Final Report
STUDY OF LASER  6ACKSCATTER BY
PARTICIPATES IN STACK EMISSIONS
By:  EDWARD E. UTHE and CHARLES E. LAPPLE
Prepared for:

ENVIRONMENTAL PROTECTION AGENCY
NATIONAL ENVIRONMENTAL RESEARCH  CENTER
DIVISION  OF CHEMISTRY AND PHYSICS
RESEARCH TRIANGLE PARK
NORTH CAROLINA 27711
CONTRACT CPA 70-173

-------
•xJ\ v  7 /jt
                                  jr Zgfifz
     F/V?a/ Report                                                January 1972
      STUDY  OF LASER BACKSCATTER  BY
      PARTICULATES IN STACK EMISSIONS
      By:  EDWARD E. UTHE and CHARLES E. LAPPLE
      Prepared for:

      ENVIRONMENTAL PROTECTION AGENCY
      NATIONAL ENVIRONMENTAL RESEARCH CENTER
      DIVISION OF CHEMISTRY AND PHYSICS
      RESEARCH TRIANGLE PARK
      NORTH CAROLINA 27711
      CONTRACT CPA 70-173
      SRI Project 8730
      Approved by:

      R. T. H. COLLIS, Director
      Atmospheric Sciences Laboratory

      RAY L. LEADABRAND, Executive Director
      Electronics and Radio Sciences Division

-------
                               ABSTRACT





     This experimental study investigates the validity of determining




 smoke plume opacity or particulate content from measurements of the




 backscatter of  laser radiation from plume particulates.  The backscatter



 experiments were conducted with the use of a specially designed aerosol




 chamber that  allowed the experimental geometry to simulate that assoc-




 iated with actual remote plume probing.





     The backscatter experiments used fly ash collected from a bituminous




 coal-burning  power plant.  The raw fly ash was classified into various




 size fractions  and pneumatically injected into the chamber at controlled




 concentrations.  The aerosol was continuously monitored by a white light




 transmissometer along the axis of the chamber.  Lidar measurements were




 made to determine the backscatter and transmission of the generated



 aerosols.





     At a wavelength of 0.7u, the backscatter-to-extinction ratio was




 essentially independent of particle size, and the backscatter-to-mass




 concentration ratio was inversely proportional to the volume-to-surface




 mean diameter.  At a wavelength of 1.06j_L, the backscatter-to-extinction




 ratio was more  dependent on particle size and the backscatter-to-mass




 concentration ratio was less dependent on particle size than at the




 0.7(j, wavelength.  Transmission data at both lidar wavelengths agreed




 with the white  light transmission data.  Mie theory computations of




 backscatter and extinction values agreed reasonably well with observed




 data.





     These results for fly ash indicate that plume opacity may be deter-




 mined from lidar backscatter measurements at 0.7^ wavelength, but that




mass concentration is better inferred from 1.06^ lidar backscatter




 measurements.  At either wavelength, lidar measurements of transmission




may be used to  determine plume opacity.




                                   iii

-------
                           ACKNOWLEDGMENTS
     The authors are grateful to Mr.  W.  D.  Conner  of  the Environmental



Protection Agency for his many valuable  suggestions at  various phases



of the study.





     We also acknowledge the assistance  of  the following Stanford Re-



search Institute personnel:  John Oblanas,  for his design  of  the aerosol



chamber optical sensors and directing a  portion of the  data collection;



William Dyer, Conrad Schadt, Earl Scribner, William Evans, Robert Allen,



Ilsabe Niemeyer and August Pijma, all of whom participated in the data



collection; David Jackson and Raymond Cummings for their discussions on



CW laser techniques; and Ronald Collis,  Director of the Atmospheric



Sciences Laboratory, for his helpful  discussions during the course of



this project.

-------
                               CONTENTS


ABSTRACT	   iii

ACKNOWLEDGMENTS  	     v

LIST OF ILLUSTRATIONS	    ix

LIST OF TABLES	xiii

  I   INTRODUCTION 	     1

 II   SUMMARY AND CONCLUSIONS	     3
      A.    Experimental Facility and Procedures  	     3
      B.    Data Analysis	     4

      C.    Results and Conclusions  	     5

III   DISCUSSION OF EXPERIMENTAL PROGRAM 	     9
      A.    Experimental Facility  	     9
      B.    Experimental Procedures  	     9
      C.    Experimental Results 	    16
      D.    Comparison of Experimental Results  and Mie Theory   .  .    25

 IV   TECHNIQUES OF PLUME EVALUATION 	    37
      A.    Comparison of Opacity and Backscatter
            Measurement Techniques 	    37
      B.    Laser Pulse Length Considerations  	    41

      C.    Polarization Measurements  	    43

      D.    CW Laser Techniques	    44
            1.    Unmodulated CW Laser System  	    45
            2.    Pulse Modulation CW Laser System  	    46
            3.    FM CW Laser System	    48
                                  VII

-------
                         CONTENTS (Continued)


  V   RECOMMENDATIONS FOR FURTHER RESEARCH 	   51

      A.    Development of a Low Cost,  Remote Plume  Sensor  ....   52

      B.    Experimental Evaluation with  Other Particulates   ...   52

      C.    Aerosol Chamber Improvement for Aerosol  Studies   ...   53

      D.    Extending the Accuracy of Backscatter Measurements
            Made with Pulsed Lidars	   54

      E.    Computations from Mie Theory	   55

      F.    Development of Multiple Wavelength Lidar
            Techniques	   55

REFERENCES	   57

Appendix A  Details of the Sighting Tunnel System   	  A-l

Appendix B  Fly Ash Preparation and Properties  .....  	  B-l

Appendix C  Dust Concentration Control  and Measurement  	  C-l

Appendix D  Aerosol Chamber Optical Components  	  D-l

Appendix E  Transmissometer Data	E-l

Appendix F  Lidar Instrumentation  	  F-l

Appendix G  Lidar Signature Analysis 	  G-l

Appendix H  Lidar Data Summary	  H-l
                                 Vlll

-------
                            ILLUSTRATIONS
Figure 1       Diagram of Experimental System

Figure 2       Photographs of Laser Beam Size
               at Chamber Location 	
                                                      10
                                                      12
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7
Figure 8
Backscatter Quantities Observed
at 0.7 (j, Wavelength as a Function
of the Extinction Coefficient Observed
by the Transmissometer  	
Volume Backscatter Coefficients ( |3)  Oberved
at a Lidar Wavelength of 0.7 p Related
to Particulate Mass Concentration  (M)
and White-Light Volume Extinction
Coefficients (o~)	
Volume Extinction Coefficients (a) Observed
at a Lidar Wavelength of 0.7 p Related
to Particulate Mass Concentrations (M)
and  White-Light Volume Extinction
Coefficients  	
Volume Backscatter Coefficients (c) Observed
at a Lidar Wavelength of 1.06 p Related
to Particulate Mass Cencentrations (M) and
and White-Light Volume Extinction Coefficients
Coefficients ( CT)	
Volume Extinction Coefficients (a) Observed
at a Lidar Wavelength of 1.06 u Related to Par-
ticulate Mass Concentration (M) and White-Light
Volume Extinction Coefficients	
Spherical Particle (Mie Theory) Backscatter Ef-
ficiency Factors (Backscatter Cross Section)
Geometrical Cross Section) as a Function
of Particle Size at 0.7 p, and 1.06 p Wavelengths
for Several Values at the Complex Refractive
Index 	
                                                                     17
                                                                     21
                                                                     22
                                                                     24
                                                                     26
                                                                     27
                                   IX

-------
                        ILLUSTRATIONS  (Continued)


 Figure 9       Observed and  Computed  Extinction and Backscatter
                Efficiency  Factors as  a Function of Size Param-
                eter Computed from the Average Sauter
                Diameters	    34

 Figure 10      Plume Opacity as a Function of Particulate Volume
                Extinction  Coefficient and Observed Plume
                Length	    39

 Figure A-l      Aerosol  Chamber Used in the Backscatter
                Experiment	A- 8

 Figure A-2      Aerosol  Chamber Details 	  A- 9

 Figure A-3      Dust Dispenser Details  	  A-ll

 Figure B-l      Flowsheet for Classification Processes  	  B- 8

 Figure B-2      MSA-Whitby Particle Size Analyses
                of Blackdog Fly Ash	B- 9

 Figure B-3      Coulter  Counter Particle Size Analyses
                of Blackdog Fly Ash	B-10

 Figure B-4      Photomicographs of Blackdog Fly Ash	B-ll

 Figure B-5      Additional Photomicrographs of Blackdog Fly Ash
                at Lower Magnification  	  B-12

 Figure B-6      Scanning Electron Micrographs
                of Blackdog Fly Ash	B-13

 Figure B-7      Scanning Electron Micrographs of Fly Ash Deposits
                on Sampling Filter Papers  	  B-14

 Figure  C-l      Sampling and Filter Arrangement 	  C-16

 Figure  C-2     Delren Filter Holder   	  C-17

Figure  C-3     Bag Sampler System	  C-18

Figure C-4     Sampling Locations   	  C-19
                                  x

-------
Figure C-5
                      ILLUSTRATIONS (Concluded)
Typical Transmissometer Strip Chart  Records
of Various Dust Sizes and Concentrations
                                                                 . C-20
Figure D-l     Optical Diagram of the Transmissometer
               and Angular Scattering Sensors   	  D-  6

Figure D-2     Data Flow Diagram of the Optical Scattering
               Sensors	D-  7

Figure D-3     Data Format of the Optical Scattering Sensors  .  .  .  D-  8

Figure E-l     Effect of Fly Ash Concentration and Size
               on Optical Transmission 	  E-10

Figure E-2     Generalized Correlation for
               Optical Transmission  	  E-ll

Figure F-l     SRI Mark V Lidar	F-  5

Figure F-2     Diagram of Mark V Lidar System  and Data Recording
               for Backscatter Experiments 	  F-  6

Figure G-l     Examples of Lidar Signatures (reproduced
               from Polaroid prints) for Various Mass Concen-
               tration of the 0  to 2.5 ^ Diameter Fly Ash
               Fraction (0.7 p, Wavelength Lidar)	G-  7
                                   XI

-------
                                TABLES
Table 1        Observed and Computed Aerosol Parameters   	   30

Table A-l      Dust Feeder Calibration Data	A-12

Table A-2      Effect of Compressed Air Pressure
               on Dust Dispersion	A-13

Table A-3      Summary of Dust Collected Inside Aerosol Duct-
               work and Tunnel	A-14

Table B-l      Summary of Size Analysis Data
               on Blackdog Fly Ash	B-15

Table B-2      Summary of Mean Particle Diameters
               of Blackdog Fly Ash	B-16

Table B-3      Log-Probability Approximations
               to Size Distribution of Each Fraction
               of Blackdog Fly Ash	B-17

Table B-4      Recommended "Average" Size Distribution
               for Each Fraction Used in Study	B-18

Table B-5      Typical Fly Ash Analyses	B-19

Table C-l      Summary of Initial Series of Sampling Runs   .... C-21

Table C-2      Summary of Second Series of Sampling Runs	C-26

Table C-3      Summary of Concentration Ratio Data for Sampling
               and Filter Arrangement "T"	C-32

Table C-4      Summary of Concentration Ratio Data
               for Second Series of Sampling Runs	C-33

Table C-5      Summary of Weather Measure Sampling Data	C-34
                                  Xlll

-------
                          TABLES (Concluded)
Table E-l

Table E-2


Table F-l


Table H-l

Table H-2
Summary of Transmissometer Data 	 E-12
Comparison of Measured
and Calculated Average Efficiency Factors  .  .

Characteristics of SRI/EPA Mark VIII
and SRI Mark V Lidar  	
E-13
F- 7
Lidar Data Summary	H- 3

Digital Data from Individual  Runs	H- 5
                                  xiv

-------
                            I  INTRODUCTION








     Enforcement of increasingly stringent air pollution laws governing




emissions from pollution sources requires the development of objective




direct- and remote-sensing techniques.  One major problem is the remote




evaluation of aerosol density emitted from stationary sources in terms




of legally acceptable parameters.  Subjective plume opacity reading by




a human observer (the Ringelmann procedure) is the method used by most




pollution control agencies.  The present study is concerned with certain




aspects of the development of an instrument for the objective measurement




of plume opacity or density.





     Remote measurement of stack effluent opacity using a pulsed laser




technique has been described and demonstrated by Evans (1967), Conner




and Hodkinson (1967), and Cook et al. (1971).  Single-ended measurements




of laser energy transmission through stack plumes are made by comparing




"clear air'  backscatter returns from the near and far side of the plume




return.  However, to obtain the necessary signal levels, a relatively




expensive pulsed lidar with high transmitter energy and a highly sensi-




tive low noise receiver is required.  In addition as Evans reported, sensi-




tive light-receiving photomultiplier tubes may become "saturated" from the




plume return and may not recover sufficiently to make accurate measure-




ments of the clear air backscatter from the far side of the plume.





     An alternative approach, the remote inference of plume density by




measurement of elastic backscatter of laser radiation from plume partic-




ulates, would place fewer requirements on the lidar components.  How-




ever, the inference of plume particulate content from the laser radiation




backscattered from these particulates is susceptible to errors originating

-------
from inherent uncertainties of particle backscatter cross sections



(Twomey and Howell, 1965).





     The objective of this experimental study was to investigate laser



backscatter from plume particulates in terms of plume opacity, trans-



mission at laser wavelengths, and mass concentration.  The collected



data are presented and analyzed in terms of developing a low cost, plume-



measuring instrument.

-------
                      II  SUMMARY AND CONCLUSIONS


     The experimental facility and lidar techniques used during  this

study and the results obtained are described briefly below.   Conclusions
of significance to the development of a low cost plume-monitoring  instru-

ment are presented.


A.   Experimental Facility and Procedures

     A major effort of the present study was to design and assemble  an

experimental facility to be used for the evaluation of laser backscatter

techniques from aerosols consisting of particulates of known character-

istics and in known concentrations.  The following experimental  work

was performed:

     (l)  A special aerosol chamber was designed and constructed.
          The chamber dimensions (20 in.  square and 30 ft  long)  are
          sufficient to enclose completely  a Q-switched laser pulse
          and allow the pulse to propagate  through the test  aerosol
          without interaction with any chamber surfaces.

     (2)  Raw fly ash was collected from a  coal-burning power plant
          and was divided into various size fractions in the 0 to
          10 u diameter range.  The classified fractions were analyzed
          by the Fisher Sub-Sieve Sizer,  MSA-Whitby Sedimentation
          Analyzer, and Coulter Counter techniques to determine their
          particle size distribution.

     (3)  A white-light transmissometer and a 90°  side-scattering
          photometer were constructed and mounted  in the aerosol
          chamber for continuous aerosol monitoring.   The optical
          instrumentation was specially designed to reduce multiple
          scattering effects and not to interfere  with propagation
          of the laser pulse through the test aerosol. (The 90  side-
          scatter photometer was not extensively used because of low-
          signal-to-noise problems during daylight conditions.)

-------
      (4)  The fly ash fractions  were  pneumatically  injected  into the
           chamber at various  concentrations by a  rotating-disk par-
           ticle feeder.   The  particle feeder was  calibrated  in terms
           of mass concentration  for each  fly ash  fraction by separately
           measuring dust  rates from the feeder at various feeder speeds.

      (5)  Extensive particle-sampling was conducted at various locations
           within the chamber  to  assess the degree of uniformity of par-
           ticulate concentration.

      (6)   The white-light transmissometer data were  calibrated in terms
           of particulate  mass concentration for each fly ash fraction
           by relating time-averaged transmissions to the calibrated
           particle  feed rates.

      (?)   Lidar observations of  the test  aerosol were made 500 ft from
           the aerosol  chamber.  The lidar backscatter signatures were
           recorded  on  Polaroid film using  several oscilloscope displays
           to  provide  a linear reading of  return signals that can vary
           over several orders of magnitude.  Concurrent lidar and
           chamber measurements of aerosol concentration were conducted
           as  a function of particulate concentration, particulate size
           (using the various fly ash size fractions), and lidar wave-
           length (0.7 and 1.06 u,).
B.   Data Analysis

     Each lidar observation (firing) produced relative measurements of:

transmitted peak power, peak power return (backscatter) from the test

aerosol, and peak power return from a black passive reflector placed

beyond the aerosol chamber.  These recorded quantities were interpreted

in terms of aerosol transmission at the lidar wavelength, relative back-

scatter from the test aerosol,  and absolute backscatter (volume back-

scatter coefficient) by assuming values for the laser pulse length and

the reflectivity of the passive reflector.  Then these lidar-derived

quantities were related to aerosol mass concentrations and white-light

extinction coefficients derived at the chamber site for each fly ash

fraction.  All data are presented as mass concentrations or optical co-

efficients that are independent of experimental geometry.

                                   4

-------
C.   Results and Conclusions





     Data plots of lidar-derived aerosol backscatter showed a linear



relationship with respect to aerosol mass concentration and volume



extinction coefficient at low concentrations,  becoming nonlinear at high



concentrations.  Although detector saturation or limited instrumentation



bandpass could explain this result, it was concluded that within-pulse



attenuation by the test aerosol caused the observed nonlinearity.   A



model expression that includes this nonlinear effect as a correction



factor to the expected linear relationship was fitted to the observed



data using a nonlinear least-squares procedure.





     Data collected at a lidar wavelength of 0.7 u showed that the rela-



tionship between backscatter and the volume extinction coefficient for



white-light is nearly independent of the particle size distribution.



Since extinction is related to the second moment of the particle size



distribution this result indicates that backscatter is similarly related,



Data collected at a lidar wavelength of 1.06 p. showed that backscatter



is less related to extinction but more nearly related to the third moment



of the size distribution than at a lidar wavelength of 0.7 u.  This sug-



gests that interpretation of lidar backscatter data in terms of plume



opacity is best achieved at a lidar wavelength of 0.7 LL and that inter-



pretation in terms of mass concentration is best achieved at a lidar



wavelength of 1.06 |i.  The degree of experimental data scatter between



aerosol backscatter and mass concentration was similar for low and high



density plumes.





     Extinction coefficients observed at the two lidar wavelengths




agreed with those observed with white light from the transmissometer



records.   Hence, valid plume opacity measurements may be made by ob-



serving the near and far side clear air returns at either lidar wave-



length.  The extinction at the two wavelengths  is  sufficiently- similar

-------
 to discourage any attempt  to  determine mass concentration by comparing
 observations of  extinction at these  two wavelengths.  However, an im-
 proved inference of mass concentrations may be possible by comparing
 backscatter from the  plume particulates at these two wavelengths.

      Observed relationships among aerosol mass concentrations, volume
 extinction coefficients and volume backscatter coefficients were compared
 with those computed by use of the Mie scattering theory for spherical
 particles  assuming the results of the particle size analysis.  In general,
 the theoretical  and experimental results agreed for each fly ash fraction.
 The close  agreement between theory and experiment encourages confidence
 in  the  experimental techniques used, which can now be extended to study
 backscatter  from particulates for which theory is not complete (such as

 backscatter  from volumes of irregularly shaped particles).

      In general,  the main  conclusion of the study is that the experi-
mental techniques developed are valid and useful for evaluation of laser
plume measurement  instrumentation  and for investigating the scattering

properties of particulate volumes.   In particular,  the experiments show
to a substantial extent (but not conclusively) for fly ash from coal-
burning power plants that:

     (l)  A lidar measurement  of plume opacity by observing the clear
          air returns from the near and far side of the plume return
          is valid; i.e.,  plume transmission at lidar wavelengths of
          0.7 \i and 1.06  p. and white light are equal within experi-
          mental error.

     (2)  A lidar measurement  of plume particulate  backscatter at
          0.7 p. wavelength  may be interpreted  in terms of opacity
          regardless  of particle size distribution.   The relationship
          between plume backscatter and opacity is  not so well defined
          at 1.06 p, wavelength.

     (3)   The relationship  of  backscatter  to particulate mass  concen-
          tration is  less dependent  on particle size at 1.06  |i than
          it is at 0.7 u..

-------
(4)  The relationship between plume backscatter and  plume  opacity
     at 0.7 u. wavelength is less dependent upon particle size
     than is the relationship between plume backscatter and  par-
     ticulate mass concentration at 1.06  p, wavelength.

(5)  Comparative measurements of backscatter at both 0.7 u and
     1.06 u. wavelength are likely to give better indications of
     particulate concentration than are single wavelength  measure-
     ments.  Dual wavelength measurements of extinction, however,
     do not appear to offer a similar advantage, due to the  rela-
     tivity small dependence of extinction upon wavelength.

-------
                Ill  DISCUSSION OF EXPERIMENTAL PROGRAM








A.   Experimental Facility





     An especially designed experimental facility was developed to study



laser backscatter from a volume of particulates.  The facility allowed



the investigation to be conducted in a manner that nearly duplicates the




geometry of actual remote plume evaluation.  Figure 1 shows the experi-




mental set up, including lidar, aerosol chamber, and particle feeder.




The chamber dimensions (20 in. x 20 in. x 30 ft) are sufficient to enclose




the effective scattering portion of a Q-switched laser pulse (typically




20 to 30 ns).  Particle-laden air curtains at both ends of the chamber



effectively isolate the test aerosol from the surrounding environment




without affecting transmission of the laser pulse as it traverses a path




through the chamber.  Additional design details of the aerosol chamber




are included in Appendix A.








B.   Experimental Procedures





     Particulates of known size distribution, shape,  and refractive prop-



erties were pneumatically injected into the chamber at controlled con-




centrations by means of a rotating disk particle feeder (see Appendix A).




The generated aerosol was continuously monitored by a white light trans-




missometer along the axis of the chamber and, during periods of low back-




ground light levels, by a photomultiplier viewing 90° sidescatter from




a small portion of the transmissometer beam.  (Further details of these




optical components are included as Appendix D.)

-------
                                                          BLACK TARGET
    LIDAR
                                                      Air intake -
                                                           FAN
                                                            TA-653583-17

                FIGURE 1   DIAGRAM OF EXPERIMENTAL SYSTEM


      The  lidar  (laser  radar) was  placed 500 ft  from  the  chamber.  The

lidar consists  basically of a  laser  transmitter that  emits  a  very brief,

high  intensity  pulse of coherent  light and a  receiver that  detects  the

energy at that  wavelength backscattered from  the  atmospheric  aerosol as

a  function of range.   The resulting  signal was  displayed on an oscilloscope

and photographed.  The lidar data records consist of  Polaroid prints that

display relative  transmitted peak power, peak power  return  from  the aero-

sol within the  chamber, and peak  power reflected  from a  black passive re-

flector placed  on the  far side of the chamber.  Further  discussion of the

lidar  characteristics  and data recording procedures  are  found in Appen-

dix F.

     Fly  ash from a large coal-burning power plant was used for  the partic-

ulate matter in the backscatter experiments.   Raw fly  ash was collected

and classified  into various size  fractions and  these  were then subjected

to various particle size analysis.  Appendix B  discusses the  results of

particle  size classification and  analysis.  Basically, the  fly ash used
                                   10

-------
in the backscatter experiments consisted of four size fractions (0-2.5,



2.5-5, 5-10, and 0-10 (i diameter).





     Valid measurements of aerosol backscatter required that the laser



pulse pass through the aerosol chamber with negligible energy incident




on the chamber surfaces.  Since light reflected from the chamber could




not be distinguished from light backscattered from the chamber aerosol



during an experimental run, extensive efforts were made to eliminate this




problem area.  The lidar van was isolated from its shock mountings so




that lidar pointing was independent of weight loads or movement within




the van.  The receiver field of view was coaligned with transmitted laser



pulses.  The lidar was then aligned with the aerosol chamber.  Initial




lidar observations made with a clean air chamber showed laser returns



from the chamber so that further modifications were necessary.





     Infrared Polaroid photographs of the laser pulse size at the chamber




location (Figure 2) showed that the laser energy could be detected over




a diameter of 18 inches (21-inch skirt).  Photographs made by inserting




neutral density filters in front of the camera optics reduced the spot




size to 16 inches for a 10-dB filter and to 4 inches for a 20-dB filter.




An upper limit to the spot size at the energy halfpower points (3 dfi)




is estimated as 6 inches.  However, the solid surfaces of the aerosol




chamber caused significant energy return from the beam skirts (from




energy at least 10 dB down from that at the beam center).  It was de-




sirable to reduce the beam size sufficiently so that negligible returns




were received from the clean air chamber.  This was accomplished by




inserting an optical stop within the laser transmitter.  Figures 2(b)




and 2(c) illustrate the laser pulse size before and after inserting an




optical stop.  Negligible energy was intercepted and reflected back to




the lidar by the chamber surfaces with the optical stop in place.
                                   11

-------
                                                (a)   PHOTOGRAPH OF LASER
                                                BEAM FOR VARIOUS DEGREES OF
                                                ATTENUATION BY NEUTRAL
                                                DENSITY FILTERS
                                               (bl   PHOTOGRAPH OF LASER
                                               BEAM FOR 10 LASER FIRINGS
                                               BEFORE PLACEMENT OF
                                               TRANSMITTER OPTICAL STOP
                                               TAPE LENGTH hUUALS 4.3 inches
                                               (c)  PHOTOGRAPH OF LASER
                                               BEAM FOR 10 LASER FIRINGS
                                               AFTER PLACEMENT OF
                                               TRANSMITTER OPTICAL STOP
                                                                 TA-8730-23
FIGURE 2   PHOTOGRAPHS OF  LASER BEAM SIZE AT CHAMBER  LOCATION
                                   11:

-------
     The photomultiplier tube was tested for its range of  output  linearity.

At high light levels the tube became saturated and behaved in a nonlinear

fashion.  A neutral density filter was placed within the receiver optical
path (see Appendix F) so that the return pulse from the passive reflector

was within the linear operating region of the photoraultiplier.  Unfortu-
nately, the filter also decreased the returns from the particulate scatters

along the path traversed by the lidar pulse.  The passive  reflector w:as

painted with a black nonglossy paint to minimize the value of the re-
ceiver optical filter required for operation within the limits of detector
linearity.

     A pair of communication lines was installed between the lidar and

aerosol chamber sites.  One line was used for voice communication and

the other line provided electrical impulses to the data recorders located
near the chamber site at times of the lidar firing (see Appendix  D).

     The backscatter experiments normally consisted of making a series

of lidar firings with a clean air chamber to establish the scattering

properties of the background air and to check the lidar/chamber alignment.

This was followed with alternating series of lidar firings for a  chamber

of clean air and various aerosol concentrations.

     The data records collected at the lidar site for each laser  firing

included relative measures of:

     (l)  Transmitted pulse peak power, P^

     (2)  Aerosol return, Pa

     (3)  Passive reflector return, P .

     The data records collected at the chamber site included:

     (l)  A continuous strip chart record of the light transmission, T,
          for a path along the length of the chamber.

     (2)  A continuous record of event marks indicating particle  feeder
          disk rotation speed.

                                   13

-------
      (3)   A record of  event marks  indicating  time of lidar firing relative
           to the above data records.

      (4)   A paper tape printout  of  transmission  at  times of the event
           marks of items  (2) and (3)  above.

      (5)   A continous  record of  relative 90°  light  scattering for times
           of low background light  levels.

      (e)   A "Nuclepore" filter sample of aerosol mass concentration, M,
           within various  parts of  the chamber for some lidar data collec-
           tion periods.

      Examples of the raw  data records are shown  in  Appendix C, Figure

 C-5,  and  in Appendix G, Figure G-l.   A detailed  listing of the lidar and

 transmissometer data collected for  each laser firing is given in Appen-

 dix H.

      The  lidar data records  were used to evaluate the following quantities:

      (l)   Relative aerosol  backscatter uncorrected  for variations of light
           energy transmission between the lidar  and aerosol chamber.

      (2)   Relative reflected energy from the  passive reflector normalized
           by the transmitted peak power.

      (3)   Relative aerosol backscatter corrected for variations of light
           energy transmission between the lidar and aerosol chamber by
           using  the target reflected  return.

      (4)   Absolute aerosol backscatter from knowledge of the passive
           target reflectivity and lidar pulse  length.

      (5)   Absolute transmission of  the laser  energy across the aerosol
           chamber  by using data from  item (2)  above and assuming a clean
           air  transmission of 1.0.

The computational procedures used to  derive these quantities are discussed

in Appendix F.

     The lidar-derived quantities of  aerosol backscatter and extinction

at the lidar wavelength were  to be related to  the white-light transmission
                                   14

-------
and aerosol mass concentration determined from the filter samples.  How-

ever, as discussed in Appendix C, difficulties were encountered with de-

riving valid aerosol samples  from the aerosol chamber.  It was concluded
that mass concentration evaluated from the particle feeder disk rotation
speed was most representative of the aerosol within the chamber over an
extended time period.  Short  term variations of the aerosol concentration

were observed for the smaller sized fly ash fractions and they probably

resulted from fly ash adhering to particle feeder components.  Accordingly,

the transmissometer output T was calibrated in terms of mass concentration

M for each fly ash fraction by the expression:
                           T  =  «p-S  ML    ,                    (1)
where the volume extinction coefficient 
-------
      The transmissometer reading at the. time of each lidar firing was used

 to derive the instantaneous aerosol volume extinction coefficient and mass

 concentration.  These values are tabulated with the lidar quantities in

 Appendix H.


 C.   Experimental Results

      Data collected from the backscatter experiments are presented and

 discussed in this section.   Experiments were conducted at two laser wave-

 lengths using the SRI/EPA*  Mark VIII (0.7 p,) lidar and the SRI  Mark V

 (1.06 |_L )  lidar.   The characteristics of these lidars are given  in Appen-

 dix F.

      Figure 3 shows aerosol backscatter quantities observed at  the 0.7-u,

 wavelength  plotted against  the  optical  volume extinction coefficients de-

 rived from  the transmissometer  data. The extinction coefficients may be

 related to  plume opacity by the graphs  in Figure  10 (shown later) for a

 plume length of  10 m.

      Although a  general  increase of  the volume backscatter coefficient

 accompanies  increases  of the optical extinction coefficient,  an unexpected

 nonlinear relation exists at the higher concentration level.  Lower values

 of backscatter than expected were observed  at high aerosol  concentrations.

 This  can be  caused by  saturation (nonlinear response)  of  the  optical  de-

 tector  and degradation of large  peak signals by limited  electronic  band-

 pass,   A third possible  cause could  be  increased within  pulse attenuation

 by the  plume  particulates at high particulate concentrations.   Experiments

 conducted under  laboratory  conditions,  although different  from  those  used

 in the  field  experiment,  indicate that  detector saturation  or electronic

 bandpass problems  would  not  account  for the  degree  of  nonlinearity observed.
*  Use of the Mark VIII lidar on this program was granted by the Division
   of Meteorology, Air Pollution Control Office, EPA.

                                   16

-------
                       10
                     CO
                     CJ
                     m
                       0.5
                       0-2
                       0.1
                      0.05
                            	1	'—'—!—'—' ' ' I
                            DUST FRACTIONS
                               x  0-2.5 M
                               +  2.5-5 At
                               D  0-10 At
                               O  5-10AI
                         1
                                         10
                                             20
                                                    50
                                                        100   200
                      EXTINCTION COEFFICIENT, TRANSMISSOMETER — km
0 -iU
V.
1 10
I
h- 5
LU
CJ
COEFFI
to
DC 1
111
H
Y~
cj 0.5
CO
CJ
CO
ui 0.2
5
o 0.1
>
| , { , , , , | , : } , : : ,
T INFERRED FROM LIDAR
_ a x _
; (Independent of Return Signal t>z'~.
~_ from the Passive Reflector) rfJ-«''
' ^ "--n* "
»«?'x
x J*
s\v'*
• ffi ~_
s
y/
s







] 2 5 10 20 50 100 200
T INFERRED FROM TRANSMISSOMETER
a
; (Dependent on the Return Signal from x* ;
" the Passive Reflector) x x_^
- x jf^:'''-:
& '')
*%i3'*
oj*
_ )*6 _
"
/
s
Xx
0 ,*a ° (c)
/°"
/ I ,,!,,,,! I , , I i , , , 1
12 5 10 20 50 100 2C
EXTINCTION COEFFICIENT, TRANSMISSOMETER — km~1 TA-8730-3
FIGURE 3   BACKSCATTER QUANTITIES OBSERVED AT 0.6943 M WAVELENGTH AS A  FUNCTION
           OF  THE EXTINCTION COEFFICIENT  OBSERVED BY THE  TRANSMISSOMETER
                                          17

-------
      The fraction of signal  decrease  as  a  result  of  attenuation within


 the laser pulse by a homogeneous  aerosol can  be expressed  as:
                          2 ,, T/2  -2ax        1-e                     /  x
                    f  =  -       e     dx   = 	     ,             (2;
                          T Jo                   ar
 where T is the pulse length and cr is  the volume extinction coefficient.


 This expression reduces to a value of one for a = 0  or  for an infinites-


 imal pulse length.   For a value of a  - 100 km   and  T = 0.01  km,  f  =  I  -  e


 « -.63,  which is approximately the amount of  nonlinearity  observed  at a =


 100 km" .   The dashed lines in Figure 3 represent nonlinear least squares


 fit of the function:
                                                                      (3)
 where



                Y = quantity plotted  along  the  ordinate


                X = quantity plotted  along  the  abscissa


           C  ,  C  = parameters  to  be  determined.  C., will  relate  to  the

                    linear  portion of  the curve  and C  will  relate to  ef-

                    fective pulse  length.



      The parameters C  and C   were determined  by the nonlinear least-


 squares procedure  of Marquardt (1963, 1964) using the supplied convergence


 criterion.  The  best fit function was determined by minimizing percent

                                                N  .       A  .     2
 residuals; i.e., finding C^ and C2 so that § =  .£ [(Y. -  Y  )/Y.]  is  a mini-


mum,  6, where Y  is  determined  from the model for each of  the N data points.


The degree of fit of the model to the data points is objectively given as

                                                ~         1 /2
 the standard error of estimate defined as S  =  [$/($ - 2)]
                                   18

-------
     The amount of data scatter varies among the backscatter quantities


presented in Figure 3.  This scatter can be caused by nonvalid assumptions


made for the data analysis or by experimental errors.  As more pieces of


experimental information are included in the analysis, data point scatter


may be reduced by eliminating certain assumptions or increased by the


experimental errors associated with the additional information.  The


relative backscatter is simply the ratio of aerosol return to transmitted


peak power; hence it ignores any variation during the experimental data


collection of the atmospheric transmission characteristics along the 500-ft


path between the lidar and aerosol chamber.  However, both signals are


relatively large and can be read from the primary data records (Polaroid


prints) with little reading error.  Mark VIII (0.7 "  wavelength) data were


collected where background conditions were similar so that variations of


attenuation of the laser pulse along the 500-ft path were probably small.


Most of the data point scatter in Figure 3(a) (s  = 0.16) is thought to


result from experimental limitations in deriving a valid measure of trans-


mitted peak power or from variations in the peak-power-to-energy relation-


ship; i.e., from variations in transmitter pulse shape (see Section IV B).



     Volume backscatter coefficient, j3,  values are shown in Figure 3(b)


and (c).  The derivation of these data is discussed in Appendix G.  As-


sumed in their derivation is knowledge of target reflectivity, laser


pulse length, and transmission of the laser pulse by the test aerosol,


T .  As discussed in Appendix G, derivation of T  from the lidar data
 a                                              a

results in an expression for p that is independent of the target reflected


pulse P .   In addition to information on relative backscatter, there is
       r

information in the form of lidar-observed quantities for conditions of


a clear air aerosol chamber.  The amount of data scatter in Figure 3(b)


(s  = 0.20) is slightly greater than that in Figure 3(a), which indicates
  e

that experimental errors associated with the additional information used
                                  19

-------
are greater  than variations of clear air transmission during the experi-



ments.  Figure  3(c)  shows volume backscatter coefficients derived by using



T  from the  transmissometer data.  Since wavelength dependence of trans-
 0.


mission is relatively minor, the more accurate transmissometer values may



have decreased  the data scatter.  However, in this analysis the target



reflected energy pulse is required and is subject to large reading errors



at high aerosol concentrations.  This explains the large data point scat-



ter (S  = 0.23) at high aerosol concentrations in Figure 3(c).
      e


     In Figure  4(a)  the volume backscatter coefficient is plotted against



the particulate mass concentration as determined from the particle feed



rates and in Figure  4(b) the volume backscatter coefficient is plotted



against the extinction coefficient as determined from the transmissometer



records  [same plot as Figure 3(b)].  A greater nonlinear distribution of



the 2.5-5 |JL  backscatter data is explained because these data, the first



collected, produced  lidar returns sufficiently large to be in the nonlinear



response region of the lidar detector, and an additional neutral density



filter was placed in the lidar receiver after this data run (see Appen-



dix H),  The other three fly ash fractions produced greater nonlinearity



at higher concentrations in proportion to their backscattering efficiency



at equal mass concentrations.




     The data in Figure 4 indicate that the backscatter coefficient is



nearly proportional  to the optical extinction coefficient independent



of particle  size, but the relationship to particulate mass is dependent



on particle  size.  This implies that at this lidar wavelength (0.7 (j,)



backscatter is proportional to second-moment quantities of the particle


size distribution.




     Figure 5(a) shows the volume extinction coefficient at a wavelength



of 0.7  jj, derived by relating the passive reflector return to the relative



transmitted peak power (see  Appendix G)  as a function of particulate mass
                                  20

-------
zw


j 10


1 5



]

3
- 2
j
3


C ,
j 1


t
j
0
j 0.5
f
a
u
>
5
D 0.2
>



n 1
| 	 1 	 1 — | — i i i i | 	 1 	 1 — i — | — i i i i | 	 1 	
DUST
FRACTIONS ,-"
	 CCS x ^
- * 0-2.5 0.100 0.007 0.18 ^s™ I
+ 2.5-5 0.061 0.011 0.16 $f*& * ^jf""
~ Q 0-10 0.036 0.006 0.15 y' +^^i
O 5-10 0.016 0.005 0.05 'V* n ,,-^+t ~/-
/ * Jr ^ v
_ ^ ^. ^"yt / /'


x .^ ^x *fli /
x^ ^ XX° /
x JK^ XH /x /
* / * '' ax ₯*
X w^ V^ *• ' t fl. JW



Xv / +^ >W D X
-* x -*1 x

*$• ^ ^ / A1^
- /•& /"$' $ ~~
/ ft. s r*
*,*,?>/
/' /' s /
- /* */ °/D% /'' M C2M / _
/ / / s
/ / ' (a)
' / /
/ \ Xlll
-------



*"
'E
* 100

a:
o
0 50
LU
1
LL
II i
cc
^
U-
2 20

_J
*
H
u 10
o
u_
u_
to uj
10 8 5
2
O
I
f)
z
1- 2
X ^
LU

1
1 • 1 | 1 1 1 l| | 1 1 | 1 I 1 l| },
/
/
**
DUST Y
- FRACTIONS *' /j
(/I) RELATION / 4. ,&''.
x 0-2.5 a = 0.82 M $o5< X 4/ttP* ^
_ + 2.5-5 a=0.39M *,' ,'/?aS-
0 0-10 a = 0.33 M * #* / / ,A
O 5-10 a = 0.21 M x/x*cv(^// '' 8
x' /a x
O X ^^ AQ >
+ / / / ^j .
— X XX*X XD XX4± fl ft ~
^^ ' Q / ^*4 V /
X XX _ * X+ y
xx + v a°^ x ' n
x v + XOB ' / /*n

~ x x/x tx^' x7 ~
X )* XX^ ^/ ^xx x
X / X x O
xX x x'"xX x °
XX / ' j S ^ 	
.M '/ / D
x xx x x. a
x xxx xx+
/ / / /
^ s ^ /
/x /xx/ xx
~ / '''/ / * (a]
/' ''/' '' *
1 '*' , D| X, , , , 1 1 , , 1 , I , , 1 1






























1 . . 1 . 1 1 1 1 1 I 1 1 1 1 1 1 1 ,
/
s

l£
X
t '
J^x '-
vp D/ :
/ 9
/6
. o t x'W x
X n
— x D x>§9x a^P ~
X n XX *
+ Q ^< v D
O * n x X^ £ O
"^^ttfX «. X D
T x X x ~
x * xx xV o x
x * D a
X «X X
X X / Of, _
/ * °
/ . D X
/
/
/
/
/
- ' o (b)
/\ °
/ 45°
/ a | \ , ,!,,,,! 1 , ,!,,,,!
                         5     10    20       50    100   200

                   MASS CONCENTRATION, FEEDER — mg/m3
                                                                                      10     20
                                  50    100     200

                                          -1
EXTINCTION COEFFICIENT, TRANSMISSOMETER — km
                                         TA-8730-28
FIGURE  5    VOLUME EXTINCTION COEFFICIENTS (a) OBSERVED AT A  LIDAR WAVELENGTH  OF 0.7 n RELATED TO PARTICULATE
            MASS CONCENTRATIONS (M) AND WHITE-LIGHT VOLUME EXTINCTION COEFFICIENTS.  The dashed lines represent the
            best fit  linear relationship  to the  observed data.

-------
concentration and Figure 5(b) shows the lidar derived volume extinction




coefficient as a function of the optical volume extinction coefficient




derived from the transmissometer records.  The relationship of the ex-



tinction coefficient at 0.7-Li lidar wavelength to mass concentration is




about equally dependent on particle size as the relationship of backscatter




to mass concentration [Figure 4(a)].  However, a greater overall scatter




of the data points occurs with decreasing scatter for increasing particle




concentrations.  Since the largest experimental error associated with




reading of the reflector return from the Polaroid records occurs at higher




particle concentrations, the data scatter trend must be associated with




the experimental technique and probably results from small variations of



the clear air transmission that become negligible at large aerosol con-




centrations.  The lidar-derived extinction coefficients (0.7 n) are only



slightly less than optical (white light) extinction coefficients [Fig-



ure 5(b)].





     Figure 6 presents volume backscatter coefficients derived with the




1.06-u wavelength lidar.  Unfortunately, data were not collected for the




5-10 11 fly ash fraction.  Data on the relationship of volume backscatter




to mass concentration were nearly half as dependent on particle size as




data collected for the 0.7-|i lidar wavelength TFigure 4(a)].  Correspond-




ingly, the scatter of data points was greater in the plot of backscatter




coefficient against optical extinction coefficient than the scatter of




0.7 u, wavelength data (a standard error of 0.26 for the 1.06-n data is




comparable to a standard error of 0.20 for the 0.7-n data, which included




data from all four fly ash fractions).  From Figure 6 it may be concluded



that, for a lidar "wavelength of 1.06 p,, the volume backscatter coefficient




is more nearly related to particulate mass concentration independent of




the particle size distribution than at the 0.7-u wavelength.  Therefore,




interpretation of single wavelength lidar data in terms of mass concentra-




tion is better achieved at 1.06-|i wavelength than at 0.7-u wavelength.
                                    23

-------
£~\J

1
1 10
r-
'E
-^
I 5

2
m
o
u. „
u. 2
LU
O
o
DC
Ul i
t
<
o
OT
^
0 0.5
<
CD
UJ
5
3
— J
0 0.2
>

0.1
| i i | i i M| | . i •! n.r - • -f
DUST
FRACTIONS
~ (M) C1 °2 Se * „-
x 0-2.5 0.115 0.016 0.17 i**'-^---
+ 2.5-5 0.090 0.009 0.16 + «f-<- " ~^^n -
p- D 0-10 0.070 0.010 0.22 2* ^^ -
r x }ffc%'0
s s'k / 9^
J&,'' s'
J*- '*^ '
j&ts /
~~~ X rj S 	
^ ^* i^ "^
*< yXM^Er ££
rfx-c xffl?
xx * s* *^
- ' » D
~ xX,^Ex/' ~
/ / s

S f S

- / ' / -
/ . f

' / '
/ ,' rf— 1 \
/ ' Of n - M | 1-e"C2 \
_x/6/ ' \ C M / —
/' ,f /' 2 /
x"x'' (^
/I i i 1 j i i i 1 1 i i 1 i i i 1 1 1
            2       5      10     20




              MASS CONCENTRATION, FEEDER — mg/m
                            50    100    200
                                                                               •*.  ,,-
                                                                            ->'•*
                                                                      o     ^
                                                                     D     X
                                                                            = 0.180 I
                                                                          S = 0.26
                                                                                   1-e
                                                                                     -0.0260\
                                                                                     0.0260
                                                                                          Ib)
                                                          i   '  I—I I M
                                                                                i  ill—i  l i
                                                                     10
                                                                           20
                                                                                    50
100     200


   -1
                                                    EXTINCTION COEFFICIENT, TRANSMISSOMETER — km


                                                                                          TA-8730-29
FIGURE 6
VOLUME BACKSCATTER COEFFICIENTS (0) OBSERVED  AT A LIDAR  WAVELENGTH OF 1.06 n RELATED TO


PARTICULATE MASS CONCENTRATIONS (M) AND WHITE-LIGHT VOLUME EXTINCTION COEFFICIENTS (a).  The


dashed line  represents the best fit to the data of the nonlinear equation indicated.

-------
However, the best universal inference would be optical extinction (trans-




mission or opacity) from data collected at a wavelength of 0.7 a.





     Figure 7 shows the volume extinction coefficient derived at 1.06-j.




wavelength as a function of particulate mass concentration and the optical




volume extinction coefficient.  The empirical relationships between ex-




tinction and mass concentration are not substantially different from those




derived at a lidar wavelength of 0.7 u,.  This indicates that extinction




measurements taken at these two wavelengths with the same experimental



accuracy as the measurements presented here could not be interpreted in




terms of particulate mass concentration, as suggested by Twomey and Howell



(1967).  However, plume transmission measurements (by the technique of




observing the clear air returns in front of and behind the plume) at




either lidar wavelength may be related to white-light transmission or




opacity.








D.   Comparison of Experimental Results and Mie Theory





     Individual particle backscatter cross sections and integrated or



volume backscatter coefficients can be investigated theoretically using




the exact Mie scattering theory for homogeneous or stratified spheres.




Figure 8 shows results of Mie theory computation in terms of the back-




scattering efficiency, which is defined as the ratio of the computed back-




scatter cross section to the geometrical cross section.  The particle size




parameter is the ratio of the particle circumference to the wavelength




of the scattered light.   These results are for homogeneous particles with




a real refractive index of 1.5 and imaginary refractive indices of 0.0




(nonabsorbing),  0.01,  0.05, and 0.10.





     The sharp resonances of the nonabsorbing particles have been explained



by surface waves, i.e.,  waves that travel along the interface between the




sphere and the medium (Bryant and Cox, 1966; van de Hulst, 1957).  It is
                                   25

-------
to
01
^uu


| 100
-it

'
CC
O 50
1-
O
LU

LU
LU
DC
1 20
Q

-1
z 10
LU
u
LU
LU
LU c
0 5
O
z
Q

p
O
S 2
X
LU


1
I 1 1 I 1 I MI 1 1 I 1 I 1 1 1, 1
X
DUST J&' /
FRACTION x ' S_
A< RELATION *x /W~
/ / „ /
x 0-2.5 a = 0.83 M / / W -
+ 2.5-5 a = 0.51 M x'|- x'X ^xB
a 0-10 a = 0.27 M x +/ V /X ''
* «t r, /

/ * / + a /
/ X +• +• -~S
X Xv ^tf'
X X *XC1
— ' / / _
x>< x x x'
x. x x
/ * X X
£^ x^a oxx
rf, y' a i*r

: a / xxx*g/B :
0 X X « X *
x x 'x
_ / X X 0 _
a ' x x
xX xX xX 0. D
x x Dx " a
XXX
XX X
/ X X +
xX x /x °
~ / X X ""
XXX D i.
xxx (a)
/ x x
XX X
IX X i i 1 i | i
, X| 1 1 1 1 i 1 1 1 1 1 1 1 1 1 1 1
                                                                                                      -1	1—I—I—I  I I
                                                                                                                 -r
                                                                                                            *
                                                                                                            *
                                                                                                       ,

                                                                                                        ?
                                                                                                                  (b)
                                                                             1
                                                                                 1   1
                    2        5     10     20




                       MASS CONCENTRATION, FEEDER — mg/m
                                                                                            50     100    200


                                                                                                     -1
50    100    200     1      2        5      10     20



        3                EXTINCTION COEFFICIENT, TRANSMISSOMETER  — km"



                                                                 TA-8730-30
    FIGURE 7
VOLUME EXTINCTION COEFFICIENTS (a) OBSERVED AT A LIDAR  WAVELENGTH OF 1.06 \i RELATED TO PARTICULATE


MASS CONCENTRATIONS  (M) AND WHITE-LIGHT  VOLUME EXTINCTION  COEFFICIENTS.  The dashed lines represent the best


fit linear relationship to the observed data.

-------
                                         10
                                       PARTICLE SIZE PARAMETER

                                        20                 30
                                                                                                 40
                                                                                                                    50
                    10"
                 o
                 z
                 uu
icr
to
                 o
                 to
                 o
                 <
                 CD
                   10
                     ,-3
                    10-4 t
                                                                                        m = 1.5 - 0.0!

                                                                    f        • \ '• '< '; |" i
                                                                   1.5 - 0.01/          I" l' ''
                                                                         m =  1.5 - 0.1/
                                                                     m = 1.5 - 0.05;
                                                                 m =  complex  index of refraction
                                         1234


                                           PARTICLE RADIUS FOR AN INCIDENT LIGHT WAVELENGTH  OF 0.7 {l
                                                                                                                   5.5
                                                                                                   _L
                                             2          3          4          5          G           7


                                           PARTICLE RADIUS FOR AN INCIDENT LIGHT WAVELENGTH OF 1.06 p
                                                                                                          TA-653581-24R1
                FIGURE 8   SPHERICAL PARTICLE (MIE THEORY)  BACKSCATTER EFFICIENCY FACTORS  (BACKSCATTER

                            CROSS SECTION/GEOMETRICAL CROSS SECTION) AS A FUNCTION OF PARTICLE SIZE AT

                            0.7  AND  1.06 v WAVELENGTHS FOR SEVERAL VALUES OF THE COMPLEX  REFRACTIVE INDEX

-------
 seen that,  as particle absorption  increases  (increase  of  the  imaginary


 part of  the refractive index),  the magnitude of  the  resonances is de-


 creased,  as are the overall  backscatter  cross sections.   However, with


 increasing  absorption  or  increasing size,  the backscatter approaches the


 Fresnel  reflection limit  at  normal incidence (McDonald, 1962) and increases


 with increasing absorption by the  particle.   The important conclusion is


 that the absence of surface  wave effects in  the  backscattered light from


 absorbing spheres could increase the correlation between  volume backscatter


 and  particulate concentration for  small  changes  in particle size distri-


 butions.  In addition, these surface waves are probably dependent on the


 smooth surface  of the  sphere for their propagation and thus may be absent


 within backscattered light from irregularly  shaped particles  regardless


 of their  orientation.  This  conclusion is  supported  by the scattering


 measurements  of Holland and  Gagne  (1970) on  randomly oriented flat plates


 that  gave less  backscatter than equivalent size  spherical particles.



      The fly  ash  particulates used  in the backscatter experiments were


mostly all  spherically shaped (see Appendix  B),  and  hence theoretical


results may relate  favorably to empirical results.   The ratios of volume


backscatter coefficient to mass concentration  and of volume extinction


coefficient to  mass concentration were experimentally determined for each


fly ash fraction  at each of the two  lidar wavelengths (0.7 and 1.06 p,).


Theoretically,  these parameters, which are independent of particulate


number densities, may be expressed as:
                              I   2
        rra Q (x,m)n( a)da
CT  _  £	e	

M  ~~    co

        f — rra pn( a)da
        'J  O
        O
                                   28

-------
 and
                               f   2
                               J ira Q (x,m)n( a)da
                        j|  _   £	b	

                        M       °°

                                 f — rra pn( a)da
                                J 3
                                o
where



           a = particle radius



           x = particle size parameter = 2rra/\



           A. = incident light wavelength



           m = complex refractive index



          Q  = extinction cross section/geometrical cross section
           e


          Q  = backscatter cross section/geometrical cross section
           b


           p = particulate density



           n = number of particles per unit volume per radius interval.



     Values of 
-------
                                                    Table  1



                                    OBSERVED  AND COMPUTED AEROSOL PARAMETERS

Size
Fraction

0-2.5
2.5-5
5-10
0-10
"Average" Sauter
Diameter
5
32
1.7
3.1
6.5
4.0

1/5
32
0.59
0.32
0.15
0.25
(1/D '.*
32'
1.00
0.55
0.26
0.43

Observed V/hite-
Light Parameters
a/H

0.943
.450
.214
.336

(cj/M1*

1.00
0.48
0.23
0.37

Q
e
av
2.73
2.46
2.12
2.23

Size
Fraction
0-2.5
2. 5-5
5-10
0-10

Size
Fraction
0-2.5
2.5-5
5-10
0-10
Mie Theory Computations (0.7-L: wavelength'1
a/U
0.990
.467
.210
.388
(CT/U)*
1.00
.47
.21
.39
3/11
0.0820
.0410
.0151
.0295
( s/ii ; *
1.00
.50
.18
.36
K
e
0.411
0.870
1.939
1.048
K
b
4.96
9.92
26.86
13.78
k
12.07
11.39
13.90
13.15
(k)*
1.00
0,94
1,15
1,09
Q
2.76
2.37
2.24
2.54
^
0.228
.208
.161
.193
X
7.70
14.01
29.44
18.09

Mie Theory Computations (1.06-^ wavelength^
C-/U
1.157
0.506
0.217
0.401
(a/in*
1.00
0.44
0.19
0.35
S/U
0.0520
.0479
.0160
.0281
(5/11)*
1.00
0.92
0.31
0.58
K
e
0.351
0.804
1.877
1.014
K
b
7.82
8.48
25.40
13.56
k
22.25
10.56
13.56
13.41
'ki*
1.00
0.47
0.61
0.60
Q
e
3.22
2.57
2.31
2.63
^
0.145
.244
.171
.184
X
5.03
9.17
19.26
11.84

Size


0-2.5
2.5-5
5-10
0-10
Lidar-Observed Aerosol Parameters
0.7-u Wavelength
C/M

0.817
.395
.207
.331
(cj/11)*

1.00
0.48
0.25
0.40
3/M

0.100
.0605
.0156
.0358
f S/1I)*

1.00
0.61
0.16
0.36
k

8.2
6.52
13.27
9.23
Q

2.28
2.01
2.20
2.17
Qb

0.279
.308
.166
.235
1.06-j, Wavelength
a/l!

0.833
.509
—
.278
( S/M}*

1.00
0.61
—
0.33
= /M

0.1145
.0896
—
.0696
(3/111*

1.00
0.78
—
0.61
k

7.28
5.68
—
3.99
Q

2.32
2.56
—
1.82
Qb

0.319
.456
—
.456
039
    - Sauter diameter ( jj. ]
  - = volume extinction coefficient (km"1^


  5 - volume backscatter coefficient (km"  ster"1;


  M = mass concentration ( mg/m^ )


  k = cj/5 'sterv,


  x - — D32/X - average size parameter


  \ = wavelength  (microns]
                                                              av
     = average extinction efficiency factor



     = average backscatter efficiency factor
  av   ( ster *• 1


  K  = K factor of Ensor and Pilat ' 1971";
   e

  K  - backscatter factor defined in the manner
   b
       of K


(P'* = parameter P normalized to a value of 1.0

       for the 0.2- to 5-j. fraction.
                                                      30

-------
is an experimentally determined quantity and hence subject to experimental
error, and that an accurate evaluation of the complex refractive index
was not available.

     Light extinction is not strongly wavelength dependent, and the cr/M
values observed with white light may be related to the monochromatic com-
putations.  Especially good agreement exists between the white-light and
0.7-(j, values for  all four fly ash fractions.  Hence, single wavelength
lidar measurements of plume transmittance at 0.7-u, wavelength can be
interpreted adequately in terms of plume opacity (l - transmission of
white light).

     Parameters of Table 1 that have been normalized to a value of 1.0
for the 0-2.5 p, fly ash fraction are enclosed within parentheses.  Quan-
tities that are independent of size distribution are constant for each
fraction; quantities that vary as the particle surface to volume ratio
are proportional  to the reciprocal of the Sauter diameter D   .  As ex-
                                                           o 2i
pected , the white-light and monochromatic extinction-coefficient-to-
mass-concentration ratios vary nearly as the reciprocal of the Sauter
diameter, 1/D  .   The computed ratios of backscatter coefficient to mass
             O^
concentration at  0.7-(_i wavelength show greater dependence on  size distri-
bution than the 1/D   variations and at 1.06-u wavelength are nearly in-
                   O^
dependent of size distribution for the 0-2.5 ij, and 2.5-5 - fly ash frac-
tions.   The 5-10  [j, fraction is approximately half as dependent on size
distribution as at the 0.7-p, wavelength.  The observed ratios are in
agreement with the computed ratios as far as establishing dependence on
size distribution.  Therefore, for the 1.06-u wavelength both observations
and theory predict less dependence on the backscatter-to-mass ratio on
particle size than for the extinction-to-mass ratio (proportional to
1/D32).
                                   31

-------
      The  other  quantities  presented  in Table 1  are  for comparison with


 the work  of  other authors.   The value K  , which is  the ratio of particulate
                                       e

 volume  concentration  to  extinction coefficient,  corresponds to the value


 K  given by Ensor  and  Pilat  (1971) and Pilat and Ensor (1970).  The rela-


 tionship  between  a/M  and K  is given  by
                    -1       2
               CT/km        m\        ,32,  ,,3.               .  .
               -I	  -  — ) =  l/K(cm /m  )p(g/cm  )    ,           (6)

                 \mg/m
                                                                       3
and the computation of K has assumed a fly ash density of p = 2.46 g/cm


for each fly ash fraction (see Appendix B).  The K  quantities are the
                                                  b

corresponding values for the ratios of particulate volume concentration


to volume backscatter coefficient.  Effective (average) efficiency factors


for volume extinction and backscatter coefficients computed from Mie Theory


are given by Q  and Q ,  where:
              e      b
                                  a Q (x,m)n(a)da
                         Q
                          e
                                    r  2 f
                                   J a n(
and
                                I  2
         a Q (x,m)n(a)da

\  •  "—^	
          r  2
          J  a nf a)da
                                   32

-------
Observed average values of Q  and Q  are determined from the following
                            e      b

expressions (Appendix E):
                          Q     =  I-  DD
                           e       \3 /   32  M
                            av
                          Q     = I - |  pD
                           b      \ 3 /  H 32 \ M
                            av
where o~/M and 9/M are experimentally determined values.



     Figure 9 presents observed and computed backscatter and extinction


efficiency factors as a function of particle size parameters computed


from the Sauter diameters (x = rrD  /X).  The lidar-observed efficiency
                                 O t~i

factors on the average are slightly less than the monochromatically com-


puted factors, however they agree well with the observed white-light values,



     The lidar-observed backscatter efficiency factors show a general max-


imum at a size parameter of approximately 10.0, which agrees with the


maximum observed in the theoretical curves for single particles (Fig-


ure 8).  The observed average values are greater than the computed average


values and exhibit more of a peak at x = 10, which indicates that the fly


ash size distributions may have been more uniform (narrower) than assumed


in the computations.   However, the use of too large a target (passive re-


flector) reflectivity or too short a laser pulse length will produce over-


estimates of volume backscatter coefficients and backscatter efficiency


factors.  In any future work to derive these absolute backscatter quan-


tities using the passive reflector technique, more emphasis must be placed


on deriving knowledge of the target reflectivity and laser pulse length


than was possible in the present study.
                                    33

-------
     10.0
IO

o
LU
O
LU
LU
LU

O

O
  X
  LU
 10
 o
 LU
 5
 LL.
 LU
 o:
 LU
U
(O
^
CJ
      5.0
      2.0
      1.0
      0.5
     0.2
     0.1
                      I     I
                          D
                          D
                                i  i   i  I             i        i     r

                                O Observed White Light (A = 0.6^ Assumed
                                     in Q Calculations)
                                A Observed 0.7 p.
                                • Observed 1.06 H
                                £± Calculated 0.7 /I
                                n Calculated 1.06 M
                                     m
                                     S3
                                                           A
                                                           i — i
                                  A °           *
                                                    D
                                                            A
                              I
                          5            10            20

                      AVERAGE PARTICLE SIZE PARAMETER. X
                                                                    50
                                                               TA-8730-31
FIGURE 9   OBSERVED  AND COMPUTED EXTINCTION AND BACKSCATTER
            EFFICIENCY FACTORS  AS A  FUNCTION  OF SIZE PARAMETER
            COMPUTED  FROM THE AVERAGE  SAUTER DIAMETERS
                                   34

-------
     The relatively good agreement between theoretical and experimental




results obtained gives support to the experimental procedures used during



this study.  Hence, further experimentation into areas for which existing




theory is incomplete (such as scattering by irregularly shaped particles!




may now proceed with some degree of confidence.
                                   35

-------
                  IV  TECHNIQUES OF PLUME EVALUATION
A.   Comparison of Opacity and Backscatter
     Measurement Techniques

     Plume evaluation from a lidar backscatter signature consisting of

the plume return, P ,  and near and far side clear air returns,  P, and P ,
                   p                                            ^      F
are the subjects of the following discussion.

     An inference of plume transmission, T, or opacity, O ,  from clear

air returns is given by the expression:
                     =  1-T  =  1	—
where R is the range to the appropriate return signal.  This equation
assumes that backscatter coefficients for the clear air regions are equal
and that the plume transmission at the lidar wavelength adequately rep-
resents the transmission of white light.  However, this technique does
not require absolute lidar calibration, and the inferred quantity is not
subject to variations of clear air transmissions over the path from lidar
to plume.

     A plume opacity measurement is related to the volume extinction
coefficient a and the observed plume length L by Bougure's law of
attenuation:
                         Jta(l - O )  =  -aL                            (8)
                                 P
                                   37

-------
 Figure 10 is  a graphic  presentation of  this  expression.   [Aerosol  opacity


 observed  during the backscatter  experiments  may  be  determined  from Fig-


 ure 10 using  the observed extinction coefficients  (Section  III)  and a


 plume  (chamber length) of 10 m.]   From Eq.  8  opacity is  related to  the


 particulate mass concentration,  M,  of a vertically  rising circular and


 homogeneous plume of radius a  observed  at  ground level  with a  lidar ele-


 vation angle  9 by the expression:
                      JLn  (1  -  0  )   =  -(-}  Ma  sec  6                     (9)
                              p        UI/
The quantity  a/M  is  independent  of  particle  concentration;  it  relates  the


aerosol optical density  (the extinction coefficient  a)  to  its  physical


density M and may  be  considered  a constant for  particulates  of a  given


type and relative  size distribution.  The relationship  of  this ratio to


various particle refractive indices and log-normal size distribution


parameters has been studied theoretically for spherical particles by


Pilat and Ensor (1970), Ensor and Pilat (l97l),  and  Conner and Hodkinson


(1967).



     The total particulate emission M  from  the  stack per  unit of time


may be expressed in terms of observed opacity as:
            M   =  77a Mv  =  -rr (-) a cos 9v  In  (l - 0  )    ,          (10)
             t                  \ a/                   P
where v is the air speed of the stack emission.  The  interpretation of  a


lidar measurement of plume opacity in terms of particulate concentration,


M, or total emission, M , requires knowledge of the optical-physical


parameter cr/M and hence is dependent on the particle  size distribution.
                                   38

-------
   500
      —     En (1 - O )  = -al_
                       2                           5
                       OBSERVED PLUME LENGTH, L — meters
       10

TA-8730-33
FIGURE 10   PLUME  OPACITY AS A FUNCTION  OF PARTICULATE VOLUME
            EXTINCTION COEFFICIENT AND OBSERVED PLUME LENGTH
                                39

-------
      The  lidar opacity  measurement  described  above  is  independent of  the

 lidar signal  from the plume  particulates,  P   (However,  the  large  sig-
                                           P
 nals  received from the  plume may  saturate  typical lidar detectors and,

 without their complete  recovery,  accurate  measurements  of P may not  be

 possible. )  The plume return P  may be  related  to plume mass concentration
                              P
 by  the expression:
                                  -2  /B \   2                           ,   .
                         P    =   K -R    [-  MT                            (ll)
                         p        p   \M/   c                           V   '
where


           K   =   a lidar constant


          T    =   clear air transmission to the plume
           c

           j3   =   volume backscatter coefficient.


The above statements applied to the ratio a/M also pertain to the quan-

tity 3/M-  Equation 11 assumes that:  particles are uniformally dis-

persed within  the laser pulse; a constant energy density exists across

the beam; attenuation by the plume particulates within the pulse may be

neglected; and P  is the maximum peak signal return, i.e., R  is the
                P                                           P
range at which the pulse becomes completely embedded within the plume.

As shown both experimentally and theoretically in Section III, the ratio

3/M is nearly independent of particle size at long wavelengths.  There-

fore,  less requirement is placed on knowledge of particle size for a mass

concentration estimate based on a backscatter measurement than for one

based on an opacity (transmission) measurement.


     Other advantages of the plume particulate backscatter technique

over the opacity technique are its adaptability to low density plumes

and its ability to take a direct reading of plume mass concentration.

In addition,  because of the relatively large plume returns, less
                                   40

-------
expensive lidar instrumentation is required.  However, disadvantages of



the particulate backscatter technique are the requirement of an absolute



lidar calibration, dependence of the plume return on the transmission



of the laser pulse along the clear air path between the lidar and the



plume return, dependence of the plume return on the emitted energy from



the lidar, and the greater sensitivity of the volume backscatter coeffi-



cient than that of the volume extinction coefficient to particle shape



and particle refractive properties.  The above operational aspects of



remote plume reading by observing backscatter of laser energy by the plume



particulates requires further investigation.







B.   Laser Pulse Length Considerations




     The lidar data analyzed in this report are the peak power returns



from both aerosol volumes and solid targets.  The data presented give in-



formation on the scattering properties of aerosol volumes that is needed



for the remote evaluation of plumes.  These data have been presented in a



manner that allows inferences about the accuracy with which remote plumes



may be evaluated without knowledge of the particle size distribution.



However, the experiments have been idealized in that the ratio of plume



length to laser pulse length has remained nearly constant.  Derivation of



plume density (mass concentration) from opacity measurements requires



information on the parameter o/M and hence depends on particle size dis-



tributions.  Backscatter has been shown to be less dependent on particle



size distribution (for long wavelengths) and a more measurable quantity



for low density plumes than either extinction or opacity.




     The lidar equation for peak power returns may be expressed in general



form as:
                   P(R  )  =   f  Y(R)P (R  - R)dR    ,                   (12
                      o      •-•       to
                             o
                                   41

-------
 where
           P(R )   =  peak power returned from range R
              o                                      o

                       —2     2
            Y(R)   =  KR  P(R)T (R)
              P   =  power  density
               T*



               K   =  a  lidar  constant.
 When the  pulse length is  much less  than the plume length (L « L ),  then


 P(R  )  sa  Y(R ) f P (R  -  R)dR.   The instantaneous power return at  a
   o        o  J   t  o                                  p m             -,

 given  range (R ) is dependent on the transmitted energy  P  P (R  -  R)dR

               o                                         [>   *  °        J
 and  the shape of the lidar backscatter signature is  nearly that of the


 transformation function Y(R).   A feature in Y(R), such as increased  scat-


 tering by a stack plume imbedded within cleaner air,  is reflected  in the


 function  P(R).   This case (L  < L )  relates  to the backscatter experiments


 where  it  was assumed in the data analysis that:
                         f  P  (R  - R)dR  =  K*P
                        J   tv o     '          t
                        o
where P   is  the observed peak  transmitted  power  and K*  is  a  constant,


The total area under the return signal (energy return)  is  given  as
                                    ^

                    f  P(R  )dR   =  K'P   f  Y(R  )dR     ,                 (13
                   «J     o   o       t «J     o   o
or the product of the pulse area and  the  area under  the  lidar  backscatter


signature  (received signal as a function  of  range).  Assuming  Y is  directly


related to aerosol concentration, the return signal  from a  plume of greater


density than its background may easily be distinguished  from the back-


ground, and the range integration of  the  return signal from the plume
                                  42

-------
would relate to total concentration of the plume after applying a correc-


tion for attenuation by the plume particulates.



     The pulse length may be larger than the plume it penetrates (L » L


In such a case Eq. (12) may be expressed as
                    P(R ) % P (R  - R) P  Y(R)dR
                       o      to      -J
and the lidar signature is more responsive to pulse shape than Y(R) vari-


ations.  Hence, for long pulse lengths it is more difficult to distinguish


the plume return superimposed on the background (clear air) return.  In


the extreme case of CW (continuous wave) lasers a single numerical value


represents the return integrated over the observed path.






C.   Polarization Measurements



     Important inferences on the nature of scattering particulates can


be made from polarization measurements of scattered light (Eiden, 1966;


Gilbert, 1970; Dave, 1969; Harris, 1970).  Since laser energy can be


emitted as polarized light, it is questioned if polarization measurements


of backscattered laser light can be used to derive particulate concentra-


tions.   The polarization of the incident light is conserved in the back-


scattered radiation from spherical particulates (except for circular


polarized light that is conserved except for reversal of the rotation


vector) and consequently is independent of the nature of the spherical


particulates.  Hence, any depolarization would be attributed to multiple


scattering and would be relatable to particulate concentration.  However,


the polarization of the incident light tends to be destroyed in the back-


scattered light from irregularly shaped particles.  Schotland etal(l97l)


has used this property to distinguish between water and ice clouds. Since
                                  43

-------
 the primary backscattered light  from smoke  plume  particulates would



 dominate over any multiple scattered light  and  since  the  particulates



 exhibit  varying degrees  of sphericity,  it is  unlikely that  backscatter



 polarization measurements could  be  related  to plume density.  A  limited



 number of  polarization measurements made during this  study  showed no



 relationship between  particulate concentration  and depolarization of the



 energy returns.   However,  some experimental error was introduced because



 of  the difficulty of  making polarization measurements with  coaxial lidar.








 D.   CW  Laser Techniques





     The present  study indicates that optical backscatter techniques,



 along with existing lidar  technology, may yield useful  information on



 mass concentration within  a stack plume.  However, several  characteristics



 of  current lidar  technology will tend to limit  lidar's  immediate useful-



 ness as a  practical,  low  cost, mobile monitoring  instrument.  The major



 limitations  are:   equipment complexity  and cost,  lack of  a  simple real-



 time data  processor within the receiver, and  high peak  power optical



 radiation  that  could  be hazardous to eye safety.





     The cost of Q-switch  crystal lasers (ruby, neodymium-glass, and



 neodymium-yag)  currently used in lidar  sensors  is typically in the



 315,000 to $20,000 bracket.  Although these prices have declined some-



 what in recent years  as the  result  of improved  technology and increased



 competition,  it is unlikely  that  the price of Q-switched crystal lasers



will decrease significantly  in the  near future.   Accordingly, most



Q-switched crystal lasers  are expected  to remain  unattractive for this



particular application in  the foreseeable future.





     In the present study  extensive use was made  of digital computer



 techniques to process  the  raw lidar data.  For  a  truly practical monitor-



ing instrument it is desirable that the lidar data be processed
                                   44

-------
automatically and analyzed in real time and that the result be displayed




with little or no intervention by the operator.  There seems to be no




adequate device for this purpose, although recent advances in integrated



circuits and memories may make feasible the use of hybrid analog and




digital computation techniques at a comparatively low cost.





     If these previously mentioned limitations can be overcome, the re-




sulting device should be a relatively simple, low cost, mobile instrument




that is capable of yielding accurate, repeatable measurements even when




used by relatively unskilled personnel.





     Application of CW lasers to the problem of remote evaluation of




plume density may overcome the high cost and hazard to eye safety prob-




lems of Q-switched crystal lasers.  Several CW laser techniques are




briefly discussed below.








     1.   Unmodulated CW Laser System





          The lowest cost and least sophisticated laser device for remote




evaluation of plume densities would use an unmodulated CW laser as its




energy source.  The received backscattered signal would represent the




backscattered light integrated over the viewing path.  This is equivalent




to path integration of the lidar backscatter signature from R = 0 to co.




Because of the inverse range-squared dependence of the lidar signature,




the nearby clear air return in a relatively polluted atmosphere may equal




the energy return by the remote plume.  In addition, the range to the




plume must be evaluated independently to correct for the range dependence




of plume returns.  Because of these problems an unmodulated CW laser back-




scatter device looks unattractive.  However, spatial separation of the




transmitter and receiver optics so that the receiver views only a remote




portion of the complete scattering volume may eliminate the clear air




scattering problem and give an estimate of plume range.  A separation
                                   45

-------
 distance of  8 to 10 ft  between  transmitter  and  receiver would  allow  the



 system to be mounted on a mobile  van or  stationwagon.  Some  effort may



 be  expended,  however,  to align  the  system for a particular remote air



 volume,  and  the system  optical  stability may limit  the accuracy of the



 plume  reading.   Another limitation  is that  the  optical receiver would be



 required to  detect  weak backscatter signals in  the  presence  of background



 sources  such as the sun,  street lights,  and the like.  Signal  discrimina-



 tion may be  possible with an  extremely narrow bandwidth optical filter



 or  by  modulating the CW laser with  a constant frequency and  including in



 the receiver electronics  a narrow bandwidth electrical filter  to reject



 the unmodulated background noise.








     2.    Pulse Modulation CW Laser System





          An  acousto-optic modulator suitable for helium-neon and argon



 lasers has been described by  Maydan (1970).  The modulator in  effect



 allows a  continuous  wave  helium-neon or argon laser to be operated in



 a pulsed  mode with minimum pulse  widths of  20 nsec and maximum pulse



 repetition rates of  2 MHz.  In addition, approximately 50 percent of the



 power within  the laser  cavity is  available  at the output, thus increasing



 the peak  power  of the pulse by a  factor of  50 to 100 over the  correspond-



 ing CW power output  of  the same laser.  The availability of a  very high



 repetition rate  laser transmitter makes it  possible for the first time



 to use various  signal-processing  techniques within the receiver.  Certain



well-known techniques from microwave  radar  technology, such as signal-to-



noise improvement by means of pulse-to-pulse averaging could be easily



employed, along with hybrid analog  and digital  real-time processing of



the resulting data.





          The use of the  acousto-optic modulator could, in theory, provide




a number of significant advantages  to lidar system performance.  These
                                   46

-------
advantages could make possible low cost, remote monitoring of smokestack

plumes.  The acousto-optic modulator offers these major advantages:

          (l)  It significantly increases the peak power output of a
               gas laser.

          (2)  It provides the ability to pulse the gas laser very
               rapidly (typically, 106 pulses/sec).

          (3)  It provides the ability to vary the width of the trans-
               mitted laser pulse.

          (4)  The rapid pulsing of the laser transmitter produces a
               received signal with a rapid data rate (e.g., a million
               measurements of plume backscatter per second).  This
               high data rate allows the use of range gating and video
               integration to increase receiver sensitivity and measure-
               ment accuracy over that derived using a single pulse.

When incorporated into a system, this technique may permit pulse (range)

measurements to be made without any danger to eyes.

          A simple system could incorporate the technique of automatically

sending out a transmit pulse shortly after detecting a receive pulse

from the plume.  The rate at which the laser transmits pulses would be

a function of the range to the plume.  The strength of the receive pulses

could then be corrected automatically for the range to the plume and
recorded; the operator would be needed only for pointing and turning on

the power switch.  Other range discrimination techniques may be applicable

and should be considered.

          Although detailed system performance calculations have not yet

been completed, preliminary calculations seem to indicate that this con-

cept is worthy of further investigation.
                                   47

-------
      3.    FM CW Laser System





           The FM CW technique  (Jackson,  1971)  provides  the most complete



 set of  capabilities and,  with  appropriate development,  at little or no



 additional cost or complexity  over  the other CW  laser systems.  Ease of



 operation and eye safety  features of  CW  systems  are advantages of the FM



 technique.   In addition,  this  technique  has  the  lowest  peak-power-to-



 average-power ratio and is  not vulnerable to the receiver overload and



 recovery  problems associated with pulsed systems.  The  sensitivity (hence,



 transmitted power required)  of a radar depends on the average transmitted



 power and,  for any given  average power required  from the laser, the FM



 technique results in  the  lowest peak  power.





           Calculations indicate that  it  will be  possible to:  modulate a



 low power CW laser beam;  detect scattered laser  light;  and measure, in



 real time,  the scattering as a function  of range, thus  deducing plume



 reflectance and plume opacity.  The unique characteristic of the technique



 is  that the laser beam is amplitude modulated by a frequency-modulated



 subcarrier.   A telescope  collects the light scattered by clear air and



 the plume,  and a photodetector recovers  the FM/CW subcarrier.  The FM



 permits the measurement of range, signal strength, and  range resolution



 by  standard FM radar techniques.  Real-time processing  can be obtained



 using frequency  filters and heterodyne techniques instead of the time-



 gating techniques  applicable to pulse systems.





          What  is  required at this time  are experiments to obtain engi-



 neering data  that  will be needed in future FM/CW lidar  development



 programs.   These  data (which are also required for other types of CW sys-



 tems) will  permit  engineering choices to be made of system parameters



 such as subcarrier frequency, subcarrier FM rate, size  and amplitude



weighting of  receiver aperture, frequency of sawtooth FM, laser type and



power,  laser  stability specifications, and photodetector specification.
                                   48

-------
Of primary interest in the experiments will be optical Doppler spreading




of the backscatter signal, conversion of laser FM noise into AM noise




owing to the depth of the scattering region, and receiver aperture




smoothing effects.





          To summarize, the unmodulated CW laser system requires the




lease amount of electronics but provides the least amount of information



required for plume evaluation.  However, the optical alignment problems



could outweigh the electronic simplicity for actual portable field systems.



The pulse-modulated CW laser system could obtain,without an eye safety




problem, the same data as a single-shot pulsed lidar.  The FM/CW technique



would require more development but could actually cost less and would pro-




vide a system capable of measuring both the backscatter and opacity of




smoke plumes.
                                   49

-------
                V  RECOMMENDATIONS FOR FURTHER RESEARCH


     The research conducted under this study has demonstrated that

measurement of laser energy backscattered from plume particulates is

useful for remote evaluation of smoke plumes and that evaluation of

backscatter techniques is possible by using an especially designed aero-

sol chamber to generate an artificial plume consisting of known particu-

lates at known concentrations.  However, a low cost laser plume-reading
instrument must still be developed and evaluated.  In addition, the

techniques used and the results obtained under this study suggest other

research areas.  Accordingly, it is suggested that research be undertaken

to (l) develop a low cost plume reading instrument, (2) extend the experi-

mental results of this study, and (3) further develop the aerosol facility

for use in a variety of aerosol studies and instrument evaluations.  In

particular, the suggested areas of further research are:

     •  Development of a low cost remote plume sensor

     •  Experimental evaluation with other particulates

     •  Aerosol chamber improvement for aerosol studies

     •  Extend the accuracy of backscatter measurements made with
        pulsed lidars

     •  Computations from Mie theory

     •  Development of multiple wavelength lidar techniques.

Each of these recommended research areas is discussed in the following

paragraphs.
                                  51

-------
A.   Development  of a Low Cost, Remote Plume Sensor





     Although high power lasers such as ruby (0.7 p,) and neodymium-glass




(l.06 p.) were used to investigate backscatter from fly ash particulates,



it  is unlikely that these lasers will find extensive application for



remote plume monitoring because of their high cost and complexity and



their potential hazard to eye safety.  Accordingly, further research



should be directed toward applying the results and techniques of the



present study to  develop a low cost laser instrument for reading coal-



burning power plant plumes.  Experimental development of CW techniques



based on bistatic, amplitude modulated and frequency modulated systems



for remote plume  reading is recommended.  Preliminary evaluation of



these laser systems could be accomplished using the experimental aerosol



facility developed under this study.








B.   Experimental Evaluation with Other Particulates





     The present  study dealt exclusively with fly ash particles.  The



results showed that optical characteristics could be characterized by a



specific average  particle size.  Fly ash particles are mostly spherical



in shape and heterogeneous in composition.  It would be desirable to



obtain optical data on other synthetic and real particulates to establish



the effects of shape and composition as a guide to developing a generalized



means for assessing the applicability of optical measurements.  Examples



of suggested synthetic powders, each of which has specific attributes,



are:  glass spheres,  aluminum spheres, flake aluminum, Cab-0-Sil, iron



oxide,  ammonium chloride, dioctyl phthalate.  Suggested actual industrial



dusts are:   cement, open hearth fume, paper mill recovery furnace dust,



limekiln dust, sulfuric acid, and incomplete hydrocarbon combustion




products.
                                   52

-------
     Where feasible, it is proposed to grade the dusts by size,  as was
done in the present study, and to evaluate both closely sized fractions
and composites.


C.   Aerosol Chamber Improvement for Aerosol Studies

     The present aerosol chamber provides a means of assessing laser
backscatter measurements in terms of aerosol properties.   In addition,
a facility to generate large volumes of easily accessible aerosols con-
sisting of known particles at known concentrations would be useful for
a variety of other studies including:

     •  Evaluation of techniques of deriving a valid aerosol sample

     •  Evaluation of direct aerosol samplers

     •  Comparison of various sampling instruments

     •  Development of optical aerosol sensors such as multiwavelength
        transmissometers and multiangle photometers

     •  Investigation of multiple scattering effects on instrument re-
        sponse functions.

     It is recommended that the aerosol chamber be further developed to
provide a unique facility for both remote and direct aerosol studies.

Some improvements include:

     •  The dust feeder system should be modified to produce a more uni-
        form particulate feed rate.

     •  Distributors should be placed at the chamber inlet to straighten
        air flow within the chamber.

     •  The problem of deriving representative aerosol mass concentration
        should be resolved.

     •  The present optical systems should be modified to eliminate drift
        in transmissometer readings and permit angular scattering measure-
        ments during daylight hours.
                                   53

-------
         Instrumentation  should be added to provide the aerosol size
         distribution measurements.
D.    Extending the Accuracy  of Backscatter Measurements
      Made  with Pulsed  Lidars

      High  powered Q-switched ruby and neodymium-glass lasers were used

to collect aerosol backscatter data for this study.  The accuracy of such

data  could be  significantly  improved by better monitoring of the trans-

mitted peak power and  the  total energy of the laser pulses, and by better

knowledge of the target reflectivity at lidar wavelengths used.

      Most Q-switched ruby  and neodymium-glass lasers exhibit significant

pulse-to-pulse variations  in output power and energy.  In a well-designed

laser cavity,  the pulse shape of the laser output is approximately Gaus-

sian  and does  not exhibit  a  significant pulse-to-pulse variation.  The

major pulse-to-pulse variations occur in the peak power output of the

laser.  For the present study, the peak power was monitored by means of

a beam splitter that directed a small portion of the output pulse onto

a diffusing surface that was viewed by a fiber optic light pickup coupled

to the receiver optics.  The output energy was assumed proportional to

the peak power; i.e.,  it was assumed that the pulse shape remained con-

stant.  The use of improved calibration techniques for assessing trans-

mitted energy  should result  in greater accuracy of the backscatter

measurements.

     The laser energy  reflected from a solid target of known reflectivity

was used in this study to derive an absolute value of the volume back-

scatter coefficient.   A black target was used rather than a standard white

target in order to reduce the target return to a level of the detector

response that  is linear.  In future backscatter experiments using this

technique,  more emphasis must be placed on evaluating the target reflec-

tivity at each lidar wavelength used.
                                   54

-------
E.   Computations  from Mie Theory




     Actual  aerosols  consist of various sized particles rather than



particles of  a  single size.  However, it is possible to characterize every



phenomenon by some effective average particle size.  The effective size



will be different  for different phenomena.  It is desirable to establish



what average  size  will characterize each of the common optical phenomena



that may be  used to determine physical aerosol characteristics:  trans-



mission, forward scattering, backscattering, and 90° scattering.  This



can be done  from the  Mie  theory for spherical particles.  Although it is



recognized that such  calculations may not apply to the irregularly shaped



particles encountered in  practice, such theoretical evolutions would be



of tremendous use  as  a guide to the type of engineering approximation that



might be useful for assessing the optical properties of such actual systems.



In the present  study, for example, experimental results indicated that the



Sauter diameter, D , was a good characteristic measure of the transmission
                   o *j


effects.  This was also confirmed by a few spot theoretical calculations



based on Mie  theory.  For backscatter, Mie theory predicts multitudinous



and abrupt changes of intensity with particle size; however,  for typical



particle size distributions, such variations tend to be evened out rapidly



so that the aerosol backscatter may be related to an average particle size.




     It is suggested, therefore, that calculations for spherical particles



be carried out  for various optical phenomena at various wavelengths,



particle refractive indexes, and particle size distributions.  The aerosol



optical parameters should then be related to aerosol physical properties.






F.   Development of Multiple Wavelength Lidar Techniques




     The results of this  study suggest that  multiple wavelength lidar



measurements  may provide  more valid estimates of particulate concentra-



tion than single wavelength lidar measurements.  It is suggested that the



two-wavelength lidar  data presented in this report be used to investigate




                                   55

-------
techniques of inferring higher moments of the size distribution from such



measurements.  Also, since the Mie theory computations relate well to



experimental data for fly ash particulates ,  it is suggested that the Mie



theory be used to investigate various aspects of multiple wavelength



lidar probing of plume particulates.   The number and position of wave-



lengths used and the manner of combining the measurements for an inference



of particulate concentration could be investigated by this means at rela-



tively low cost.
                                  56

-------
                              REFERENCES
Bryant, H. C., and A. J. Cox, 1966:   Mie  theory  and  the glory.  J. Opt.
    Soc. Am., 56, 1529-1532.

Conner, W. D., and J. R. Hodkinson,  1967:  A  study of  the optical prop-
    erties and visual effects of smoke-stack  plumes.   Cooperative Study
    Project, Edison Electric  Institute and U.S.  Public Health Service.

Cook, C. S., G. W. Bethke, and W.  D.  Conner,  1971:   Development of a
    mobile lidar system to measure smoke  plume opacity, paper presented
    at the Joint Conference on Sensing of Environmental Pollutants,
    Palo Alto, California, November  8-10, 1971.

Dave, J. V., 1969:  Scattering of  visible light  by large water spheres.
    Applied Optics, _8, 155-164.

Eiden, R. , 1966:  The elliptical polarization of light scattered by a
    volume of atmospheric air.  Applied Optics,  £, 569-576.

Ensor, D. S., and M. J. Pilat, 1971:   Calculation of smoke plume opacity
    from particulate air pollutant properties.   J. Air Pollution Control
    Association, 21, 496-501.

Evans, W. E. , 1967:  Development of  lidar stack  effluent opacity measur-
    ing system.  Final Report, Edison Electric Institute, SRI Project
    6529, Stanford Research Institute, Menlo  Park, California.

Gilbert, G. D., 1970:  The effects of particle size  on contrast improve-
    ment by polarization discrimination for underwater targets.  Applied
    Optics, 9, 421-428.

Harris, F. S., 1970:  Water-and-ice  cloud discrimination by  laser-beam
    scattering.  Paper presented at  the 1970  Spring  Meeting  of the
    Optical Society of America.

Holland, A. C. and G. Gagne,  1970:  The scattering of  polarized light by
    polydisperse systems of irregular particles. Applied Optics, £),
    1113-1121.
                                  57

-------
Jackson, D. VI., 1971:  New development of a CW laser radar.   Paper  pre-
     sented at Joint Conference on Sensing of Environmental Pollutants,
     Palo Alto, California, November 8-10.

Marquardt, D. W. , 1963:  An algorithm for least-squares estimation  of
     nonlinear parameters. J. Soc. Ind. Appl.  Math.,  11, 431-441.

Marquardt, D. W. , 1964:  Least squares estimation of nonlinear parameters.
     IBM Computer Library SHARE program 3094.

Maydan, D., 1970:  Fast modulator for extraction of internal laser  power.
     J. of Applied Physics, 41, 1552-1559.

McDonald, J. E., 1962:  Large-sphere limit of the radar back-scattering
     coefficient, Quart. J. Roy. Met. Soc., 88, 183-186.

Pilat, M. J., and D. S. Ensor, 1970:  Plume opacity  and particulate mass
     concentration.  Atmos. Environment, 4, 163-173.

Schotland, R. M.,  K. Sassen and R. Stone, 1971:  Observations by  lidar
     of linear depolarization ratios for hydrometeors.   J. Applied
    Meteorology, 10, 1011-1017.

Twomey, S., and H. B. Howell,  1965:  The relative merit of white  and
    monochromatic light for the determination of visibility  by back-
     scattering measurements.  Applied Optics, 4, 501-506.

Twomey, S., and H. B. Howell,  1967:  Some aspects of the optical  esti-
    mation of microstructure in fog and cloud.  Applied Optics, 6^,
    2125-2131.

van de Hulst, H. C., 1957:  Light scattering by small  particles.  John
    Wiley and Sons,  Inc., New  York, New York,
                                   58

-------
             Appendix A





DETAILS OF THE SIGHTING TUNNEL SYSTEM

-------
                               Appendix A





                  DETAILS OF THE SIGHTING TUNNEL SYSTEM








     Details of the sighting tunnel system, diagrammed in Figure 1 of the




main text, are discussed in this appendix.  A photograph of the system



is shown in Figure A-l.





     Air was supplied by means of a Buffalo Forge Type BL,  Model No.  365,




Arrangement 9, Class II centrifugal blower, belt driven at  2,080 rpm by



means of a 5-hp motor.  This blower discharged into the horizontal duct




system shown in Figure A-2.  The pressure difference across the 12-in.-




diameter nozzle located after the blower was used to measure fan air flow



rate.








Dust Feeder





     Dust was metered out by a proprietary grooved-disk feeder.  The dust




feed rate was determined by the size of the groove and the  disk rotational




speed.   The dust was exhausted pneumatically from the groove and fed into




the sonic velocity stream as shown in Figure A-3.  This sonic stream also




provided the suction for pulling conveying air through the  feeder.  The




pressure upstream of the sonic nozzle was 50 psig during a  normal run.




The compressed air flow was calculated as 33 CFM and the induced feeder




conveying air was estimated to be on the order of 2 CFM.





     Separate feeder calibrations were carried out to establish the feed




rate as a function of disk speed for each dust.   These were conducted by




pressurizing the feeder with a metered amount of compressed air (1.3 CFM)




that carried the dust from the groove into a separate collecting bag.  The
                                   A-3

-------
 collecting bag was weighed over timed  intervals.   The  results  of these




 calibrations are shown in Table A-l.   The dust  feed  rate was usually




 linear with disk speed;  in these cases the  average value of the grams/



 revolution measured at all dust speeds was  used  in all subsequent calcu-



 lations of feeder rates.   In a  few  cases a  significant variation in grams/




 revolution was observed;  thus no average value  is  shown for them in Ta-




 ble A-l,  and the calibration factor established  for  each disk  speed was




 used in subsequent calculations.





      Special tests were  conducted to observe  the effect of the compressed




 air pressure on dust  dispersion.  In the tests  the transmissometer reading




 was recorded on a strip  chart for a timed interval as the compressed air




 pressure  was changed  in  steps for a given feeder disk speed setting.  Typ-




 ical results are shown in Table A-2.   The test  results would indicate that




 a  reproducible state  of  dust  dispersion is  obtained  for pressures over




 30 psig.   At lower pressures  the apparent dispersion is progressively




 poorer, as  reflected  by  the greater transmissometer  readings.  The sharp




 drop in apparent dispersion at  pressures below  30  psig for the speed con-




 trol  setting of  100 was  partially due  to a  failure of the feeder groove




 to be unloaded  completely  at  that speed because of the lower induced air




 at  the  lower air pressures.   At  a speed control setting of 40 or less,




 the  groove was  essentially unloaded  at  all  air pressures.   Even at a set-




 ting  of 100,  no  more  than 20% of the groove was not unloaded.  Thus, any




 increase  in  transmissometer reading must reflect primarily the effect of




 dust  dispersion  and not any actual  reduction in dust feed rate.








Aerosol Chamber





     The dust-laden air stream is gradually decelerated in the expanding




ductwork and  allowed to enter the plenum chamber beneath the sighting




tunnel.   In  the original equipment Arrangement A, the top of the plenum
                                   A-4

-------
chamber was equipped with three rectangular nozzles pointing upward.   Two



were  at either end as shown in Figure A-2.  These two nozzles each had a




throat dimension of 3 in. x 20 in.  The third nozzle was located in the



center and had a throat dimension of 2 in. x 20 in.  The sighting tunnel




was filled with a controlled aerosol by the dust-laden air coining into




it through the central nozzle.  The two end nozzles served as an aerosol




curtain to carry away any outside air blown into the tunnel and hence to



establish a fixed boundary to the dust-laden air through which the lidar




would sight.





      Because sampling problems had become obvious in the first few runs,




the central nozzle was replaced after Sampling Run 8 with the multihole



distributor shown in Figure A-2 to minimize any effects on sampling that




the high velocity from the central jet might have.  This arrangement




(Arrangement B) was used throughout the rest of the study.








Air Rates





     With Arrangement A the air flow could be measured by either the pres-




sure differences across the 12-in. nozzle after the fan or by the drop




across the rectangular nozzles.  The total flow measured by these two




methods agreed within 2%.  The pressure drop across the plenum chamber




nozzles was slightly less with Arrangement B than with Arrangement A and




the system air flow increased slightly (approx. 2rc) for Arrangement B.




The total flow with Arrangement B was taken as 5800 CFM, including com-




pressed air introduced at the feeder.  It is believed that this value




should be correct within ±3%; the value also agrees well with the flow




to be expected from the fan performance curve provided by the manufacturer.




Although the air flow through the system could have been varied by pro-




viding a throttle at the fan inlet, this was not done.  All data were taken




at this same flow rate.   The flow rate also changed slightly with air
                                   A-5

-------
 temperature.   Since this variation should  have  been  less  than ±3%, it was



 not allowed for.   Because air rate variations with temperature were also



 not allowed for in the sampling system,  the  effect of temperature is



 largely self-compensating in so far as  overall  concentration comparisons



 are concerned.








 Flow Pattern





      The air  flow pattern in the sighting  tunnel was established qualita-



 tively by probing with short streamers  on  a  rod.  Apparently a marked



 rotary circulation in  the tunnel is induced  because the entrance to the



 plenum chamber  is located on one side.   The  central 55-hole distributor



 alone is apparently not sufficient  to eliminate the side  velocity com-



 ponent.   Viewed from the inlet  end  (east end) of the tunnel, the rotary



 flow was clockwise in  both halves  of the tunnel.








 Dust Deposition





      During the course  of the test  two inspections were made of the inside



 of  the  aerosol  chamber  system by disconnecting the 4-ft X 18-in. transition



 piece.   In both cases  it was observed that there was an obvious deposit



 on  the  floor of the  plenum chamber,  the sighting tunnel,  and the transition



 piece.   The sidewalls  and roof had  only a  slight coating  that was negligible



 compared with that on  the floor.  The ductwork upstream of the transition



 piece and including  the  3-ft upstream portion of the transition piece was



 essentially totally  void of  any  dust deposit.





      In the inspection on February  2, 1971 (after Sampling Run 8),  a por-




 tion of the deposit on the floor of  the transition piece was collected by



brushing and was weighed.  The weight indicated a deposit density of ap-



proximately (l/3) gram/sq ft.  Because this was the region of heaviest



deposit, it was estimated that the deposit in the entire chamber system
                                   A-6

-------
totaled less than 50 g.  Since this represented the entire accumulation




after an estimated 2 to 3 Ib of dust had been fed to the system,  it may




be  concluded that wall deposit within the chamber amounted to less than




5%  of the dust feed.  These measurements were made rather crudely, however,



and it was possible that the (l/3) gram/sq ft value was low because of



windage losses.





     On October 4, 1971, the chamber was opened for the second time and




was cleaned out completely with a small hand-vacuum cleaner.   The dust



from the various parts of the system was collected separately in a bag



filter and weighed.  The results are shown in Table A-3.  These measure-




ments also indicated that less than 5% of the dust feed was retained on




the chamber system walls.





     The material collected on both of the above occasions had a Fisher




diameter (corrected for slip flow) of 6.7 and 6.0 LL ,  respectively.  Exam-




ination under the microscope showed it to be substantially all in the 5-




to 10-|j, range.





     Because the chamber was subject to random vibrations and to deliberate




rapping during the experimental program, it is not possible to be certain




that a greater wall deposit had not accumulated but was dislodged because




of this mechanical action.  Considering the low velocities (4 to 8 ft/sec)




in the various parts of the chamber, it is not likely that significant




reentrainment of deposited dust occurred.








Transmissometer





     A transmissometer was mounted in the sighting tunnel.  The collimated




beam generator or light source was located at the upper right of the inlet




end of the tunnel and the detector at the lower left end of the outlet end.




Both light source and detector were located outside the aerosol zone in




ambient air.   The transmissometer details are given in Appendix D.
                                   A-7

-------
               (a)  FRONT VIEW OF AEROSOL CHAMBER SHOWING
                  LASER PULSE ENTRANCE
               (b)  SIDE VIEW OF AEROSOL CHAMBER WITH
                   PARTICLE FEEDER AND INSTRUMENTATION
                   FOR THE OPTICAL SENSORS
FIGURE A-1   AEROSOL CHAMBER USED  IN THE BACKSCATTER EXPERIMENTS
                                   A-8

-------
           EXHAUST AIR
         TO ATMOSPHERE
         6'
     20"
SOUTHEAST
      SIDE '
               20"
    -NORTHWEST SIDE


      'OPEN ENDS  (20" x 20")
                       2" HALF COUPLING
                                                                   ,DUST FROM FEEDER
                                                                     COMPRESSED
                                                                     AIR
                                                               PRESSURE TAPS
                                                               (1/8" Hole,
                                                               1/8" Nipple)
                                                                           FLEXIBLE
                                                                          COUPLING
V.
      SAMPLE HOLE NO. 8
WOODEN STAND
                                                  (a)  ELEVATION
                                                         12" o.d.
                                                                              18" o.d.
                                                                                                          FAN
                                                                                                         DRIVE
                                                                                       FAN OUTLET
                                                                                       20" x 14 3/8"
                                                                                                 	7   l_	1
                                                                                                  /  AA
                                                                                                       '   IN
                                                  (b)  TOP VIEW
                       18"
                               -6" R
                                                               32'-
                              (c)  INSIDE ELEVATION OF TUNNEL (Section A-A, Northwest Side)
   COLLIMATED
    LIGHT BEAM
                     EXHAUST
                       18"
                                6'
                             20"i
                                                                     AEROSOL
                                                                     CURTAIN
                                                     SIGHTING TUNNEi.

                                                                                         ATMOSPHERIC
                                                                                         AIR
               12"
u^v. i [—-34--— | 1(r
36" 55 1 1/4" HOLES 	 /\ } jl k
| ~"" 	 ~ " ~ PLENUM CHAMBER
00'

J\.
^^^



                                                                                                                DETECTOR
                         (d)   SECTIONAL VIEW OF CONTROLLED AEROSOL CHAMBER (Section B-B)
                                                                                                                TB-8730-19
                                         FIGURE  A-2   AEROSOL CHAMBER DETAILS

-------
AIR TO
PLENUM
            STAINLESS STEEL TUBE
                 5/16" o.d., 1/4" i.d. '
                RUBBER STOPPER
              12" o.d. DUCT
              1" PLUG WITH
                 3/8" HOLE
                                   1
                                                             AIR  AND DUST
                                                             FROM FEEDER
                                                             2 CFM
1/4" i.d. RUBBER HOSE
                                                  1" PIPE CAP
                                               -1" PIPE
                                                1" ELL
                                                            50 psig
                                                            COMPRESSED AIR
                                                            33 CFM
              AIR FROM FAN
                 5750 CFM
          FIGURE  A-3   DUST  DISPERSER  DETAILS
                                                                  TA-8730-20
                              A-ll

-------
                                                Table A-l

                                       DUST FEEDER CALIBRATION DATA
Data Taken
Groove Dimensions
Centerline Fly Speed
Diameter Depth Width Ash Control
(in.) (in.) (in.) Fraction Setting
3.50 1/16 1/8 0-10 t 15
30
40
(Groove volume = 50
1.408 cc/rev.) 60
80
100
200

5-10 u. 20

40

100


2.5-5 14 20
40
100
0-2.5 (i 20
40
100
3.50 1/8 1/2 0-10 p 20
20
(Groove volume = 40
11.27 cc/rev.) 100
5-10 p 20
40
100
2.5-5 |i 20
40
100
0-2.5 u. 20
40
100
Number
Revolutions
of
Disk
	
	
	

3
18
3
18
3
18

18
18
18
18
18
18
1
3
3
3
3
3
3
3
3
3
3
3
3
Powder
Collected
( Grams )
32.5
42.9
40.9
54.5
89.2
90.2
86.1
70.9

5.1
29.4
5.2
29.3
4.8
28.4

26.9
26.3
25.2
18.1
16.1
15.3
12.3
36.9
36.1
35.8
39.3
36.1
33.1
34.7
34.5
34.1
22.7
23.2
22.5

Time
(Sec.)
4620
1980
1200
1200
1800
1200
900
600

432
2882
152
911
57
327

2844
913
317
2840
943
320
157
486
154
54
490
155
57
492
156
54
480
160
53
Disk
Speed
( RPM )
0.255*
o.sost
1.224t
1.643t
2.05 *
2.81 *
3.58 t
4.20 t
Av,
0.417
0.374
1.183
1.183
3.16
3.30
Av,
0.379
1.183
3.40
0.380
1.145
3.37
0.382
0.370
1.17
3.34
0.367
1.16
3.16
0.366
1.15
3.33
Av.
0.375
1.13
3.40
Av.
Feed
Rate*
(g/rev)
1.66
1.61
1.67
1.66
1.45
1.61
1.60
1.69
. 1.64
1.70
1.63
1.73
1.63
1.60
1.58
. 1.63
1.50
1.46
1.40
1.005
0.895
0.850
12.3
12.3
12.0
11.9
13.1
12.0
11.0
11.6
11.5
11.4
11.5
7.58
7.73
7.50
7.60

Date
11/17/70



9/23/71






9/23/71


9/23/71


2/17/71



2/26/71


2/18/71


2/26/71


 Determined by collecting dust for timed period (10 to 80 min) and timing 1 rev.  of disk  for tests conducted
 on 11/17/70; all other determinations made by timing and weighing dust collected in  a precise number of
 disk revolutions.  For sp. gr. of 2.46 g/cc the following are calculated volume voidages,  e:

                       Groove             Dust             g/rev.               s
                     1/16 x 1/8          0-10 n            1.64              0.527
                                        0-2.5 u.            0.850             0.755
                      1/8 x 1/2          OrlO p,           12.0               0.567
                                        0-2.5 ti            7.60              0.726


.Measured by duplicate determination of time for 1 revolution.
                                                   A-12

-------
                       Table A-2

 EFFECT OF COMPRESSED AIR PRESSURE ON DUST DISPERSION*


                                         Transmissometer
 Speed           Compressed Air            Reading at
Control             Pressure               Equilibrium
Setting              (psig)	             (percent)
  100                  50                      38
                       40                      38

                       30                      41

                       20                      55

                       10                      63

   40                  50                      72

                       40                      71

                       30                      71

                       20                      77

                       10                      90

   20                  10                      92
*  General conditions:  Fraction dust fed:  2.5 to 5
                        Disk groove:  1/8 in.  x 1/2 in.
                          A-13

-------
                               Table A-3

     SUMMARY OF DUST COLLECTED INSIDE AEROSOL DUCTWORK AND TUNNEL*
                                                  Weight Collected
                                                    g [g/sq ft]
Transition Section                                 41     [2.5]

Plenum Chamber                                    311     [5.8]
     Central Section (6 ft long)                   74     [7.4]
     3 ft NE and adjacent to
        Central Section                            30     [6.0]
     3 ft SW and adjacent to
        Central Section                            46     [9.2]
     10 ft balance of NE section                   89     [5.3]
     10 ft balance of SW section                   72     [4.3]

Wind Tunnel                                       155     [3.1]
     Central distributor (3 ft long)              (20)t    [4]
     10 1/2 ft NE and adjacent to
        Central distributor                        67     [3.8]
     10 1/2 ft SW and adjacent to
        Central distributor                        51     [2.9]
     3 ft inside of NE  aerosol curtain              6     [1.2]
     3 ft inside of SW  aerosol curtain             n     [2.2]

Total Collected*                                  507 = 1.1 Ib
  *
    Substantially no deposit in duct upstream of transition section;
    upstream 3 ft of transition section also almost free of deposit.
    Weights given above were swept up from bottom with a vacuum
    cleaner.  The sides contained a coating of dust but the quantity
    of dust was negligible (< 20%) compared with deposit on bottom.
    All weights in grams.  Values in brackets are approximately
    g/sq ft.

  t Not measured or collected but estimated by noting depth and nature
    of deposit relative to other measured sections.

  * Total dust fed into system prior to these measurements was
    approximately as follows:
                                           Amount Fed
          Dust Fraction                       (Ib)
            0-2.5 ^                            9.0
          2.5-5   |j,                            7.7
            5-10  u.                           15.1
            0-10  u                            9.2
                                    Total     41.0

                                  A-14

-------
            Appendix B





FLY ASH PREPARATION AND PROPERTIES

-------
                              Appendix B





                  FLY ASH PREPARATION AND PROPERTIES








Source





     The raw fly ash used in this project was supplied  by  Diamond  Aggregate



and Fly Ash Company (501 Eleventh Avenue South,  Minneapolis,  Minn.  55415)




to the Donaldson Company, Inc. (1400 West 9th Street, Minneapolis,  Minn.




55431) for further treatment.   Diamond Aggregate had  this  material in




storage, but had originally obtained it from Northern States  Power



Company's Blackdog Power Plant in Minneapolis.   It  represents combustion



of Southern Illinois bituminous coal during January 1969.   The Blackdog



power plant, built in 1954, has 4 boilers and 4  stacks  and is equipped




with electrostatic precipitators.  It has an output of  475 megavolts when




all four boilers are in use.  Loading in the winter time is about  65-^.




It is the 4th largest power plant in the area and the only one from which




fly ash is readily available.   The plant generates  approximately 50,000 tons




of fly ash per year.








Treatment





     The raw fly ash was delivered to the Donaldson Company,  Inc.  (later




designated DCl) for size classification in their proprietary centrifugal




air classifier.  This classifier was originally  developed for Donaldson




by SRI.  Its unique features are the provisions  for achieving dust dis-




persion, which is normally a limiting factor in  achieving sharp size




classification in the range below 10 |j,.  It was  also developed to be




capable of production capacities corresponding to powder feed rates up




to 50 Ib/hr.





                                 B-3

-------
     Prior to classification, approximately 1000 Ib of the fly ash were



 screened through a Tyler 8-mesh sieve to remove traces of paper, gravel,



 and other superfluous debris.  The screened material was then passed



 successively through the classifier at settings designed to give various



 splits.  The complete flowsheets for all classification runs are given



 in Figure B-l together with resultant material balances.





     The first lot was successively classified at cut sizes of 10, 2.5,



 and 5u, to yield the fractions indicated.  A second lot was then classi-



 fied at 10 u, to yield a 0 to 10-u, fraction.  The 0 to 10-|i fraction ob-



 tained for the first lot has been designated as 0 to 10 u, (F) to dis-



 tinguish it from the corresponding fraction obtained from the second lot.



 All the 0 to 10-p, (F) material was used up for the successive classifica-



 tions.  The 0 to 10-|j, and the 0 to 10-p, (F) fraction should have been



 nominally identical.  There was, however, a slight difference in the



 operating conditions at which each was obtained with a corresponding



 slight difference in results, the 0 to 10-u, fraction being cut at a



 slightly finer size than the 0 to 10-|j, (F) fraction.





     In the classification system, the fines are collected in a cyclone.



Any fines escaping the cyclone are collected in a filter.  Because of the



 small amount of material involved , all filter fractions from all the



 classification runs were combined into a single fraction called "Super-



 fines," as indicated in Figure B-l.  Although a total of 9.0 Ib were col-



 lected in the filter as determined by direct weighing, only approximately



 5 Ib were physically recovered , the remainder being either left in the



pores of the filter medium or lost during shaking of the filter to remove



the dust deposit.
                                  B-4

-------
Properties


    Particle size analyses were conducted by DCI using the Fisher Sub-


Sieve Sizer, the Coulter Counter, and the MSA-Miitby sedimentation


techniques.  The Fisher analysis was checked by Stanford Research Insti-


tute (SRl).  The results of all analyses are summarized in Table B-l and


plotted  in Figures B-2 and B-3.  The mass media diameters given in


Table B-l were obtained from the 50% intercept of the curves shown in


Figures  B-2 and B-3.  The sp gr of material in all fractions was assumed


to be 2.46 for purposes of any calculations for interpretation of


results  of size analyses.  It  is likely that the sp gr will vary with


the fractions although such measurements were not made during the study.


A summary of the mean sizes derived from each method is given in Table


B-2.



       The data shown in Figures B-2 and B-3 can be approximated by a


log-probability relationship.  This was done by visually fitting the best


straight line to the data of Figures B-2 and B-3 giving emphasis to the


data in  the 10- to 90-cumulative percent range and ignoring those outside


of that  range.  These straight lines can then be described in terms of


two parameters, a mass media diameter and a standard geometric deviation.


These parameters are summarized for each dust fraction in Table B-3

                                      >r=
together with corresponding calculated  values of other mean sizes


(D   and D  ).  Since the Fisher Sub-Sieve Sizer supposedly measures
  32      mO

D   values, the Fisher data have also been included in Table B-3 for
 32

comparison.


    As might be expected all methods yield different particle size


results  since different particle properties are assessed by each method.
*                                                          M
 The calculation methods are described by C. E. Lapple in  Particle-

 Size Analysis and Analyzers," Chem. Eng., 7_5  (11), 149-156,  (Way  20,

1968)
                                  B-5

-------
Table  B-4 lists suggested effective values in terms of optical properties


for each of the dusts employed in this study.  These were derived as a


matter of personal judgment assuming that:  (1) the optical effects


are most nearly related to D   values; (2) the Fisher Sub-Sieve Sizer
                            o 2t

value comes closest to measuring D  ;  and (3) the distribution (in terms
                                  O &

of standard geometric deviation) can be taken from the MSA and Coulter


Counter data.  Because of these brash assumptions, these effective


values cannot be considered as having any great precision but they are a


as good as can be obtained in the present state of the art.  The values


of D   and D   given in Table B-4 have  been calculated from the quoted
    m3      mO
                                       *
value of D   and a by standard methods.
          32


    The fractions were examined at SRI both in the optical microscope


and in the scanning electron microscope.  Photomicrographs of the various


fractions are shown in Figures B-4, B-5, B-6, and B-7.



    No chemical analyses were made on the fly ash used in the project.


However, Diamond Aggregate and Fly Ash Company in 1966 supplied the


data given in Table B-5 on fly ash samples from two other Northern


States Power generating plants in the  Minneapolis/St.  Paul area.


These analyses are averages from samples taken over more than twelve


months of operation.



     The superfine,  0-2.5^1,  2.5-5|i,  and 5-lOp, fractions were prepared


by classifying the 0-10(1 fraction.  The size distribution of the 0-10(1


fraction can be calculated from the mass balance data of Figure B-l and


the Fisher diameter of each of the composite fractions, assuming that the


Fisher diameter is a reasonably good measure of the mass median diameter


of each of the composite fractions.   Since the standard geometric devia-


tion of these composite fractions  is of the order of 1.5, it can be shown
 Ibid
                                  B-6

-------
from the methods described by Lapple* that the mass median diameter will

be only about 9% greater than the D   diameter (which is presumably the
                                   O iL,
Fisher diameter).  On this basis the 0-10- fraction would have the follow-

ing size distribution:

               Particle Diameter             Cum.  tt't. To
               	microns	            Finer Than

               6.5 x 1.09 = 1.7                67.8
               3.1 x 1.09 = 3.4                22.4
               1.7x1.09 = 1.85                5.5
              0.77 x 1.09 = 0.84                0.8

A plot of these data yields a mass median diameter of 5.4 microns.

     It should also be noted that the Fisher size of the 0-10,. fraction

calculated from the measured Fisher sizes of the component fractions is

4.0^.   This checks exactly with the measured Fisher size for the 0-10^-

fraction.  In all the above calculations the SRI measurements of Fisher

size shown in Table B-l have been used.
*Ibid
                                 B-7

-------
     LOT
   NUMBER
    FIRST
RAW FEED
  INTER-
 MEDIATE
FRACTIONS
AND FEEDS
                              LOSS'
CLASSIFI-
 CATION
PROCESS f
 PRODUCT
FRACTIONS
   SECOND
*  Loss obtained by difference between feed weight and total weight of recovered fractions.
t  Mechanical failure occurred early in run and both coarse and fine fractions obtained prior
   to failure were discarded.
t  Rates given are powder feed rates to classifier.
 FILTER
MATERIAL
                                                                                       TA-8730-11
             FIGURE  B-1     FLOWSHEET FOR  CLASSIFICATION  PROCESSES
                                                B-8

-------
                         CUMULATIVE WEIGHT PERCENT OVERSIZE
             99  98   95   90   80   70  60 50  40  30  20    10   5    21  0.5 0.20.1
                O Raw fly ash
                O 0-10/J (F)
                • 0-10M
                O 5-10M
                A 2.5-5M
                D 0-2.5/J
                V Superfines
0.3
  0.1 0.2  0.5  1   2     5   10    20   30  40 50  60  70  80    90   95   98  99    99.8 99.9
                        CUMULATIVE WEIGHT  PERCENT  UNDERSIZE                  o,_ ,„
                                                                             TA—8730-1 2

  FIGURE  B-2  MSA-WHITBY PARTICLE SIZE ANALYSES OF BLACKDOG FLY ASH
                                       B-9

-------
              99  98
   CUMULATIVE WEIGHT PERCENT OVERSIZE
95  90    80  70 60  50  40  30  20    10
                                                                        1  0.5 0.2 0.1
                 O Raw fly ash
                         (F)
                 • 0-
                 O 5-10H
                 A 2.5-5M
                 O 0-2.5M
                 V Superfines
   0.10.2  0.5  1   2    5   10    20   30 40 50 60  70  80    90   95 9698  99    99.899.9
                         CUMULATIVE WEIGHT PERCENT UNDERSIZE             T  o,on „„
                                                                           T A-873Q- i 3

FIGURE 8-3  COULTER  COUNTER PARTICLE SIZE ANALYSES OF BLACKDOG  FLY ASH
                                      B-10

-------

                      ,~<
                                    -°
                                    >-v
                                                      O-OQ
                                                      •



                                                   iQ M
                                                          ••


                                                         O

         (a)  RAW FLY ASH
                                         (b)  0-1 Op FRACTION

        *
             «f b     ™
              ,    •

             •

           P      *



     •    *^

        |c)  5-1 Op FRACTION
                         ,



•   *              .     - •







          ."    "
                                 •



                                 IF

                                                    -
                           ^
                           -

                                                         0
             o a  °  ° ,0
                        '

              «#

                                      ^4n? 5
                                                - • -.   ' o p € '
                                                     o
                                                          «  O
                                                   •    >
                                         Uj-v-mol

                                     (d)  2.5-5.0P FRACTION




                                     r     J>J ^- "   ,W***^
                                  ,
                                                    -
    ^> •

^r
                                              '
"
   ' - '        -          •        .'TI-I'F r^.rri\ftl
                   1  .      -           -,   '•*.    -      *

    (e) 0-2.5P FRACTION                      If)  SUPERFINE

                SMALLEST DIVISION ON SCALE = 6.7/J



      FIGURE B-4   PHOTOMICROGRAPHS OF BLACKDOG FLY ASH
                                                       TA-8730-14
                              B-ll

-------
                                                  29 .D
                                      r
I
              i

    £r- ""-. ?*   ^V-'*V>:. &- ,
             .0   \  0*.??  #&f.
            .*   •  »  - . *   4  •'  -./''•
         .     ^      q - • .
MP
    —

                         V-i
                           (a)  RAW FLY ASH

                                                                . -
                                                                 •.
                                                                  -  '
    ;gr  o'f
                            o

                              > {.A|
                           •  •      /
                               ,
                           k V*  •  -
                                 .  •'
                                                           •  •  . • •*
                                                            ^     'i •*
 *-
         ft  •
     ,    .^.   v ,_v  %-,  ^ ;vv- . /
     *.•'*'•      '«.^    *    *A'~  '**,' '   * +
      *    "•    -' '_»•». ' " f           » s
-<                           . »
          v,   •  ..- .            ••  1  .
                          (c)  5-Wp. FRACTION
            NUMBER ABOVE EACH COLUMN GIVES MAGNITUDE OF SMALLEST
            SCALE DIVISION IN EACH PHOTOGRAPH IN THAT COLUMN
                                                             TA-8730-15
 FIGURE B-5  ADDITIONAL PHOTOMICROGRAPHS OF BLACKDOG FLY ASH AT LOWER
            MAGNIFICATION

                                B-12

-------
 (a(  0-10M FRACTION
                                                  (b)  5-10JU FRACTION
                                                                         -*
(c)  2.5-5M FRACTION
                                                 (d)  0-2.5M FRACTION
                               o      O
                            0



                           2.5 H     5jU       10/J





FIGURE B-6   SCANNING ELECTRON MICROGRAPHS OF  BLACKDOG FLY ASH




                                   8-13

-------
! '
a-
                          (a)  RUN 74A (0-2.5-fI Fly Ash)
(b)  RUN 80A (2.5-5-M Fly Ash)
                                                                    3 |U
              NOTE:  Filter paper:  G.E. "Nuclepore", '\~ll pore size.
                                                                                                                 TA-8730-21
                        FIGURE  B-7    SCANNING ELECTRON MICROGRAPHS OF  FLY  ASH  DEPOSITS ON  SAMPLING
                                       FILTER  PAPERS

-------
                                                                Table B-l

                                          SUMMARY OF SIZE ANALYSIS DATA  ON BLACKDOG FLY  ASH
W
l
        • 0 - !U[i (F)
        9.0 u
        7.2 .>.
        5.69 >
                              9.8 (}i!. 1 HI . 1
                              9 . -I 17.1 y 7 .
                                                                                                    ,1.61
                                                                                                    3.U7
99.5
SH.3
93.9
                                                                                                        .1. fi 10 . 0  5.3  3 .
                                                                                                        ] . 2  43  1.0  1 .
                                                                                                                       2, 46
                                                                                                                       0. 51

-------
                                                    Table B-2
                            SUMMARY OF MEAN PARTICLE DIAMETERS OF BLACKDOG FLY ASH
Size Analysis Method
Fisher Subsieve Sizer, diameter
corrected for slip-flow , u,
DCI
SRI
Coulter Counter, mass median
diameter , |_i
DCI
MS A- Whit by Sedimentation, mass
median diameter, u,
DCI
Fraction
Raw
Fly Ash
8.4
8.11
13.7
5.0
o-io u, (F)
4.5
4.00*
6.2
2.75
0-10 u,
4.4
3.95
5.0
2.70
5-10 u,
7.7
6.47
8.8
5.7
2. 5-5 |j,
3.9
3.10
4.1
2.97
0-2.5 p,
2.1
1.70
1.93
1.85
Superfine
0.64
0.77
1.43
1.77
w
i
M
         Calculated from analyses of component  fractions.

-------
                               Table B-3

                    LOG-PROBABILITY APPROXIMATIONS TO
                  SIZE DISTRIBUTION OF EACH FRACTION OF
                           BLACKDOG FLY ASH
                                              Diameter, micron
Fraction

Raw Fly
Ash

0-10i_i(F)


0-lOu


5-10u


2.5-5[a


0-2. 5u


Superfine


*
Size
Analysis
Method

CC
MSA
F
CC
MSA
F
CC
MSA
F
CC
MSA
F
CC
MSA
F
CC
MSA
F
CC
MSA
F
Standard
Geometric
Deviation

2
2

2
1

2
1

1
1

1
1

1
1

1
1


.74
.50
-
.41
.78
-
.48
.93
-
.51
.70
-
.50
.62
-
.52
.46
-
.64
.61
-
+
Mass '
Median
D

13
5

5
2

4
2

8
5

4
2

1
1

1
1

m3
. 5
.0
-
.80
.75
-
.75
.70
-
.60
.40
-
.00
.90
-
.85
.80
-
.46
.80
-
j.
Sauter "
D

8
3
8
3
2
4
3
2
4
7
4
6
3
2
3
1
1
1
1
1
0
32
.12
.28
.1
.94
.33
.0
.13
.17
.0
.90
.74
. 5
.68
.58
.1
.70
.67
.70
.29
.61
.77
j.
Number "
Median
D
mO
0
0

0
1

0
0

5
2

2
1

1
1

0
0

.64
.40
-
.57
.01
-
.40
.74
-
.17
.32
-
.44
.44
-
.09
.17
-
.70
.91
-
 CC = Coulter Counter; MSA = MSA-\Vhitby Sedimentation;  F  - Fisher
 Sub-Sieve Sizer
'Obtained from best straight line through data  in  log-probability paper
 in 10 to 90 cumulative percent range
*.
 Calculated from 3 and D
                        m3
                                   B-17

-------
               Table B-4


RECOMMENDED "AVERAGE" SIZE DISTRIBUTION
    FOR EACH FRACTION USED IN STUDY
Fraction
0-lOp.
5-10)0.
2.5-5p,
0-2.5)1
Standard
Geometric
Deviation
2.2
1.5
1.5
1.5
Diameter, microns
Mass
Median
ra3
5.45
7.06
3.36
1.85
Sauter
D
32
4.0
6.5
3.1
1.7'
Number
Median
mO
0.85
4.31
2.06
1.13
                  B-18

-------
        Table  B-5
TYPICAL FLY ASH ANALYSES
Generating Plant (Station)
Location
Physical Analyses
Specific gravity
Sieve analysis
"3 passing 200 mesh
73 passing 325 mesh
Blaine, surface area, sq cm/g
equiv diam, |_L
Hygroscopic moisture, /&
Chemical Analysis, $
A12°3
Fe2°3
CaO
MgO
S°3
Ignition Loss, 79
High Bridge
St. Paul

2.36

92
81
3110
8.15
0.02
49.6
Riverside
Minneapolis

2.50

97
91
3491
6.87
0.11
47.6
18.0 | 16.9
18.5 21.9
4.8 4.3
0.8 0.8
1.3 1.0
3.4 3.3
           B-19

-------
                Appendix C





DUST CONCENTRATION CONTROL AND MEASUREMENT

-------
                              Appendix C





              DUST CONCENTRATION CONTROL AND .MEASUREMENT








Dust Concentration Control





     The dust concentration in the system was controlled by establishing



known rates at which both air and dust were admitted to the chamber sys-




tem.  The dust feed rate was controlled by the size of groove in and the



rotational speed of the disk used in the feeder described in Appendix A.




The total air flow rate was essentially constant throughout all the tests,




a given fan being operated at a fixed speed on a system of fixed geometry.



The air rate was measured and monitored as described in Appendix A.








Calculation of Dust Concentration





     Since both dust and air were admitted to the aerosol chamber at fixed




and known rates, the dust concentration existing in the sighting tunnel




can be calculated if it is assumed that there is no deposition or accumu-



lation of dust within the chamber.  This calculated concentration (termed




"feeder concentration") is obtained by dividing the dust feed rate by the




total air rate.   The dust feed rate is established from the feeder calibra-



tion data given in Appendix A and the feeder disk rotational speed.  Al-




though the disk speed was determined by the speed control setting, some




small speed differences were noted at a given setting presumably because




of differences in frictional resistance in the drive.  For this reason




the disk speed was measured in each run by a microswitch that was acti-




vated momentarily once every revolution of the disk.  This generated a




mark on the transmissometer strip-chart recorder from which the actual




disk speed could be calculated for each run.






                                   C-3

-------
      Recognizing  at  the  start of  the program  that dust deposition in the



 aerosol  chamber could be a problem, the dust  sizes and air rates were



 chosen so  as  to minimize this difficulty.  Although it was not feasible



 to  eliminate  entirely deposition  with any reasonable system, it was ex-



 pected that such  problems should  be minor, reflecting a change in dust



 concentration of  less than 10%.   To achieve this, however, it was also



 necessary  to  disperse completely  the dusts used.  The compressed-air



 disperser  described  in Appendix A was designed to provide as good a dis-



 persion  as possible with known technology.








 Concentration Measurements Techniques





     To  provide a separate check  on concentration, separate measurements



 of  concentration were made by sampling the dust-laden stream from the



 sighting tunnel.  During the course of the program three different sam-



 pling  devices were used.





     In  the beginning of the program a simple probe inserted into the



 aerosol  chamber at right  angles to the nominal flow was used as shown in



 Figure C-l.  This arrangement was designated  as the "T" arrangement.  Al-



 though it  is customary to use isokinetic sampling when dealing with par-



 ticulates  to avoid sampling bias  caused by particle inertia, this would



 have been  very difficult  to achieve because of the low air velocities in



 the chamber.  It was also felt that with most of the dusts the inertial



 sampling bias should not be too severe and hence the simple form of sam-



 pling probe was used.





     Air was pulled through the sampling train at a constant and metered



 rate by connecting the sampling system to a high vacuum line through a



 critical flow orifice (approximately 0.025-in. diameter).   This orifice



was calibrated against a gas prover.  Sampling rate was held constant at



a predetermined value by maintaining a given upstream pressure on the
                                   C-4

-------
orifice by means of a valve located upstream of the orifice.   The pres-




sure downstream of the orifice was always recorded to be certain that



critical flow was being realized.  A sampling rate of 2.55 x/min was used.




The filter medium was a 1-in.-diameter General Electric ''Xuclepore" me-



dium of 1-u. pore size.  One side of this medium is smooth and the other




is matte.  The dust was allowed to deposit on the matte side.  This filter




medium was used because of its light weight (5 to 6 mg) and low moisture



retention (less than 0.01 mg, as established in another study).  Before




and after sampling, the filter disks were stored in an aluminum cup that




in turn was placed in a closed plastic box.  The aluminum cup was used




to mitigate the possible repulsive forces arising from electrostatic



charges on the filter medium.





     The second sampling technique used the Delrin filter holders shown



in Figure C-2.  These filter holders were placed over the 1-in. sampling




ports or located directly in the aerosol stream.  Two such units were al-




ways operated at the same time (although not at the same sampling loca-




tion) and air was drawn through each at a rate of 2.55 x/min by a separate




critical-flow orifice similar to that used for the first sampling tech-




nique.  The same "Nuclepore" filter medium was also used.  The filter




holders were used in four orientations:  (l) pressed over the 1-in.




sampling ports with the filter medium parallel to the chamber walls;



(2) vertical,facing into the nominal direction of air flow, so that the




filter face is normal to the nominal air flow; (3) horizontal with the




filter medium facing upward and parallel to the air flow; and (4) hori-




zontal with the filter medium facing downward and parallel to the nominal




air flow.  As will be noted from Figure C-2, there were two types of




holders:  (l) an "open" filter in which essentially the entire filter



medium (13/16-in. diameter free area) was exposed directly to the air stream;




and (2) a "closed" filter in which the air must pass through a nominal
                                   C-5

-------
 5/16-in.-diameter hole before arriving at the filter medium.  It should




 be  noted that  the air velocity in the sighting tunnel was nominally



 4.4 ft/sec  (133 cm/sec) whereas, at a sampling rate of 2.55 £/min, it



 was almost  the same [2.8 ft/sec (86 cm/sec)] in the inlet opening of the




 closed  filter.  In the "open" filter with an inlet opening of 13/16 in.,




 the velocity was 0.42 ft/sec (13 cm/sec).





     A  few  runs were made with the sampling arrangement shown in Fig-




 ure C-3.  This entire unit was inserted into the sighting tunnel with




 the nozzle  mouth facing into the air stream.  Large volumes of air were




 thus sampled by a vacuum cleaner, and dust was collected in a paper vac-




 uum cleaner bag in relatively large quantity.  The air flow was measured




 by  the  inlet nozzle to the system.  The velocity at the mouth of the




 nozzle  (2-in. diameter) was of the order of 15 ft/sec (460 cm/sec).  The




 air flow in this case was determined largely by the resistance of the




 bag.  Although the air flow was reasonably constant during a run, it




 varied markedly with the way in which the bag was inserted in the holder




because of  the close clearance between wall and bag.





     With all the sampling arrangements,  samples were taken at various




 locations,  indentified in Figure C-4.








Sampling Results





     The details of data taken during the series of samplings with sam-




pling arrangement "T" are given in Table  C-l.  In these runs the sampling




arrangement was held constant and samples were taken at a variety of lo-




cations with three of the dust fractions.   It was during the course of




these measurements that the Mark V lidar  shots were made.
                                  C-6

-------
     Since it was obvious that there were problems in obtaining a correct




sample, at a later date a second series of runs were made in which the



feed rate was held essentially constant (at the same speed control setting),




and the type of sampler and the sampler orientation and location were var-



ied with each of the dusts.  These results are given in Table C-2.





     To give a simple indication of the sampling runs,  the data were




reported as a ratio of the concentration calculated from the feeder speed



to that measured on the sampling filter.  These ratios  are summarized in



Table C-3 for the data of Table C-l and in Table C-4 for the data of



Table C-2.





     With the "T" sampling arrangement, the data of Table C-3 shows a




distinct trend of the concentration ratio with dust fraction.  Although




there is some trend with sample location, this is overshadowed by the




effect of dust fraction.  In the second series of runs  (Table C-4), the




concentration ratio also shows a distinct trend with dust fraction although



some effect of sampling arrangement is demonstrated.  In both cases this




is best illustrated by the following arithmetic averages of all concentra-




tion ratios measured for each dust fraction, leaving out concentration




ratio values that are obviously grossly in error.
AVERAGE CONCENTRATION RATIO (Feeder/Filter)
Dust
Fraction
0 -.2.5 u,
2.5 - 5 u.
5 - 10 u.
0 - 10 p.
For "T" Sampling
Arrangement
(Table C-3)
1.28
2.08
3.61
For Second Series of Runs
(Table C-4)
All
Data
0.98
1.53
1.75
1.83
Data for CHU
Arrangement Only
1.02
1.65
2. 39
2.10
Data for OHU
Arrangement Only
0.94
1.42
1.48
1.66
                                  C-7

-------
      It  is  apparent  that the concentration ratio is essentially greater


 than  unity  and  increases with  increasing dust particle size.  A concen-


 tration  ratio greater  than unity means that  the concentration measured


 with  the filter is less than would be calculated from the dust feed rate.





 Uniformity  of Aerosol  Concentration



      The best indication of the overall degree of uniformity of the aero-


 sol concentration in the chamber is through  the transmissometer strip-


 chart readout.   Figure C-5 shows typical recordings obtained for the


 various  dusts at a variety of  dust concentration levels.  The run numbers


 correspond  to those cited in Tables C-l and C-2.  Figure C-5 gives the


 nominal  transmissometer readout as a function of time.  The range in the


 transmissometer  scale  is nominally 20% to 100%, with 100% being at the


 left.  The  chart was calibrated during each  group of runs since there


 was some drift  in calibration  and the readout was not quite linear.  It


 is for this reason that the transmissometer  scale is termed nominal.  The


 100%  level  is almost correct on an absolute  basis but the indicated 20%


 reading  may have been  as high  as 25% on an  absolute basis.  The strip


 chart was operated at  its lowest speed of 1  mm/sec in all cases,  with the


 smallest time-scale divisions being 5 mm (hence,  5 seconds) apart.



     From Beer's law,  the aerosol mass concentration is related to trans-


mission by



                           w   =  K
                            P



where


                                    3
     w   =  mass concentration, mg/m
      P


      T  =  fractional light transmission,  dimensionless


                                        3
      K  =  proportionality factor,  mg/m
                                  C-8

-------
By  differentiation
Aw
P
w
P




P -,
AT
T £n(l/T)

Here  Aw /w   is  the  fractional change in aerosol concentration with a change



in  transmission, AT.  For values of T greater than 70<>, a close approxima-



tion  is:
Aw
P
w
P



AT
1 - T
-



~A(1 - T)"
(1 - T)

This says that for transmissions greater than 7CTc, the fractional change




in aerosol concentrations should be proportional to the fractional change



in opacity, (l - T).   In the region of T between 10 and 70% a reasonable



approximation is
                                  =  3 AT
     From the above relationships, the variations in concentration with




time can be estimated from the variation in transmission with time shown




in Figure C-5.  It will be noted that for the coarsest dust fraction, 5 -




10 M., which is relatively free flowing,  the concentration variation dur-




ing a run is less than ±5%.  For the finest and least free-flowing dust,




0 - 2.5 (J., the concentration variation is only some ±15%.





     The traces in Figure C-5 indicate occasional spurts in concentration,




especially with the finest dust.  These were demonstrated to result from tap-




ping or vibration of the dust feed line to the aerosol disperser; either




would momentarily dislodge a buildup of dust in this line.  This could also




                                   C-9

-------
be duduced from the fact that such a spurt is always in the direction of
increased concentration and the fact that a time of about 6 seconds

(twice the nominal aerosol chamber retention time)  is required  for this

perturbation in concentration essentially to disappear.

     It may, therefore, be concluded that the aerosol concentration  in
the tunnel remained uniform to better than ±15% throughout a run except

for an occasional short-lived increase.


Sources of Sampling Error

     The following are possible sources  of sampling error:

     (l)  Error in concentration of coarse particles because of failure
          to achieve isokinetic sampling conditions.   With sampling  ve-
          locities higher than those of  the stream  being sampled,  a  low
          apparent concentration of coarse particles will be measured;
          the reverse will be true if sampling velocities are too  low.
          The following is a summary of  approximate nominal velocities
          under the test conditions employed:

            In sighting tunnel                4.4 ft/sec (133 cm/sec)
            In plenum chamber                 7.3 ft/sec (223 cm/sec)
            Entering open "Delrin" filter     0.42  ft/sec (13 cm/sec)
            Entering closed "Delrin" filter   2.8 ft/sec (86 cm/sec)
            Entering "T" "arrangement          2.8 ft/sec (86 cm/sec)
            Entering bag sampler nozzle        15 ft/sec (460 cm/sec)

The above statements assume  that  the sampling port points into the gas

stream.  If the gas stream flow direction is other than directly into
the sampling port,  there will be  a  tendency toward measuring low ap-

parent concentrations.

     (2)  Deposition in sampling lead lines.   This  was probably a  major
          factor in the "T"  arrangement  in addition to any sampling  bias
         and  would result in the radically  lower measured concentrations
         observed with that  arrangement.
                                  C-10

-------
(3)  Electrostatic repulsion at the probe intake.   Particles  sub-
     jected to the violent impact and shearing in  the compressed-
     air disperser can acquire electrostatic charges,  the sign of
     which can be asymmetric.  When such particles deposit on an
     insulated wall, an electrostatic field is set up that will tend
     to repel those particles bearing the predominant sign of charge.
     Such electrostatic effects are likely to be more pronounced with
     the coarser particles.

(4)  Failure to retain particles on the filter medium.   This  can
     happen because of electrostatic charges or because of mechani-
     cal action.  The "Nuclepore" medium can acquire very large
     charges that could repel particles of similar charge and keep
     them from landing.  Mechanical vibration or flexure in handling
     the filter papers could result in a loss of deposited particles.
     Indeed, in a few cases it was observed that particles were pro-
     pelled into space when the filter paper was removed from the
     holder, although the magnitude of this action was  not assessed
     quantitatively.  This behavior could be caused by  either elec-
     trostatic or vibrational effects.  In either  case, the action
     would be expected to be most pronounced with  the coarsest par-
     ticles because of the much greater tendency of fine particles
     to adhere to the surface.  It was established that such  losses
     did not occur during weighing.  A few filters were weighed by
     successively removing them from and replacing them in their
     storage containers between weighings.  The loss in dust  deposit
     after several such manipulations was negligible (less than 5%
     of total dust deposit) when dealing with the  5 - 10 u dust frac-
     tion.  Such weighings, however, were made a relatively long time
     (hours) after sampling.  Thus, strong adhesive forces may have
     developed and electrostatic charges may have  leaked off.

(5)  Error in sample weighings.  The precision of  the weighing tech-
     niques was checked on numerous occasions and  weighings were
     always reproduced within ±0.020 rag.  Such an  error would be less
     than lOTc with all but the smallest filter deposits.  Although
     human error in weighing the clean filter papers could result,
     such error should be very infrequent as judged by the reproduc-
     ibility achieved in reweighing of dust-laden  filter papers.

(6)  Deposition of particles on the walls, roof, or floor of  the
     tunnel.  Under the test conditions such wall  deposition  would
     be most pronounced the coarser the particles.  The tunnel and
     dust conditions were selected to minimize such effects although
                             C-ll

-------
          such effects could not be completely eliminated.   As discussed
          in Appendix A, measurements of wall deposits in the entire
          tunnel after many sampling runs indicated that such deposition
          was probably less than 5% of the dust feed.   This conclusion,
          however, is not incontrovertible because of  the possibility  of
          dislodgement of wall deposits by vibration during periods be-
          tween samplings.  It was observed that what  little dust deposit
          occurred on the tunnel walls was almost entirely  on the hori-
          zontal surfaces.  However, theoretical calculations confirm
          that gravity settling could account for only a small amount  of
          deposition on the tunnel floor.
Discussion of Sampling Results

     The following is a summary of the results indicated by the sampling

data summarized in Tables C-l to C-4:

     (l)  The concentration ratio (feeder/filter)  is close to  unity for
          all filter arrangements and  sampling locations when  dealing
          with the finest dust fraction,  0 - 2.5 p,.

     (2)  For the coarser dust fractions the concentration ratio becomes
          progressively larger.   In the  case of filter arrangement "T"
          the concentration ratio increases steadily with increasing
          particle size and becomes quite large (3.6)  for the  coarsest
          size.   Although dust deposits  were observed in the "T" arrange-
          ment lead lines,  their magnitudes could not be established quan-
          titatively.   Wi-th the other  filter arrangements the  trend is
          the same, but there appears  to be a major  increase when going
          from the finest (0 - 2.5 u)  to the next  finest (2.5  - 5 (j, dust
          fraction..  The concentration ratio then  increased only slightly
          more for the coarsest fractions.
     (3)   Although  there  was  some trend of  concentration ratio  with  some
          filter  locations, the  trends  were somewhat  erratic with  the
          "T"  arrangement.  In the other arrangements the concentration
          ratio was more  dependent on the sampling  technique than  on the
          sampling  location.  The open  "Delrin"  filter collected more
          dust (gave lower concentration ratio)  with  horizontal filter
          medium  (OH , OH ) than did  the closed  filter (CH , CH ).   For
                    u    d                                 u    d
          the  vertical medium facing  into the  stream,  there was little
          difference in concentration ratio between open and closed  filter
          (OS, CS).
                                  C-12

-------
     (4)  Runs in which the filters contained a heavy deposit (86,  91-94)
          tended to give a lower concentration ratio but the trend  was
          not consistent.

     (5)  Runs in which a double filter mat was used (87 - 90)  with the
          deposit between the mats gave a lower concentration ratio
          (about 1.3 for the 5 - 10 p. fraction) than did most of the
          other arrangements.

     (6)  With all dust fractions a significant number of samples indi-
          cated a concentration ratio of close to unity for most sampling
          arrangements, and sampling locations.

     It is felt that the concentration determined by sampling is subject
to the greatest error and that actual concentrations are close to those

calculated from dust feed rates, although the actual concentrations must
be somewhat lower than those so calculated.  The following tabulation

represents an "educated guess" on the relationship between actual and

calculated concentrations.
                                    Concentration Ratio
                Dust Fractions        (Feeder/Actual)

                 0   - 2.5 p.               1.00
                 2.5 - 2.5 U               1.05
                 5   - 10 U                1.10
                 0   - 10 u                1.10
These values cannot be derived rigorously but are based on a considered
judgment of all the above considerations.

     Although particle inertia undoubtedly accounts for some of the error

in the sampling measurements, it is felt that a major factor is associated

with loss of dust from the filter paper due to either electrostatic effects

during deposition or dust loss while removing the filter paper from the

holder.  The fact that the particle was reasonably spherical could be a
significant factor in such accidental loss from the filter medium.  This
                                   C-13

-------
 conjecture has  not been proved but is suggested by the poor reproduc-

 ibility of the  sampling measurements.  A major consideration in this

 conjecture is the fact that the transmissometer readings showed excel-

 lent  agreement  with  feeder speed settings and did not correspond to the

 wide  variations in apparent concentration indicated by the sampling mea-

 surements for a given feeder speed setting.



 Sampling with Weathermeasure Instrument


      A few high-spot evaluations of the dust concentrations in the tunnel

 system were made by  sampling through an instrument produced by Weather-

 measure Corporation  (Division of KMS Industries), P.O. Box 41257, Sacra-

 mento, California  95841.  This instrument measures light scattered at

 90  from an incident light beam.  The light detector current is allowed

 to accumulate in a condenser that pulses whenever a given charge accumula-

 tion  is reached.  The count rate, which is proportional to aerosol con-

 centration, is  obtained by separately timing the number of pulses.  Each

 such  discharge  is registered as a count.  The instrument will also read

 count rate directly for rates over 100 counts/min.


      In all these runs air was sampled by means of the "l" arrangement

 shown in Figure C-l.   However, the Weathermeasure instrument was inter-

 posed between the sampling line and the filter.  All the measurements

 were made in the course of the sampling series recorded in Table C-l.

The data so obtained are summarized in Table C-5.


     The Weathermeasure instrument requires calibration for each dust;

hence, the results can be considered only relative.   The instrument has

been calibrated with 0.32-^1 stearic acid aerosol by the manufacturer;

for this aerosol 1 count/min is reported to correspond to a particle
                          3
concentration of 0.01 mg/m .   From the data of Table C-5, 1 count/min

for 0 - 10 (I fly ash would appear to correspond more nearly to a dust
                                 C-14

-------
                       3
concentration of 1 mg/m .   With the 2.5 - 5 |i fly ash,  1 count/min appears

                               3
to correspond to about 0.5 mg/m .   Since these results  are presumably also


subject to the sampling errors previously discussed for sampling with the
"T" arrangement, the actual instrument equivalents are probably about 1/2

                  .                        3
those cited above (i.e., 0.5 and 0.25 mg/m  for the 0 - 10 p and the


2.5 - 5 n fractions, respectively).



     In general, resonable reproducibility was obtained although there


was some problem with a drift in the instrument calibration setting.  The


manner in which the instrument was held also seemed to affect results.


It was held horizontally in all the runs reported in Table C-5.  The trend


with sample location corresponds reasonably well to those observed in the


filter sampling runs with the "T" sampling arrangement.
                                  C-15

-------
                                                           TUNNEL WALL.

                                                                1" HOLE.
                1/4" i.d. TYGON TUBE
                  AEROSOL IN
                                                                  AIR  OUT
POROUS METAL, PALL CORP
         GRADE F POROUS
         STAINLESS STEEL,
              1/16" THICK
                         /RUBBER STOPPER

                          COPPER TUBE
                          3/8" o.d., 5/16" i.d.
                                                   1/4" i.d. x 5/16" o.d.
                                                   FRICTION FIT
                                                     1/16" 'O' RING

                                                      3/32" 'O'  RING

                                                    1/8"
                                                           I
                                  SHT.,   ,rj$ys:i/i6"T  i 1-1/2-
                                            >
 QUICK OPENING  LATCHES
.TO CLAMP  TOP AND BOTTOM
 PIECES TOGETHER
                             BOTTOM VIEW OF FILTER
                                                                     TA-8290-14R
              FIGURE C-1   SAMPLING AND FILTER ARRANGEMENT
                                            C-16

-------



4


4



\
\
"---





^•.^

O"~-
-, ~-
;<;




>^H« 	
f(\V
?-:\
1 T/ft"



- 7/8" 	 *~V>*.

I I'*.



•^

^ v




;





i
1/8
i
1 1

3/4"
 1/4"
_L
                                                          9/16"
                                                      3/8"
      3/16"
                       -9/16-
                       - 7/8"-
                                                    ' —
                                                        3/32"
          (a)   OPEN GELMAN FILTER HOLDER (DELREN)
40 MESH SCREEN
           (b)   CLOSED GELMAN FILTER HOLDER (DELREN)
                                                      TA-8730-22
           FIGURE C-2   DELREN  FILTER  HOLDER
                             C-17

-------
O
K-1
00
              1" i.d.
         1 1/8"  o.d. TUBE
                 \
                     TO
                 . VACUUM      x RUBBER STOPPER
                  CLEANER
                                                                            12 1/2"-
   PAPER VACUUM CLEANER BAG
     FOR HOOVER MODEL 2900
         PART No. C-15631
         PACKAGE No. 16446
  (Approximately 1.0 sq. ft. Filter Area)
3 1/2" o.d. 16 ga Al TUBE
                                                                                                               PRESSURE
                                                                                                             1/8" HALF COUPLING
RUBBER STOPPER
.4"	^
                                                                                                                                                      -1/2"
                                                                                                                                                TA-8730-26
                                                             FIGURE C-3   BAG SAMPLER  SYSTEM

-------
n
i
                  10"
                  18"
                                           •—- 2'-
                                                                 •4'-
\
                                            3A
                                                  38
                                              • 3AA
                            30" •»-
                                  5A
                                    12"
                      4"
                   10"
                                                                ELEVATION
                                                               PLAN VIEW



                                                  FIGURE C-4   SAMPLING LOCATIONS
IT
J_
r
4"



1 3A2 22 |
1A t ^ t f
S3Al "Cl2l
2 \
V
1
/ ,2"
"P


                                                                              TA-8730-25

-------
                 0 - 2.5-
                                                            2.5 - 5.0L
                5.0 - 1CU
                                                             0 - 10L
                                Nominal Transmission, percent
DUST
FRACTION
0-2. 5t


2.5-5M


5-1 DM


0-lOfi


CHART
NUMBER
1
2
3
4
5
6
7
8
9
10
1 1
12
BUST COf/CENmriON
(Based en Feeder)
jig m^
16
48
146
27
74
222
32
73
206
35
84
241
SAMPLING RUN
NUMBER
75
51
48
79
34
33
91
100
101
65
27
29
JVEHAGE TBJNSHISSOXETEfl
READING CORING RON
85
66
29
84
74
42
95
S3
67
91
76
43
FIGURE C-5   TYPICAL TRANSMISSOMETER  STRIP CHART RECORDS OF VARIOUS  DUST
              SIZES AND CONCENTRATIONS
                                         C-20

-------
                                                                                  Table  C-l

                                                                SUMMARY  OP INITIAL SERIES  OF SAMPLING RUNS*
                 Run Number
       Date                             1/25/71      1/2G/71     1/2G/71      1/2C
       Equipment
         Arrangement  number            A            A            A            A

         Pressure differentials, mm
          xylenc
           Across nozzle               00           92           92           92
           Across plenum (2^G)         —           —           /19           .19
       Atmospheric  condition
         Temp,  '' C
         Barometer, mm II;;              770          772          772          772

       Oust type (fly ash)              0-10[i        0-lOu        0-10,1        0-W       O-lOjo,
       Feeder
         G roove .s i ze
^P       Va r i at: setting
to       RPM
h-1
       Filter
         Sampler type                  T            T            T            T
         Initial filter wei|;lil, mi;     5.707        5.591        5.7H7        5.80S
         Dual collected, nn;            0.2:19        0.205        0.100        0.117
         Time,  see                      600          (Hid          000          600
         Sampl in;; position              2266
         Samp 1 i nj; rai e ,  ,0,/m i n

       Aerosol  eoncent ra 1 i on,  m;1;,'m'
         From ti-ed  rate
         I''rom filler  samp 1 e
         Rat 10:  feeder/I' i HIM'

       Average Iransmissometer
        read i n;;
5
1/26/71
A
92
49
772
6
1/26/71
A
92
49
772
7
1/26/71
A
92
49
1H.5
770
8
1/2C/71
A
94
51
19.0
770
9
2/4/71
n
94
46
M.5
772
10
2/4/71
li
94
46
15.0
772

-------
                                                                       Table  C-l  (Continued)
O
I
to
Run Number
Date
Equipment
Arrangement number
Pressure differentials, mm
xylene
Across nozzle
Across plenum (2-6)
Atmospheric condition
Temp, °C
Barometer, mm Ilg
Dust type (fly ash)
Feeder
Groove size
Variac setting
RPM
Filter
Sampler type
Initial filter weight, rng
Dust collected, mg
Time, sec
Sampling position
Sampling rate, 4/min
Aerosol concent rat ion, mg/m^
From feed rate
From filter sample
Ratio: feeder/filter
Average transmissomcter
reading
Notes
11
2/4/71

B


94
46

15.5
772
0-10(1

1/16 x 1/8
100
3.27

T
5.887
0.081
GOO
3A
2.55

32.6
3.2
9.82

0.008
3
12
2/4/71

D


94
46

16.0
772
O-10H

1/16 x 1/8
100
3.21

T
5.586
0.261
600
3A
2.55

32.0
10.2
3.14

0.905

13
2/4/71

B


94
46

15.2
772
0-10(1

1/16 X 1/8
100
3.27

T
5.626
0.195
GOO
2
2.55

32.6
7.65
4.26

0.902

14
2/4/71

B


94
46

15.0
772
0-10(1

1/36 X 1/8
100
3.21

T
5.836
0.383
600
8
2.55

32.3
15.0
2.15

0.906

15
2/4/71

B


94
46

15.2
772
0-10(1

1/16 X 1/8
100
3.27

T
6.294
0.280
600
3
2.55

32. C
11.0
2.96

0.906
1
16
2/4/71

B


94
46

14.8
772
0-10(1

1/16 X 1/8
100
3.24

T
6.070
0.276
600
5
2.55

32.3
10.6
3.05

0.911

17
2/4/71

B


94
46

14.2
772
0-10(1

1/16 x 1/8
100
3.2-1

T
6.314
0.2G7
600
6
2.55

32.3
10.5
3.08

0,912

IS
2/4/71

B


94
46

14.0
772
0-10(1

1/16 x 1/8
100
3.21

T
6.255
0.221
600
6
2.55

32.3
8.7
3.71

0.912

19
2/4/71

B


94
46

13.5
772
0-10|l

1/16 x 1/8
100
3.27

T
6.157
0.323
600
5
2.55

32.6
12.7
2.57

0.914

20
2/4/71

B


94
46

13.0
772
0-10(1

1/16 xl/8
100
3.24

T
5.451
0.315
600
3
2.55

32.6
12.3
2.65

0.909


-------
           Table C-l (Continued)














0
1
to
w









Run Number
Date
Equipment
Arrangement number
Pressure differentials, mm
xylcnc
Across nozzle
Aeross plenum (2-G)
Atmospheric condition
Temp , ° C
Barometer, mm IU;
Dust typo (fly ash)
Feeder
Groove Si7.o
Vari ac sot t intf
RPM
Filter
Sampler type
Initial filter weif.ht, mi,e.
Dust colli'd eel , mi;
Time, sec
Sainpl i UK pos i t i on
Sampling rate, ^/iniu
Aerosol concent rat ion , mj.;/m-5
From feed rate
F rom i i 1 1 er samp 1 e
llat io: feeder/r i 1 1 er
21
2/5/71

I!


04
-10
10.0
70S
0-lOu,

1/10 x 1/8
20
0. 292

T
0.335
0. 171
3G02
:;A
2.55

2. '.11
1.11
2 . 55
22
2/5/71

1!


0-1
'10
17.0
7GB
0- 1 On

1/1G x 1/8
20
—

T
5. -1-12
o. Ir>o
:)Goo
3A
2.55

2. ill
0.9S
2. 97
23
2/5/71

1!


0-1
4G
1 5 . 0
7(iH
n-lOu,

1/16 x 1/8
•1C)
1 . 07

1'
G.2:!1
0.240
1KOO
3 A
2 . 5 5

1 0 . 7
3. 1 1
3.40
24
2/5/71

B


01
40
11 .8
7GH
0-1 0|i

1/1G x 1/8
40
—

T
G.:n4
o.2:iK
1800
•J,\
2.55

10.7
3.11
3.4 I
25
2/11/71

li


93
40
20.5
771
0-1 0|j,

1/8 x 1/2
20
0.321

T
(i.lHl
0.01G
830
3A
2.r.r>

21.2
0. 15
53.H
20
2/11/71

13


93
40
25.5
771
0-1 0|J,

1/8 x 1/2
20
0.332

T
G.284
0.217
liOO
3A
2.55

21.8
8 . 5
2 . 92
27
2/11/71

I)


93
10
') r •>
771
0-10(1

1/8 x 1/2
40
]. 15

T
G.025
0. 751
GOO
3A
2 . 55

X 1.0
20.4
L'.Kl
28
2/11/71

D


93
4G
21.8
771
0-lOp.

1/8 x 1/2
40
1.15

T
li.190
O.G58
GOO
3A
2.55

81.0
25.8
3.2G
29
2/11/71

li


93
40
25.0
771
()-10|i

1/H x 1/2
100
3 . 33

T
G.311
1 .GOb
GOO
3A
2.55

241.
G3.0
3. 82
30
2/11/71

U


93
10
24.8
771
0-lOu.

1/8 x 1/2
100
3.37

T
G.334
1 . 537
GOO
3 A
2.55

214.
GO. 2
1 .05
O.ilGG       0.9M9        0.918        O.921       0.7GI       0.770        0.182

-------
                                                                       Table C-l (Continued)
n
I
to
Run Number
Date
Equipment
Arrangement number
Pressure differentials, mm
xylene
Across Nozzle
Across plenum (2-G)
Atmospheric condition
Temp, °C
Barometer, mm Hg
Dust type (fly ash)
Feeder
Groove size
Variac setting
HPM
Filter
Sampler type
Initial filter weight, me
Dust collected, mg
Time , sec
Sampling position
Sampling rate
Aerosol concentration, mg/m^
From food rate
From filter sample
Ratio: feeder/filter
Average transmissometcr
read ing
31
2/18/71

B


93
46
15.8
770
2.5-5(1
1/8 x 1/2
40
1.05

T
6.200
1.302
600
8
2.55

73.5
51.0
1.44

0.727
32
2/18/71

B


93
46
15.8
770
2.5-5(1
1/8 x 1/2
40
1.05

T
6.176
1.301
eoo
8
2.55

73.5
51.0
1.44

0.734
33
2/18/71

B


93
46
15.4
770
2.5-5|l
1/8 x 1/2
40
1.05

T
5.969
0.731
6OO
2
2.55

73.5
28.6
2.57

0.740
34
2/18/71

B


93
46
15.4
770
2.5-5(1
1/8 x 1/2
40
1.05

T
6.171
0.770
000
2
2.55

73.5
30.2
2.43

0.738
35
2/18/71

B


93
46
16.2
770
2.5-5(1
1/8 x 1/2
40
1.07

T
6.195
0.843
600
3A
2.55

74.9
33.0
2.27

0.734
36
2/18/71

B


93
46
16.2
770
2.5-5|l
1/8 x 1/2
40
1.08

T
6.164
0.996
600
3A
2.55

75.6
39.0
1.94

0.737
37
2/18/71

B


93
46
16.0
770
2.5-5(1
1/8 x 1/2
20
0.297

T
6.130
0.281
600
3A
2.55

21.0
11.0
1.91

0.917
38
2/18/71

B


93
46
15.7
770
2.5-5(1
1/8 x 1/2
20
0.278

T
6.222
0.287
600
3A
2.55

39.6
11.3
1.74

0.920
39
2/18/71

B


93
46
15.2
770
2.5-5(1
1/8 X 1/2
100
3.20

T
5.980
2.733
600
3A
2.55

222.
107
2.08

0.417
40
2/18/71

B


93
40
15.0
770
2.5-5^
1/8 x 1/2
100
3.23

T
6.167
1.808
600
3A
2.55

224 .
74.4
3.01

0.387
      Notes

-------
                                                                          Table c-1  (Concluded)
















o
1
to
Ol












Run Number
Date
Kqu ipment
Arrangement number
Pressure, differential mm
xylene
Across nozzle
Across Plenum (2-6)
Atmospheric condition
Temp , ° C
Barometer, mm 1[^
Dust t ype ( fly ash)
Feeder
G roove s i ze
Vari ac sett 'nix
KI'M

Filter
Sampl el1 t ype
Initial filter weip,lit, mr;
Dust collected, mr;
T i me , .sec
Samp 1 i n)', pos i t i oa
Sampling rale, £/min
Aerosol conren 1 ra ! i on, i;i;;-/!a:}
From food rate
F rum filter samp 1 e
Rat 10: feoiler/l i 1 1 or
A ve ra[;e t ransmi ssor.u • 1 e i'
road i IH:
No! os
41
2/25/71

B


93
4G

9.8
773
0-2.5(1

1/8 x 1/2
20
0. 255


T
G.085
0.292
GOO
3 A
2 . 5 5

11.8
11.1
1 .03

0.902

•12
2/25/71

B


93
4G

10.8
773
0-2.5',!

1/8 x 1/2
20
0.2G3


T
G. 153
0.301
GOO
3A
2 . 55

12.2
11.8
1.03

0.883

43 44
2/25/71 2/25/71

B B


93 93
46 46

11.4 12.0
773 773
0-2.5(1 0-2.511

1/8x1/2 1/8x1/2
40 100
0.983 2.92


T T
G.115 6.180
0.927 3.029
600 GOO
3A 3A
2.55 2.55

45.5 135.2
3G.3 118.8
1.25 1.11

0.652 0.311

45 40 17 18 49 51 51
2/25/71 2/25/71 2/25/71 2/25/71 2/20/71 2/25/71 2/25/71

n B B B B B B


93 93 93 93 93 93 93
46 46 4G 4G 46 4G 46

12.5 — 13.0 13.3 12.8 12.3 12.6
773 773 773 773 773 773 773
0-2.511 0-2.511 0-2.5(1 0-2.5(1 0-2.5(1 O-2.5ji 0-2.5(1

1/8 x 1/2 1/8 x 1/2 1/8 x 1/2 1/8 x 1/2 1/8 x 1/2 1/8 x 1/2 1/8 x 1/2
100 40 40 100 10 -10 40
2.92 1.05 1.02 3.16 1.04 1.03 1.03


T T T T T T T
G.lll 6.213 6.087 6.179 6.161 0.201 G.22G
2.712 1.210 1.001 2.877 1.008 0.785 0.780
600 liOO (500 GOO GOO 605 GOO
3A 8 3A 3A 8 2 2
2.55 2.55 2.55 2.55 2.55 2.55 2.55

133.2 18. 5 17.2 11G.3 18.2 17.7 17.7
10G.3 18. G 39.1 112.8 39. G 30.5 30. G
1.27 1.00 1.20 1.30 1.22 1 . 5G 1.56

0.315 0.586 O.G53 0.291 O.G31 o.ii:')] O.GG1

Not.(.>H :   1.   1'i 11 cr  \ips i (It-  (in\ui i n pl;u;t i r  box be Con1  \vi> i r,-h i nj;.
         2.   Some rlust  ;uJlK>rini; to  in.si/Jt.- 1 i(J  of  nili.-r holder box,   Hriu-e dust  collected rni;ht  be recorded as  tno low.
         .'t.   Under m i crosi-opr  1 h i s  filter eont a ined  nboul 1 /•!  ;u; much null or i :il  as filter from  Hun 1^ by  v i r-uia 1  rst iniat e .
         •t.   I'eed rat o  est iniat ed ,  not  measured .

-------
                                                                         Tnble C-2

                                                        SUMMARY OK SECOND SERIES OF SAMPLING RUNS*
Run Number
Date
Equipment
Arrnngement number
Pressure differentials, mm
xylene
Across nozzle
Across plenum*"" '
Atmospheric conditions
Temp, °C
Barometer, mmllg
Dust type (fly ash)
Feeder
Groove size
_ Variac setting
1 HPM
to
C* Filter
Sampler type
Initial filter weight, nig
Dust collected, nig
Time, sec
Sampling position
Sampling rate, £/min
Aerosol concentration mg/m
From feed rate
From filter sample
Ratio: feeder/filter
Average transmissometor
reading
Notes
52A 52B
8/31/71

B


89
41

25.2
--
0-lOu

1/10 x 1/8
100
3.31


T OS
6.315 0.400
0.155 0.525
707 707
2 2 1
2.55 2.55

33 . 3
5.10 17.5
6.45 1.90
0.913


53 A 53B
9/2/71

B


89
13

27.8
702
0-lOp.

1/10 x 1/8
100
3.28


OF OS
0.235 G.OG8
1 .051 0.838
832 832
2 2 1
2.55 2.55

32.9
29.7 23.7
1.11 1.39
0.908


54 A 54 B
9/2/71

B


89
43

27.8
702
0-lOu

1/10 x 1/8
100
3.33


OF OS
0.891 0.877
0.691 0.710
GO 3 003
2 2,
1
2.55 2.55

33 .2
27.0 28.9
1.23 1.15
0 . 903


55A 551!
9/3/71

B


88
13

28.5
701
0-10u

1/16 x 1/8
100
3.10


OF OS
0.215 G.095
0.158 0.3G5
059 059
3 A 3Aj
2.55 2.55

31 .0
1 0 . 4 13.0
2.07 2.02
0.905


56 A 56 B
9/3/71

B


87
42

—
761
0-10u

1/10 x 1/8
100
3.44


OF OS
6. 020 G.200
0.080 0.505
016 016
3 A 3A2
2.55 2.55

34.3
25.9 21.6
1.32 1.59
0 . 900


57A 57B
9/3/71

B


80
41

3 1 . 9
761
0-lOu.

1/10 x 1/8
100
3.43


OF OS
6.203 0.099
1.058 0.520
017 017
2 22
2.55 2.55

34 . 2
40.3 20.0
0.85 1.71
0 . 900


58A 58B
9/9/71

B


91
45

21.1
7G2
0-lOp,

1/16 x 1/8
100
3.33


OF OS
C.I 66 6.000
0.010 0.170
600 000
2 2y
2.55 2.55

33.2
25.1 18.1
1.32 1.80
0.905


59A flnil 60A GOB
9/9/71 9/9/71

B B


91 90
1 5 1 5

22.2 23.2
762 702
0-lOu, 0-lOu,

1/10 x 1/8 1/16 x 1/8
100 100
3.33 3.33


OF OS OF OF
0.135 0.129 G.9CO G.2.14
0.895 0.199 O.OG3 0.221
600 GOO GOO 600
2 2C 8 7
2.55 2.55 2.55 2.55

33.2 33.2
3 5.1 19. 5 2.5 8.8
0.95 1.70 13.3 3.77
0.908 0.905

1
*  Common conditions:  tunnel air flow 5800 CFM including compressed air; compressed air:  50 psi.
   T  =  copper tube with 90° bend
   0  =  open "Delrin"
   Hu =  filter horizontal facing up
   S  =  filter facing into stream
F  =  filter flush with wall
C  -  closed "Delrin"
11^ =  filter horizontal facing down
V  =  vacuum cleaner bag sampler.

-------
n
                                                                                     Table  C-2  (Continued)


                  Run Number	    61A     6in     62A     621?     G.'iA     G3li         G4        G5A     G5I1     (iGA     G6H     67A     67I1     (iSA     6HJ1    G9A     G9W
Date
Kcju i pmen t
Arrangement number
Pressure differentials, min
xylene
Across noxzle
Across plenum^""*' '
Atmospher ic concl i t. ions
Temp, C
Haromcl er , mmlli;
Dust typo (fly as]))
Feeder
Groove' s i 7,e
Var i ac .sot t i nf^
HPM
Fi 1 tor
Sampler type**
Initial I'i 1 ter \voiu lit . niu'
9/9/71 9/9/71

1) 1!


89 89
•1 3 1 3

28.1
702 702
o-iou o-jou

I/Hi x 1/K I/Hi x 1/H
100 100
3 .33 3.11

OK ())!„ OF Ol,,
(i. 109 7(i] 7(il 701
0-lOn O-10(J, 0-lOu O-IOU 0-JOu O-K'U

I/Hi x 1/8 I/Hi x 1/8 1/10 x 1/8 1/10 x 1/8 I/Hi x 1/K I/Hi x 1/K
100 100 100 10O 100 IOO
3.11 3.T)7 3.T>1 'i.37 3.37 3.37

0,S OH,, ("11,, OH,, CHU OH,, HI,, OH,, <'!lu Oil,, Til,,
G.2M ri.riio r>.7iiH 5.009 ri.nni r..2in <;.o79 0.2. 1:1 (;.I33 fi 775 r>.70<;
          Dust eolleeted,  tn^             O.17T)   O.(i91   O.19K  O . ;>:r/  O . :!2O   1I.3H1   O.O11         O.fiiO  O.fiKi  O . Ti 1 8   0.112   O.3S7   O.:':V  O.171  O.IS(i  O.liMd   (l..r)39
          T i me,  sec                       (iOO     (iOO     000     (iOO    (ion     (>oo     Hi            730     73(i     (ioo     lion        (ion             (id!)             (ioo
          Sampling  position              3 A      3A,     3A      .'iA,     7        8       3A(              3 \c             3A(.             3A(.             3A(.             '>c


        Aei-osol  eoneen I r a t i on  ms1,/nr
          From feed rate                      33.2           31.0            31.0        31.0             3."i.0            3:",.2            33.(i            33.0            33.0
          From filter sample              18.(i    27.1    19 . li    20.7   K.ii     IT..(I    —            2H..'i    1 7 . r,    2O.3     10.2    1 '.'< . 2    9.9     1 S , r,    1 •) . I    27.3    21.1
          liatio:    feeder Ti Her          1.79    1.'.!:!    1.71    I.01   3.91,    2.20    --            1.71    2.01    1.7'!     2.1H    2.21    3.39    1 . HL!    1.70    1.23    l.:',9
        Not r

-------
                                                                          Table C-2 (Continued)
O

to
00
Run Number
Date
Equipment
Arrangement number
Pressure differentials, mm
xylene
Across nozzle
Across plcnum(2-6)
Atmospheric conditions
Temp, °C
Barometer, mmllK
Dust typo (fly ash)
Feeder
Groove size
Variac setting
HPM
Filter
Sampler type**
Initial filter weight, mg
Dust collected, m(?
Time, sec
Sampling position
Snmplinp rate, ,^/min
Aerosol concentration mg/m
From feed rate
From filter sample
Ilatio: feeder/filter
Average transmissometer
70A 70B
9/20/71

D


87
43
23.0
763
5-lOp,

1/16 x 1/8
100
3.26

°"u C1IU
6.170 5.708
0.438 0.413
600
3AC

31.3
17.2 16.2
1.82 1.93
0.955
71A 71B
9/20/71

n


87
43
23.0
763
5-lOp,

1/16 x 1/8
100
3.26

5.560 0.1 «3
0.492 0.32G
600
3 Ac

31.3
19.3 12.8
1.G2 2.44
0.945
72A 72D
9/20/71

n


87
43
23.0
7G3
G-lOp,

1/16 x 1/8
100
3.26

°»u C"u
5 . 70 1 6 . 070
0.761 0.299
600
2c

31.3
29.9 11.7
1.05 2.08
0.950
73A 73B
9/20/71

B


86
44
22 5
763
G-lOp.

1/16 x 1/8
100
3.27

°»u C"u
5.567 5.953
0.705 0.345
600
2
c

31.4
27.6 13.9
1.14 2.20
0.945
74A 74B
9/27/71

B


90
44
21.6
7GG
0-2. GH

1/16 x 1/8
100
3.22

Oil CII
5.930 C.045
0.425 0.413
596
3Ac

10. 7
16.8 16.3
1,00 1.02
0.863
75A 75B
9/27/71

B


—
—
21.9
766
0-2. Dp,

1/1G x 1/8
100
.3.17

OH CII
u u
5.539 0.0 16
0.532 0.449
638
o

16.4
19.6 16.5
0.84 1.00
0.860
76A 7611
9/27/71

n


89
44
21.9
76G
0-2. 5u,

1/16 x 1/8
100
3 . 27

0.094 5.509
0.4G2 0.436
653
:IAc

16.9
16.6 15.7
1.02 1.08
0.863
77A 77B
9/27/71

n


—
—
21.9
766
0-2. 5p,

1/16 x 1/8
100
3 . 23

OH CII
u u
5.951 5.678
0.494 0.451
611
2C

16.7
19.0 17.3
0.88 0.97
0.855
7SA 78B
9/28/71

B


89
44
18.8
764
2.5-5U

1/16 x 1/8
100
,"! . 20

Oil CII,.
u »
0.124 5.058
0.400 0.440
638
:!Ac

27.2
17.2 16.2
1.58 1.G8
0.840
        reading

       Notes

-------
                  Table C-2 (Continued)
















0
1
to
to








Hun Number
Date
Kqu ipment
Arrangement number
Pressure d 1 1' ferent ials , mm
xylene
Across nozzle
Aeross plenum '
Almospher ic coiuli t ions
Temp, °C
Barometer, mmlln
Dust typo (rly ash)
Feeder
(Iroove s i ye
Var ine set t in^
11PM

F i U e r
Sampler type**
Initial f i Her \iein lit. , mi;
Oust ro 1 lee t ed, in);
'I'ime, see
Samp! i n;p, pns i t ion
Sampling; I ' a t f ' , f / Iti i 1 1
Aerosol ronrentral ion nir,/'"'
From t'eeil rale
From niter sample
Hat i<>: IVeder/l i 1 t < r
79A 7011 80A 8015
9/28/71 9/28/71

n B


89 89
'1 5 4 f>

18.:!
7G4 7(11
2.5-510, 2.S-5U.

I/Hi x 1/K 1/1G x 1/8
100 10(1
.'!.!!) 3.28


on en on en
u u " »
5.750 5.(i21 (i.051 5.711
0.5:i(i 0.4G1 0.447 0.423
(115 (100
:'-,. :IA(.

27.2 27.!)
19.5 1(1. '1 17.5 1(1.11
1 .:!'.! 1 .(11 1 .59 1 .(18
81A SIB
9/29/71

B


89
43

22.5
7(i2
5-lOp.

1/10 x 1/8
100
:i . 21


Oil C\\
U I!
5. (11 2 5.5(19
0.153 0.:iGl
(100
3A
c

:i2. i
17.7 11 . 1
1.81 2 .28
82A 82D 83A 83B 84A 8111
9/29/71 9/29/71 9/29/71

11 B H


89
44

23. 0
—
5-lOu 5-lOp. 5-lOu

1/1G x 1/8 1/1G x 1/8 1/1G x 1/8
100 100 100
3.25 3.28 3.2(1


Oil HI Oil Cll Oil Cll
11 U (1 (1 ll tl
5.G23 5.1115 G.OR2 5.1185 5. (151 5.58G
0.173 0.315 (1.521 O.322 0.112 0.211
(127 (100 (115
3A _ 3A(. 3 A

32.2 32.5 32.3
17.7 11.X 20.5 12.11 1(1.9 9.3
1 .82 2 .73 1 .5S 2 .5H 1.91 .'! . 1 7
85A 8511 8BA 8611 87A 87B
9/29/71 9/29/71 10/8/71

11 B B


89 00
•11 43

21.0 20.5 21.4
7GT>
5-lOp, S-ldu f)-I()n

1 /Ifi x 1/8 1 /Id x 1 '8 1 'Hi x 1 '8
100 100 100
3.30 3.30 3.25


OS CS OS CM OS OS
5.974 5. (115 (1.037 5,925 5.73(1 12.913
(1.884 0.899 1.190 1.2(10 O.(i59 0.519
1 200 12OO (i 1 L'
3A 3A 3A
'' rr . '' .,-''.

32.7 32.7 32.3
17.3 1 '! .11 23 .3 21.7 35.3 21.1
1 . H9 1 .8(1 1 . 1(1 1 .32 1 .28 1 , 53
O.K75

-------
                                                                        Table C-2  (Continued)
               Run Number
03
a
Equipment
  Arrangement number
  Pressure d ifferentials,  mm
  xylcne
    Across nozzle
    Across plenum'^-6)

Atmospheric conditions
  Temp, °C
  Barometer, mmHg

Dust type (fly ash)
Feeder
  Groove size
  Variac setting
  RPM
Filter
  Sampler type
  Initial filter weight, mg
  Dust collected, mg
  Time, sec
  Sampling position
  Sampling rate, J>/min
Aerosol concentration mg/m
  From feed rate
  From filter sample
  Ratio:  feeder/filter
Average transmissometer
 reading
Notes
10/8/71
13
90
43
23.3
765
5-10|J,
1/16 x 1/8
100
3.31
OS OS
5.936 12.282
0.608 0.648
606
3AC
2.55 2.55
32.8
23.6 25.1
1.39 1.31
0.946
5
10/8/71
n
90
43
21.9
765
5-lOp,
1/16 x 1/8
100
3.27
6.029 17.813
0.587 0.626
600
3Ac
2 . 55 2 . 55
32.4
23.0 24.5
1.41 1.32
0.948
5
10/8/71
n
90
43
21.2
765
5-10,
1/16 x 1/8
100
3.27
5.558 12.506
0.597 0.829
600
3Ac
2.55 2.55
32.4
23.4 32.4
1.38 1.00
0.948
5
10/12/71
D
91
44
18.0
763
5-lOp,
1/16 x 1/8
1OO
3.23
OS OH
5.724 5.960
1.891 1.360
1800
3AC
2.55 2.55
32.0
24.7 17.8
1.30 1.80
0 . 94 9

10/12/71
n
91
44
20.9
763
5-10,
1/16 x 1/8
1OO
3.27
OS OI1U
5.908 5.914
2.692 1.557
1800
3 A
2.55 2.55
32.4
35.2 20.3
0.92 1.60
0.942

10/12/71
n
87
42
32.9
704
2.5-5y,
1/16 x 1/8
100
3.37
OS Oil
5.924 5.593
2.021 1.868
1800
3AC
2.55 2.55
28.7
26.4 24.4
1.09 1.17
0.860

10/12/71
D
88
42
32.7
764
2.D-5U,
1/16 x 1/8
100
3.43
OS OIIU
6.105 5.546
1.280 1.639
1800
3AC
2.55 2.55
29.2
16.7 21.4
1.75 1.36
0.850

10/18/71
n
90
44
18.8
764
2.5-Su,
1/16 x 1/8
]00
3.17
V
13710
240
1800
3AC
490
27.0
16.3
1.66
0.90
6,7
10/18/71
n
90
44
18.8
764
2.5-Su,
1/16 x 1/8
100
3 .23
V
13680
270
1800
3AC
613
27.5
14 .7
1.87
0.88
6

-------
                                                                            Table C-2  (Concludes)
o
Hun Number
97
Date 10/19/71
F.qu i pmcn t
Arrangement number li
Pressure d i f ferent in IK , mm
xylcno
Across nozzle
Across plenum (2-(>)
Atmospheric conditions
Temp, "t:
Barometer, nmillt;-
Dust type (fly ash)
Feeder
Ciroove s i /e
Var iae se 1 1 i n^
ItPM

91
•M

18.8
70 n
5-lOu,

I/Hi x 1/8
100
3.30
98
10/19/71
H

91
-11

21.0
7Gn
fj-lOn.

1/1G x 1/8
100
3 . 33
99
10/26/71
13

92
•1-1

1-1.0
703
fl-lOp,

1/8 x 1/2
20
0.227
                                                                                              101
                                                                              10/26/71     10/26/71    10/26/71
                                                                              7G3
                                                                              5-lOu,
                                                                                                7G3

                                                                                                5-lOn.
                                                                                          100
                                                                                          3 .08
Fi1ter
  Sampler  type
  Initial  r i 1 ler v.'oit;ht.,  mp,
  Dus t eo 1 lee ted,  mp;
  Time, see
  Sampl i nf;- pos i t ion
  Samp L i n^ rate, 1/m in

Aerosol eoncenlrat ion IIIK/ni'
  From teed  rate
  From f i 1 t or  san.p 1 e
  Hat io:   (Yodel- Ti I tor

                   imotor
                                                     13790
                                                    320
                                                     18(10
                                                    3-V
                                                     107
                                                                 18.1
                                                                             ().S7fi
1.  Filter  mount.ed over  side  of  tunm-l (not over  a  hole).
2 ,  Dummy  run in wh i rh filler was moved  into  t unne t  faring st ream with  no air f I ov. through  filter.   1
3,  Manomet er across filler 1)1 ew out when run  was  1 /2 to li/.'i over.
•1.  Major drift  exper i enced  in  t ransmi ssion met or  7ero .
r>.  In  the  "ll" runs ,  a spec i a 1  composi te  filter was  used :  1) Lop of  Nuclepore cont a i ni nsc  four  1 /H" 1
    Nuclepore with mat te s i de up ; 3} the  two  spot  Allied  with Duco cement  and separat ed at  cent er  will
G ,  Sain pi e  t a ken with vacuum  c 1 eaner with 1   rounded no/./ ! c po i nt ed  into  air st ream 1 ead i n£  d i reel 1 y
    ( appro x.  1 sq . f t. . area) .   Pressure d rop  across  inlet no// 1 e used to  me t e r air f 1 ow  rat e .
7.  Feed rate dropped erratically during  last  half of run,  apparently due to brid^in..-  in feeder hopp<
                                                                                                                         nles V, Mil
                                                                                                                          small pi i
                                                                                                                         to paper A
                                                                                                                             sh i ny side up
                                                                                                                             re  of Nurlepo
                                                                                                                             aeuiun el eane r

-------
                                                Table C-3





                SUMMARY OF CONCENTRATION RATIO  DATA  FOR SAMPLING AND FILTER ARRANGEMENT "l"
Concentration 1
Feeder Conditions At Locatic
Dust
Fraction
0-2.5 p


2.5-5 p,


0-10



Groove
Size
1/8 x 1/2


1/8 x 1/2


1/16 x 1/8

1/8 x 1/2


Speed
Control
Setting
20
40
100
20
40
100
20
40
100
20
40
100
2

1.56
1.56


2.57
2.43

6.45
3.35
3.94
4.38
4.26



3







2.18
2.09
2.96
2.65



3A
1.03
1.03
1.25
1.20
1.14
1.27
1.30
1.91
1.74
2.27
1.94
2.08
3.01
2.55
2.97
3.40
3.44
9.82
3.14
53.8
2.92
2.84
3.26
3.82
4.05
/ Feeder \
>n Number
5







3.43
2.91
3.05
2.57



6







5.00
5.53
3.08
3.71



Sampling Run
Number
8

1.00
1.22


1.44
1.44


2.35
2.15




41
42
50, 43, 46
51, 47, 49
44
45
48
37
38
33, 35, 31
34, 36, 32
39
40
52A, 21
22
23
24
1, 5, 11, 7, 3, 9
2, 6, 12, 8, 4, 14
10, 15, 16, 17
13, 20, 19, 18
25
26
27
28
29
30
See Figure C-4 for position  corresponding to each location number.
                                                 C-32

-------
                                                 Table C-l

                  SU-'SIARY OF CONCENTRATION RATIO DATA FOR SECOND SERIES OF SA.VPLI.NC RL'.NS
                         Condit ions Cormon to all Runs—Groove Size:  1, 16 x 1/8
                                                        Speed Control Setting:   100
uoncentra
Sampling
Dust Location For Sampler
Fraction Number OF CF OS
0-2.5- 2
c

3A
c

2.5-5^. 2
c
3A 1.09
c
1.75


5-10- 2
c

3A 1.89
c
1.40
1.28
1.53
1.39
1.31
1.30
0.92


0-10- 2 1.11
1.23
0.85
1.32
0.95
2 1.90
1.39
1. 15
2 1.71
2
1.80
2 1.70
c
3A 2.07
1.32
1.79
1.74
3A 2.62
3A 1.53
2
3A
C


7 3.77 3.95
8 13. 3t 2.26
tion Katio I iecacr,, i liter;
Tvpe and Arrangement NLirr.ber*
CS OKU CH,, OH.-J CHd
0.84 1.00

0.88 0.97
1.00 1.02

1.02 1.08
1.39 1.61

1.53 1.68 1.65

1.59 1.63 1.87
1.17
1.36
1.05 2.68

1.14 2.26
1.86 1.82 1.93 1.58 2.58 1.66

1.32 1.62 2.14 1.91 3.47 1.53
1.81 2.28
1.82 2.73
1.41
1.32
1.38
1.00
1.80
1.60











1.23 1.59





1.22 1.64


1 . 74 2.04
1.73 2.18
2.21 3.39
1.S2 1.76




Sa-.plini; P.Ljn Number
-51 7"*

77A, 77B
74A, 7 IB

7GA, 76B
79A, 79B

93A , 7SA , 73B . 95

91A, 80A, SOB, 96
933
94B
7"\ 7"S

73A, 73B
85A, 85B, 70A, 70B , 83A, 83B, 97

86A, SJ6E, 71A, 71B, 84A, 84B. 98
87A, 81A. SIB
87B, 82A, 82E
83A. S9A
883, S9B
D1A, 90A
92A, 903
91B
92B
53A
51A
57A
53A
59A
52B
533
5 IB
573

583
59B, 69A, 69B

55A
56A
61A
62A
553, 61B, 62B
56B

65A.65B
66A, 66B
67A, 67B
68A, 68B
603, 63A
60A, 633
*  Symbols used as follows:
     Sampler Orientation:    F  =  flush with wall
                            S  -  facing into nominal air st rean (medium vert i cal}
                            H, =  filter medium horizontal facing down
                            H^ =  filter medium, horizontal facing up.

     Type sampler:          0  -  open ''Delrin"
                            C  =  closed "Delrin"
                            V  -  vacuum cleaner bags (vert ical int ake}.

t  Que s t i o nab1e run.
                                                    C-33

-------
                                                                                               T«ble C-5


                                                                               SUMMARY OF WEATHER MEASURE SAMPLING DATA
O
 I
co
Date
Equipment Arrangement Number
Dust Fraction (Fly Ash)
Feeder, Groove Size
Speed Control Settirg
Disperser Air Pressure, psig
Sampling Rate, liters/min
Nominal Dust Concentration, mg/m3
Weather Measure Headings, counts/min
At Sampling location





























Equipment

Sec t ton
Supply Line





PlenuQ






Sighting Tunnel












Number

9
10
7

8

12
13
6
5
5A
16
4
11
1
14
2

15
3B
3A

3AA

3
Atmospheric Air
1/28/71
A
0-!0u
1/16 X 1/8
100 100
50 50
40** 2.55
32 32







42

41,42 57,47,48,50



37,38 40,40
43,41


42 49
44


21,25,29







48
3,2
2/3/71
II
0-10H
1/16 X 1/8
100
50 10 20 30 50
2.55 2.55 2.55 2.55 2.55
32 32 32 32 32





0

35

46,55,43 14 49 49 47,51

45
40
42
48 48
46
44,48

36
25,27
20
28 29

28
35,28,26
30,27,21 9 34,21,40

37,39,32
42,29,39
46,38,37

2/11/71
E
0-10n
J/8 X 1/2
20 40 100
50 50 50
2.55 2.55 2.55
24 84 242









34,32 97(106) 322(336)
102(96) 294(276)














31,29 92 272(286)
90(86) 294(256)




2/18/71
B
2.5-5|i
1/8 X 1/2
20 100
50 50
2.55 2.55
20 222







32 376(376)
342(326)
52,47 474(420)
502(470)



51 508(516)



52


31,34 322(246)
390(316)


44,45 448(446)





* Except where otherwise noted, Weather Measure Instrument exhaust was connected to the same metcred vacuum system as was used for the filter sampling runs.
** In this case instrument was exhausted to the atmosphere with the plenum pressure acting to cause flow through the instrument; the flow rate given is estimated.

-------
            Appendix D






AEROSOL CHAMBER OPTICAL COMPONENTS

-------
                              Appendix D

                  AEROSOL CHAMBER OPTICAL COMPONENTS


     The fly ash aerosol generated with the especially designed and as-

sembled equipment described in Appendix A was continuously monitored by a

white-light source transmissometer and, during periods of low background

conditions, by 9CT light scatter from a small portion of the transmissome-

ter light beam (Figure D-l).  The transmissometer was designed to minimize

the forward-scattering problem associated with light extinction measure-

ments (Zuev, 1966).  Light transmission T is defined in terms of Bougure's

(Beer's) Law as:
                             T   =  —   =
                                   o
where

          I  = light intensity in absence of an attenuating medium
           o
          I  = measured light intensity

          L  = path length along the interacting medium

          o  = volume extinction coefficient.

However, the light intensity I may contain some light that has been scat-

tered by the attenuating medium,  resulting in an apparent extinction co-

efficient (G" ) smaller than the true coefficient (cr).  The ratio of these
            a
*  ZUEV, V. E., 1966:  Atmospheric transparency in the visible and infrared,
   Translated from Russian, available from the Clearinghouse for Federal
   Scientific and Technical Information (TT69-55102).
                                  D-3

-------
 coefficients  depends  on  the  transmissometer geometry  (primarily path length


 and receiver  aperture) and the angular scattering function of the aerosol.


 Since  larger  particles (with respect to the light wavelength) are more ef-


 ficient  forward  scatters, 
-------
Event marks on the opposite side of the chart indicate both lidar firings



and pulses signifying one revolution of the particle feeder disk.  For




each of these event marks, a value of transmission (voltage) was printed




on the paper tape record.





     The 90  side-scatter data were recorded in relative logarithmic units




on a second channel of the strip chart recorder.   However,  because of high




background light levels during daytime,  these data were of only limited



use.  Since the transmissometer views a much larger aerosol volume than




the side-scatter photometer,  the relative variations between these outputs




are indicators of aerosol homogenity.  Larger variations were normally



observed in the transmissometer output indicating the aerosol was well



dispersed within the chamber.
                                  D-5

-------
o
O'l
                       LIGHT TRAPS
               TUNGSTEN
              'FILAMENT
               LAMP
IRIS
         COLLIMATED LIGHT SOURCE

                                                                  FIELD OF VIEW
                                                                   Adjustable from
                                                                   6 to  20 mrad

PROVISION
FOR FILTERS ~~--fc-
FIELD STOP^"
PHOTOMULTIPLIER--

-------
110
          POWER
          SUPPLY
                                         90  LIGHT
                                         SCATTERING
                                         RECEIVER
                                                          TRANSMISSOMETER
                                                              RECEIVER
                                                      AEROSOL TUNNEL
                            HIGH-
                          VOLTAGE
                        POWER SUPPLY
          LIDAR  FIRE PULSE
                                                                                           OPERATIONAL
                                                                                            AMPLIFIER
                           DIGITAL
                         VOLTMETER
                                                                                 PRINT PULSE
                          PRINTER
                                                            EVENT
                                                            MARK
  STRIP
  CHART
RECORDER
 DIRECT
 DISPLAY OF
 TRANSMISSION
PRINTOUT OF
TRANSMISSION
GOAL = 1% ACCURACY
                                                                                                                   TA-8730-5
                       FIGURE D-2   DATA FLOW  DIAGRAM  OF THE OPTICAL SCATTERING  SENSORS

-------
    STRIP CHART OUTPUT
                                             LIDAR FIRING
                                             AND DISK
                                             ROTATION PULSES
     0.0--
           TRANSMISSION
           PRINTOUT
                                               1 mm/s
                           r-  O  CM i-
                           m  ir)  in cn
                           o  6  ci d
                                                      TA-873O-6
FIGURE D-3  DATA FORMAT OF THE OPTICAL SCATTERING  SENSORS
                              D-8

-------
     Appendix E





TRANSMISSOMETER DATA

-------
                                Appendix E



                         TRAXSMISSOMETER DATA



     Transmissometer readings were taken at all times when the sighting


tunnel was in operation.  This appendix presents and analyzes all data


showing the relationship between the optical transmission measurements


made in the tunnel and the aerosol concentration for the various aerosols


Details of the transmissometer construction and operation are given in


Appendix D while details of simultaneous concentration and transmission


measurements are given in Appendix C.





Background Theory



     Basic Optical Transmission Equations




     Optical performance is usually measured in terms of a scattering


coefficient defined by the Beer's relationship:




                       dl  =  -GldL    .                            (E-l}~




The gross geometry of an aerosol system can be allowed for by introducing


an efficiency factor, Q , defined as
                       e
                              n =n

                        a  =   PS P Q (iD"/4)an      .               (E-2)
                                     e   p     p
                               n=o
*  A table of nomenclature is given at the end of this appendix.
                                   E-3

-------
The  factor, Q  , is a function of the ratio of particle size to wavelength


of light used (D A) and the refractive index of the particle (m).  For
                P
spherical particles this functional relationship is given by the well-


known Mie theory.


     Thus, although Q  is a function of particle size, for many prac-
                     e

tical purposes  it is convenient to write Equation E-2 in terms of an


average efficiency factor
                                 n=n
a  =  Q
                              av  n=o
 S  (TTD /4)An
       P     P
                                                                   (E-3)
where, by definition,
                           av
S Q D 'An
   e p  p
    2
 Z D An
    P  P
                                                                   (E-4)
The advantage of so defining Q0   is that, as will be shown later, <3
                              eav                                   eav

will be dependent primarily on some mean particle diameter when dealing


with a mixture of particle sizes.
     Mean Particle Diameters


     A mixture of particle sizes can be represented by any one of an


infinite number of mean sizes.  Mean size can generally be defined in


the following format:
                      D
                                 DqAn  '
                                  P  P
                              L   P   P-
                                 [E-5)
     Certain specific mean sizes will be of special significance in ex-


pressing transmission performance.  The "surface mean  diameter (sometimes


                                   E-4

-------
called the  surface-to-number  mean) corresponds to values of 2 and 0

for q and j, respectively,
                         D
                          20
      L, D
                                     2
                                      an
                                     P  P
               1/2
                                       ;E-G)
The  Sauter  diameter (sometimes called the "volume-to-surface" mean"

corresponds to values of 3 and 2 for q and j,  respectively,
                                    £ D /in
                                       P  P
                            '32        2
                                    £ D An
                                       P  P
D
 E-7,
But, since, by definition,
                         w   =  Z(rrp /6)D An
                          P         P    P  P
                                       (E-S'
Equation E-7 may also be written as
                           D
       (6w /up )
          P   P
32         2,
        £ D an
           P  P
(E-9)
if all particles have the same density.   If the particles'density varies,

p  may be defined as an "average" particle density.
 P
     Applied Optical Equations


     By combining Equations E-l and E-3 with Equation E-6 and E-9, re-

spectively, optical performance can be written in two alternate formats:
                                  E-5

-------
           a)   In terms  of  particle number  concentration:
                        dl   =  -(n/4)q    n D"  IdL                    (E-lo)
                                     e   p  20
                                      av
 or
                J2n(l/x)   =  j£n(l /I)   =   (rr/4)Q   n D  L
                               o              e   p 20
                                               av
               In  terms  of particle mass concentration:
                    dl   =  -(3/2)Q    (w /p D n)ldL                 (E-12)
                                  e    p  p 32
                                   av
or
                     =  Jin(l /I)  =   (3/2)(Q   w /p D   )L     .      (E-13;
                            o               e   p  p 32
                                             av
     It will be observed that when concentrations are measured in terms


of particle number, the optical performance is dependent on the square


of the mean diameter D  ; when measured in terms of mass, it is dependent
                      &(J

only on the first power of the mean diameter D   .  In this study mass


concentration of aerosol was determined as well as the Sauter diameter,


D  , for each fraction.  Consequently, these values in conjunction with
 O*j

the measured values of light transmission can be used to evaluate Q
                                                                   e
                                                                    av



Experimental Results



     The runs in which both aerosol concentration and transmission values


were established are discussed in Appendix C.  These results are summarized


in Table E-l.  Average values of aerosol concentration and optical trans-


mission are given in Table E-l for all runs made at nominally the same


conditions and with the same aerosol.

                                  E-6

-------
     As discussed in  Appendix C,  aerosol  concentrations  calculated  from

the measured feeder disk speed are believed  to be the most  reliable  mea-

sure of aerosol concentration.  Consequently,  for this comparison  with

optical transmission, only the aerosol  concentration  values established

from feeder disk speeds will be used.


     The data of Table E-l are shown plotted in alternate forms  in

Figures E-l and E-2.  In Figure E-l the transmission  (on  a  log scale)

is plotted against the mean dust concentration (on an arithmetic scale)

for each dust fraction.  As expected from  Equation E-13,  straight  lines

are obtained for each dust, with the finest  dust being optically the

most active.  The upper scale expresses concentration in  grains/cu ft

to permit each comparison with units used  in practice.


     In order to permit a more quantitative  evaluation of the effect of

particle size, the data are plotted in  Figure E-2 as  transmission (on  a

log scale) against the grouping 3Lw /2p D    (termed effective opacity)
                                   p   p 32
on an arithmetic scale.  The slope of such a plot is  the  average effi-

ciency factor.  Lines for values of Q    of  1, 2, and 3 are shown as
                                      av
dashed.  The data can all be correlated within the actual precision by

the single solid line.  This line corresponds to an average efficiency

factor of approximately 2.4.  Although  there appears  to be a trend in

the data for the various fractions toward  higher Q    values with reduced
                                                   av
particle size (as expected from theory  for this particle  size range),

these trends are believed to be outside of the overall intrinsic precision.


     The average efficiency factor calculated for each data point is

shown in the last column of Table E-l.   Because of errors  in the indivi-

dual values that go into the calculation of  Qe   , these values for the
                                              av
individual points will tend to exaggerate discrepancies.    In general,

however, these values agree well with the value of 2.4 deduced from

Figure E-2.
                                  E-7

-------
     Values of Qe   have been calculated for each of the dust fractions.
                 av

To do this, the size distributions given in Appendix B, Table B-4 were


used in conjunction with Mie theory to define Q  and the operation in-
                                               e

dicated by Equation E-4 was made on a CDC-6400 computer.  Such calcula-


tions were made for two assumed wavelengths of light, 0.7 and 1.06^.


The results of the calculations, shown in Table E-2, show excellent agree-


ment with the experimental results for the transmissometer.
                                 E-8

-------
                             NOMENCLATURE
  D   =  particle diameter,  cm
   P


 D    -  any mean particle diameter,  cm
  qj


 D    =  surface-mean particle diameter,  cm



 D    =  Sauter diameter,  cm
  o*^


   I  =  incident light intensity, lumens/sq. cm



  I   =  initial incident light intensity, lumens/sq. cm
   o


   L  =  optical path length, cm



   m  =  refractive index, dimensionless


                                                 3
  n   =  particle number concentration, number/cm
   P


 q,j  —  dimensionless integers



  Q   =  cross section efficiency factor, dimensionless
   e


Q     -  average efficiency factor, dimensionless
 e
  av


   T  =  fractional light transmission = I/I
                                            o

                                            /  *•?
  w   =  particle mass concentration  grams/cm
   P

                                        3
  p   =  true particle density, grams/cm
   P

                                   -1
   0"  =  scattering coefficient, cm



  0"   =  standard geometric deviation, dimensionless
  ' g


   X  =  wavelength of light, cm
                                   E-9

-------
w
I
                                  0.01
0.02
 T
FLY ASH CONCENTRATION — mg/m  (from feeder rate)
 0.03     0.04     0.05     0.06     0.07     0.08
                                                                                                          0.1
                                    A = 0-2.5 p.
                                    m = 2.5-5M
                                    V = 5-10/V
                                    • = 0-10U
                                  20
                                         40
             80     100    120     140
              FLY ASH CONCENTRATION
                                 160
                                  grams/ft
180
3
                                                                                                200
                                                             220
240    260

 TA-8730-36
                         FIGURE E-1    EFFECT OF  FLY  ASH  CONCENTRATION AND  SIZE ON OPTICAL TRANSMISSION

-------
M
I
                                  A = 0-2.5 /i
                                    - 2.5-5/LI
                                  T = 5-
                                  • = 0-10M
                                          0.1
                                              EFFECTIVE OPACITY = 3L wp/2 Pp D32 — dimensionloss
                                                                                                            TA-8730-37
                                 FIGURE E-2   GENERALIZED CORRELATION FOR OPTICAL TRANSMISSION

-------
                                                                              Table E-l

                                                                   SUAIMAHV OF TKA.VSMISSOMETER DATA
Fly Ash
Fraction
Used1'

0-2.5 |i.



2.5-5 |L



5-10 (i



0-10 n




Mean
Diameter'
532
Microns
1.7



3.1



6.5



•1.0




Feeder Conditions
Groove Speed
Size Control
Sotting

1/16 x 1/8
1/8 x 1/2


1/1G x 1/8
1/8 x 1/2


1/16 x 1/8
1/8 x 1/2


1/16 x 1/8

1/8 x 1/2


100
20
40
100
100
20
40
100
100
20
40
100
20
100
20
40
100

Feeder
Disk Speed
RPM
3.22
0.259
l.OO
3.00
3.27
0.288
1.06
3.22
3.27
0,227
1.02
3.08
0.292
3.33
0.328
1.15
3.35
AvcraRC Results*
Dust-
Concentration
3
mg/m
16.7
12.0
47.0
138.9
27.8
20.3
74.1
223.
32.3
18.1
74.5
206.
2.91
33.2
24.5
84.0
243.
Optical
Transmission
%
86.0
89.2
65.5
30.8
86.4
91.8
73.5
40.2
94.6
96.9
87.3
66.6
98.5
90.8
92.1
76.7
47.7
Data from
Run Numbers

74-77
41-42
43, 46, 47, 49-51
44, 45, 48
78-80, 93-96
37, 38
31-36
39, 40
7O-73, 81-92, 97, 98
99
100, 102
101
21, 22
9-20, 52-69
25, 26
27, 28
29, 30
Effective
Opacity
(3Lw /2p T5 )
p p 32
Dimcnsioaless
0.0550
0.0395
0.1515
0.456
0.0501
0.0366
0.1336
0.401
0.0278
0.0156
0.0641
0.1773
0.0040
0.0463
0.0342
0.1172
0.339
Average
El f iciency
Factor *
Q
e
av
Dimcnsionlcss
2.73
2.89
2.73
2.58
2.91
2.35
2.30
2.27
1.99
2.08
2.11
2.29
3.75
2.08
2.39
2.26
2.18
M
1
          *  Details  of  runs  given in Appendix C.
          t  Details  of  fly ash  properties given in Appendix IS.
          *  Calculated_from  measured feeder disk speed and feeder disk dust rate calibration data.
          5  Qn    = 2p D  in (l/T)/3Lw ,  using p  = 2.-16 g/cm ,  1, -= 30 ft.
               av     P  32              P         I1

-------
                              Table E-2
  COMPARISON OF MEASURED AND CALCULATED AVERAGE  EFFICIENCY FACTORS

Average Efficiency Factor

Fly Ash
Fraction




0-2.51^
2.5-51-1
5-lOlJ.
0- 10M.

Measured
Q
av
( Transmissometer )

2.73
2.46
2.12
2.53
(2.23)T
Calculated^"
Q
e

A. = 0.711

2.76
2.37
2.24
2.54


A. = 1.061-L

3.22
2.57
2.31
2.63

Particle Size
Distribution
Parameters
Geometric
Standard
Deviation
r~
g
1.5
1.5
1.5
2 2

Sauter
Diameter
D
32
Microns
1.7
3.1
6.5
4.0

*  Average value from last column of Table E-l.
t  Calculated from Q  =
                    6
 Q D2An
   - -
Z D An
   P  P
                                  for an assumed log-probability
   distribution with parameters taken from the last two columns of
   Table B-4, as summarized in the last two columns above.

   Omitting one value in Table E-l that was abnormally high and of
   questionable precision.
                                E-13

-------
     Appendix F





LIBAR INSTRUMENTATION

-------
                               Appendix F





                         LIDAR INSTRUMENTATION








     The SRI Mark V lidar system (Figure F-l)  was used to collect back-



scatter data at a wavelength of 1.06P- (neodymium laser),  and the SRI/EPA




Mark VIII lidar was used to collect backscatter data at a wavelength of




0.69M. (ruby laser).  Table F-l gives the lidar characteristics and




Figure F-2 is a diagram of the transmitter/receiver optics and of the




data recording method used with the Mark V lidar.  The transmitter and




receiver optics are coaxial so that complete beam convergence occurs at




a short range from the lidar (approximately 50 m).   A small portion of



the beam energy is reflected by means of a glass plate beam splitter onto



a diffusing surface that is viewed by a light  pipe with a wide acceptance




angle.  The output from the light pipe is reflected into one of the photo-




multiplier tubes, and the resulting pulse amplitude is used to normalize




the pulse amplitude returns from both the chamber aerosol and the passive




reflector.  Since the peak return from the aerosol is proportional to




transmitted energy and the return from the reflector is proportional to




transmitted peak power,  it is assumed that the ratio between these quan-




tities remains constant from pulse to pulse, i.e.,  the transmitter pulse




shape does not vary.   Because of the high bandpass of the lidar signals,




a photographic technique is employed to record the data records.  Linear




signal processing and displaying was used because of normal uncertainties




of the transfer characteristics of the video logarithmic amplifier.




However, because the lidar return signals may vary over several orders




of magnitude, a multistage display scheme was used.  The output from




one detector is input to three separate oscilloscope displays, each set
                                  F-3

-------
at a different voltage gain.   The output from the other detector was



logarithmically amplified and displayed on an oscilloscope to provide



a single trace monitor of the energy return along the beam path.  Data



recorded with the Mark VIII lidar, which has only one detector, was



recorded using a single higher bandpass Tektronix 556,  dual-beam oscil-



loscope (30 MHz) with two linear displays.
                                 F-4

-------
                                                   TA-728&-7
FIGURE F-1    SRI MARK V LIDAR
               F-5

-------
D1
02
D3
D4
                                                                      FILTERS
                                                                     •»—-^
                                                             NARROW BAND
                                  LOG
                               AMPLIFIER
                                         LIGHT PIPE
                                                                             NEUTRAL
                                                                             DENSITY
 TEKTRONIX 555 DUAL-BEAM
      OSCILLOSCOPES
                                                                                            TA-8730-24


  FIGURE F-2   DIAGRAM OF MARK V LIDAR SYSTEM AND DATA RECORDING FOR  BACKSCATTER EXPERIMENTS

-------
                                       Table F-l

                CHARACTERISTICS OF  SRI/EPA MARK VIII AND SRI MARK V LIDAR
                                                            Charac ter1st ics
SRI/EPA Mark VIII  Lidar

  Transmitter

    Laser Rod
    Wavelength (A)
    Beamwidth (mrad)

    Optics

    Pulse Energy (joules)
    Pulse Length (nsec)
    Q-Switch
    Cavity Cooling

  Receiver

    Optics
    Field of View (mrad)
    Predetection Filter  Passband Width (A)
    Neutral Density Filter
    Detector
    Post Detection Filter Bandwidth
Ruby
6943.0
Measured as 2 nrad and reduced to 1 mrad
 during this study
2-inch refractor coaxial with receiver
 telescope
0.9
Measured as 50 ns during this study
Pockels cell
Refrigerated water
Newtonian reflector (6-in.,
2 and 3 mrad used during this study
10
16 dB
RCA-7265 photomultiplier (S-20 cathode)
30 MHz
SRI Mark V Lidar

  Transmitter

    Laser Rod
    Wavelength (A;
    Spectral Line Width (A)
    Beamwidth (mrad)
    Optics

    Pulse Energy (joules)
    Pulse Length (nsec)
    Q-Switched

  Receiver

    Optics
    Field of View (mrad)
    Predetection Filter Passband Width (-•
    Neutral Density Filter
    Detector
    Post Detection Filter Bandwidth
Neodynium glass
10,600
90
0.4
2-inch refractor, coaxial with receiver
 telescope
1.0
Measured as 23 nsec during this study
Rotating prism
Newtonian reflector  (6  in.;
1.0
100
30 dB
RCA  7102 photor.ultiplier  ', S-l  cathode)
30 MHz
                                          F-7

-------
       Appendix G





LIBAR SIGNATURE ANALYSIS

-------
                              Appendix G
                       LIDAR SIGNATURE ANALYSIS
     Figure G-l is an example of the lidar signatures recorded for various


fly ash concentrations (Mark VIII lidar, 0-2.5 pm diameter fly ash).  The


amplitude of the pulse at zero range is a measure of the peak power


emitted by the laser.  As the aerosol concentration is increased, the


return from the passive reflector decreases because of the round-trip


attenuation of the laser pulse by the aerosol.  The amplitude of the


aerosol return signal (normalized by the transmitted energy) increases


with increases of aerosol concentration.  The primary data records con-


sist of three pulses  P , P , and P , that are proportional to the peak
                       t   a       r

transmitted power (P')> the peak aerosol return (P')> and the peak passive
                    t                             a

reflector return (P').  The relationship among these quantities  and the


derivation of their information content are best approached by use of


lidar equations for solid and volume targets that may be expressed  in


the form:
                  Solid target:  P'
                                  r
                              -2 r  22
                          KP'R   - T T
                            t r  'i  c a
G-l
                 Volume target:  P'
                                  a
                              -2  T  2
                          KP'R  B - T
                            t a   2  c
G-2
where
           K  =  lidar constant
          P'
           t
=  peak power transmitted
                                  G-3

-------
           R   =   range  to the target



           r   =   target reflectivity



           |3   =   volume backscatter coefficient



           T   =   pulse  length



          T    =   transmission along the path of clear air
           c


          T    =   transmission along the path of the test (chamber) aerosol.
           cl


The above expressions assume single scattering, no geometrical effects


over  the beam  path  (coaxial lidar), negligible attenuation over one


laser pulse length, uniformly distributed scatters within the pulse, and


a Lambert (cosine)  reflector.  For a given lidar, the quantities K, R  ,


R , r, and T are  assumed to remain constant.  The constant T implies that
 a

the ratio of transmitted energy to peak power (or pulse shape) remains


constant from  pulse to pulse.



     Relative  backscatter of the chamber aerosol is defined as a. quantity


normalized by  transmitted energy, not corrected, however, for variations


of the clear air  transmission, i.e.,
                                                 P
                                                  a
                 6   =  relative backscatter  =  —    .               G-3
                  R                              P
This quantity is dependent on the lidar optical and electrical efficiencies


and hence its numerical values would vary from lidar to lidar.  An abso-


lute value of backscatter, the volume backscatter coefficient, requires


calibration of the lidar, which may be accomplished by use of a solid


target (passive reflector) of known reflectivity.  The ratio of returns


from a solid target to a volume target leads to the expression:
                                  G-4

-------
                              P  R
                               a  a  r  2

                              n	9  ~ T                              G"4
                              P   2  n  a
                               r R
                                  r
The backscatter experiments required a black passive reflector so that


received signals were not sufficiently large to fall in a nonlinear


operating region of the detector (detector saturation).  The target re-


flectivity (0.07) was determined from the ratio of reflected target re-


turns from the black target and a standard white target (0.98 reflectivity)


with a neutral density filter placed in the receiver to prevent detector


saturation.  Pulse lengths were measured before the backscatter experi-


ments using a photodiode and a high bandpass oscilloscope (Tektronix 556).


Pulse shapes were not monitored during the experiments.



     Transmission of the laser energy pulse through the aerosol chamber


T , may be derived from the transmissometer data or inferred from the
 a

lidar data.  Using the data recorded from a lidar observation made with


a clear air aerosol chamber c, the transmission for a test aerosol may


be evaluated from the expression:
                           a     P     P
                                  re    t
                                                                      G-5
where the transmission through a "clear" chamber, T   , was assumed to be
                                                   ac

unity.  Using the lidar derived value of T , the expression for 0 becomes:
                                          cl
                                       2
                              P   P   R
                               a   tc  a  r 2
                        R  =  —  	 —  — —                          G-6
                        P     P   P    2  TT T
                               t   rc R
                                       r
                                      P
                                       a
                           =  const  • —                               G-7




                                  G-5

-------
and  3  is  independent of the target return for lidar firings with a test


aerosol within  the chamber.  This is important for the backscatter experi-


ments  since high aerosol concentrations may result in weak target returns


that cannot accurately be read from the primary data records.  A relative


backscatter quantity corrected for variations of the clear air transmis-


sion over the 500-ft path, T , can be derived (Eq. G-6) from the expression
                            c
                            P   P           P
                             a   tc          tc
                     B '  =  —  	  =  8   	    .                   G-8
                     PR     P   P       PR  P
                             t   re          re
     The value of T  in Eq.  (G-4) can be replaced by the transmission
                   a

determined with the white light transmissometer.  This transmission mea-


surement is probably a better indicator of the true laser energy trans-


mission (because of less experimental error); however, Eq.  (G-4) then


requires a reading of target return energy, P , at times of high aerosol


concentration.  Hence, enhanced experimental errors could be expected at


large aerosol concentrations.



     The quantities P , P , and P  were read from the Polaroid prints
                     t   a.       r

as voltages above the receiver noise level that, because of the black


target, was electronic noise limited rather than atmospheric background


limited.  The data were placed on computer input cards and  subjected to


the above analysis and then plotted in various ways by a computer micro-


film plotting system.  The data are tabulated in Appendix H and shown as


data plots in Section III C of the main text.
                                  G-6

-------
        UJ
        oc
        Q
        _l
                                 LIDAR
                                                      REFLECTOR
                  T = 0.99
                  CLEAR AIR
                  T = 0.88
                  M = 11  mg/m'
                  T = 0.60
                  M = 18 mg/m3
                       AEROSOL RETURN
                                                         REFLECTOR
                                                         RETURN
                  T  =  0.32
                  M  =  110 mg/m
                     MEASURE OF
                    TRANSMITTED
                         ENERGY
                                       MEASURE OF
                                    RETURN ENERGY  J
                                     FROM AEROSOL
                                            MEASURE
                                            OF RETURN
                                            ENERGY
                                            FROM
                                            REFLECTOR
                                      RANGE
                                                               TA-8730-35
FIGURE  G-1
EXAMPLES OF  LIDAR SIGNATURES (REPRODUCED FROM POLAROID PRINTS)
FOR  VARIOUS  MASS CONCENTRATIONS OF THE 0- TO 2.5-fi DIAMETER FLY
ASH  FRACTION  (0.6943 ju WAVELENGTH LIDAR)
                                      G-7

-------
    Appendix H





LIDAR DATA SUMMARY

-------
                              Appendix H





                          LIDAR DATA SUMMARY








     The lidar data collected and analyzed under the present study are



summarized in Table H-l in digital form.  Absolute power levels were not




measured because the photomultiplier gain and/or receiver attenuation




for various experimental phases had to be varied.  System studies applied




to other lidar techniques are best approached using the absolute scatter-




ing and extinction coefficients presented in Section III-C of the main




text.  The relative returns (voltages) shown in Table H-l may be applied




to any lidar by calibrating the lidar in terms of received signal from




a black diffuse passive reflector.
                               Table H-l




                          LIDAR DATA SUMMARY
Run No.
1
2
3
4
6
7
8
9
10
11
12
13
14
Lidar
(wavelength )
Mk V (1.06 u.)
Mk V (1.06 u)
Mk V (1.06 p.)
Mk V (1.06 p.)
Mk VIII (0.69 u.)
Mk VIII (0.69 u.)
Mk VIII (0.69 p,)
Mk VIII (0.69 p,)
Mk VIII (0.69 |o,)
Mk VIII (0.69 (o,)
Mk VIII (0.69 |i)
Mk VIII (0.69 jo,)
Mk VIII (0.69 u.)
Fly Ash
Fraction
0-10
0-10
0-2.5
2.5-5
2.5-5
0-2.5
0-2.5
0-10
0-10
Feeder
Groove
Size
1/16 X 1/8
1/8 x 1/2
1/8 x 1/2
1/8 X 1/2
1/8 x 1/2
1/8 X 1/2
1/8 x 1/2
1/8 x 1/2
1/8 X 1/2
Screen Experiment
5-10
0-2.5
0-10
1/8 X 1/2
1/16 X 1/8
1/16 X 1/8
                                  H-3

-------
     Table H-2 contains printouts of computer runs.  A glossary of terms

used in the column headings precedes the table.
                    Glossary for Terms in Table H-2



         D  =  particle feeder speed control setting



        P   =  relative transmitted peak power (mV)



        P   =  relative aerosol peak power return (raV)
         3,


        P   =  relative passive reflector (black) return (mV)
         r


        P   =  relative passive reflector (white) return (mV)



        P   =  relative screen return (mV)
         s


         T  =  transmission (white light)



         a  =  volume  extinction coefficient (km  )


                                          .     3,
         M  =  aerosol mass concentration (mg/m )
                                  H-4

-------
LTDAB FIELD DATA, si
D

0
n
n
0
P
0
n
0
AVG
20
20
20
20
20
AVG
40
40
*n
40
40
40
40
40
40
40
40
40
AVG
100
ion
100
100
AVG
P-.
t
1500
1550
1450
1310
1030
1300
1650
1450
1414
1550
1400
1500
1580
1650
1536
1520
1700
1S50
1810
1450
1650
1350
1380
1550
1520
1?PO
1290
1512
1530
1480
1530
1710
1567
P.
a
-0
-S
• n
-1
-n
-ft
• ft
-0
^
18
la
2?
22
24
21
75
9i
an
80
12i
135
in
105
ITS
170
Un
160
119
22n
23ft
240
27ft
24ft
p
r
390
350
300
320
250
300
420
300
327
320
300
350
310
400
336
300
400
360
360
340
390
295
300
340
360
260
280
332
300
300
300
350
313
p /P
a t
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
.012
.013
.015
.014
.015
.014
.046
.053
.048
.044
.083
.082
.076
.076
.110
.112
.109
.124
.080
.144
.155
.155
.158
.153
J* NO.
P /P
r t
.253
.226
.207
.246
.243
.231
.255
.207
.233
.206
.214
.233
.196
.242
.219
.197
.?15
.218
.199
.214
.236
.219
.217
.219
.237
.213
.217
.219
.196
.203
.194
.205
.199
1
T

,997
.997
,997
.997
.997
.997
.997
,997
.997
.992
.990
,988
,9P6
.986
,9fl8
.967
.966
.967
,965
.948
.945
.945
.947
,929
.933
.912
.935
.948
.9(19
.699
.901
.899
.902



.33
.33
.33
.33
.33
.33
.33
.33
.33
.93
1.10
1.32
1.54
1.54
1 .27
3.66
3.77
3.66
3.89
5.82
6.17
6.17
5.94
8.03
7,56
7,68
7.33
5. PI
10.40
11.61
11.37
11.61
11.25

.,

•11
.11
.11
.11
.11
.1;
• 11
.11
.11
.29
.37
.44
.52
.52
.43
1.23
1.27
1.23
1.31
1.96
2.07
2.07
2.00
2.70
2.54
2.58
2.46
1.95
3. 5Q
3.90
3.R2
3.90
3.78
LTD'R FIELD D*TAi BUM MO.
D

n
rt
0
0
n
0
0
AVG
20
20
20
2n
20
20
20
20
20
20
AVO
P
t
800
880
12*0
800
1250
1150
1250
1047
1050
uno
sno
1150
1280
1380
1320
500
910
1000
1078
P
a
-n
-n
•5
• n
*9
-0
• ^
5
25i
350
lei
3Po
38j
363
340
lOfl
24g
240
274
p
r
540
750
900
660
1100
900
1100
850
700
1140
650
890
1000
1200
1120
340
660
800
850
P /P
a t
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0,000
.238
,250
.225
.261
.297
.261
.258
.200
.267
.240
.250
p /p
r t
.675
.852
.750
,825
.880
.783
.880
.806
.667
.814
.813
.774
.781
.870
.848
.680
.733
.800
.778
T

.995
.995
.995
.995
.995
.955
.995
.995
,9?8
.922
.'523
.923
.924
,928
.920
.924
.912
.916
.922
-

.55
.55
.55
.55
.55
.55
.55
.55
8.15
fl,86
8.74
8.74
8.62
8.15
9.09
8.62
10.05
9.57
8.86
M

,1 8
.18
.18
.18
.1 8
.13
.18
.18
2.74
2.98
2.94
2.94
2.95
2.74
3.06
2.90
3.38
3.21
2.98
               H-5

-------
           Table  H-2  Continued '
LIDAR FIELD DATA,  RUN'  NO.  2  (CONTINUED)
D

0
0
0
0
AVG
40
40
40
40
40
AVG
100
100
100
100
100
AVG
0
0
AVG
100
100
100
100
100
AVG

D

0
0
1
AV3
20
2(1
20
2(1
20
20
20
20
20
2"
2o
*VG
r>
n
0
0
AVG
100
100
100
AVQ
p
t
1200
1400
2000
1200
1450
800
850
1450
1250
1300
1130
1000
1000
880
700
8SO
886
300
1100
700
600
1350
1610
1350
1640
1310

P
t
1620
1400
1450
1490
1600
1250
1310
1500
1580
1380
1240
1550
1400
1480
1800
1462
1200
1150
1060
900
1H77
1500
1490
1800
1597
P
a
•5
-0
•2

0
5po
510
85?
75i

686
119n
loon
880
80S
85S
944
-J)
• n
S
6on
1150
155*
1305
1500
1220
LTCl'
p
a
— 1
•I
-n
P
600
400
535
46o
Son
son
395
40n
4on
4o5
535
46?
• n
-n
-n
-A
"
16fln
1635
16oo
161S
p
r
950
1000
1620
810
1095
310
400
610
610
600
506
200
150
140
120
100
142
150
650
400
100
160
260
200
300
294
»B FIELD
P
r
1480
1300
1300
1360
1200
910
860
1100
1150
910
800
1250
900
1100
1400
1053
1050
920
930
740
910
200
250
300
250
p /P
a t
0.000
0.000
0.000
0,000
0.000
.625
.600
.566
.600
.631
,6H8
1.190
i.ono
1.000
1.143
1.000
1.067
0.000
0,000
0.000
I. 000
.852
.963
,963
.915
.938
DATA,
P /P
a t
0.000
o.ooo
O'OOO
0.000
.375
.3?n
.408
.307
.316
.362
.319
,258
.286
.270
.294
.320
0.000
0.000
0.000
0.000
o.ooo
1.067
1.094
,809
1.017
P /P
r t
.792
.714
.810
.675
.748
.387
.471
.421
.488
.462
.446
.200
.150
.159
.171
.118
.160
.500
.591
.545
.167
.119
.161
.148
.183
.156
RUKi NO.
P /P
r t
.914
.929
.897
.913
.750
.728
.662
.733
.728
,659
.645
.806
.643
.743
.778
.716
.875
.800
,877
.822
.844
.133
.168
.167
.156
T

-0.000
.0.000
.0.000
-0.000
n.ooo
.758
.765
.767
.760
.751
.76f>
.494
.488
.491
,489
.491
.491
-0.000
.0.000
0.000
.496
.497
.498
.496
.493
.496
3
T

.998
.998
.998
.998
.9(12
.913
«9r7
.900
,881
,896
,892
.867
,881
.892
.861
.890
.998
.998
.998
.998
.998
.430
.420
.422
,4?4
.'

-0.00
.0.00
-0,00
-0.00
0.00
30.22
29.21
28.93
29,93
31.23
29.90
76.91
78. ?4
77,57
78.01
77.57
77.66
-0.00
-0.00
0,00
76.46
76,24
76.03
76.46
77.13
76.47

-

.22
• 22
• 22
.22
11.25
9.93
10.64
11.49
13.82
11.98
12.46
15.56
13. «2
12.46
16.32
12.70
.22
.22
.22
.22
.22
92.04
94.60
94.fl8
93.57
SI

-0.00
-o.oo
-O.OQ
-0,00
0.00
10.15
9.82
9.77
10.06
10.49
10.05
25. e*
26.29
26.16
86.21
26.06
26. ng
-0.00
.0.00
0.00
25.69
25.62
25.54
25.69
25.91
25.69

M

• 21
•21
•21
• 21
10.61
9.36
10*04
1".83
13.03
11.29
11.75
14.68
13.03
11.75
15.39
11.98
• 21
•21
• 21
•21
• 21
86.79
89.21
88.72
88. ?4
                  H-6

-------
            p /p
                    p /p
                              T
40
4o
41
»n
40
40
40
40
40
4n
AVG
o
n
0
0
0
AVG
100
100
100
100
100
100
100
AVQ
7?0
1100
1350
1300
UOO
1650
1600
1910
1300
1200
13T5
400
600
450
B50
1 000
660
651
1150
300
100
100
160
ooo
1066
son
son
n on
lloo
125o
13flfi
12no
1500
85o
900
1059
-0
•o
*2
•o
• n
n
900
145o
1700
I4no
1400
145o
1300
1371
250
460
600
500
600
700
750
850
7QO
450
586
300
450
350
700
900
540
50
100
80
90
SO
SO
BO
80
.667
.727
.815
.846
.781
.836
.750
.789
.654
.750
.762
0.000
0.000
a. ooo
0.000
0.000
0.000
1.385
1.261
1.308
1.273
1«273
1«250
1«3PO
1-293
.333
.418
.444
.385
.375
.474
.469
.447
.538
.375
.4?!
.750
.750
.778
.874
.900
.800
.077
.087
.062
.082
.073
• 069
•080
• 076
.630
.680
.630
.631
.628
.537
.628
.6?5
.625
.731
.634
.998
.998
.998
.998
.998
.998
.259
.3nl
.266
.326
.3p4
• 293
.296
.292
50.39
42.06
SO. 39
50.21
50.^3
67. 8n
50.73
51.25
51.25
34.17
49.90
.22
.22
.22
.22
.22
.22
147.32
13Q.93
144.4]
122.23
129.85
1 33«87
132*76
134.48
47.51
39.66
47.51
47.35
47.84
63.94
47.84
48.33
48.31
3?. 22
47.05
.21
•21
•21
•21
.21
•21
138,92
123.47
136.18
115.26
122.45
126«24
125.19
126-82
LTOAB FIELD DATA, RUM NO.  4
            p /p
                    p /p
0
n
0
n
AVG
40
40
40
40
40
4n
40
40
+ 0
40
40
40
40
*0
AVQ
n
0
0
AVG
1900
1660
1600
1900
1765
1160
1550
1200
1300
1650
1300
1730
1600
1550
900
1100
1400
1300
1680
1389
BOO
1400
1600
1Z67
-0
•n
-0
-n
n
750
lf)5n
80o
900
1200
9fn
lion
1000
lion
60o
65o
9oii
1000
118o
938
-0
•o
-o
n
1400
1100
1100
1490
1272
300
500
400
400
6QO
430
600
500
400
250
300
400
450
600
418
400
750
1000
717
0.000
1.000
0.000
0.000
n.OOO
.647
.677
.667
.692
.727
.692
.629
.625
.710
.667
.591
.643
.769
.702
.674
0.000
0.000
0.000
0.000
.737
.663
.688
.764
.71 8
.259
.3?3
.333
.308
.364
.331
.343
.313
.258
.278
.273
.286
.346
.357
.312
.500
.536
.635
.554
.995
.995
,995
.995
.995
.735
.7*8
.747
.7^4
.752
.732
.752
.817
.740
.740
.740
.740
.740
.740
.748
.995
.995
.995
.995
.55
.55
.55
.55
.55
33.58
31.66
31.81
30.79
31.08
34.02
31.08
22.04
32-84
32.04
32.84
32.84
32.84
32.84
31.65
.55
.55
.55
.55
.?5
.25
.25
.25
• 25
15.11
14.25
14,31
13.86
13.99
15.31
13.99
9.9?
14.78
14.78
14.78
14.7R
14.78
14.78
14.24
.25
•25
• 25
.25
              H-7

-------
          Table H-2  .Cent inucd ^
LIDAR FIELD DATA, RUN NO.  -1  (CONTINUED)
D

40
40
40
40
*0
40
AVG
t
o
AVG
40
4n
*0
40
AVG
0
0
AVG
20
20
20
20
20
2"
AVG
0
n
0
AVG
20
20
20
20
AVG
0
0
0
AVG
100
100
100
100
100
100
100
100
100
AVG
0
0
n
AVQ
P
t
1400
1400
1550
1450
1350
1700
1475
1300
1500
1400
1400
1700
1600
1SOO
1550
1300
loco
1150
1380
1200
1*10
1600
1060
920
1295
1210
11?0
1300
1220
1000
1430
lino
13(50
1207
1250
950
1200
1133
1150
1400
1151
10*0
lino
1200
1410
1550
1500
12*2
1080
mo
1108
1113
P
a
75*
100P
iioii
980
loon
lion
988
•0
• o
fl
loon
I25r
lino
llor
111?
-r
-0
P
400
36o
5Pn
500
256
2fto
36»
-0
-p
«A
n
29S
30ii
35n
33o
317
-n
-0
"2
n
142n
175«
1600
Uftfl
16no
168n
172?
1800
2100
16P1
-0
• 0
-n
n
p
r
350
500
450
450
450
570
462
700
900
800
400
500
560
500
490
TOO
600
650
700
610
810
900
540
500
677
ft90
700
900
763
480
800
600
650
632
700
600
800
700
100
150
120
120
130
150
150
160
200
142
680
710
760
717
P /P
& t
.536
.714
.710
.676
.741
.647
.671
0.000
0.000
0.000
.714
.735
.688
.733
.718
0.000
0.000
0.000
.290
.300
.311
.313
.236
.217
.278
0.000
0.000
0.000
0.000
.290
.210
.318
.254
.268
0,000
0.000
0.000
0.000
.235
.250
.391
.352
.455
.400
.220
.161
1.400
1.318
0.000
0.000
0.000
0.000
p /p
r t
.250
.357
.290
.310
.333
.335
.313
.538
.600
.569
.286
.294
.350
.3^3
.316
.538
,600
.569
.507
.508
.503
.563
.509
.54.1
.522
.570
,609
.692
.624
.480
.559
.545
.500
.521
.560
.632
.667
.619
.087
.107
.104
.111
.118
.125
.106
.103
.133
.111
.630
.612
.691
.644
T

.738
.731
.724
.742
.751
.715
.733
.995
.995
.995
.740
.728
.72?
.726
.730
.995
.995
.995
.890
.927
.922
,9?4
.928
.91P
.918
.995
.995
.995
.995
.918
.927
.910
,911
,916
.995
.995
.995
.995
.424
.418
.428
,4n6
.418
.417
.413
.415
.417
.417
.995
.995
.995
.995
-

33.13
34.17
35.22
32.54
31.23
36.58
33.81
.55
.55
.55
32.94
34.62
34.62
34.92
34.25
.55
.55
.55
12.71
8.27
R.86
8.6?
8.15
9,33
9.32
.55
.55
.55
.55
9.33
a. 27
10.28
10.16
9.51
.55
.55
.55
.55
93.57
95.12
92.54
98.30
95.12
95.38
96.43
95.91
95.38
95.31
.55
.55
.55
.55
M

14.91
15.38
15.85
14.64
14. ns
16.46
15.22
.25
.25
.25
14.78
15.58
15.58
15.71
15.41
.25
.25
.25
5.72
3.72
3.99
3.89
3.67
4,2?
4.19
.25
.25
.?5
.25
4.20
3.72
4,63
4.57
4.28
.?5
.25
.25
.25
42.11
42.81
41 .64
44.23
42.81
42.92
43.40
43.16
4?. 92
42.89
.25
.?5
i?5
• 25
                   H-8

-------
       Table i[-2
LTHAR FIELD  DATA. RUN NO.   <•
             P /P
                      P /P
0
0
0
0
AVG
2n
20
20
20
2n
20
2r
2o
AVG
0
r,
0
AVG
40
40
40
40
40
40
40
40
40
40
AVQ
0
n
0
0
AVG
100
100
100
100
100
100
»VG
315
3*8
360
420
3*6
352
375
377
427
375
4?0
375
450
393
315
344
3«2
347
315
372
339
3<57
420
420
427
398
360
337
378
24n
3"7
360
375
320
342
300
270
285
3*0
340
319
-^
«*
-0
"5
*
152
28fl
23?
232
18S
2P?.
22*
2ls
226
-n
-n
•n
n
597
630
54*
566
617
69ft
607
6i«;
49«;
55i
59, •
-f<
• «
-o
* n
n
77?
727
756
65?
78n
96»
77*
3«2
407
330
502
405
255
352
330
360
322
372
390
465
356
372
360
360
364
234
282
225
225
279
315
330
288
210
255
264
307
330
375
450
365
140
105
8o
75
169
142
llfl
0.000
o.ooo
0.000
0.000
0.000
.432
.768
.624
.543
.480
.671
• 600
.484
.575
0.000
0.000
0.000
0.000
1.895
1 .694
1 .593
1 .426
1 .469
1.657
1.422
1.545
1.375
1.635
1.571
0.000
0-000
0.000
0.000
0.000
2.257
2-423
2.800
2-268
2.167
2.689
2.437
1 .213
1.116
.917
1 .195
1.108
.7?4
.Q39
.RS7
.S43
.859
.886
1.040
1.033
.901
1.191
1 .047
.942
1.057
.743
• 7S8
.664
.567
.664
.750
.773
.7?4
.583
.757
.698
1.279
1.075
1.042
1.200
1.149
.409
.350
.296
.263
.469
.39*
.364
.995
.995
,994
.994
.994
.925
.921
.9?3
.916
.9?!
.915
.9?4
.9!9
.920
.993
.991
.995
.993
.764
.767
.763
.771
.779
.782
.796
.770
.776
.780
.775
.992
.991
.992
.991
.991
.474
,478
.498
.534
.509
.545
.507
.55
.55
.66
.66
.60
8.50
8.97
fl.74
9.57
a. 97
9.69
8.62

-------
LTD*" FIELD  DATA»  RUNi  NO,  7
            p /P
                    P /P
0
(1
0
n
AVG
2i
20
20
20
20
20
20
2o
AVG
0
0
0
c
AVG
40
40
40
+ 0
40
4n
40
4n
AVG
0
0
n
AVG
100
100
10P
100
100
100
100
100
100
100
100
AVG
150
195
175
US
171
140
130
150
150
183
185
inn
135
1*7
135
flO
140
145
125
1*0
164
155
135
165
160
115
138
149
95
125
13«
117
93
1"5
128
124
128
125
1*2
125
115
140
128
119
•2
«n
-n
-0
P
12-5
7?
9^
Us
155
I7n
*5
lln
121
• n
• fl
-n
• 6
6
34*
38«
309
3on
40?
40?
266
31?
340
V A
»«
•0
"
278
294
as?
36ft
357
377
3l«
3SJ,
369
37*
37n
337
130
215
160
160
166
95
88
125
140
135
US
75
105
113
135
55
1*5
105
110
65
65
55
42
65
58
45
65
58
95
115
UO
117
25
35
60
39
41
35
28
40
35
40
40
38
0.000
0.000
0.000
0.000
c.ooc
.893
.554
.633
.967
.838
.919
.950
.815
.821
0.000
0.000
0.000
n.ooo
0.000
2.125
2.366
1.994
2.222
2.455
2.500
2.313
2.261
2.279
0.000
0.000
0.000
n.ooo
2.989
2.800
I .992
2.9"52
2.789
3.01*
3.088
2.848
3.209
2.671
2.891
2.840
.867
1.103
.914
.970
.963
,679
.677
,S33
.933
.730
.784
.750
.778
.770
1.000
.688
1.036
.724
.862
.406
.396
.355
.311
.394
.362
.391
.471
.3*6
1.000
.920
1.077
.999
,?69
.333
.469
.315
.320
.280
.275
.3?0
.304
.286
t313
.317
.993
.993
.993
.993
.993
.865
.935
,9?4
.898
.864
.871
.865
.877
.886
.991
.991
,991
.990
.991
.628
.634
.625
.««*
.60!
.623
.631
.641
.624
.999
.991
.995
.995
.432
.565
,7"8
.547
.514
.515
.514
.513
.482
,5P8
.561
.534
.77
.77
.77
.77
.77
15. «2
7.33
8.62
12.95
15.94
15.06
15. «2
14.31
13.23
.99
.99
.99
1.10
1.01
50.73
49.70
51. ?5
54.26
55.52
51,6.1
50.21
48.50
51 .47
.11
.99
.55
.55
91.53
62.26
37.66
65,79
72.58
72.37
72.58
72.79
79.59
69,65
63.04
69.17
.72
.7?
.72
.72
.72
14.91
6.91
8.13
12.22
15.03
14.20
14,91
13.50
12.48
.93
.93
.93
1.03
.96
47.84
46.86
48,33
51,17
52,36
48.66
47.35
45.73
48.54
.10
,93
.5?
,52
86.31
58.71
35,51
62. n*
68.44
68, 24
68,44
68.64
75,^5
65.68
59.44
65.14
            H-10

-------
LTHAR FIELD oiia, c
D

2n
20
20
20
20
20
AVG
40
40
40
40
40
AVG
100
100
100
100
100
100
100
AVG
P
t
73
114
75
118
113
95
98
95
120
134
129
120
120
95
100
95
124
108
125
130
111
P_

5s
6P
73
6f
9c
9s
7?
19?
32?
384
37?
3nS
31?
37?
37?
487
49?
46?
52?
54fl
466
P
r
75
93
75
100
105
85
89
60
44
62
55
50
54
8
13
8
13
8
15
10
11
p /P
a t
.753
.596
.973
.559
.841
1 .000
.787
2.053
2.708
2.866
2.884
2.500
2.602
3.947
3,750
5.126
3.992
4.306
4,176
4.154
4.207
'UK NO,
P /P
r t
1.027
,816
1 .000
.847
.929
.895
.919
.632
.367
.463
.426
.417
.461
.0*4
.130
.OS*
.105
,074
,1?0
.077
,096
Q
f

.911
.872
.875
.883
.875
,870
,R8l
.630
,6P4
.612
.592
,606
.609
.231
.357
.270
,3?0
.276
.310
.257
.289

„

10.16
14,94
14,56
13.57
14.56
15.19
13.83
50,39
54.98
53.55
57.17
54.62
54.14
159.80
1 12,32
142.78
124.26
140.39
127.72
148,17
136.49



9.59
14. rg
13.73
12.80
13.73
14.32
13,04
47,51
51 .85
50, 49
53.91
51.51
51.05
150.69
105.92
134,65
117.17
132.39
120.44
139,72
128.71
LTD'R FIELD DATA. HUN NO,   9
            P /P
0
0
0
0
0
AVG
20
20
20
2n
20
20
20
20
AVG
0
0
0
r
AVr,
40
40
40
40
40
40
40
40
AVG
95
125
125
114
128
117
108
155
132
110
158
133
160
139
137
120
120
115
115
117
115
120
138
140
120
132
145
118
128
-n
• n
-n
• ^
-0
0
85
7a
loo
6 =
8P
9o
7(1
8?
84
»i
• A
• n
• rt
*
20=;
2o«
2li
IB?
16-»
19?
19?
Ifl?
191
100
125
93
78
110
lo.l
95
1P9
110
70
1*5
108
120
115
108
1*5
100
92
110
112
85
SO
90
80
62
100
88
85
84
0.000
0.000
0.000
0.000
0.000
0.000
.787
.503
.758
.591
.557
.677
.487
.612
.621
0 ,000
0.000
0.000
0.000
0.000
1.783
1.667
1.522
1.300
1.358
1,477
1.324
1.568
1.500
1 .053
1.000
.744
.684
.859
.868
,7S7
.703
.833
.636
.919
.812
.750
.827
.783
1 .208
.833
.800
.957
.950
.739
.667
.652
.571
.517
.758
.607
.720
.654
.909
.999
.999
.997
.999
.999
.900
.913
.9no
.907
.915
,894
.896
.892
.902
.993
.993
.992
.993
.993
.749
.750
.754
.761
.759
.759
.767
.755
.757
.11
• 11
• 11
.33
.11
.15
11 .49
9.93
11 .49
10.64
9.69
12.2?.
11 .98
12.46
11.24
.77
.77
,88
.77
. 79
31.52
31.37
30.79
29.78
30.07
30«07
28.93
30.65
30,40
.04
.04
.04
•11
.04
.05
3.H6
3.34
3.86
3.S8
3.25
4.1]
4.02
4.19
3,78
• ?6
.26
.29
.26
.27
10.59
10.54
10.35
10.01
10.10
10. 10
9.7J
10.30
10.21
             H-ll

-------
       Table U-2   Continued
LIDAR FIELD DATA, RUX NO.  9 (CONTI.NTED)
D

0
n
0
AVG
100
100
100
100
100
100
100
100
AVG

D

n
0
0
0
AVG
20
20
20
20
A«G
40
41
*0
»0
AVG
100
100
100
100
100
AVG
p
t
8!
131
128
115
120
135
109
125
175
115
130
119
122

P
t
99
1(15
140
120
116
100
125
115
115
114
98
1 18
115
135
116
98
1?4
128
105
130
117
p
a
• ft
•ft
• ft
A
36ft
405
338
34=;
36"
345
375
318
35*
LTPAB
p
a
•* *
•ft
•ft
• ft
ft
45
55
45
5ft
4Q
u;
125
14ft
16*
147
300
3l«
314
311
308
31?
P
r
Io5
140
122
122
31
45
35
30
35
35
39
29
35
FIELD
P
r
99
120
145
132
124
94
110
100
98
100
70
70
70
ln2
78
42
38
46
30
35
38
P /P
a t
0.000
o.ooo
0.000
0.000
3.000
3.000
3.101
2.760
2.880
3.000
2.885
2.672
2.912
DATA,
P /P
a t
0.000
0.000
o.ooo
0.000
0.000
.450
.440
.391
.435
.429
1.429
1.059
1.217
1.222
1.232
3.061
2.565
2.453
2.962
2.369
2.682
p /P
r t
1.235
1.069
.953
1.086
.258
.333
.321
.240
.280
.304
.300
.244
.285
"UNI NO. i
p /p
r t
1.000
1.143
1.036
1.100
1 .070
.940
.880
.870
.852
.885
.714
.593
.609
.756
.668
.429
.306
.359
.286
.269
.33(1
T

.993
.994
.994
.994
.494
.497
.490
.491
.487
,*91
.495
.489
.492
0
T

.993
.994
.993
.993
.993
.922
.937
.931
.920
.927
.774
.798
.781
.788
,7P5
.542
.543
.542
.504
.514
.579
j

.77
• 66
.66
• 69
76.91
76,74
77.79
77.57
78.46
77.57
76,68
78.01
77.41

„

.77
.66
.77
.77
.74
8,86
7.10
7.8(1
9.09
8.21
27.94
24.61
26.96
25.98
26.37
66.79
66.59
66.79
74.72
72.58
69.49
M

.26
•22
• 22
• 23
25.84
25.62
26.14
26.06
26.36
26.06
25.77
26.21
26.01

\(

.26
•72
•7.6
• 76
• 25
?.98
2.38
2«62
3.06
2.76
9.39
8.27
9.06
8.73
8.86
22.44
22.37
22.44
25.11
24.39
23.35
 LTOAR  FIELD DATA,  RUN NO.  11
D

0
0
0
0
0
0
0
0
0
0
0
AVG
p
t
139
130
148
144
138
128
145
135
140
132
139
138
P
s
• n
•5
-1
• ft
-n
-0
• ft
• n
• n
• ft
•n
-0
p
r
155
145
152
165
175
130
142
138
150
148
140
149
P /P
s t
0.000
0.000
o.ooo
1.000
0.000
0.000
0.000
0.000
0.000
o.OOO
0.000
0.000
p /p
r t
1.115
1.115
1 .027
1.146
1.268
1 .OlA
.979
1.022
1.071
1.121
1.007
1.081
             H-12

-------
D

0
0
1
n
0
i
n
0
0
AVG
D

(1
0
0
n
0
0
n
n
0
0
0
0
AVG
P
t
lAfl
135
12?
124
140
125
131
1"5
125
127
P
t
130
125
125
12n
130
108
119
135
12n
135
136
130
126
P
s
-0
• n
— n
-n
-n
• n
• n
« n
-0
-0
p
5
434
44)
44f
481
42*
33T
36S
494
479
47
-------
         Table H-2  Continued
LIDAR FIELD DATA, RUN NO.  12  (CONTINUED)
D

*0
40
40
4n
40
*0
*0
AVG
0
(1
0
0
AVG
100
100
100
ion
100
100
100
100
AV4
17.01
.55
.77
• 55
.55
.60
49.87

-------
D

100
100
ion
101
100
:oo
ion
100
AVO
40
41
4C
4,1
40
41
40
AVC

D

•>
n
0
1
AVG
20
2'J
2fl
20
2')
AVG
40
41
4n
40
41
40
40
40
AVG
0
0
n
0
AVC,
100
100
100
ion
100
100
100
100
AVG
n
0
0
0
AVG
P
t
no
104
110
110
130
105
98
104
119
102
115
11"
139
110
122
110
115

P
t
89
1"5
95
qo
95
85
»5
inn
119
]n(j
99
sn
72
89
lift
90
119
115
113
99
114
90
115
112
113
100
115
<=9
no
94
110
109
li
• 1
.;
I?
s
q
li
9
9
2"
2i
2l
2*
2ft
24
2P
3*
26
« r>
-0
-0
• n
n
65
5?
5?
61
40
65
67
60
59
• n
.0
.n
-ft
rt
P
r
85
88
85
98
112
82
80
100
91
118
U)5
If 9
142
l'':5
149
11 5
119
FIELD
p
r
90
115
82
100
97
106
85
115
135
130
112
95
qn
88
122
95
125
129
126
l-<9
115
100
130
128
118
115
ion
95
110
78
102
U:9
110
102
130
110
159
132
133
p /p
a t
.955
.9?3
.909
.955
.985
.848
.827
1.231
.954
.333
.270
.264
.281
.264
.32"
.236
.281
DATA.
P /P
a t
0.000
0.000
0.000
0.000
0.000
.118
.094
.090
.OPS
.083
.094
.222
.292
.236
.245
.222
.202
.243
.265
.241
0.000
0.000
0.000
0.000
O.OPO
.620
.452
.584
.555
.511
.591
.569
.639
.565
0.000
0.000
0.000
0.000
0.000
p /p
r t
.773
.846
.773
.891
.862
.781
.81 6
.962
.833
1 .157
.913
.98?
I.0?2
.955
!.??!
.955
I ,0?9
SUN NO. 14
p /r
r t
l.hi i
1.005
."63
l.ll 1
1 .021
l.?47
1 .000
1.050
1 .144
1.193
1.1?7
1.056
1 .250
.989
1.151
1 .056
1 .050
1.1?2
1.115
1 .099
1.106
1.111
1 ,?33
1.143
1 .149
1.1 = 0
.S70
1.0*7
1 .010
.830
.9?7
I .000
1 .019
.983
1.193
1,058
1.233
1,200
1.171
T

.851
.792
,8»2
.840
.795
.874
.869
.842
.839
.951
,9*1
.959
.949
.953
.954
,958
.954

T

.993
.992
.993
,995
.993
.97fl
.986
.981
.9°0
.983
.982
.962
.959
.961
.962
.9*2
.965
.967
.969
.963
.997
.997
.997
.997
,997
.9i6
.906
,916
,9"8
.904
.916
.908
.917
,916
.9*1
,994
,993
.992
,990


17.59
25.43
18.75
19.01
25.02
14.69
15.31
18.75
19.32
5.4?
5.48
4.57
5.71
5. '5
5.14
4.6B
5.19

-

.77
.88
.77
.55
.74
?.43
1 ,C4
2.19
2.21
1 .07
2.13
4.22
4.57
4 ,34
4.22
4.22
3.«9
3.66
3.43
4.17
.33
.33
.33
.33
.33
10.77
10.77
10,77
10.5?
11.01
1C. 77
10. =2
10.64
10.72
2.09
.66
.77
.88
1 .10


16.59
23.98
17.69
17.93
23.59
13.85
14,44
17.69
18.22
5.17
5.17
4.31
5.39
4.95
4.84
4.4\
4.S9



.26
.'9
.26
.18
.'5
.8?
.52
.71
.74
.63
.69
1.42
1.53
I .46
1.4?
1 .4?
1.31
1.23
1.15
1.37
.11
.11
.11
.11
.11
3.62
3.62
3.62
3.54
3.75
3.62
3.5»
3.58
3.60
.7n
.'2
.26
.29
.37
H-15

-------