the filter provided
smoothing of the higher spatial frequencies in the beam without adding appreciable
beam spreading.
Figures 25 through 30 show the same one-dimensional plots using
the external optics for the ILAMS system shown in Figure 33. These plots
show the effects of the spatial filter (cleanup aperture), laser intracavity mode
(IRIS) stop and the beamsplitter on three wavelengths, 90305, 9.504 and
10.532 microns. Table 8 shows the conditions for each of the experiments.
Comparison of the odd figures with the even figures shows the effect of the
cleanup aperture. In this case the spatial frequency cutoff of the aperture is
seven cycles per meter which is sufficient to produce some spreading of the
beam at the beam expander output. In spite of this fact the aperture seems to
have negligible effect on the high spatial frequency fluctuations of the laser
power density. The conclusion therefore is that the spatial fluctuations
observed are produced by the optical elements between the spatial filter and
the beam expander output. Comparison of Figures 25 and 26 with Figures 27
and 28 shows that the smaller mode stop size has relatively little effect on the
fluctuations in the relative power density which simply says that the 8. 7 mm
mode stop aperture was sufficient to maintain fundamental mode operation in
the laser. Comparison of Figures 29 and 30 with the other figures shows the
effect of the beamsplitter (Figure 33) in producing the laser power density
fluctuations in the output. An important fact to note here is that pattern
changes with wavelength. Note also that some spatial variations remain
which must be attributed to the other optical elements between the cleanup
aperture and the beam expander output.
In response to the results of beam mapping experiments a radical
beam propagation optics configuration change was tried. The experimental
68
-------
Table 8
EXPEHIMENTAL CONDITIONS FOR ONE-DIMENSIONAL PLOTS
gure
#
25
26
27
28
29
30
0.9 mm
Diameter
Clean Up
Aperture
Out
In
Out
In
Out
In
Mode Stop
Size
8.7 mm
8.7 mm
6.4 mm
6.4 mm
6.4 mm
6Q4 mm
Beam-
splitter
In
In
In
In
Out
Out
69
-------
10.532
-q
o
9.504]u
9.305
-3
-2
-1 0 1
Distance (cm) from Center of Aperture
Figure 25 HORIZONTAL SCAN WITH 8. 7 mm MODE STOP AND BEAMSPLITTER IN PLACE
-------
10.532
Laser off.
Turned on at
higher power
9.504f*
9.305M
-3
-2
-1
Distance (cm) from Center of Aperture
Figure 26 HORIZONTAL SCAN WITH 0.9 mm CLEANUP APERTURE, 8. 7 mm
MODE STOP AND BEAMSPLITTER IN PLACE
-------
-q
to
10.532
9.504
9.305
1
-3
—r~
-2
T r
-1
1 1 r
Distance (cm) from Center of Aperture
Figure 27 HORIZONTAL SCAN WITH 6.4 mm MODE STOP AND BEAMSPLITTER IN PLACE
-------
CO
10.532
9.504
9.305^
-3
Laser output t
power increased
-2
-1
I
0
I I
Distance (cm) from Center of Aperture
Figure 28 HORIZONTAL SCAN WITH 0.9 mm CLEANUP APERTURE, 6.4 mm
MODE STOP AND BEAMSPLITTER IN PLACE
-------
10.532
9.504M
9.305^
I 1 1 1 1 r
-3 -2 -1
T r-
1
0
-i 1 1 r
2 3
Distance (cm) from Center of Aperture
Figure 29 HORIZONTAL SCAN WITH 6.4 mm MODE STOP AND BEAMSPLITTER REMOVED
-------
tn
10.532
9.305{i
-3
1
-2
1
-1
1
1
T r
Distance (cm) from Center of Aperture
-r r
Figure 30 HORIZONTAL SCAN WITH 0.9 mm CLEANUP APERTURE, 6.4 mm
MODE STOP AND BEAMSPLITTER REMOVED
-------
configuration is shown in Figure 31. The laser beam from the coupling mirror
goes to a beamsplitter. A portion of the beam passes through It to a reference
detector. The balance is reflected off the beamsplitter to a spatial filter,
through a hole in an angled mirror thence to the off-axis parabola (colllmator)
and out to the retroreflector0 Returned energy comes back through the colli-
mator and Is reflected off the angled mirror with a hole in it to the signal
detector. It was thought that this approach, eliminating a beam splitter after
the spatial filter would improve system performance. But in the absence of
clearly superior Initial performance as determined by the stability of the tele-
type printout, further study of this configuration was halted to conserve time
and funds. Beam propagation optics were returned to the original configuration
OTHER OBSERVATIONS
An uncertainty In St. Louis measurements was the calibration of
the chemllumlnescent point monitor. Reference 5 reports significant concerns
in this area. The portable monitors used in the reported study were made
available for this program. The study concludes In part:
"The major problem encountered In this study Is the variation in the
sensitivities of the monitors. The RAMS (Regional Air Monitoring System)
monitors varied In total from 360 ppb to 120 ppb full scale. Of course,
once the stations are fully operational this may not be the case. However,
one of the mobile monitors once varied from 200 ppb at 10 AM to 133 ppb
at 4 PM. As a consequence of the wide fluctuations in sensitivity, at
least half of the measurement and data reduction efforts were spent
calibrating and determining sensltivltlesi"
The determination of the cause (an apparent temperature sensitivity was noted
In the course of the work) of the portable (mobile) monitor fluctuations is the
object of a current Investigation. The report estimated that "In addition to the
76
-------
Signal |
Detector 'rpr
i »
i i
Collimator / '
(Off-axis Parabola) "^j — T*^
Mirror 1 1
V/l '~~~^
li1 (Mirror
1 ' m Surface
' m I Return
^^m t T-i
Reference!
Spatial
Filter
^ \
Coupling j—
Mirror
p-»
Beamsplitter
\
Laser
Beam
II?
/
I m I Energy
>'•!•
Figure 31 EXPERIMENTAL BEAM PROPAGATION OPTICS CONFIGURATION
-------
uncertainties arising due to the fluctuation In the ozone concentration, there Is
an additional error of at least ±5% due to uncertainties In calibration."
78
-------
REFERENCES
1. Snowman, L. R., et al. "Development of a Gas Laser System to Measure
Trace Gases by Long Path Absorption Techniques" Final Report,
EPA Contract 68-02-0757, June 1974.
2. Morgan, D. R. and Roberts, D. A0, "Computer Signal Processing Study",
Dept. of the Army DDEL, Final Report, Vol 1: Analytical Results
Contract DAAA15-71-C-0186, September 1972.
3. Kreuzer, L. B., Kenyon, N. D., Patel, C. K. N., "Air Pollution:
Sensitive Detection of Ten Pollutant Gases by Carbon Monoxide and
Carbon Dioxide Lasers" SCIENCE VOL 177, pp 347-349, 28 July 1972.
4. Snowman, L. R., "Field Study on Application of Laser Coincidence
Absorption Measurement Techniques" Final Report of EPA Contract
EHSD 71-8.
5. Chaney, L. W. and McClenny, W. A., "St. Louis Regional Air Pollution
Study Sub-Grid Scale Characterization Pollutant Distribution Study"
Preliminary Draft.
79
-------
80
-------
Appendix A
ILAMS SYSTEM DESCRIPTION
The Infrared Laser Atmospheric Monitoring System (ILAMS) operates In the
middle region of the Infrared spectrum and Identifies atmospheric constltutents by
absorption spectroscopy. It measures average pollutant concentrations (total burden)
over Its optical path. Laser operation Is at relatively low, safe power densities of
-,..__ .2
. 001 to . 01 watts/cm In the spectral region where the eye does not transmit.
Figure 32 shows a block diagram of ILAMS. The output power from the laser is
directed to a 50% beamsplitter via a 1 mm spatial filter (cleanup aperture). The energy
reflected from the beamsplitter Is focused down to a 0.1 mm aperture which serves as
an attenuator. Behind this aperture Is the reference energy detector. The transmitted
power through the beamsplitter goes to a germanium lens which focuses the energy
near the focal point of an off-axis parabolic mirror, and the expanded, nearly collimated
beam Is transmitted to the retroreflector. The return energy from the retroreflector
retraces the path through the beam-expanding parabolic mirror and the germanium lens
to the beamsplitter. The return energy reflected from the beamsplitter is collected by
a germanium lens doublet and focused on the signal detector. Preamplifiers are mounted
directly behind the signal and reference detectors and their outputs go to the signal
processor. The detectors used In the system are thermistor bolometers, operating
at ambient temperature (uncooled), having a characteristic flat response across the
middle IK spectral region.
In the signal processor, the natural logarithm of the return to reference signal
ratio is taken for each wavelength. These logs are weighted and summed to produce
an output proportional to pollutant concentration In accordance with Beer's Law
principles. Transmitted wavelengths are selected with the aid of computer programs
using system parameters and absorption coefficient data for target gases and expected
interferences, to maximize return signal to noise ratios. Linear weights are
computer-calculated to discriminate pollutant effects from interferences.
81
-------
^ _ _ : ______'______ _____ _ \
"** — . __ Polluted Atmosphere /
.-1-- --- ^ ------ ----- /
^^ '
N /
Lens
Reference
Deteo or
Attenuator
____ sj ___
Spatial
Filter
Spectrally
Scanning
Laser
Linear Weights
Digital Signal Processor
Retroreflector
Signal
Figure 32. ILAMS BLOCK DIAGRAM
82
-------
The CO laser in the system is alignable to six wavelengths. These wavelengths
12 -i &
may be selected from the more than 70 available with CO. The laser is presently
i
designed to detect ozone, ethylene and ammonia in an environment that is expected to
contain neutral attenuation, carbon dioxide and water vapor as spectral interferences.
With other isotopes of CO or other gas lasers the system has the potential for
£t
detecting practically all important pollutants. The spectrally scanning laser itself
includes a high-gain flowing gas CO laser as the radiation source and a wavelength
A
selection mechanism, which periodically (40 Hz) scans through a series of six laser
wavelengths. The laser optical configuration is shown in Figure 33. The laser cavity
consists of a "V" shaped plasma tube and an external spectral tuner. A relatively long
laser cavity is used for sufficient gain to overcome the losses inherent in the spectral
tuner and to obtain lasing action on a large number of spectral lines. A beam travels
through the plasma tube with aid of a mirror at the point of the "V". Leaving the tube
through a germanium Brewster window, the beam is directed by mirrors through an
iris (for mode control) and onto a 105 lines/mm diffraction grating, which disperses
the beam spectrally and spatially. The six wavelengths of interest are then relayed
through holes in the chopper wheel to the four end mirrors of the laser cavity. These
holes in the chopper are so located that, as the wheel turns, only one wavelength at a
time is permitted to pass through to the end mirrors. The six end or wavelength
selection mirrors are adjusted so that the beams are directed back on themselves
through the laser cavity. In this way, selected laser wavelengths are transmitted
sequentially.
The mini-computer signal processor includes a general purpose (stored program)
, mini-computer and appropriate interface electronics. The collection and reduction of
data is entirely under computer, i. e., program control; results are displayed on simple
displays incorporated in the equipment, and on an optional teletype, which need not be
used (or even be connected) during field or test range exercise of the system.
83
-------
Coupling Mirror
Concave Mirror
/
Spatial
Filter
<» Signal
Bolometer
Reference
Beam Bolometer
Splitter
Beam to/from
Ge Focussing Lens I Collimator
Figure 33. "V" LASER OPTICAL LAYOUT
-------
The use of the stored program control and data reduction means:
• changes in system design, or variations in data reduction algorithms,
may be accommodated without alteration of the data collection or re-
duction hardware; only changes in the control program will be required.
• modification of signal processor parameters such as number of wave-
lengths (up to 8), gate locations, system response time, weighting
factors, etc., do not even require software changes, these parameters
are expediently entered by the teletype input.
• the precision of data processing may be made as accurate as desired;
similarly the impact of imprecise calculations may be assessed by direct
simulation for purposes of evaluating future low cost special purpose
instruments.
• additional data, e.g., environmental conditions, time, date, signal
variability, laser parameters, etc., may be measured and recorded
without modification of or addition to the existing system hardware.
• the performance of one or more data processing and display systems
can be directly analyzed, e.g., data from several ozone monitors
could be crosscorrelated and recorded.
The data collection and reduction system is sketched in Figure 34. A Digital
Equipment Corporation PDP 11/05 is used for the central processor. The data
collection and reduction equipment in Figure 34 consists of three major subsystems:
Interface Subsystem
This subsystem includes an 8 input analog signal multiplexer, which is
followed by a sample-and-hold amplifier and an analog-to-digital converter at
10-bit precision. (The analysis path detector preamplifier output is connected to
one multiplexer input, the reference path to a second multiplexer input, the re-
maining 6 are available for sensing other voltage levels of interest). Additional
subsystem elements include an AGC attenuator, a wheel position counter and
demultiplexer/storage capability for analog data displays like the meters shown
in Figure 34.
85
-------
Central Processor Subsystem ~ -••--_
The central processor and its own control panel form this subsystem. Power
supplies for this equipment are contained within the CPU cabinet proper. The
central processor control panel ordinarily is disabled during operation.
Program Input and Data Logging Subsystem
A Teletype Corporation ASR-33 teletype with appropriate interface circuits
constitutes this subsystem. As indicated, it plays two roles. First, it permits
entry (ordinarily via paper tape) of the control program. Second, it permits detailed
reporting of directly measured quantities, or derived (computed) quantities.
86
-------
Analysis Path
Amplifier Output.
Reference Path
Amplifier Output
Chopper Wheel
Encoder Timing »—
Signals
Interface
00
Multiplexer,
Sample and
Hold, A/D
AGC
Attenuators,
D/A Converter
Storage Registers
Teletype and Paper Tape Reader
(Teletype Not Required for Equipment
Operation; May Be Removed After
Control Program Has Been Loaded
0
Meters
Central Processor
POP 11/05
CPU
Program Input and
Data Logging
Data Display
Figure 34. DATA COLLECTION AND REDUCTION SYSTEM
-------
The central processor is designed so that programs stored in its core memory
may be caused to remain intact during periods of no primary power. This option is
exercised, so that once a control program has been entered in the CPU, it need not be
reentered until there is a need to change it, regardless of whether the CPU remains
energized or not. The control program is designed so that it will run properly re-
gardless of whether the teletype is connected or not. Thus the teletype unit is an
optional data display device, not an essential component of the system once the control
program has been entered.
The present system is in a breadboard configuration and is housed in a 2.4 by
4.8 meter (8 x 16 ft.) mobile office-type trailer. Beam steering capability allows the
laser beam to be positioned 360° in azimuth and ±10° in elevation. The system is shown
in Figure 35. The laser is in the background and before the seated operator is the PDF
11/05 Computer (central processor) and output meters. The interface unit is out of
sight behind the computer and output meters. The ASR 33 teletype is in the foreground.
Figure 36 shows the beam steering optics, the present location of the laser in
the trailer, and the structure over the laser extending through the trailer roof to
support the upper beam steering mirror. The long gray cylinder whose base is prominent
near the center of the figure houses the off-axis parabolic mirror used as the transmit-
receive optics. Behind the gray cylinder can be seen the lower beam steering mirror
and in it, the reflection from the upper mirror and a portion of the fixture (black ring)
supporting it.
88
-------
Figure 35. BREADBOARD SYSTEM
-------
Figure 36. BEAM STEERING OPTICS
90
-------
Appendix B
POLLUTANT CONCENTRATIONS DETECTABLE
WITH GAS LASER LONG PATH SENSORS
An indication of pollutant concentrations detectable at present and
projected ILAMS sensitivities can be obtained from calculations. The results
of this work have been summarized in Table 9. System sensitivity to 1%
absorptions means that all Important pollutants are detectable in average
(integrated) concentrations of 1 to 15 ppb over a 1 kilometer range. If another
order of magnitude (0.1% level) improvement in sensitivity can be attained,
average concentrations varying from 0.1 to 1.5 ppb of such pollutants as CO,
NO, NO2, SO2 and 03 are detectable over a 1 kilometer range. This distance
(2 kilometers out and back) represents the minimum "optical thickness"
required for the indicated concentration levels., A longer monitoring range will
also enhance sensitivity In direct relation to the Increased path length. (There
Is, however, a limit to this means of sensitivity improvement as will be seen
later.)
The information In Table 9 is calculated with pollutant spectral
data using Beer's Law relationships„ Assuming that the line width of the
resonance absorption Is broad compared to the transmitted laser line width,
and assuming also that there is no saturation in the absorbing media, then,
for a uniform concentration of the absorber over the path, the transmission
at each discrete wavelength Is of the form:
T = exp(-A C.L) (1)
m ^ m A ' v '
where A Is the absorption coefficient of absorber A at transmitted wave-
in
length m, C Is the concentration, of absorber, A, over the total optical path,
and L is the total optical path through the sample region. If the concentration
is non-uniform over the path, as is usually the case, then C.L can be
A
91
-------
Table 9
Indicated ILAMS Pollutant Detectabllity for a
Laser to Retroreflector Range of One Kilometer
(Two kilometers total path)
Minimum Detectable
Concentration in Parts per Billion
Pollutant
S°2
CO
°3
N02
NO
C2H4
HC1
HN03
HCHO
PAN
PBzN
1.0% m
4
94
15
4
13
10
2
2
1345
1515
4
27
11
1
0.1% (1)
0. 4 -HeNe
9.4-C120218
1.5
0.4
1.3
1.0
0.2
0.2
134.5
152
0.4
2.7
1.1
0.1
(1) Projected system sensitivity in terms of .percent absorption
92
-------
replaced by the Integrated concentration over the total path. Typically, C has
units of gramsAiter or atmospheres of partial pressure, and L is in centi-
meters. A Is in units to make A C. L dlmensionless.
m m A
A little manipulation simplifies calculations and converts equation
(1) to a form consistent with sensitivity terms:
ACL (2)
m
In this expression, the righthand side is equivalent to another frequently used
sensitivity term, aCL, which is usually displayed directly on the system's
teletype printout for each target gas0 In this expression or Is the absorption
coefficient, C the average concentration and L path length. Also, since
transmission = 1 - absorption, small absorption values are approximately
equal to ln(— ), and translate almost directly Into aCL or A C L. Thus,
T m A
cvCL's of 0.1, 0.01, and 0.001 are treated as equivalent respectively to
absorptions of 10%, 1%, and 0.1%.
It can be seen from equation (2) how the Table 9 concentration
levels (C ) for a 1 km system range (2 km total path) are calculated for 1%
A
and 0. 1% absorptions once A Is determined. The absorption coefficient
m
values used in Table 9 calculations appear in Table 10 . Because laser line
widths are orders of magnitude narrower than the resolution of spectra
generated by conventional spectrophotometers, it is desirable to obtain
absorption coefficients from actual laser measurements with the gas of Inter-
est at the selected line (laser wavelength). Wherever known, such data was
used. For some pollutants, however, it was necessary to' calculate A using
m
equation (2) and scaling an approximate value of T at the selected laser
m
wavelength from a conventional absorption spectrum. Data sources for each
gas are noted in the Remarks column of Table 10 and referenced at the end of
this Appendix.
93
-------
Table 10
Absorption Coefficient Data for Indicated ILAMS
Pollutant Detectability
Pollutant
CO
°3
N02
NO
NH3
C H
2 4
H S
2
HC1
HNO
\J
HCHO
PAN
PBzN
A
m
Absorption
Coefficient
(atm )
13.23
.53
3.329
12.7
3.965
5.123
21.9
29.10
.0372
.033
14.25
1.842
4.750
40.99
Wavelength
(Mlnrnns)
7.427
9.024
4.6110
9.501
3.448
5.4048
10.331
10.529
7.6510
3.3344
11.1732
3.5080
10.7250
10.120
Laser
HeNe
12 18
C °2
HeXe
C02
HeNe
HeNe
C°2
C°o
2
HeNe
HeNe
13 16
c13o216
HeXe
13 16
C °2
£j
c'V6
Remarks
Laser measurement
(1)
Laser measurement
(2)
Laser line from Table 5-15
Spectrum from Beckman IR-9
Laser measurement
Laser measurement
Laser measurement
Laser measurement
Laser measurement
(2)
Laser line from Table 5-11
Spectrum #37 p. 1231 (4)
Laser measurement
,g.
Laser line Table A-2 p. 48 ^ '
Spectrum #36 p. 1231 (4)
/o\
Laser from Table 5-15 v '
Spectrum Fig. 12 p. 115 (6)
.-.
Laser line Table A-2 p. 48
Spectrum Fig. 29 p. 134 (6)
/g\
Laser line Table A-2. p. 48 v '
Base 10 absorption coefficient7)
94
-------
The effect of an Increase In monitoring range can also be seen
from equation (2). For given values of T and A , a larger L means a lower
m m
detectable concentration. A one-way optical path of a kilometer, produced by
1 kilometer separation of laser and retro reflector or by path folding over
shorter distances, represents the minimum optical thickness (concentration-
path length product) required to produce sufficient signal for detection of the
Indicated concentration levels. Monitoring range can thus be Increased to the
limit of atmospheric attenuation and, ultimately, detector noise, for further
sensitivity improvement (reduction in threshold concentration levels). The
reduction factor equals the ratio of monitoring range to the 1 kilometer range
used In calculations. For example, the 3 mile range over which we operated
in Syracuse would reduce system threshold (minimum detectable) concentra-
tion levels by a factor of about five.
95
-------
REFERENCES
1. Menzies, Robert T., "Remote Detection of SOZ and CO2 with a Hetero-
dyne Radiometer", Appl. Phys. Lett, Vol 22, No. 11, 1 June 1973, p 592.
2. Smith, W. V., and Sorokin, P. Pe, "The Laser", McGraw-Hill Book Co.,
1966.
3. Patty, R. R. et al., "CO2 Laser Absorption Coefficients for Determining
Ambient Levels of 03, NHg and C2H4", Applied Optics Vol 13 No. 12
December 1974, p 2850.
4. Pierson, R. H., Fletcher, A. N., and Gantz, E. St» C., "Catalog of
Infrared Spectra for Qualitative Analysis of Gases", Analytical Chemistry,
Vol. 28, August 1956, p. 1218.
5. Jacobs, G. B., Morgan, D. R0, Snowman, L0 R., and Ware, D. A.,
"Spectrally Scanning CC>2 Laser Design Considerations", G. E. Technical
Information Series Report R67ELS-94, December 1967.
6. Hanst, P. L., "Spectroscopic Methods for Air Pollution Measurement",
in Advances in Environmental Science and Technology, Vol. Two,
John Wiley & Sons, Inc. 1971.
7. Heuss, J. M., and Glasson, W. A., "Hydrocarbon Reactivity and Eye
Irritation", Environ. Sci. & Tech.. Vol. 2, No. 12, p 1109, Dec. 1968.
96
-------
Appendix C
LASER LONG PATH MONITORING APPLICATIONS
The results obtained to date suggest ILAMS has utility for both atmospheric
monitoring and research. A block diagram of the general system concept for multi-
pollutant monitoring of a large urban area Is shown in Figure 37. It adds beam rotation
or steering and multiple retroreflectors to the concept previously described. A spec-
trally scanning laser, or lasers,capable of emitting multiple wavelengths would be em-
ployed in the system. They would provide IR energy over the spectral range defined by
the detection problem (3-12 microns for the pollutants listed below). While spectrally
scanning lasers with CO , Isotoplc CO , and noble gas fill mixtures have been used
tt &
In laboratory and field experiments, the system is adaptable to tunable lasers em-
ploying other media. Rapid developments in laser technology Indicate mat three or
perhaps only two lasers would offer sufficient lasing lines to identify important urban
atmospheric pollutants as discussed in Appendix B.
A basic monitoring configuration is shown In Figure 38. An urban area, with retro-
reflectors located about Its periphery, is swept by the optically rotated laser beam of a
centrally located transceiver. With such an arrangement, the system would provide a
warning when pollutant concentrations exceeded standards. At the same time, It would
be gathering data useful for establishing ambient or baseline pollutant levels for the
monitored area. Similarly,finer data granularity In certain portions of a monitored
: area can be obtained with Intermediate retroreflectors. With pollutant concentration
. data from the long and short paths, the pollutant contribution of the area In between
ban be determined.
97
-------
TRANSMITTER
SPECTRALLY
SCANNING LASERS
ISOTOPIC CO2
NOBLE GAS
OPTICS
SIGNAL
DETECTOR
' ELECTRONICS
I
SIGNAL PROCESSING
CONCENTRATION
READOUT
TRANSMIT
RECEIVE
BEAM
EXPANDER
BEAM
ROTATION
DETECTOR
POLLUTED
ATMOSPHERE
DISTANT
RETROREFLECTORS
REFERENCE
I LINEAR WEIGHTING MATRIX
Figure 37. SYSTEM CONCEPT FOR MULfiPOLLUTANT URBAN AREA MONITORING
LASER SCANNER
RETROREFLECTORS
Figure 38, AREA MONITOR CONFIGURATION
98
-------
It is of interest to note how the detectable pollutant concentrations of Appendix B
relate to the monitoring requirements. Table 11 compares Appendix B Table 9 values
with some pollution monitoring criteria. Detectable concentrations even at the 1%
sensitivity level are comfortably below federal standards. The Request for Proposal
leading to our field evaluation contract with EPA included suggested pollutants and con-
centration ranges for long path monitoring as noted in the table. It can be seen here,
too, that 1% sensitivity levels are adequate. In all but one case, 1% levels are at, or
below,the low end of these ranges. And even in this instance, NO2, the discrepancy is
minor, 10 ppb versus 13 ppb.
Such results can have substantial impact on monitoring philosophies. It has been
suggested that long path data, average concentration over an optical path, may be more
representative of pollution, especially ambient air levels. Being able to operate at the
indicated low concentration levels may serve to prove this point. When applied to vali-
dation of meteorological models, long path monitor data is particularly attractive, just
because it is "average." One kilometer is about the resolution limit of current models.
"Resolution1* in this context is the distance over which pollutant concentrations are
assumed to be constant.
The comparative strengths and weaknesses of point and long path sensors for urban
(1 2)
area monitoring have been discussed. ' Both of the referenced works cite the need
for the orders of magnitude sensitivity improvement in long path monitors which the work
proposed here expects to achieve. Achieving this improved performance, along with
weighing the two sensor types in terms of not initial, but total cost, can resolve current
cost effectiveness questions. If total costs are considered when point and long path
monitoring networks are compared, then operating expenses become an Important factor
and the unattended operation feature of optical sensors is significant. It may well tip
the balance of cost effectiveness in favor of long path sensors.
99
-------
Table 11
Comparison of Indicated ILAMS Pollutant Detectabillty
with Pollution Monitoring Criteria
(All data in parts per billion)
Minimum Detectable
Concentration
Pollutant 1.0%(1) 0.1%(1)
SO,
CO
15
0.4
1.5
Suggested Monitoring
Range (2)
10 - 1,000
500 - 50,000
°3
NO0
2
NO
NH3
C2H4
4
13
10
2
2
0.4
1.3
1.0
0.2
0.2
10
10
10
50
10
- 1,000
- 1,000
- 1,000
- 500
- 500
Federal
Standard
140<3>
(max. 24 hr. cone.)
9000(3)
(max. 8 hr. cone.)
80<3)(4)
(max. 1 hr. cone.)
50
(annual arith. mean)
(1) Projected ILAMS sensitivity in terms of percent absorption
(2) From RFP No. CPA - Neg. 221, "Field Study on Application of Laser
Coincidence Absorption Measurement Techniques."
(3) Not to be exceeded more than once per year
(4) Photochemical oxidants
100
-------
To meet particular monitoring and research requirements, a number of config-
urations are possible. Figure 39 illustrates a configuration to obtain finer data gran-
ularity with retroreflectors positioned to obtain long and short path data. For survey
work, a trailer configuration such as that shown in Figure 40 is convenient. For the
urban portion of the EPA evaluation.program and subsequent programs, the system
was employed in this way. Emplaced or portable retroreflectors could be used, de-
pending upon the nature of the survey. Perimeter monitoring of a high-emission
multiple source area is envisioned with the component arrangement of Figure 41. The
concept shown in Figure 42 proposes vertical monitoring with an airborne retroreflect-
pr and the laser transceiver on the ground. Other arrangements for vertical monitor-
ing are also possible.
A folded path configuration, Figure 43, is useful for plume studies or to detect
pollutants whose absorption strengths are not compatible with available monitoring
range (line of sight) for a particular installation. Folded path configurations provide
ILAMS with quasi-point monitoring capability to localize pollution sources either as
a separate unit or as part of a larger network similar to the intermediate retroreflect-
ors of Figure 39.
101
-------
Figure 39. SENSITIVE SECTOR MONITOR
Figure 40. MOBILE MONITOR
Figure 42. VERTICAL MONITOR
Figure 41. PERIMETER MONITOR
•Sj
Figure 43. FOLDED PATH FOR ATMOSPHERIC STUDIES
102
-------
REFERENCES
1. BIbbero, R. J., and Young, I. G., Systems Approach To Atr Pollution
Control John Wiley & Sons Inc., 1974.
2. Lueck, D. W. and Tschupp, E. J., "User Survey of Air Pollution
Monitoring Systems," G. E. Technical Information Series Report,
72TMP-3, January 1972.
103
-------
TECHNICAL REPORT DATA
(Please read InUructions on the reverse before completing}
1. REPORT NO.
EPA-600/2-77-009
2.
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
CARBON DIOXIDE LASER SYSTEM TO MEASURE
GASEOUS POLLUTANTS
6. REPORT DATE
January 1977
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO
R.J. Glllmeister and L.R. Snowman
9. PERFORMING ORGANIZATION NAME AND ADDRESS
General Electric, Ordnance Systems
Electronic Systems Division
100 Plastics Avenue
Pittsfield, Massachusetts 01201
10. PROGRAM ELEMENT NO.
1AD605
11. CONTRACT/GRANT NO.
68-02-1290
12. SPONSORING AGENCY NAME AND ADDRESS
Environmental Sciences Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final, May 1974-1975
14. SPONSORING AGENCY CODE
EPA - ORD
15. SUPPLEMENTARY NOTES
16. ABSTRACT
The report concerns the continuation of work in the
development of a gas laser system for air pollution monitoring
over long paths, a kilometer or more, using infrared absorption.
Modifications to a bread-board system for simultaneous detection
of 03, NH , C-H, and the addition of beam steering optics to give
the system area monitoring capability are discussed. Operation
for a two month period in St. Louis in conjunction with the
RAPS program is also discussed. During this period 0« and NH_
were monitored at Site 103 in RAMS. Data comparing system
performance with that of conventional monitors is presented
along with the results of problem investigations.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COS AT I Field/Group
*Air pollution
*0zone
*Ammonia
*Ethylene
*Analyzers
Design
*Carbon dioxide
*Infrared radiation
Evaluation
Field tests
lasers
13B
07B
07G
14B
20E
20F
8. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (ThisReport)
UNCLASSIFIED
21. NO. OF PAGES
111
20. SECURITY CLASS (Thispage)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (9-73)
104
------- |