United States     Office of Air Quality       EPA-450/4-81-034
          Environmental Protection  Planning and Standards     November 1981
          Agency        Research Triangle Park NC 27711

          Air
&EFA     The Effects Of Chemistry
            And  Meteorology On
               Ozone Control
          Calculations Using Simple
           Trajectory Models And
            The EKMA Procedure

-------
                              EPA-450/4-81-034
    The Effects Of Chemistry
       And  Meteorology  On
   Ozone Control  Calculations
Using  Simple Trajectory Models
    And The EKMA  Procedure
                     by

             H.E. Jeffries, K.G. Sexton and
                  C.N. Salmi
       U.S. Environmental Protection Agency

       Region V, Library
       230 Soul'n Dearborn Street
       Chicago, Illinois  60604
              Contract No. 68-02-3523
          EPA Project Officer: Edwin L. Meyer, Jr.
                  Prepared for

           U.S. Environmental Protection Agency
            Office of Air, Noise and Radiation
         Office of Air Quality Planning and Standards
         Research Triangle Park, North Carolina 27711

                 November 1981

-------
      This report was furnished to the U.S. Environmental Protection
      Agency by the University of North Carolina, Chapel Hill,
      North Carolina in fulfillment of Contract No. 68-02-3523.  The
      contents of this report are reproduced herein as received from
      the University of North Carolina.  The opinions, findings, and
      conclusions expressed are those of the author and not necessarily
      those of the Environmental Protection Agency.  Mention of
      company or product names is not to be considered as an endorse-
      ment by the Environmental Protection Agency.
U,S. Er"'	---*-'
                                     ii

-------
ACKNOWLEDGEMENTS

     The authors gratefully acknowledge the time, efforts, and thoughtful -
ness of their friends and co-workers who contributed significantly to this
project.  At UNC:  Douglas Case generated almost all the plots.   Amy
Gerhardt managed the business details of the project, produced plots,
reviewed and proofed text, and dealt  effectively with the myriad of details
necessary for the project's success.   Mario DeSousa generated some, and
modified most,of the new computer code produced in this project.   He
effectively managed the computer libraries and disk files.  Kenneth Crossen
developed the histogram fitting algorithm, implemented the OZIPP  plotting
packages, and generated and supplied OAQPS with FORTRAN code to correct
OZIPP; Richard Kamens and Michael  Gery provided technical  dialogue.  Donna
Simmons typed and corrected the report and tables;  Phyllis Carlton assisted
in typing many sections.  We are especially grateful to Suzanne Spencer
who applied the finishing touches to the figures and the text.

     At EPA:
     The project would not have been  possible without the  continual
cooperation of Gerry Gipson at OAQPS  who supplied large quantities of data;
this required attention to meticulous detail  and patience  for which we are
grateful. Robert Kelly and Edwin Meyer at OAQPS, and Joan  Novak at the
Meteorology Branch also assisted with data and discussion.

     The most significant input to the project was that of the model
developers.  We gratefully acknowledge the willingness to  cooperate, the
significant time spent in discussion, and their patience with us  while we
sometimes mistreated their creative work.  We thank Jim Killus and Gary
Whitten of System Applications Inc.,  the source of the Carbon Bond mechanism.
We thank Kenneth Demerjian of EPA not only for supplying his mechanism, but
also for significant time spent discussing its use and the RAPS emission
inventory.   We  thank Marcia Dodge for her cooperation and  tolerance.  We
thank Gregory McRae and John Seinfeld of Cal  Tech for their willingness to
share their work before they had completed their efforts.

     Portions of this work were partially supported by EPA Grant  R807762-01,
"Modeling Observed Hydrocarbon Mixture Reactivity Effects".
                                   iii

-------
                              Table of Contents
List of Figures                                                        vii

List of Tables                                                         xiv

Glossary of Terms                                                     xvii

1.0  SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS                           1
     1.1  SUMMARY                                                        i
          1.1.1  Purpose
          1.1.2  Need
          1.1.3  Approach
          1.1.4  Results
     1.2  CONCLUSIONS                                                   H
     1.3  RECOMMENDATIONS                                               15

2.0  INTRODUCTION                                                       17
     2.1  THE REQUIREMENT FOR MODELLING                                 17
     2.2  THE MODELS OF INTEREST TO THIS STUDY                          18
          2.2.1  Simple Trajectory Models
          2.2.2  Control Calculations
                 2.2.2.1  Isopleth diagrams
                 2.2.2.2  Application of EKMA
     2.3  PURPOSE                                                       24

3.0  BACKGROUND AND APPROACH                                            29
     3.1  PREVIOUS STUDIES                                              29
          3.1.1  UNC mechanism modelling study
          3.1.2  EPA/OAQPS OZIPP/EKMA Study
     3.2  APPROACH                                                      4!
          3.2.1  General Concept
          3.2.2  Tasks Undertaken

4.0  CHEMICAL KINETICS MECHANISMS                                       44
     4.1  SELECTION                                                     44
     4.?  MECHANISM REACTIONS                                           44
     4.3  PHOTOLYTIC RATES                                              59
     4.4  HYDROCARBON COMPOSITION INPUT CALCULATIONS                    65

5.0  BUREAU OF MINES SMOG CHAMBER SIMULATIONS                           70
     5.1  THE BUREAU OF MINES DATABASE                                  70
     5.2  ASSUMPTIONS                                                   75
     5.3  SIMULATION RESULTS                                            77
     5.4  BOM ISOPLETHS AND CROSS SECTIONS                              82
     5.5  SUMMARY                                                       88

-------
 6.0  REGIONAL  AIR POLLUTION  STUDIES  DATABASE:   EPA SUPPLIED  INFORMATION  89
      6.1   GENERAL DESCRIPTION                                            39
      6.2   DAYS SELECTED                                                 94
      6.3   INFORMATION SUPPLIED  BY  EPA                                    97

 7.0  INPUT DATA FOR SIMPLE TRAJECTORY MODELS  FOR RAPS DAYS               98
      7.1   TRAJECTORIES                                                  98
      7.2   AMBIENT DATA ALONG TRAJECTORIES                               122
      7.3   EMISSIONS DATA  ALONG  TRAJECTORIES                             136
           7.3.1   Procedure for  Processing  Emissions
           7.3.2  Results
           7.3.3  Method of Treating  Hydrocarbon Composition
           7.3.4  Sensitivity of Demerjian  and Carbon Bond Mechanisms
                  to RAPS  Ambient Hydrocarbon  Composition in  Smog
                  Chamber  Simulations
      7.4   MIXING HEIGHT PROFILES                                       167
      7.5   OZONE ALOFT                                                  178

 8.0  RAPS SIMULATIONS WITH SIMPLE  TRAJECTORY  MODELS                     180
      8.1   METEOROLOGY ONLY SIMPLE  TRAJECTORY  MODEL SIMULATIONS OF
           RAPS DAYS                                                    188
      8.2   PHOTOCHEMICAL SIMPLE  TRAJECTORY  MODEL SIMULATIONS  OF RAPS
           DAYS                                                         199
      5.3   RESPONSE OF SELECTED  PHOTOCHEMICAL  SIMULATIONS TO  DIFFERENT
           INPUT ASSUMPTIONS                                             219
           8.3.1   Carbon Bond Mechanisms' Response to Increased
                  Aldehydes and  Photolytic  Rates
           8.3.2  Demerjian and  Carbon Bond Mechanisms'  Response
                  to Hydrocarbon Composition Changes
           8.3.3  Effect of Trajectory Assumptions for Selected Days
           8.3.4  Effect of Emission  Pattern for Selected Days
           8.3.5  Effect of Ozone Aloft  Assumptions for  Selected  Days
      8.4   SUMMARY                                                      265

 9.0  ISOPLETHS AND CONTROL CALCULATIONS                                270
      9.1   OZONE ISOPLETHS GENERATED  FROM PHOTOCHEMICAL  SIMPLE
           TRAJECTORY MODEL SIMULATIONS                                  270
      9.2   OZONE PRECURSOR CONTROL  CALCULATIONS  FROM PKSTM  ISOPLETHS
           AND  THE EKMA PROCEDURE                                       285

10.0  RAPS SIMULATION WITH LEVEL III  MODEL                               297
      10.1  INPUT DATA                                                  297
      10.2  SIMULATION RESULTS                                           304
      10.3  OZONE-PRECURSOR CONTROL CALCULATIONS USING THE  EKMA
            PROCEDURE                                                   309

REFERENCES                                                               327

APPENDIX A:  THE HISFIT ALGORITHM:   AN  AREA TRUE HISTOGRAM               330
APPENDIX B:  SPECIAL EKMA  COMPARISONS                                    336
APPENDIX C:  SIMULATIONS OF   INDIVIDUAL  BUREAU  OF MINES  EXPERIMENTS
             USING CARBON  BOND,  DEMERJIAN,  DODGE,  AND  CIT
             MECHANISMS                                                  342

-------
                             List of Figures


Figure

  1      Components of a Simple Trajectory Model                          20

  2      The Standard OZIPP/DODGE Isopleth Diagram                       23

  3      An EKMA Example                                                 25

  4      Experimental  Data  from UNC Outdoor Smog  Chamber,                 31
        August 8,  1977

  5      Models of  August 8,  1977                                        32

  6      Experimental  Data  from UNC Outdoor Smog  Chamber,                 33
        June 19, 1978

  7      Models of  June 19,  1978                                         34

  8      Experimental  Data  from UNC Outdoor Smog  Chamber                  35
        September  18,  1976

  9      Models of  September  18, 1976                                    36

 10      Experimental  Data  from UNC Outdoor Smog  Chamber  for  Mix          37
        vs Mix and Toluene,  November 19,  1973

 11      Effects of Hydrocarbon Composition Change on  Carbon  Bond  II      38
        Mechanism  Predictions

 12      Effects of Hydrocarbon Composition Change on  Demerjian           39
        Mechanism  Predictions

 13      Scatter Diagram for  OAQPS OZIPP/DODGE  Simulation  of  RAPS  Days    40

 14      Comparison of Cal  Tech Simulation Using  the Cal  Tech Mechanism   57
        with UNC Simulation  Using the Cel  Tech Mechanism

 15      Simulation of University of California-Riverside  Smog Chamber    58
        Experiment SUR-119J  with the CIT  Mechanism

 16      Examples of Experimental  Data from the Bureau of  Mines Smog      71
        Chamber Auto  Exhaust Data Set

 17      Bureau of  Mines Run  Locations and Ozone  Maxima                   72
                                    vii

-------
Figure

  18  •   Bureau of Mines Replicate  Run  Comparison                         73

  19     Scatter Diagrams for Mechanism Simulations  and  BOM  Data,         79
         Ozone Maxima

  20     Scatter Diagrams for Mechanism Simulations  and  BOM  Data,         80
         Time to Ozone Maxima

  21     Scatter Diagrams for Two Versions of Demerjian  Mechanism  and
         BOM Data, Ozone Maxima                                          81

  22     Scatter Diagrams for Two Versions of Demerjian  Mechanism  and
         BOM Data, Time for Ozone Maxima                                 81

  23     Ozone Isopleths for the BOM Simulations                         83

  24     Ozone-Hydrocarbon Cross Sections at Constant NO  for BOM
         Simulations                                                     84

  25     Ozone-NO  Cross Sections at Constant Hydrocarbon for BOM
         Simulations                                                     85

  26     Comparison of  Hydrocarbon and NO  Ozone Cross Sections for
         BOM Simulations                                                 86

  27     The Regional Air Pollution Study Area at  St.  Louis               91

  28     The Regional Air Monitoring Station  UTM Co-ordinates            92

  29     RAPS Major Point Sources                                        95

  30     RAPS Grid System Overlay for Area Emissions Inventory           96

  31     Trajectory Map, Day 159, June  7                                 101

  32     Individual  Trajectories, Day 159                                102

  33     Boxes  Generated from Trajectories,  Day 159                       102

  34     Trajectory Map, Day 160, Site  115,  June 8                       103

  35     Individual  Trajectories, Day 160, Site 115                       104

  36     Boxes  Generated from Trajectories,  Day 160, Site 115            1°4

  37     Trajectory Map,  Day 160,  Site 103, June 8                        1°5

  38     Individual  Trajectories, Day 160, Site 103                       106

  39     Boxes  Generated from Trajectories,  Day 160, Site 103            106

  40     Comparison  of Individual Trajectories, Both Sites,  Day 160      108

                                     viii

-------
Figure                                                                  Page

  41       Trajectory Map,  Day 195,  Site  114,  July  13                      109

  42       Boxes Generated  from Trajectories,  Day 195                      110

  43       Trajectory Map,  Day 201,  July  19                                111

  44       Individual  Trajectories,  Day 201                                112

  45       Trajectory Map,  Day 221,  August 8                               113

  46       Individual  Trajectories,  Day 221                                114

  47       Trajectory Map,  Day 238,  August 25                              115

  48       Individual  Trajectories,  Day 238                                116

  49       Boxes Generated  from Trajectories,  Day 338                      116

  50       Trajectory Map,  Day 261,  September  17                           117

  51       Individual  Trajectories,  Day 261                                118

  52       Trajectory Map,  Day 275,  October 1                              119

  53       Boxes Generated  by  Inspection                                   120

  54       Trajectory Map,  Day 276,  October 2                              121

  55       Ambient  Data,  Weighted  Mean, Max and Min  for Day  159            126
          Trajectory, June 7

  56       Ambient  Data,  Weighted  Mean, Max and Min  for Day  160/115        127
          Trajectory, June 8

  57       Ambient  Data,  Weighted  Mean, Max and Min  for Day  160/103        128
          Trajectory,  June 8

  58       Ambient  Data,  Weighted  Mean, Max and Min  for Day  195            129
          Trajectory, July 13

  59       Ambient  Data,  Weighted  Mean, Max and Min  for Day  201            130
          Trajectory, July 19

  60       Ambient  Data,  Weighted  Mean, Max and Min  for Day  221            131
          Trajectory, August  8

  61       Ambient  Data,  Weighted  Mean, Max and Min  for Day  238            132
          Trajectory, August  25
                                    IX

-------
Figure                                                                 Page

  62      Ambient Data,  Weighted  Mean,  Max  and Min  for  Day 261 Trajectory
          September 17                                                   133

  63      Ambient Data,  Weighted  Mean,  Max  and Min  for  Day 275 Trajectory,
          October 1                                                      134

  64      Ambient Data,  Weighted  Mean,  Max  and Min  for  Day 276 Trajectory,
          October 2                                                     135

  65      Example Segment of RAPS Gridded  Inventory                     137

  66      Distribution  of CO,  NO  ,  and  HC  Emissions for Selected RAPS
          Days                 x                                       147

  67      Carbon  Distribution  (Percent)  of  NMHC  for Selected  RAPS Days   147

  68      Examples of Fitting  Equal-Area Smooth  Curve to Hourly Averaged
          Emission Values                                               149

  69      Absolute Emisson Rates  for  CO, NO , NMHC  used in Simple
          Trajectory Model  Simulations,  AllxDays                         152

  70      Effect  of HC  Composition  and  Concentration on Outdoor Smog
          Chamber Type  Simulations  with the Demerjian Mechanism          163

  71      Effect  of HC  Composition  and  Concentration on Outdoor Smog
          Chamber Type  Simulations  with the Carbon  Bond II Mechanism     164

  72      Effect  of Aldehyde Photolytic Rate with HC Composition and
          Concentration  in Outdoor  Smog Chamber  Type Simulation with
          the Demerjian  Mechanism                                       166

  73      The Characteristic Growth Curve                                170

  74      Mixing  Height Profiles, All Days                               173

  75      Dilution Rates Resulting  From Selected Mixing Height Profiles  177

  76      Relationship  Between Ozone  Aloft  and Post-Inversion Breakup
          Ground  Concentrations                                         178

  77      MOSTM Simulation Results  for  NO  . HC,  CO, and 0,; All Days     183
                                        X                *3

  78      Scatter-Diagram MOSTM Predictions and  Ambient 03 Maxima        197

  79      Photochemical  Simple Trajectory  Model  Simulation Results
          for NO, N02,  and 03; All  Days                                 201

  80      Scatter Diagrams:   Mechanism  Predictions  and  Ambient 0^ Maxima 217

-------
Figure                                                                  Page

  81       Scatter Diagrams:   Mixing Height Predictions and Ambient Oq
          Maxima                                                    d    218

  82       Effect of Increased Aldehyde Emission Fraction for Carbon Bond
          Mechanism;  Day 275 (8)                                          221

  83       Effect of Increased Aldehyde Emission Fraction for Carbon Bond
          Mechanism;  Day 159                                             222

  84       Effect of Increased Aldehyde Emisson Fraction for Carbon Bond
          Mechanism;  Day 221                                             223

  85       Effect of Increased Aldehyde Emissions Fraction:   Effect of
          Changing From "OLD" to  "NEW" Photolytic Rates; Day 261          225

  86       Effect of Changing from "OLD"  to "NEW" Photolytic Rates  and
          Increasing  Both the Radical  Photolysis and Aldehyde Emissions
          Fraction; Day 275                                               226

  87       Effect of Changing from "OLD"  to "NEW" Photolytic Rates  for
          Dodge  Mechanism;  Day 275                                       227

  88       Effect of Changing Hydrocarbon Composition on Demerjian
          Mechanism's  Simulation  of Day 275                              229

  89       Effect of Changing Hydrocarbon Composition on Carbon Bond's
          Simulation  of Day  275                                          231

  90       Supplied Column Path for Day 238                               234

  91       Simulations  with  Different Trajectories;  Day 238                237

  92       Instantaneous Emission  Rates Along  Trajectories;  Day 238       243

  93       Supplied Column Path for Day 201                                245

  94       Simulations  with  Different Trajectories,  Day 201                247

  95       Instantaneous Emission  Rates Along  Trajectories,  Day 201        253

  96       Effects of  Ozone Aloft  on Simulations, Day 275                 261

  97       Effects of  Ozone Aloft  on Simulations, Day 159                 262

  98       Example of  Locations for Simulations Used to Produce an  Ozone
          Isopleth Surface                                                271

  99       An  Ozone Isopleth  Surface (A); Surface Sliced at  Constant NO
          (B); Surface  Sliced at  Constant  HC(C); Simple Trajectory Model
          Simulation  for Day (D);  "*"  Identification Coordinates of Initial
          Conditions  and 0^  Maximum                                      272
                                     XI

-------
Figure                                                                 Page

  100     A  Ozone-HC Cross  Sections,  October 1  Simulations              274
          B  Ozone Isopleths,  October  1  Simulations                      275
          C  Ozone-NO  Cross Sections,  October 1  Simulations             276
          D  Ozone Isopleth  and Cross  Sections for CIT October 1
             Characteristic  Curve Simulation                            277
          E  Ozone Isopleths over Larger Range for Carbon  Bond
             Simulations                                                278

  101     A  Ozone-HC Cross  Sections,  June 7 Simulations                280
          B  Ozone Isopleths,  June 7 Simulations                        281
          C  Ozone-NO  Cross Sections,  June 7 Simulations                282
          D  Ozone Isopleths and Cross  Section for CIT June  7
             Characteristic  Curve Simulation                            283

  102     The EKMA Procedure                                            286

  103     Ozone-HC Fraction  Plots for  October 1  Simulations              289

  104     Ozone-HC Fraction  Plots for  June 7 Simulations                290

  105     Ozone Control  Diagrams for October 4                          293

  106     Ozone Control  Diagrams for June 7                             294

  107     Control  Predictions  by Day by Mixing Height, by  Mechanism     295

  108     Level  III Simulation Profiles for Three RAPS Days with  250 m
          Initial  Mixing Height                                         301

  109     Level  III Simulation Profiles for Three RAPS Days with  350 m
          Initial  Mixing Height                                         302

  110     D159 Level  III Isopleth Diagrams and Ozone-HC Fraction  Diagrams
          with 250 m Initial Mixing Height                              310

  111     D195 Level  III Isopleth Diagrams and Ozone-HC Fraction  Diagrams
          with 250 m Initial Mixing Height                              311

  112     D275 Level  III Isopleth Diagrams and Ozone-HC Fraction  Diagrams
          with 250 m Initial Mixing Height                              312

  113     Level  III Isopleth Diagrams  and Ozone-HC Fraction  Diagrams  with
          350 m Initial  Mixing Height  for Three RAPS  Days                313

  114     Level  III Ozone Reduction Diagrams for Three RAPS  Days  with
          250 m Initial  Mixing Height                                    318

  115     Level  III Ozone Reduction Diagrams for Three iRAPS  Days  with
          350 m Initial  Mixing Height                                    319
                                    xn

-------
Figure                                                                 Page

  116      Effect of Increasing  Simulation  Time  from G  to  10  Hours
          for D275                                                       324

  117      Control  Predictions by Model  Level, by  Day,  by  Mechanism       326

  A.I      Comparison of Equal Area  Histogram  Algorithms Utilized by  OZIPP
          and by UNC                                                   333

  A.2      Comparison of Equal Area  Histogram  Algorithms with OZIPP showing
          Strong Oscillation                                            334

  B.I      Effect of Assumptions of  Aldehyde Carbon  Mole Fraction on  EKMA
          for RAPS  October  1, 1976                                      338

  B.2      Effect of changing from a Level  II  to a Level III  Application
       •   on  EKMA for RAPS  October  1,  1976                             341

   C       Simulation and Experimental  Data Profiles for Bureau of Mines
          Study                                                        343

-------
                             List of Tables


Table                                                                  Page

  1      Mathematical  Equations of A Simple Trajectory  Model               21

  2     Nitrogen Balance at Ozone Maximum for Models and  Data,
        August 8, 1977                                                   31

  3     Nitrogen Balance at Ozone Maximum for Models,
        June 19, 1978                                                    33

  4     Nitrogen Balance at Ozone Maximum for Models,
        September 18, 1976                                               35

  5     The General  Characteristics of the Selected Mechanisms            45

  6     The Dodge Mechanism                                              47

  7     The Demerjian Mechanism                                          50

  8     The Carbon Bond II Mechanism                                     52

  9     The California Institute of Technology Mechanism                  55

 10     Old and New Photolytic Rates at Selected Solar Zenith Angles      61

 11      Photolytic Rates Used in Bureau of Mines Simulations              64

 12     Hydrocarbon Species Characteristics  in Each Mechanism             66

 13     Factors Used in BOM Simulations                                  68

 14     Bureau of Mines Model Simulation Results                          74

 15     Days from RAPS Database Selected for Simple Trajectory  Model
        Simulations                                                       94

 16     Sample Ambient Data Printout                                     123

 17     Initial Conditions of HC, NO. CO Chosen for  Simulations         124
                                    X

 18     Sample Area  Emission Inventory Printout                         138

 19     Sample  Point Source Emissions Inventory Printout                139

 20     Species Molecular Weights Used in Emissions  Calcuation            140

 21      Hydrocarbon Carbon Numbers Used in Emissions  Calcuations         140
                                    XIV

-------
Table                                                                    Pare

  22     Sample  Emissions Calculation for Point and Area Sources         143

  23     Distributions of Emissions for Selected RAPS Days                145

  24     Distributions of Emissions for the St.  Louis AQCR                145

  25     Carbon Distribution of NMHC for Selected RAPS Days               146

  26     Computed Emissions Rates for Selected RAPS Days                  150

  27     HC Fractions for Each Mechanism                                  161

  28     The Demerjian/Schere Characteristic Mixing Height
         Growth Curve                                                     170

  29     Mixing Height Patterns Used to Model  Selected RAPS Days          172

  30     Assumed Ozone Aloft Used in Models of Selected RAPS Days         179

  31     Effects of Mixing Height Profile on the Contribution of Each
         Source to Final  Concentrations in MOSTM Simulations, All Days    192

  32     Ozone Maxima Predicted by MOSTM and PKSTM Using Four Mechanisms
         and Three Mixing Heights                                          216

  33     Effects of Trajectory on Ozone Maxima,  Day 238
                                                                          241
  34     Emission Rates and HC Fractions Resulting from Different
         Trajectory Assumptions, Day 238
                                                                          242
  35     Effects of Trajectory on Ozone Maxima,  Day 201
                                                                          251
  36     Emission Rates and HC Fractions Resulting From Different
         Trajectory Assumptions, Day 201
                                                                          252
  37     Random Emissions Patterns and Total Mass of Emissions
                                                                          255
  38     Predicted Ozone  Maxima Using Three Different Emissions
         Patterns
                                                                          256
  39     Deviations of Ozone Maxima Predicted from the Supplied           257
         Emission Pattern

  40   •  Effect of Ozone  Aloft on the Prediction of Ozone Maxima          259

  41     Effects of Chemistry on Constant Ozone  Aloft Survival            263

  42     Effects of Chemistry on Variable Ozone  Aloft Survival            264

  43     EKMA Solution Points for October 1, 1976 Simulations             287
                                     XV

-------
Table                                                                  Page

 44     EKMA Solution Points for June 7,  1976 Simulations               288

 45     Percent HC Reduction to Achieve 0.12 ppm CU Predicted  by Models
        for October 1, 1976 and June 7, 1976                            292

 46     Level  III Initial  Conditions and Observed Ozone Maxima          299

 47     Emission Densities for Level III  Simulations                    300

 48     NMHC Composition Carbon Fractions for Level  III Simulations     300

 49     Predicted Ozone Maxima for Level  III Simulations                303

 50     Comparison of Initial and Emissions Mass for Level II and Level III
        Simulations                                                     306

 51     Level  III  % HC Control Estimates                               315

 52     Effect of Initial  Aldehyde Fraction on 03 Maxima for Selected
        Level  III Simulations                                           32~|

 P.I    EKMA Solution Points for October 1, 1976
        (with Demerjian Mechanism)                                      339
                                    xvi

-------
                              GLOSSARY OF TERMS
EKMA - Empirical Kinetics Modeling Approach

       A procedure applied to 0, isopleth diagrams to estimate the impact of
       controlling organic and/8r NO  emissions on peak hourly 0, concentra-
       tions.  The method allows for scaling to permit use of isopleths which
       do not exactly match observed ambient conditions of maximum observed
       [OJ, and initial [HC] and [NOJ.
         O                           f\

OZIPP - Ozone Isopleth Plotting Package

       A plotting program which generates 03 isopleths from results of
       repeated simulations using a simple trajectory model utilizing the
       Dodge mechanism.

SIP - State Implementation Plan

       A report to the EPA from a state which describes what actions the
       state will take to meet National Ambient Air Quality Standards.

Mechanism - A representation of chemical processes thought to be necessary for
       Oo formation; can be complex or simple; a mechanism is necessary for
       tne estimation of the emissions control required to achieve the Oo
       standard; example mechanisms are:  Dodge, Demerjian, Carbon Bond II,
       and CIT.

BOM - Bureau of Mines database

       A 1967-68 set of auto exhaust and NO  constant light and temperature
       smog chamber experiments that were tfte basis for adjusting the Dodge
       mechanism

STM - Simple Trajectory Model

       A model  which is based on the assumption that an air parcel can be
       tracked for at least one day with simplifying assumptions of no hori-
       zontal  exchange of mass, no vertical wind shear, and no vertical
       concentration gradients within the air parcel.

MOSTM - Meteorology Only Simple Trajectory Model

       A simple  trajectory model with no simulation of chemistry.

PKSTM - Photochemical Kinetics Simple Trajectory Model

       A simple  trajectory model with the functions to simulate photochemical
       systems that produce O.
                                     xv ii

-------
RAPS - Regional Air Pollution Study

       A 1975-76 EPA study centered in St.  Louis and designed to collect data
       needed to test urban air quality simulation models.

MH - Mixing Height

       The height to which complete mixing  in the air parcel  is assumed to
       occur; the height of the air parcel  in the simple trajectory model.

LIN - Linear Mixing Height

       The mixing height obtained from linear interpolation between soundings,

HP - Hand-picked Mixing Height

       The mixing height estimated by a meterologist using  surface tempera-
       tures and soundings.

CHAR - The "Characteristic Curve" Mixing Height

       Technique developed by Demerjian and Schere (1980) for describing
       a mixing height profile.

NO  - Nitrogen oxides
  J\.

       Refers to both nitric oxide (NO) and nitrogen dioxide (NOp).

HC - Hydrocarbon

       Usually refers to non-methane hydrocarbon in this text.

ALD - Aldehyde

       Compounds containing a carbonyl group capable of photolyzing and
       producing radicals.

CO - Carbon monoxide

Level III - Refers to a simple trajectory model with a particular set of
       assumptions described in "Guideline  for Use of City-Specific EKMA in
       Preparing Ozone SIPs".  Model assumes trajectory path from mid-city
       to point of highest hour-average 0-,  concentration; requires only
       county emission inventory, general wind data, and mid-city initial
       [HC] and [NOl.
                   A

Level II - Refers to a simple trajectory model with a particular set of
       assumptions and guidelines described in  "Data  Collection for 1982
       Ozone  Implementation Plan  Submittals."    Model  requires that a back-
       trajectory from point of highest hour-average Oo concentration be
       calculated to morning position.  A gridded emission inventory with HC
       composition detail adequate for the  mechanism utilized is required.
                                    xvm

-------
       Initial conditions of [HC] and  [NO ] are needed at the morning time
       and position of the air parcel.   A comprehensive monitoring network
       is required for Level II.

DOD - Dodge Mechanism

       An explicit mechanism for n-butane and propylene used in OZIPP.

DEM - Demerjian Mechanism

       A highly compressed mechanism used in the Photochemical BOX Model.

CIT - California Institute of Technology Mechanism

       A new lumped species mechanism for use in the CIT airshed model.

CB2 - Carbon Bond II Mechanism

       A new mechanism based on bonding types used in SAI airshed model.
                                     xix

-------
                 1.0  SUMMARY, CONCLUSIONS, RECOMMENDATIONS
1.1  SUMMARY
1.1.1  Purpose
     The purpose of this study was to examine the effects of input choices
and modeling methods likely to be used in "Level  II" and "Level  III"-type
SIP applications on the predicted ozone precursor control requirements.   A
"Level II" analysis requires that a simple trajectory model incorporating
effects of pathway, emissions, mixing height changes, entrainment from aloft,
and chemistry be used to simulate days of high 03 formation.  The model  is
then used to generate an ozone isopleth diagram as a function of initial  and
emitted precursors from which precursor control requirements are estimated
by procedures such as EKMA.  A "Level III" analysis uses further simplifi-
cations of the inputs. It uses no actual trajectory information,  center city
initial conditions, and areawide averaged emissions to generate an isopleth
diagram for use with EKMA procedure.
1.1.2  Need
     Various methods for choosing the inputs to the simple trajectory model
exist; different investigators make different choices.  Two previous studies,
one dealing mainly with the chemical  mechanisms and one dealing  mainly with
the meteorological  inputs,  indicated  that the effects of these choices might
be large (Jeffries and Sexton, 1981,  Gipson, 1980).  This study  combines  both
of these studies and extends the mechanism testing to include comparison
with the entire Bureau of Mines auto-exhaust smog chamber database.   A part
of this database was used in developing the Dodge Mechanism incorporated
in the current EPA calculation tool,  OZIPP.
                                     1

-------
  1.1.3  Approach
      Four chemical mechanisms (Dodge, Demerjian, Carbon Bond II and CIT)
  were chosen for both smog chamber and ambient data comparisons (see Table 5,
  page  45  for mechanism details).  Each mechanism was used to simulate the
  Bureau of Mines smog chamber database.
     The ten RAPS days that had been selected for the previous EPA study were
then simulated with an OZIPP-like, simple trajectory model  incorporating each
of the mechanisms     for each of three choices of mixing height profiles.
These were:   1) linear interpretation of the heights found  from the three
radiosonde soundings;  2) use of surface temperatures each hour to subjectively
interpolate the soundings (called the "hand-picked" profile); and 3) calcu-
lated heights based on the "characteristic curve" of Demerjian and Schere (1980)
and the morning and afternoon soundings.  The first two of these had been
used in the previous OAQPS  study.
     Selected days were then used to generate isopleth diagrams and these
were subsequently used to perform EKMA control  calculations.
     Three of the ten  days were selected for simulation with  Level  III
assumptions  using only the characteristic curve mixing height profiles.   Iso-
pleth diagrams were generated and subsequently used to perform EKMA control
calculations.
     The basic concept for these simulations was to use methods and
procedures similar to  those that might be used by state officials in making
SIP calculations.
1.1.4  Results
     The experimental  uncertainties in the BOM database are greater than the
differences in predictions among three of the four mechanisms (Demerjian,
Dodge and Carbon Bond II), all of which have large differences in fundamental
                                     2

-------
assumptions.  The fourth mechanism (CIT) did not adequately represent the



BOM results at higher NO. but was reasonable at lower NOV levels.  The more
                        A                                X


recent, and presumably more kinetically correct mechanisms, suggest a stronger



dependance of Oo upon NOV and HC changes than the older Dodge mechanism.
               *J        X


     Thirty meteorological-only, simple trajectory model  simulations were



run:  three mixing height profiles for each of ten days from the 1976 RAPS



database.  There was no'chemistry in these simulations; they were used to



examine the reasonableness of the meteorological, emission, and boundary



condition input information used in the previous OAQPS study.  Comparison of



predicted CO, HC, and NOV to ambient values suggested that there were sub-
                        X


stantial discrepancies for some days regardless of the mixing height profile



used, and for others, poor predictions occurred for selected emissions-mixing



height combinations.   These results suggested that there  would be



non-mechanism related problems in simulating 0^ formation for as many as six



of the  days selected.



     Ninety-six simple trajectory model simulations using standard assumptions



were run:  three mechanisms (Demerjian, Dodge,  and Carbon Bond II) at each of



three mixing height profiles for each of 10 days from the 1976 RAPS database;



because of its poor results in the BOM simulations, the CIT mechanism was



used for each of three mixing height profiles on only two of the ten days.



As expected from the meteorological-only simulations many days' predictions



were poor compared to ambient data.



     The choice of  the chemical mechanism and the mixing height profile can



have a  very large effect on 03 prediction.  For the highest observed 03 day,



October 1, 1976, the choice of mechanism and input assumptions resulted in a



range of 5% overprediction to more than 70% underprediction for the linear



mixing  height profile.  With the assumptions used  in the simulations, the



Carbon  Bond  II mechanism underpredicted eight of the 10 days by 50% or more



and  underpredicted nine days by more than 25% for  all mixing height profiles.




                                      3

-------
     Seven days were underpredicted by more than 25% for all  mechanisms for
the handpicked mixing height profile.   On one day, however, it was the only
profile that produced any reasonable result.  Only two of the 96 simulations
overpredicted the observed ozone; these were both for the Demerjian mechanism
at the linear interpolated mixing height profile.
     Twenty-five additional simulations were performed with the Carbon Bond
II and Demerjian mechanisms to investigate their sensitivity to the input
assumptions.  On some days, the initial aldehyde concentration and aldehyde
fraction in the emissions greatly affect the Carbon Bond mechanism's pre-
dictions.  (These serve as radical  sources in the mechanism.)  Likewise, for
an initial  aldehyde fraction increase from 2% to 5%, the Demerjian mechanism's
Oo prediction increased by 20% for October 1, the high 0^ day.  The Demer-
jian mechanism has 0^ + olefin chemistry that provides significantly more
radicals than current theory suggests would occur.  This probably incorrect
chemistry serves as an enhanced  radical source once Oo formation starts and
therefore improves the predictions.  The Carbon Bond II mechanism uses a
presumably more kinetically correct 03 + olefin chemistry which results in
a much reduced radical source by this path.  The Carbon Bond mechanism is
therefore more dependent  upon aldehydes in  the emissions than the Demerjian
mechanism.  The aldehyde  fraction is very important for some days, not very
important on other days and its effect varies with mechanism.
     Twenty simulations using static smog chamber conditions, and 12 simple
trajectory model simulations using Oct 1 conditions demonstrated the Demer-
jian and Carbon Bond II mechanisms' response to compositional changes  in
hydrocarbon classes within the emission inventories for the  tne RAPS days.
The mechanisms are very sensitive to compositional changes in hydrocarbon
classes within the emission inventories for ten RAPS days; more than 80%

-------
 reduction in predicted On occurred for the range of most olefin to most



 paraffin composition change with no change in HC concentration.  Smog chamber



 data suggest that this response is excessive.




     An additional 45 simple trajectory model  simulations were used to



investigate possible effects of trajectory choices (21 simulations),  emission



patterns (27 simulations), and ozone aloft profiles (18 simulations).



     Simulations starting at 8 AM versus those starting at 7 AM or 6  AM



generally result in better agreement between predictions and ambient  data.



This is because the usually higher, later initial  conditions are being



inputted based on ambient data instead of being dynamically generated  by the



correct combination of early morning emissions and early morning mixing



heights.  The 6 AM start on October 1, for example, resulted in a large over-



prediction of the 8 AM NO  and HC values for the handpicked and character-
                         /\


istic curve mixing heights.




     No one mechanism nor one mixing height profile was superior at producing



the "best fit" over all the days.



     For the two best simulated days,  October  1 and June 7, 1976, twenty ozone



isopleth diagrams were generated using the simple trajectory models: for the



Demerjian, Dodge, and Carbon Bond II mechanisms for the three mixing  heights,



and for the CIT mechanism for the characteristic curve mixing height.   In



addition to the isopleth diagrams^ross section diagrams showing 0, as  a



function of HC at five constant NO  levels and 0^  as a function of NO   at
                                  X             %3                     X


five constant HC levels were produced.  These  figures show large differences



among mechanisms, mixing heights, and  days.  The Dodge isopleths are



characteristically u-shaped.  The other mechanisms exhibit more v-shaped



isopleths in which an isopleth may parallel  a  particular HC-to-NO  ratio
                                                                 /\


line.

-------
     The EKMA procedure for calculating ozone-precursor control  requirements



from isopleth diagrams was applied to the twenty isopleth diagrams  described



above.  EKilA solutions were possible for only five of the ten isopleth  dia-



grams produced for the highest 03 day, October 1.   That is,  there was no inter-



section between the HC-to-NO  ratio line and the ozone isopleth  of  interest
                            X


on five isopleth diagrams. For mechanisms other than Dodge,  the  simple



trajectory model simulation must be fairly close to the observed 0^ for an



EKMA solution to be possible.  This is because of the v-shaped nature of these



isopleths and their tendency to parallel the observed HC-to-NO  ratio line
                                                              /\


when the mechanism underpredicts.  For the Demerjian mechanism,  a small



change in input assumptions (initial aldehyde fraction decreased from 5% to



2%) moved the EKMA solution point from 1.15 times the observed initial  con-



ditions to 2.1 times the observed initial conditions.



     Control diagrams showing the percent ozone reduction as a function of



percent hydrocarbon reduction at three NO  control conditions were  generated
                                         A


for the nine EKMA solutions.  On October 1, the models predict nearly a



linear ozone change for HC change and the percent change in  03 is more than



1.5 times greater than the percent HC change.  The predicted reductions



required to achieve the 0^ standard are nearly twice as large for June 7 as



for October 1, the day of highest 0^.  This is because June  7 had 0.12 ppm



Oo aloft which resulted in 0.07 to 0.10 ppm 0-, entrained into the mixed



layer; 52 to 72% of the entrained 03 survived to add to the  modelled maximum.



October 1 had only 0.06 ppm 03 aloft; only 0.01 to 0.04 ppm of this was



added to the modelled Oo maximum.  The control diagram slopes on June 7 were



less than one.


     For four mechanisms,  three mixing profiles and two days, where a solution



was possible, the calculated Level II HC control requirements (with no change



in NO ) range from 15% reduction to 75% reduction.  The most important factor
     X

-------
in the magnitude of the reduction is the amount of entrained 0^ from aloft
relative to the amount of emissions generated CL; this ratio is day specific.
The second most important factor is the particular mechanism.  The last factor
is the mixing height profile.
      Three of the ten  RAPS  days,  October  1,  June 7,  and  July 13 were  selected
 for Level  III simulations.   Level  III  assumptions include  use of  the
 characteristic  curve mixing height  profile  but  with  a minimum 8 AM mixing
 height of 250 meters.   In addition  to  simulations with this  minimum,  each
 day was simulated with a  350 meter  minimum  for  one mechanism.  The  Demerjian
 and Carbon Bond mechanisms  were  used.   OAQPS/EPA supplied  the Dodge mechanism
 results.
      All  Level  III simulations overpredicted (by factors of  1.18  to  2.71)
 the 03 maximum  for the day.   This  is due  to  the assumptions  in the Level III
 approach:   (1)  an origin  in the  center city  has high initial  conditions;
 (2)  movement  to outer  stations is usually through higher emission areas than
 those in  Level  II trajectories;  (3) high  initial  mixing  heights increase
 the initial mass  and decrease dilution and early emission  contributions;
 and (4) average aldehyde  initial and emissions  fractions are higher than
 specific  cases  in the  Level  II approach.  Because of these differences, all
 Level  III  isopleth diagrams provide an EKMA  solution, in contrast to  Level  II
isopleth diagrams  for the  same day,  same mechanism where  an EKMA solution
 was impossible  or was  very  far out  on  each axis.
      At Level II,  the  EKMA  procedure most often required that  existing
 initial conditions and existing  emission  inventories be  scaled upward to
 obtain a  solution while,  at Level  III,  the EKMA procedure  always required
 that these precursors  be  scaled  downward  to  obtain a solution; the scale
 factors over  both levels  range from 2.10  to  0.29.

-------
o

J 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 '__

-„„. J._. „.„ <^ -1- t-m.nn-
_« , _.„ f* t __
A_— - _f\ 	 ^JL -*.

_ _„-.., j^ 	 _-fV- __l- __«


^_-T. JL- 	 „., , r\ ,,_ j_ —
— +e — i- —

| i | 1 | 1 | 1 1 1 1 1 1 1 |
• •*••••••*


a
o
•-D
Q
°






Q
°

Z>


CM
fTl
Lf)

Lf>
CM

in

cr>




O) J_
-a c o
O II
QO
II 0
Q • •!-
- = ». 42
«3 CJ i.
-r- J3  0 •(->
C. S- 4-> O
~" O) II rO
QUO JC
ii r-~ o
•~3 CM
• •« X
£00 O
fO r^ O
-C 3 CVJ
004-
O) II II
5= in  C7)
>,^- c:
•« i-
m t-
-O CD QJ
>> =1 O
II
•^ cr» • *
o LO x
-c i— o
CO •
|II >, LO CM
C II
? c ^: a)
m O O C
-> -^^2
0 "~
•r- XI t. r
T3 C CO '
QJ O Q. v
i- CQ Q.
O. =S
C
r— 0 •«
O -Q X*
s- S- ca j
•>-> rn -^-
C O
03 0 •!- J
r-
QJ
CD
                           OH %




                             8
                                                          a>

-------
     The control requirements predicted for three days, three chemical
mechanisms, and two levels of modelling sophistication are shown in Figure
117, which has been repeated here.  Summarizing Figure 117:
     The range of results is very large:  12% to 79% reduction.
     The largest difference among results is accounted for by the day
     modelled.  The highest control requirement is not for the day with
     the highest observed 0  maximum, but was for the fifth highest day of
     the ten studied.  This day had the highest 03 aloft of the ten days
     investigated.
     Within a day, choice of mechanism is the next most significant factor.
     At Level III, mechanism substitution gave a ratio of 1.8 in control
     requirements; at Level II, mechanism substitution gave a ratio of 2.2
     in control  requirements.   When it gives a solution, the Carbon Bond
     mechanism always predicts the lowest control  requirement.  The Dodge
     mechanism generally predicts the highest control requirement.
     For the same day, same mechanism, Level III control  predictions were
     larger than Level II control predictions by as much as 2.2. to 1.
The effects of mixing height profiles on control predictions for the Level  II
approach is shown in Figure 107.  At Level  II, for the same day and same
mechanism,  changing the mixing height profile can  result in control require-
ments differing  by a factor of 1.47.

-------
CD
CD
CD
en
co
CD
r-
               CD
       CD
       ID
              CD
              CO
              CD-

              C\J
                                                    CD
J l 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 '__





-- J. f\ «
- - _ -1 	 ^V J



f. . ...,Q-.^


«** «
"",1,1,1,1,1,1,1.1,1,""

Q
•"3
a_
a:
««
C_5

O
a
Q
Q
Q
t— i
a.
a:
o
z:
i — i
CT)
LO
it!
0
P-.
0)
c
3
LO
OJ
rC
Q
S-
ai
.a
0
0
o
                                                                                         -a
                                                                                          c
                                                                                          o
                                                                                         CO 4->
                                                                                            _C
                                                                                          c en
                                                                                          O •<-
                                                                                         -Q  Cn
                                                                                          II C
                                                                                         O ••-
                                                                                            X
                                                                                          en
                                                                                         •a  u
                                                                                          o •!-
                                                                                         a -i->
                                                                                          II  00
                                                                                         O T-
                                                                                          cr 4->
                                                                                          us  o
                                                                                          •r-  ra
                                                                                          •r-> i-
                                                                                          t.  T3
                                                                                          ai j^
                                                                                          CO
                                                                                             "O
                                                                                             CU
                                                                                   03
                                                                                  j^:
                                                                                   o
                                                                                   O)
                                                                                             <-)
                                                                                             •!-
                                                                                             Q-
                                                                                             c:
                                                                                             ra
                                                                                             tl
                                                                                             Q.
                                                                                          (T3
                                                                                          -a
                                                                                             OJ
                                                                                             r—
                                                                                              II
                                                                                   to
                                                                                   C
                                                                                   O
                                                                                   •r-
                                                                                   -t->
                                                                                   O

                                                                                   •5
                                                                                            -
                                                                                           CX E
                                                                                    o
                                                                                   o
                                                                                             -r-  4)
                                                                                              C r—
                                                                                              
-------
1.2  CONCLUSIONS
     The range of choices likely to be used by different investigators to
perform a "Level II" or "Level III" state implementation plan ozone control
calculation will result in substantial differences in calculated control
requirements.
     The range of choices include:  which level of model sophistication to
use; which days to model; which chemical  representation to use; which repre-
sentation of the extent of vertical mixing to use; which hydrocarbon compo-
sition to use; what assumptions to use for aldehyde emissions; what assump-
tions to use for Oo aloft profiles; and what methods to use for calculating
air parcel history and content.
     Clearly, the present EPA documents provide inadequate guidance in these
areas and it may be unreasonable to expect that EPA could provide the
answers for each individual  case.
     It is not the "modelling approach" that is at fault in this situation.
The model merely tells the user the outcome that would occur given his
supplied descriptions and inputs; it is an outstanding tool for codifying
and testing assumptions about the real world.  What is inadequate, then,  are
the descriptions of the processes involved in 0., formation and/or the input
data used in this study that were related to these processes.
     The chemical mechanisms used in this study are probably inadequate
descriptions of the real  atmospheric processes because they were not
developed by comparison with databases which included the type and range  of
variations which apparently occur in the atmosphere.  That is, none of the
mechanisms have been tested against databases in which the hydrocarbon
composition, including the aldehyde mole fraction, had been systematically
varied, and in which dynamic dilution and injection of new reactants had
                                    11

-------
occurred, and in which dynamic addition of 0^ such as from entrainment aloft
had occurred, or in which other factors similar to atmospheric processes  had
been used.  These databases do not exist, thus modellers must use the 12-year
old Bureau of Mines database.  It will  require years to develop the needed
databases to insure that chemical mechanisms are giving adequate descriptions
of the atmospheric chemical processes.
     Based on meteorology only simulations, the methods of generating meteoro-
logical inputs to the simple trajectory model used in the previous OAQPS
study and subsequently incorporated into this study are too simple to give
adequate results for a range of days.   Areas requiring work include:
better methods to determine trajectories from surface wind data; better
methods to incorporate point source emissions into the moving cell; and
better methods to characterize the mixing process.  These meteorological
problems essentially invalidate any chemical mechanism results for the days
affected.
     The trajectories used in this study are more "jagged" and show greater
regional variation than is likely in the real world.  "Blending" functions,
              2
other than 1/r , may be better at producing more consistent wind fields.
                                o
Such a function might be exp(-ar ) where "a" is an adjustable small number,
say 0.01.  More than three stations may be blended.  Additional improvement
may also result from allowing for more frequent variation in vectors (say
every 15 minutes).  The network pattern in RAPS may be inadequate in
defining trajectories over the areas needed for examining 0^ formation.
     In this study and the OAQPS study, point sources were treated like area
sources.  Agreement between ambient data and predictions could often be
improved by assuming less dispersion of point sources.  A simple point dis-
                                      12

-------
persion model applied to sources near the trajectory may significantly
improve predictions.
     Three mixing height soundings, without improved descriptions of atmos-
phericdynamic processes, are inadequate in describing the dilution
experienced by the urban atmosphere.  The characteristic curve is an improve-
ment.  However, it always gives the same rise rate regardless of atmospheric
stability.
     The results of this study are in disagreement with a fundamental
assumption of the city specific EKMA method.   The EKMA procedure is needed
when the model predictions are not in agreement with observations.   As stated
in "Guidelines for Use of City-Specific EKMA in Preparing Ozone SIPs"  (p.  4):
          ... sensitivity studies have shown that differences resulting
     from incomplete input data or  gross assumptions employed by the
     model tend to exhibit proportional impacts on the positions of
     various ozone isopleths on an isopleth diagram.  Thus, the model
     should perform satisfactorily when applied in a relative sense
     (e.g., to estimate a change in peak ozone accompanying changes in
     local precursor levels).
These sensitivity studies were conducted over a narrow range of conditions,
varying only one factor at a time, with an inappropriate mixing height pro-
file, using only one mechanism.
     This study shows that the EKMA method, as a process for correcting the
model's inadequate representation of CL formation, does not produce consistent
results.  In fact, using the EKMA procedure,  it has been UNC's experience
that the way to generate the lowest control requirement is to manipulate the
model conditions to underpredict to the largest extent that still  results  in
a possible EKMA solution.  This moves the solution point away from the
origin and gives an apparently greater effect for HC control.  These results

                                    13

-------
Imply that to be useful  the models must give better absolute  descriptions  of
the Oo formation on each day or that more reliable and  adequate methods  for
adjusting the model results when they do not describe the  03  formation must
be found.
                                    14

-------
1.3  RECOMMENDATIONS
     - Allow SIP authors substantial freedom in modelling choices to find
those that give the best description, and require extensive documentation and
demonstration that the choices made were reasonable.
     - Alternatives to EKMA  as a procedure to deal with model predictions
that are significantly different from atmospheric conditions, need to be
investigated.
     - Because of 0^ aloft assumptions, the day with the highest observed
0^ maximum may not be the day with the highest control requirement; the
validity and consequences of these assumptions should be investigated fully.
Some areas requiring further investigation are the effects of aloft profile
assumptions, differential treatment of injected Oo in chemical mecahnisms,
and interaction of mixing height profile shapes with 0-, aloft profiles.
     - Until descriptions that clearly are more adequate than those used
herein are available, further Level II and Level Ill-type analysis should
be performed with more than one chemical mechanism, and sensitivity to other
major assumptions should be investigated as a requirement in the control
calculation; in particular, at least ten days should be simulated.
     - A new auto-exhaust smog chamber study using static and dynamic
operating conditions and varying HC composition is needed to test mechanisms.
     - A better wind-field model applied to the RAPS database is needed for
trajectory determinations.
     - A submodel for the wind-field model  is needed to treat point source
emissions.
     - The characteristic curve for mixing height profiles should be made
dependent upon more factors than the morning minimum and afternoon maximum
mixing heights.
                                    15

-------
     - Alternate methods for formulating the vertical  mixing  process  in
simple models should be investigated (e.g.,  as a  dilution  rate  rather than
a mixing height rise).
     - Significant errors in some of the algorithms  used in OZIPP  need to be
fixed.  It is recommended that the code be substantially revised to  increase
its portability, flexibility, efficiency, and maintainability.
                                     16

-------
                             2.0  INTRODUCTION
2.1  THE REQUIREMENT FOR MODELLING
     The National Ambient Air Quality Standard (NAAQS)  for ozone (03)  states
that the expected number of days during a year with a maximum one-hour
average ozone concentration above 0.12 parts per million (ppm)  must be less
than or equal to one.  Because the NAAQS is widely violated,  a  significant
number of localities will have to implement control programs  to reduce
ambient ozone levels.  Therefore, methods are needed for determining the
degree of emissions reductions of the ozone precursors,  hydrocarbons and
oxides of nitrogen, necessary to achieve the NAAQS.  Because  of the complex
chemical and meteorological  factors involved in ozone production mathematical
simulation models are thought to provide the best methods for estimating the
needed reductions.
     In the November 14, 1979 Federal  Register, EPA published a "Guide to
Data Collection for 1982 State Implementation Plan Revisions" in which the
data needs were divided into four levels "based upon four generic types of
models:  (1)  photochemical  dispersion models, (2)  simplified  trajectory
models, (3)  city-specific  EKMA, and (4)  standard EKMA." These are  also
referred to as "Level I", "Level  II",  "Level  III",  and  "Level  IV"  methods.
     Also included in the November 14 Federal  Register was a  list of antici-
pated levels of modeling for major urban areas:  nine areas for Level  I,
ten areas for Level II, twenty-two areas for Level III,  and none for Level  IV.
     A major investigation of the models expected to be  used  at Level  I was
already underway at EPA.  Level  II and Level  III analyses were  expected to
                                   17

-------
be based on EPA's Ozone Isopleth Plotting Package (OZIPP)  used as a simple
trajectory model  and on EPA's Emperical  Kinetics Modeling  Approach (EKMA)
applied to 0^ isopleth diagrams.  Some assumptions,  such as the mixing
height profile, in OZIPP were somewhat arbitrary.  Some of the assumptions
underlying the EKMA procedure required further validation.  Thus, OAQPS/EPA
determined that further work was needed on these two methods.
2.2  MODELS OF INTEREST TO THIS STUDY
2.2.1  Simple Trajectory Models (STM)
     Trajectory models are based on the concept of an "air parcel".  An
"air parcel" assumes that there is some continuity of identity for a body  of
air and therefore that this body can be followed or  continuously identified
within a given region.  Field studies such as the Los Angeles  Reactive
Pollutant Program (LARPP) in which tetroons were used to "tag" air motion
have indicated that the concept can be a reasonable  representation of
measured atmospheric events.
     Simple means applying simplifying assumptions to complex  atmospheric
processes.  The usual assumptions involved are:  no  horizontal exchange of
mass; no vertical wind shear;  and rapid vertical mixing of mass, i.e. no
vertical concentration gradients.  Thus the volume of air is a well mixed
column that does not exchange material with adjacent columns.   These
simplifying assumptions allow a Lagrangian based mathematical  model of the
chemical and meteorological processes to be readily  formulated.  These sets
of equations are relatively simple to solve and the  input data requirements
appear to be likewise simple.  They therefore hold promise for providing
useful calculating tools for state and local air pollution control officials
faced with making decisions for State Implementation Plans (SIPs).
                                    18

-------
     A simple trajectory model has the following components (see Figure 1):
a) method of estimating where an air parcel  is going or where it has been
(i.e. determining a trajectory); b) a method of estimating the concentration
and physical conditions along the estimated trajectory; c) a method of esti-
mating the vertical extent of the air parcel (i.e.  the mixing height)  along
the trajectory; d) a method of estimating the amount of material  above the
mixing volume that gets mixed into the air parcel  as the mixing height
rises; e) a method of estimating the magnitude and composition of emission
sources encountered along the trajectory; f) a method for estimating the
changes caused by chemistry along the pathway.  The various methods may be
in the form of separate computer programs or manual  procedures.  In the
more complex models they may all be combined into  one computer program.
The various methods usually result in tables of values as a function of
time.  These tables then become input to a computer program designed to
simulate chemical changes under varous physical  conditions.  A major  input
to this program is a description of what chemical  reactions are thought to
take place in the urban atmosphere (a chemical mechanism).
     The set of mathematical  equations used in both EPA's OZIPP and the
PKSTM program used herein are given in Table 1.  In the programs,  the  tabular
information discussed above is converted into a continuous function of time
by fitting linear or non-linear functions to the table entries (for example,
see the emissions plot in Figure 1).   In Table 1,  those fitted functions are:
f, for the mixing height; g,  for the concentrations aloft, and h  for the
emissions; the initial species concentrations and  the functions that describe
the chemical production and chemical  loss of each  species as functions of
the other species are the other inputs.  The latter functions are  usually
                                    19

-------
                       Figure 1.

      COMPONENTS OF A SIMPLE TRAJECTORY MODEL
Methods  for  Estimating:
            -  Trajectory
              Ambient Concentrations
              Along Trajectory
              Extent of Vertical
              Mixing
              Material Above  Mixing
              Height
              Magnitude and             r'.'t
              Composition of  Emission  ll',[
              Sources                   i"!

              Changes Caused  by
              Chemistry
                                       NO   * PS
                                       NOJ   * N02
1.00 NO

1.00 03

1.00 NO2

1.00 N03

1.00 NO

2.00 NO2

1.00 N20S

1.00 N02

2.00 WHN03
                        20

-------
      Table 1.  Mathematical  Equations of A Simple Trajectory Model
      C. = the concentrations of the ith chemical  species in  the mixed
           volume at time = t

  C.(@0) 5 the initial  concentration of the ith  species  in  the  mixed
           volume at time = 0

 C-a °   = the concentration of the ith species  above the mixed volume
           at time = t
MH = f(t)


     f(t)
                              mixing height at time  =  t


                              rate of rise at time = t
   _ ,1_, dMH
   ' (m>  dt
dC.
           i/
        "
                              dilution  rate  at  time  =  t
                              decrease in species  C.  caused  by  mixing  height
                              rise
C?loft - g(t)
                              concentration  of species  aloft  at  time  =  t
dCi,
                 aloft
                              increase  in  species  C.  caused  by  entrainment  due
                              to  mixing height  rise
   = h(t)
                              mass  of species  Ci  emitted  per  unit  time at
                              time  =  t (concentration  -  height units)
dC,
                              increase  in  species  C.  caused by emissions  into
                              a  mixing  volume  with  unit  area  and height,  MH,
                              at time t
   = func (C,j=l,n)
                              production  rate  of  species  C. due  to  chemical
                              formation at  time t
   = func (C,,j=l,n)
                              loss  rate  constant  for species C. due to
                              chemical reaction at  time  t
dC.
     h
    chem
                              increase or decrease in species C.  caused by
                              chemical reactions of all  species at time t
^i,
  dt1
    TOTAL
            dC.       dC.       dC.         dC,
              ll    +   T. |     + 	L|      + 	-\
            dt  dil       ent   dt  emiss   dt  chem
 SOLUTION:
                t  dC
= C.(@0)  + /
   1        o
                      TOTAL
                              numerical  integration used to solve initial
                              value, differential equation problem
                                        21

-------
constructed by the program by internally processing a reaction list such as
that at the bottom of Figure 1.   The rate of change of each  species concen-
tration can be generated by using all  of these functions.
     The concentrations of each  species that would exist if  the model
(formulation and input data) gave proper descriptions of atmospheric events
can be found by a numerical solution of these equations.
2.2.2  Control Calculations
     Ozone Isopleth Diagrams.  Once we have a simple trajectory model  that
presumably adequately describes  atmospheric and chemical processes for a
particular set of conditions, it is possible to simulate various emission
control scenarios.  The modelling computer programs can be easily modified
to determine and display the maximum concentration outcomes  from all  possible
emission control combinations within a given range (e.g. OZIPP, the Ozone
Isopleth Plotting Package).  That is,  the trajectory model is run repeatedly
with different combinations of emissions and a response surface, for 03, is
drawn as a function of initial  and emission conditions.  That is, all  the
conditions that generate the same maximum 03 are connected together by a
line.  By comparing the coordinates of two locations on the  diagram, the
effects of various emission changes on the model's predicted 03 can be
quickly assessed.
     Applications of EKMA.  Given techniques for showing the effects of
emission changes on STM's predicted 03, how can these be translated into
atmospheric control requirements?  One procedure adopted by  EPA is Empirical
Kinetics Modelling Approach  (EKMA).  The fundamental assumption in EKMA is
that an 03-precursor response surface (an 03 isopleth diagram) exhibits
essentially constant relative shape.  That is, it is assumed that two points
                                    22

-------
            OtS'.O  ^   OOZ^O
  091'0      OEt'O      0080'0
4—h	1   H    i   i   ' +
                                                                                                  

                                                                                                  E
                                                                                                  (O

                                                                                                  O)

                                                                                                 'r-
                                                                                                 T3

                                                                                                 ^:

                                                                                                  O)

                                                                                                  Q.
                                                                                                  O  •
                                                                                                  to E
                                                                                                 •r- 03
                                                                                                     L.
                                                                                                  O) D1
                                                                                                  CD O3
                                                                                                 -a -r-
                                                                                                  O "O
                                                                                                 Q
                                                                                                  S_ 3
                                                                                                   O
                                                                                                 00 
-------
on the surface move in a constant relative relationship to each  other as  the



generating conditions of the surface are changed,  therefore,  it  is  not



necessary to have an exact STM fit on a given day  to calculate control



requirements for the conditions of that day.



     If a method is available for relating the observed ambient  high 03 con-



centration to some location on the Oo isopleth surface, then  because of the



relative relationship assumption, the change  in 03 for change in precursors



is assumed to be established.  The EKMA procedure  requires that  the location



of the starting point for control calculations on  the surface is the inter-



cept between the observed HC to NOV ratio line and the isopleth  equal to
                                  X


the observed ambient high (K.  For an ^sopleth surface developed from a STM,



this point may be above, at, or below the actual coordinates  of  the precur-



sor conditions for the day simulated.  If the point is not at the actual



coordinates, then the effect is to apply a linear  scaling factor to the



axes.    An alternative might be to apply a linear scaling factor to the



isopleth levels.



     The EKMA procedure was primarily based upon the extrapolation  of smog



chamber data to the atmosphere.  Its ability to perform under more realistic



simulations of atmospheric conditions and different assumptions  about the



chemistry needed to be examined.



2.3  PURPOSE



     Several methods for formulating the components of simple trajectory



models and for estimating the input values to these models are in current



use.  Which of these formulations and input data generation methods are



likely to give proper descriptions of atmospheric events and yet  remain



sin.pleand reasonably available to local control agencies?  Furthermore,



                                     24

-------
GIVEN:  03 Daily Design Value     = .24


        Design NMOC/NOY           = 8:1
                      y\

        Anticipated Change in NOV = -20%
                                A

        Base case diagram shown below



FIND:   Percent reduction in VOC emissions needed to reduce ozone from .24

        to 0.12 ppm



SOLUTION:
                                05  2 t   : . o   : . 2
                            BflSl  :PC
                                        ?,f. 03 R'.
                                                   - .12
STEP 1: The base case point is found by the intersection of the 8:1 NMOC/NO

        ratio line with the .24 ozone isopleth (Point 1) on the diagram.  At

        Point 1, (NMOC)i = 1.64 and (NOYh = .205
                                       A


STEP 2: The post-control NO  coordinate is calcualted as follows:
                           /\
(NOX)2 = (.205)  x (1  -
                                           = .164
STEP 3: The post-control point is located at the intersection of the .164 NO

        coordinate and the 0.12 ppm ozone isopleth (Point 2).  At Point 2,

        (NMOC)2 = 0.46



STEP 4: The VOC emission reduction is calculated as



               % reduction = (1 -    *) x 100 = 72%
Figure   3 .   Example emission reduction calculation using a single ozone

             isopleth diagram
                                     25

-------
if choice is allowed, how are the results of control  calculations likely to
differ from those using the currently standard EPA techniques?  Two previous
studies (Jeffries and Sexton, 1980; Gipson, 1980), one dealing mainly with
the chemical mechanisms and one dealing mainly with the meteorological inputs,
indicated that the effects of this  range of choices might be large.
     The purpose of this study was to examine the effects of input choices
and modeling methods likely to be used in "Level  II"-type State Implementation
Plan revision applications on the predicted ozone-precursor control require-
ments.  A "Level II" analysis requires that a simple trajectory model incor-
porating effects of pathway, emissions, mixing height changes, entrainment
from aloft, and chemistry be used to simulate days of high Oo formation.  The
model is then used to generate an ozone isopleth  diagram as a function of
initial and emitted precursors from which precursor control requirements
are estimated by procedures such as EKMA.
     In the previous studies, some key areas were identified.  Because the
meteorological processes influence the concentration which determines the
outcomes of the non-linear chemical processes, it was not possible to predict
the combined effects of the processes previously studied.  Two main areas
were singled out, however.
     Extent of Vertical Mixing.  Dilution is a significant atmospheric pro-
cess.  In STM's the model volume is moving with the wind, does not exchange
material with surrounding parcels, and therefore, dilution is caused only
by increase in the height of the moving mixing volume.
     Methods for estimating mixing heights and their temporal changes range
from the application of statistical data (Hoizworth,1972 ) when no other
information is available to the use of an acoustical echo sounding equipment
                                     26

-------
which second by second echo-ranges the depth of the mixed layer.   Different
researchers generate different temporal  profiles when given the problem of
specifying the mixed layer.   In the previous EPA study (Gipson, 1980)  several
1ikely-to-be-used methods were identified.
     The effects of these choices on photochemical  simple trajectory model
predictions and on subsequent control  calculations  was a major concern for
this study.
     Chemical  Mechanisms.  A chemical  mechanism is  a collection of the path-
ways and their associated rates for the  chemical transformations  thought (by
their developers) to be necessary to describe the chemical  behavior of air
pollutants.  In collecting these pathways and in selecting  the level of
detail  and nature of the representation, different  researchers have made
different choices.  In addition, newer information  often results  in changed
descriptions,  so if a mechanism is submitted for publication,  it  frequently
can be outdated before it actually appears.   It is  not surprising,  therefore,
that there is  more than one mechanism intended for  use in modeling urban
air pollution  chemistry.
     Four mechanisms, their sources, and history of development will be
presented in a later chapter.  The majority  of OAQPS/EPA control  calculations
have been made with the Dodge mechanism  which was published in 1977.   Even
though it has  some known kinetic deficiencies, it reportedly described the
Oq production  of the Bureau of Mines auto exhaust smog chamber quite well.
Several  newer  mechanisms are likely to be used for  SIP calculations.   The
ability of these newer mechanisms to describe the Bureau of Mines database,
as well  as their ability to perform in ambient situations was  of  interest
to EPA.
                                    27

-------
     This study,  therefore, was extended to include testing all  the mecha-



nisms by comparison with the entire Bureau of Mines auto exhaust smog cham-



ber database, rather than the limited set used by Dodge (1977).
                                    28

-------
                       3.0  BACKGROUND AND APPROACH
3.1  PREVIOUS STUDIES
3.1.1  UNC Mechanism Study
     As part of a previous study for EPA, (Jeffries and Sexton, 1980) UNC
investigated the various descriptions of nitrogen chemistry used in six
photochemical kinetics models.  This was done by implementing each model
and simulating identical situations.  The test situations were based on
three case days in the UNC outdoor smog chamber.  Since ozone production
does not depend upon absolute oxides of nitrogen concentrations, but rather
upon a ratio of N0£ to NO, mechanisms  can predict reasonable ozone peaks
while having significantly different chemistries for the nitrogen species.
Nitrogen balances therefore can illustrate important differences among
mechanisms.  The results of these comparisons for three of the mechanisms
that were tested will be described here.  The three mechanisms are:  the
Dodge mechanism currently used in OZIPP and most likely to be used in Level
II and Level  III applications, the Carbon Bond II mechanism used in the SAI
airshed grid model  and still  undergoing development on both UCR and UNC
smog chamber data;  and the Demerjian mechanism used in his Photochemical
Box Model  that has  been applied in St.  Louis and Houston simulations.  These
mechanisms will be described in more detail in Chapter 4.0.
     The comparison was conducted by modeling increasing complex hydrocarbon
systems with each model.  These were:  propylene (the August 8, 1977 experi-
ment, Figure 4), a  propylene-butane mixture (the June 19, 1978 experiment,
Figure 6),  a 13-component olefin-paraffin mixture (also June 19, 1978), and
                                    29

-------
real  auto-exhaust (the September 18,  1976  experiment,  Figure  8).  The model
predictions are reproduced here as Figures 5,  7,  and  9.   The  nitrogen
balances are given in Tables 3, 4 and 5.
     Comparison of Figure 5 with Figure  4  and  review  of  the data  of  Table  3
shows that the Carbon Bond mechanism  is  clearly much  better at  describing
the August 8 propylene experiment. The  Dodge  mechanism  incorrectly  repro-
duces the DO and PAN greatly exceeds  its measured concentration.  The
Demerjian mechanism also underpredicts the CL  and incorrectly ends with  18%
of the nitrogen as peroxynitric acid  (HCUNCL).
     For the June 19 propylene-butane  experiment,  the  plots in Figure 7 show
that the Dodge mechanism most nearly  describes  the N0? and CL profiles of
Figure 6, the experimental data.  Carbon Bond  and Demerjian are nearly
identical and both overestimate the 03.  Table  4  shows that Dodge makes  far
too much PAN and Demerjian again has  a high (14%) prediction  of peroxynitric
acid.
     In the auto-exhaust simulation of September  18,  Carbon Bond  and Demer-
jian mechanisms were again very similar in N02 and 03 performance; both,
however, underestimated the observed  0^.   The  Dodge predictions differ
obviously from the observed data.  The Carbon  Bond mechanism  predicts a
substantial quantity of aromatic nitrates; again  the  Demerjian  mechanism
predicts an incorrect, large amount (12%)  of peroxynitric acid.  Likewise,
Dodge predicts too much PAN, a characteristic  of  her  mechanism  since the
PAN chemistry is now know to be incorrect.
     A final set of simulations for the Carbon Bond II and Demerjian
mechanism was based on the UNC smoy chamber experiment of November 19,  1973,
in which the hydrocarbon composition  was drastically  altered  in the  two
                                    30

-------
                                      Table 2
             Nitrogen Balance at Ozone Maximum for Models and Data, August 8, 1977.
             Initial Conditions:  0.495 ppm NOX,  1.5 ppmC propylene.  (% nitrogen)
SPECIES
NO,
NO
WHN03a
HN03
H02N02
RN03
PAN
TOTAL
03 max,
ppm
DATA
8.2
0.4
(24.4)
39.3
-
-
27.6
100.0
*>.90
1 ) DODGE
0.00
0.00
4.60
35.84
-
-
60.04
100.00
0.584
2EBII/VT
5.42
0.07
25.75
40.15
-
3.08
25.54
100.01-
0.949
3J.DERMJ
5.24
0.11
18.22
24.07
17.77
3.. 57
30.95
99.93
0.784
                *WHN03 1s formed from
   1,
   0
   0.
   0.
 to.
 Q.
 ; o.
o
 » 0.
o
•z.
  . 0.
o
~ 0,
   0
   0
00
90
80
70
60
SO
10
30
20
10
00
I     I  I  I
I  I   >  I  I  I  I  I
 flUGUST 8, 1917

                                                                       CO
1.00
0.90
0.80 F
0.70 I
      z
0.60 "3
0.50 P
                                                                 0.30
                                                                 0.20
                                                                 0.10
                                                                 0.00
                             -o
                             z
                             •o
                             3
         5    6    7   8    9   10   11  12   13   I*   IS  16   17   18  19
                                   "HOURS,  EOT
                                      Figure  4
          Experimental data from UMC outdoor smog chamber.  Initial conditions:  NO 0.395 ppm,
          N0 0.100  ppm,  propylene 1.5 ppmC.
                                       31

-------
           HNO,. PflN.  ppm
                000
                                                   .
                                                 O o
                                                 «M
                                                 Ot*


                                                 if
                                                 Si

                                                JL
           00
                                                                                 JJ
                                                                                 CO
                                                                                  60
                                                                                 o
                                                                                 en
                                                                                 O)
                                                                                 T3
                                                                                 LO

                                                                                 Q)
                                                                                 00
—4   O   O
               uidd
           HNO,. PflN.  ppm
HNO,.  PflN.  ppm
                                                           oooooooooo
                                                                                                            -°rT
                                                                                                            i* I O
                                                                                                           Is 5
                                                                                                            qj". 3
                                                                                                            goo
                                                                                                              5u3
                                                                                                             tO LO
                                                                                                            esd
                                                                                                            O o c
               wdd
                                                                          uidd  '«o  '«ON  'ON
                                                      32

-------
                               Table  3

       Nitrogen Balance at Ozone Maximum for Models for June  19, 1978.
       Initial Conditions:  0.352 ppm NO , 2.07 ppmC propylene/butane;
       0.352 ppm NO , 2.14 ppmC UNC mix.
       SPECIES
       NO

       WHN03

       HN03

       H02N02

       RN03

       PAN
DATA
<19





S24
1 ) DODGE
16.27
0.45
3.27
45.18


35.17
2JCBII/VT
17.01
0.23
17.87
47.98

4.86
12.07
3 )DERI
9.13
0.19
20.72
38.12
14.22
2.30
15.78
       TOTAL
100.34
100.02
100.45
       03 max,

       PPm        0.435
 0.476
 0.645
 0.626
                                                                          1.00


                                                                          0.90
                          JUNE  19,   1978 -
               8    9   10   11   12    13   H   15   16   17   i8   19
                           _ HOURS,  EOT

                            Figure  6

Experimental data from UNC outdoor smog chamber.  Comparison of 25S propylene,
75% n-butane mixture ( —), with  11 component mix (	).   Initial conditions:
NO 0.261 ppm, N02 0.091 ppm. HC 2.07 ppmV ( —), HC 2.14  ppmV (	).
                                 33

-------
           HNO,. PflN. ppm
                                                  2- 3
                                                  'go *
                i-H

                 0)

                 3


                u-i
                 O

                 en
                r-H
                 0)
                ID
                 o
                                                                                       0)
                                                                                       S-i

                                                                                       60
                                                                                      •H
                             -ON
            HXOj.  PflN.  ppm
   HNO.. PflN. ppm
                                                                                     =r  cr>  CJ  *-*
                                                   S.-.
C3  O  CD  O  O  O
                                                                                                                 2
                                                                                                                w» > u a.
               udd -«Q  ''ON  'ON
O   O   O

       uidd
                                                          34

-------
   1.00
   0.80  -
 Ct
 o.
o


 N
O
   0.60
   0.
-------
           HNO,.  PflN. ppm
—•  o   o
                               o  o  o
                                                 ss
                                                  IS
                                                                                      cx
                                                                                      0)
                                                                                      I/I
                                                                                      o

                                                                                      en
                                                                                      iH
                                                                                      O>
                                                                                      T3
                                                                                      O
                                                                                      0)
                                                                                      t-l

                                                                                      oc
                                                                                      •H
               add  "o  -'ON  'ON
            HMOs.  PflN.  ppro
HNO,. PfiN.  ppm
                                                  II
                                                  t/i Q
                                                  gl
                                                  od

                                                  eTx
                                                  ae o

                                                  S^r
        CD  O  O
                                                  £.  - Q
                                                                                                               ft e 0
                                                                                                               > a
                                                                        O  O  O
               uidd
        C»   O   O   O  O

        '-o  '«ON  'ON

-------
chamber halves  with little effect on  03  production (see Figure  10).  When
this same  kind  of compositional change was simulated by the  two mechanisms
(Figure 11 and  12) large changes in Oo were predicted.  For  the Carbon Bond
mechanism, the  effects reverse and increase with increasing  concentration,
whereas for  Demerjian, the effect is  decreased with increasing  concentration.
     Considering  its overall behavior and  its more recent  development history,
the Carbon Bond II mechanism would appear  to be the more reasonably choice
for describing  the chemistry in urban simulations; it does,  however, probably
overrespond  to  hydrocarbon compositional changes.  Except  for  its  consistent
overprediction  (12% of the nitrogen)  of  peroxynitric acid, the  Demerjian
mechanism  is quite close to the Carbon Bond mechanism.  Both provide better
descriptions than the older Dodge mechanism in these outdoor smog  chamber
comparisons.
V . OU
0.15
O.HO
e 0.35
Q.
°; 0.30
° 0.25
rw
i 0.20
i 0.15
0.10
0.05
n nn
_ 1 | I 1 | ! | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 _
_ NOVEMBER 19. 1973 _
— —
— —
— —
_ _
—
— /^ ~-
r~^>>^_ :
Ci i I i ! i -1 — • — ti ! i 1 i 1 i '-I"'" 1 i
V . OU
O.H5
0.40
0.35
0.30
0.25
0.20
0.15
0.10
0.05
n nn
                            8
10   11   12   13   H   15   16   17
HOURS. EOT
  Figure  10.   Experimental data for mix, mix-toluene run.  Initial HC:
               (	) 2.28 ppmC mix, 8.8% acetylene, 64.3?, paraffins, 27.055 olefins, 0% aromatics;
               (---) 2.34 ppmC mix, 5.3S acetylene, 38.n paraffins, 16.3S olefins, 39.5£ toluene.
                                      37

-------
                  I  I   ,  I  I  I   I  I  I  ,   I  ,  j  I   I  I  I  .   I  ,  I  ,   I  ,
                                  CflRBON  BOND
                                                            2'L7% RRO
              5   6    7   8    9  10   11  12   13  If   15  16   17
                                       HOURS.  EOT
Figure 1 la.   Effects of compositional change on Carbon Bond II mechanism predictions.
               Initial HC:  3.1 pproC.
                                                                       18   19
                         1  I  '   I  '  I  '   I  '  I  '   I  '  I  '   I  '  r-r
                                  CflRBON  BOND
                                                            .•21.7% BRO   -
              5   6    7   8    9   10   11   12   13  H   15  16   17  18
                                       HOURS.  EOT
Figure lib.   Effects of compositional change on Carbon Bond II mechanism predictions.
               Initial HC:  1.5 ppmC.
                                                                            19
                                                                              0.50
                                                                              0.45
                                                                              0.40
                                                                              0.35
                                                                              0.30  P
                                                                              0.25  5
                                                                                    3
                                                                              0.20
                                                                              0.15
                                                                              0.10
                                                                              0.05
                                                                              0.00
                                          38

-------
                                     10   11   12   13  If
                                         HOURS.  EOT
                                              15   16  17   18
Figure  123.  Effects of compositional  change on Demerjian mechanism predictions.
               Initial HC:  3.1 ppmC.
         1.00
         0.90
         0.80
         0.70
      |  0.60
      Jo.50
      z  O.HO
      z  0.30
         0.20
         0.10
         0.00
     I  '  I   '  I  '   I
         I CD .
5    6
               Effects
               Initial
              1  I   '  I  '   I
              DEMERJIAN
  7    8   9   10   11  12   13   H  15   16   17   18
                   HOURS.  EOT
of compositional  change on Demerjian mechanism predictions.
HC:  1.5 ppmC.
                                                                              19
                                          39

-------
3.1.2  OAQPS/EPA OZIPP/EKMA  Study

     In early 1980 the Mr Management Technology Branch of OAQPS conducted

an evaluation of EKMA using  the  St.  Louis  RAPS database (Gipson, 1980).

The purpose of the study was  to  assess how well  the kinetics model imbedded

in OZIPP (i.e. the Dodge mechanism)  performs when atmospheric, rather than

smog chamber data, were used  for comparison.  In the actual  performance of

of the study, OZIPP was operated as  a "Level II" type model  for ten days

in St. Louis in 1976.  The ambient data came from the RAPS database.   In

effect, the study was the prototype  for Level  11-type SIP revision model

applications, and emphasis was placed on the models'ability  to predict

absolute ozone concentrations.   A large number of methods for estimating

input data had to be developed.   Severe"*  specific computer programs to manip-

ulate the extensive RAPS database were written and substantial amounts of
                 0.25 -
                 0.20
                . 0.15 -
               d
               Q
                 0.10 -
                 0.05
                 0.00
                  "6.00    0.05    0.10    0.15

                                OBSERVED 0,. ppm
                                               0.20
                                                      0.25
                                                           0.25
                                                           0.20
                                                           0.15
                                                          0.10
                                                           0.05
                                                           0.00
Figure 13.
             Scatter diagram for OAQPS OZIPP/DODGE simulation of RAPS  days.
             See Table 15 for day identification.  E=exponential , HP=  hand-
             picked, and L=linear mixing height profiles.
                                      40

-------
data processing were performed for the ten selected days.  The choices of
the researchers were tempered, however, by the fact that the methods would
have to be usable by state control agencies in performing SIP revisions.
     The same ten RAPS days and much of the data already produced became
the basis of the RAPS portion of this study.   The procedures used to gene-
rate the data and the data themselves, therefore, will  be presented in
detail  in Chapter 7.0.
     One of the factors investigated in the OAQPS study was choice of
mixing height profiles; the performance with  the OZIPP  default exponential
dilution pattern was very poor.  A linear interpolation method supposedly
similar to that in the SAI airshed model  and  values determined by a meteoro-
logist  calculating from surface temperatures and soundings were the other
methods.  The peak 0^ and subsequent EKMA control calculations were found
to be quite sensitive to the dilution pattern (see Figure 13).  A conclusion
was that a more appropriate mixing height profile method was needed and
that other sensitivity studies would need to  consider the mixing height
specification  problem.
     It was difficult to judge how much of the underprediction problem that
occurred on many days was associated with the Dodge chemistry.  Combined
with an assumption that different groups would choose different chemical
mechanisms, the underprediction problems described above suggested that
the study described herein was required to verify the effect of chemistry.
3.2  APPROACH
3.2.1  General  Concept
     Based on the previous described studies  and other  on-going developments,
several mechanisms were considered:    the Dodge mechanism, the Carbon Bond
II mechanism, the Demerjian mechanism,  the California Institute of
                                    41

-------
Technology mechanism (expected to be used in California's SIP),  and an
updated propylene-butane mechanism based on the most recent smog chamber
work.  The updated propylene-butane mechanism was deleted because of the
extensive work that would be needed to determine a new propylene-butane
carbon fraction split needed to model the BOM data.
     Because the standard EPA method (OZIPP) used the Dodge mechanism and
it had been validated against the BOM data, it was decided to compare all
the mechanisms using the BOM conditions.  Furthermore, it was decided to
use all 31 BOM runs instead of the 14 used by Dodge in the 1977  study.   In
the BOM experiments, the light intensity was constant, and the chamber
operation was static (i.e. all material  present initially, no additions  of
material during the experiment and little dilution).
     Because the information was readily available or could be determined
by slight modifications of existing programs, the same days as were being
studied in-house at OAQPS would be used for the ambient database.  In the
ambient RAPS data, the intensity and spectral distribution of the light
changed, the hydrocarbon composition changed, the majority of the reactant
was added throughout the day, and the dilution was often very large.
     The exponential mixing height rise that is standard in OZIPP was deter-
mined by OAQPS to be unrealistic and they were seeking simple alternatives
in their study.  Two methods were being examined:  a linear interpolation
of soundings similar to that in the SAI airshed model and a method in which
the mixing height was estimated each hour by surface temperatures and
soundings.  To these was added the "characteristic curve" method of
Demerjian and Schere used in the photochemical box model.
     This study, therefore, used four mechanisms, three mixing height pro-
file methods, and two databases for comparison.
                                     42

-------
3.2.2  Project Tasks
     1)   Acquire BOM data  base  and determine  assumptions  necessary  for
         modeling.
     2)   Simulate BOM experiments  with  mechanisms  selected  for study and
         compare.
     3)   Examine OZIPM program  and deduce  algorithms  - write  program for
         simple trajectory model with expanded  (and corrected, if necessary)
         algorithms  allowing  for features  and capabilities  to aid in the
         study.
     4)   Determine  for each mechanism how  to  convert  emission inventory
         to inputs  for each mechanism and  provide  documentation  of  this
         process.
     5)   Perform simulations  along the  trajectories with  each of the mechanisms
         for each of ten days,  the days  and data to be supplied  by  EPA.
     6)   Compare mechanism simulations with each other.
     7)   Provide analysis  of  the results in 6 above which may require
         additional  simulations and comparisons.
     8)   Generate isopleth diagrams for  selected days and calculate control
         requirements  for  each  mechanism for  each  day selected,  for three
         choices of  mixing height  data.
     9)   Provide analysis  of  the results in 8 above which may require
         additional  isopleth  diagrams be generated for comparison.
    10)   Generate a  report documenting the work performed and  providing
         conclusions and recommendations.
                                    43

-------
                     4.0  CHEMICAL KINETICS MECHANISMS



4.1   SELECTION



     Many mechanisms have been developed for modeling photochemical  oxi-



dant systems.   Four mechanisms were selected for this study on the basis



of the likelihood that they might be utilized in Level  II  or III  applica-



tions.



     The mechanisms selected were:



     1) Dodge  - currently used in OZIPP



     2) Carbon Bond II - used in the SAI airshed grid model



     3) Demerjian - used in the Photochemical Box model



     4) California Institute of Technology - used in Cal  Tech airshed



        model.



     Each mechanism also reflects a different approach to  representing the



photochemical  smog system.   The Dodge mechanism consists  of explicit



chemistry of propylene and butane and NO.,, and is tuned to simulate the



reactivity of urban NMHC mixtures by assuming the initial  NMHC consists of



a specific mixture of butane and propylene.  The Carbon Bond II mechanism



has  reactions  for carbon bond classes (i.e. sigma, pi, aromatic,  and



carbonyl) and an inorganic NOV section.  The Demerjian mechanism has reac-
                             A


tions for HC classes (olefin, paraffin, aromatic, aldehyde) and inorganic



NOY  chemistry.  The CIT mechanism also uses hydrocarbon classes and is
  A


generally more explicit than Demerjian, with a separate set of ethylene



reactions, and separate formaldehyde chemistry.



4.2   MECHANISM REACTIONS


     Table 5 lists the characteristics of each mechanism.   Besides the



                                    44

-------
      o
       I—
       H—4
       O
                 CM LO
                 LO OJ
                              CO
   CO



CO CO
CM r—
    O
    1C
o  o
01  o:

 E
•i—

4->
-C
 cn
                                                                                                             S-
                                                                                                             o
       Q
 co
 O
 U
 ro
 S_
 rO
x:
 CO
•r-
 E
 ro
x:
 
LO     CQ
.a
 ro
                    CO
                    r— COi
       LU
       CD
       Q
       O
       Q
                                                                                          X
                                                                                          O
                                                                                          CQ
                                                                                          ro






f^
1 —
^^^
CO CTi
CO i — O i — CO



CM ^r
0 0
rc rc
f—
[i «\ «*
^.i— CO
CO CM C_5 O
CM r— 1C 1C
CM O
o 2:
coze O
O CM 1C
•z. o »— '
rc ic
#>
« « CO
x^ o 2:
o  O)
o -o
x: o
a. E
                CO CO
                                                                     i.
                                                                     o
                                                                     CL
                                                                     ro


                                                                     S-
                                                                     
                                                                     ro
                                                                                                             QJ
                                                                                                             CL
                                                                                                             aj
O
C£
<:
r>
CD
LU C£
_J  c_> s:









CO CO
O O 0
2: ^ 2:
1C Di CO ID
o
** ** 1f^ *^
xz: o o;
o 
•r- JO
C E
rO

C
OJ
T3
C
CD
CL
O)
-o

CO
4->
E
ro
-»->
CO
C
o
o
CD
4->
rO
S-

x:
+J
•1 —
2

CO
c
0
•i —
4->
O
ro
O)
S-
E
cn
C
ro

O
QJ
E

>)
cn
O
r— .
O
C.
-C
O
Ol
1 — •
M-
O

CD
4->
3
4->

-(->
CO
C
I — t

ro
•i —
C
i-
o
CO
E
0
•r—
+J
O
ro
QJ
o;
E
rs
E

-4->
0
*->



J
J2
ro <~i o
0 E O
**v» ^) CN.
•r- +J rc



n
4->
J^
cn

i—



• •
CO i —
QJ ro
•r- +-> O
U O \
QJ +J •!—
CL
OO
CJ
•i —
S- E
O rO
•o cn
ro S-
E 0


CO
0)
•1 —
U
1)
CL
CO
1
•ZL


<*-
QJ
ja
E
rO
-E
C_)
O
co
•r-
S-
ro
Q.
E
0
U

co
-!->
E
QJ

pS!
O
O

s_ s_
QJ QJ
JD -Q
E E
3 3
E E

ro jQ
rO
<_>

t\
It—
i — i
CJ

CJ
                                                           45

-------
differences mentioned above, these mechanisms also vary widely in the
number of NO*, species represented and the number of water, temperature, and
light dependent reactions.
     Tables 6, 7, 8, and 9 list the mechanisms as actually represented in
the models.  Light dependent reactions are indicated by "HV" in the arrow
of the reaction.   Water dependent reactions are indicated with "H^O" in
parentheses before the reaction rate constant listed under the A factor
column.  Temperature dependent reactions have an activation energy value
listed under the appropriate column.
     The Dodge mechanism was obtained from the computer code for OZIPP.
The other mechanisms were obtained from the developers directly.  Instruc-
tions and guidance for mechanism use were obtained from publications and
personal communication with the developers.  Final verification was per-
formed by matching supplied example simulations.  Figures 14 and 15 are
example verification model  runs performed for the CIT mechanism.  Figure 15
shows the simulations of chamber experiment SUR-119J (supplied by McRae)
by both the CIT model and program and the UNC implemented model and program.
The numerical results were identical to 4 decimal places.  Figure 14 shows
the chamber data and the simulation performed by the CIT mechanism with the
UNC implemented model and program.  These are identical to the results sent
to us by McRae as an example of his mechanism.
     The tests clearly indicated that each mechanism had been implemented
correctly.
                                     46

-------
 E
 10
 o
 ai
 O)
 CD
T3

8

 a;
* O
 O
• O
03
                                                             47

-------
 
-------
        o   o  o   o
 cu
 3
 C
 o
 o
 


 re
 O)
 CD
T3
 O
O


 O)
JZ
VO

 CU


 ro
                                                       49

-------
                       •-<    .-i    
-------
QJ
3
£1
 C
 o
 o
 CO
 o
 Ol
••-5
 S-
 O)
 
-------
 0
 O)
 O
DQ
 c:
 o
_a
oo
 CD
-Q
 fO
                                          C   O  IT
                                                                52

-------
XJ
 01
 £1


+J
 o
 o
 1/5
•r—
 c
 o
 OJ
 o
CO
 c:
 o
 Ol
co

 QJ
 fO
t-
                                                                  53

-------
            UJ  UJ  LJ   LJ
 O)
 3
 c:
 c:
 a
 o
 C

 IO
 u
 OJ
-o
 c:
 o
CO

 c
 o
.a
 s_
 (O
o
OO


 OJ
-Q

 (t3
                                                           54

-------
                                                          .H   r4   fM   rl
 £
 to
 o
 O)
 CD
 o
 o
 CD
 O
 (LI
 rO
 e
 o
4-
r-*
fcj
 01
-C
CTl
 0)
-Q
 fC
                                                                            55

-------
                                                                       •H   (V   -<
                                                                                                               t\l    rl   r-i
T3

 

 Z3

 C
 o
 o
 en
 o
 o
 CL)
 O

 OJ
 to
 £=
 
-------
CD
co
                 ppm

                 CD
                                   CD
                                   CD
                                                            UJ
CD
CO
CD
C\J
                                   CD
CO
CD
                 CD               CD

               NUd   '2ON   'ON
                                                                     en
                                                                  QJ c:
                                                                  CD
                                                                  1= C
                                                                  •r- O
                                                                  I/) -i—
                                                                  3 •»->
                                                                     fO
                                                                  C i—
                                                                  O 3
                                                 CD-.
                                                 O  3=
                                                 r—
                                                 O  E
                                                 C  i^t

                                                 O  c:
                                                 QJ  fO

                                                    u
                                                 4-  QJ
                                                 O  E

                                                 QJ  >> E
                                                 4->  CD to
                                                 3  O T-
                                                 4-> r— C
                                                 •i-  O IB

                                                 in -C U
                                                 c:  <_> QJ
                                                 i—i  QJ S

                                                 re    QJ
                                                 .r. 4- +J
                                                 c  o ^
                                                 S-    +->
                                                 O  QJ -i-
                                                                  O -M
                                                                     l^  fO
                                                                  M- C -r-
                                                                   O i—i  C
                                                                        S-
                                                                   C H3  O
                                                                   O T- 4-
                                                                   l/l C •!-
                                                                  •i- S_ i—
                                                                   S- O  ro
                                                                   rd M- CJ
                                                                   Q.T-
                                                                   E r-  QJ
                                                                   o ra -c:
                                                                   QJ
                                                                   S-
                            57

-------
  O.HO
  0.30
           CflL TECH MECHflNISM - UNC  SIMULflTION


           CHflMBER DRTfl - EXP. SUR-119J
  0.00
                               200.

                             MINUTES
                                                             O.fO
                                                             0.30
                                                            0.20
                                                            0.10
                                                             0.00
  0.08
,0.06
d
O-
O-
o

or



P-
LLJ


H
o
  0.0a „-
  0.02 —
  0.00
                               200.

                              MINUTES
                                             300.
                                                             0.80
                                                          -  0  60
             CflL  TECH  MECHflNISM  -  UNC  SIMULflTION

             CHflMBER DflTfl  -  EXP. SUR-119J

               RRO
                                                                 -a
                                                                 •a
                                                                 3
                                                          — 0.20
   0.00
too.
   Figure  15.   Simulation of University of California-Riverside
               smog chamber experiment SUR-1190 with  the  CIT
               mechanism.."  Model results plotted  as solid lines
                           -56

-------
4.3  PHOTOLYSIS RATES
4.3.1  Introduction
     In constructing generalized, simple mechanisms different investigators
make different assumptions about the photolytic (and other) rates used.
Since the photolytic reactions are the driving force of photochemical
mechanisms, the choice of these rates can be critical;  these rate choices
are where "tuning" of mechanisms is often accomplished.
     Demerjian and Schere(1976) first provided algorithms for computing
theoretical rate constants for key photolyzing species.  These rate constants
are calculated from:  1) the theoretical actinic flux of sunlight as a
function of wavelength and zenith angle; 2)  the absorption crossection for
each specie as a function of wavelength; and 3) the quantum yield for each
specie  for each photolytic process as a function of wavelength.   The theo-
retical  actinic flux is derived from a multilayer atmospheric scattering and
absorption model with an assumed Oo aloft, aerosol  distribution,  surface
albedo,  and no clouds.  It is an expensive model to run; the output of this
model is the usual starting point for all  theoretical calculations.   Data for
items 2 and 3 are determined experimentally and are thus subject  to frequent
revision as techniques improve.
     Dodge used Demerjian and Schere's 1976  values  of photolytic  rates when
she developed her mechanism.  When the Dodge mechanism  was incorporated  into
the OZIPP program, the Demerjian and Schere  code for calculating  rate  con-
stants as a function of zenith angle, the theoretical actinic flux numbers,
the absorption coefficients, and the quantum yields were built into the
OZIPP program.  These were the 1976 values.   They can only be changed  by
changing the source code constants and recompiling  the  program.
                                     59

-------
      In 1980 Demerjian, Schere, and Peterson  published  a  set  of  new  values of
 photolytic rates based upon new experimental  values of absorption coefficients
and quantum yields.  Furthermore,  these new values were to be used with the
Demerjian mechanism and had been supplied to  Whitten (Carbon  Bond mechanism)
and McRae (CIT mechanism).
     UNC implemented new and more  efficient computer code to  compute photo-
lytic rates.   In this program,  actual  total  solar radiation measurements
can be used to adjust the theoretical  actinic values before the  rate calcu-
lation occurs and it is very simple to change absorption  coefficients and
quantum yields.
     The OZIPP source computer  code was jsed  to obtain  the values for use with
the Dodge mechanism.  Demerjian and Schere's  new values were  obtained
directly from them.
     Table 10 presents both the "new"  values  and the "old" (OZIPP) values for
the common photolyzing species  as  a function  of solar zenith  angle.   There
are some significant differences in the formaldehyde (HCHO) photolysis rates.
Both sets of values in Table 10 were used in  the simulations  performed herein.
     The application of the data in Table 10, however is  not  straightforward
because not all  mechanisms  include all  these  photolytic species  and  reactions.
The photolytic reactions of 03, HONO,  and FLO? are most often omitted.  The
photolytic reactions of aldehydes  are often combined and/or adjusted by:
1) representing the aldehydes assumed to be present or  to be  important, by
one or more general species; 2) assuming a certain composition of the alde-
hydes present; 3) assuming  a representative set of rate constants for the
composition assumed; and 4) adjusting the stoichiometry effecting the ratio
of radical products/photolyzing species and the ratio of radical products/
stable products for each photolyzing specie.
                                     60

-------
a>
       O-
       o_
       M
       O
       JO

       -o
       cu  co
       4->  CU

       Q. O10O
       ECO
       o <  •—
       o
 CO J->
 

-O ro

o 'o
    OO
-a
 c -o
 ra a>

.—. o
O O)
CO •—
CTi OJ
i— 00
       C  ra
       ro
       •t-  to
       •<-> cu

       a>  ro
       EQ;
       >>•!->
       JD >,
       cu-t->
       +J O
               -a
                cu
                a.

                •p
                c
                CD
                CU
                cu

                (O
         CO
         cu
         3

         rO
                C

                E
        nn

        o
       to
       CU
       rO
       CU


















f 	 s
O)
QJ

CU

*. 	 X

CU
'01
C
 r—
C O

r\
CM
o
"Z.

ro O
cn ro
ro i —

^
ro oo
r-~' P^
i— LO
sr LO
o cn
LO r—
ro r—

CO r—
co ro
r-» cn
LO i —


sr sr
cn oo
O ro
r~- CM



ro LO
r— CO
i — r--
00 CM



O LO
O r*^
co r-
CO CM

ro i —
st- ro
CM r—
cn ro

r— cn
cn i—
sr CM
cn ro


LO CO
r-~ sr
CO CM
cn oo



2 TO
CU r—
c o

A
o
~^-
o
re

t— LO
LO C
r— O

oo st-
ro oo
cn o
'
CM O
CO O
LO r—
r—

LO OO
r— r—
i— CM
CM


cn i—
LO ro
sr ro
CM



r— ST
o oo
!•-- sT
OvJ



CM LO
CO r-
CO LO
CM

•— ro
f— r-^
O LO
00

cn LO
r^ cr>
O LO
00


O LO
0 r-
i — LO
OO




2 "O
0) ,—
c o

*
ro
OO
O
ro r->
o o
O O
0 O

OO LO
o o
0 O

LO CO
r— CM
o o


r-, sr
sr oo
O 0



OO r—
cn LO
0 r—




LO CM
•* sr
i — CM




LO LO
cn r—
r— OO


LO OO
oo r*^
CM OO


OO i —
LO i—
CM si"



t 	 ) 	
!->. CM
CM sr





2 "O

-------
     The Dodge mechanism is an explicit mechanism in which the species repre-
sent themselves and the rate constants are generally those in Table 10 "old"
values.  For the Bureau of Mines simulations, Dodge assumed the light
intensity and spectral distribution to be constant.  The values of the
photolytic rates were in the OZIPP computer code as constants.  These were
used for the Bureau of Mines Dodge simulations.
     The copy of the Demerjian mechanism supplied to us included the rates
used by Demerjian to model the Bureau of Mines data.  In this mechanism there
is a single aldehyde,therefore some choice of rate constants is needed.
     A copy of the OZIPPM implementation of the Carbon E5ond II mechanism
and discussion of the values in Table 10 with VJhitten led UNC to assume that
essentially "old" values were used in the development of the Carbon Bond
mechanism and some assumptions about mixtures are needed.
     From the copy of the CIT mechanism supplied to UNC., it appears that the
CIT mechanism used a mixture of "old" and "new" theoretical photolysis rate
constants.  For the Bureau of Mines simulations the values supplied in the
simulation deck by McRae were used directly (except for the N02 photolysis
rate).
4.3.2  Bureau of Mines Photolytic Rates
     Table 11 summarizes the photolytic rates used in the Bureau of Mines
simulations.  There were two assumptions regarding the actual light itensity
(see later discussion, Section 5.0).  The Dodge assumption was used for all
mechanisms and in addition the Demerjian assumption was used for his
mechanism.
4.3.3  Simple Trajectory Model Photolytic Rates
     For each RAPS day that was simulated, the appropriate values (old or new)
from Table 10 were used in the UNC photolytic rate program to compute  the
                                     62

-------
theoretical photolysis rates for each day.  The total solar radiation was
multiplied by an empirical scale factor determined by Saeger (1977) to be
         1         21
0.40 min" /(cal-cirT -min~ ) for the NOp photolysis rate.  These values were
compared with the theoretically computed values and where necessary a daily
correction for cloud cover and other atmospheric factors was computed.  All
mechanisms used the NOo photolysis rate derived from the actual sunlight
measurements.  Other rates were computed from Table 10 based theoretical
rates and the daily correction factor from above.
     For the Demerjian mechanism, a 50/50 average of the HCHOC and CHoCHO
"new" values in Table 10 was multiplied by the daily correction factor to
give the ALD photolysis rate.  The MONO photolysis rates were the "new"
values in Table 10 scaled to the N0? photolysis rate.
     For the Dodge mechanism, the HCHOC "old" photolysis rates in Table 10
were multiplied by the daily correction factor.  For photolysis of HCHO  to
stable products, Table 10 and the above corrected  HCHOC  rates were used for
the HCHO photolysis to stable products.  Likewise, for the photolysis of
CH3CHO to radicals, the 1000-1300 LOT average of the "old" CH3CHO to HCHOC
rates were used for the CH-XHO photolysis rates.  The fixed rate constants
for the other aldehyde reactions were the same as  those  used for the BOM
simulations (and in OZIPP).   The other photolysis  rates  (03, H202, and HONO)
were made dependent on the "observed" N02 photolysis rate with the ratios
the same as those used for the BOM simulations.
     For the Carbon Bond II  mechanism, a 50/50 average of the HCHO0 and
CHoCHO "old" values in Table 10 was multiplied by  the daily correction factor
to give the CARB photolysis  to radicals.  The CARB photolysis to CO was
assumed to be all  HCHO -> CO  + H2 and hence the rate was  based on the "old"
valu3s of HCHOd in Table 10.
                                     63

-------
     For the CIT mechanism all  photolytic rate constants were varied with
the diurnal  N02 photolytic rate constant using the same ratios as utilized
in the BOM simulations.
             Table  11.   Photolytic  Rates  Used  in  BOM  Simulations
                                                 .,-„-!
Reaction
N02 +
HONO +
RONO ->
03 •* 01D
03 + 03P
H202->
HCHO + Rad
HCHO -> CO
CH3CHO + Rad
CARB -»• Rad
CARB -j- CO
ALD -> Rad
GL.Y -> Rad
	 n,n i
Demer Dodge
0.434 0.35
8.10-E2 0.133

9.49E-4
1.90E-2
8.40E-4
9.80E-4
3.00E-3
9.80E-4


7.90E-4

t ounomm, mill 	
Carbon
Bond CIT
0.35 0.35
6.44E-2
1.75E-2

2.52E-2
8.96E-4
1.31E-3
2.84E-3
1.09E--3
9.10E-4
1.33E-3

1.26E-2
Demer (Revised)
0.35
6.53E-2









6.37E-4

                                     64

-------
4.4  HYDROCARBON COMPOSITION INPUT CALCULATIONS
     Mechanisms differ in many ways, including  in  the species  used to  represent
the hydrocarbon present in the photochemical systems to be simulated.  Hydro-
carbon composition data usually need  to be converted to the species repre-
sentation used in each mechanism.  The actual conversion depends on the
degree of detail of the original HC composition data available, and the
characteristics of each mechanisms' HC species as defined by the mechanism
developer.  The conversion calculations used by each mechanism utilized in
this study are described in the following sections.
 4.4.1  Dodge  Mechanism
     The Dodge mechanism uses a mix of butane and propylene to represent
the hydrocarbon species present in a photochemical oxidant system.  Dodge has
determined for her mechanism that a carbon mole fraction of 75% butane and
25% propylene yielded a system which has a photochemical reactivity similar
to that in the Bureau of Mines smog chamber experiments.  The Dodge mechanism,
therefore, does not utilize HC composition data, except for the aldehyde
carbon mole fractions and aldehyde carbon numbers.
     In the Bureau of Mines experiments, initial formaldehyde concentrations
were determined for each run in the study, and the Dodge simulations of the
BOM experiments utilized these initial  conditions.  The HC input values must
be in units of ppmV, therefore the initial conditions for each HC specie are
divided by the carbon number of the specie.   The calculations are:
     Prop (ppmV)   = NMHC (ppmC)* 0.25/3.0
     But  (ppmV)   = NMHC (ppmC)* 0.75/4.0
     HCHO (ppmV)   = measured initial  HCHO
4.4.2  Demerjian  Mechanism.
     Demerjian mechanism used three HC  classes (OLE,  PAR, ARO) and one
aldehyde  specie to represent all  aldehydes (Table 12).   The Demerjian mechanism
                                     65

-------
                  TABLE 12
Mechanism Hydrocarbon Species Characteristics
Mechanism
Dodge



Demerjian



CB2




CIT





Symbol
Prop
But
HCHO
ALD2
ALD3
ALD4
OLE
PAR
ARO
ALD
OLEC
ETHC
AROC
PARC
CARB
OLE
ETH
ARO
ALK
HCHO
RCHO
HC Species
propylene
butane
formaldehyde
acetaldehyde
propionaldehyde
n-butyraldehyde
olefin
paraffin
aromatic
aldehydes
olefin double
bond (C = C)
ethyl ene
aromatic carbons
paraffin carbon
carbonyl carbon
(C=0)
olefin
ethylene
aromatic
alkane
formaldehyde
aldehydes
Characteristics
25% propylene/75% butane
represents urban
ambient hydrocarbon react-
ivity
explicit aldehydes of
propylene and butane smog
system

olefin plus ethylene, carbon
number=3
paraffin with four carbons
and higher, carbon number=6
aromatics not including
benzene, carbon number=8.1
carbon number=l .0
carbon number=2.0
carbon number=2.0
carbon number=6.0
all paraffin carbons and all
side chain; carbons in ole-
fins and aromatics carbon
number=l .0
carbon number=l .0
mole fraction weighted rate
constant and carbon number
olefin not including
ethylene

mole fraction weighted rate
rate constant and carbon
number aromatic
mole fraction weighted rate
rate constant and carbon
number paraffins

not including formaldehyde

-------
utilizes HC composition data of carbon mole fraction only.  The carbon number
for each HC and aldehyde class was defined by Demerjian.  Ethylene was
lumped into the olefin group.  The characteristics of the mechanism HC
species were based upon a generalized composition for the auto exhaust used
in the Bureau of Mines experiments.  The analysis generated by Demerjian is
shown in Table 13.
     Demerjian divided the carbon mole fraction by the carbon number in each
class in Table 13 to obtain factors for converting NMHC (in ppmC units) to
mechanism species concentrations in ppmV units.    The equations are:
     OLE (ppmV) - NMHC (ppmC)* 0.173/3.0
     PAR (ppmV) = NMHC (ppmC)* 0.417/6.0
     ARO (ppmV) - NMHC (ppmC)* 0.287/8.1
     ALD (ppmV) = NMHC (ppmC)* 0.053/1.0
4.4.3  Carbon Bond II Mechanism
     The Carbon Bond II mechanisms'  HC species  represent carbon bond types
rather than hydrocarbon classes, with ethylene  explicitly represented as the
exception.   The CARB specie represents the carbonyl  carbon present in
aldehydes,  ketones and other oxygenates (Table  12  lists  the species charac-
teristics).  Both carbon mole fraction and carbon  number are needed in HC
composition data.   Each carbon bond type  also has  a  defined carbon number
(i.e.,  olefins: C -  C has 2 carbons).   Portions  of each HC composition class
are added to the paraffin carbon bond specie, depending on the  extent  that
the HC  composition carbon numbers  are in  excess  of the corresponding carbon
bond type carbon number.   The fraction of the olefin portion (of the HC
composition data), which is in ethylene,  needs  to  be determined as well as the
non-ethylene-olefin  carbon number.   For the  BOM  simulations,  the ethylene
fraction was known.   The  non-ethylene-olefin carbon  number was  determined  to
be 3.71.  The formulas to determine  the conversion factors are:

                                     67

-------
           Table 13.   Average Bureau of Mines Exhaust Composition
                      (after Demerjian)
                                              Lumped
                              Carbon Mole     Carbon Mole     Average
Component                     Fraction        Fraction	     Carbon No.

Ethane                          .007
Propane
Acetylene                       .063
Benzene                         .035
Non-reactive 	  .105             3
Butanes                         .034
Pentanes                        .107
Hexanes & Heptanes              .088
Octanes & Higher Order          .188
Paraffins C  & Up	  .417
Ethylene                        .0720
Propylene                       .0457
Butenes                         .0419
Pentenes & Higher Order         .0137
Olefins 	  .173
Toluene                         .1131
M,P-Xylene                      .0350
0-Xylene                        .0156
Higher Order Alkyl Aromatics    .1233
Aromatics 	  .2870
                                       TOTAL    0.982
Aldehydes 	-0.053
(not measured, estimated)
                                   68

-------
     OLEC (ppmV) = NMHC (ppmC)* 0.101/3.71
     ETHC (ppmV) = NMHC (ppmC)* 0.072/2.0
     AROC (ppmV) = NMHC (ppmC)* 0.287/8
     PARC (ppmV) - NMHC *ppmC)* (0.417 +
                                (0.101/3.7  * 1.71)  +
                                (0.287/8 *  2))
     CARB (ppmV) = NMHC (ppmC)*(0.053/1.0)
4.4.4  California Institute of Technology (CIT) Mechanism
     The CIT mechanism utilizes four HC species:  olefine,  ethylene,  aromatic,
and alkane,  and two aldehyde species:   formaldehyde and a general  aldehyde
species.  (See Table 12).   Both carbon mole fraction and carbon number data
are required to implement  this mechanism.   The Demerjian HC composition
analysis reported for the  BOM study was reorganized slightly,  to include
ethane and propane into the alkane  group, and benzene into  the aromatics
group.   Ethylene was subtracted from the olefin value.   To  calculate  carbon
numbers for  each general HC class,  carbon numbers  for each  sub-classification
shown in the BOM HC analysis needed to be approximated.   The factors  used
for the CIT  mechanism to simulate the  BOM experiments are:
     OLE (ppmV) - NMHC (ppmC)* 0.101/3.83
     ETH (ppmV) = NMHC (ppmC)* 0.072/2.0
     ARO (ppmV) - NMHC (ppmC)* 0.322/7.88
     ALK (ppmV) = NMHC (ppmC)* 0.487/5.12
                                     69

-------
                     5.0  THE BUREAU OF MINES SIMULATIONS



5.1  THE BUREAU OF MINES DATABASE



     The Bureau of Mines (BOM) auto exhaust database has been  used  for 12



years to validate relationships among NO ,  HC, and 0^.   This database was
                                        A           O


generated by Basil Dimitriades during 1967-1968 at the  Bureau  of Mines



Research Center at Bartlesville, Oklahoma.



     At the time, the database was considered one of the best; it was reason-



ably complete, at concentrations close to those in the  atmosphere,  and most



importantly it used real automobile exhaust.  The study used only automobile

                                                     q

exhaust from a 1963 automobile equipped with a 283 in  engine  and used the



7-mode California cycle.  Hydrocarbon concentrations were 0.25 and  5.0 ppmC



and NO  concentrations were 0.10 to 1.4 ppm.  Thirty-one experiments  were
      A


performed at approximately constant light intensity, temperature, and



relative humidity.  The experiments were six hours in duration.



     The BOM data were obtained from Dimitriades in the form of copies of  the



hourly value printouts.  These were plotted and Figure  16 is an example of



two experiments.  These are the first replicate run pair in the data  set.



Not included in the data set were other conditions, such as, the NOp  photo-



lytic rate, spectral distribution and accuracy of measurement  systems.  The



previous Dodge study (1977) was used as a source of some of this information.



     An isopleth-like view of the 31 BOM runs is shown  in Figure 17.   The



runs are numbered in chronological order and the maximum 0., produced  is



indicated for each point.  Large numbers are used to order the seven  dupli-



cate run pairs.  A scatter diagram comparing the duplicate ozone maxima is



shown in Figure 18.  In six of the seven pairs, the first run  was substan-



tially higher.  The first six columns of Table 14 give  the BOM information



in  numerical  form.



                                     70

-------
   0.

   CL
   cn
   a_
   o

I
0
0
0
0
0
0
0
0
0
0
00 (
90
80
70
60
50
40
30
20
10
00
                                              NOVEMBER  6,  1908
   EF 0
   Q.
   0-
     0
  cr
  o_.
  o
  o
  ^
  c5
  o
00


90


80 (—


70


60


50 h-


40


30


20


10


00
                                                                   0.00
               0   '10
                       120   160  200  240  280  320  360   400

                        FINE, MINUTES
                                          ~i—i—i—i   i  |   r

                                             NOVEMBER 27,  1968
                                                      HCHO
                                      ™IH	*	^s—rte==i
                                      t  i II i-nV-_ i-A—'	'  ' ^ '  I
3 00



2 70 ^
     -<
     O

2.40 g
     o
     33
     ;U
     CO
     O
     z
     CO
2. 10


1 .80
1.50 ^
     2

1 20 5
     m

0 90 ^
0 60 ^
     "O
     3

0.30 °
                                                                  0  00
Figure 16.
        0   40   80   120  160  200  240  280  320  3GO  400

                        TIME, MINUTES


     Examples  of experimental  data  from the  Bureau  of  Mines  smog
     chamber auto exhaust  data set.   These were the first replicate
     pair of experiments  (Nos. 1  and  8).
                                    71

-------
CD
^t-
          CD
                    CD
                              CD
                              CD
                                       CD
                                       CD
                                                 CD
                                                           CD
                                                                     CD
                                                                     CDCD
                                                                        CD
      CD
      CO
                                                              -L. «*•' ' —
                                                             —T ^- ^
                                                                 + - _
                                                                        LO

                                                                        CD
                                                                        LO
                                                                        <=>
                                                                        CD
                                                                        LO

                                                                        OO

                                                                        CD
                                                                        CD
                                                                        CD   Q_
                                                                        LO   CL
                                                                             O
                                                                        CD  ^
                                                                        CD

                                                                        CM

                                                                        CD
                                                                        LO
                                                                        LO

                                                                        CD

                                                                        CD
                                                                        CD
CD
=r
          CD
                    CD
                              CD
                                        CD
                                                  CD
                                                           CD
                              C^       d^

                               LU d d   ' >•
                                                                     CDC?
                                                                     CD
                                                                                           XI

                                                                                           >4-
                                                                                           Ol
                                                                                           -a
                                                                                           
-------
     In 1976 the BOM data  were chosen as the basis for empirically adjusting
the Dodge mechanism's  propylene/n-butane fractions (the Dodge mechanism has
no aromatics) to represent an  appropriate 0,-reactivity.  Subsequently, the
Dodge mechanism and the assumptions  of reactivity derived from the BOM data
were incorporated into a computer  program for producing Oo isopleth diagrams
(OZIPP).  These isopleth diagrams  are,  in turn,  used in an EKMA procedure to
estimate HC reduction requirements for  urban areas.   The assumption is that
the urban atmosphere has the same  03-reactivity  as the BOM smog chamber.
The BOM HC composition, however, was  high in aromatics compared to current
emission inventories for urban  areas.
             0.80
0.8C
                                                           - o.io
                                                           - 0.60
                                                           - 0.50
                                                           - 0.10
                                                           - 0.30
                                                             0.20
                                                           - 0.10
             0.00
0.00
               0.00   0.10  0.20   0.30  O.fO   0.50  0.60  0.70   0.80
                      FIRST OBSERVED 0,. ppm REPORTED'X U.9
    Figure  18.  Bureau  of Mines replicate run comparison.  The run numbers
               are  identified  in Table 14.   Dashed  lines  are +  25%.
                                      73

-------
                 ooooooooooooooooocDOCDOooooocDOOocD
                 ^i^UD'^^UDVOvO^'^^tO^^'^O'-O^O'-O^O'-O'^'-O'-O^O'-D^C^'^'^D'^'^JUD
                 co ro co co co co ro co  rocorococococococncncofococor^rot^tocococococo
          CO
        O
TD
 OJ
     •r-j
      S_  X
      OJ  tJ
      E  E
      OJ
     CD   o-
        O
          en
        O
                                                                         LO    LO    IDOULOLO    LO r*.. LO CD
                                                                                  P — . — OOJOOOOOOJCD
                                                     OOOOOOOOOO
                            co "-O co LO «^- vo i—
                                            OOOOOOOOOO
                                                                                            OOOOOOOO
                  en CM CM ro i— coc^coc\)cococ\JOJCOcoc\jcococorococo<^cocococococococo
                                            OOOOOOOOOOOOOOOOOOOOOOO
                                               ! — cxj^a-roc\jc\jir)
                                                                                                                             
-------
5.2  ASSUMPTIONS
     Below is a list of key factors requiring assumption in the BOM
modelling, the problem with each factor, the assumption made by previous BOM
modellers (Dodge and Demerjian) and the choices made in this study.
     Assumptions necessary to model the Bureau of Mines smog chamber data-
base:
1)   Light Intensity
     - only pseudo-first order N09 photolysis rate
                              1
     - Dodge assumed 0.35 min   photolysis rate for N09
                                 -1
     - Demerjian assumed 0.43 min   photolysis rate for NOp
     - spectral distribution not measured assumed to be same as sunlight,
       but two different sets of quantum yields have been used   (see
       Section 4.3, Table 10)
     - this study tested both assumptions for Demerjian (see Section 5.3),
       used "old" (OZIPP) values for Dodge, Carbon Bond, and CIT supplied
       values for CIT.
2)   Temperature
     - wall  temperature measured, air temperature gradients not characterized
     - assumed to be constant at 85 F
     - this study assumed a constant temperature of 85 F in all mechanisms
3)   HC Composition
     - known to vary from test to test
     - detailed data not available during run
     - was assumed by Demerjian to be constant over range of experiments (see
       Section 4.4, Table 13)
     - Dodge "tuned" model  composition fractions of n-butane and propylene
       to obtain good fit
     - this  study accepted Demerjian's values and applied them in Carbon Bond
       and CIT (see Section 4.4)
4)   03 Concentrations
     - known to be subject to interference
     - Dodge assumed truefO,] = 0.9 [03] measured
     - Demerjian assumed true[03J = [03] measured
     - this  study accepted Dodge assumption
                                    75

-------
5)   Aldehyde Initial  Values
     - large variation in initial  measured aldehyde to measured HC ratio
     - Dodge assumed true aldehyde - measured[aldehyde]
     - Demerjian assumed true aldehyde = 0.05 initial[NHMC]
     - this study used Dodge assumption for Dodge and Demerjian's assumption
       for Demerjian,  Carbon Bond and CIT simulations
6)   Wall  Effects and  Background Reactivity
     - Dodge assumed wall was a source of propylene
     - Demerjian did not assume wall effects
     - this study used Dodge's assumption for Dodge and Demerjian's
       assumption for Demerjian, Carbon Bond, and CIT simulations
                                     76

-------
5.3  SIMULATION RESULTS
     Appendix C shows the NO, NO-, PAN, and 03 data and model predictions
for each BOM run for each mechanism.  In these plots the reported BOM 0^
values are used (i.e. the data were not adjusted) and the original Demerjian
mechanism and light intensity assumptions were used.  The simulated ozone
maximum and the time of occurrence for each mechanism for each experiment
are given in Table 14.  Figure 19 shows the scatter diagrams for 0^ max for
all four mechanisms.  (In Figure 19, the second set of assumptions was used
for the Demerjian mechanism.)  These figures assume that the BOM Oo was 90%
of what was reported.  Figure 20 shows the scatter diagrams for time to 0-
max.
     These figures show  that none of the four mechanisms is obviously
superior.  The CIT mechanism, however, had a large number of low predictions,
late times, and more scatter than the others.
     Runs 25, 27, and 28 for Carbon Bond and Demerjian were high compared to
the Dodge mechanism.  This was a direct consequence of the initial aldehyde
assumptions for the simulations.  Dodge assumed that the measured initial
aldehyde was the actual aldehyde and Demerjian assumed that the aldehydes
were proportional to the initial NMHC.  UNC used Demerjian's assumption for
Carbon Bond.   For runs 25, 27, and 28 these assumptions resulted in a 5-fold
difference in initial aldehydes (Dodge being less).   Run 25 for Carbon Bond
was resimulated with the same aldehyde as Dodge and the predicted 0, was 2.5
times lower and similar to Dodge's prediction.  These runs are at high NO
                                                                         A
concentrations and are very sensitive to radical  sources and thus the results
are more a reflection of the assumptions than of the mechanism behavior.
     At the request of Demerjian, two sets of simulations were performed
with the Demerjian mechanism.   The first set used the rate constants from

                                    77

-------
the original  mechanism (including some now known to be wrong)  and an  assumed
light intensity equivalent to an NOo photolysis rate of 0.43 min~.  The
second set used lower rate constants for reactions 11  and 12 in  Table 7 (the
peroxynitric acid formation and decomposition reactions) resulting  in less
NCL loss and thus requiring a lower NCL photolysis rate to avoid overpre-
dicting the 03 formation.  UNC assumed the same value for light intensity as
Dodge had assumed (0.35 min~ ) for this second set of Demerjian  simulations.
     The scatter diagrams  for these two sets are shown in Figure 21.   The
older rates and assumptions produced consistent overpredictions.  The newer
rates and assumptions still result in a slight overprediction.  The BOM
observed 0~ in these plots, however, ib only 90% of the reported Oo value
(using the DODGE assumption).  Therefore, if the reported 03 is  the true 03
than the Demerjian predictions would be more centered on the data set.   The
newer  rates and lower light intensity assumption were used for all
subsequent  work.
                                    78

-------
 0.80
                                           71 O.BO
                                          - 0.70
                                          - 0.60
                                            O.SO
         I  '   I  '   I  '  I   '     '


         CflRIION BUND MCC
                                                              DEMERJJflN MCCHfiNJSM

                                                              LIGHT INTENSITY - 0.3S
                                                                                                  - 0.10
0.00  0.10  0.20   0.30  0.10  0.50  0.60  0.70  0.8

         BOM CHflMBER OBSERVED Q, MflX
                                                          0.00  0.10  0.21)  0.30  0.10  0.50   0.60  0.70  0 to

                                                                   BOM CHflMBER OBSERVED 0, MflX. PPM
                                                                                                    0.00
0.80
0.70 -
      DODGE MECHflNISn

      BOH CHflMBER DRTH
0.00
                                         - 0.30
                                         - 0.20
                                         - 0.10
                                         •—' 0.00
  0.00  0.10  0.20  0 30  0.10  0.00  0.60  0.70  0 80

           BUM nllllllll.H UBiUiVLI) 0, I1IIX
                                                                                                    0.30
                                                                                                  - 0.70
                                                        0.00
                                                       0.00  0.10  0.20  0.30  C.40  0 50  0.60   0 70  O.bB

                                                                BOM CiiniillLR OBSERVED G, mX
                                                                                                   0.00
   Figure  19.   Scatter  diagrams  for mechanism  simulations  and  BOM  data,
                  ozone maxima.   Run numbers  are  identified  in  Table  14.   Replicates
                  are connected  by broken line.   Dashed  lines are +25%.   It  is
                  assumed  that the observed  0^ maximum  is 90% of the  oxidant value
                 reported.
                                                79

-------
100
350. -
CmiBON BOM) MtCllilMSM
BOM CHMMBLR L'fHfl
                                                           100.
                                                           350.  -
DEMERJIfW MECHflNISM
LIGHT INTENSITY - 0.35
BOM CHHM8ER DflTfl
    0.   50.   100.   150.   200.  250.  300.  350.  100.
         BOM CHfllffitR UOSERVED TIME 0, MflX
                                                     0.   50.   100.  150.  200.  250.  300.   350.   100.
                                                          BOM CHRMBER OBSERVED TIME 0, MflX. MIN
         DODGE MECHflNISM
         BUM CHflKBER DflTfl
                                              100.
                                     350.
                                              300.
                                              250.
                                              ISO.
                                              ICO.
                                               SO.
                                                          100. I—
    0.   SO.   101).   11)0.  200.  250.  300.  350.   100.
         BUM CHflMIIER UliSERVtn TlrE 0, mx
                                                                 1     '     '  I   '     '     •  •   '      ' /
                                                     _    CPU TECH MECHflNISM  ti  l
                                                            0. t
                                                     0,   50.   100.   150.  200.  250.  300.  350.   100.
                                                          6011 CHflimiH CRStRVED TIKE 0> I'fiX. MIN
   Figure  20.   Scatter diagrams  for mechanism  simulations and  BOM  data,
                  time to ozone  maxima.   Run  numbers  are identified  in  Table  14.
                   Replicates  are connected  by  broken  lines.   Dashed lines are  +25%.
                                                 80

-------
                                               £

                                               c5
                                               P     \-
                                                                                                                       "-  Jo
NIW 'XbW  "0 3WI1 Q31DI03ad 1300W
                                                                                               isaou
                                                                                                                        > 13

                                                                                                                          E
                                                                                                                        O •'-
   KdJ  'XHW '0 a310IC3a=) 13QOW
                                                                                            1300UI
                                                       81

-------
5.4  BOM ISOPLETH AND CROSS SECTION DIAGRAMS
     In addition to simulating the individual  BOM runs,  the mechanisms were
used to generate isopleth diagrams for the BOM conditions.   These are shown
in Figure 23.  The BOM datapoints have been superimposed on the isopleths.
     The differences among the mechanisms are  more evident in this figure.
The HC-NOX-ratio for maximum 03 formation is much higher for the CIT mechan-
ism than for the others which are reasonably similar.   This accounts for the
substantial  number of underpredictions by the  CIT mechanism.  The top
isopleth for Carbon Bond and Demerjian simulations is  0.65 ppm compared to
0.60 ppm for Dodge and CIT.
     To obtain more insight into the differences among the mechanisms, cross
    sections of the 0.,-precursor surface were  generated  at constant NO
                     •3                                                X
levels and at constant HC levels.  These are shown in  Figures 24 and 25.
These figures suggest that Carbon Bond and Demerjian mechanisms are more
sensitive to both HC and NOV change than is Dodge.  The CIT mechanism is
                           X
less sensitive to HC change and more sensitive to NO  change than the Dodge
""""- •--™"                               • '                 X
mechanism.
     In Figure 26 all four mechanisms are compared in  two sets of cross sec-
tional diagrams, one at the maximum condition  and one  at lower, more urban-
like conditions.  Also shown at the lower conditions are BOM experimental
datapoints and replicate experiments.
     At NO  above about 0.2 ppm the CIT mechanism is inconsistent with the
          X
other mechanisms and with the experimental data:  it appears to be very
inhibited by NO .  An examination of the mechanism's reactions (Table 9)
               /\
shows that one olefin molecule only gives one  aldehyde molecule (the usual
practice is two aldehydes) and that the aromatic molecules only produce a
higher aldehyde which has a lower photolysis rate than formaldehyde.  These
                                     82

-------
OOt 0    009 0   DOS 0   OOti  0    CQE 0   OQZ 0    001 0       0
                                                                              OOi 0   009 0    DOS  0   OOh 0   DOC 0    002 0   001 0      '0
     0 700   0  600   0 500   0 tOO   0  300   0 20J   0 100    0

                                NOX.PPM
                                                                                                                                                                o

                                                                                                                                                               +->
                                                                                                                                                                re
 OOi 0   009 0   OOS 0    OOh 0   OOC 0   002 0   001  0
                                                                               OOi.  0    009 0   DOS 0   OOh  0    OOE  0   002 0   001 0
     0 700   0  600   0 500   0 100   0  300   0 200   0 100    0

                                NOX  PPM
                                                                                    0  700    0 tiOO   0 jOO   0  400    0 300   0 200   0 IC?    0
O
CO
                                                                                                                                                                o
                                                                                                                                                               M-

                                                                                                                                                                

                                                                                                                                                               -i->
                                                                                                                                                                O)

                                                                                                                                                                Q_
                                                                                                                                                                o
                                                                                                                                                                01
                                                                                                                                                                c:
                                                                                                                                                                o
                                                                                                                                                                M
                                                                                                                                                               CZ3
                                                                                                                                                               OO
                                                                                                                                                               CSJ
                                                                                                                                                                OJ
                                                                                                                                                                i.
                                                                                                                                                                cn
                                                                      83

-------
001 0  D0£ 0  009'0  OOS'O OOt 0  OOC 0  002 0  00! 0     0 '
                                                                                                                                          CO
                                                                                                                                          c
                                                                                                                                          o

                                                                                                                                         4J
                                                                                                                                          0}

                                                                                                                                          3
                                                                                                                                          E
                                                                                                                                         O
                                                                                                                                         ca
                                                                                                                                          o
                                                                                                                                          o
                                                                                                                                          +J
                                                                                                                                           10
                                                                                                                                           ts>
                                                                                                                                           c
                                                                                                                                           o
                                                                                                                                           Ol
                                                                                                                                           co

                                                                                                                                           CO
                                                                                                                                           CO
                                                                                                                                           O

                                                                                                                                           O
                                                                                                                                           s_
                                                                                                                                           (O
                                                                                                                                           o
                                                                                                                                           o
                                                                                                                                           -o
                                                                                                                                            i
                                                                                                                                           O)
                                                                                                                                           c
                                                                                                                                           o
                                                                                                                                           N
                                                                                                                                           O
                                                                                                                                           CO

                                                                                                                                            0)
                                                                                                                                            3
                                                                                                                                            o>

-------
008*0   004. 0  009 0  DOS 0  OOi 0  DOC 0  OOZ 0  OOl'O
                                                                         008 0  OOL  0   009 0  DOS 0  OOfc 0  OOC 0  00"; 0  001 0      0
    0 800  0 700  0.600   0 500  0.100  0 300  0  200  0 100  0

                               03.PPM
  0 600  0 700  0 600   0  SOO  0 
                                                                                                                                                      ra

                                                                                                                                                      3
O
CO
                                                                                                                                                      c
                                                                                                                                                      o
                                                                                                                                                      o
                                                                                                                                                      D
                                                                         008 0   00' 0  009 0  OOS 0  OOi  0   OOC 0  00£ 0  001 0      0   .
    0 800   0.700  0 600  0 SOO  0 '

   ;                            03.PPM
 0 800  0 700  0 600   0 SOO  0 100  0 300  0  200   0 100  0.

|                            03.PPM
                                                                                                                                                      O
                                                                                                                                                      u
                                                                                                                                                      to
                                                                                                                                                      c
                                                                                                                                                      o
                                                                                                                                                      o
                                                                                                                                                      Ol
                                                                                                                                                      to

                                                                                                                                                      O

                                                                                                                                                      O

                                                                                                                                                       X
                                                                                                                                                     o
                                                                                                                                                     z:

                                                                                                                                                      O)
                                                                                                                                                      c:
                                                                                                                                                      o
                                                                                                                                                      IM
                                                                                                                                                     O
                                                                                                                                                     If)
                                                                                                                                                     CVJ

                                                                                                                                                     cu
                                                                                                                                                     s-
                                                                                                                                                     3
                                                                                                                                                     en
                                                                     85

-------
COS 0  OOt D  (IDt 0  005 0  00k 0  DOC 0  OOE 0 001 •     0
                                                                        1 t  001 0  004 0  DOS t  DOl'O OOC'O  002 0  001 0     0
                                                                                                                                              CO
                                                                                                                                              c:
                                                                                                                                              o
                                                                                                                                              (0


                                                                                                                                             ll
                                                                                                                                                   to
                                                                                                                                             g  -!->
                                                                                                                                             O   C
                                                                                                                                             CO  -i-


                                                                                                                                              t-   i.
                                                                                                                                              O
                                                                                                                                             •+-  r—
                                                                                                                                                   10
                                                                                                                                              I/)  4J
                                                                                                                                              C   C
                                                                                                                                              O   Ol
                                                                                                                                              <-»   i.
                                                                                                                                              O)   O)
                                                                                                                                              C/>   Q.
                                                                                                                                                   X
                                                                                                                                              CO   Q>
                                                                                                                                              CO
                                                                                                                                              O   0)
                                                                                                                                              i-   S.
                                                                                                                                              O   ns
                                                                                                                                              o   c
                                                                                                                                              N  -r-
                                                                                                                                              O  r-

                                                                                                                                                XC
                                                                                                                                             O   O)
        X
       o
                                                                               Q.
                                                                               Q.
                                                                              00
                                                                              CM
  X
o
                                                                                                                                             •a
                                                                                                                                              o
                                                                                                                                              -Q   T3
                                                                                                                                              S-   O)
                                                                                                                                              ns   -*J
                                                                                                                                              CJ   o
                                                                                                                                              O   Q)
                                                                                                                                              s-   c:
                                                                                                                                              -o   c:
                                       0 200  0 100  0
                                                                                                                200  0 100  0
                                                                                                                                                    o
                                                                                                                                                    O
                                                                                                                                              «*-    CO
                                                                                                                                              O   -
                                                                 o  -o  
                                                                •>-   to  re
                                                                 s-       o
                                                                 re   + -r-
                                                                 Q.      r—
                                                                 E   cu  o.
                                                                 O  .C  OJ
                                                                                                                                               QJ
                                                                                                                                               t-

-------
differences suggest that the initial  reactivity in the mechanism would



decrease very rapidly in the simulation.   On the other hand,  there would  not



be enough NO  loss at high HC-to-NO  ratios, so that although the 0,  may  be
            XX                                o


reasonably predicted, the other secondary products would  not  be.   Careful



adjustment of the mechanism's rates and compositional  inputs, however,  could



lead to good fits over a narrow range of  conditions such  as those shown in



Figure 15.




     In Figure 26 it can be seen that Carbon Bond and Dodge converge  at low



HC and at high NO  for medium HC.   Carbon Bond and Demerjian  converge at
                 y\


high HC and at high NOV for high HC.   Compared to the experimental  points,
                      X


however, it would seem to be difficult to say which mechanism is a better



description of BOM results.
                                    87

-------
5.5  SUMMARY



     Four chemical mechanisms, potentially useful  in photochemical  simple



trajectory models to calculate hydrocarbon and NOV control  requirements,
                                                 A


have been compared with the Bureau of Mines auto exhaust database.   The



mechanisms were:  the Dodge propylene-butane mechanism used in the  OZIPP



program; the Demerjian mechanism used in  his photochemical box model;



the Carbon Bond II mechanism (CB2) used in the SAI airshed  grid model;



the California Institute of Technology mechanism (CIT) designed for use



in their airshed grid model.



     The BOM database is the only complete and well  documented set  of



auto exhaust smog chamber experiments.  It does, however,  contain enough



uncertainties in experimental  conditions,  analytical methods,  and replicate



experiments that comparisons of models and data are not unambiguous.



That is, the "noise" in the experimental results is greater than the



differences shown among the Dodge, Demerjian, and Carbon Bond  II mechanism



simulations which have large differences in construction assumptions. At



lower NOV concentrations, this is also true for the CIT mechanism.   At
        X


higher NO  concentrations, the CIT mechanism 0-, reactivity fall s signifi-
         X                                    o


cantly below that of the other three mechanisms and the BOM chamber.  Thus,



the suitability of the CIT mechanism for control strategy calculations at



medium  to high initial NO   conditions may be questionable.   Given the range
                         X


of  uncertainty in the BOM data (e.g. 03 maxima and initial  aldehydes) and



the sensitivity of mechanisms to necessary assumptions, it is  concluded



that the Demerjian, Dodge,  and CB2 mechanisms were all capable of providing



adequate descriptions of the 03  formation in the BOM chamber;  the CB2



mechanism is probably slightly better at describing the other measured



secondary products.




                                   88

-------
              6.0  REGIONAL AIR POLLUTION STUDIES DATABASE:
                         EPA SUPPLIED INFORMATION
6.1  GENERAL DESCRIPTION
     The Regional Air Pollution Study (RAPS) was conceived early in 1970 to
provide a rational, scientific basis for the management of air quality, as
mandated by the Clean Air Act (as amended).  The basic premise of the Act
is that desired air quality standards can be obtained by setting appropriate
emission standards.  The development of Implementation Plans, called for by
the Act, assumes that existing knowledge was at least minimally adequate
for planning.
       A basic tool for the development of air quality management is the
simulation model, a mathematical  description of the complex  relationship
between emissions, atmospheric dispersion and transformation, and ambient
concentration.  The development of any model presupposes:   1) a detailed
understanding of the physical, chemical  and meteorological process involved;
and 2) availability of adequate emission data, meteorological information
and measurements of ambient concentrations of the pollutants under investi-
gation.
     At the beginning of the RAPS study, a number of simulation models had
been developed, but few -if any-  had been verified in the  field.  The pri-
mary reason for this was the absence of an adequate database, which would
contain accurate, high resolution data covering a sufficiently large area.
Ambient data were available with  adequate time resolution  -one hour or less-
but the stations providing such data were typically too few  and improperly
spaced to cover a given area.   Meteorological data were usually available
                                    89

-------
only at very few points in a given area, such as an airport, where they
were gathered for other purposes.   Micrometeorological  data related directly
to ambient measurements were generally unavailable.
     Emission inventories have been in existence for some years,  and owing
to the efforts of the National Air Data Branch of OAQPS,  were being collected
in a uniform, machine readable format known as National  Emission  Data System
(NEDS).  However, the NEDS inventory contained essentially only annual  data,
which cannot readily and reliably  be converted to hourly  values over the
two year RAPS program period.
     Clearly, what was needed as a first step in the development  of a rational
approach to the management of air  pollution was an extensive, detailed data-
base containing all  these elements:  emission, meteorological, and ambient
data, with a resolution in time and space and an accuracy adequate to pro-
vide an input to simulation models.
     The St. Louis Interstate Air  Quality Control Region  (AQCR 70) was chosen
as the site for RAPS (Figure  27).  The selection was based on the  need to
find a large city within the continental United States, which was away from
oceans and mountains and which typified the coal-burning  industrial nature
of many urban areas, yet which lay in an extended region  of rural country.
Of the 33 Standard Metropolitan Statistical Areas larger  than 400,000 popu-
lation, St. Louis emerged as  the clear choice based on the following criteria:
     - Surrounding Area
     - Heterogeneous Emission
     - Area Size
     - Pollution Control Program
     - Historical Information
     - Climate
                                     90

-------
 ircles denote radius
in km from Jefferson
Arch Memorial in down-
town St. Louis.
50
          FIGURE 27. THE REGIONAL AIR MONITORING STATIONS NETWORK
                                     91

-------
CD
CD
CD
IO

UJ

O

s:
TOTU
4330
4320
4310
4300
4290
4280
4270
4260
4250
4240
*o^n
RftMS STRTION LOCflTJONS -
~~ +22 ~~
— —
»44 ~~
+14 +15
• O
— +13 *Q +16 —
+20 +2 23 + _
+ 25 +7 +3
A"\ 0* 1 *-7 *"""""
6^* *r ^4
+ 11 +10 +17
- +19 —
»42
+ 18
« 43
+ - RflMS STflTIQN
— o - URN STRTIQN —
+24
— «••
,1,1,1.1,1,1.1,1.
I OTU
4330
4320
4310
4300
4290
4280
4270
4260
4250
4240
493 0
       690   700    710    720   730    740    750    760   770    780

                            UTM ZONE 15   1000M
        Figure  28.  The  Regional Air Monitoring Station UTM coordinates.
                                    92

-------
     In the RAPS program 25 regional  air monitoring stations (RAMS) were
placed concentrically throughout the  study region (see Figure 27).   The
stations, which were numbered 101 to  125, were thought to be located where
they would riot be unduly influenced by any one source or group of sources.
The network was operated for two years, 1975 and 1976.
     The RAPS database is maintained  by EPA in several forms.   Most readily
available for each station are hourly average concentrations for:  ozone,
nitrogen dioxide, nitric oxide,  total oxides ot nitrogen, nonmethane hydro-
carbons, sulfur dioxide, and carbon monoxide.  Each station also measured
wind speed, wind direction, and temperature; these too are available as
hourly averaged values.   The RAPS program also included an Upper Air Sounding
Network (UASN).  In this network, radiosondes were released three times per
day, five times per week, at at least two stations.  From these soundings  a
vertical temperature profile could be derived and the extent of atmospheric
mixiny could be estimated.
     Figure 28 gives the locations of the RAMS stations and the four UASN
stations on a UTM grid coordinate system (the RAMS stations have been
numbered 1 to 25, the UASN stations are numbered 41 to 44).  Station 1  was at
the Arch.  This plot will be the basis for showing the trajectories and
emission areas in Section 7.1.2.   The RAPS program also included a  new  and
hourly resolved area and point source emissions inventory for a number  of
pollutants (Littman, 1979).  Figure 29 identifies the major point sources.
The area source inventory was spatially resolved using a variable-sized grid
                                            2                           2
system in which the smallest grids were 1  km  and the largest were  100 km .
The grid system is shown in Figure 30.   A detailed section of the grid
system near the center of the city will be shown later as Figure 65.
Estimates of emission rates from both area and point source's are available
by hour and by grid for any day in 1975 and 1976.
                                     93

-------
6.2  DAYS SELECTED

     The days used in this  study were the same  days  selected  by  OAQPS/EPA  in

their evaluation of EKMA (Gipson, 1980).   These days were  selected  to  include

the days with the highest ozone levels and to include a  range of atmospheric

conditions (season, wind direction,  mixing heights,  ozone  aloft,  etc.)-   In

addition, a few days with lower ozone concentrations were  included  to  test

for performance near the 03 standard.   Table  15 identifies  the days  selected.
                       Table  15.  Model Test Days
                                      Time of  Peak 0,
Peak
Concentration,
Date
10/1/76
7/13/76
6/8/76
6/7/76
6/8/76
8/25/76
10/2/76
9/17/76
7/19/76
8/8/76
Julian Day
275
195
160
159
160
238
276
261
201
221
RAMS Site
102
114
115
122
103
115
115
118
122
125
Local Daylight Time
1500-1600
1600-1700
1700-1800
1600-1700
1400-1500
1400-1500
1700-1800
1300-1400
1300-1400
1800-1900
ppm
.24
.22
.22
.20
.19
.19
.19
.15
.15
.12
                                     94

-------
ii.iu.ii! ,!...!„!!,
5 H . ? -- I  fM^f'ill-S.si"'* !
"*"i?1«35i;-s's-!?*M3aifii;
5fSgii°E~ls
-------
96

-------
6.3  INFORMATION SUPPLIED BY EPA

     For each of the ten days during the 1976 RAPS study, EPA supplied to UNC
the data used to run the DODGE model (OZIPP)  and additional  information that
was needed for the different mechanisms.  This information included:
     a)  trajectories for each day as both UTM coordinates and maps.
     b)  the emissions along the trajectory as used in  the DODGE/OZIPP
         simulations and also in as much detail  as is  present in  the  emissions
         inventory.
     c)  the boundary conditions, both transport and aloft,used in  all  the
         DODGE simulations and the data used  to  generate  these conditions.
     d)  the ambient concentrations estimated along each  trajectory and the
         individual  data points used to estimate these  values.
     e)  the mixing  height data used in the DODGE simulation  and  the  sounding
         data used to generate this data.
     f)  example results of the DODGE/EKMA simulations  for selected days  and
         copies  of each  isopleth figure generated.
     g)  other data  needed to accomplish the  tasks.
                                    97

-------
         7.0  INPUT DATA FOR SIMPLE TRAJECTORY  MODELS  FOR  RAPS  DAYS
7.1  TRAJECTORIES
7.1.1  Methods of Determining Trajectories
     The determination of a trajectory is an implicit  input  to  a  simple  tra-
jectory model.  An accurate trajectory is important  because  ambient  data  and
emissions into the parcel are determined along  its path.   The first  problem
in determining the trajectory is to select  the  height  at which  it is to  be
determined.   In most cases the 10-m wind is chosen;  this is  the usual  obser-
vational height.  Being in the surface layer, however,  this  height may not be
the best one.
     The surface layer is generally characterized  by strong  gradients of  wind,
temperature, and humidity.  Thus a level in the higher mixed layer where
the gradients  are more uniform may describe the true winds more accurately.
Measurements of these winds can be obtained using  bistatic acoustic  sounder,
lidar, towers, balloons and aircraft.
     In most cases the 10-m wind and possibly one  or two rawinsonde  obser-
vations may be available.  From this information,  models and smoothing
techniques can give a mass consistent wind  field from  which  a trajectory  can
be determined  (Dickerson, 1978,  Goodin, McRae  and Seinfeld, 1980).   The
concept of a mass consistent wind field is  important since in a convective
boundary layer strong convergence can occur (Shreffler, 1978).  The
neglecting of  the vertical wind component  in a  trajectory  model can  produce
significant errors (Liu, Seinfeld, 1975).
                                     98

-------
     An even more simple technique of calculating trajectories is to use a
time averaged wind vector from the observing stations, weighing the closest
                                        2
stations to the point of interest by 1/r.
     If sufficient wind data are not available but there is sufficient atmo-
sphere pressure data available for a region, then a wind flow may be estimated
assuming a geostrophic balance.  In an urban area, however, the difference
in pressure measurement may not be resolvable within the accuracies of the
instrumentation.
7.1.2  Trajectories for RAPS Days
     The trajectories for the RAPS days were determined by OAOPS/EPA personnel
and were supplied to UNC on request.  The general procedure used by EPA was
to perform backward trajectory calculations from the vicinity of the RAPS
station with the high CL, starting at the time of the observed high CL.  Two
different computer programs were  used depending upon station density (Gipson,
                                                  2
1980).   In general, these programs performed a 1/r  weighting of the three
closest wind stations to generate a wind vector, then moved backwards along
this vector and repeated the process.
     Five trajectories were run for each site-day, one from the station
itself and four displaced 5 Km North, South, East, and West around the
station.   As will  be shown below, these trajectories sometimes converged,
sometimes crossed, and sometimes  diverged,  leading to a large uncertainty
in the initial  location and path  history of the air parcel.
     Given the uncertainties in the trajectories, OAQPS/EPA personnel  gene-
rated a "box" for each hour by connecting the extreme positions of the
trajectories for that hour.  Thus,  the probable trajectory was assumed to
be contained somewhere in the box.   The "box" was used to select and average
the emissions sources for that hour.  This  process will  be described in the
                                     99

-------
next section.   On some days,  a bigger "box"  was drawn  around  multiple  hours



of trajectories when little motion had taken place.



     In the following discussion,  the maps with boxes  were supplied  by



OAQPS/EPA.  They also supplied tables of trajectory  and box coordinates.



These have been plotted to more clearly demonstrate  the basis of the



calculations.
                                      100

-------
4340
4330
4320
4310
4300
4290
4280
4270
4260
 4250
 4240
            700      710     720     730      740     750     760       770      780





       Figure 31-   Trajectory for Day 159, Juno 7/Site 122
                                       101

-------
  4330
                                            1330
      730
Figure  32.
                 750

         UTM ZONE 15  1000M


Individual trajectories, Day 159.
                                                      -I ^320
                               I

                                    DRY  159 JUNE 7  1976
                                                             M
                                                             D
                                                             rn

                                                             I-*
                                                             in
  1330
  4 r) q
      40
                              /SO


                          ^QNC  il)  lOOQM
                                                       no
Figure 33.   Boxes  generated from  trajectories, Day  159.

                              102

-------
 4340
 4330
4320
4310
4300
4290
4280	/>-'f
4270
4260
4250
 4240
            700       710      720      730    740     750






        Figure  34. Trajectory for Day 160, June 8/Site 115
760     770     780
                                       103

-------
   1330
   1320
   1310
S  1300
M
  1290
                              I    'i^T1I   '    I    ^
                              DRY 160 JUNE 8  1316 SITE  115
                                                  44
                                   13
                                                       16
                         * 20
                                 12
            f7    3*  i
            I    I    I
                                    ^330
                                    1310

                                                                   n
                                    f2SO
      700     710     720     730     740     750     760     710

                        UTM  ZONE  15   1000M

 Figure 35.  Individual trajectories,  Day 160, Site  115.
                                                              1280
   1320
o
a
CD
In
UJ
a  ^300
                               |   '    | *22 '    I    '    I   >

                               DRY 160 JUKE  8  1316 SITE 115
                                                  »44
           J.
                   L
J	I  12»f
                                      r
f    I  .  r
                                                         J_
                                                              1-310
                                                              1-230
                    1-280
       700     710     720     730     "HO     750     760     110

                         urn ZONE  is  IOOOM


Figure 36.   Boxes generated from trajectories,  Day  160, Site 115.
                                   104

-------
4340
4330
4320
4310
4300
4290
 4280
 4270
 4260
 4250
 4240
             700     710     720       730     740     750      760    770      780






       Figure37-  Trajectory for Day  160, June 3/Sitc 103
                                         105

-------
   H320
   1310  -
                             DRY  160 JUNE 8 13"76 SITE  103
                                                        -  1210
   1260
700
                       720      730      710
                       UTM 7QNE 15  10QQM
750      160
                                                    4260
Figure  38.   Individual  trajectories,  Day 160,  Site 103.
   1320
   13U
   1300
-  1230
Q
h-4
   1280
   1210
                             DRY  I €0 JUNE 8 1416 SITE 103
                                                     440
                                   21
                                             14
                                                4-8
               o 42
                                                      15,
                                                     ^300  =
                                                          rz
                                                          o
                                                          -z.
                                                          rn
                                                     ^280
      700      710      720      730      7tO
                       UTM 7QNE 15  1DOOM
                                          750
         160
                                                           4260
Figure  39.   Boxes generated from trajectories,  Day 160, Site  103.
                                  106

-------
     D159 Trajectory.  Figure 31    shows the map of the derived  trajectory
path for June 7.  Figure  32   shows the coordinates of each backward trajec-
tory for each hour and the nearest RAPS stations for which the wind and
ambient data were derived.  Figure 33  shows how the "boxes" were generated
from the trajectories.  Only one box was used for the first four hours of
these trajectories.
     D160/103 and D160/115 Trajectories.  Two different sites were used on
June 8, one in the center of the city (102) and one just north of the city
(115).   Trajectory calculations  for these two sites illustrate some of the
problems with the method used.   The backward trajectories from site 115
(Figure 35  ) generally maintained about a 10 Km spread showing some narrowing
from 10 am to 9 am.  Most of the close wind stations were south of the
trajectory.  The 103 site, however, is surrounded by wind stations (Figure 38).
After three hours of back calculations these trajectories diverge dramatically,
ultimately becoming wider than  the whole city!
     Figure 40 shows both sets  of trajectories  on the same plot.
     Other Days Trajectories.   Maps,  trajectories (where supplied) and "boxes"
for other RAPS days used in this study are shown in Figures 41   to  54.
     The trajectories and "boxes" for October 1 (Figures 52  and  53)   were
not done by computer program but were done by manual  methods and detailed
examination of data printouts  (it was the first day to  be studied by OAQPS).
UNC was supplied no supporting  documentation for the shapes and locations
of the "boxes" shown in Figure   53.  These boxes were used to select
emissions sources.
                                     107

-------
                        UTM  ZONE  15   1000M
CD
co
CO
CD
CM
CO
CD


CO
CD
CD
CO
si-
cn
CM
CD
CO
C\J
CD
r-
C\J
CD
UD
CM
    cn
                                                                           CD
                                                                           UD
                                                                           r-
                                                                           CD
                                                                           LO
                                                                           C3
                                                                           CO
                                                                           CD
                                                                           CM
                                                                           CD
                                                                           CD
CD
CO
co
CD
CM
CO
CD
<—t
CO
CD
CD
CO
CD
cn
CM
CD
CO
CM
CD
r-
CM
CD
^o
OJ
                                                                               CD
                                                                               CD
                                                                               CD
                                                                         00
                                                                         O)
                                                                         -M
                                                                                    00
                                                                                    0)
                                                                               £   I
                                                                                    O)
                                                                               o
                                                                          to

                                                                     E  I
                                                                     TD   >
                                                                         "O
                                                                                    c
                                                                                    o
                                                                                    l
                                                                                    (U
                                                                                    Q.
                                                               O
                                                               ^1-

                                                               Ol
                                                               S-
                                                               3
                                                               cn
                          WOOD!    ST  JNOZ  NIP
                                        108

-------
4340
4330
4320
4310
4300
4290
4280
4270
4260
4250
4240
           700      710      720      730     740     750      760     770      780






       Figure  41.    Trajectory for Day 195, July 13/Site 114
                                           109

-------
   1310
   1300
   1250
z:
o
o
o
l/J
UJ


Q

M
   1280
   1210
   1260
   *250
                                       DflV 155 JULY 13 1376
           + 21
       730
                13*",
                                  16
740
750
760
                                            W10
                                            1300
                              ^230
                                                                     O

                                                                     rn

                                                 o
                                                 o
                                            1210
                                            *260
  1250
770
                          UTM  20NE 15  1000M
   Figure  42.  Boxes generated from trajectories,  Day 195,
                                110

-------
4340
4330
4320
4310
4300
4290
4280
4270
4260
4250
4240
            700       710     720
730
740     750       760     770
780
         Figure43-   Column Path for Day 201, July 19
                                      111

-------
     13 n
     1330
     1320
     1310
     1300




   i
   § 1290
   ui
     1280
     1210
     1250
     12tQ
           20
        720
                                     I    '     I
                                       211 JULY 11
                             122
                       13
                 19
                                18
                                          44
                         »

                         15
                               16
                             17
                                              43
730
7ft
7SO
7*0
                        DTM TONE 15  1IO«M
Figure  44.   Individual trajectories,  Day 201


                           112
                                      mo
                                      124 ft
                                                            en
                                      1281 £

                                           I
T70

-------
4340
4330
4320
4310
4300
4290
4280
4270
 4260
 4250
 4240
            700      710     720      730      740       750     760      770     780
       Figure45-  Trajectory for Day 221,  August 8/Si'te  125
                                     113

-------
UTM  ZONE  15  1000M
                                     o
                                     CD
                                     CD
                                     LiJ
                                     O
                                         CM
                                         CM
                                         CO
                                         OJ
                                         S_
                                         O
                                         O
                                         CU
                                         to
                                         3
                                         T3
                                         •r—
                                         >

                                         TD
                                         c:
                                         cu
  WOOOI   SI 3NOZ Win
             114

-------
4340
4330
4320
4310
4300
4290
 4280
 4270
 4260
 4250
 4240	~r~
           700     710       720      730    740      750      760      770     780






      Figure47.   Trajectory for Day 238, August 25/Site 115
                                         115

-------
   1310
   f300
   1290
   1280
   1270
  4260
                        44
                                  DRY 238 flUGUST 25  1976
                                            23
           + 1
         41
           18
               + 3
              10
                            17
                                                           1310
                                                           1300
                                                           1290
                                                                en
                                                           f280 g
                                                           t270
                750        760        770        780


                       UTM ZONE 15  1000M


Figure 48.   Individual  trajectories,  Day 238.
                                                         790
   1310
   1300
   12SO
iu



3  1280
   12"? 0
                           I     '     I      'I    ^

                                  DRY 238 flUGUST 25  1916
           » 14
                      _U
                                                           1300
                                                                c:



                                                                M
                                                                o
                                                           1280
      740       750        760        T70


                        UTM "ZONE 15  1000M
                                               160
ISO
Figure 49.   Boxes generated from  trajectories,  Day 238.



                                 116

-------
 4340
 4330
 4320
 4310
 4300 r
 4290
 4280
 4270
4 260
 4250
4 240

"A^w^life--
         700     710    720    730     740    750     760    770     780
      Figure 50.  Trajectory for Day 261, September 17/Site 118
                                  117

-------
  1310
  1300
   1230
z:
o
o
2  1280

UJ

Q
M
   1270
   1260
 «•
20
       720
                                  I
                                 DfiY  261  SEPTEMBER 17
                                        *14
          730
740
750
                         UTM 70NE 15  1000M


       Figure 51.  Individual trajectories,  Day 261.

                                  118
                                                         15
                                             *
                                             9
                                                    WOB
                                                    1230
                                  c


                                  M
                                  0

                                  rn
                             1280
                                                    *270
                                                    1260
760
                                                               1250

-------
4340
4330
4320
4310
4300
4290
4280
4270
4260
4250
4240
          700     710        720      730     740      750      760    770     780






         Figure 52  Trajectory for Day 275,  October I/Site 102
                                       119

-------
UTM  ZONE  15  1000M
                                               2:
                                               o
                                               o
                                               o
                                               LO


                                                  CL
                                                  -
                                                  -o
                                                  O)
                                                  s_
                                                  cu
                                                  c
                                                  0)
                                                  CD

                                                  l/l
                                                  O)
                                                  X
                                                  o
                                                  CO
                                                  ro
                                                  LO

                                                  QJ
                                                  S-
                                                  3
  WOOOT   ST 3NOZ Win
           120

-------
 4340
4330
4320
4310
4300
4290
4280
4270
 4260
 4250
 4240
             700      710     720      730    740      750       760     770     780






         Figure 54  Trajectory for Day 276,  October 2/Site  115
                                        121

-------
7.2  AMBIENT DATA ALONG TRAJECTORIES
     The ambient data alonq the trajectories  were supplied  by  OAQPS/EPA.  An
example of the ambient data printout is shown in Table  16.  The  table
information was generated as follows.   Given the midpoint  of  the trajectory
path for each hour, the program searched the RAPS database for the  three
nearest RAMS station (see the second and third columns  of  Table  16).   The
hourly averaged concentration of each species and the number  of  minutes used to
compute the average is shown in columns 4 to 7.   The hourly average in the
hour before and the hour after the calculating time were averaged to
estimate the value on the hour.  That is, to calculate  the 0900  LOT values
for 03, the 0800-0900 LOT average and the 0900-1000 LOT average  were averaged
to give the value at 0900 LOT for each of the 3 closest stations.   To
estimate the probable value at the center line of the trajectories, the
average values at the three closest stations were combined into  a weighted
average using the square of the reciprical  of the distance between  the
trajectory location and the station location (column three, Table 16). This
weighted average is shown in the last column in Table 16.
     The average values for each day were used to select the  simulation
initial conditions.  These values are given in Table 17.
     Figures 55 to 64 show the ambient concentrations along the  calculated
trajectories as a function of time.  Also included are  the temperature and
total solar radiation for each day.  Three lines are shown in each  plot.  The
                           2
solid line connects the 1/r  weighted average points.   The bottom dashed
line connects the value from the station with the lowest value of the  three
used to compute the weighted average and the upper dashed  line connects the
value from the station with the highest value of the three used  to  compute
the weighted average.  The hourly averaged values are centered (plotted on
                                    122

-------
1 UJ
a
a
1 u^
1 .


t-Ll
1 0
i
, ' >
•*
1 i i

8
0
1 1
p"
^ i
^ s '•: [
o - • o ! i
-4-1 UJ !» O O
c sc ;» <*• o
il >- ,* i

« 'I
-t-> ' i
re ! i i

1
+J -*
C 1 **7
OJ i *
'.5 ! , : *~
c ' w.
< . a


ir-
r.




- ;




0



-
r*
0


_ <
•^
*ri '°
• •



o o

0
p -^
3 , -

O pl-
^rt
o
C) "&
eno
. •:




=f U'
* -T
- (V




i C









j
i
1
1

































i
c
i







~0
t (X
. ".
• •



f r~

'"•':
'•** j
^ A

r ^

c. o
•• o




V 3
•f CC
"^ ^<
,_,



,*.




<
3



^
f*1
^


.,
.,«
^
•



o


• t

"V
>,
n
>





CT
r\
•^



a








(M
1 O
h^ '
1 1


;
~v r*- f-
.1 o i r, m
m tx . j
• • . • •
i • j
1

- o- ! '<>• r-
1
b o r- 'V.
13 o i .
. (.I, -. -0 ^,
!-» r^ l.i f

1
O '^C rx ^ ' -I
r pj
P Q *" 'V
O O '
r, ID- in \~- IT
o u. K ! '<"  0
1 (") i -n !**•
;:

^ icv - ' ro r- :.- ! r- 0
^ r- ^j ^c r- o ' < r- 3
0 0 0 t ^ 1C. 7iS
i ' | " i_ •..
| r.. r. ri
1 , j
•Ju I J

'" r < i >H '~ r
•-1 »-4 ^-( i / ' C\. -X C\J
1,1 + 4' V 4 V *
-. Iv -; fc ,; rj 1C ^ f"
; | i
*-< p r\ K c : ^ r~ r^
«f, .,, „-,,
-.!-.. r> r-. p (. iv .,
o p o r- ^ n ; f> i- o
•_ \T O 0 C- " "- -^
f» r- it cr- o p> O ( *-

! , 1 '
i

r\ ct ^ 
-------
                    Table  17



           Simulation  Initial  Conditions
DAY
159
160/103
160/115
195
201
221
238
261
275/6
275/8
276
NMHC
0.49
0.26
0.10a
0.10a
0.10a
0.10a
0.22
0.20
1.10
1.89
2.98
NO
0.0082
0.0445
0.0105
0.0027
0.0052
0.0051
0.011
0.0199
0.0265
0.1466
0.1219
NO 2
0.029
0.0121
0.0037
0.0028
0.0026
0.0047
0.0096
0.0307
0.0886
0.0553
0.3644
°3
0.034
0.011
0.062
0.059
0.056
0.016
0.010
0.008
0.005
0.005
0.005
CO
0.731
0.484
0.200
0.115
0.159
0.185
0.581
0.872
1.319
3.153
9.408
assumed minimum value
                         124

-------
the half-hour tic mark)  and  the  average  value between the hours is plotted on



the hour tic mark.   Because  of the  range of conditions for these 10 days,



three different scale factors had to  be  used for NOV, NMHC, and CO plots.
                                                  X


The 0, plots used one scale  factor.
                                    125

-------
o 20
o oo
                                     I  ' I  '  I '  I  ' J
                                      JUNC 7.  1976 —
                  rt
                                               2 00
                                               1 80
                                               1 GO
                                               1 ^0
                                               1 20
                                               1 00
                                               0 80
                                               0 60
5  6   /   8  9  10  11   1?  13  1J   15  16  I/  18  19
                   HOURS. TDT
0 20
0 00
  1.00
  0.90
  O.BO
  0.70
e 0 60
0.
^ 0.50
O
I o.io
  0 30
  0.20
  0. 10
  0.00
                          1  i  '  I '  I '  I  ' I  '  I '  I  '  I '  I '  I  ' 1  '
                                                    JUNE 7. 1976 -
                                                                      5  6
                                                                                   9  10  11  12  13  11  15  16  17  18
                                                                                         HOURS, EOT
                                                                                                                    19
1.00
0.90
0.80
0.70
0 60
0.50
0.10
0.30
0.20
0.10
0.00
                                 --]—r—|—r
                                      JIM.  1  19/6 _.
                 9  10  11   12  13  11  15  16  I/  18  19
                       HOURS. EDF
120
100
096
084
0/2
060
018
036
024

012
nnn
250 r-r i , i -T-
225
-
200 -
175
150
8 125
0 I0°
075
050
-
-
-
-
-
	
-
025
000
" , 1 , 1 ,
                                                                                  -y-ry-r-pry-T
                                                                                                   |  i |  i   ,   T
                                                                                                    JJNC 7.  19/6
                                                                      56/8
                                                                                  10  11  12  13  14  15  16  17  18
                                                                                     HOURS  TOT
                                                                                                                    19
                                                                   250
                                                                   225
                                                                   200
                                                                   175
                                                                   ISO
                                                                   125
                                                                   100
                                                                   075
                                                                   050
                                                                   025
                                                                   000
                 9  10  11  12  13  H  15  16   II  IE
                       HOURS  TDI
                                                  19
                                                     120
                                                     108
                                                     0<36
                                                     081
                                                     072
                                                     060
                                                     048
                                                     036
                                                     021
                                                     012
                                                     000
                                                                        I '  I '  I  '  I
                                                    i r r^ i  ^r
                                                    JUNE 7.  1976 _
                                                                                                   I  ,  I i  I  . I
                                                                                  10  11  12  13  11  15  16
                                                                                     HOURS, LOT
                                                                                                              17  18  19
                                                      100.
                                                       90.
                                                       80.
                                                       70.  _,
                                                           rn
                                                       60-  1
                                                           •yo
                                                       50.  |
                                                       10.  m
                                                       30.  •"
                                                       20.
                                                       10.
                                                        0.
    Figure  55.   Ambient  data,  weighted mean,  max  and  min  for Day  159  trajectory.
                                                          126

-------
2 UU
1 80
1 60
1 to
; 20
1 00
r- 80
0 C 0
0 to
0 ^5
0 00
-r-n-r-n i-m

-

-
;-
-


' 1 ' 1 '














*
;'Vxi

5 6 7 8 9 10 11 12 13 1

HOURS r:
JUN






. * ,-^-<

•4 l^ [
: a








6 1
' 1 '_
1976 _
_
-
-

-i
•/\ "

-
L '_!_•
7 18 1
1 80 0 45
1 60 o 40
! "0 o 35
1 20 " r, 30
1 00 0 2r
0 80 £ Q 20
0 GO o IS
0 to Q ,0
0 20 o 05
0 00 o 00
9
_'' 1 ' I ' 1 '
-
-
-
-
-
•-
-
i 1 , 1 , 1 ,
5 6 ; a







•-»-•
,







-"

9101
1





P
S/
i



^


A

,
• , | 'I'






\^
, i
JUM






.--?-*
,
— ] —
L 8






•~~»-J
,
1 1 '
1976 _
—
—
-
-
-
—
,-Z-i -
, 1 ,
1 12 13 la 15 16 17 18 1
r HOURS
U 3D
0 
-------




CJ
F
£
O
i_)









0.
0.
o
^









c.
r.
0



2 00 r-r p- | r
1 80 j-
1 60
1 ^ 0
1 20
i 00
0 BO
0 60
0 d 0
0 20
0 00
-
-
-
-
7 /

-
_
-, i , T ;
567


10«
0%
OB'l
072
000
0J8
030
024
012
000

-
-

—
-
-
"
*/*
",!,*«
5 6 7
J20 -r-, _,--,
,03 -
096 —
OB4 —
L
o;2 L
rjeo
r -5
03C
Ci2
-
>

	

| -, T-T-J - r-[ -r-T-r-pr- [ -i-pr-pr -pr-pr-
JUNf 8 197G -
"A"
/A
/ \ ~
/ A'
\ A
Xy A
• V-^-\^:^-'
-
i , i , i . i , i , i , i . i , i , , j.. , "
8 9 10 11 12 13 1« 15 16 17 18 1
2 00
1 80
1 00
1 "0
1 20
1 00
0 80
0 00
0 «0
0 20
0 00
9
HOURS. EDr

JUNE B, 1970 _
-
—
-
-
-
-
"*x«
* i PTf-*-»-*-?-»-<-4— *-+-<-i , 1 _j 1 , I ,
8 9 10 11 12 13 11 IS 16 17 18 1
HOURS EOT
~f n j l | l |~T— ] i p i | , | , | i | , | ,
JUNC 8 1976 _
-
-
-
-
\ ;
V "-
\^A < o-^V
* \ fi~~~-/
* . 1 , 1 . 1 V 1 , l T 1 ' 1 , 1 . 1 .
100
096
08'<
072
000
018
036
024
012
000
9
120
108
096
081
072
060
O-'B
030
024
012
nnn
r 1 r1 i ' i ~r i r~i T T i ' r n i ' ' ' ' '
0 90
0 80
0 70
0 60
0 50

0 "0
0 30

0 20
0 10
n nn
JUNE 8. 1976 _
r ~

_.
- —
A t -
l\* f
- ^J Y \** - J
m^^^* * -i \ / \ H *,&—£ -i
V"; ^\/ \
l . - V . ' . j
• , , , i , j . i , i , , i , i , i . i , . , '
i UU
0 90
0 80
0 70
0 60
0 50

0 ^0
0 30

0 20
0 10
n nn
                                                                       10  11  12 13 11 IS  16  17  18
                                                                          HOURS. EDr
                                                                                                 19
, £ -J V
225
200
175
150
125
ill
075
050
025
-i | , | , | , | , | , f , | , | , | , | , | , | , | , _
JUNE 8. 1976 _
7" f ~
' sJ-V"
t-^^i -
1 T
I ' ~
JT 1.
S -
~ * ^S ~-
nnn , 1 , T ' 1 , ! • 1 , .1.1 	 I.I,
dO 
-------
                                                                                            J'JL'
                                     -,-, 0 SO


                                    76 J 0 15


                                      J 0 to


                                      -| 0 3S


                                      - o «
                                                                                                        I
OHO
                                                           250 r,-T


                                                           225 -
                                                           200 h
                                                           100
                           FT' r
                           JJLY 13. 1976
                                                                            10 11  12  13  H  15  16  17  18

                                                                               HOURS FCT
                                                                                                       19
250


225


200


175


150


125


100


0/5


050


025


000
                                         18  19
                                                               n '  i '  i '  i '  i '  i '  i ' i  ' i  ' i  '  i '  i
                    HOURS ft) I
6  1  8  9 10 U  12  13  11  15  16  17  18

              HOURS. LOT
                                                                                                       19
  Figure  58.   Ambient data, weighted  mean, max and min for Day 195.


                                                   129

-------
\J »J V
0 IS
o to
0 35
't 0 30
Q.
D.
. 0 25
f 0 20
0 15
01 A
1 U
0 05
n on
_ ' | i | i | i
-
-
i-
—
-
-
— y
-^^»
1
",1,1,1,
' TT I ' I '





V-
Kj
« ^

,1,1,1,
1 i '

< X
A.
/
/




, ! .
1 1 '
ULY 19.








. 1 i
!
197








p
— !
6 _
-
-
—
-
-
-
~
	
I ,
'J J\J
0.45
o to
0 35
0 30
0 25
0 20
0 15
01 n
i U
0 Q5
n nn
                                                           5   6   7   8   9  10  H  12  13  1 '<  15  16 17 18
                                                                            HOURS.
                                                                                                   19
015

036
032
028
f 024
" 020
r?
^ 016
012
003
004
000


	
1 '
—
-
-
-
-

—
-

" , 1 ,
5 6


-r-








\-^^
,
7


-r | , | , | . [ r-p- -p-r-|-r-| p-r-y-r-j
JUL' 19 1976 __
-
-
-
_

• - A J
< fl! \
>-«-/ ^k, -
. 1 , 1 , 1 , I . I , I , I , I , J , | '
B 9 10 11 12 13 1" 15 16 17 18 1
HOU^S. EDT
nun ->c n
y T u . 6 j j
03b 225
032 200
028 175
024 150
020 | 125

016 0: 100
012 .075
008 050
004 025

9



-
-
-
-
-

-
~-
'
L



1







t


G


— i —







-A-,


7


i -| , )-, | ,






*•/
--A' r
-AA/
. '

a 9 10 n i
HOURS

" ' 1 '



^

/



, 1 ,
2 13 1
, CDT

1 ^ — j — i — '
JULY








• ! .
•* 15 i


!
19








,
6 1


' 1 '.
1976 _
-
-
_
	 ,

-
:
_
, i , "
7 18 1


250
225
200
175
ISO
125

100
075
050
025

000
9

5  6  ;  8  9 10 11  12  13  1'  15  1G  \1  18
                                                     0 o
                                                     £ o

                                                             I  ' 1  ' I  ' I '  1 •' I '  I '  I '  I '  I '  1 '  I
                1  r1  i '
                  JULY IS.  1976 _
3 10 11  12  13  14  15  16  17  18  19
     HOURS. LOT
100.
 90.
 80.
 70. -H
    rn
 60- I
    33
 50. |
 40. S

 30. ""
 20.
 ID.
  0.
Figure  59.   Ambient data,  weighted  mean, max and min for Day  201.

                                                130

-------
2.00
1 00
1 60
1 -»0
1 2')
1 00
0 SO
0 CO
0 J0
0 21

0 "i '.'
036
032
028
02"
Ir'O
110
112
,,08
;')'



136
132
VO
'124
"20
.116
"12
008
>JOJ
000
-i y-r-pT- p-| -r-rrT-T^rTT-T-r-r-T--p-r-r-r-r-r-
_ flUGUST 8 19/6 _
--
-
"..
_
_
- ' - * ~
""•X
y ~^^~^'~A"-« -*-< * * ^^^t"^'^^ *
I 1,1 l,t, ",_],!,, j 	 !__, _[ , 1 , 1 .
S 6 7 8 9 10 i; 12 13 !•' 15 15 17 18 1
FlUGUST 8, 1976 _
-
_
-

-
_
D
" , 1 , 1 , 1 . 1 , 1 , 1 , ! , 1 , 1 . 1 , 1 , 1 , 1 .
5 G / 8 110 11 12 131' 15 16 17 181
H3UP1 for
. ' 1 ' ' 1 ' ! ' i ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' I ' .
_ flUGUSF 8 1976 _
-
_
-
-
" « A < " ' ~~
X * * ft
** * ' A^
J* 	 ^-f/ ^
^T^^^^^'t^^TT^-t;^-""" ^ , ._:
' ' 1 1— ' 1 ' J-L i_l_J_!_L_l_l_J_J_Ll_l J _ iJ .i.A- 1 \
2 00
80
1 60
1 .-o
1 20
1 00
0 CO
0 Gl
0 ")
• 0 20
0 00
9
nan
036
032
028
02'
020
OlG
0 ; ^
00"
00'

9

03G
032
023
02-i
020
OlG
012
009
00'
000
225
200
175
150
125
100
075
050
025
000
/r~l 	 ' ' '
-
-
-
-
-
-
A-
- A/^
" , 1 , 1 • » -
5678
r~ j — '"j — ' — | — i [ — I — | — i — | — i | i — j r^
RUGUSr 8 1976 -
-
-
-
' "-^
^S*^' ~
j^f*-^"*^
' ' ' ' '
-
1 , . ,1,1, ,1
. d J'V
225
200
175
150
125
100
075
050
025
nr, n
                                                                       11  12  13  H 15  16  17  18 19
                                                                       HOUR",  for
        8  9 10  11  12  13 14  15  16  17 18
                HOURS  FOT
                                      19
   12  13  It 15 16  17  18
HOURS, LOT
Figure 60.   Ambient data,  weighted mean, max and  min  for Day  221.

                                             131

-------
JU
1 80
1 60
1 "0
o 1 20
a 1 00
S 0 80
0 60
0 "0
0 20
n nn
. ' ' ' ' ' I ' 1 r 1 ' 1 ' 1 ' 1 ' 1 ' "~^l [ I ' .
_ flUSUST 25 1976 _
-
-
-
X A X X
— x —
A - /\T -
\ * / \'
V— fvi-« / ^
* ' , • * • J -
" , i , i , i , i ; i , , i , i , i , i , i , i , i , •
C \J\J
1 PO
1 60
1 10
1 20
1 00
0 80
0 60
0 "0
0 20
n nn
 5   6   7   8  9 10  11  12  13  1»  15  16  I/ 18

                  HOURS. CDT
                                            19
                              flUbUSF  25  1976 _
     010

     036

     032

     028

     024

     020

     016

     012

     008

     004
                               15  16 17 18  19
                                   p-
y-r-
                              RUGUST 25  19 76 _
                  V
                    -
i  . !  , I  , !.. I  i  I .  I i  i . .i_i_Lj-J  . I
7  8  9  10  II  12  13  pi  15  16  17  18
            HOURS, COT
                                            19
     036

     032

     028

     024

     020



     012

     008

     004

     000
U J U
0 ^5
0 J0

0 35
C.J
° 0 25
CJ
j 0 20
0 15
0 10
0 OS
0 00



25 0
225
200
175
150
n 125
o I™
0/5
050
025


QOO

2 00

1.80
1.60
t 1.40
f 1.20
'{=
r i.oo
o
u 0.80
° 0.60
1C
°= 0.10
	 I
8 0.20
n nn
T-pr-p-r— p— | ' i , , j i | , | , i | i i—
- ^ RUGUST 25 1976 _
^ 	


r\
^A \ -
=< V \
\ / 1
\ * I \

1 . . ^-5S»J ._, J

" , 1 , 1 1 < 1 , , 1 , ,1,1,1 , 1 , , "
5 6 7 8 '1 10 11 12 13 la 15 16 17 18 1
HOURS COT

--.-pry-i'I ' | ' 1 ' 1 ' 1 ' F ' 1 ' 'I'l1 ' _
1 RUGUST 25. 19/6 _
- -
/^ -
/ -
/ -
S'* ' ~
- /
?''

-------
                                                                 I r-p-r
' I ' ! ' 1 ' 1
CR 17. 1976 _
-
-
-
-
-
—
-
i 1 . 1 , 1 i "
3 60
3 20
2 GO
2 -"0
2 00
I 6"
I 20
0 £>0
0 ~G
0 00
. r i ' i : * i i ' i ' i ' i ' i ' ' i '• ' i ' i '_
A SCPTEPfCK 17 1976 _
/ \ -
:. . \
/ V
/ • --
•• / - •
/ ' '•>
"- . '-V,, -
. 1 J I • 1 . 1 . 1 , 1 1 "« 'f *-• -+ . . I , I I I , J
5 6 / e 9 i o i j i .' 13 it i s 16 i ; i a i
HOUR': nu
120 «'•>
108 225
096 200
OBJ 175
0/2 'SO
C60 R 125
r^s o 100
C.'6 0/5
Q; i 050
012 025
q

-' i ' i n~' i '' i ' ""'' ' i ' i ' f ' 'MM.
_ S£PFEM3CR 17 1976 _
-
X -
/
/
x ;^ '
^^^/^ 	 3
5 6 / 8 9 JO 11 12 13 H 15 16 17 IE 1
HQJRS TDT
225
200
175
150
125
100
OJ5
050
025
000
g

                                                                10  11 12  13  11 IS  16 17  18
                                                                   HOURS. LOT
                                                                                        19
Figure 62.   Ambient  data weighted  mean, max  and min  for Day 261.

                                           133

-------

1 ' 1 '






V"
>V
V

L: 1 .
Oil!
HOURS

1








N^
-
1 i 1

-r-p-p-IT -P-P~
OCTOBER 1 1976 _
-
_
_
	

-
-
XyV -:

, 1 ' 1 , t , 1 , 1 , •
2 13 1-4 15 16 17 18 1
ED
10
9
8
7
6
5

4
3
2
1
0
9
r
0
0
0
0
0
0

0
0
0
0
0

4
3
3
2
I 2
° 2
CJ
F i
~Z
i
0
0
0

oo p-p-p-pnm
60
-
20 •-
80 1-
UQ [_
00

60
20
80
J0
00
- y\ -,
•
-------
                              OC1QRCR 2  1976 _.
10 n
 9 i
 8 0
 7 0
 6 0
 5 0
 •* 0
 3 0
                                                         QQ
                                                                                      9CTC3ER 2  1976 _
i  CO
3  60
3  20
2  CO
2  J0
2  00
1  60
o "0 .-
0 3t
                                                                      9 10  11  12  13 It 15  16  17  18
                                                                           HOURS. LOT
                                                                                                 19
    Figure  64.   Ambient data,  weighted mean, max and min  for  Day 276.
                                                135

-------
7.3  EMISSIONS ALONG TRAJECTORIES
7.3.1  Procedure for Processing Emissions
     The procedure adopted by OAQPS/EPA for estimating  emissions  was  to
calculate an average emission rate over the entire uncertainty  band of the
trajectories for each hour.   Thus the coordinates  of the extreme  trajectory
locations at each hour were used to define the area over which  the average
would be computed.  These areas were shown in  Figures 31-54.  A computer
program was used to identify and process the emission inventory grid  areas
that fell inside each of these areas each hour.   A detail of the  grid near
the center city (RAMS stations are plotted as  *)  is shown in Figure 65.
     Table 18 is an example of the area source printout from the  computer
program for hour 10 (LST) of day 159, June 7,  1976.  Listed  are the grid
squares that were totally or partially included in the  trajectory area, their
area, the fraction of the area included and the proportional  amount of the
emission rate for the pollutantb.  Table 19 is a  printout for all the point
sources that were in the area of trajectory between hour 10  (LST) and hour  11.
     The emissions database used was the so-called "modellers tape" in which
the hydrocarbon specialion of each of the source  types  has already been
accounted for and the composition has been converted to an appropriately
magnitude weighted,five level, lumped species inventory.   These  lumped species
are:  NR, non-reactive; PAR, paraffins; OLE, olefins; ARO, aromatics; and ALD,
aldehydes.  Tables 20 and 21 give the inventory average molecular weights and
carbon numbers for the hydrocarbon species.  These were obtained  from
Demerjian (1980).
     In their in-house study (Gipson, 1980), OAQPS had  converted  the  emissions
to units relative to initial concentrations and initial  mixing  heights for
use in the OZIPP program.  In addition OAQPS had  treated point  sources as if
                                     136

-------
 o
 c
 O)
-a
 ai
 s-
 cn
CO
a_
 c
 (U
 OJ
 1/1
 ai

 o.
 
-------
O
cu
CO

i
ro
O)
OI

Q.
E
re
CO
oo

K
? C
LxJ «
-1
O
ac
o
u
I/- G
UJ
U
u.
0-
l/l
CC
o a
03
V
2
CC
c t)
f
c
t-> 1^
X ^
X.
U">
^
NOX
MS EH IS SIC
0 is
t- u-
t.
LJ
is
^ c;
LLJ
H
5
;
n
r\.
j
ru
r*-
— PJ
i~
K1
T •
^
— P.
T »
V C
c: m
x *•<
^ cc
CT
£* r-
f> P,
J .
-j
P>
n
f
Kl
ry
C)

c*. t*.
0,0
fn
f\» o>
rvj p^
PJ
- if
r?2
"
nj  ,0
• «
a __,
cc or
0- a)
""» in
a IT*
co a
in
0- m
O u
-c »
x r-

«• -e
f, (*~i
PJ f\
r c.

o r-
r- f
in a
PO C\
••«
PJ ir
» o
o c\
r- -c
ft C\
m •-
h- r,
M
m 0
i"j *n
•^ oj
CO P"
f*- CV)
m in
o- cc
CJ '">
cc fv-
o cc
._, fl
iH
c> "-
cr
i j fi
-0 N
-rfl -C
p. r-

J
a- tr
CO 0
* 4
o* a
m *
r- f-
j» *-
a o
o- m
ru =*
o- c-
in tr
,0 ,0
o '-^
c . CJ
it.'
a tr
n. o-
t\ P-

«o
PJ "
COK
_« p.
(E P-
in t
U> »"
* CO
,0 0
CO r;
-0 -0
m cvj
0 PJ
tc r-
rj D
n PJ
.-» r-
,0 r-
r> O-
P. P.
C f-

(-1 U
oa
•" P
•o ^
in F.
ro c
in *
-O ,;
«-• o
^ a
CO (T
-O rv
1 CC
-c rv
* *
O o
t~> -i
* •
0 o

K> fr,
K r<-\
r- c

.,
to «
:]
2*
eo 0*
££
n ir
!C '""
O f>
«
- o
0 h-
fX r-
0 f»-
C^ O-
o r-,
. »>
P, I**
en K\
K* K
r- C

J
*
r^ ^
r»
(-1 .«
CM n
.
LT r*
a .*
w%
m *
m
fM ^
^ r*
in
•o »
a- r-.
• *
CO 0
T- »
f^i
m ,T
.'i 0
~« 0
in ->
0 ,-i
a p,
Kl ^
K Kl
*"i C.

PJ O-
o  U^
f , u-
1^ ^
0 O
_j r-j
f3 0
I-1 *— »

fo a
^ -c
l»-| K\
C- r-i

J
*) fO
^ a
PJ in
o -
o> «-•
PJ
in o
-^
0- *
PJ
on •-•
o cr
~« a
r* o-
*\ PJ
* ~;
> Kl
1 »^
. .
-2
rj a
•_* i*"*
f> O1
-' t^

m -o
^i 
CC fM
•C t_
0 0
ri o
0 o
ri C
-» o
ft if\
{-• c_

a o-
o* «
K1 CO
« J
in f^-
^
IT (T
» t-,
CO (O
let
u> r»
r. a
 f^
IT
ft P-
•-4 a
^3 r>
ot,
in cc
PJ
K> CO
-0 0
PJ
a inJ
O PJ
M
a 4
K, CO
o* r*
Kl
••* P.
a co
PJ
r^ - 1
IT f\
f^ O
PJ
-0 0
•0 ri
CO O
IT f~-

CC O
o- -.
(•5 /-i

C C

CO Pi
in
•A
~Q
m ~*
PJ *n
co o
,0 P-
a *
r- r-
4A ^
O* r->
r* 0"
a P'
O* C
(\i PJ
0 a

1-1 P.
• '
-•:
r*. i**)
arJ
* r3
in «C
r^ r-
PJ a
r* >n
^4 00
-o **
tn ^
•C f**
et CJ
fx in
a O
a PJ

in •&
--3
» i
50 S
o- *#•
re tc
-S
ru ;»•
~2
r- tf\
r^ ro
cr
r> O-
-j 0
CO -'
.A OJ
K>
frt f
a IT
PJ PJ
in CM
K\ f*-
«1 0
a or

0 PJ
o ' *
P. K
C r

o if
r«
-. IT
^- rv
in «£
*
^1 «0
c\
in. o
**
^ CM
f- a
a
AJ *>
> IP
in cr
J -0
a
in '-j
r- ,1
•o '"
"^
K 'C
K4 K

r» PJ
i-.
CTiO
o
f*^ m
f. M
a-
o ru
f\i
v"> »-i
in
^1 M
o- PJ
•_i a
• •
U1
J» N>
O »-t
•_ o-
iC M
a
.n a-
-j -0
i a i*\
w r-
~> •-!
o> fo
f-l f
c c

-c r-
• •
-* (ft
r* a-
*- a
PJ 0-
in »i
^
> 0)
•« in
ro
M »p
KI m
o- m
m ^
in KI
o o*
"1 M
p.
r* LP
CO 0
^ rv.
*r -O
a ru
• •
a m
a a
tr t*
r; c


o
M


r-
PJ
in
XI
r~
ra
a
rvi

-C
I

»
m
in
1*1
CO
r-
^
Jl
O
PJ
o
r*
O
r*
!"*\
«H
a
(E
PJ
Jl
in
m
•^
?
CO
fo
rA
a
EC



3





0- «
^ Q
PJ *
a »o
a- ^,
C\J o
-c a-
• •
m
PJ
•» 3C
M JiH/S3TOW 3)1
(UX TS XH>£N >
- >-
* i/;
•^ ^
-i |,,
O
- O
/I ^
' w
- 1^
»- t-l
>' K-
UJ


                                                     138

-------




















_!_._!

0
C
i.
Q.

£*>
O

c
Qj
C
HH

CO
O
i/l
1/1
•p-
LU
0)
I,
^3
O
CO
£
£X
0)
CL
E
ro
00


CTt
1 •
OJ
-Q





c.
i ' <
rr
T
_j or
o <
in
wn
Ul
0 -1
u- O
Q_
l/l

O
u K
< o <
•a: a
X

UJ
i CC
I 1
' Z
1
tt !
" ' 2
o a —
U1 ' l/l T
'*-t IS
h- ' UJ —

!£ '<_>
•— !
O ' 1/1
a 'o ~
/i X
»— c
U- **
1
' o

•s
y tv
'^ V
i: t"
U, *:
2
'.*-



UJ
^^_

t-
f ^.
s

i T
I


00


o o
G f^i






,-, r-
cr


0 -0
' ~





^ tr>
(*> tn
K>





8 A
0 ,1





C <_.









•\J "•'
3 ^
_ r t
3 er j
3d,
5 O C
!


0 C




CM OC





m CO
A, -


nr ^
a *





0 P*
O (^
f^ fu





r* i T
"' '"





c- r-









» 5

r
? o
/} -C

, a



0 C




,**





^ ^
tr CC


,r, «
,; ^





-<3 -C
"-1 —





LJ X
(.1 -t





Ci O









if IT

" r
cr, o
rf> -i
^ *»
"l



!c o
i



-O f^





«
CO ,0


in vi
LA r\;





•fl _-
*-• CM





' •» O
,1 „-





D Ci
•> I-'






t


r. ^

i t
D DJ
W ^
1
J 0 1

1
1
i
0 0




^'«;





3 ^
•O O


-o ^
o! j;





in f^,
nr, >~i
f^ if>





( i (_ i
T . 1





= 5









- 2'

f' C
rf o-
a 4
i
1
1
1
1
«cv




•





„ „



OJ f-
i m





0. O






S™
-. uO





-o r-
r^- a









- *


r^ ,
ff ff
E H
^ r.



1 * *


1







=r ^



» r
*-. o





". *:






^ CO
p~ '\
* r-
!•' r,




if <





? n
^ HI


^ n

r c.
tr tr.
f *
0 0

'
] i
^ — i .0 c*
•«

1
|
1


[ \

i i
i_ j,l.j
1 •-<
!
1
i ;
i
1 i
|

r- f\'





' ") o ^*






f> .-»• fv
^ CM •>
*•' u" er




1
a r\ r-




l
i
"";""


?-~ -

»- t_ r
or cr o
<; j- <:
H ^ H
>
;










































^
1














|C



1 *
1 —






^
«

fM
-O





^n
ji
^




;.
">











'



J

^_
f -
a
3

^ ,


0 0


n o

1-4 -O






K> -.
f\j

o m
-0 -





rt "
^ wl
~











...
j O
j i"-i






M 
M r-t
i t"



0 0


0 C

^ sfl






J, ^
:"

. rr






~ ~
". *\
•. -c





,0
*. LJ
*> rj





-> C_
3 0






N.S


- J
f ^

El
S S3
« r«
j O



0 0


f^ CO
•A
* tt







&7

^ en
CC,





s!
" *

^




~l C_J
r, ^





"» CJ






M rv,


» n

~"
>-2
£ j;
^ »
""
-j r-,



0 0


0 0

a ^






« ^
"~


O f>





*• o-
r\j ^r
M ro




.?§
t*5 .-j





3 O
"1 U,









^ r-



3 ft
-» M
-' 0



0 0


r- o

* ^






W I-*
M t>


r> r--





-O 0,1
^
lOj tC
•^




j « i
r i '_
3 ~>





= S






"~


~ ?
a a

i J>
a =
;;



0 0




^ ^






03 ;»
S"


M^





--t £
^ c-»
rr 53
nj




"V C1
" o





•j C






^ ?,


r ^
*• r^
t» a

•< ( ,
*t <
t? o-
I
»• »~l


0 0


0 0


















sc
ra 0






-IS
O t"!)





D C
^ 0






nS


0" O
r* r-

__
f!
c» -J
T O
- ri



0 0


O 0

(M

"*




u^ '
~



00




r\i (->
^
•\j





* o





D C.









- «_

— (. '
-i <-;
B" Of
* 3
-"
•-*•


0 0


C CO
--
r-'o
fM (M





r-' <0
^ o


* a-
-. —




"> ec
r^ in
-t -0
— "




•> o
n ^





~ 0
3 L-_>






is


~> o

3 a
C S
' 2
:
t— ^^


„
i o
<-•
0
CO
o
i r-
^
r^*
o




cs.
J
OJ

1 K
J
r~
OJ
l

u
r* ' •
in O-
£
M 1


j
in cr
X ^
-• ^
I
1


:fc

f\j in
^r

1
rv "
*~ *j i—
(V C -I
T IT C
XX
l- C1

t/1 i/ */"
c. c o
u1 1/1 tr
L.I H UJ
— 1 —^ — '
** -* <
c c o
H- ^ »—
139

-------
      Table 20.   Species Molecular Weights




Species                         Molecular Weight



  HC                                 14.5



  CH4                                16.0



  NR                                 68.0



  PAR                                88.6



  OLE                                42.3



  ARO                               100.0



  ALD                                46.1
     Table 21.   Hydrocarbon Carbon Numbers





Species                         Molecular Height



  CH4                                1.0



  NR                                 3.0



  PAR                                6.3



  OLE                                3.0



  ARO                                7.1



  ALD                                2.0
                       140

-------
they were area sources.  That is, the summed point source emissions  for each
hour (the values at the bottom of Table 19)  were divided by the total  area of
the "box" for the hour (the value near the bottom of Table 18).  These
averaged point source emissions were then summed with the area source
emissions to give a single emissions value for each hour.
     In this study, UNC decided to use absolute emission units of ppm-km/
  2
km -hr in the models instead of the relative units used by OAQPS.  An  emission
rate of 0.25 ppm-km/hr would result in 0.25  ppm of material  accumulating in
          2
a box 1 km  and 1 km high in one hour and would result in 0.50 ppm
                           2
accumulating  in a box 1  km  and 0.5 km high in one hour assuming no
reaction of the material.  The value is simply divided by the current  mixing
                                                            2
height to get the rate of change of concentration for a 1  km  box.
     To be consistent with the OAQPS study,  UNC decided to apply the same
assumptions as OAQPS did  to the point sources but to determine their
contribution to the total emissions.   The OZIPPM model program allows  for  a
single emissions composition to be used for  the entire day.   To be consistent
with this approach, UNC decided to compute this compositon from each day's
trajectory's emission inventory.
     To perform all the above tasks,  a computer program was  written.   The
program accepts as input  the area and point  source averaged  emission values
such as those at the bottom of Tables 18 and 19 in units of  kg/hr  or moles/hr.
These are converted by use of the area, molecular weight,  and average  carbon
number into ppm(C)-km/hr  units.
                                    141

-------
The basic equations are:

                          Ei  = (1000*factor/MWi)*(Esupi/A)

                          Ei  = Ei*CNi

                          ET = EEi


                    where Ei  = emission rate in ppm-km/hr for the ith species
                                                       o
                     factor  = conversion from moles-km ' to ppm-km (at 25°C)

                               factor = 24.46656 x 10~6

                        MWi = molecular weight for the ith species

                      Esupi = supplied emission rate in units of kg/hr
                                                                        o
                          A = area covered by trajectory in one hour (km )

                        CNi = carbon number for ith hydrocarbon species

                         Ei = emission rate in ppmC-km/hr for the ith

                              hydrocarbon species

                         Ey = total emission rate in ppmC-km/hr for all

                              hydrocarbons and aldehydes.


In addition the program computes the carbon fraction composition for each

hours emissions and the running carbon fraction composition for point sources,

area sources, and total emissions.   An example printout for the 1100 LOT

(1000 LSI) hour of June 7, Day 159, is shown in Table 22'.  One printout page

is produced for every hour of each day.
                                     142

-------
                                   o
                                   o.
                                                    o c
                                                    O rt
                                               0  CN

                                               O  CO
r*>
o
o
 c
 o
                                                           o
                                                           o
                                                           o
                                                                                                            o o o

                                                                                                            coo
                                                                                         o o o

                                                                                         o o o
                                                                                             C  -H "
                                                                                             o  o o
                                                                                             o  c c-
                                                                                             o  o o

                                                                                             o  o o
                                                                                                                                ^ ri  "
                                                                                                                                o c j  TJ
                                                                                                                                o o  o
                                                                                                                                o o  o
                                                    i- ro
                                                    o   •
                            o  r.
                            o
                                                                                         000
                                                                                         -.' O O
                                                                                             o  e  c
                                                                                             o  c  c-
                                                                                             o  •:<  o
                                                                                             o  c  c

                                                                                             coo
                                                      c -• c
                                                      O -I r-
                                                      o o o
                                                      o o o

                                                      o o o
 0}
 O
 i,
 IXJ
 OJ
o
u:
<:

LJ
_!




O
oc
IN
r-i
o
<

0
rj
Ch
0
o
«
o
fN
O

.
M
O-
0-
o -
O -i
o «
c
-o .-.
o •
o o-
o n

o
n
c
o
C'
o
CK
o
0
o

o
o*
CO
-c

0-

o
o


r^1
6
o
c
<•

o
o

o
r- ra
o o
o ^
o c
r- c-
T-H ro
0 Q
0 fl

0 O
IN
0
o
c
-o
fx
c-j

o
*c-
o
o
c
-0
c.

^

0
C- 0
-'"' I*
u~ lj~
o c
O 0
CN CN
OS CC
o- o
O 0

0 0
TH ri cc
r. LI -7
O x; *C
0 O O
0 O O
fi IN O
<* r. o
c K re


coo
 O
Q_
 s,
 o
o

-------
7.3.?  Results of Emissions Calculations
     Table 23 summarizes  the calculated distribution  of emissions  among  point
and area sources for the  10 days;  these values  are illustrated  in  Figure 66.
For comparison, Table 24  gives  the distributions  for  the entire St.  Louis
area emissions inventory.
     Figure 66 can be interpreted  readily by reviewing  the  location  of major
point sources, Figure 29,  and the  trajectory paths for  the  days.   The  tra-
jectories for days 159 and 160/15  were both  north of  the central business
district (CBD) and spent  significant time near  the Alton/Wood River  group of
point sources.  The trajectories for days 160/3,  221, and 261,  which have
similar distributions in  Figure 66, were  all  west of  CBD, and day  238  was
east of CBD, areas reasonably low  in major point  sources. Days  195,  201, and
275, which also have similar distributions in Figure  66, had trajectory
paths near the CBD (stations 101,  103, and 109).   Day 276 was northeast,  of
the CBD.
     Since the point sources were  uniformly  distributed into the entire
trajectory area for one hour, it is difficult to  estimate whether  their
influence has been over or under estimated.   For  example, a large  point
source located just inside the  upwind edge of a trajectory  "box" contributes
for nearly an hour in advance to its being encountered.   On the other  hand,
it is dispersed over the  entire box volume (in  effect,  it diffuses "upwind")
thus greatly reducing its  contribution to areas near  it. Sometimes  point
sources located 20 km perpendicular to the trajectory could make significant
contributions to a given  hour's emissions.
     Table 25 gives the computed percent carbon distribution among the five
nonmethane hydrocarbon classes  in  the emission inventory for point sources,
area sources and the total for the ten trajectory paths. The  total  percen-
                                    144

-------
                          Table 23

      Distributions of Emissions  for Selected RAPS Days
                      (Percent)
                   CO
                                   ,  NO.
HC
Day
160/3
160
159
221
261
201
195
276
275
238
Date
Jun
Jim
June
Aug
8
8
7
8
Sept 17
Jul
Jul
Oct
Oct
Aug
19
13
2
1
25
Area
89
53
25
90
94
78
85
47
73
99
.8
.9
.1
.1
.5
.6
.7
.3
.7
.7
Point
10
46
74
9
5
21
14
52
26
0
.2
.1
.9
.9
.5
.4
.3
.7
.3
.3
Area
86.9
72.0
34.6
84.3
90.5
58.7
66.0
65.1
62.8
83.0
Point
13
28
65
15
9
41
34
34
37
17
.1
.0
.4
.7
.5
.3
.0
.9
.2
.0
Area
81.6
64.5
27.8
78.7
76.0
81.5
85.8
84.5
88.1
65.7
Point
18.4
35.5
72.2
21.3
24.0
18.5
14.2
15.5
11.9
34.3
TABLE 24.  TOTAL EMISSIONS  FOR  THE  ST.  LOUIS  AQCR  (TONS PER YEAR)

Point Sources
(% of Total)
Area Sources
(% of Total)
Total
Particulates
45,224
(3%)
1,299,782
(97%)
1,345,006
so2
1,007,530
(97%)
30,813
(3%)
1,038,334
NOX
322,730
(72%)
125,567
(28%)
448,297
HC
47,610
(23%)
157,204
(77%)
204,814
CO
164,331
(11%)
1,325,556
(89%)
1,489,887
                             145

-------
      Table 25.   Carbon Distribution  of NMHC  from  Emission  Inventory for
                 Selected RAPS  Days  (percent)
 Day     Date

159      June 7



160/103  June 8



160/115  Jun 8
195
201
221
238
261
275
276
July 13
July 19
Aug 8
Aug 25
Sept 17
Oct 1
Oct 2
Source
Area
Point
Total
Area
Point
Total
Area
Point
Total
Area
Point
Total
Area
Point
Total
Area
Point
Total
Area
Point
Total
Area
Point
Total
Area
Point
Total
Area
Point
Total
NR
13.7
5.1
7.7
11.2
3.8
9.9
15.3
4.7
12.0
12.0
10.1
11.7
11.8
12.0
11.8
15.6
4.8
13.7
6.8
75.9
6.75
11.6
4.0
10.0
11.8
15.2
12.2
15.2
5.3
14.0
PAR
50.2
79.6
70.6
57.0
46.4
55.1
44.3
80.1
55.5
54.3
47.5
53.4
54.3
45.9
52.9
42.0
79.0
48.3
75.5
11.1
75.5
55.0
56.1
55.2
55.5
39.0
53.7
44.2
78.0
48.5
OLE
22.3
10.2
13.9
17.4
4.2
15.2
26.1
10.1
21.1
19.1
5.3
17.4
18.7
6.0
16.6
28.9
7.5
25.1
10.1
5.3
10.1
18.6
1.8
15.0
18.4
6.8
17.2
26.2
4.1
23.3
ARO
11.2
5.1
7.0
12.4
45.6
17.9
11.3
5.0
9.3
12.4
37.0
15.5
12.8
36.0
16.7
10.1
8.6
9.9
6.6
4.2
6.6
12.7
38.1
18.1
12.0
38.9
14.9
10.6
12.5
10.8
ALD
2.95
0.00
0.82
2.09
0.07
1.72
3.07
0.00
1.98
2.31
0.00
1.98
2.38
0.00
1.94
3.43
0.69
2.86
1.01
3.03
1.08
2.21
0.00
1.68
2.29
0.00
2.02
4.05
0.00
3.42
Fraction
of Total
27.8
72.2
81.6
18.4
64.5
35.5
85.8
14.2
81.5
18.5
78.7
21.3
96.4
3.6
76.0
24.0
88.1
11.9
84.5
15.5
                                     146

-------
 UJ
 O-
OQ

CO
2
O
I— I
CO
CO
o
o
100


 90


 80


 70


 60


 50


 40


 30


 20


 10


  0
           159   160/3  160/15  195    201    221    238    261    275


                DISTRIBUTIONS OF EMISSIONS FOR  RRPS  DRYS
                                                           276
 100


  90


  80


  70


  60


  50


  HO


  30


  20


  10


   0
Figure 66.   Distribution  of CO, NO  ,  and  HC emissions  for trajectory
             paths through the RAPS  emission inventory.
 LU
 CJ
 ce
 LU
 Q_
o
I—I
»—
ZD
00
t—I
ct:
i—
CO
•—i
C3

Z
O
CO
ct
cc
o
100


 90


 80


 70


 60


 50

 to


 30

 20


 10


  0
          ALD
 Figure 67.
              PAR
              159  160/3  160/5  195  201   221   238  261   275   276

                   TOTflL NMHC FOR SELECTED RflPS DflYS
100


 90


 80


 70


 60

 50

 to


 30


 20


 10

  0
         Carbon distribution  (percent)  of NMHC for trajectory
         paths  through the  RAPS emission  inventory.
                                 147

-------
tages are plotted in Figure 67.   Day 238 (Aug 25), in which the trajectory



came from east of the CBD and curved north, had the highest paraffin frac-



tion, 75%.  Day 221  (Aug 8), in which the trajectory carne from northeast



of the CBD, swept through the city and ended west of the CBD, had the highest




olefin carbon fraction, 25%.  Day 261 (Sept 17), in which the trajectory



started slightly north of the CBD and went south, had the highest aromatic



fraction, 18%.  The significance of these compositional  changes will be



examined more fully in later sections (see 7.3.4 and 8.3).



     The computed hourly averaged emissions rates for each  selected day are



given in Table 26.   These rates exhibit significant changes from hour to



hour and the solution methods in the computer programs (such as the Gear




routine in OZIPP and PKSTM) do not work well with such changes.



     One method for dealing with this problem is to substitute a smoothly



changing curve for the hourly averaged values.   Figure 68 shows an example



of NO  and HC hourly averaged emission values and a smooth  curve fitted
     X


through these values.   An important criterion for such a  curve is that the



average emission computed from the curve each hour must  be  equal to the



original constant hourly value.   Another way of stating  this requirement is



that when the average hourly values are plotted in a histogram-like plot,



then the area under the smooth curve each hour must be equal to the area of



the histogram bar each hour.



     The algorithm for producing these equal-area, smooth-curve histogram



fits is not simple.   Unfortunately, the algorithm in the OZIPP program does



not always work properly and sometimes it produces negative emissions rates;



this was true for several days in this study.  Appendix   A  discusses the



problem more fully and describes an algorithm developed  by  UNC to solve the



problem without giving negative rates.




                                     148

-------
1 . -J
1.3
1.1
0.9
0.5
0.3
0.1
-0.1
-0.3
-0.5
-n ~i
. ' 1 '
i
A
i
i
1 1 '
i
1
i
- DflY/22\l fiUGUST 8, 1976 NO, HISFIT — 1
—
—

—
-


,
/




1




\
v — -



•x. -/•*




^



,-,
' \




\
^v^




" , 1 •
|
i
i
1
i
,
, i



— x
\




i

—

—
—
_
S ••>

—
—
, "
1 . 
-------
                              CD O O O O O
                                 O CD O O O
                                                                                 O O O CD O O O
                                                                                 10 co o r-.  (
                                                                                 en cri 01 o
                                                                                 O O O —  r
                                                                                                                                         CM
                                                                                                                                            E
                                                                                                                                           _i1 -rj- I_T) ID
                                                   o o o o  o o o
                                                   o o o o  o o o
                                                   r^. cO en CD  .— cxj n
                                                   O O O -—  r- r— ,—
CXJ
                              roncnr
                              OCDO.
                              CDOO
                              OCDOCDOCDOOO
CVlCXJLO
OOO
ooo
OOO
                                       LOuOCOtMCDCO
                                       i — i — O t — UDO
                                       oooooo
                                       CDOOOOO
                              OOCDOOOOCDO
                              OOCDOOCDOOO
                                                                                                                                                                          CD
                                                                                                                                                                          t.
                                                             .
                                                   in i.n  '— UDf^-CNJ^
CDOt'Oi— •— i— LnOOQO
CDOCsjOCDOOOOOO

CDOOOOCDOOCDCDO
                                                                                 CD O ro r^ *3~  ro  O CT  OOCDOOOOCD
                                                                                                          CDC'OC  OOOOOOOCD

                                                                                                          o* o o cr o o o' o o o o o
                                                                                                              v

                                                                                                          f->rouDc\'U->OJO.— i— OOO
                                                                                                          CDOCXI.— OOCDOOOOCD

                                                                                                          OOOCTCDCDCDOCDOOO
                                                                                                                                         OCDOCTOOCDOOOO'
                                                                                                                                         uD r-. CO O". O .— cxjrota-iD
                                                                                                                                         OCDOO-— r-r— ^-r— r-
                              i~O CT> CTv <— CM CSJ i— CD IT1
                              i— r— i— cxi ixj CM m ro o
                              onrororoo-jroi— o CD
                              ooooooooo

                              OOCDOCDOOOO
                              ooooooooo
                              OOOOOOO CDO
                                                   CD OJ r^- UD -^T O CXJ
                                                   CD O f"> 'JD i— .— CD
                                                   O O O O O CD CD

                                                   CD O O O O O O
                                                                                 CD O O O CD O O
                                                                                 CD O O O  CD O O
                                                       cxicxji-OCQinrocor-    tn
                                                       O co f*^ r-- cvi <^j- LD LO •—•  ^D
                                                       ro o O txi i— 
-------
     Figure 69 illustrates,  on  the  same  scales,  the  emissions  rates  used  in



the UNC simulations.   The  equal-area  histogram   fitting  procedure  was  applied



to the values in Table 26  to produce  the rates  plotted  in  Figure 69.
                                   151

-------
NOi.  ppm»km/mm  (XI IT* 3
                                                           On.  ppm«km/mln   CXI0^3
   U-OtX)  "
                                                                     o  o  cz>  o
                                                             (r-OIX)  uiui/uj>|«"id  l/l
fO  >>
S-  (O
   ID
                                                                                                                CO  rO
                                                                                                                CO


                                                                                                                E  to
                                                                                                                Ol  C
                                                                                                                   O
                                                                                                                01  -i-
                                                                                                               +->  4->
                                                                                                                13  (O


                                                                                                                O  3

                                                                                                               ££
                                                                                                               =t  to
                                                                                                               CTl
                                                                                                               ID
                                                                                                                O)
                                                                                                                I.
   C,-01X)   ui
                                                              (c-OIX)  u|iu/"i>l«l''d<1  '03  'OH
                                                   152

-------
 N0».  ppm»km/m!n

o   o   o   a
                                                           Oi.  ppm»km/mln  CX10")
                                                                                                                |«uid|«ujdd '03  'OH
s- c
O O
                                                                                                                O)   •
                                                                                                               •!-> to
                                                                                                                ro >,
                                                                                                                1- rt3
                                                                                                                   T3
                                                                                                                c:
                                                                                                                o i—
                                                                                                                E  01
                                                                                                                ai  c
                                                                                                                   o
                                                                                                                O)
                                                                                                                i-
                                                                                                                O1
                                                                     "iuj/ui>(*ujdd  'o
                                                  153

-------
NO,. ppmokm/mln   CX10"1)
                                                                                                        O>
                                                                                                       -&
                                                                                                        a
                                                                                                       o
                                                                                                       o
                                                                                                       OJ
                                                                                                       OJ

                                                                                                       "o.
                                                                                                       "D
                                                                                                       O)
                                                                                                       CO
                                                                                                       13


                                                                                                       CJ
NO,,
                  CXIO")



                                                                                                       CO
                                                                                                       O)
                                                                                                           CO
                                                                                                           >,
                                                                                                           ro
                                                                                                          -a
                                                                                                       CO  03
                                                                                                       CO


                                                                                                       f=  CO
                                                                                                       a>  c
                                                                                                           o
                                                                                                       OJ -i-
                                                                                                       4-> 4->
                                                                                                       Z3  ra
                                                                                                      vo
                                                                                                       OJ
                                                                                                       s-
                                                                                                       CD
   C.-OIX3   uiuj/iij>|.iiidd '03  'JH
                                                 154

-------
 7.3.3  Method of Treating HC Composition in Each Mechanism
     Mechanisms differ in the number and characteristics of the species used
 to represent the hydrocarbons present in the photochemical systems to be
 simulated.  Hydrocarbon composition data available from the RAPS emission
 inventory (see Table 25 and Figure 66) usually needed to be converted to the
 appropriate representation used in each mechanism.  The actual conversion
 depends on the characteristics of each mechanisms' hydrocarbon species as
 defined by the mechanism developer.
     In simple trajectory models, HC composition data are utilized in calcu-
 lating emission composition along the path, in addition to calculation of the
 initial morning HC concentrations.
     The HC concentrations needed internally in the model  are in ppmV.  The
 values for each HC specie used in the mechanism are calculated by multiplying
 the measured NMHC value from RAPS (or the NMHC values calculated from emission
 inventory) times the calculated carbon mole fraction for each specie, divided
 by the determined carbon number for each specie.  The conversion calculations
 used in this study for each mechanism will  be described.   Table 27 lists
 the HC fractions.
     7.3.3.1  Dodge Mechanism.   The Dodge mechanism used a mix of butane and
 propylene to represent the hydrocarbon species present in  a photochemical
 oxidant system.   Dodge had determined that  a carbon mole fraction of 75%
 butane and 25% propylene yielded a system which had a photochemical  reactivity
 equivalent to the Bureau of Mines smog chamber experiments.   The Dodge
mechanism, therefore,  does not  utilize HC composition data.
     In the standard OZIPP program the assumption  has been made that aldehyde
 fraction would be  5%:   2% formaldehyde and  3% acetaldehyde.
                                     155

-------
              Species             Fraction           Carbon No.
               Prop                 0.25                 3.0
               But                  0.75                 4.0
               HCHO                 0.02                 1.0
               ALD2                 0.03                 2.0
     7.3.3.2  Demerjian Mechanism.   The Demerjian mechanism uses  three HC
classes (OLE, PAR, ARO) and one aldehyde species to represent all  aldehydes.
The Demerjian mechanism utilizes HC composition data of carbon mole fraction
only.   The carbon number for each HC and aldehyde class is  defined by
Demerjian.  Therefore, the RAPS carbon number data in Table 21  is  not
utilized.   Ethylene is included in the olefin group.  These conditions allow
direct use of the RAPS emission inventory derived carbon mole fractions
(Table 25).  These fractions were utilized for initial  condition  determina-
tions  as well as  for the emission fractions.
               Species            Fraction           Carbon No.
OLE
ALK
ARO
ALD
RAPS OLE
RAPS PAR
RAPS ARO
RAPD ALD
3.0
6.0
8.1
1.0
     7.3.3.3  Carbon Bond II  Mechanism.   The Carbon  Bond  II  mechanism's  HC
species represent carbon bond types rather than hydrocarbon  classes,  with
ethylene explicitly represented as the exception.  The CARB  specie  represents
the carbonyl carbon present in aldehydes, ketones  and  other  oxygenates.  Both
carbon mole fraction and carbon number are needed  from the  RAPS  HC  composition
data.   Each carbon bond type  also has a  defined carbon number (e.g.,  olefins:
C - C has 2 carbons).   Portions of each  HC composition class are added to  the
paraffin carbon bond specie,  depending on the degree that the HC composition
carbon numbers exceed the corresponding  carbon bond  type  carbon  number.  The
carbon mole fraction of the olefin portion (of the HC  composition data)  which
                                     156

-------
is ethylene needs to be determined (or assumed), as well as the non-ethylene-
olefin carbon number.  For this study both a 50/50 or 33/66 ethylene/olefin
carbon mole fraction was used.  Little effect was observed when one split
was substituted for the other.  The carbon number for OLE as determined in the
St. Louis emission inventory was 3.0.  Since the carbon number for ethylene
is 2.0, the eth/ole split and the inventory carbon number determine the
carbon number for ole without ethylene.  The formula below is used to
calculate the Carbon Bond II's Olefin # from the RAPS total Olefin fraction
(includes ethylene) and the assumed ethylene fraction/total olefin fraction
ratio  (E).

     CBII Olefin # =           ] " E
                             1
                         (RAPS Ole #)
       Ethylene carbon.      F..  „      n,   „         Inventory
       Olefin carbon        "n ff      uie ff       Total  Olefin #
          50/50              2.0        6.0             3.0
          33/66              2.0        4.0             3.0
     The carbon number for Ole in the CB2 mechanism is  2.0.  Therefore to
arrive at the CB2 emission fraction for olefin from the St. Louis  inventory:
                    (1  -  r— )(ole carbon frac particular day)
     CB2 ole frac = - — - - • - - — — - - -  x 2.0
                              4.0 carbons/inventory
                                    olefin

The initial  olefin value is:  CB2 init olec = NMHC * CB2 ole  frac
Likewise for ethylene:
     CB2 eth frac =       x  (ole  frac  St.  Louis)
                               270      ~~~   ~~  X   >U
                                     157

-------
and


                     NMHC x CB2 eth  frac
     roo •  -4.  +u
     CB2 mit ethc -
                             ?~0
     The aromatic carbon number is  7.1  in the emission  inventory  and  6.0  in



the CB2 mechanism.



     Therefore,




     rno      ..   f      (arom frac St.  Louis)    r  n
     CB2 aromatic frac = •* - 7— ^ - ^ x  6.0




     roo •  -4-        NMHC x CB2 aromatic frac
     CB2 init a roc  = - F-~H --
                                o .u





     The aldehyde carbon number is  2.0  in the emission  inventory  and  1.0  in



the CB2 mechanism:   the carbonyl  carbon  atom (CARB).  Therefore
     CB2 CARB frac -          -=      x 1.0




                          x CB2  CARB frac
     CB2 init CARB =
                              Q
     The paraffin carbon number is  1.0 in  the CB2  mechanism.   Therefore  the



St. Louis fractin for paraffin can  be used directly.   However,  several other



species contain paraffin carbon atoms in side chains  which  must be  added to



the paraffin class.



     The number of carbon atoms over two in the average  olefin molecule  are



added to the paraffin fraction.  The carbon atoms  in  the aromatic side chains



are added.  The carbon atoms in excess of  the carbonyl carbon  in the  aldehydes



are also added to the paraffin fraction.



     Therefore





     CB2 par frac = (par fraction St. Louis)



                  + (4.0-2)(l-£~)(ole frac St. Louis)
                                  __





                                     158

-------
           + (7.l-6)(aro frac St. Louis)
             _    .      __,
           + (2.0-1.0)(aid frac St. Louis)
                         2.0
     CB2 init par =
     The non reactives were treated as follows
     CB2 NR frac = NR frac St. Louis.  x 3>Q
     CB2 1n1t NR =
                           « U
     7.3.3.4  California Institute of Technology (CIT) Mechanism.  The CIT
mechanism utilizes four HC species:  olefin, ethylene, aromatic, and alkane,
and two aldehyde species:  formaldehyde and a general  aldehyde species.  Both
carbon mole fraction and carbon number data are required to implement this
mechanism.
     Generally the RAPS emission inventory derived HC  carbon mole fractions
can be used directly with the CIT mechanism.  However, the fraction of the
RAPS Olefin fraction which is ethylene must be determined or assumed.  A 50/50
split was assumed for the Ethyl ene/Total  olefin fraction ratio for the Level
II simulations (The BOM autoexhaust analysis indicates 42% of Olefin is
ethylene).  Therefore, the carbon number for the non-ethyl ene olefin fraction
must be approximated.  The value used for this study was 4.0.  Also the HCHO/
RCHO fraction needs to be determined or assumed, as well as the carbon number
for RCHO.  A value of 33%/66% was assumed.   Since the  aldehyde carbon
number determined from the emission inventory was 2.0, the carbon number
for RCHO was assumed to be 2.5.
                                     159

-------
Species             Fraction           Carbon No.

 OLE              (non-ethylene         RAPS OLE
                   olefin factor)
 used 4.0                                 RAPS
                                       non-ethylene
                                          olefin
 ETH              (1  - non-ethylene    RAPS OLE 2.0
                    olefin factor)
 ALK              RAPS PAR             RAPS PAR

 ARO              RAPS ARO             RAPS ARO
 HCHO             1/3 (RAPS ALD)        1.0

 RCHO             2/3 (RAPS ALD)        2.5
                       160

-------

V_Q
(S,^
OJ


ID
I*-**.
> °J
Q
OO
o_
5? 00
co
«= Q
cu Q
cu
-C
^~^ I—
S— *^

4- Q
C
O
T3 z r—
o  0 O •— O
o o o o o
en o m o «*•
in m r--- co co
CM r^ co in o
ID O O i — O
o o o o o

en in co oo «d-
in o in in in
en in CM in o
r~- o o o o
o o o o o

«3- m oo r~v oo
in in CM co ^j-
r~^ CM 10 co i —
LO •— 0 O O
o o o o o

i — o Ln i — r~~
ID co r— i— en
CD co ^ «=r o
U3 O O r— O
O O O O 0 1—
1—4
o

*a- o in o en
i — i^ co § — en
i — CO «3" CO O
 r— O O O
CD O O O O


r^* o CD co ID
CO ID CO i — OO
CM f-^ co in o
VD O O i— O
CD O O O O

to LO r-. CM r-
in en *a- en «d~
in ID co in o
r^ o o o O
                                                            in in    i—m
                                                         r^. CM CM en ID co
                                                         co en en «d- o i—
                                                         in o o r— o o

                                                         o o o o o o
        o o o o
                               o o o o o
                                                         ID in in    i-- in
                                                         o en en o CM in
                                                         r--. us VD i—o o
                                                         o o o o o o

                                                         O O C5 O O O
C  5-
O
•e
(0
o



E
£
13

CU
o
cu
Q.
O
VD


«£
Q-

O
CO


LU
cl

r- O
00 r—


O O
o; _i

-------
7.3.4  Sensitivity of Demerjian  and  Carbon  Bond  II  Mechanisms  to  RAPS
       Ambient HC Composition  in Smog  Chamber Type  Simulations
     Figures 70 and 71  show simulations  of  BOM-type outdoor  smog  chamber
experiments at 0.2 ppm NO  and 1.0 or  2.0 ppmC NMHC initially,  using  the
                         /\
Demerjian and CB2 mechanisms and four  HC compositions:   1) BOM  auto exhaust,
2) RAPS D261, 3) RAPS D221, and 4) RAPS  D238.
     The compositions of the BOM exhaust and the three  RAPS  days,  representing
the days of highest aromatic,  olefin,  and paraffin  fractions were:
            Day     NR      PAR      OLE     ARO
            261    0.100   0.552    0.150   0.181
            221    0.137   0.483    0.251   0.099
            238    0.0675  0.755    0.101   0.066
            BOM    0.105   0.417    0.173   0.287

     Each figure shows simulations performed with a single mechanism.  In
each figure, columns show the same initial NMHC concentration and each row
shows a different HC composition.
     Generally the BOM composition yielded the highest predicted Oq concen-
trations for a given mechanism and initial NMHC condition.  Although the
two mechanisms agreed within 20% for both levels of initial HC utilizing
the BOM auto exhaust composition, the mechanisms respond differently to
changes of composition.  The CB2 mechanism, for example, predicted that the
high  aromatics day (D261) at 2.0 ppmC NMHC, would produce as much Oq as
the CB2 2.0 ppmC BOM-composition system, while the Demerjian mechanism shows
a substantially reduced Oq yield for the same change in composition.  The
same change in composition would also have a dramatic effect at the 1.0 ppmC
initial level, with the Demerjian mechanism again showing the greatest effect,
yielding less  than half the 03 than predicted by the CB2 mechanisms.
                                    162

-------
                                                          0.50
             "' I  ' I '  I''  I '  I ' I "r I/1  ' HJ '  I ' I  ' i  ' I  ' ] °-5:
                 Df-M - 2.0 PPMC NMHC        ^             '
                      i | i  | i  p-] ' |  ' I  ' I '  I '  rr
                                                                                 I,  n-r+ i  I i I -t I  i I  i I  i J
                                                                                                       - 0.10
                                                                                                       - 0.30 P
                                                                                                       ~ 0.20
                                                                                                       - 0.10
  0.00
                                               l-J 0.00
         6  7  8  9  10 11 12 13  11  15  16 17 18  19  20
                       HOURS, LDT
            5  6  7  8   9  10  11  12 13 11  IS  16  17 18 19 20
                              HOURS. LDT
  O.SO
  0.10
  0.30
  0.20
  0.10
  0.00
          Trn ' i  • i '  i
          DEM - 1.0 PPMC NMHC
      -   HIGH RRO - DflY 23B
         6   7  8  S 10 11 1?  13  11  lli  16 17 18  Ifl  20
                       HOURS, LOT
0.50      0.50
         0.15
0.10      0.10
       e 0.35
0.30 P  n 0.30
    I  I 0.25
o.rc    j 0.20
       B-»-«
0.10      0.10
         o.os
0.00      0.00
  I  ' I  ' I '  I '  I '  I ' I  ' I
    DEM - 2.0 PPMC NMHC
-  HIGH flRO - DflY 261
                                                                  NO
                                                                                                         0.50
                                           0.10
                                                                                                         0.30
                                                                                                         0.20
                                                                                                         0,10
            S  6  7  8  9  10  11  12  13 11 IS  16  17  18 19 20
                              HOURS. LDT
  C.50
  0.10
  0.30
c5 0-20
 .
  0.10
  o.oo
          Di-M - 1.0 PPMC NMKC
          HICi( OLE - DRY 238
                                    PflN
0.50     0.50
        0.15
0.10
                                                                                                         0.50
                 9 10 11  12  13  11 IS 16 17  18  19 20
                       HOURS. LOT
                                                 0.00
            5  6   7  8  9  10 11 12 13  11  IS  16 17 18 19  20
                             HOURS. LDT
                                                                                                        0.00
V.3U
0.10
0.30
0.20
0.10
0.00
"' 1 ' i r 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 '
DEM - 1.0 PPMf NMHC
- HIGH PRR - DRY 238
-
NO
~~ ^^> ~
i- HO, ______——-
,i ,i ,i , i i.i . _i_, i- ,_i .t4- .--I'-.-i-L i , i , i i
5 6 7 8 9 10 11 12 13 11 15 16 17 18 13 2
u.ou
0.10
0.30
0.20
0.10
0.00
0
HOURS. LDT
        0.50
        0.15
                                                                                                        O.SO
                                                         0 00
                                                             S  6  7  8  9 10  11  12  13  11 IS 16 17  18  19 20
                                                                              HOURS. LDT
                                                                                                        0.00
     Figure  70.    Effect  of  HC  composition  and  concentration  on  outdoor  smog
                      chamber  type  simulations  with  the  Demerjian  Mechanism.
                                                      163

-------
                 I  I I .  I , I  I I I-1 ,  , . I  ,-J
                      -l7r-4-i-l-5jnrrT. I  , I .  I . I
                                              - 0 30 P    0 30 -
 rf 0 20  -
   0. 10  -
                                              -1 0 20    j 0 20 -
                                                                                                    - g 10 P
  0 00
      5  6  1  8  S 10 11  12  13 II  15  16 11 18  19 20
                       HOURS  LOT
                                                         0.00
                                                            5  6  1   8  9 10 II  12  13 11 IS  16 11 18 19  20
  0 50
                      p~r pr-pr
          CB2 - 1 0 PPMC l.'MHf
                                                                CB2 - 2 0 PPHC N1HC
          HIGH flRO - BUY 261
                                                                HIOH flRO - DflY 261
      5  6   1  8  9 10  11  12  13 II  15  16 1!  16  1« 2C
                                                            5  6   7  8  9 10  II  12 .3 M  15  16 11 18  15 ?0
  0 50
  0 10
d ° 2°
  0  10
         CB2 - 1  0 PPMC NMHC

         HIGH OLE - OflY 221
  0 00
5  6   7  8  9 10  11  12 13 II
                 HOURS  IDT
                            =._-J^-]_i_L_i_l_i_
                              it n  is  19 /
                                                0 SO
                                                  0 SO i  i r i  | i |  i | i  [ i  |-i~
                                                          C82 - 3 0 PPMC NM<'C
                                                  0 10
                                                 e
                                                 £ 0 30
                                              i   I
                                          020    g°2°

                                          0 !0    So,o
                                                            -   HIGH OLE - OrtY 221
                                               1 f.
                                                        0 00
                                                                NO
                                                               6   1  8  9  10  11  12 13 H  15 16 II
                                                                             HOURS  i nr
                                                                                                  15 20
   0 50
   0 10
         I '  I ' I  '' I  ' I '  I "H
          CB2 - 1 0 PPM NMHC

          HIGH PflR - DOY 238
   0.30  -
 rf 0 20
   0 00
0 50     0 50



0 10     0 10

       E
       a
0 30 P  " 0 30



       2T


r, in     0 10
                                                         |  I | I  | ' |  ' |  ' I I  | ' I  ' |  I
                                                           CB2 - 2 0 PPMC NMHC

                                                           HIGH PflR - DflY 258
                                                                 HO
                                                                                                      0 30 P
                                                                                                      0 ?3
                                                                                                      0 IT
      Figure  T\ .    Effect  of  HC  composition  and  concentration  on  outdoor  smog
                      chamber  type  simulations  with  the  Carbon  Bond  II  Mechanism
                                                      164

-------
     The switch from BOM composition  to the  high  OLE  day composition  (D221)



also illustrates the difference in directional  results.   Both  mechanisms  at



the 1.0 ppmC NMHC system showed similar decreases in  03  yield, but at the



higher 2.0 ppmC level systems the Demerjian  mechanism indicated that  the



high OLE would be slightly less reactive system than  the BOM composition



system and the CB2 would be slightly  more reactive  than  predicted  with



the BOM composition.



     Similar effects are seen with the high  PAR day (D238).  The initial



1.0 ppmC NMHC system predictions were most dramatic for  both mechanisms.



     Figure 72 shows the effect of changing  from a  summer to fall  (October)



distribution of photolytic rates, for the Demerjian mechanism  at 2 ppmC



initial HC.  The aldehyde photolysis  rate constants change most dramatically:



decreasing about 40%.  There is little effect shown for  the  more reactive HC



compositions (BOM and high olefin) but there is considerable effect in the



high paraffin composition.   Similar studies  with  the  Carbon  Bond II mechanism



showed less of an effect and are not shown here.
                                     165

-------
  0.50
       1  ' ' 1  ' I  ' I '  I '  1 ' I/1  "1-U '  I ' I  ' i '  I
          DEM - 2.0 PPMC NMHC/       \v
  0.00
         6  7  8  S  10 II  12  13  11  15 16  U  18  19 20
                       HOURS. LOT
                                                 o.oa
                                                         o so -,
                                                         0.10
                                                         0.30 -
                                                         0.20 -
                                                         0.10
                                                         0.00
                                                                     I  ' I '  I '  p T1 I' U-U ' I ' I  ' I '  I '
                                                                                                     - 0.10
                                                       6  1  8  9  10 11 12 13  11  IS 16 17  18  19 20
                                                                     HOURS. LOT
                                                                                                       0.00
  0.50
  0.55
  0.10 -
e 0.35
S0.30 h
I 0.25
c7 c-20
•j.
 -,..5
  o.io
  0.05
  0.00
I'I'I' I' I  ' I '  I '  I' I' I '  I '  I ' I' I
 DEM - 2.0 PPMC NMHC
 HIGH flRO - DflY 261
                                                O.EO
 NO
                                                                                                       O.SO
                                     - 0.10
                                                                                             - 0.30
                                                                                             - 0.20
                                                                                             - 0.10
        6  7  8  9  10  11  12 13 11  IS  16  17 18  19  20
                       HOURS. LOT
                                                    5  6  7   8  9 10 11  12  13  11  IS 16 17  18  19 20
                                                                     HOURS. LOT
                                                                                                       0.00
                                                0.50
                                              - 0.10
                                              - 0.30
                                              - 0.10
                   10  11  12
                       HOURS,
                   13  11  IS 16 17  18  IS 20
                    LOT
                                                0.00
                                                        HIGH OLE - DRY 221
                                                        PHOT RRTES FOR OCT LIGftT
                                                                                                        O.SO
                                                                                                      - 0.10
                                                                                                      - 0.30 P
                                                                                                      - 0.20
                                                                  7  8
       9  10 11  12  13  11 15 16 17  18  19 20
             HOURS, LOT
                                                                                                        0 00
  0.50
                                                O.SO
                                              - 0.10
                                              - 0.30
                                              - 0.20
  0 00
        6  7  B   9  10  11 12 13 11  IS  16 17 18  IS  20
                       HOURS. LOT
                                                0.00
                                                 0.51)
                                                 0.15
                                                 0 10
                                               E 0.35
                                               S 0.30
                                               1 0.25
                                               g 0.20
                                               * 0.15
                                                 0.10
                                                 o.os
                                                 0.00
1  1 ' I  ' 1  ' I '  I '
DEM - 2.0 PPMC NMHC
HIGH PHR - DflY 238
PHOT RflTES FOR OCT LIGHT
                                                                                    i  I i  I i I  I I I  I i
                                                                                              f1 I '  I
                                                                                                       0.50
                                                                                                        0.10
                                                                                                        0.30 P
                                                                                                        0.20
                                                                                                        0.10
                                                             567
                                                                9  10  11 12 13  11  IS  16 17 18  19
                                                                      HOURS. LOT
      Figure  72.    Effect of aldehyde  photolytic rate with HC  composition and
                       concentration  in  outdoor smog chamber  type  simulation  with
                       the  Demerjian  Mechanism.
                                                        166

-------
7.4  MIXING HEIGHT PROFILES
     The atmospheric boundary layer (ABL) is generally classified by its
stability.  The classification systems generally involve the depth of the
ABL, h and the Obukhov length (a buoyancy length scale), L.   Using this
system,the atmosphere can be divided into three general  categories:   unstable
or convective,    neutral or near neutral, and stable.  The  convective
boundary layer is usually observed during the daytime over land while stable
boundary layers are generally observed during the nighttime.   The neutral  or
near neutral boundary layer is rarely seen over land, except for momentary
episodes during the transition from the daytime to the nighttime boundary
layer and vice versa.
     In convective boundary layer regime, the buoyant  production of turbu-
lence dominates over the mechanical production.   The convective boundary
layer is characterized by three layers, a mixed layer, a transition  layer
and a cloud layer.  The mixed layer is characterized by  a uniform mean velo-
city profile, with no significant vertical gradients.  This  is due to a large
amount of mixing.   The transition layer is slightly stable due to the entrain-
ment downward of warmer and drier air from the cloud layer.   The transition
layer occurs around the height of the capping inversion, in  which there is
a noticeable jump in the wind speed across the inversion, with significant
wind shears in the cloud layer above.   The wind velocity in  the mixed layer
is nearly geostrophic (Yamada, 1976).
      The simplest methods to determine the depth of the boundary,  h, are
direct measurements, i.e., using Sondar (acoustic sounder) and Lidar (the
                                     167

-------
optical equivalent of radar).   The echoes detected by a monostatic acoustic
sounder are produced entirely by temperature fluctuations; with a bistatic
system the echoes are produced by both wind and temperature fluctuations.
The principal scattering sources for lidar echoes are aerosols and light
molecules.  During the daytime, these direct measuring devices may not have
enough power to reach the top of the boundary layer.   Other direct measure-
ments include turbulence sensors mounted on research  aircraft, high towers
or on tethered balloons.  These provide a means of measuring h directly.
Generally, these types of turbulence measurements are rarely made.  More
often, the ABL depth is estimated from the measured temperature and humidity
profiles obtained from radiosonde observations.
     If direct measurements are not available then climatic or historical
methods must be relied on.  These generally involve'the use of an early
morning radiosonde observation from the nearest National Weather Service
Radiosonde station.  Proposed methods include Holzworth (1972), Benkley and
Schulman  (1979) and EPA (1981).
     During the nighttime, the boundary layer becomes stable, and the depth
reaches a minimum.  The ABL is generally identified with a surface inversion.
The height defined by the observed mean potential temperature profile may
not correspond to the actual height of the ABL due to turbulent exchanges --
thus the need for direct measurements.
     The boundary layer height is highly variable in  both time and space.
This is due to horizontal roll vortices and convective cells.  Thus, the
average height of the interface between the turbulent ABL and the free
atmosphere should be determined by the averaging or smoothing of continuous
data; a single sounding can only give an instantaneous value (Arya, 1979).
     Methods for choosing mixing height profiles given radiosonde data are
                                     168

-------
likely to vary from investigator to investigator.  One objective of this
study was to examine how sensitive the control requirements might be to some
likely-to-be-used methods for estimating the mixing height profiles.
     The method that is built in to OZIPP assumes an exponential rise between
the 8 A.M.'mixing height and the afternoon mixing height.  The reason is
simple:  this profile results in a constant dilution rate.  It is an inade-
quate representation of the ABL behavior, however, and was not used in this
study.
     In the OAQPS EKMA study (Gipson, 1980), two alternative mixing height
profile methods were explored.   The first method, Linear, consisted of a
simple linear interpolation between the mixing height calculated from the
near-sunrise sounding and the mixing height calculated from the late-morning
soundings and between the late-morning mixing height and that calculated
from the afternoon sounding.  In the second method, Handpicked, the hourly
mixing heights were manually calculated by a meteorologist using the same
three soundings and hourly surface temperatures.  It was decided to include
these two methods in this study.  The data for the ten days were supplied
by OAQPS/EPA.
     The third mixing height profile method used in this study was the
Characteristic curve method.  The method is based upon a piecewise curve
thought to describe the characteristics of the mixing height rise.  The
numerical values for the curve  were derived from detailed data from about
40 days in the RAPS study by Schere and Demerjian (1980).  Table 28 gives the
numerical values of the curves  break points and Figure 73 shows the norma-
lized curve  plotted.  The curve has been normalized by fraction of total
growth in the mixing height versus fraction of the total daylight period.
Use of the curve requires the sunrise mixing height and the maximum afternoon
                                      169

-------
           Table  28.   The Demerjian/Schere  Characteristic Mixing
                       Height Growth Curve
Interval:

Fraction of
day:

Fraction of
growth:

Growth
rate:
= 0.0    0.07   0.14    0.33   0.50   0.70   0.90    1.0    2.0
= 0.0    0.02   0.10    0.588  0.85   1.00   1.00    0.5     0.5
=0.286  1.143  2.526   1.588  0.750  0.0    -5.0     6.0    0.0
        0.00
                0   0.1   0.2 0.3  0.4  0.5  0.6  0.7  O.d  0.9  1.0

               TSR                                     TSS

                                 FRACTION OF DAYLIGHT
                                                      1.00

                                                      0.90

                                                      0.80

                                                      0.70

                                                      0.60

                                                      0.50

                                                      0.10

                                                      0.30

                                                      0.20

                                                      0.10

                                                      0.00
                                                   TSR2
       Figure 73.   The characteristic growth  curve.

                                       170

-------
mixing height.  The maximum mixing height is assumed to occur at 70% of the
total daylight period.  Note that 10% of the growth occurs in 14% of the day
and that 85% of the growth occurs in 50% of the day.  Morning and afternoon
mixing heights for use with the curve were supplied by OAQPS/EPA.
       Table  29 lists the hourly values of mixing height for all three
methods for the ten days studied.
       The effect of a mixing height rise, assuming uniform mixing to the
mixing height, is three-fold:  1) the volume of the "box" is bigger and the
same mass in  a bigger volume means a decrease in concentration, i.e., dilu-
tion; 2) the  rate of increase of concentration due to emissions is less
because the volume is bigger; and 3) material that was above the mixing
height is entrained into the mixed layer and is diluted uniformly throughout.
As described  in Table 1, the first and third effects are implemented in the
model as a dilution rate while the second effect is accounted for by dividing
the emissions rate by the current mixing height.
     To calculate a dilution rate, the rate of change of the mixing height
is needed, as well as the mixing height; that is, the dilution rate is given
 •y'  k , = -r-  -nr , where k. = dilution rate,  h =  mixing  height
To obtain a smooth and continuous value for the mixing height and its rate
of change,rational spline functions with variable tension were used to fit
the values in Table 29.  The resulting mixing height profiles must be mono-
tonically increasing functions,that is, the derivative must be zero or posi-
tive everywhere.   The mixing height values were initially fitted with a low
tension which results in a cubic spline-like fit.  If the derivatives any-
where were negative, the fit was repeated with increasing tension until the
derivatives were positive everywhere.  The resulting curves for each of the
                                     171

-------
                                 i— r-~ r— O> CO LO ro r-^
                                 ro in vj- ro r— CD cr> CD
                                 i— CM LO co i— ro  en  ,— OOO
                                 <^t- «J LO UD  CO Cn i— CM

                                 OOOOOOr- r-^
                                                                                      O O CT> "X> •— •^f'-OOi— CMCTiCTiCTt
                                                                                      OOLooov-crocTiLOCMcrirororo
                                                                                      r— i— .— <^-GUDCnroir>'^>r-^r---r~-

                                                                                      O O O O r— i— •— CMCMCMCMCMCM


                                                                                      OOOOCOOOOOOOO
                                                                                      LOLOLOOOLOLOLOLni-OLOLOI-O
                                                                                      OO>Or— CMCOr^r— r— r— !— .— r—

                                                                                      OOOOOOOronororororo
                                                                                                                   CM CM CM CM CM CM OJ
                                 OOOOOOOO
                                 oooooooo
                                 r~- CO en o r— CM ro rj-
                                 OOOr— r— r-r-,—
                                                                                      ooooooooooooo
                                                                                      ooooooooooooo
                                                                                                OOOOr
                                         - i—  ro in UD
                                         - LD  no ,— c\j
                                          ix)  r^. co 03
                                 O O O r-  r— .-  r-

                                                                                                                i— CDcnr-—
                                                                                                OOOOOOOCDOOO


                                                                                                ooooooooooo
                                                                                                OOOCDOOCDOOOO
                                                                                                O«—  i— cxjcNjro^Lnin^or*-

                                                                                                ooooooooooo
                                 O O O O O O O
                                 O O O O O O O
                                 i— co ro ^j- LO 1^3 r-^
                                                                                      ooooooooooo
                                                                                      OOOOOOOOOOO
CTi
C\J
          a>
          'QJ
-ME     H
                                                    oocvioo
LOUOOOOOOOOO
O^O^CTiCTiCTlCriCTlCTiCTt
i— i— OOOOOOOO

OOC\JC\JC\JCMCXJC\JC\JC\J
                                 OOOOCDOOOOO
                                 ooOLncrtu~)kDooororo
                                 Oi— i— CNJ«d-l£)Oti— ,— ,—

                                 OOOOOOOf— i— .—
                                                                                                OUDCOi —
                                                                                                OOOOr
                                                                                      OOOOr— r— r— r-



                                                                                      COLOr— CO«^-COOr—
                                                                                      ^O«=tCMCnr~-CMCMr—
                                                                                      CM- COCO


                       OOOi— r—'^i—^r-^r—"r-^



                       OLOr—OOOOOO»O
                       LO «^- LO <~ ~
                       O r— i	 •


                       O O O r
                                                                                                r— r— O LO LO O i - . —
                                                                                                OOOOr— ,— r— r—
                                                                                                                ooo
                                                                                                                (IT)i— CM

                                                                                                OOOOOOi— .—
                                 oooooooooo
                                 oooooooooo
                                 cocnOr—cMm^j-LOLor--
                                 OO'	 >	  "	 r— ,	 i	 .	 r—
                                                                                       OOCDOO'^OO
                                                                                       OOOOO'OOO
                                                                                                OOO-— r— -—
                                                                           172

-------
                                                                                                         u
                                              !  '1H3I3H ONIXIW
                                                             '1HOI3H ONIXIW
                                                                                                         CO
                                                                                                         >>
                                                                                                         to
I  '   I
              I  '  I
                                                                                                         LO
                                                                                                         01
                                                                                                         o
                                                                                                         S-
                                                                                                         Q.
                                                                                                        CD

                                                                                                        OJ
                                                                                                        O)  •
                                                                                                        c:  
                                                                                                        X  S-
                                                                                                        •r-  3
                                                                                                        s:  o
                                                                                                        01
                                              (OCMCNJOJ—«—"C>
""(  '1H3I3H ONIXIW
                                                         ^'i  'J.HOI3H ONIXIW
                                            173

-------
                                                                                                  T3
                                                                                                   01
"I '1HOI3H ONIXIW
                                                         '1HOI3H ONIXIW
                                                                                                   o
                                                                                                   u
                                                                                                   to
                                                                                                  -a
«t-    O   O3    OJ   CD

CVJ    C\J   —'    —'   O

 ""I  'J.H2I3H  ONIXIW
                                                 I  '   I  '  I  '   I  '  I  '   I  '  I
                                                                                                   OJ
                                                                                                  4-
                                                                                                   O
                                                                                                   1-
                                                                                                   Q.
                                                                                                   CD


                                                                                                   O)
                                                                                                   en
                                                                                                   c:
                                                                                                   O)
                                                                                                   s-
                                                                                                   Z3
                                                                                                   CD
                                                          '1HOI3H ONIXIW
                                           174

-------
     ""(  '1HOI3H ONJXIW
                                                                                                             •O
                                                                                                             
-------
profile methods(linear, handpicked, and characteristic curve)  for each of
the days are shown in Figure 74, all  on the same scale.
     The different methods of estimating the mixing height profile do pro-
duce large temporal differences in the profiles.
     In the solution of the model  it is the dilution rate, rather than the
mixing height itself, that is of interest.   As has  been explained, both the
height and the rate of change determine the dilution rate.   Thus, if the
rate of change is constant, then the dilution rate  will decrease over time.
If this period is followed by one in which  the rate changes,  then the dilu-
tion rate can rapidly increase before decreasing  again.  To illustrate this
process, the dilution rates resulting from  the three profiles  for two days
are shown in Figure 75.  Obviously, the different dilution patterns will
have a strong influence on the concentration profiles.  Such  discontinuities
in dilution probably do not occur in the real  world.
                                      176

-------
DILUTION  RflTE.   /win
                                                DILUTION RflTE.   /»
                                                                                                  DILUTION RflTE.
                                                      '1HOI3H ONIXIU
                                                                                                    ""I  '1HOI3N ONIXIU
                                                                                                                                                 to
                                                                                                                                                 OJ
                                                                                                                                                 O
                                                                                                                                                 S-
                                                                                                                                                 Q.
                                                                                                                                                 CD

                                                                                                                                                 
 O
 O)
                                                                                                                                                 cn
DILUTION RrtTE.
                                                 DILUTION RflTE.   /mn
                                              I  .  I  .  I  .   I
                                                                                                  DILUTION RBTE
4J

 3
 
 CD
 s_

 
 Ol
+J
 fO
 s_

 c
 O
                                                                                                                                                 CD
                                                                                                                                                 en
  "1 'iHOI3H ONIXIU
                                                                                                    "1 '1HOI3H ONIXIW
                                                                        177

-------
7.5  OZONE ALOFT

     The  last  data needed to perform the  simulations are estimates of  63

aloft or  ozone that is above the mixing height in the morning that would  be

entrained  into the mixed layer as the mixing  height rises.  Direct measure-

ments of  this  value are rare.  It has been  shown that there is a relationship

between the  average 0^ aloft as measured  by helicopters and subsequent

reading of surface 63 monitors following  inversion breakup.  One example  of

this relationship is shown in Figure 76 (Evans,  1979).  The recommended  pro-

cedure for estimating the 03 aloft values is  described in the  EPA   guidance

document  (EPA, 1981).  The procedure basically consists of using 03  measure-

ments from surface monitoring sites upwind  of the city during the first  hour

after breakup  of the nocturnal inversion  as an estimate of the ozone aloft.

This technique was used by OAQPS/EPA to estimate the 63 aloft values for  the

days studied herein; these studies were supplied to UNC and are listed in

Table 30.

     In addition to 63 aloft, all simulations assumed a natural background

value of  0.20  ppm carbon monoxide aloft.

     Figure  76   AVERAGE OZONE CONCENTRATION IN EARLY VERTICAL PROFILE
       3             VS UPWIND POST-BREAKUP CONCENTRATION
                     8.15 —|	"~~~	?
                                         A DAY 201
                                         O D»Y 305
                                         a DAY 216
                                         * OM217
                                         O DAY 22i
                                         V DAY 225
                                         X DAY 228
                               a.05      8.10      e.15
                      SURfACE UPWIND HOUR AVERAGE OZONE CONCENTRATION
                            FOLLOWING INVERSION BREAKUP
                                       178

-------
                   Table 30

       Assumed Ozone Aloft Used in Models
              of Selected RAPS Days
                       (ppm)
       Day                      Concentration


159 - June 7                        0.12

160 - June 8                        0.09

195 - July 13                       0.07

201 - July 19                       0.08

221 - August 8                      0.07

238 - August 25                     0.09

261 - September 17                  0.06

275 - October 1                     0.06

276 - October 2                     0.08

-------
             8.0 RAPS SIMULATIONS WITH SIMPLE TRAJECTORY MODELS



8.1 METEOROLOGY-ONLY SIMPLE TRAJECTORY MODEL SIMULATIONS OF RAPS DAYS


8.1.1  Brief Description



     The Meteorology-Only Simple Trajectory Model (MOSTM) has no chemistry.



That is, the concentrations of the species of interest are effected by the


initial concentrations, their boundary values (the concentrations transported


and entrained in the mixed layer) and their emissions; all of which are di-



luted as the mixing height rises.  Thus, the rate of change in the concentra-



tion is prescribed by the equation:





               dci = - 1  dh  r     1  dh /r_nn  „„ .  +  Ei
               	     -  —  C-j +  -  — (FRA-;  CA,-)     —

               dt      h  dt        h  dt               h




      where,




               dC-j
               —   is the rate of change of the concentration of the


                     ith species



            -  —    is the dilution rate

            h  dt


                h    is the mixing height



               C-j    is the concentration of the ith species



             FRAj    is the fraction aloft of the ith species at time t



              CA-j    is the concentration aloft of the ith species


               Ej    is the emission flux of the ith species



                                      180

-------
     A meteorology-only simulation gives an indication of the reasonableness



of the input data.  If the precursors are not predicted"correctly, then the



Oo is not likely to be predicted correctly.  Since adding chemistry to the



simulation acts as a sink for NO  and HC and a source for 0.,, the predictions
                                X                          o


of NO  and HC concentrations in MOSTM should be greater than the observed
     /\


concentrations along the trajectory.  Carbon monoxide predictions should be



close to the ambient data if the emissions and mixing height data



were accurate.  The magnitude of 0. aloft entrainment can also be seen.



When the meteorology-only simulation underpredicts the observed NO  and HC
                                                                  X


concentrations, the input data should be reexamined.



8.1.2  Results of MOSTM Simulations



     Figure 77 shows the model predictions as solid lines and the ambient



data that were shown in Figures  55 - 64 are shown here as symbols and



dashed lines.  One mixing height profile was used for each row and a single



day is in each column.   One page shows the NMHC and CO predictions and the



facing page shows the NO  and 0.,.  Remember that the 0, model predictions
                        X      j                      o


are the result of initial concentrations and entrainment from aloft as the



mixing height rises and that, since there was no chemistry, there is no



reaction of the NO with the 0-.



     In the June 7 (D159) model  predictions, the first column of Figure 77,



the only differences in input data were the three mixing height profiles



(see Figure 74).  These do result in significant differences among the simu-



lations.   In the LIN simulation, the CO is nearly constant at about 0.7; in



the HP (handpicked) simulation,  the CO decreases sharply then rises to a peak



greater than 1 ppm and  then falls to about 0.4 ppm; in the CHAR simulation



the CO decreases slowly all  day.



     A comparison of the emission pattern for this day (Figure 69) with the



                                      181

-------
mixing height profiles (Figure 74) reveals the causes of these different



predictions.  The highest CO emissions were for the 8-9 A.M.  hour;  from



9 to 13 the emissions were almost constant and they decreased after 1300



LOT.



     The linear mixing height profile was the highest at 8 and it changed



only a small amount over the 8-9 hour.  The dilution caused by the  rise was



almost exactly compensated for by the emissions so the concentration did not



change very much.  After 9, the CO emissions decreased and likewise the



dilution rate decreased all day, resulting in little change in the  CO



concentration.



     The HP mixing height profile was tiie lowest at 8 and it  tripled its



height in 8-9 hour, resulting in a factor of three dilution for the initial



CO concentration.  This was more than compensated for by the  CO emissions.



Even though these were the same magnitude as in the LIN case, they  went into



a volume that was initially 15 times smaller and at 0900 LOT, was still six



times smaller. Thus, the emissions into this smaller volume were able to



rapidly increase the CO concentration.  In the hour from 10 to 11,  however,



the volume increased by nearly a factor of 10, resulting in a rapid and



dramatic decrease in CO concentration.  After 1100 LOT, there was no change



in mixing height and therefore no dilution.   The CO emissions into  the large



volume were able to slightly increase the CO concentration.



     Since the CHAR m'xing height profile is in between the LIN and HP it



exhibits some characteristics of both.  Its  initial height and rate of rise



prevents the large increase in CO concentration that was seen in the HP case.



The dilution rate, however, is larger than in the LIN case, resulting in a



rapid initial drop.  Because the initial volume was smaller than in the LIN



case, less CO mass was present,   even though both simulations had  the same



                                      182

-------
                                                          -o
                                                           o
                                                           O)
                                                           S-
                                                           fC
                                                           01
                                                           c:
                                                           O
                                                          oo
                                                           rd
                                                          -o
                                                           ro
                                                          o
                                                          C_3
                                                          rc
                                                           x
                                                          o
                                                           o
                                                          <+-

                                                           
                                                          O  i-
                                                           
-------
                                                 O)
                                                 rs
                                                 c:
                                                 c:
                                                 o
                                                 o
                                                 O)
                                                 S-
184

-------
            1  S  I

I , I I I I I | I I I  I I  , I
                            5  S  5 S3
  I ' I ' I ' I ' I ' I ' I
                                                                                         QJ
                                                                                         3
                                                                                         c
                                                                                         o
                                                                                         u
                                                                                         (U
                                                                                         i.
                                                                                         3
                                                                                         CD

                                           185

-------
                                                           L_LJ__LJ
                                                                              -£
                                                                                  i I .  I i I  i I   1,1,1,1,1,,
                         -I-
;-


f-
                                                                                                                          -o
                                                                                                                           cu
                                                                                                                           O
                                                                                                                           
-------
                                                                         1:
            -1-
s  I

                                                                            : 1
                                                                                     -a
                                                                                     cu
                                                                                     c
                                                                                     o
                                                                                     o
                                                                                     
-------
                                       0,. ppm
                   0:
i . i . i .  i . i . i
                                         • r'-i ' i • r
                                                                                                        o
                                                                                                        o
                                                                                                        
-------
     *-t-' I ' I ' I ' I ' I ' 1 -
              if
$~\~^'
'Y i i i i i i i i i i i i i i i i i ,'
                                                     CO. ppmC
                                                                    -a
                                                                     O)
                                                                     o
                                                                     o
                                                                     
-------
         0.. ppm
rr-p-| ' i ' i
                                                                                                            O)
                                                                                                            rs
                                                                                                            o
                                                                                                            o
                                                          190

-------
initial CO concentration.  The final  volumes, however, were similar,  there-



fore, the concentration was less in the CHAR simulation.



     To aid in these interpretations, Table 31  has been prepared.   In MOSTM



simulations there are three sources of mass:  initial  material, emissions,



and material aloft that is entrained.  There are no losses of mass,  only



increase in volume which results in decrease of concentrations (mass/unit



volume).  MOSTM keeps track of how much of the  final  concentration resulted



from which process; these values are  shown in Table 31.  For Day 159



(June 7), for example,  46.6% of the final  CO concentration was the result



of the initial amount of CO specified.   That is, if there had been no emis-



sions and no CO aloft,  the final CO concentration would have been  0.304 ppm.



     The amount of material that results from entrainment varies with mixing



height profile (see Table 31).   This  is because the ratio of the difference



between final  and initial heights to"  the final  height  is  quite different



for each mixing height  profile.   That is,  the mass of  ozone at the end of



the simulation is given by:







     ozone mass = [03]f* MHf= [03]A[_  *  (MHf - m.) +  [03].  * m.






     where, [O^L - final ozone concentration in mixed region




            [O-L.  = aloft ozone concentration  (assumed constant)





            [0,].  = initial  ozone  concentration in mixed region
              v5 1




            MHf    = mixing height  final





            MH.    - mixing height  initial




                                     191

-------
Table 31.   Effects  of  Mixing Height  Profile on  the  Contribution of  Ea..h
               Source to Final  Concentrations  in  MOSTM  Simulations, All  Days


                                             Day  159
                                              (ppm)
 Mixing Height
 Profile
                Spec IBS
     INITIAL
Amount  Percentage
    EMISSIONS
Amount  Percentage
   ENTRAPMENT
Amount  Percentage
    Final
Concentration
 Linear
CO
MO
N02
HC
03
0.304
0.003
0.012
0.023
0.014
46.6
16.6
86.3
14.9
16.8
0.290
0.017
0.002
0.130
0.000
44.4
83.4
13.7
85.1
0.00
0.058
0.000
0.000
0.000
0.070
9.0
0.0
0.0
0.0
83.2
                                                                                       0.652
                                                                                       0.021
                                                                                       0.014
                                                                                       0.152
                                                                                       0.084
 Handpickea
CO
NO
N02
HC
03
0.018
0.000
0.001
0.001
O.Q01
4.8
1.3
29.5
1.2
0.7
0.264
0.016
0.002
0.118
0.000
69.5
98.7
70.5
98.9
0.0
0.098
0.000
0.000
0.000
0.117
25.7
0.0
0.0
0.0
99.3
                                                             0.380
                                                             0.016
                                                             0.003
                                                             0.119
                                                             0.118
 Characteristic
CO
NO
N02
HC
03
0.105
0.001
0.004
0.008
0.005
22.5
6.7
69.5
5.9
4.6
0.273
0.017
0.002
0.124
0.000
59.3
93.3
30.5
94.1
0.0
0.086
0.000
0.000
0 . 000
0.103
18.3
0.0
0.0
0.0
95.5
                                                                                       0.469
                                                                                       0.018
                                                                                       0.006
                                                                                       0.132
                                                                                       0.108
                                           Day 160 SUe 103
                                                (ppm)
 Mixing Height
 Profile
                 Sp_e_c_i_es
     INITIAL
Amount  Percentage
     EMISSIONS
Amount  Percentage
   ENTRAPMENT
Amount  Percentage
    Final
Concentration
 Linear
CO
NO
N02
HC
03
0.044
0.001
0.004
0.003
0.001
4.7
1.7
36.8
0.6
1.2
0.810
0.063
0.007
0.460
0.000
85.7
98.3
63.2
99.5
0.0
0.091
0.000
0.000
0.000
0.082
9.6
0.0
0.0
0.0
98.8
                                                              0.945
                                                              0.064
                                                              0.011
                                                              0.463
                                                              0.083
 Handpicked
CO
NO
N02
HC
03
0.002
0.000
0.001
0.001
0.000
2.6
1.1
2.7
0.4
0.3
0.341
0.026
0.003
0.194
0.000
75.6
98.9
73.3
99.6
0.0
0.098
0.000
0.000
0.000
0.088
21.7
0.0
0.0
0.0
94.7
                                                              0.450
                                                              0.027
                                                              0.004
                                                              0.194
                                                              0.118
 Characteristic
CO
NO
N02
HC
03
0.037
0.001
0.003
0.002
0.001
6.9
2.9
49.4
1.0
1.0
0.410
0.032
0.004
0.233
0.000
76.0
97.1
50.7
99.0
0.0
0.092
0 000
0.000
0 000
0 083
17.1
0.0
0.0
0.0
99.0
                                                              0.539
                                                              0.033
                                                              0.007
                                                              0.235
                                                              0.084
                                           192

-------
Table  31.   continued
                                               Day 160 Site 115
 Mixing Height
 Profile
     INITIAL
Amount  Percentage
     EMISSIONS
Amount   Percentage
   ENTRAPMENT
Amount  Percentage
    Final
Concentration
 Linear
CO
NO
N02
HC
03
0.014
0.000
0.001
0.001
0.000
3.7
2.7
41.7
0.6
5.0
0.279
0.009
0.001
0.129
0.000
72.3
97.3
58.2
99.4
0.0
0.093
0.000
0.000
0.000
0.084
24.1
0.0
0.0
0.0
95.0
                                                                   0.386
                                                                   0.010
                                                                   0.002
                                                                   0.130
                                                                   0.088
 Handplcked
CO
NO
N02
HC
03
0.010
0.000
0.001
0.001
0.003
3.9
3.6
48.5
0.8
3.5
0.151
0.005
0.001
0.770
0.000
59.0
96.4
51.5
99.2
0.0
0.095
0.000
0.000
0.000
0.086
37.1
0.0
0.0
0.0
96.5
                                                                   0.256
                                                                   0.005
                                                                   0.001
                                                                   0.011
                                                                   0.089
 Characteristic
CO
NO
N02
HC
03
0.061
0.001
0.003
0.004
0.019
20.2
16.5
83.5
4.3
23.1
0.170
0.006
0.001
0.079
0 000
56.6
83.5
16.5
95.7
0.0
0.079
0.000
0.000
0.000
0.063
23.2
0.0
0.0
0.0
76.9
                                                                   0.301
                                                                   0.007
                                                                   0.004
                                                                   0.082
                                                                   0.082
                                                    Day  195
Mixing Height
Profile
Linear




§pecies
CO
NO
N02
HC
03
INITIAL
Amount
0.027
0.001
0.001
0.003
0.014
Percentage
8.9
3.5
23.1
2.5
20.8
EMISSIONS
Amount
0.205
0.019
0.002
0.106
0.000
Percentage
66.4
96.5
77.9
97.5
0.0
ENTRAPMENT
Amount
0.076
0.000
0.000
0.000
0.053
Percentage
24.7
0.0
0.0
0.0
79.2
Final
Concentration
0.308
0.019
0.003
0.108
0.067
 Handpicked
CO
NO
N02
HC
03
0.042
0.001
0.001
0.004
0.022
10.9
3.9
25.2
2.9
32.7
0.281
0.025
0.003
0.150
0.000
72.7
96.1
74.8
97.2
0.0
0.063
0.000
0.000
0.000
0.044
16.4
0.0
0.0
0.0
67.3
                                                                                               0.386
                                                                                               0.027
                                                                                               0.004
                                                                                               0.149
                                                                                               0.066
 Characteristic
CO
NO
N02
HC
03
0.075
0.002
0.002
0.008
0.039
23.8
8.9
44.9
6.6
61.7
0.208
0.019
0.002
0. 10?
0.000
65.4
91.1
55.1
93.4
0.0
0.034
0.000
0.000
0.000
0.024
10.8
0.0
0.0
0.0
38.4
                                                                                               0.318
                                                                                               0.021
                                                                                               0.004
                                                                                               0.115
                                                                                               0.063
                                               193

-------
Table  31.   continued
                                                     Day 201
Mixing Height
Profile
 Linear
 Handpicked
 Characteristic
                INITIAL
Species    Amount  Percentage
CO
NO
H02
HC
03
0.058
0.001
0.002
0.004
0.020
16.2
3.8
41.5
3.5
28.6
0.236
0.024
0.003
0.117
0.000
66.0
96.2
58.6
96.5
0.0
0.064
0.000
0.000
0.000
0.051
17.8
0.0
0.0
0.0
71.4
0.058
0.001
0.002
0.004
0.020
0.011
0.000
0.000
0.001
0.004
0.013
0.000
0.001
0.001
0.005
16.2
3.8
41.5
3.5
28.6
3.9
1.0
15.1
0.9
4.9
4.6
1.2
17.5
1.1
5.9
                                                       EMISSIONS
                                                  Amount   Percentage
                                             ENTRAINMENT
                                          Amount  Percentage
CO
NO
N02
HC
03
0.011
0.000
0.000
0.001
0.004
3.9
1.0
15.1
0.9
4.9
0.176
0.018
0.002
0.087
0.000
62.8
99.0
84.9
99.1
0.0
0.093
0.000
0.000
0.000
0.075
33.3
0.0
0.0
0.0
95.1
CO
NO
N02
HC
03
0.013
0.000
0.001
0.001
0.005
4.6
1.2
17.5
1.1
5.9
0.176
0.018
0.002
0.087
0.000
62.7
98.8
82.5
48.9
0.0
0.092
0.000
0.000
0.000
0.074
32.7
0.0
0.0
0.0
94.2
                                               Final
                                           Concentration
                                                                                                 0.357
                                                                                                 0.025
                                                                                                 0.005
                                                                                                 0.121
                                                                                                 0.0071
                                                                   0.280
                                                                   0.018
                                                                   0.002
                                                                   0.088
                                                                   0.078
                                                                   0.281
                                                                   0.018
                                                                   0.002
                                                                   0.088
                                                                   0.078
                                                      Day  221
 Mixing Height
 Profile
    INITIAL
N jou nt_  Percen tn y_e
     EMISSIONS
Ainou rit_  Percentage
                                                                          ENTRAINMENT
                                                                         o^Tt_  Percentage
                                                                               Final
                                                                           Concentration
Linear
CO
NO
N02
HC
03
0.068
0.002
0.002
0.004
0.006
16.6
7.4
43.8
4.2
11.7
0.278
0.022
0.002
0.097
0.000
67.9
92.6
56.2
95.8
0.0
0.063
0.000
0 000
0.000
0.044
15.5
0.0
0.0
0.0
88.3
                                                                                                 0.409
                                                                                                 0.023
                                                                                                 0.004
                                                                                                 0.102
                                                                                                 0.050
Handpicked
CO
NO
N02
HC
03
0.006
0.000
0.000
0.000
0.001
1.7
0.8
7.1
0.4
0.8
0.249
0.019
0.002
0.087
0.000
70.8
99.2
92.9
99.6
0.0
0.097
0 000
0 000
0 000
0 068
27.5
0.0
0.0
0.0
99.2
                                                                    0.352
                                                                    0.020
                                                                    0.002
                                                                    0.088
                                                                    0.068
Characteristic
CO
NO
N02
HC
03
0.027
0.001
0.001
0.002
0.002
6.9
3.1
23.8
1.9
0.037
0.272
0.021
0.002
0.095
0.000
70.8
96.9
76.2
98.3
0.0
0.086
0,000
0.000
0 000
0.060
22.3
0.0
0.0
0.0
96.3
                                                                                                 0.304
                                                                                                 0.022
                                                                                                 0.003
                                                                                                 0.097
                                                                                                 0.062
                                                194

-------
Table  31.    continued
                                                   Day  238
 Mixing Height
 Profile
 Linear

Species
CO
NO
NO
HC
03
INITIAL
Amount Percentage
0.081 34.6
0.001 23.0
0.002 75.5
0.003 4.7
0.001 1.8
EMISSIONS
Amount Percentage
0.067 28.6
0.005 77.0
0.001 24.5
0.068 95.3
0.000 0.0
ENTRAINMENT
Amount Percentage
0.086 36.8
0.000 0.0
0.000 0.0
0.000 0.0
0.078 98.2
Final
Concentration
0.234
0.006
0.002
0.071
0.079
 Handpicked
CO
NO
N02
HC
03
0.020
0.000
0.000
0.001
0.000
11.4
7.6
45.8
1.3
0.4
0.061
0.004
0.001
0.062
0.000
34.3
92.4
54.2
98.7
0.0
0.097
0.000
0.000
0.000
0.087
54.3
0.0
0.0
0.0
99.6
0.1778
0.004
0.001
0.063
0.087
 Characteristic
CO
NO
N02
HC
03
0.072
0.001
0.001
0.003
0.001
33.6
24.9
77.4
5.1
0.5
0.053
0.004
0.000
0.054
0.000
25.1
75.1
22.6
94.8
0.0
0.088
0.000
0.000
0.000
0.079
41.3
0.0
0.0
0.0
98.5
0.213
0.05
0.002
0.057
0.080
                                                    Day 261
 Mixing Height
 Profile
  Linear

Spec ?os
CO
NO
N02
HC
OC
INITIAL
Amount Percentage
0.078 21.4
0.006 13.4
0.004 47.5
0.005 1.4
0.002 3.3
EMISSIONS
Amount Percentage
0.575 69.0
0.041 86.6
0.005 52.5
0.331 98.6
0.000 0.0
ENTRAINMENT
Amount Percentage
0.080 9.6
0.000 0.0
0.000 0.0
0.000 0.0
0.048 96.7
Final
Concentration
0.833
0.047
0.009
0.336
0.049
  Handpicked
CO
NO
N02
HC
03
0.040
0.001
0.001
0.001
0.000
4.8
2.8
14.5
0.3
0.6
0.682
0.048
0.005
0.393
0.000
83.5
97.1
85.6
49.7
010
0.010
0.000
0.000
0.000
0.057
11.7
0.0
0.0
0.0
99.4
0.817
0.049
0.006
0.394
0.053
  Characteristic
CO
NO
N02
HC
03
0.058
0.002
0.001
0.002
0.001
9.1
5.6
25.8
0.5
10.9
0.048
0.034
0.004
0.280
0.000
76.3
94.4
74.2
99.5
0.0
0.093
0.000
0.000
0.000
0.056
14.6
0.0
0.0
0.0
99.1
                                                                                              0.638
                                                                                              0.036
                                                                                              0.005
                                                                                              0.282
                                                                                              0.057
                                               195

-------
Table  31.   continued
                                               Day  275 @ 8 am
 Mixing Height
 Profile
               INITIAL
Species    Amount  Percentage
     EMISSIONS
Amount  Percentage
   ENTRAINMENT
AmouiVt  Percentage
    Final
Concentration
 Linear
CO
NO
N02
HC
03
0.712
0.033
0.013
0.043
0.001
38.4
25.1
53.3
10.4
2.4
1.067
0.099
0.011
0.368
0.000
57.5
74.9
46.7
89.6
0.0
O.C77
0.000
O.COO
0.000
0.045
4.2
0.0
0.0
0.0
97.6
                                                                              1.856
                                                                              0.132
                                                                              0.0234
                                                                              0.411
                                                                              0.048
 Handpicked
CO
NO
N02
HC
03
0.177
0.008
0.003
0.011
0.000
14.6
8.7
24.4
3.2
0.5
0.940
0.087
0.010
0.324
0.000
77.6
91.3
75.6
96.8
0.0
0.094
0.000
0.000
0.000
0.057
7.8
0.0
0.0
0.0
1.0
                                                                              1.211
                                                                              0.095
                                                                              0.013
                                                                              0.335
                                                                              0.057
 Characteristic
CO
NO
N02
HC
02
0.645
0.030
0.011
0.039
0.001
28.1
17.1
41.3
6.7
2.1
1.569
0.145
0.016
0.541
0.000
68.4
82.9
58.7
93.3
0.0
0.080
0.000
0.000
0.000
0.048
3.5
0.0
0.0
0.0
97.9
                                                                              2.293
                                                                              0.175
                                                                              0.027
                                                                              0.580
                                                                              0.049
                                                     Day 275
 Mixing Height
 Profile
  Linear
INITIAL
Species
CO
NO
N02
HC
03
Amount Percentage
0.108
0.007
0.002
0.006
0.000
6.8
5.4
13.4
1.3
0.7
EMISSIONS
Amount Percentage
1.387
0.126
0.014
0.493
0.000
87.4
94.6
86.6
98.7
0.0
ENTRAINMENT
Amount Percentage
0.092
0.000
0.000
0.000
0.055
5.8
0.0
0.0
0.0
99.3
Final
Concentration
1.587
0.133
0.016
0.500
0.056
  Handpicked
CO
NO
N02
HC
03
0.074
0.005
0.002
0.004
0.000
5.3
4.3
10.8
1.0
0.5
1.222
0.111
0.012
0.435
0.000
87.9
95.7
89.2
99.0
0.0
0.094
0.000
0.000
0.000
0.057
6.8
0.0
0.0
0.0
99.5
                                                                              1.391
                                                                              0.1556
                                                                              0.014
                                                                              0.439
                                                                              0.057
  Characteristic
CO
NO
N02
HC
03
0.248
0.017
0.005
0.015
0.001
10.5
8.3
19.5
2.0
1.9
2.041
0.185
0.021
0.726
0.000
86.1
91.8
80.5
98.0
0.0
0.081
0.000
0.000
0.000
0.049
3.4
0.0
0.0
0.0
98.1
                                                                                                2.369
                                                                                                0.202
                                                                                                0.026
                                                                                                0.740
                                                                                                0.050
                                                196

-------
Table  31.   continued
CO
NO
N02
HC
03
0.213
0.008
0.003
0.008
0.000
40.5
36.7
63.5
17.6
0.1
0.21E
0.014
0.002
0.037
0.000
40.9
63.3
36.5
82.4
0.0
                                                       Day 276
  Mixing Height                     INITIAL               EMISSIONS           ENTRAPMENT            Final
  Profile	      iP££iSi   Amount  Percentage    Amount  Percentage.   Amoun_t  Percentage    Concentration


  Linear              CO        0.213      40.5        0.21E     40.9        0.010      18.6          0.526
                                                                           0.000       0.0          0.0223
                                                                           0.000       0.0          0.004
                                                                           0.000       0.0          0.045
                                                                           0.078      99.9          0.078


  Handpicked          CO        0.149      34.3        0.187     43.0        0.098      22.6          0.435
                      NO        0.006      31.8        0.012     68.2        0.000       0.0          0.018
                      N02       0.002      58.4        0.001     41.6        0.000       0.0          0.003
                      HC        0.006      14.7        0.032     85.3        0.000       0.0          0.038
                      03        0.000      0.0        0.000      0.0        0.079      99.9          0.079


  Characteristic      CO        0.344      52.4        0.215     32.9        0.096      14.7          0.655
                      NO        0.013      48.3        0.014     51.7        0.000       0.0          0.028
                      M02       0.004      73.7        0.002     26.3        0.000       0.0          0.006
                      HC        0.013      25.6        0.037     74.4        0.000       0.0          0.050
                      03        0.000      0.2        0.000      0.0        0.077      99.8          0.077
                                                1.37

-------
     Thus, for the LIN case for Day 159,  the 03 concentration  resulting from
entrapment was 0.070 ppm with [0,]..  = 0.12 ppm and (MH, -  MH.)/MH,  = 0.58,
                                 •J ML                   T     It
while for the HP case (MHf - MH1)/MHf = 0.97, resulting in 0.117  ppm  Q^ from
entrainment.   In examining the plots in  Figure 77,  it  is evident that the
detail in the ambient data is not always  simulated  very well by the model.
For example, the CO ambient data for Day  159 shows  some very sharp spikes
in the afternoon(1330 to 1530 LOT) that do not occur in any  of the models.
The  NMHC data exhibits similar rapid jumps for June 8  (D 160/03)  and in the
morning the NO  shows a quadrupling of concentration in one  hour.  Particu-
              A
larly bad cases are July 13 and July 19,  where the  trajectories presumably
went through the center of town and a dense part of the RAMS network. It  is
evident from the simulations in Figure 77 that the  emissions and  mixing
height data for these two days are not consistent with  the ambient data.
This is especially troublesome on July 19 which should  be an ideal case for
a simple trajectory model approach.  The  July 19 case will  be  examined in
detail later in this report (section 8.3.3 and 8.3.4).   The  oversimplistic
treatment of point sources and the large  amount of  "noise" in  the trajectory
leading to spacial averages of the area emissions may be major causes of the
poor agreement.
8.1.3  Summary of MOSTM simulations
     Based on meteorology-only simulations the days  expected to perform
poorly in the photochemical simulations are:  June  8, D'160/15; July  13,
D195; July 19, D201; August 25, D238; LIN and CHAR  for  Sept  17, D261;
starting at 0600 LOT on October 1, D275;  and LIN and CHAR of October  2,
D276.  More than one-half of the test cases appear  to be so  poorly charac-
terized by input data that they are probably not suitable for  examining
the effects of chemical mechanism choices on controls.
                                      198

-------
8.2  PHOTOCHEMICAL SIMPLE TRAJECTORY MODEL SIMULATIONS OF RAPS DAYS
8.2.1 Brief Description
     The Photochemical Kinetics Simple Trajectory Model (PKSTM) is a substan-
tially expanded OZIPP-like model more suitable for research use.  The
equations solved by the model were given in Table 1.  Special features of
the model include:
     • easy entry of different chemical mechanisms
     • easy entry of photolytic processes
     • use of variable temperature and temperature dependent rate constants
     . use of variable humidity and water vapor dependent rate constants
     . allows for CO emissions
     . allows for CO aloft (constant or variable)
     . allows for variable 03 aloft
     . use of absolute emission units
     . use of smooth, equal-area emission histogram fits
     . use of smooth mixing height fits
     . simple, fast solution algorithms
     . large amount of output information
     • simple repeat functions for isopleth diagram generation.
Of course the model  incorporates all of the processes included in the
meterology-only (MOSTM) model and adds to these the processes needed to
represent the chemistry.
     Each of the 10 days  was simulated with each of the three mixing height
profiles (linear, hand picked, or characteristic) with each of three
mechanisms (Demerjian, Dodge, or Carbon Bond).  Because of its performance
on the Bureau of Mines smog chamber, the CIT mechanism was used to simulate
                                       199

-------
June 7 (D159) and October 1  (D275) only.



8.2.2  Results of PKSTM Simulations



     The comparison of predictions and ambient data for all  days is shown



in Figure 79.  Figure 79 shows one day per page, a single mixing height



per row, and a single mechanism per column.  The top row is  always the



linear (LIN) mixing height profile results; the bottom row is always CHAR.



The order from left-to-right is always the Demerjian (DEM),  Dodge (DOD),



and Carbon Bond II (CB2) mechanisms results.  The CIT simulations appear



on a separate page by themselves.



     Only the NO, N0?, and 0^ data are shown since the CO data would be



essentially the same as the MOSTM simulations already shown.  Of course,



the mechanisms do not use NMHC, but represent individual molecules of



various hydrocarbon species.  These were not plotted because there would



be no ambient data for comparison.



     June 7 D159.  As expected from the MOSTM simulations, this day was



reasonably well simulated by all four mechanisms.  Of course this day had



0.12 ppm 0.-, aloft and significant quantities of 0^ were entrained.



     June 8 D160/03.  Only DEM and DOD  CHAR simulations were close to



ambient data.



     June 8 D160/15.  As expected from MOSTM results, no simulations were



satisfactory.



     July 13 D195.  As expected from MOSTM results, no simulations were



satisfactory.



     July 19 D201.  As expected from MOSTM results, no simulations were



satisfactory.  Although the DEM and DOD LIN simulated 03 values were



similar to the ambient data for the first few hours, the failure to track



                                      200

-------
             0.  PP.

otn»meir>ois>o
   Sryof^txl(vor-irt
   (M      to
                                                                                                 O    >,
                                                                                                 Ol       X>
                                                                                                      c
                                                                                                I—   
                                                                                                 o  s-
                                                                                                -c  o
                                                                                                £X  <4-
                                                                                                CTl
                                                                                                 
-------
S S
                                                           •a
                                                           a;
                                                           o
                                                           o
                                                           a>
             202

-------
   1,1,1.1.1,1.1,1.1

S 8  I  S S S  £ = = I

            •OH ON
g a:  s  s s s  s s

         •*< ""OH 'OH
                       - 3
  Jn*-
                                       1,1,1.1,1,1.1.1,1

                                    ssssssissli

                                             •« "ON 'OK
                                                                                                                   13
                                                                                                                    O)
                                                                                                                    o
                                                                                                                    o
                                                                                                                    cu
                                                                                                                    s-
                                                                                                                    3
                                                                                                                    O)
                                                         203

-------
 11 ' i' n i ' i ' i ' i ' i ' i '\-
• 1-:         '   d=
                                                                                          OJ
                                                                                          3
                                                                                          C
                                                                                           O
                                                                                           o
                                                                                          CTl
                                                                                          r-.


                                                                                           t
                                                                                           3
                                             204

-------
0.. m
SRSSSKgio

-a
-SJ

:\ *-. "f
- ••••-•.::::'•'•: (
i .
-
'i 1 1 1 1 .1.1.1.1.1,

s s

1 1 1 1 1
f
/

s/
\\
a
-

s

m
^

"
f

=s
-


»«•«>•----£
1 1 ' 1 ' 1 ' I ' 1 • 1 'T
-a
-s> 11
2« -I
:s*--*-,. I
.••*"*•*
""***«• .

jj _, i , 1 . r , I . I . LJ

• «

""'I ' 1 '

r

s'j~
Nv
i £
.1.1.'


s
-
"

2
-
- c
" £
= 4
-
•"« "W 'OH
                                       •" 'IW 'ON
                                                                                     1. „.
                                                                          2SSE2"-3s5™8
                                                                            Ti' i  • i'  i' i • i' i'  i' i'  =
                                                                                                       = 2
   1  • i •  i •'  1111 1111, i .I,,,
SssssssssSs
          "• "ON 'OH
      SPSS.
           IS  S S
                             SasESSISss
                              rK
                                                '\    I
                                                                          K  S
                                                                                                      - s
     ' i -1' i  * i •  i

                                r • i' i " ' i'  i'''' i"'
                              -8
                                 i, 11111 11 • i  • i •  i .  i •-'-
                              i s  s  s  s  s  s  = \  s  1
                                         *•'  "OK 'M
                                                                                                                                 SZ
                                                                                                                                 O
                                                                                                                                 O
                                                                                                                                 cu
                                                                                                                                 3
                                                                                                                                 a>
                                                          205

-------
                                               -o
                                               O)

                                               c
                                               o
                                               o
                                               s_
                                               3
                                               CD
206

-------

                  •M -ON 'ON
                                                       -o
                                                        OJ

                                                        c
                                                        c
                                                        o
                                                        o
                                                       en
                                                        a)
                                                        s_
                                                        3
207

-------
                                                §
                                                   
-------
  s
Pf a


 s
  i ' i  ' i ' i' i ' i • i' r
                                                                                   £  o 10 •
             s •• V-1!


i i i li i i  l r l• ' • ' ' i • ' •'"



       •« ••on 'on
                                                                    -s :
                                                                     K
                                                                      I • I ' I '  I ' I ' I ' I ' I '  I 'J-
                                                                                  £ "•
                                                                            »«« "ON 'W
                                                                                                               -o

                                                                                                               S
                                                                                                               c
                                                                                                               o
                                                                                                               o
                                                                                                               cr>
                                                                                                               0)

                                                                                                               S-
                                                                                                               CD
                                                        209

-------
•"" "ON ON
                                                                              -= a
                                                                                                    -a
                                                                                                    a>
                                                                                                    e:
                                                                                                    o
                                                                                                    o
                                                                                                    CTl
                                                                                                    cu
                                                                                                    O1

                                                                                                    •r—

                                                                                                    u_
                                              210

-------
•"  '•on •en
                                                                                 S -.       3

                                                                        L. I ,  I , I  , I  . I  . I .  |
                                                                                   '•!» 'OH
                                                                                                                       c

                                                                                                                       8
                                                                                                                      en
                                                                                                                      r-.
                                                                                                                       OJ
                                                                                                                       CD
                                                      211

-------
                                                         -a
                                                         O)
                                                         c
                                                         o
                                                         O
                                                         ai
                                                         s_
                                                          en
                            •
-------
                                    S S
s  K
          •" '-CH '0,1
                                                                                                                                T3
                                                                                                                                OJ
                                                                                                                                3
                                                                                                                                •(->
                                                                                                                                o
                                                                                                                                CT>
                                                                                                                                r-
                                                                                                                                0)
                                                                                                                                cr>
                                                              213

-------
the NO  and NMHC for this day suggest that the actual processes
      X


for this day were not well represented.



     August 8 D221.  The MOSTM results suggested that this day had poten-



tial for being reasonably simulated.  DEM simulations were reasonable;



DOD simulations were slightly low; and CB2 failed to remove the NCU and



form 00.



     August 25 D238.  The MOSTM results showed that the afternoon CO,



NMHC, and NO  were poorly predicted.  Since the NO  and HC rise in the
            X                                     X


13-1500 LOT period was not predicted, the 0^ predictions were significantly



low.



     Sept. 17 D261.  As suggested by the MOSTM the hand picked mixing



height profile appears to be the most appropriate for this day.  Both DEM



and DOD predicted not only the 03 but also the NO, N02, and NMHC (and the



shape for CO) very closely.  CB2 did not make much 0^ in the afternoon.



     Oct. 1 D275 at SAM.  The early start on this day is a good test of



having to simultaneously match the mixing height and its rise with the



emissions rates.  The MOSTM LIN simulation suggested that CO would not be



predicted well.  The HC and NO  values, considering chemical losses,
                              A


might be a little low.  The HP MOSTM results gave large NO  overpredictions
                                                          X


and the CHAR simulation also gave a somewhat smaller but still large 8-9



morning spike.



     In the PKSTM simulations, only DOD for CHAR was able to produce a



reasonable simulation.



     Oct.__1 D275 at 8 AM.  This was the high 0~ day.  Comparison of the



8 AM simulations with the 6 AM simulations shows significant differences.



A major difference is that the 8 AM initial NO is specified in this



                                      214

-------
simulation (and is therefore correct) whereas it was computed from two



hours of meteorology, emissions, and chemistry in the 6 AM starting



simulation.  DEM and DOD did reasonably well; CB2 was lower.  This was



the second day simulated with the CIT mechanism.  As expected from the



high NO  initial conditions and the BOM simulation results, the CIT
       A


mechanism performed poorly on this day.



     Oct. 2 D276 at 6 AM.  The MOSTM simulations suggested that the mixing



heights specified for this day resulted in too much dilution and that the



NO  emissions used were inappropriate for the 8-11 LOT period.  As shown
  /\


in the PKSTM these problems  resulted in too little mass to continue the



03 production.





8.2.3.  Summary of PKSTM Simulations



     Table 32 summarizes the ozone maxima predictions by the photochemical



kinetics simple trejectory model for the eleven selected days.  Figures 80



and 81 are scatter diagrams organized by mechanism and by mixing height



profiles.  See Section 8.4 for discussion of "fits".
                                     215

-------
       Table 32.   Ozone Maxima Predicted  by  MOSTM  and  PKSTM  Using  Four
                  Mechanisms and Three  Mixing Height Profiles  for
                  RAPS Days  (ppm 03)
                                                   PKSTM
159
June 7
160/03
June 8
160/15
June 8
195
July 13


201
July 19
221
Aug 8
238
Aug 25
261
Sept 17
275/6
Oct 1
275/8
276
Oct 2
 Obs

0.192
0.194
0.206
0.210
0.138
0.119
0.190
0.151
0.244
0.244
0.181
MHIP

 L
 HP
 C

 L
 HP
 C

 L
 HP
 C

 L
 HP
 C

 L
 HP
 C

 L
 HP
 C

 L
 HP
 C

 L
 HP
 C

 L
 HP
 C

 L
 HP
 C

 L
 HP
 C
MOSTM

0.084
0.118
0.108

0.083
0.084
0.084

0.088
0.089
0.082

0.067
0.066
0.063

0.071
0.078
0.078

0.050
0.068
0.062

0.079
0.087
0.080

0.049
0.058
0.057

0.056
0.057
0.050

0.048
0.057
0.049

0.078
0.079
0.077
DEMER
0.2519
0.1642
0.1965
0.1782
0.1338
0.1457
0.1258
0.1059
0.1245
0.0982
0.1127
0.1325
0.1127
0.0918
0.0933
0.1337
0.1002
0.1137
0.0960
0.0951
0.0943
0.1032
0.1227
0.1042
0.0783
0.0708
0.1175
0.1970
0.0772
0.1680
0.1206
0.1012
0.1399
DODGE
0.1478
0.1143
0.1279
0.1822
0.1221
0.1300
0.0872
0.0564
0.0747
0.0937
0.1022
0.0988
0.1107
0.0954
0.0962
0.0851
0.0732
0.0807
0.0888
0.0864
0.0811
0.1382
0.1425
0.1210
0.1429
0.1254
0.1820
0.2138
0.1298
0.2100
0.0990
0.0997
0.1138
CARB BOND
0.1970
0.1320
0.1540
0.0904
0.0876
0.0825
0.1046
0.0935
0.1006
0.0719
0.0756
0.0986
0.0717
0.0706
0.0709
0.0721
0.0628
0.0651
0.0820
0.0875
0.0823
0.0491
0.0580
0.0545
0.0241
0.0297
0.0270
0.1160
0.0450
0.0808
0.0709
0.0723
0.0739
CIT
0.1963
0.1348
0.1534








0.0771
0.0385
0.0546

                                      216

-------
                       en
                       O)
                       -C
                     'E e
                       O

                       *°
                       -i£
                        O
                     It
                        =
                     c.
                     to
                       **
                        -f-l
                     (O
                          OJ
                     O)
                        to O
                        tO t-
                     o
                     00

                     Ol

                     3
                     ov
217

-------
                                      en

                                      CD
                                        QJ
                                        c
                                        O
'-0 OH3I03HJ
                                      -0
                                      C
                                      H
                                      O
                                      £5
                                      <_)  "3
                                      O)
                                        + 1

                                        0)
                                      T3

                                      S_
                                      QJ
                                      S-
                                      Z3
                                      cn
                     218

-------
8.3  RESPONSE OF SELECTED PHOTOCHEMICAL SIMULATIONS TO DIFFERENT INPUT
     ASSUMPTIONS
8.3.1  Carbon Bond and Demerjian Mechanism's Response to Increased Aldehyde
       Emissions.
     In the RAPS emission inventory (Table 25), aldehydes represented from
0.8% to 3% of the carbon for the days included in this study.  The OZIPP
program with the Dodge mechanism assumes that aldehydes are 5% of the
NMHC   and this assumption was maintained in this study for the Dodge
mechanism.  The aldehyde fraction in the emission inventory may be a highly
uncertain number; there were no systematic aldehyde measurements conducted
in RAPS.  Aldehydes are significant radical  sources in the chemical
mechanisms.
     The Carbon Bond mechanism performed very well in the Bureau of Mines
smog chamber simulations in which Demerjian's value of -5% aldehydes was
used and it performed poorly in the RAPS simulations in which -2% aldehydes
were used.  Furthermore, Demerjian had assumed an average aldehyde carbon
number of about one for the BOM data and the RAPS emission inventory average
aldehyde carbon number was reported by Demerjian as being about 2.0.  In the
Demerjian mechanism this shift is not important since there is no real  effort
to have carbon accountability in the Demerjian mechanism.  In the Carbon Bond
mechanism, where carbon accountability is a feature, this shift could be
critical.  The extra carbon on the aldehyde would have to be counted as a
paraffin carbon in CB2 and thus the actual  number of photolytic carbons would
be decreased by a factor of two when compared to Demerjian's mechanism in
which the single aldehyde species represents some combination of 1-carbon,
2-carbon, and higher-carbon number aldehydes.  For Carbon Bond then, the
difference in assuming that the aldehyde carbon number is one or two repre-
                                      219

-------
ents a doubling of the amount of carbon that is carbony!  and thus a  photo-
lytic source of radicals.
     In Figure 82, the October 1, D275/8 standard Carbon  Bond simulation is
compared to a Carbon Bond simulation in which the aldehyde initial and
emission carbon fractions have been doubled.  Although the simulation pre-
dictions remain under the observed CL, there is a substantial improvement in
the predictions.  Other simulations, not shown here,  with higher aldehyde
fractions show that large aldehyde fraction assumptions can give nearly
perfect fits, suggesting that for this day the Carbon Bond mechanism is a
little short of radicals.
     A similar experiment was conducted with the Demerjian mechanism.  In
Figure 83, the October 1, D275/8 standard Demerjian simulation is compared  to
a Demerjian simulation in which only the initial aldehyde carbon fraction was
increased from 2.0% to 5.0%; the emission aldehyde carbon fraction remained
at 2.0%.  The effect was dramatic; the 0~ maximum increased by 20%!  The cause
for this large response will be discussed below.
     The effect of increased aldehydes was tested on  a day that was  modelled
well by Carbon Bond with the standard assumptions, June 7, D159.  In Figure
84, the standard Carbon Bond simulation is compared to one in which  both
the initial and the emission aldehyde carbon fraction had been increased
from 0.82% to 5%, a fivefold increase.  Less than a ten percent effect on 0,
maxima was observed.  However, this day had 0.12 ppm  Og aloft and October 1
had only 0.06 ppm 0^ aloft.
     The reaction of aloft 03 with emissions can serve as a  radical  source.  The
addition of two radical sources does not have a linear effect in mechanisms;
if there are sufficient radicals, the effect of a second  source may  not be
very discernible.  Furthermore, a detailed analysis of the October 1 simu-
                                     220

-------
a.. PP.
"«  'MN 'ON
0.. ppm
                                             '-ON 'OH
                                         0.. ppm
                                                                                                                E*
                                                                                                                c  cu
                                                                                                                (O  T3
                                                                                                                ^:  r—
                                                                                                                o  re
                                                                                                                O)
                                                                                                                E  -o
                                                                                                                    O)
                                                                                                                    CU
•">«  'TO 'OH
                                                                                                                o
                                                                                                                CO
                                                                                                                c
                                                                                                                o
                                                                                                                re  re
 en  co
 c  c
 O  ••-

'•Jj  c
 O  T-
 re  co
 s-  QJ
If-  ^2

 o;  c
-o  o
 >>••-
x:  +•>
 o>  re
T3  I—
r—  rs
 re  E

 O)  i/>
j=
+->  to
    D.
 cncC
 c  S

r—  IO  C
J3  r~  E
 3  CTi  3
 O  r- •—
T3      O
 O    4-)
    s- ^:
4-> <]j  en
 O^ T-
 o» o  s-
M- -(->
M- O  c
LjJ O -i-
                                                                                                                CNJ
                                                                                                                00

                                                                                                                 O)

                                                                                                                 3
                                                                                                                 co
                                                     221

-------
                                                                                           p
                                                                                           E 4^
                                                                                           Ss
                                                                                           s- 
                                                                                           o to
                                                                                           +J CD
                                                                                           o c
                                                                                           £'£
0.. ppin
   '"ON 'ON
                                                                                           Oi en
                                                                                           T3 <1)
                                                                                           O)
                                                                                           fO -P
                                                                                              rtJ
                                                                                           •r- (/)
                                                                                           E
                                                                                           •I- CO
                                                                                              Q_
                                                                                           S« et
                                                                                           LO CC.

                                                                                           CD IO
4J O)
O J2
O) O
f- -M
4- O
LLJ O
                                                                                           CO
                                                                                           CO
                                                                                            CD
                                                                                           •I—
                                                                                           U-
                                          222

-------
                                                                                           E
                                                                                           V)
                                                I .  I .  I .  I  . I  • I  •  I •  I •  I •  '"•
          ««  'MN  'ON
                                                            '-ON 'ON
                                                                                           u
                                                                                           O)
                                                                                           -o
                                                                                            c
                                                                                            o
                                                                                           CO

                                                                                            c
                                                                                            o
                                                                                            (O
                                                                                           o
                                                                                            c
                                                                                            o
                                                                                            o
                                                                                            (D
                                                c
                                                o
                                                •f—
                                                                                           •i-  C

                                                                                            E  O
 1,1,1,1.1,1,1,1.1


sssssssll

         «« '-ON 'ON
                              - 3
                                                        Oi.
                                               1,1,1.1,1,1,1,1,1
                                                 cu ••-
                                                XJ tn

                                                 (B I/)
                                                    o.
                                                -o <
                                                 a; a:
                                                 to
                                                 to UD
                                                 oj r*~
                                                 s- en
                                                 o i —
                                                 c
                                                 O -3
                                                 a>
                                                <+-  s_
                                                4-  o
                                                                                            oo
                                                                                            a>
                                                                                            s-
s  s  s
                                                        •« '-OH 'ON

-------
lations showed that In the Demerjian mechanism the (L + OLE reaction



(Reaction 20, Table 7) gives two radicals (one R02 and one H02)  and is a



significant source of new radicals (-5% of the radicals) for this mechanism.



The Carbon Bond mechanism, on the other hand, uses newer (and presumably more



kinetically correct) 03 +  OLE   chemistry in which the radical  yield is



only about 0.1 (compared to Demerjian's 2.0) and this process is not a signi-



ficant source of radicals (<0.1% of the radicals) in the CB2 mechanism.   There



is an additional  radical source in the Carbon Bond mechanism and it is



photolysis of unique carbonyl products resulting from aromatic reactions.



Demerjian has no unique carbonyl aromatic products; arornatics are treated



similar to olefins with an extra standard aldehyde being produced.



     Thus, for Demerjian, once 0, production begins and olefins  are present



in the hydrocarbon emissions, new radicals are abundantly available to



sustain the 0^ production process.  This may not be a good representation of



the actual process.



     Extra aldehydes are not always adequate to improve the Carbon Bond



mechanism's performance.  Figures 85 and 86 compare standard Carbon Bond



simulations to those with increased aldehyde emissions; not much improvement



occurred  for  these  days which were  simulated better by  the other



mechanisms.



     For the Dodge mechanism, the aldehyde emission fraction assumption is



constant.  As was shown in Table 10, however, new photolytic rate data are



available which, if used, might affect the radical production in the Dodge



simulations.  Figure 87 compares the standard Dodge October 1, D275/8 simu-



lation using the original ("OLD") photolytic rates from the OZIPP program



with a simulation using the newer photolytic rates ("NEW").  The effect was



very small.





                                     224

-------
                                        »"« ''ON  'ON
                                                                                                                     e
                                                                                                                     to
                                                                                                                     •r~
                                                                                                                     C
                                                                                                                     ro
                                                                                                                     .C
                                                                                                                     u
                                                                                                                     OJ
                                                                                                                     e
                                                                                                                      c
                                                                                                                      o
                                                                                                                     CO

                                                                                                                      c
                                                                                                                      o
                                                                                                                     -Q
                                                                                                                      s_
                                                                                                                      to
                                                                                                                     o

                                                                                                                      s_
                                                                                                                      o
                                                                                                                      c
                                                                                                                      o
                                                                                                                      o
                                                                                                                      (O
                                                                                                                      s_
                                                                                                                      c
                                                                                                                      o
                                                                                                                      CO  C
                                                                                                                      oo  O
                                                                                                                      QJ  (O


                                                                                                                      OJ  3
                                                    S g
'-ON 'ON
                                        *" "ON 'ON
                                                                                                                      JT  00
                                                                                                                      OJ
                                                                                                                      •O 
                                                                                                                      O> r—
                                                                                                                      S_
                                                                                                                      O   «
                                                                                                                      C CO
                                                                                                                          cn
                                                                                                                      O
                                                                                                                      O>
                                                                                                                      UJ
                                                                                                                      LO
                                                                                                                      00


                                                                                                                      O)
                                                                                                                          S-
                                                                                                                          O
                                                    225

-------
                  LO


                  O
                  oo
                  o
                  c
                  o
                   o
                   (O
                   LO  C
                   c:  , c
                  -C  O
                   >
                   +-»  (O
                   <_> a
                   a>
                   4-  i-
                   M-  O
                   00
                   CD
                   i-
226

-------
         s s  s
««H  ''ON 'ON
                                          ""I 'MN 'ON
                                                                                                                     O

                                                                                                                    •SJ

                                                                                                                     O)
                                                                                                                     Ol
                                                                                                                    X3
                                                                                                                     (O
                                                                                                                     0)
                                                                                                                     o c
                                                                                                                    •M O
                                                                                                                     O •<-
                                                                                                                    ^: •(->
                                                                                                                     CL to


                                                                                                                    "2 i
                                                                                                                     QJ •!-
                                                                                                                     C in

                                                                                                                        CO
                                                                                                                    •O Q.
                                                                                                                     C <
                                                                                                                     (Q o:

                                                                                                                    =   VO
                                                                                                                    •a r^
                                                                                                                    r— <7>
                                                                                                                     o •—
                                                                                                                     O J3
                                                                                                                     01  o
                                                                                                                    4-  4->
                                                                                                                    M-  O
                                                                                                                    LU O
                                                                                                                    CO


                                                                                                                     S-
   'ON  OK
                                                      221

-------
8.3.2  Demerjian and Carbon Bond  Mechanism  Response  to  Hydrocarbon
       Compositional Changes

     The Carbon Bond mechanism did  well  in  the Bureau of  Mines  smog chamber

simulations in which the aromatics  carbon  fraction was  nearly 30%, the

aldehyde carbon fraction was assumed  to  be  about 5%, and  the aldehyde carbon

number was assumed to be one.   The  mechanism did poorly in  the  RAPS simula-

tions in which the aromatics carbon fraction was 6%  to  18%, the aldehyde

carbon fraction was 0.8% to 3%, and the  aldehyde carbon number  was 2.0.

     The effect of switching from an  aldehyde carbon number of  2.0 to an

aldehyde carbon number of 1.0 has already  been demonstrated and discussed

(Figure 82).   The effect of using the BOM  HC composition  in the RAPS

October 1 simulation will  now be  investigated.  To see  the  effects of com-

position changes only, all  simulations assumed an aldehyde  carbon number of

1.0.

     The left column of Figure 88 shows  the October  1 simulation using  the

October 1 HC composition and an aldehyde carbon number  of 1.0;  this is  the

same as the right column of Figure  82.  The center column of Figure 88  shows

the October 1 simulation with the BOM HC composition but  using  the October 1

aldehyde emission carbon fraction and an aldehyde carbon  number of 1.0.  The

right column of Figure 88 shows the October 1 simulation  with the BOM compo-

sition and the BOM aldehyde carbon  fraction (5%) and an aldehyde carbon

number of 1.0.  The simulations with  the full BOM assumptions nearly  per-

fectly fit this day.

     The HC compositional  change  was  (%  carbon):

                     NR        PAR         OLE       ARO        ALJJ

       Oct. 1       12.2       53.7       17.2       14.9      2.0

       BOM          10.5       41.7       17.3       28.7      5.3

                                     228

-------
g  8  a  E  5 «"• 8  5  8  s
                                                   0*. PP«
                                                 in o v^ o  v>  »  v>
                                                 r>mfw«>r>tnn
                                                 — — — —  000
§  I
               "ON 'OH
                                                                                                                       (/I
                                                                                                                       o
                                                                                                                       Q.
                                                                                                               - 3
                                                                                       •»  W  IM
                           a. PP»
                     sSSSlEss
                                                     o

                                                     3
                                                     (O


                                                     3
                                                     CO

                                                     o
                                                     -M
                                                                 229

-------
That is, about 12% of the carbon was shifted from paraffins to aromatics
between the simulations.
     The change in 03 resulting from the various assumptions were (ppm 03):
Composition
Aldehyde Fraction
Aid. Carb. No.
Mixing Heiqht:
LIN
HP
CHAR
Oct. 1
0.020
2.0
0.116
0.045
0.081
Oct. 1
0.020
1.0
0.147
0.060
0.117
BOM
0.020
1.0
0.192
0.079
0.164
BOM
0.053
1.0
0.252
0.118
0.233
     That is, the Oct.  1 to BOM compositional  change increased the 0-,  by
0.045 ppm (31% increase) and the Oct.  1 to BOM change with BOM aldehydes
increased the 03 by 0.105 ppm (71% increase).   The total  change in going
from standard Oct.  1 conditions to full BOM conditions,  however,  increased
the On by 0.135 ppm (117%).  As described above,  aromatic products are sig-
nificant sources of radicals for the Carbon Bond  II mechanism and it has
some difficulties in dealing with these RAPS days with low aromatic content
as determined from the emission inventory and the trajectory paths.
     Another demonstration of HC compositional effects on mechanism pre-
dictions, similar to that discussed in Section 7.3.4, is  shown in Figure 89.
The left column is the standard Demerjian Oct. 1, D275/8  simulation from
Figure 79.  The center column shows the effect of substituting the HC  compo-
sition fractions from the day with the highest olefin content (August  8,
Day 221) for the October 1 HC composition fractions.  The right column shows
the effect of substituting the HC composition fractions from the day with
the highest paraffin content (August 25, Day 238) for those of October 1.
This compositional  change was (% carbon):
                                     230

-------
                            •" '"ON 'ON
231

-------
                           NR & PAR      OLE        ARO         ALD
Oct. 1
Aug. 25
65.9
86.0
17.2
10.0
14.9
6.6
2.0
1.0
     These modelling results suggest that the control  strategy for St.  Louis



would be to replace 15% of the olefin and aromatic carbon  with paraffin carbon



and have no_ reduction in total emissions; smog chamber data,  however,  suggests



that this would not be true.
                                          232

-------
8.3.3  Effects of Trajectory Assumptions for Selected Days
     The purpose of this section is to show the effects of emission pattern
variations that might result from trajectory assumptions.  The emission rate
pattern is determined by the column path of the air parcel.  The column path
is defined as the area over which the emission rates are determined.  For
this study the column path was defined by the area encompassed by five back
trajectories from the site of observed ozone maximum.  Since the meteoro-
logical parameters are assumed to be uniform throughout the urban area, the
emission rate pattern variations show the effects of column path variations.
     To examine the effects of different emission rate patterns two approaches
will be taken.  First days 238 and 201 will be examined, since each exhibited
similar ozone profiles for the three different mixing height profiles when
modeled with the three different chemical mechanisms (see Figure 79 of
section 8.2).  The other approach is to see the effects of altering the
pattern of the hourly emission rates while keeping the total amount of the
emitted mass constant.
     Day 238 - August 25.  The column path from the supplied information is
repeated as Figure 90.  The back trajectories from the site of the highest
ozone maximum indicate that the column enters the RAPS data region from the
southeast then turns approximately 90°, just southeast of the Wood River-
Alton point source region (see Figure 29).  Figure 91 shows the modelled and
ambient concentrations of NOV, Oo, CO and HC.  The ambient concentrations of
                            X   o
these species along the column path are plotted as the dashed lines.  These
represent the weighted average concentrations from the three closest
monitoring sites to the centerline of the column path.   The solid lines
represent the predicted concentration of the species by the Demerjian, (DEM)
                                     233

-------
4340
4330
4320
4310
430Q
4290
 4280
4270
 4260
 4250
 4240
           700     710
                             720
730    740      750       760      770     780
      Figure  90.  Trajectory for Day  238, August  25/Si te 115
                                         234

-------
mechanism.
     Since the ozone maximum predicted by all  three mechanisms for the three
different mixing height profiles was considerably less than the observed
ozone for the day (0.190 ppm), and since the CO concentration is so poorly
predicted, one can assume that the emission pattern of the column path
reaching the observation station is different than that used in the model
run.
     For Day 238, August 17, three alternative column paths were examined:
     A.   The first, Tl, uses the centerline trajectory of the supplied
column path and places a two kilometer box around it, one km radius.  Figure
91 shows that the ozone predicted varies little from the supplied column
path.  Table 33 contains a list of the ozone predicted for each column path.
Table 34 contains the emission rates in units of ppm-km used for each column
path, Figure 92 represents the instantaneous fit of these emission rates.
The emission rates for T3 and TH column paths were selected from the emission
rates from the supplied column path for Day 201.
     Notice the rise in ambient NO  concentration in Figure 91, during the
                                  /\
last hour and a half,and the rise and fall  of the hydrocarbon and CO concen-
tration  (also in Figure 91), during the last three hours of the simulation.
These changes in the ambient concentrations are not predicted usiny either
the supplied or Tl column paths.
     B.   The T3 column path considers a discontinuous column path, during
the last three hours of the simulation.  The column path considered does not
curl around from the southeast, but comes down from the northwest passing
                                    235

-------
through the Wood River-Alton Point source region, Figure 29. The T3 emission



rates consist of the Tl  rates for the first four hours and the major emissions



from the Wood River-Alton region for the last three hours of the simulation.



     Using this discontinuous column path, the DEM mechanism predicts much



more ozone than either the supplied or the Tl column paths, Table 33.  The



N02 peak is now present, Figure 91, and the ozone concentration profile



appears to be best predicted using the characteristic or handpicked mixing



height profiles.  A hydrocarbon peak is predicted, but in an excessive



amount, while the CO concentration is not modeled well.



     C.  Since there was promise from the T3 results,a hypothetical column



path was determined passing through the Wood River-Alton region, TH.  The



TH column path was constructed from a straight line hypothetical path through



this region using a two kilometer wide box surrounding this line.  The



emission rates for the TH column path consist of the T"! emission rates for



the first four hours and these new rates for the last three hours.  Figure



91  and Table   33 show that this column path is an inadequate description of



the processes occurring on this day.  The NOo levels are excessive; while



the hydrocarbon and CO predictions are better, they do not agree with the



observed ambient data.



     The turning of the wind back trajectories in the region of the observed



ozone maximum may indicate aconvergence zone just outside the  center city.



If this is the case then the air parcels reaching thic region came from



many paths.  To adequately establish if this was the case a more detailed



wind field analysis must be performed.  The determination of a column path



is not a simple process, as established by the TH column path selection.



                                    236

-------
                                                CO.  pp»C
                                                                                       CO. pp»C
                             - 3
                                                                                       8   5
                                                                                                           -3
v>  e>  i/>
                   228
                                                                          ca  u>  a  irt
           CO. pp»C
jut"  'DWJN

CO. ,pmC
 >»« 'MM

CO. >P«C
                                                                               ce  r-  to  10
                          TS§
                                                                                                         = 3
                                                                                                                     £0
                                                                                                                     CO
                                                                                                                     (M
                                                                                                                      (O
                                                                                                                      CO
                                                                                                                      
-------
                                        -a
                                         a>
                                         O
                                         u
                                         CTi


                                         O)

                                         5-

                                         :3

                                         CD
238

-------
         co,
                                                CO. PBmC
                                                                                     CO. ppmC
                            - 3
                                     •E


                                     • 3
                                      2%
                                                                                           OJ  —  —
                                                                              CD  r-  jj
- 8
                                    L- 3


                                         *
1,1,1,1,1,1,1,1,1,

inainoipoinotn
                                                                         -" 5
                                                                                                                    -o
                                                                                                                     0)
                                                                                                                     c.
                                                                                                                     o
                                                                                                                     o
                                                                                                                     en

                                                                                                                     a>
                                                                                                                     &.
                                                                                                                     =j
                                                                                                                     O)
                                                             239

-------
1.1.1,1.1.1.1,1,1,  ...    LuJ_L
                                                                                             "O
                                                                                              O)
                                                                                              Z3
                                                                                              O
                                                                                              O
                                                                                             CTl

                                                                                              CD
                                                 240

-------
  Table 33.  Effects of Trajectory on Ozone Maxima;  Day 238

                           DAY 238


Observed ozone = 0.190 ppm

Trajectory                 Sup       T1         T3        TH
Mixing height profile

Linear                    0.095     0.092     0.137     0.068

Handpicked                0.095     0.093     0.130     0.086

Characteristic            0.095     0.092     0.127     0.079
                             241

-------










co
c~
O
••~
(/)
to
E
UJ

(_}
m










nr
I—



oo
1—



I 	
I—




Q.
* /N
V J


.
•1-3
ro
s_
h-
o
00
o
o
o
o
co
o
o
o
o
co
o
0
o

CO
oo
o
0
o



o> en
E i
•r- CO
^-
r-.
"^3~
o
0
o
r-
«sj-
o
0
o
r-^
•=3-
O
O
O

o
un
0
o
o


o
1 —
1
en

O CM
^j- en
0 O
o o
o o
O CM
«3- en
0 0
0 O
o o
O CM
«* en
0 0
0 O
0 0

CO CM
CM O
*^J~ r—
0 O
0 0


r— CM
1 — ^—
1 1
0 r-
r— i 	
en
oo
CM
O
0
un
CM
0
oo
o
i —
0
CM
O
o

*^~
CM
i —
0
0



1 —
1
CM
1 —
CO
00

0
o
oo
r^^
un
CM
0
un
oo
i —
o
o

CM
i —
r—
O
o



CM
I
i —

OO
un
(^
i —
o
t 	
f-*v
un
•—
o
en
^j-
o
o
o

0
"Cj~
o
o
o



00
i s:
CM UJ
Q
un
r—
un
o
o
en
CO
CO
i—
o
en
CO
CO

0

o

o

co O
c:
O
4->
0
ro UJ
S- _l
u. o
CO
un
CO
CO
o
un
en
CO

o
un
en
00
un
o

0
un
un
r-~
o




ry"
c£
Q.
un
*^j-
, —
CM
O
un
en
o

o
un
en
o

0

o
CO
CO
0
0




o
Qi

00 UO
r- CO
CO O
0 0
O 0
r— O
CM CM
i — CM
•— o
0 0
r- 0
CM CM
i — CM
i— O
0 0

un oo
r^. o
CO i—
O 0
o o




f~^
cy t— i
•Z. «=C
                                                                                                                 o
                                                                                                                 u
                                                                                                                 OJ
                                                                                                                 fO
                                                                                                                 O)

                                                                                                                 O)
                                                                                                                 o
                                                                                                                 en
oo
OO
CM
>
                                          n: co
                                             oo
                                          X) CM
                                           sz
                                           fO  >i
                                              fC
                                           CO Q
                                           cu
                                          •4->   -
                                           fO  CO
                                          o;  cz
                                              O
                                           C T-
                                           O 4->
                                          •i-  Q.
                                           to  E
                                           CO  3
                                          •r-  tO
                                           E  tO
                                          UJ c£
                             O r— CM
                    CL)    CTli— i— i— i— CMOO
                    E     I   I   I   I   I    I   I
                    •i-    cocnoi— cMi— CM
         O i— CM
CD    CTl r— i— i— r— CMOO
E      	
•i-    co en o i— CM i— CM
oo

 CD
                                                                                                                _a
                                                                                                                 ro
                                                           242

-------
  NO,.  ppm*km/mJn
     C^OIX)  uiui/ms|»uidd  '03 'D
NO,, jpm'km/min   CXIO"")
-I
                                                       NO,.  ppm«km/mln-





                                                     1  I  ' I  '  I  '  I  '  I  ' I  '  I  '  I  '  I '
                                                            C..OIX)   uiui/cuij
                                                                                 -Oc  'O
                                                         NO,.  ppm»km/m(n   CX1CT1)
                                                         I  i I  i  I  ,  I  i  I i  I  i  I  i  I  i I  .
                                                    o  en   OD
                                                                                                                CO
                                                                                                                co
                                                                                                                CVJ
                                                                                                                fO
                                                                                                                Q
                                                                                                                CL)

                                                                                                                il
                                                                                                                O
                                                                                                                (1)
                                                                                                                to
                                                                                                                O


                                                                                                               <


                                                                                                                CO
                                                                                                                (O
                                                                                                               cc:

                                                                                                                c
                                                                                                                o

                                                                                                                in
                                                                                                                10


                                                                                                               LU
                                                                                                                o
                                                                                                                CJ
                                                                                                                c
                                                                                                                c:
                                                                                                                fO
                                                                                                               CM
                                                                                                               cn
                                                                                                                CD
                                                 243

-------
     Day 201 - July 19.  The column path for this day from the supplied
information is shown in Figure 93.  On this day,  the column path runs north-
south through the city from the south.   In comparison with Figure 29, note
that the supplied column path encompasses the two major clusters of point
sources in the RAPS database, the St. Louis-East St. Louis region and the
Wood River-Alton region.
     Again the dotted lines in Figure 94       represent the ambient concen-
tration interpolated along the centerline of the column path, and solid
lines represent the predicted concentrations by the Demerjian mechanism.
For the supplied column path, neither the Demerjian (shown), Dodge, or
Carbon Bond II mechanisms did well  in predicting the ozone maximum.  For
the three different mixing height profiles the results were very similar
for each mechanism.  The observed ozone maximum on this day was 0.144 ppm.
Again different emission patterns or column paths will be examined to see
their effects on the ozone prediction.
     For day 201, July 19, three alternative column paths were examined.
     A.  The first, Tl, uses the centerline trajectory of the supplied
column path and places a two kilometer box around it, one kilometer radius.
Table 35  and Figure 94 indicate that the predicted ozone concentration was
not any better than the predicted concentration using the supplied column
path.  The predicted ozone precursor concentrations (Figure   94)    for
the Tl column path, do not agree with the observed ambient concentrations
for this day.  In determining the emission rates for this column path, two
point sources were encountered in the region, but their locations could not
be established from the information at hand.  The Tl column path does not
consider either of the sources to be within the column path.  The remaining
two column paths  consider both of the point sources to be within the column
                                     244

-------
4340
4330
4320
4310
4300
4290
4280
4270
4260
4250
4240
           700
710    720
730
740     750      760    770
780
        Figure  93.   Column  Path  for  Day 201, July 19.
                                  245

-------
path.  Both of these point sources are hydrocarbon sources  only.   With



their inclusion into the column path,  the hydrocarbon fractions  for the



DEM mechanism are shifted.  The column path TAB is the Tl  column  path for



emission rates, with these new hydrocarbon fractions.  Table 35   and Figure



94      indicate that shifting to these new fractions has  a small  effect on



the predicted concentrations.



     The remaining two column  paths are a variation of the  Tl  column path.



The emission rates represent levels which were increased or decreased



slightly,  based upon the comparison of the predicted concentrations with



the observed ambient concentrations.



     B.   In the column path  TA,  the emission rates for the hour  between 10



and 11   were increased five fold  from the Tl  emission rates.  Table 35  and



Figure94 show that the ozone predicted is much closer to the observed con-



centration than either the Tl  or  the supplied column paths.  The  ozone



concentration is overpredicted with the linear mixing height profile.



Figure   94  shows    that the  prediction of the precursor  concentrations



are also overpredicted, with the  greatest overprediction for N09.



     C.   In the column path  TB,  the emission rate  for the oxides of nitro-



gen was  decreased by approximately 15% from the TA column  path,  Figure 95.



This decrease causes the model to predict an increased ozone maximum.  The



ozone concentration predicted  using the handpicked and characteristic mixing



height  profiles increases to a level which is very close to the  observed



maximum; with  the linear mixing  height profile, the predicted ozone con-



centration also increases by 10%.



     The fact that the predicted  ozone concentration rose  by 10% with a 15%



decrease in oxides of nitrogen emissions indicates that the timing and/or the



mass of the emission rates is  important to the amount of ozone predicted.




                                    246

-------
--5
                         5  8
                         X

   I  , I  I  I I. L-l-1  I I  .  I I I  I I
                                                     Outdd  3HWH


                                                     CO. ppmC
                                                                                                 CO.  ppmC
                                                                                                                      = 3
                                                                                                                 s  -
                                                                                              CO.  ppmC
                                                                     S  J
                                                                                  -   s
                                                                                                              8  -
                                                                                        I  i U I  i  1 i  I  i I  i  I i  I
                                                                                                                   - S
                                                                                                                                O
                                                                                                                                CvJ
                                                                                                                                 0)
                                                                                                                                 i.
                                                                                                                                 O
                                                                                                                                O
                                                                                                                                0)
                                                                                                                                 (O
                                                                                                                                 S-
                                                                                                                                 cr
                                                                                                                                 O)
                                                                                                                                 S-
                                                                                                                                 0>
                                                                                                                                 2:

                                                                                                                                 in
                                                                                                                                 c.
                                                                                                                                 O

                                                                                                                                4->
                                                                                                                                 03

                                                                                                                                 ZJ
                                                                                                                                O>
                                                                                                                                s-
                                                                                                                                 CT>
                  'MIH
                                                                  247

-------
                                         01
                                         c
                                         o
                                         o
                                         en
                                         5-
                                         ZJ
248

-------
 fe»
                       !  
-------
                                                  T3
                                                   O)
                                                   c:
                                                   O
                                                   O
                                                   0)
                                                   i.
                                                   cr>
                       "">" 'MN 'ON
250

-------
                               Table 35

            Effects of Trajectory on Ozone Maxima, Day 201

                                DAY 201


Observed ozone = 0.144 ppm

Trajectory                  S        TI         TAB       TA       TB	
Mixing height profile

Linear                    0.113     0.086     0.084     0.154     0.172

Handpicked                0.091     0.076     0.074     0.128     0.140

Characteristic            0.093     0.077     0.076     0.128     0.141
                                  251

-------
o
CM
ro
0
ft
C
O
CL
3
l/>
to
Trajectc
4->


SUOLSS
L§
0
re

h-
<
h-
P
Q.

CM
O
O
O
O
CM
O
o
o
o
CM
o
o
o
o
0
o
o
o
UD
o
o
o
o
UD
o
o
o
0
<
to
o
o
o
c
to
CM
o
o
o
CTl
0
0
CTl
LO
O
CD
C C
cn
0
o
cn
co
CO
o
o
cn
o
cn
LO
0
Q
cn
CO
o
o
to
o
0
r-~
0
0
r~
o
o
0
o
o
to
o
o
CO
o
o
CO
o
0
CO
o
o
o
o
CM
o
o
0
CM
r—
o
o
o
CM
0
o
o
r-.
CM
o
o
o
    CD


    CU
UD  O  t—

CO  i-  O
    4-  CM
 ~
 O  C~  ecL"
 Oi—  CMi—



























to
c:
o
•r—
4_I
O
ra
L.
u.

CM
CO
to
f—
•
0

OQ
c-f
CM
CO
to
r~
O

-)c
CO
CM
r"^
•—
o

o
to
to
<—
o





LU
	 i
0

r^
<^J-
r^
LO
•
O



f-*^
«^J-
r*^
LO
o


o

to
LO
o

o
cn
CM
LO
O





o;
di
D_

CM
O
CO
i—~
•
O



CM
O
CO
•—
o


o
CO
CM
i —
O

o
(•" —
<£>
1 — "
o





o

—
o






o:
z

to

(—
o
•
o



to
r-x
r—
O
o


CO
00

o
o

*^~
CTt
r^
o
0





Q
t—H
"^
c
•r—
O
r>

C
0

^j
ro
O
O
— 1

3
0
c
\/
c
—^

0)
-I-J
• 1 —
O)
f~

to
s~

fQ
O
o
— 1
c
^
0
c:
\s
C
^
-C
o
xi

to
^1 1
tL>
•f—
I/)
c
0
o
CO
*%
"^




























to

                         ca
    o
    ra
   CJ)
   zc

   -a
    to
    a>
   D;
    10
    to
                  o
                  o
 a.
 ^
oo
          l£> CM l£> LO «^" CO CO
          O i— to r^ r~- r^ ^j-
          o o cri «d- oo i— o
          o O o to o o o

          o o o o o o o
          UD CM  UD IT) «d~ CO CO
          o i—  to r~~ r- r^~ -3-
          o o  cr> "3- co r— o
          o o  o to o o o

          o o  o o o o o
to CM to LO  «3- co co
o i— to cr>  r^ t~^ •=]-
o o cri CM  co i— o
o o o r—  o o o

o o o o  o o o

CTI LO r^. 01  §— §— LO
o LO en LO  i— to co
o o to ^-  co CM o
o o o i—  o o o

o o o o  o o o
                                     to
                                     c
                                     o
                                                                        to
                                                                        to
                                                                        O
 a.
 3
to
                                                       i— CM LO O tO CTl -3-
                                                       O O O O CT) O O
                                                       O O r— LO CM O O
                                                       o o o o o o o

                                                       o o o o o o o
          i— CM LO LO to cr> ^t-
          o o o co en o o
          O O i— LO CM O O
          o o o o o o o

          o o o o o o o
                                                                 r— CM LO (--. UD CTI ^d"
                                                                 O O O r— O1 O O
                                                                 O O r— ,— CM O O
                                                                 o o o o o o o


                                                                 o o o o o o o
CM to "vt- r- CO i— co
o o r-~. c\j <^- UD o
O O O r— O O O
o o o o o o o

o' o o o o o o
                                         O r— CM
                             CO    COCTli— i— i—  i— CM
                             £     I    I   I   I    I   I   I
                             •i-    r-~cocr>Oi—  CM.—
                                                 01    coaii
                                                 E
                                                                               i — i — CM
                                                                                       i
                                                                               i — CM. —
                                                                          252

-------
                                       NO,  ppm-km/min  CX10")
                                                                                NO,  ppm.km/mi
NO,, ppm-km/nun-
                                                          JL—	^
                                                         S^-J\

                                                                                          UI«J/»>(.«<1<1  03 'OH
                                        NO,.  ppm-km/IBIn  CX10-)
                                                                                NO,  ppm»km/mln  [XICT*)
                                           i'  i  '  i'  i'  i' i'  i '  i  i-q-      (.' i' i'  i •  i' i
                                                                                                                           QJ
                                                                                                                           O
                                                                                                                           0)
                                                                                                                           en
                                                                                                                           c.
                                                                                                                           O

                                                                                                                           
-------
8.3.4  Effects of Emissions Pattern for Selected Days
     The purpose of this section is to examine the effects  of changing the
pattern of the emission rates while keeping the total emission mass  constant.
     On three days, Day 159-June 7, Day 261-September 17,  and Day 275-October
1 starting at 8 o'clock, three different emission patterns  were examined.
The emission rate patterns were determined using a random  number table.
Each hour:s emission rate was assigned a value corresponding to the  hour of
its occurrence, the first hour being one, the second, two,  etc.  (Table 37).
Then these numbers were reordered by going down the random  number table until
all the numbers had appeared.  Table 37 indicates the order of the emission
rates for the three different patterns, A, B, and C,  and the total emission
mass in ppm-km for the three days.
     Day 159, June 7 is the day with the lowest oxides of  nitrogen emissions
while day 275 starting at 8 AM has  the highest.  Both of these days  have
similar hydrocarbon emission rates.  The results are  shown  in Table  38.
Table 39 shows the absolute and relative deviations as percent of each
emission pattern from the supplied  pattern.  Day 275  shows  the largest devia-
tions while day 159 has the smallest deviations.
     The results suggest that the order in which the  column path intersects
the emission inventory is important.
8.3.5  Effects of Ozone Aloft Assumptions
     The purpose of this section is to examine the effects  of two different
ozone aloft profiles on the predicted ozone maximum.   Three days from the St.
Louis RAPS study will be used to examine the effects, using the Demerjian
chemical mechanism.
     The two profiles are a constant profile and a variable profile.  In the
constant ozone aloft profile a constant concentration of ozone is entrained
into the mixed layer for a given mixing height rise.   For  the variable ozone
                                     254

-------
                         Table 37



   Random Emissions Patterns and Total  Mass of Emissions






      Pattern         Hour Order of Emissions
      Supplied        1   23456789



         A            345863719



         B            536479128



         C            364985217
                                   CO        NOV         HC
                                               X
Day 159,  June 7                  0.5248     0.0346     0.2347



Day 261,  September 17             0.7535     0.0589     0.4344



Day 275,  October 1  at  8 AM       0.8364     0.0859     0.2884
                           255

-------
                            Table 38

        Predicted Ozone Maximum Considering Three  Different
                         Emission Patterns
                                       Emission  Pattern

Day 159, June 7                   S        A       B        C
Mixing Height Profile            	^—   ———	
              Linear             0.252   0.248   0.243   0.239
              Handpicked         0.164   0.157   0.157   0.152
              Characteristic     0.197   0.104   0.189   0.184

Day 261, September 17
Mixing Height Profile

              Linear             0.103   0.118   0.113   0.119
              Handpicked         0.123   0.136   0.137   0.137
              Characteristic     0.104   0.118   0.114   0.115

Day 275, October 1 at 8 am
Mixing Height Profile
              Linear             0.261   0.233   0.228   0.238
              Handpicked         0.096   0.070   0.067   0.070
              Characteristic     0.234   0.186   0.179   0.189
                                256

-------
                      Table 39

   Deviations of Ozone Maximum Prediction from the
               Supplied Emission Pattern
                                  Emission Pattern
Day 159, June 7

Mixing Height Profile
      Linear

      Handpicked

      Characteristic
                                A
-0.004*
 1.59
-0.0007
 4.27
-0.093
47.21
             B
-0.009
 3.57
-0.0007
 4.27
-0.0008
 4.06
-0.0013
 5.16
-0.012
 7.32
-0.013
 6.60
Day 261,  September 17

Mixing Height Profile
      Linear

      Handpicked

      Characteristic
0.015
14.56
0.013
10.57
0.014
13.46
0.010
9.71
0.014
11.38
0.010
9.62
0.016
15.53
0.014
11.38
0.011
10.58
Day 275, October 1 at 8 AM
Mixing Height Profile
Linear

Handpicked

Characteristic



-0.028
10.73
-0.026
27.08
-0.048
20.51


-0.033
12.64
-0.029
30.21
-0.055
23.50


-0.023
8.81
-0.026
27.08
-0.045
19.29
*The top number represents the absolute deviations,  calculated
 as difference in ozone maximum between the variation pattern
 and the supplied pattern.  The lower number represents  the
 relative deviation,  represented as percent; computed as
 absolute deviation divided by the predicted ozone maximum
 from the supplied emission pattern.
                         257

-------
aloft profile, the concentration of ozone entrained into the mixed layer
depends on the fraction of the total  rise of the mixing height for  that day.
The fractional multiplier is prescribed by the equation:
                           FAL - 2FR3 - 3FR2 + 1
     where  FAL is the fraction of the aloft concentration
             FR is the fractional  rise in mixing height,  given by
                   the current mixing height minus the initial  mixing
                   height divided  by the total  rise in mixing  height
                   for that day.

     As one might expect, the maximum ozone concentration decreases  as  the
concentration of ozone entrained into the mixed layer decreases,
i. e. a constant aloft profile provides the greatest entrained concentration,
and the variable aloft profile provides a lesser concentration, Table    40.
     The constant aloft profile is considered in Table  41.  The  first  column
represents the effective concentration of the entrained ozone.  This was
determined by subtracting the maximum ozone concentration with ozone aloft
from the maximum ozone concentration using the constant ozone  aloft  profile.
The second column  represents the  concentration of the ozone entrained
considering only dilution, i. e.  no chemistry.   The third column  can be
interpreted as the effect of chemistry on the entrained ozone  concentration,
expressed as a percentage.  A value of zero indicates the entrained  ozone
concentration is not affected by chemistry.  For the three days examined,
no pattern of the affect of chemistry with respect to the different  mixing
height profiles can be determined.  For day 261, the ozone maximum is
enhanced by the chemistry and the  entrainment ozone, with the  characteristic
mixing height profile.
                                     258

-------
                            Table 40


            Variation of Ozone Aloft on the Prediction
                        of Ozone Maximum
Day 159, June
Mixing Height Profile
               Linear
               Handpicked
               Characteristic
          Aloft Profile

Constant     Variable     None


  0.252        0.232      0.215
  0.164        0.117      0.072
  0.197        0.159      0.123
Day 261,September 17
Mixing Height Profile
               Linear
               Handpicked
               Characteristic
  0.103
  0.123
  0.104
0.081
0.096
0.072
0.067
0.070
0.058
Day 275, October 1  at 8 am
Mixing Height Profile

               Linear               0.261
               Handpicked           0.095
               Characteristic       0.233
               0.241
               0.070
               0.209
           0.218
           0.053
           0.193

-------
     The results from using a variable aloft profile are presented in



Table  42.  The three columns contain the same information as  in  Table   41,



considering a variable ozone aloft profile.   Again no pattern  can be  deter-



mined with respect to the mixing height profile.



     In comparing the days for the two different  aloft profiles,  specifically



days 159 and 275, a trend of the effects of  chemistry are similar,  with  the



magnitudes being close in most cases.  For day 261,  the enhancement of the



ozone maximum with constant ozone aloft profile was  not reproduced with  the



variable aloft profile.



     In conclusion, the predicted ozone maximum decreases as less material



is entrained from aloft, while the effects of chemistry on this entrained



ozone vary widely between days and mixing height  profiles.
                                    260

-------
                                          Oi. ppm
man  TO 'ON
                                                                                IT)
                                                                                 re
                                                                                a
                                                                                 c
                                                                                 o
                                                                                _E

                                                                                 to
                                                                                 o

                                                                                 (O

                                                                                 CD
                                                                                 C
                                                                                 o
                                                                                 O

                                                                                 ifl
                                                                                 O
                                                                                 O)
                                                                                 CT>


                                                                                 S-
                261

-------
                                                           _->•.   t
                                                                                 CT1
                                                                                  (O
                                                                                 Q
                                                                                  o
                                                                                 •r—
                                                                                 +J
                                                                                  fC
ul<1<1 'GN UN
                      hl-1 I I I I I J L.I 1 1  J.U ! ..I ! U
                                                                                 it-
                                                                                  CD
                                                                                  to
                                                                                  
-------
                                Table 41

                Effects of Chemistry on a Constant Ozone
                             Aloft Profile
Effective^"'
Entrained
Ozone Cone
Entrainedv ' Percentage of
Ozone Cone the Entrained
(dilution only) Ozone Survival
Day/Mixing Height Profile
Day 159,

Day 261,

Day 275,

June 7
Linear
Handpicked
Characteristic*
September 17
Linear
Handpicked*
Characteristic
October 1 at 8 AM
Linear*
Handpicked
Characteristic*

0.037
0.092
0.074

0.036
0.053
0.058

0.043
0.042
0.040

0.070
0.117
0.108

0.048
0.057
0.056

0.045
0.057
0.048

52.9
78.6
71.8

75.0
93.0
103.6

95.6
73.7
83.3
                                                                         (C)
(A) Computed as the difference between the predicted ozone maximum with
    no ozone aloft from the predicted ozone maximum using the constant
    ozone aloft profile.

(B) Considers only the effect of dilution on the entrained ozone.

(C) Computed as A/B *100.

*
 Good prediction of ozone  maximum (See PKSTM Section).
                                Z63

-------
                               Table 42

                 Effects  of Chemistry  on  a  Variable
                          Ozone Aloft  Profile
Effective'"'
Entrained
Ozone Cone
Entrained '
Ozone
Cone
Percentage of '
the Entrained
Ozone Survival
Day/Mixing Height Profile
Day 159,

Day 261,

Day 275,

June 7
Linear
Handpicked
Characteristic*
September 17
Linear
Handpicked*
Characteristic
October 1 at 8 AM
Linear*
Handpicked
Characteristic*

0.017
0.045
0.036

0.014
0.026
0.014

0.023
0.017
0.016

0.035
0.058
0.051

0.024
0.029
0.028

0.023
0.028
0.024

48.6
77.6
70.6

58.3
89.7
50.0

100.0
60.7
66.7
(A)  Computed as  the difference  between  the  predicted  ozone maximum with
    no ozone aloft from the predicted ozone maximum using the  variable
    ozone aloft  profile.

(B)  Considers only the effect of dilution on  the  entrained ozone.

(C)  Computed as  A/B *100.

 Good prediction of ozone  maximum (See  PKSTM  Section).
                                264

-------
8.4  SUMMARY
     In analyzing the RAPS data and in generating input data and  reference
data for photochemical  simple trajectory models,  significant uncertainties
were encountered.  Although used in this study,  the simple inverse distance
weighing of hourly wind data at the closest RAPS  stations   was  judged  not
adequate for defining the origin and history of  air arriving at a particular
station.  In some cases this procedure resulted  in variations of  origin
that encompassed the whole city of St. Louis;  thus divergence and convergence
problems were significant.  A mass conservative  wind field model  should be
developed for the RAPS database and trajectories  determined from  this  model.
     The three closest RAPS stations to the assumed trajectory location at
each hour often showed large differences in measured concentrations resulting
in uncertainties in starting concentrations and  in values  along the trajec-
tories.
     The uncertainties in the trajectories  and the conceptual  basis of the
simple trajectory model lead to an averaging technique  for the  emissions
inventory that resulted in a large uncertainty   for the contributions of
point sources.  Hence a significant amount  of "tuning"  could be necessary
for any particular day.  A sub-model  in the above proposed wind field  model
could provide better treatment of the impact of  significant point sources.
     The calculated composition of hydrocarbons  along the  trajectories had
the following ranges (in percent carbon):   6.7 to 14.0,  nonreactive; 4.8 -
75.5, paraffin; 10.1 to 25.1,  olefin;  6.6 to 18.1,  aromatic;  1.08 to 2.86,
aldehydes.   There are no smog chamber databases  in which the hydrocarbon
composition has been systematically varied  in  a  complex mixture to test or
develop chemical  mechanisms.  The Dodge and Demerjian mechanisms  were  both
developed on the BOM auto exhaust composition.  The CB? and CIT mechanism
were developed on synthetic mixture smog chamber  experiments.
                                    265

-------
     For the Demerjian and Carbon Bond II  mechanisms,  in static  diurnal



light simulations, changing the hydrocarbon composition over the ranges



given above without changing the hydrocarbon concentration  resulted  in 03



decreases of more than a factor of 5 (80%  reduction).   A few UNC outdoor



smog chamber experiments indicate that the models greatly overrespond  to



compositional changes in complex mixtures.



     Significant uncertainties existed in  calculating  the mixing height



profiles.  Typically, soundings were taken at 0500,  1100, and 1700 LOT.   In



this study, three techniques for converting these soundings to temporal



variation in mixing height, and thus dilution rates, were compared.  The



first method was a linear interpolation between the  soundings.   The  second



was to have a meteorologist examine each day and generate a "handpicked"



mixing height profile.  The third method was the application of  Demerjian



and Schere's Characteristic Curve developed for their  photochemical  box



model.  Each of these profiles were used for each mechanism for  each day.



     A meteorology-only simple trajectory model (MOSTM), that is, a model



incorporating initial conditions, emissions, mixing  height  rise, and entrain-



ment from aloft but excluding any chemistry, was used  to examine the rea-



sonableness  of  the non-chemistry assumptions, especially  mixing heights



and emissions patterns.



     MOSTM simulation results were that:  a) the "handpicked" mixing height



profile generally gave poor results for most days in that it was usually



too low, too long, resulting in too high early concentrations and then rose



too fast, resulting in too much dilution and low final concentrations;



b) the probable parcel pathways were poorly represented by  the "trajectory



boxes" used and therefore the emission input figures are inadequate  for



June 8, Site 15, July 13, July 19, and the last 2 hours of  August 25.  A  mass



                                    266

-------
balance for each species in MOSTM simulations showed that:   a) 4-47% of CO
mass, 2-25% of NOV mass, and 1-15% of HC mass was the result of material
                 A
present initially, b) 44-86% of CO mass, 83-98% of NOV mass, and 85-99% of
                                                     X
HC mass was due to emissions, and c)  9-24% of the CO mass  was entrained
from aloft.  This suggests that the chemistry should be emission dominated.
     A photochemical  simple trajectory model (PKSTM), equivalent to the
OZIPP "calculate" mode but having more flexibility, was used to simulate the
ten RAPS days.  There were 9 simulations for each day:  three mechanisms at
each of three mixing height profiles.  The mechanisms were  Demerjian, Dodge,
and CB2.  Only June 7 and October 1
        were simulated with the CIT mechanism because of its previous
poor performance on the BOM simulations.
     PKSTM Results.  Different chemistries dp_ give large differences in
results for the same conditions of emissions and meteorology.  The CB2
mechanism generally predicted the lowest 03 values, the DODGE mechanism
generally predicted a middle value,  and the DEM  : mechanism generally
predicted the highest value.
     For four days, 03 was underpredicted by more than 25%  by all  mechanisms
for all mixing height profiles.  On a fifth day such underprediction also
occurred for all  mechanisms for the handpicked and characteristic  mixing
height profiles.   These underpredictions are associated with very  poor
predictions of observed NO  and HC concentrations.  The MOSTM simulations
                          /\
had already suggested that these days had problems with the trajectories
and possibly the large averaging process applied to the emissions.
     For three days,  including the days with the highest and lowest observed
0^, Oq was overpredicted from  2 to 23% by the DEM   mechanism for the linear
                                    267

-------
mixing height profile.   These same three days were predicted  within  +5%  for



the DEM   mechanism and the characteristic  mixing  height profile.  The other



two days were underpredicted by 10-15% by the DEM    mechanism for  the linear



and characteristic mixing height profile.




     On different days  different combinations of mechaniams  and mixing



height profiles gave the "best fits" using  the standard  assumptions.



Slight modification in  the standard assumptions can dramatically improve the



fits.  Below is a summary of "best fits" based on  examination of CO,  NMHC,



NO  and Oo profiles:
  A      3


                         Using Standard Assumptions



         Mechanism  Mixing Height  Comment
Day



159      DEMER           CHAR



160/03   DEMER           CHAR



160/15



195



201      DEMER,DODGE     : IN



221      DEMER           CHAR






238



261      DODGE,DEM       HP



275/6am  DODGE           CHAR





275/8am  DEMER           LIN





276/6am
                                   CB2 LIN also good;



                                   LIN better in 03,  but N02 bad;  CO bad



                                   None good



                                   None good



                                   03 18% low and N02  bad; HC bad  at end



                                   N02 probably within measurement error

                                   (lowest day); missed HC peak



                                   None good



                                   DEM low on 0^, but  better shape



                                   Oo 25% low but good shape until last hr.;

                                   CO good



                                   Oq 19%, low but good shape until last hr.;

                                   CO bad



                                   None good
                                    268

-------
Day      Mechanism

275/8      DEMER


275/8    CARB BOND



238



201
Using Modified Assumptions

      Mixing Height       Comments

      CHAR
      CHAR
      LIN
      HP
5% initial aid gives
excellent fit.

BOM HC composition
including aldehyde gives
excellent fit

Modification of
trajectory/emissions still
25% low only DEM tested

Modification of
trajectory/emissions only
DEM used
                                   269

-------
                   9.0  ISOPLETHS AMD CONTROL  CALCULATIONS



9.1  OZONE ISOPLETHS GENERATED FROM PHOTOCHEMICAL  SIMPLE.  TRAJECTORY

     MODEL SIMULATIONS



     The purpose of generating an 03 isopleth  diagram  is  to  illustrate  how



the simple trajectory model's maximum predicted  0~ varies as  a  function of



increases and decreases in HC and NO  initial  conditions  and  emission values.
                                    A


To accomplish this, the trajectory models  in the previous chapter were  run



repeatedly; in each run, the HC and NO  initial  concentrations  and the
                                      X


emissions of the standard run were multipled by  scale  factors.   Figure  98



shows an example of the locations of the run points as a  function of  initial



condition of HC and NO .  These initial  conditions also represent the emis-
                      A


sions since the same factor was used for each.   That is,  ".hen the initial



conditions were halved, so were the emission values.



     When all of the simulations had been  run, a bicubic  spline surface was



fitted to the maximum 0^ and initial conditions.   The  locations of points of



constant Oo were found by interpolation  over this  surface.   These isopleths



are shown as a "level  set" by viewing the  surface  from above  ((A) in  Figure



 99).  To aid in comparison, the surface was also  "sliced" and  its cross-



section shown for various conditions (the  dotted lines in (A) of Figure 99 ).



Figures will be presented for 03 as a function of  HC 1.0, 0.8,  0.6, 0.4, and



0.2 of maximum MO  and as a function of  NO  at the same values  of maximum
                 X                        X


HC ((B) and (C) of Figure  99).  The location  of the ambient  data point,



that is, the initial ambient HC, initial ambient NO ,  and maximum ambient 0^,
                                                   A                      ^


is shown on the diagrams as a "+" ((D) and arrows  of Figure   99).



     Ten isopleth surfaces were done for each  of two days:   the Demerjian,



                                     270

-------
OOE'O
  '   0.300
                                                                                     c:
                                                                                     03
                                                                                     Ol
                                                                                     U
                          0.200
                                               0.100
                                    NOX.PPM
                                                                                     -O
                                                                                     0)
                                                                                     to
                                                                                     c
                                                                                     O
                                                                                     (O
                                                                                     r-~
                                                                                     =3

                                                                                     •r~
                                                                                     OO

                                                                                     i-

                                                                                     M- Cl)
                                                                                        O
                                                                                     (/) to
                                                                                     C 4-
                                                                                     o s-
                                                                                     O .C
                                                                                     O •(->
                                                                                    _l O)
                                                                                        r—
                                                                                    4- CL
                                                                                     O O
                                                                                        l/>
                                                                                     .O) i— i

                                                                                     Q. O)
                                                                                     E C
                                                                                     fO O
                                                                                     X N
                                                                                    uj o
CO
CD
                                                                                     OJ
                          271

-------
                                                          E
                                                          3
                                                       E
                                                       o   co
                                                      •r- O
                                                   .«4_)
                                                  ^^  CO T3
                                                  CO .—  E
                                                  ---  3    O)  4->
                                                   C "O  -i-
                                                   (D  O  "O
                                                  +-> ET'c
                                                   ^C_>
                                                   O  i-
                                                  O  O i—
                                                       -t->  re
                                                   -M  O T-
                                                   fO  OJ -4->
                                                       •1-3-1—
                                                   T3  ro E
                                                   O)  S- i— t
                                                   i— O)  O
                                                   OO r—
                                                       D- l/l
                                                    (U E  CD
                                                    U T- +•>
                                                    na oo    O)
                                                    O) E -r-
                                                    O  ro  <4-
                                                    (T3 •»-> -r-
                                                    <4-  tO -M
                                                    S-  E E
                                                    Z3  O O)
                                                    CO C_) -O
                                                     OJ
                                                     Q. O)
                                                     O O  •"
                                                     E  d)  >•
                                                     O  O  ra
                                                     NJ  n3 Q
                                                     E  3  O
                                                    eC CO M-
                                                    CTt
                                                     OJ
272

-------
Dodge, and CB2 mechanism at each of three mixing height profiles and the CIT



mechanism at the characteristic curve mixing height profile, for October 1,



1976, Day 275 at 8 am and for June 7, 1976, Day 159.   October 1  was the high



0- day for the year (0.244 ppm 0~).  The initial HC-to-NOv ratio was 9.4 and
 O                              O                        X


the emissions HC-to-NO  ratio (determined for MOSTM simulation)  was 3.35.
                      A


Of the ten days studied, June 7, 1976 had the highest initial HC-to-NO  ratio,
                                                                      X


13.2,and the highest 0., aloft (0.12 ppm).  It was also the only  day simulated



well by all mechanisms.  The isopleth levels used were:  0.08, 0.12, 0.20,



0.24, 0.28, 0.30, 0.32, 0.34 and 0.40.



     The results for October 1, Day 275, are shown in Figure 100 A,B,C,D, E.



Recall that this day wasunderpredicted by DEM2758.LIN simulation, was under-



predicted slightly by DEM2758.CHAR simulation, was underestimated about 10%



by both the DOD2758.LIN and .CHAR simulations, and was underpredicted greatly



by all other simulations (all  .HP, all  CB2, and CIT.CHAR).



     In Figure 100B, there are obvious  large difference in the isopleths for



different mechanisms-mixing height combinations, even where there were less



obvious differences among the PKSTM simulations.  The Demerjian, Carbon Bond,



and CIT mechanisms have "steeper" sides than the Dodge mechanism.  That is,



the Oo formation is more responsive to  HC and NO  in  these mechanisms than in
     *5                                          X


the Dodge mechanism.  The response to NO  is so great in the CB2 and CIT
                                        A


mechanisms that they cannot simulate this day.   At low NO , they exhibit a
                                                         X


higher 0., dependence on HC than Dodge and therefore have more, and more



closely spaced, isopleths than Dodge.



     The "flatness" in response to HC and NO  gives the Dodge isopleths a
                                            X


characteristics "U" shape.   The other mechanisms have more "V" shaped iso-



pleths.   These shapes have important implications for EKMA control  calcula-



tions which will  be discussed  in the next section.   The U-shaped isopleths




                                     273

-------
                                                            o 600   o son    e 100    a 300
                                                                                   Di.rrr
                                                                                           a 200    4 voo    o            0 toe   t
I (    111 I    «0. D    IOC I   HE •    Bit (
                                                                                                                                                                                    in
                                                                                                                                                                                    c
                                                                                                                                                                                    o
                                                                                                                                                                                    03
                                                                                                                                                                                    cu
                                                                                                                                                                                   -0
                                                                                                                                                                                    O
                                                                                                                                                                                   •(->
                                                                                                                                                                                    O
                                                                                                                                                                                   O
I 0    ooE'o    oo» o    oat a    ooe 0    oot o        0
                                                        ooe 0   oot I    oot  0    ooc 0    OOE 0    ool 0       0 "      001 0    las o    00* 0    Oot 0   00: 0    OOt 0
                                                            0 too    0 MO   0 *00   0 >00    D 100    0 tod
o
CD
                                                                                                                                                                                    CO
                                                                                                                                                                                    CO
                                                                                                                                                                                    o
                                                                                                                                                                                    i.
                                                                                                                                                                                    CJ

                                                                                                                                                                                    CJ
                                                                                                                                                                                     a>
                                                                                                                                                                                     c:
                                                                                                                                                                                     o
                                                                                                                                                                                     N
                                                                                                                                                                                    o
                                                                                                                                                                                    o
                                                                                                                                                                                    o
                                                                                                                                                                                     ai
                                                                                                                                                                                     i_
                                                                                                                                                                                     3
                                                                                                                                                                                     O)
                                                                                                                        D tOO    0 500    0 II
                                                                                                                                               0 300    0 200

                                                                                                                                               09 ffH
                                                                                          274

-------
got o        coz
                                                                                                                    in
                                                                                                                    C
                                                                                                                    O
                                                                                                                    ro
                                                                                                                    I.
                                                                                                                    Ol
                                                                                                                   -D
                                                                                                                    O
                                                          275

-------
                                             got 0     D
                                                                if)
                                                                C
                                                                o

                                                               +j
                                                                IB
                                                                s_
                                                                Ol
                                                                o
                                                               o
                                                                co
                                                                c
                                                                o
                                                               -
                                                                CJ
                                                                CO
                                                                to
                                                                O
                                                                 X
                                                               o
                                                                O)
                                                                £=
                                                                o
                                                                tsl
                                                               o
                                                               CJ
                                                               o
                                                               CD
                                                                CD


                                                                3
                                                                CD
276

-------
                                                                                                                                                             l/l
                                                                                                                                                             c
                                                                                                                                                             o
                                                                                                                                                             CO

                                                                                                                                                             a>

                                                                                                                                                             i.


                                                                                                                                                             O

                                                                                                                                                             o
                                                                                                                                                             CD
                                                                                                                                                            -t->
                                                                                                                                                             O

                                                                                                                                                             n3


                                                                                                                                                             rC
flat o    is: I   III I   lit I   oil o   lita a
                                                 101 I   IDS I   lot 0
                                                                             to: o   lot I
                                                                                                                                                             S-
                                                                                                                                                             a>
                                                                                                                                                             o
                                                                                                                                                            o
                                                                                                                                                            •
                                                                                                                                                             o
                                                                                                                                                             a>
                                                                                                                                                             1/1
                                                                                                                                                             O
                                                                                                                                                             i,
                                                                                                                                                             o
                                                                                                                                                             O)
                                                                                                                                                             Q.

                                                                                                                                                             O
                                                                                                                                                             cu
                                                                                                                                                             c
                                                                                                                                                             o
                                                                                                                                                             M
                                                                                                                                                            O
                                                                                                                                                            CD
                                                                                                                                                            O
                                                                                                                                                            O
                                                                                                                                                             a>
                                                                                                                                                             s_
                                                                                                                                                             3
                                                                                                                                                             en
                                                                               277

-------
                                                              o
                                                              •I—

                                                              (O
                                                              i.
                                                              o>
                                                              .0
                                                              o
                                                              4J
                                                              o
                                                              o
                                                               s-
                                                               0)
                                                              o

                                                              T3
                                                               O
                                                              DO
                                                               S-
                                                               10
                                                              CJ

                                                               s.
                                                               o
                                                               Ol
                                                               CD
                                                               c
                                                               rt)
                                                               s_

                                                               s_
                                                               Ol
                                                               Dl
                                                               S-
                                                               i_
                                                               0)

                                                               o

                                                               to
                                                               .C
                                                               4->
                                                               Ol

                                                               CL
                                                               O

                                                               •r-

                                                               Ol


                                                               O
                                                               IM
                                                               O
                                                               LU
                                                               o
                                                               o
                                                                O)
                                                                S-
                                                                cn
278

-------
generally "spread" as the HC and NO  inputs are increased,  that is,  the  iso-
                                   X


pleth location moves to lower and higher HC-to-NO  ratios with increasing
                                                 A


HC and NO  values.  The V-shaped isopleths, however,  tend to parallel  a  par-
         X


ticular HC-to-NO  ratio as the HC and NO  inputs are  increased, that is  a
                X                       X


particular 0, value may never occur at a particular HC-to-NO  ratio  no matter
            %5                                               X


how much HC and NO  input is used.   Figure 100E illustrates  CB2 isopleths
                  A


over a larger range of HC and NO  than those of Figure 100B.  It is evident
                                A


that the fifth isopleth (0.24 ppm Oo) will  never intercept  a line drawn



through the origin  and the "+" which marks the HC and NO  for October 1.
                                                         X


Therefore, mechanisms which exhibit V-shaped isopleths must predict  the



observed 03 maximum for present emission conditions (the  PKSTM simulation)



within reasonably narrow limits (maybe +10 or 15%)  if a standard EKMA  pro-



cedure is going to be used.



     The results for June 7, 1976,  Day 159, are shown in  Figure 101A,B,C,D.



This day was overpredicted by DEM159.LIN simulation,  predicted within  5% by



DEM159.CHAR, CB2159.LIN, and CIT159.LIN simulations,  and  was 12% underpre-



dicted by DEM159.HP, was 25% underpredicted by the DOD159.LIN, CB2159.CHAR,



and CIT159.CHAR simulations.  The rest were more than 25% underpredicted.



     The first observation to be made about the June  7th  isopleths is  that



the Demerjian handpicked isopleth is missing.   There  was  apparently  a  closed



isopleth for this condition and the computer code to  find isopleths  on the



On response surface was not designed to deal  with closed  isopleths,  therefore,



no surface was depicted.



     The second observation  is that the 03  value at the origin was nearly



0.10 ppm Oo for most surfaces.   This is because of the very large value  of



Oo aloft on this day.   Since the upper left and lower right simulations



(zero HC, 0.07 NO  and 1.0 HC and zero NO )  interacts with  the entrained 0,
                 A                       A                                 O


                                      279

-------
           DOC 0        OPE t        DP! 0
    g 400        0 30C
                                                          t 400        o loo
oot o       oorc       BOC o
                          0 200
                          M.rrn
                                                                                                              00i D       OOC 0       001 0
                                                          0 100       D 300
100       0 SOU       0 300        0 10(
                    0) rrn
                                                                                                                                                                           C
                                                                                                                                                                           O
                                                                                                                                                                           (O

                                                                                                                                                                           Z3
                                                                                                                                                                           E
                                                                                                                                                                           CL)
                                                                                                                                                                           c
                                                                                                                                                                           13
                                                                                                              tut o       ooc 0       ooz 0       oat c
                                                       o
                                                       QJ
                                                       to

                                                       CO
                                                       I/)
                                                       O
                                                       s_
                                                       o

                                                       C_3
                                                       n:
                                                         i
                                                        cu
                                                        c
                                                        o
                                                        N
                                                       O
                                                                                                                                                                            O)
                                                                                                                                                                            5-
                                                                                                                                                                            Z3
                                                                                                                                                                            CD
                                                                                       280

-------
•ow o eoso a  oe*o o ooco o  oozo o  so
                                                       oow o  one g  »to 0  «io-o  o«o o  0010 o      o •
                                                            0 MOO 0 MOO  0 0100 0 WOO  0 0200 I 0100  9
                                                                            not rrN
                                                                                                                                                                           CO
                                                                                                                                                                           C
                                                                                                                                                                           O
 (O


 I
•I—
 (/)

r-»

 
                                                                                                                     0 0600  0 OSOI
                                                                                     281

-------
•ot •       tie «
                                                     tOt I       OBI «       001 I
                                                                                                          eot o       oat a       eo; •       out e
                                                                                                       .
                                                                                                     is
  t g       BBC «
                                                               oot o       DO; a       oai o
                                                                                                           tot i       oct o       eoi a       oot <
                                                                                                                                     a 209       0 100
                                                                                                                                                          V  I
                                                                                                                                                                        (O
                                                                                                                                                                       r—

                                                                                                                                                                        3
 0)
 c
 ^
o


 l/l

 o
                                                                                                              • 0 »BO       0 301
                                                                                                                                                                         o
                                                                                                                                                                         ai
                                                                                                                                                                         I/)

                                                                                                                                                                         CO
                                                                                                                                                                         co
                                                                                                                                                                         o

                                                                                                                                                                         o
                                                                                                                                                                          I
                                                                                                                                                                         CD
                                                                                                                                                                         c
                                                                                                                                                                         o
                                                                                                                                                                         N
                                                                                                                                                                        O
                                                                                                                                                                        O
                                                                                                                                                                        i—

                                                                                                                                                                        O
                                                                                                                                                                         a>
                                                                                                                                                                         s-
                                                                                    282

-------
                    o
                   •r—
                   4->
                    CO
                   •r—
                    i~
                    OJ
                   -l->
                    o
                    (O
                    S_
                    (O
                   .c:
                    o

                   r^.

                    O)
                    c
                   o
                    o   •
                   M- C
                       o
                    QJ  (T3

                   'o-'o
                    o  E
                    CO  •!-
                   •I—  CO

                    0)  
                    o  s_
                    N  3
                   O  
-------
differently for each mechanism,  the CL  values  in the corners  are  different.
This same isopleth surfaces have a 0.08 ppm 0, isopleth  that  intercepts the
edge and some have no 0.08 ppm 03 isopleth at  all.   Because of a  high  03
photolysis reaction (R23 in Table 9  )  in the  CIT mechanism,  it consumes
03 near the origin (Figure 101D), resulting in a 0.08 ppm 0~  isopleth  that
only touches the HC axis.
     The Dodge surfaces exhibit almost  no response  to HC and  only a  low
response to NO .  That is, the 0^ vs HC at constant NO  cross-section  plots
              X                 -J                     A
(Figure 101A),the lines are nearly horizontal  but do show increasing Oo
(only slightly for HP and  CHAR), increasing constant NO   levels.   As in the
                                                      X
isopleths for Day 275, the Demerjian and CB2 surfaces show a much  stronger
dependence upon both HC and NO  than the Dodge surfaces.
                              A
     On Day 275, the LIN surfaces are very similar  in appearance  to  the CHAR
surfaces.  On Day 159, the LIN surfaces and the CHAR surfaces differ substan-
tially in appearance.  These similarities and  dissimilarities will be
examined further in the EKMA control calculation discussion.
     Although somewhat affected  by the  high 0,, aloft values,  the  U-shaped
versus the V-shaped character of the Dodge versus the Demerjian and  CB2 iso-
pleths is still evident.  It is  expected, therefore, that, if  the  upper right
corner values of DOD159.HP and DOD159.CHAR were made high enough, the  corner
03 value would exceed 0.20 ppm 03, the  value needed for  an EKMA type calcu-
lation for Day 159.  These high concentration  simulations were not done.
                                     284

-------
9.2  OZONE CONTROL CALCULATIONS FROM PKSTM ISOPLETHS AND EKMA PROCEDURE



     The isopleths in the previous section show the effects of emission



changes on simple trajectory model's 0,, predictions.  The Emperical Kinetics



Modelling Approach (EKMA) is a method for translating the models'  response



to emission changes into atmospheric control  requirements.



     Stated in its simplest form, the EKMA procedure is (Figure 102):



a) to find the point of interception between  a radial line at the  initial



HC-to-NO  ratio and the isopleth for the ambient 0? maximum for this day;
        X                                         v


b) divide the values on the NO  axis by the NO  values at the point of inter-
                              X               X


cept; c) draw a line at constant NO  through  the intersection point, draw  a
                                   X


line on which NO  increased 20% for 50% HC decrease, and draw a line on which
                /\


NO  decreased 20% for 50% HC decrease; d) the values of 0,, along these lines
  /\                                                      O


as a function of HC fraction are assumed to represent the 0- values in the



atmosphere as a function of ambient HC emission reductions.



     Note that in the example in Figure 102, the model was not perfect.  That



is, the DOD2758.LIN PKSTM simulation, with our best estimates of initial



concentrations (0.202 ppm NO  and 1.89 ppmC HC) and HC NO  emissions, only
                            X                            X


produced 0.214 ppm 03 instead of the observed 0.244 ppm Oo (observed-to-pre-



dicted ratio 1.14).   The EKMA solution point  occurred at 2.38 ppmC HC and



0.254 ppm NO , a ratio of calculated-to-observed initial conditions and
            X


emissions of 1.26.   In other words, the model's emissions had to be increased



by 26% to get 14% more Oo-   These increased emissions become the starting



point for the control calculations.  Following down the 0,-HC fraction line



of Figure 102D shows that 0.12 ppm 0~ is reached at a HC fraction of 0.62,



that is, to achieve the Oq  standard with no change in NO , a 38% reduction
                         O                               A


in HC is required by this model.



     The EKMA solution technique was incorporated into the isopleth generation




                                      285

-------
     NOX


  Solution
                              Actual 8 AM HC and NOV Values
                                                   X
The EKMA Solution

Point
                                               The Isopleth for

                                               the Observed [Ojmax
                                               (A)
                                               (B) and (C)
                                                     (D)
                                                  Standard
                                           .20%  NOV  increase
                                                  A

                                          —No NO  change


                                            20% NO  decrease
                                                  A
Figure 102.   The EKMA Procedure



                         286

-------




Q
2:
0
'JO
O
CD
•Si







in


o
LO

5
UJ



CM

X
0


1 1
LO CO

1 1


:r







\—
GO
CL.





• —
*"
O i —
U3 «*
ro r~-»



CO
o

C5O CO
^jO ro
0 0

0 0
O
<0
*^
E

CO
Ol

_
,-_


cu
-d UJ

4-> 10 kf Q
U m ro OJ O
O QJ O • O

t- i —


xp
in •»-> o OJ
+-> c z: •
C QJ lO

O -0
a. E s- o
i — ro
CM i—
0 0

+-*
^_
o
CO



UJ


ro
^T*

cu z:

_Q >— .
ro *~D
h— a:
UJ
E:
UJ

















_Q
•z.
—I
o
CO
rf*

*s<
UJ



0
C\J


ro


t-H

X)
X


en i
CO 00
OJ Z
1
o



o
31

ro
o
i —
OJ







3£

LO
i«i
CL




u
,_- .


en
ro «at>
OJ i— 1

^ CO





o

r^H ^~-

<-> o

o o






CU

C JZ T-
•r- Q1H-
X T- O
— (D i-
s; :c o-

S-
(T3
OJ
C
•r- 0.
— i 31

i
CO

1





in
Cn
IT)

, 	
•y
o

d


o
en
og
p™

lO
OJ
Q


cn
CO
•3-

OJ



OJ

, 	




0
OJ
o




o


OJ


•a-
OJ
"I
d



^o
Ol
ro



OJ
in
«^

r-H


00
lO


o









s-


o








































l/l

p_

1/1
OJ
CO
E T3
O- O
0- E
CU >i O

'O O O
*J C
w o a.
•4-> OJ CL

C t)

+-> 'cu
0 OJ 0
"O CL
c E in


X -l-J
O i — OJ
U 'r-

r~~ E
IO O) •—

f. U U


OL O OJ
0- JC Q.
Q- E
OJ UJ
i- OJ
fa .£: oj
•*-> c:
to -*->
•*-> to
•i- c tn
C 0) C
Z3 OJ 03
E OJ
0 £
3; s:
i — «f
I— CO S

5 CX UJ


ro -Q





























in
QJ

r—



-M
C
CU

JQ
E


CO

-o
OJ m
+-> c
o o

•a •*-»
s- "a
o. c

t—
CO !—

D. •-—

5 I

o -a
cu
OJ •>->
3 TO

> U


C U
QJ

•S §
E 2
(O UJ

M- *4-~
0 O

o o

4-J 4 >
n3 ry
!_ t.

CU  •*->

in i/)


i— OJ
s- s~


o







































-a
c
13




O
c
c
o

•*-*
,H
o

t/1
c

QJ
E
I
u_
i

• «•
QJ

JO
In

0
u.

c
0

•f-*


o


o
c

to
c
aj
OJ
e

i


i


-o
287

-------
o

4->

-------
ggf a   oo* a    oot a    eoc o    DDE o    ooi o
           3 5:0   9 'GO
009 S    ODE 0    OOi 0    OOt 0    001 0    DD1
    D too    e 500    e HO   a 100   0200   o i
                                                                                                                                                                    to
                                                                                                                                                                    c
                                                                                                                                                                    o
                                                                                                                                                                    E

                                                                                                                                                                    c/1
                                                                                                                                                                    0)
                                                                                                                                                                   .a
                                                                                                                                                                    o
                                                                                                                                                                   -i->
                                                                                                                                                                    o
                                                                                                                                                                   o
 to
4->
 O

"a.

c;
o
                                                                                                                                                                    o
                                                                                                                                                                    13
                                                                                                                                                                    0)
                                                                                                                                                                    c
                                                                                                                                                                    o
                                                                                                                                                                    N
                                                                                                                                                                   O
                                                                                                                                                                   CO
                                                                                                                                                                   o
                                                                                                                                                                   O)
                                                                                                                                                                   i.
                                                                                  289

-------
                                                                                          * I      601 I      001 0      001 0
tot o      ooc a      oo; o      oo
                                                                                                                                           13
                                                                                                                                           E
                                                                                                                                           s_
                                                                                                                                           o
                                                                                                                                           1/5

                                                                                                                                           O

                                                                                                                                          'o.

                                                                                                                                           c
                                                                                                                                           o

                                                                                                                                          +J
                                                                                                                                           o
                                                                                                                                           Ol
                                                                                                                                           c
                                                                                                                                           o
                                                                                                                                           N
                                                                                                                                           O
                                                                                                                                           •=}•
                                                                                                                                           O
                                                                                                                                           O)
                                                                                                                                           s_
                                                                                                                                           3
                                                                                                                                           en
                                                                       290

-------
program, so that the EKMA solution point could be found and Oo-HC fraction



plots automatically produced.   The procedure was  applied to the ten isopleths



for October 1  (Day 275)  and the nine isopleths for June 7 (Day 159) and  the



results are shown in Table 43 and 44 and Figures  103and 104.   As had been



indicated in the isopleth discussion of the previous  chapter,  not all  iso-



pleths produced solutions, and some could no^, even if 100 times the initial



conditions were used.



     The data  in Figures 103and 104 have been converted into control diagrams,



that is, plots of percentage 0~ reduction versus  percentage HC reduction, and



these are shown in Figures 105to 106-   The HC reduction necessary to achieve



the 0.12 ppm 03 value are summarized in Table 45  and  in Figure 107,



     From Figure 107 it is evident that the most  important factor in the



amount of HC reduction needed was the choice of day.   Day 159  (June 7) was



not the day of highest 03; it was the day with maximum 03 aloft (0.12  ppm).



This Oo aloft  is not reduced by emissions reductions  in the same manner  that



Oo produced from emissions is reduced by emissions reductions.   This resulted



in "flatter" ozone control diagrams (Figure 106).



     The next  most important factor in the amount of  HC reduction needed was



the chemical mechanism used.  When it could give  a solution, the Carbon  Bond



mechanism gave the least HC reduction needed (other conditions  the same).



The Demerjian  mechanism, when it could give a solution, always  required  less



HC reduction than Dodge  (other conditions the same).



     The third most important factor was the choice of mixing  height profiles.



The characteristic curve profile, in general, required less HC  reduction than



the linear profile.  Except for the Dodge mechanism,  solutions  were not



possible with  the handpicked profile.   For the Dodge  mechanism,  the hand-




                                      291

-------















>}
_Q

T3
CU
CJ
•r—
"O
CU
S-
D-
°C?i
Q.1
Q- «
r^.
CM
r— CU
• c:
O 3
cu
> T3
CU E
•t- ro
-c
O UD
*-C I*"**
Ch
0 i—
-l_)
*
£ZZ r—
O
•r™ S*.
-^ CU
O _Q
3 O
•0 4J
CU U
a; o
O i-
31 O
^
4-3
C~ (/)
CU r-
O CU
i- -0
0) O
Q- s:

•

LO
^f-

cu
1 —
t~}
CO
1—






















-o
c
0
CO

c
o
.a
^_
re
o







X
0
z
0
o
CM



LO CO
OO CM





X
o
^"

^i
o

III 1
CO CO CO i — CO LO
z: -z. -ZL *& -z. oo
iii i



X
o
z
&5
o
CM
1



O LO
LO *J-













O)

ID
O
Q











X
o



*sj~ CM ^i" r —















C
•r-
*' — D
s-
cu
e

Q









X
o
"^

0^
0
CM
-f-



r~~ CNJ o o
1— 1 i — LO LO




X
o
z:

cH5
O

1 1
CM co LO CD co r^
CM 2: i — LO sr LO
i i



X
o
TZL

^s>
0
CM
1




O CM LO (^
OO CM l£> CO








0)
C7)4-> r—
•r- O>M-
X -r- 0
•i— CU S-
21 nr ex.
r—
S_ S-
s- to res
CU CU n- r~-. CU rv
o •>- D- x: cu -r- a. re
-u—jnrcj c_i3co
U 3
O 0












































0)
CD

rO


C
O
•r—
4->
S_
I ^
C
cu
CJ

o


0) C
3 """
•r- -O
to c
LO !3
0 0
Q. H-

c a
0 C
**~ »i —
4-> -t->
3 13
i — i — -
O O
CO CO

0 O
c c:
c/i to
c c
(O res
cu cu
e e
i i
CO Lu
Z 2:
1 I
n3
292

-------
                                                                                             NOI10n03M "0 I
                                                                                                                              o
                                                                                                                              to
                                                                                                                              S-

                                                                                                                              o
                                                                                                                             -o
                                                                                                                              CD
                                                                                                                              O)
                                                                                                                              O)
                                                                                                                              o
                                                                                                                             •i —
                                                                                                                             •t->
                                                                                                                              o
                                                                                                                              3
                                                                                                                             -o
                                                                                                                              CD
                                                                                                                             a>
                                                                                                                             sr
                                                                                                                             •a
                                                                                                                             O)
                                                                                                                             fO
                                                                                                                             Q
                                                                                                                             

                                                                                                                             O
                                                                                                                            O
                                                                                                                             cn
                                                                                                                             03
S   3
                                                        293

-------
                                                              CO
                                                             o
                                                             a.
                                                             a.
                                                             o
                                                             ITS
                                                             a>
                                                             s-
                                                             OJ
                                                             -o
                                                             a>
                                                             O)
                                                             c

                                                             E
                                                             o
                                                             o
                                                             -a
                                                             CD
                                                             OJ
                                                             c
                                                             O)
                                                             ^:
                                                             CO
                                                             
-------
D CD CD C
3 cn oo r
-1
_' 1 ' 1 '
—

—
	 4—G 	 1
_
	

	

~, 1 , 1 :
• * •
3 O CD C
D cn co r
-»
:> CD CD C
a? in =t
. , . | .
^_ 	
1 fi i

h
i n I

^.

4. 	

.1,1.
* • •
3 CD CD C
\T> LO =1
3 C
a
i
Q






O-l-

1
3 C
r c
3 0 CD C
:> c\j -r-4
1 1 ' 1 '_
—

—
—

	 J. _„
». Q -I-
«•! • 1
O4-
.1,1,"
• • «
D CD CD C
0 C\J *-•
3
n

O

C_5
a

o
i~}
*^j



D

Q_
~T~

cC
1C

2T
i—)
_1
Q.
_L
 cn
II C
CO ••-
X
CTi OJ E
i — -a o
O -i-
ro II ai
Q Q -r-
s_
^ QJ
r-- c: 4->
ro <_>
a> -1—03
c -i— > s-
zs s- ro
•-3 CD -C
E= 0
CD II
Q C£
II <
•-5 in
E -
01 -a
•r~ O!
C -il=C
(a
LD -a -
rv s_
CM >, 03
jQ , C
Q C r-
O II
i — -4-> i— (
O 	 1
S- •<-
QJ -a
X3 CJ
O S- l/>
4-> Q. £ •
U (/)(/>
o • — -r- at
S_ ra -r-
4-> -C 4—
O CU t-
cj 2: cx
r-x
0

-------
picked profile gave lower control  requirements than  the other profile.   These




solution points, however, were often at very large HC  and NO  concentrations.
                                                            /\
                                      296

-------
                10.0  RAPS SIMULATION WITH LEVEL III MODEL


     A Level III analysis has much less stringent data requirements than a
Level II analysis because of simplifying assumptions inherent in  the Level
III model.  The Level  III model  assumes that the air parcel  containing the
measured 0^ maximum originated in mid-city at 8:00 am, traveled in a
straight path at a uniform speed such that the parcel  would  arrive at the
monitoring station at the time that the ozone maximum was measured.
Therefore a back trajectory need not be calculated and  an
extensive wind direction-speed monitoring network is not required.   Wind
direction data is only required to insure positive identification of the
correct downwind monitoring station.   A gridded emission inventory is not
required.  Only city or county wide average emissions are utilized, since
the actual trajectory is not determined.
     The air quality data requirements are also much less for Level III.
Only three 0^ monitoring sites and one or two THC/CH.  and NO  monitoring
sites are recommended, as compared to the 7-11 03 and 4-6 THC/CH. and NOX
sites recommended for Level II  modeling.
     Control requirements are estimated using procedures described in
EPA 1981.  These are estimated using the ozone design value
and prevailing 6-9 am NMHC/NOV ratio to identify a starting  point on the
                             A
isopleth diagram.


10.1  LEVEL III INPUTS AND CONDITIONS
     Three of the ten  days previously simulated with the Level  II approach
were picked to be simulated with the Level  III approach:   D159, D195, and
                                   297

-------
D275.  Only the Demerjian and Carbon Bond II mechanisms were used.   Level  III



isopleths utilizing the Dodge mechanism for these days  were  supplied by



OAQPS/EPA.



     As explained in the EPA EKMA guidance document (EPA 1981),



Level III assumptions include an initial  morning 8:00 am mixing  height of



250 meters.  Additional simulations  and isopleths were  performed with the



Demerjian mechanism with assumed initial  mixing heights of 350 meters for



comparison.  Only the characteristic curve for mixing height profile was



utilized.  In each case, the measured final mixing heights were  used in  the



characteristic curve calculation.   The input data were  supplied  by  OAQPS/EPA.



     The initial conditions of NMHC  and NO. the observed Oo maxima, the
                                          X               «j


initial  and final mixing heights,  and the assumed county/hr  path for each



day are given in Table 46.   The county emission densities are given in



Table 47.  The NMHC distribution and carbon fractions are shown  in  Table 48.



The latter values were recommended by Demerjian (personal  communication,



1981).



     For comparison with Level  II  simulations (Figure 79), Figures  108 and



109 show the plots of NO, N02»  and Oo profiles along the assumed trajectories



for each simulation.  Note  that two  mixing heights (250 meters and  350 meters)



were used for the Demerjian simulations.



     Table 49 lists the 0^  maxima predicted for each simulation, the observed



Oo maxima for the day, and  the ratio of the two values.  Also included in



Table 49 are the Level II simulations results for comparison with the Level



III results.



     An important point is  that although the Level II  simulations end at the



time (hour) of the observed 0,, maximum the guidance suggests that Level  III



simulations be commonly performed for 10 hours regardless of the time of




                                    298

-------
Table 46.   Initial  Conditions  and Observed [0^]  Max
           for Level  III Simulations
Item
[NMHC]Q ,
[NOx]o '
[03] max ,
Max Time
[03] aloft
D159
ppm 1.79
ppm 0.205
ppm 0.192
1700
, ppm 0.12
Init Mix Ht. , m 250,350
Final Mix
Trajectory
Hour
800
900
1000
1100
1200
1300
1400
1500
1600
Ht. , m 1900


St. Louis City
11
11
St. Louis Co.
St. Charles Co.
11
Madison Co.
11
Jersey Co.
D195 D275
0.26 1.90
0.048 0.236
0.210 0.244
1700 1600
0.074 0.06
250,350 250,350
1800 900


St. Louis City St. 1
II
II
II
M
II
M
n
St. Louis Co.
                         299

-------
                Table 47.   Area Wide Emission Densities for
                           Level  III Simulations
     County
St. Louis City

St. Louis County

Madison

St. Charles

Jersey
                        Kg-mole/(Kmz-hr)
                           ppm(C)-Km/hr
HC

2.716
0.533
0.163
0.070
0.014
NOV
X
0.465
0.148
0.044
0.092
0.003
HC

0.06641
0.01305
0.00406
0.00174
0.00029
NOV
X
0.011362
0.003634
0.002254
0.001058
0.000092
             Table 48.  NflHC Compositional Carbon Fraction3
                        for Morning Initial Conditions and
                        Emissions for Level III Simulations
                        (from Demerjian 1981)
       Species
Initial
Emissions
Avg. Carbon Number
NR
OLE
PAR
ARO
ALD
0.12186
0.18195
0.43596
0.18549
0.07478
0.12791
0.19089
0.47965
0.17054
0.03101
3.0
3.0
6.3
7.1
2.0
       aTo obtain ppmV for each class, multiply total NMHC  in ppmC
        units by fraction and divide by average carbon number.
                                   300

-------
0.. PPM
 ">« 'MN 'OH
                                                                                       o
                                                                                       LO
                                                                                       c\j

                                                                                        en
                                                                                        c
                                                                                       •r—
                                                                                        CO
                                                                                        3
                                                                                       LT)
                                                                                       t^
                                                                                       CM   •

                                                                                       -oi
                                                                                        c to
                                                                                        03 T-
                                                                                        LT>
 to  o
 >>DQ
 (O
•o  c
    o
CO -Q
Q-  S-
eT  (O
c2 o

 oi -o
 cu  c
 S-  fO

4->  C
    to
 S- -i-
 O T-)
M-  t.
    CU
 to  E
 c  cu
 O Q
                                                                                        to  en
                                                                                            CD
                                                                                        r—  C
                                                                                        O) -r-
                                                                                        >  X
                                                                                        CO
                                                                                        o
                                                                                        cu

                                                                                        3
                                                                                        CO
                                                      '"ON 'ON
                                       301

-------
                                              (C
                                             o
                                             tn
                                             CO

                                              cr>
                                              c

                                              to
                                              3


                                             Lf>
                                             -o
                                              c
                                              (T3
                                             un
                                             01
                                             IT)
                                             T3
                                              CO CO

                                              O- T-
                                                 to
                                                 ^:
                                                 u
                                              o -o
                                              co  E
                                              c  cu
                                              -P  S-
                                              res  o
                                              CO  CD
                                              i-H -E
                                              I — i

                                                  cn
                                              r-  C
                                              (1) -1-
                                              >  X
                                              en
                                              o
                                               cu

                                               3
'•OH  'ON
                             302

-------
     Table 49.  Ozone Maxima Predicted in Level  III Simulations of
                Selected RAPS Days.   Comparison  with Level  II  Simulations


      Item            June 7 (159)        July 13 (195)        Oct 1  (275)

ACTUAL DATA

  [0,] max, ppm          0.192               0.210             0.244
  Time max               1700                1700              1600


LEVEL III SIMS.

  DEM/250 Ma
    [Oq] pred         0.302              0.216  0.226       0.418  0.418
    Time pred         1700               1700   1800        1600   1800
    Rlc               0.636              0.972  0.929       0.584  0.584

  CB2/250 Ma
    [0,] pred         0.260  0.262        0.143  0.156       0.250  0.282
    Time pred         1700   1800        1700   1800        1600   1800
    Rl                0.738  0.733        1.469  1.346       0.976  0.865

  DEM/350 Ma
    [OJ pred         0.332              0.210              0.458
    Time pred         1700               1700               1600
    Rl                0.578              1.00               0.533


LEVEL II SIMS.
  DEM/CHARb
    [Oo] pred         0.197              0.133              0.168
    Time pred         1700               1700               1600
    Rl                0.975              1.579              1.452

  CB2/CHAR
    [0-1 pred         0.154              0.097              0.081
    Time pred         1700               1700               1600
    Rl                1.247              2.165              3.012
  Code is DEM = Demerjian Mechanism;  CB2  =  Carbon Bond  II  Mechanism
  number after / is initial  mixing height in meters;  characteristic
  curve used in all level  III  simulations.

  Code is same as in a;  CHAR = characteristic curve  using  individual
  day's initial mixing height.
p
  Rl  is ratio of actual  day's  Oo to simulation predicted 0-,.

                                   303

-------
 the observed 03 maximum.   Therefore,  Table 49 lists  two  sets  of values  for



some simulations to indicate the effects  of the extra time.








10.2  DISCUSSION OF DAILY  SIMULATIONS



     Level  III simulations predicted much higher [0-j] max than the  Level  II



counterparts (see Table 49).  Only one simulation,  CB2/D195  underpredicted



the observed [0^] maximum. For the Level  II simulations of these three days,



only one (D159, a day dominated by high 03 aloft) was overpredicted (the



Demerjian mechanism^by -2%).



     There are three major reasons for these results:



     1)  Level III simulations start in center city with  250 meter  mixing



         heights and maximum center-city concentrations resulting in



         more mass initially present than in Level  II;



     2)  the average hourly emissions  of HC and NOV are often  higher in
                                                  A


         Level III simulations than average emission  values  in Level II



         simulations, because the assumed trajectories start in a high



         emissions density region; and



     3)  differences occurred in assumptions of emission  inventory



         composition in Level III as compared to Level II.








     Initial concentrations in Level III simulation are commonly higher



than Level  II since it is  assumed that the trajectory originates in the



middle of the city.  The morning mixing height assumed (250  m) is often



higher than the morning mixing height  used in Level II.   These conditions



tend to result in greater amounts of mass initially present  in the Level  III



air parcel. This can be illustrated by conversion of  the  initial conditions



for HC and NO  and the assumed mixing  height into equivalent emission terms
             A


                                    304

-------
(i.e. ppm-km units)  and comparison of the values with  the post 8 am emissions.



This has been done in Table 50.   The table shows that  two of the Level  III



simulations are clearly initial  condition dominated (D159 and D275).



     Since the Level  III approach assumes that the trajectory travels  from



mid-city straight to  the downwind station which measured the [Ck] max,  the



average hourly emissions of HC and NO  can be higher than average emissions
                                     A


in Level  II simulations (D195 in Table 50).   This would be especially  true



when the trajectory in the Level II simulations traversed the edge of  the



city as was the case  for D195.   Consequently, Level  III simulations can have



more mass input from  emissions into the parcel than Level II simulations.



     The third reason for differences between these Level III and Level II



simulations is related to the emission inventory assumptions of these  two



versions.  Level II  simulations  require a rather extensive emission inven-



tory.  Level  III does not.  It has been shown that the aldehyde fraction of



both the initial HC concentrations and consequent emissions is a key



parameter in the photochemistry  represented in this study.  The Level  II



inventory provided aldehyde fractions which  were used  in the Level  II



simulations.   The fraction ranged from -1-3% for the 10 days studied.   These



may be rather low values.   The fact is, however, that  the inventory was



conducted and these  values were  reported.   A Level III simulation does  not



require an extensive  inventory,  but rather uses county  wide average data.



It is not probable that an inventory of this type would include satisfactory



aldehyde data.   Therefore this value would need to be  approximated.



     Based on his extensive knowledge of the  emissions  inventory for St.  Louis,



Demerjian estimated the values given in Table 48.   He  used morning,  ambient,



detailed hydrocarbon  analysis to aid in estimating the initial  conditions.



                                    305

-------
to
CZ
o
•r—
4->
fO
r—
Z3
•r—
00
t—t I
Q.
i— Q-
0)
> "
CU X
_J O
-O
fO

i — i
i — i

r—
0)
c
^- 0
fO -r-

o to
i •rn>
j— i
UJ
O) O
C7)-r-
03 00
S- oo
CU -r-
> E

•i —
CT
1— H
c:
i— O
fO •!-
4-) OO
O OO
I— T-
E
UJ


^- ^J-
ro oo
CM CM
• •
o o



c
cu o
CD •' —
f<3 00
S- oo
CU T-
> E
cC ui

to LO
CM C\J
0 0

o o



2!
-)c
U
C
o
o
oo op r^
OO «T CM
r— t* tO
• • •
O O CD





OO LO
CO «*•
oo cr>
0 O
0 O

«3- LO
to o
o <—
O CD
. ,
o o


tO CM
O r-
o o

O 0






LO ^3-
CO «3-
i— LO

o o




•— o
oo to
0 O

o o



O LO ,—
CM tO CT)
r- O O

000





to cr»
LO O
co cr>
0 O
0 O


r— i — ->
t— t—
O 0

0 O


CM cr>
CM LO
o o

0 0






CO ,—
00 OO
CM LO

o o




to to
oo to
0 0

o o



tO LO LO
O r*-^ to
CM -vf tO
• • •
000


 O
 oo

 S-
 tV
 Q.

 E
 O
o
LO
 Ol
-Q
 (O
                   E
                   •r~
                   OO
                   rtJ
CM O O
r-. LO LO
CM CM OO
o
o o o
i— LO LO
r— C\J OO
 O O
O LO LO
r— CM 00
O1
LO
LO
CT>
LO
1^
CM
                                                     306

-------
The morning initial aldehyde fraction is an estimate made by Demerjian.
These values were used in the Level III simulations herein.


D159
     Discussion for June 7.  The observed [03] max was 0.192 ppm.  Table 49
gives the Level II and III simulation 03 maxima.  Both Carbon Bond II and
Demerjian mechanisms were used with a 250 m initial morning mixing height
and the Demerjian mechanism was also used with a 350 m mixing height.  All
the Level III simulations overpredicted (36-73%) while only Demerjian
overpredicted at Level II (~2%); and CB2 underpredicted in the Level  II
simulation by 20%.  The reasons for these differences in simulations are
readily explained.  The overriding factor for D159 was the difference in
the assumed initial location and initial conditions of HC and NO .  The
                                                                A
initial mixing height in both cases were almost identical:  250 m in Level III
and 272 m in Level II (characteristic curve). The initial HC, however, was
0.49 ppmC for the Level II simulation and 1.79 ppmC for the Level III
because of the differences in the assumed starting point of the trajectories
(see Figure 31).  The NO  initial  conditions were also higher by a factor
                        /\
of -5.5  The emissions in both simulations were almost identical  (because
the trajectory traveled very far).   Therefore the Level  III simulation
predicted more 03 than the Level II simulation because of more mass  initially
in the system.  The same effect can be seen between the  two DEM simulations.
The 350 m simulation made 0.332 ppm 03 as  opposed to 0.302 ppm at the 250 m
initial  mixing height.  These simulations  also show that for the  Level  III
approach DEM produces more 03 than  CB2 (e.g.  0.302 ppm vs  0.260 ppm  at 250 m
initial  height).

                                   307

-------
D195
     Discussion for July 13.   This day illustrates the range  of assumptions
possible in switching between Level  II and Level  III  approaches.  This range
makes it difficult to compare.  The  Level  II trajectory analysis  for this
day (Figure 41) shows that the parcel  originated  in the far south east of
the city, so much so that the earliest hour that  can  be assigned  a position
is 1100 LOT (with an initial  mixing  height - 1198 m).  The Level  III
assumption, however, originates the  parcel  in the center of St. Louis with
250 m at 0800 LOT.  The initial conditions (concentration and mixing height)
were such that in either case, the mass they contributed was  low.  Thus,
this day is dominated by emissions,  both for Level II and Level III approaches.
But since the Level III trajectory originates in  the  city, the emission's
contribution to the Level III simulation is much  greater than the emission's
contribution to the Level II simulation.  Hence,  for  the Level III simulations,
the Demerjian mechanism predicted 1.03 times the  observed ozone maxima;
the CB2 mechanism predicted 0.68 times the observed.   For the Level II
simulations, the Demerjian mechanism predicted 0.63 times the observed;
the CB2 mechanism predicted 0.46 times the observed ozone maxima.

D275
     Discussion for October 1.  Both effects, greater emissions and greater
initial mass, played a part in Day 275 DEM and CB2 simulations.  Both of
the simulations predicted greater [03] maxima in the  Level III than Level  II
simulations.   Initial concentrations were nearly the same at the two
different  levels.  The Level  II initial mixing height, however, was 109 m
(for Characteristic curve).  Therefore the  initial condition has -2.5 times
                                    3 OR

-------
the mass in the Level III simulation  as   in the Level II simulation.  Also

in the Level II simulation the trajectory was on the eastern border area of

the city.  Therefore the Level III emissions resulted in more than twice

the mass input into the parcel.

     The resulting difference between Level  II and Level III in [CL] max

prediction is most dramatic for this day.  The observed was 0.244 ppm.  In

the Level II simulation with characteristic  curve, the Demerjian simulation

yielded 0.168 and the CB2 simulation 0.081  ppm 0^.  The CU maxima for Level

III simulations were:  (DEM/250 m) 0.418 ppm; (DEM/350 m) 0.45 pom; and

(CB2/250 m) 0.250 ppm.

     A basic difference therefore between the Level II and Level III

approaches, is that the Level III assumptions often result in more total

mass being injected into the system than Level II assumptions, resulting

in more ozone predicted in Level  III than in Level II simulations.  The

increased aldehyde initial condition and emissions factor also has a major

impact on these simulations.



10.3  ISOPLETH DIAGRAMS AND EKMA CONTROL CALCULATIONS FOR
      LEVEL III SIMULATIONS

     Isopleth diagrams for the Level III simulations for use with the EKMA

procedure were obtained in the same manner  as for the Level  II simulations.

Figures 110 to 113 show the isopleth  diagrams and Oo as a function of

relative HC for three levels of NOV control  for the three days simulated
                                  /\

with the Level III approach.  Table 51 lists the %HC reduction estimates

obtained by EKMA procedure.  Finally, plots  of %03 reduction vs %HC

reduction estimates       for these Level III simulations are presented

as Figures 114 and 115.

                                    309

-------
                                                                cr>
                                                                LT>
                                                                 -P
                                                                r-  O
                                                                 3  -4->
                                                                -O  -C
                                                                    en
                                                                 S-  -r-
                                                                 o  
                                                                 E  C
                                                                 fO  -r-
                                                                 i-  X
                                                                 CD-r-
                                                                 to  E
                                                                •1^
                                                                TD  r-
                                                                    tO
OOt 0        00£ 0        002 0        001 0
                                                                -(-> C
                                                                 CJ -i-
                                                                 fO
                                                                 s- s_   .
                                                                f- CU  E

                                                                o   o
                                                                 o c\j    S-
                                                                •o  to  re
                                                                    r— C_5
                                                                J=  =5
                                                                -P  £ T3
                                                                 QJ T-  C
                                                                i —  (/)  (O
                                                                 ex
                                                                 o •-*  c
                                                                 C  CU  OJ
                                                                 O  >  E
                                                                 M  OJ  O)
                                                                 O _J Q
                                                                  a>
                                                                  rs
                                                                  en
      3>0

-------
oot o        oot o        ooz'o
                                                                 in
                                                                 CTl
                                                                      O)
                                                                 CO  o
                                                                 i — -Q
                                                                  s_ ^:
                                                                  O  en
                                                                 <+- •!-
                                                                      ai
                                                                  CO J^
                                                                  E
                                                                  (t5  CD
                                                                  S_  C
                                                                  CD-r-
                                                                  (O  X
                                                                 •r— *r—
                                                                 -a  E

                                                                  C t—
                                                                  O  IB
            one o        on: a        oot o
                            0 200

                            03.ffft
 O -i-
 i"O  ^—
 i- -r-    •


oEl
-c. o  c
 i  un  to
 o> CM  .£:
 c     o
 o  re  cu

 o  o>
     C  I—(
-a -r-  i->

 ta  =s  T3
     CO  d
 CO  CO  O
 E    S_

    r—  <_5
^:  3
-M  E  "O
 Q) -i—  c
I—•  CO  fO
 ex
 o >--<  c
 CO I—I  tO
                                                                  CD •—  $_
                                                                  C  0}  CD
                                                                  O  >  E
                                                                  N  O)  O)
                                                                 O —I O
                                                                  CU
                                                                  S-
                                                                  C7I
     311

-------
            ooc'O        oos'O        ooT'O
                                                                 IT)
                                                                 1^.
                                                                 CVJ
                                                                 IO
                                                                      OJ
                                                                  s-  o
                                                                  (Li ja
                                                                 jD
                                                                  O  S-
                                                                 4->  O
                                                                  o f-
                                                                 o
                                                                     4J
                                                                  S- -C
                                                                  O  Dl
                                                                 4- T-
                                                                      O)
                                                                  CO -C

                                                                  (t>  O>
                                                                  s-  c:
                                                                  cn-r-
                                                                  (O  X
                                                                 •r" *r--
                                                                 •O  E

                                                                  C r—
                                                                  o  to
                                                                 O  •!-
                                                                 fO  C
                                                                 S-  -r-
OOfc 0        QQC 0        OQZ 0        001'C
O     'r-
rr: o  c
  I  LT>  (O
 cu CM -c:
 c      o
 C3  fO  QJ
 tsl      £
 O  cn
     C H-I

 c  i I~H
 (B  3 "D
     10  C
   S-

    r— C_>
                                                                 CU  -I-  C
                                                                i—  oo  ro
                                                                 CL
                                                                 o  i—<  c
                                                                 to  I—l    E
                                                                 [si  Ol  CU
                                                                O —I Q
                                                                 cu
                                                                 S-
                 0.300        0 200        0 100
                             oa.rrn
  312

-------
                                                                        oot o      oos o      ooz a
                                                                                                                                        c:
                                                                                                                                       •r-
                                                                                                                                        CO
                                                                                                                                    O «3
                                                                                                                                   •r- I —
                                                                                                                                   +J CT1
                                                                                                                                    3  S_
                                                                                                                                    E  a;
                                                                                                                                   •r- -Q
                                                                                                                                    oo  O
                                                                                                                                       -t->
                                                                                                                                   i— i  o
                                                                                                                                   i— t O
                                                                                                                                   cu   ro
                                                                                                                                   >
                                                                                                                                   CU    A
                                                                                                                                   —I  CO
                                                                                            ocz o      ooi a
   0 300         0 200
OOt 0          OOE 0
                                                                        oat o       oot 0       OOE o       ooi 0
                                                                                                                                   i.  r--
                                                                                                                                   C31
                                                                                                                                   rtj  (1)
                                                                                                                                  •i-  C
                                                                                                                                  -o  rj
                                                                                                                                   O  S-
                                                                                                                                  •r-  O
                                                                                                                                   o
                                                                                                                                   ro
                                                                                                                                       en
                                                                                                                                    l
                                                                                                                                   CU  CD
                                                                                                                                   c  c:
                                                                                                                                   O -r-
                                                                                                                                   N  X
                                                                                                                                   O -r-
 nf  co

 CO  +J  £
 F  •"-  «/)
 fO  C •>-
 i-  T-  C
 O>     O3
 ca  i- .c:
•i-  tu  o
T3  +J  (V
     0)  E
-c  E
•M      C
 CU  O  ro
i —  IT) -r-
 Q-CO -r-j
 O      t.
 co  en CO


 CD  £= Q
 C  3
 O  <^  CL>
 M  to -d
o  to -t->
                                                                           9 100      0 300      a to I
                                                                       313

-------
10.3.1  Discussion of June 7, Dl59








     Figure 110 shows the isopleth diagrams and 03-HC fraction diagrams for



the Demerjian mechanism and the Carbon Bond II mechanism simulations



starting at 250 m initial mixing height.   Figure 113 shows  the Demerjian



mechanism simulation starting at 350 m initial mixing height.   It  should be



recalled that all three simulations overpredicted the observed ambient 03



maximum.  The EKMA solution points were lower than the measured initial



HC and NO  for all three isopleths:  the DEM simulations; were  about l/3>and
         /\


CB2 were about 1/2 of initial HC and NO.  D159 has the highest 0., aloft
                                       X                         O


values:  0.120 ppm and results with diagrams showing only a portion of the



0.08 ppm isopleth line.  The 0~ value at the origin was greater than  0.080



ppm.




     Although the DEM/250M and 350M isopleth diagrams do appear different,



the EKMA procedure when applied to each resulted in remarkably similar



HC control  estimates, 79.6 and 77.6% for a -20% change in NOV  (Table  52).
                                                            A


The CB2/250M isopleth diagram resulted in a much lower control estimate:



57.3% (-20% NOV).  For the Level II simulations the HC emission reduction
              X


for the corresponding  conditions were much less.  DEM Level  II was 67%



reduction and Level III was 79.6% reduction; CB2 Level II was  45%  reduction



and Level III was 57.3% reduction.








10.3.2  Discussion of July 13 ,  D195







     Figure  111 shows the Demerjian and the Carbon Bond II 250 m isopleth



diagrams prepared for this day.  This day can be compared with D275 in that



the absolute emissions rates were the same except for the last hour.




                                   314

-------
   Table 51.  Level III %HC Control Estimates
                             NO  Control
                               X
Day     Mech/Mix Ht     I2Q%_
159     DEM/250          79     71      64
        DEM/350          78     69     64
        CB2/250          57     47     41
        Dodge/250        74     72     67
275     DEM/250          47     38     32
        DEM/350          53     42     34
        CB2/250          38     29     23
        Dodge/250        56     48     38
195     DEM/250          45     35     29
        DEM/350          43     34     29
        CB2/250          32     24     19
        Dodge/250        59     55     46
                      315

-------
The initial HC values, however, were 7.5 times less than Day 275.   Also



the final mixing height for D195 (1800 m) is twice that of Day 275 (900 M).



The Oo aloft was similar for the two days.



     The isopleth diagrams for these days were plotted on axis with the same



magnitudes used for D159 and D275.  The isopleths suggest that this day



would produce ozone dramatically.  What must be remembered is that the



emissions as well as the initial conditions are multiplied in generating



the diagrams.  Therefore at the point on the D195 isopleth diagram which



matches the initial conditions of D275 (1.9 HC, 0.236 NO ) the emissions
                                                        /\


are 7.5 times as great as the design point  emission rate (observed initial



conditions and emissions).   This is a great amount of material compared to



the XI emission rate which is of course at  the design point of D275.



     Of further interest is that the control estimates for the DEM/250M



day 195 (44.6%) closely match those of DEM/250M day 275 (47.4%) while the



350 m control estimates differ:  DEM 195/305M was 42.6% and DEM 275/350 M



was 52.5% reduction.



     The isopleth diagram generated with the CB2 mechanism (250 m), which



appears to be similar to the DEM isopleth diagram for this day, yielded a



significantly lower control estimate (32.2% vs 44.6% for -20% NOV).
                                                                A






10.3.3  Discussion of October 1 , D 275







     Figure 11 shows the isopleth diagram and the O^-HC fraction diagrams



for the Demerjian and Carbon Bond II simulations using 250 m initial



mixing height.  This day contrasts with D159 in that the initial  conditions



on this day were similar to those of D159,  yet the emissions were much



larger on D275.  This was true in the absolute sense and the effect was




                                   316

-------
further enhanced by the 900 m final height on D275 compared to the 1900 m



final mixing height on D159.  These factors resulted in Day 275 having more



Oo producing ability for both DEM and CB2 mechanisms than on Day 159.



     Since the DEM/250M simulation predicted much more than the CB2/250M



simulation the isopleth diagrams also show similar trends.



         On   the DEM/250M isopleth diagram, the EKMA solution point was



found at 52% less_ than the day's initial condition and on the CB2/250M



isopleth diagram the EKMA solution point  was found at 8% less thanttie



day's initial conditions.  In spite of these differences in absolute pre-



dictions, the EKMA HC control estimates are 38.2% for CB2 and 47.4% for



DEM (at 250 m and -20% NO  reduction).  The DEM/350M isopleth diagram,
                         A


like the day's simulation,predicts the most ozone formation for this



day's conditions.  The EKMA control estimate  using this isopleth diagram



 is only 5% more (52.5%) than that using the DEM 250 M isopleth diagram.








10.3.4  Discussion of Oo Reduction Diagrams



     Figures 114 and 115 show the percent Oo reduction as a function of the



percent HC reduction for the Demerjian and Carbon Bond II mechanisms for



the three days simulated.  Except for June 7, Day 159, which had 0.12 ppm



03 aloft, the results suggest a greater than 1:1 control effect.  That is,



a given %HC reduction results in more than that % reduction in Oo (e.g. a



25% HC reduction gives a 50% 03 reduction).



     The ratio of % 03 reduction to %HC reduction is about 2:1 for Carbon



Bond II and about 1.3:1  for Demerjian (independent of initial mixing



height) in the Level III simulations with small  amounts of 0-, aloft.  For



Day 159 with 0.12 ppm 0^ aloft, the ratios are less than one (i.e. 0.8:1 for



Carbon Bond II and 0.5:1 for Demerjian).  Recall that in the Level II



                                   317

-------
-H   01   at
                                             - 5  5
                                                                                                            _ o  O
                                                                                                                            O
                                                                                                                            LO  CU
                                                                                                                            CM  >

                                                                                                                             05 CT>
                                                                                                                             c
                                                                                                                            •i-  O
                                                                                                                             in  4->
                                                                                                                             3
                                                                                                                                T3
                                                                                                                             CO  d)
                                                                                                                             c:  -a
                                                                                                                             O  O)
                                                                                                                            •,-  
                                                                                                                             CO  O

                                                                                                                            i—i  "O
                                                                                                                            >—(  OJ

                                                                                                                            "O
                                                                                                                             c  


                                                                                                                             O  -r-
                                                                                                                             (O  tO
                                                                                                                                +J
                                                                                                                             c  c:
                                                                                                                             ro  o
                                                                                                                            •r-  N
                                                                                                                            •1-5 .,-
                                                                                                                             s^  s-
                                                                                                                             cu  o
                                                                                                                                -o
                                                                                                                                O)
                                                                                                                             o  co
                                                                                                                             co     O.
                                                                                                                             E     Q-
                                                                                                                             t>   •

                                                                                                                             01-C  i—
                                                                                                                             03  CD   •
                                                                                                                            •1- T-  O
                                                                                                                            -£3  qj *	

                                                                                                                             c     -a
                                                                                                                             o  en s-
                                                                                                                            •i-  C  (O
                                                                                                                             qj
                                                                                                                                    (O
                                                                                                                                    +j
                                                                                                                                    CO
    
 O  ••—  03
 N  c  ^:
O  -r-  -t->
                                                                                                                             a>
                                                                                                                             i.
                                                                                                                             =3
                                                                                                                             O1
                                                                  318

-------
-Is  I
                                    4-
                                 CT> ra
                                 C. -O
                                -i-  c
                                 x  re
                                •i- +J
                                 E  to

                                i —   CO
                                 m o
                                •r—
                                -l->  OJ
                                    O)
                                E  >
                               oo  o
                                   +J
                                CT)
                                C "O
                               •i-  -J
                                S-  to
                                01  -i-
                               Q  c
                               S- i
                               o
                                   (D
                               re  o
                               S.  N
                               ro  S-
                              •r-  o
-t-> to
 o re
 3 Q '
•O
 (U
 S-

 01
c:
 o -
                                      Q.

                                     c\j
                                  O) O
                              O)
                              S-
                              01
  319

-------
simulations for Day 275, the ratio was about 2:1  for  Demerjian  characteristic
curve and that the Carbon Bond  mechanism gave no  solution  because of  under
prediction (the Dodge simulation ratio was  about  1.6:1).   For Level  II
simulations of Day 159,  the ratio was  about 0.7:1  for Demerjian and 1:1  for
Carbon Bond II.  Thus it appears that  the Carbon  Bond II mechanism con-
sistently suggests greater sensitivity of ozone to HC control than  either
Demerjian or Dodge and that this sensitivity is more  than  (up to twice)
that of simple linear roll  back.

10.3.5  Effect of Initial Aldehyde Fraction
     It was stated earlier that the Level  ill approach does  not require  more
than a county-wide emission inventory.  Aldehydes  are generally not well
characterized in these inventories and therefore  must be approximated by
the modeler.
     Based on suggestions by Demerjian,  values of 3%  Carbon  and 8% Carbon
were used for the ALD emissions  fractions and ALD initial conditions fractions
in this study.  A value  of 3% for the  aldehyde carbon fraction  marks  the
upper limit of the values determined from detailed emission  inventory
calculations for the 10  days studied in the Level  II  section.   It  is
a reasonable value.
     In the Level II discussion, the effects of variation  of the aldehyde
fraction on the simulations of certain days were  demonstrated.   To examine
the effect for the simpler Level  III approach, two level  III simulations
with the Demerjian mechanism were redone with an  aldehyde  fraction for  the
iniital conditions equal to the    emissions aldehyde fraction:  a change
from 8% to 3%.  The results are shown  in Table 52.
                                  320

-------
3% Initial Aid.,
ppm
0.359
0.208
Ratio
3% to 8%
0.859
0.963
       Table 52.  Effect of Initial Aldehyde Fraction on (k Maximum
             for Level III Simulations with Demerjian Mechanism

            [CU] max for            [Oo] max for
            8% Initial Aid.,
   Day           ppm	

   275           0.418

   195           0.216
     The magnitude of the effect of the variation on [03] maxima depends

on the relative contribution of the initial  conditions to the total  mass

of pollutants emitted into the parcel, hence a sizable effect for Day 275

and a small effect for Day 195,which was dominated by emissions.

     The aldehyde fraction effects the 03 maximum concentration mainly on the

left-hand side of the ridge onanisopleth diagram.  For those days in which

the EKMA solution points are located on this left-hand side,  the effects

of aldehyde initial  carbon fraction on the HC reduction are probably

minimal.    For those days in which the solution points are on the right-hand

side (which is not much effected by aldehyde fraction) the crossing  of the

ridgeline in a control calculation will probably result in a significant

effect on the HC reduction required.



10.3.6  Effect of Simulation Length on Control Requirements

     The  guidance supplied by EPA implies that Level III simulations (for

isopleth  generation) would  commonly be performed for a 10 hour day,

regardless of the time that the [Oo] max was measured.  For example, the

days simulated by the Level III approach in this study  ( D159, D195, D275 )

had maximum ozone concentrations measured during the 9th or 8th hour

(after 0800 LST). Emissions data were supplied only for the 8 or 9 hours it

                                   321

-------
took to reach the maximum measured (see Table 46).   The Dodge isopleths



generated by OAQPS for these days used 10 hours of simulation time and



8 or 9 hours of emissions data and used zero emissions for the extra one



or two hours until  the end of the 10-hour simulation period.   Isopleth



diagrams generated in this report for the Demerjian or Carbon Bond II



mechanisms used the actual time to [O.J max, 8 or  9 hours, depending on the
                                     O


RAPS day in question.  Concern for the possible effects from these



additional  assumptions (10 hour default run time and zero emissions  past



supplied data) suggested that simulations should be compared with  these



different conditions.  Five additional simulations and an additional isopleth



were obtained.  These additional  simulations are reported in Table 49 as the



second column under each day.



     For D159 and D195, simulations were performed for an additional hour



(nine to ten hours).   The one simulation for D159 using the CB2 mechanism



showed only a 1% increase in 03 maximum indicating that the 03 profile



had almost peaked at 1700.



     Two simulations were performed for D195 using both the DEM and CB2



mechanisms.  DEM showed a 5% increase and CB2 showed a 9% increase.



     Two simulations showing the most dramatic effects were performed for



D275.  The DEM mechanism shows that the 03 peak is closer to 1600, so that



the effect of simulating an additional two hours actually yields a lower



final Oo maximum.  The CB2 mechanism, however, having not reached  03



maximum at 1600, yields a 13% higher 03 maximum after simulation for an



additional two hours (0.250 to 0.282 ppm).



     For Level III,  the effect of simulation time is important  not for



absolute [03] predictions but for the HC reduction estimates obtained



                                   322

-------
from different isopleths generated for a day using different simulation times.
To obtain the greatest possible effect, one isopleth diagram was generated
with the conditions showing the largest change in 0^ maximum in the tests
just described:  C82 for D725, simulating 10 rather than 8-hours.
     The isopleth and 03 reduction diagrams for these two conditions are
shown in Figure 116.  The isopleths on the left-side of the ridge have been
shifted to the left (the initial  condition point occurs between 0.28 and
0.30 instead of between 0.24 and 0.28) for the longer simulation.  The
control diagrams, however, are essentially identical.  For no change in NOV,
                                                                          .A
the required HC reductions are 29.0 and 29.8% for 8 and 10 hours of simu-
lation.  The EKMA solution points, however, are on the left-hand side of the
ridgelines  and an effect might occur if the EKMA solution points were on
the right-hand side and the required HC reduction crossed the ridgeline.


10.4  SUMMARY OF LEVEL III SIMULATION RESULTS
     The Level III simulations predicted higher 03 maxima than the Level
II simulations.                          This is due to the assumptions in
the Level  III approach;  a trajectory originating in mid-city has higher
initial conditions, higher emissions and hence more mass in the air parcel.
Assumptions of initial  mixing height and the aldehyde fraction are also
parameters that effect the magnitude of the ozone maxima in the Level III
approach.
     Figure 117 is a graphic   summary of the control requirements predicted
by three mechanisms (Demerjian, Carbon Bond II, and Dodge) for three RAPS
days (159,  195,  and 275)  using a  Level III modeling approach as described
in the recent EPA draft guidance  (EPA, 1981).   Included in Figure 117 are
                                   323

-------
I 200            0.100

       NOX.PPM
                                                                   Nonano3» •o i
                                                                                                           rtS
                                                                                                           c:.
                                                                                                           o i
                                                                                                               >>
                                                                                                               10
                                                                                                              Q
 c: to

-C en
+-> •—

 2   •

 to


 O -O
-d  O

o  o
i— O

 O  4-
4->  O

oo
     E

 O •!-
 s-  c:
f-  to

 CD  O
 E  ^

+J
                                                                                                           o
                                                                                                          •r- T3
                                                                                                           +J  c:
                                                                                                           rc(  O
                                                                                                          r— DQ
                                                                                                           Z3
                                                                                                           E  C
                                                                                                          •r-  O
                                                                                                           c/1 -Q
                                                                                                               s_
                                                                                                           cn tc
                                                                                                           c: c_3

                                                                                                           
                                                                                                           s_
                                                                                                           O  i-
                                                                                                           c:  o
                                                                                                           1-  in
                                                                                                            O  c:
                                                                                                               O
                                                                                                           +J  >r-
                                                                                                            CJ  CO
                                                                                                            a>  co
                                                                                                            a>
                                                                                                            s_
                                                                                                            ZJ
0 200            0.100

        HOX.PPM
                                                 324

-------
the control requirements predicted in the Level  II  modeling study (see Chapter



9 for comparison).



     The Carbon Bond II mechanism always gives the  lowest control  require-



ment by a substantial margin (ratios of CB2 to DEM  are 0.66:1,  0.67:1, and



0.70:1, and ratios  of CB2 to Dodge are 0.65:1, 0.43:1, and 0.56:1  for days



159, 195, and 275).  This mechanism is the newest,  most complex,  and  most



validated mechanism of the four studied.  Except for Day 159 in which the



Demerjian and Dodge control  requirements were similar, Demerjian  always  gave



lower control requirements than Dodge at both Level  III and II.  For  the



days studied, Dodge always gives the highest control  requirement.



     Secondly, regardless of mechanism, the Level  III  control  requirement



predicitons are significantly higher than Level  II  predictions  (ratios of



Level III to Level  II requirements are 2.4:1, 1.5:1,  and 1.3:1  for Demerjian,



Dodge,  and CB2).



   Level  III, like  Level  II, gives the maximum control  requirement not for



the day with the  highest observed 0^ maximum, but  for  the day  with the



maximum O  aloft  and a lower contribution from local  emissions.
                                    325

-------
OCDCDCDCDCDCDCDOCDCD
OO")OOr^LD=fr-09OJ»—'
                                       CD

1
«M>^




	


— ••

	
	




	






1 I ' 1






















1 1 ' 1












• « J










' 1 ' 1



i ri







-j- Q









el


1 |











-










QJ X
1 1 1
__




	


— —

	
	




— —



	
In .4. 	








	
	
	


	
	
	
|












4-0-
11^
" O
, 1 ,
t











— f
».
•
, 1 ,
i n •
1



4-









e,i

, 1 ,
















1
	






1

i- 	






, i r
\^ ^j i
c c E
*° j~ O)
- u >


Q
CJ

0

LO
i —
C\J



Q
O

,—J

en
tn
•~


r^ (/)
« •!—
n! c s_ s-
03 OJ i 3 O
-Q 0 r-
11
-a en .«
o Ln x
-C r- 0

> c _c cn
5 0 U C
J] •!- _OJ (C
u ^ *"
•r- -a s_ r
-o c: (U ;
ai o a.1*
s- ca Q.
'o -Q >d
<_ r <^\ J
-M ro Z:
C (_)
one
ej t_> -r- J
r-.
IDCDCDOOCDOOCDOO ^
:D CD co r
- vO LT> =3- CO CM •— • =>
en
GH  %
326

-------
                                References
 1.  Uses, Limitations, and Technical Basis of Procedures for Quantifying
        Relationships Between Photochemical Oxidants and Precursors,
        EPA-450/2-77-021a, U. S. Environmental  Protection Agency,  Research
        Triangle Park, North Carolina, November 1977.

 2.  Arya, S.P.S.  (1979)  Atmospheric Boundary Layers,  to be published in
        Engineering Meteorology, Elsevier.

 3.  Benkley, C.W., L. L. Schulman  (1979)  "Estimating Hourly Mixing
        Depths from Historical Meteorological  Data,"  JAM, Vol. 18,
        pp. 772-78.

 4.  Demerjian, K. L., K. L. Schere  (1979)  Applications of a Photochemical
        Box Model for Ozone Air Quality in Houston, Texas,  In Proceedings:
        Ozone/oxidants Interactions with the Total  Environment II,
        Oct.  14-17, Houston, Texas, pp.  414-21.

 5.  Demerjian, K. L. and K. L. Schere  (1976)   personal communication.

 6.  Demerjian, K. L.  (1980)  personal  communications.

 7.  Demerjian, K. L., K. L. Schere and J. Peterson  (1980)  "Theoretical
        Estimates of Actinic (Spherically Integrated)  Flux and Photolytic
        Rate  Constants of Atmospheric Species in the Lower Troposphere,"
        Advances in Environmental  Science and  Technology, Vol. 10,
        John  Wiley and Sons, New York, New York.

 8.  Demerjian, K. L.  (1981)  personal  communication.

 9.  Dickerson, M. H.  (1978)  MASCON - A Mass  Consistent Atmospheric Flux
        Model for Regions with Complex Terrain, JAM, Vol. 17, pp.  241-53.

10.  Dodge, M.  C.  (1977)  Combined Use of Modeling Techniques and  Smog
        Chamber Data to Derive Ozone-Precursor  Relationships
        EPA-6007^77^001 b.

11.  Dodge, M.  C.  Effect of Selected Parameters  on Predictions of a
        Photochemical Model  EPA-600/3-77-048,  U.  S. Environmental
        Protection Agency, Research Triangle Park,  North Carolina,, June 1977,

12.  Evans, R.  B.  (1979)  The Contribution of  Ozone Aloft to Surface
        Ozone Maxima, Ph.D. Thesis, University  of North Carolina,  Chapel
        Hill, NC.

13.  Federal  Register, "Data Collection for 1982  Ozone  Implementation Plan
        Submittals" Nov.  14, 1979, 44, (221) 65669-65670.
                                   327

-------
14.   Gipson,  G.  L.   (1980)   personal  communication:   OAQPS  study.

15.   Gipson,  G.  L.,  W.  Freas,  R.  Kelly,  E.  Meyer  (1981)  Guideline for
        Use of City-Specific EKMA in  Preparing Ozone SIPs,
        EPA-450/4-80-027,  U. S.  Environmental  Protection  Agency,  Research
        Triangle Park,  N.C.,  March 1981.

16.   Goodin,  McRae,  Seinfeld  (1980)   An Objective Analysis for
        Construction Three Dimensional  Urban  Scale Wind  Fields,
        JAM,  Vol.  19, pp.  98-108.

17.   Holzworth,  G.  C.  (1972)   Mixing Heights, Wind  Speeds, and Potential
        for Urban  Air Pollution throughout  the Contiguous United  States,
        AP-101,  USEPA Research Triangle  Park,  NC.

18.   Jeffries, H.  E. and K.  G. Sexton  (1981)   Modeling Aspects of Nitrogen
        oxides Using Smog  chamber data,  in  workshop  proceedings on
        Formation  and Fate of Atmospheric Nitrates,  EPA-600/9-81-025,
        U.  S. Environmental  Protection Agency, Research Triangle  Park,  NC.

19.  Jeffries, H. E., J. E.  Sickles, M.  Saeger, and M.  A.  Carpenter
        "Experimental Determination of the  Specific  Photolysis  Rate of
        Nitrogen Dioxide,"  Draft Final  Report, Bruce Gay, Jr.,  Project
        Officer, Environmental Protection Agency,  Research  Triangle
        Park, NC,  Feb.  1981.

20.   Littman, F. E.  (1979)   "Regional Air Pollution  Study:   Emission
        Inventory  Summarization",  EPA-600/4-79-004,  Rockwell  International
        Creve Coeur, Mo.

21.   Liu,  M.  K,, J.  H.  Seinfeld  (1975)   On the Validity  of Grod  and
        Trajectory Models  of Urban Air Pollution,  Atm.  Env.,  Vol.  9,
        pp. 555-74.

22.   Schere,  K.  L. and K. L.  Demerjian  (1980)   personal  communication:
        characteristic curve.

23.   Shreffler,  J.  H.  (1978)   Detection of Centripetal  Heat Island
        Circulation from Tower Data in St.  Louis,  BLM, Vol. 15,
        pp. 229-42.

24.   Spath, H.  Translated by Hoskins,  W.  D.,  H. W.  Sager  (1974)
        Spline Algorithms  for Curves  and Surfaces, UTILTAS  Mathematica
        Publishing Incorporates, Winnipge.

25.   Whitten, G. Z.  and H. Hugo  User's  Manual for Kinetics Model  and
        Ozone Isopleth Plotting Package, EPA-600/8-78-014a, U.  S.
        Environmental Protection Agency, Research Triangle  Park,  NC,
        July 1978.
                                     328

-------
26.  Whitten, G.  A. and H.  Hugo  U. S.  Environmental Protection Agency,
        EPA-600/8-78-014b,  Research Triangle Park, NC, July 1, 1978.

27.  Whitten, G.  Z., J. P.  Killus, and H.  Hugo  Modeling of Simulated
        Photochemical Smog with Kinetic Mechanisms, EPA-600/3-80-028a,
        U. S. Environmental Protection Agency, Research Triangle Park,
        NC, February 1980.

28.  Yamad, T.   (1976)   On  the Similarity Functions A, B,  and C of the
        Planetary Boundary Layer,  JAS,  Vol.  33, pp. 781-93.
                                   329

-------
                      APPENDIX  A
THE HISFIT ALGORITHM:   AN AREA TRUE  HISTOGRAM FUNCTION
                         330

-------
                                Appendix A

The HISFIT Algorithm:  An Area-True Histogram Function

     It is frequently desirable to represent a series of interval histogram

values as a smooth function having the area-true property.  Consider a

series of n intervals and associated histogram values h-j, \\2  •• h .  We

define CQ, e-|  .. e  to be the abscissa values (x) at the edges of the intervals

Thus e-j .. e   I are common edges, eg and e-| are the edges of  the first

interval, e-j and 62 are the edges of the second interval and  so on.  For any

abscissa value x we assume the function produces an ordinate  value y.  Then

the function is defined to be smooth (differentiate):
     lim fP(e.   ) = lim fP(e.   ) = y[ (p = 0,1 and k = l..n-l)
     e-K)     K c    e->0     K"£-     k
The area true property may be expressed as:


      /k f (x)dx - hk(ek - ek  )      (k = 1 .. n)
     ek-l
     An algorithm for finding the coefficients of a fourth degree spline

with these properties is presented in Spath (1974).In the present situation,

negative function values are physically meaningless and so we must ensure that

the function nas no negative values.   Cubic splines (as originally used in

OZIPP) have fewer degrees of freedom and when a cubic spline is constrained to

have no negative values, very poor fits often result.   Oscillating histogram

values also lead to poor fits with excessive over- and under- shoot.

Comparative examples are presented in Figures Al  and A 2 below.

                                   331

-------
     The STAIR algorithm (Spath,  1974)  solves the equation system over a



domain given the ordinate values  at the edges and boundary values for the



first derivative.   Thus a solution domain consists of a contiguous series of



intervals a, a+l,...,b (l^a^b^n), the given ordinate values  at the edges



y  ,, y , .., y,  and the two boundary values dy  -, and dy, .
 u~i    a       o                               a~ i       D


     It was discovered that some  histograms resulted in spline  functions



having negative values when the solution domain is the entire histogram.



In particular when either of the  ratios:
           or
     hk+l
is too large or too small, negative function values result.   In either case,



we control the maximum (minimum) value of the first derivative at the



various edges.  Examining the histogram, we apply a  heuristic to find those



edges which are likely to have an excessively large (or small) first



derivative, and we break the solution domain at those points.   Each such



edge then becomes a common boundary for two solution domains,  and we can



define the edge first derivative as a boundary condition such  that the



resulting function has no negative values.  Any h.= 0 must always be a



boundary of at least one solution domain.



     MISFIT finds the minimum set of solution domains and defines the ordinate



and boundary values as follows:
                                    332

-------
  §  0.
I . (-
1.0
0. 8
0.6
0.1
0.2
0.2
0.1
n ft
i
i
i
i
i
_ DRY 160
—
—

—

, 1


	 f


, 1


^-


i

/--"
X


i

-\
V


i
i
1 '
i

i
JUNE 8. 1376 m _2


^_^



,

/
/
^_^/

/







\ —
I "•


i
i

\J
—
1
1 . ^
1.0
0.8
0.6
0.1
0.2
. 0
-0.2
-0.1
_n F,
                                1     5

                              INTERVflLS
  o


  X
  to
  CO
  z:
  UJ
1 . d
1.0
0.8
0.6
0.1
0.2
0.0
0.2
0.1
n P,
I



i
i


i
I
DRY
P—^

i
i
160

y

i
i
JUNE 8



i
' ' i
. 1976
v ^


i
I
NOX
/

i i
H/sVlT J

' \
—
^:
\-
—
i
, I ,
1 . C
1.0
0.8
0.6
0.2
0.0
-0.2
-0.1
-n f,
                                15

                              INTERVflLS
Figure A.I.   Comparison  of equal area histogram fit as produced
             by OZIPP algorithm giving negative area (top) and by

             HISFIT algorithm (bottom).
                              333

-------
t
o
t—I

X
 co
 CO
 LU


 O
               I  I   I  I  1   I  I   I  I  I   I  I  I   I  I   I  I  I   I
                                 Y 221  flUGUST 8,  1976  HC
          t   _   i .         i       i  i       i          i     i
        0123456789    10   11
 CO
 CO
 UJ
o . v
5.0
n n
H. U
3.0
2.0
1.0
On
. U
-1.0
-2.0
T n
1 1 '
i

i


i

1
I
1 A DflY 221 flUGUST 8, 1976 HC HISFIT _I
—
_


1
\s*—-J\
^__




^ — ^




^ /-\




\
\




\ j
^-^

/ \
\


\_ J
"• — "
—
—
r , i ,


,
L i _j
i

1
,
—
, 	
A-
V
—
—
!
U . U
5.0
3.0
2.0
1.0
n n
-1.0
-2.0
-3 n
                                5    6    7

                              INTERVRLS
10   11
Figure A.2.  Comparison of equal area histogram  fit as  produced

             by OZIPP algorithm showing strong oscillation  (top)

             and by HISFIT algorithm  (bottom).
                             334

-------
     hk =   0
       or
      k+l
                                     dy  =0
     5.01 > nk+] > !____
            Fk""   5.01
                                         h,    0.5
                                                            \>hk+i
     'k+1
          >  5.01
         or
          >  5.01
     'k+1
2.5 hk.
                                          5.25 h
      k
                                         -5.25 h,
hk+l > hk
It Is assumed that h  ,  = 0.  There are two alternative assumptions about

the value of h«.   In one case (mode = 0) we assume  h« = 0.  Otherwise

(mode = 1), we assume h0>h,  and we set yQ = 2 h-, and dy^ = -yQ/(e,-6Q).
                                    33 R

-------
       APPENDIX B
SPECIAL EKMA COMPARISONS
         336

-------
                                  APPENDIX B


 Special EKMA Comparisons.

      EKMA is a technique which is used to obtain HC control  estimates from

 03 isopleth diagrams which do not necessarily agree in the absolute sense to

 the observed atmospheric  conditions to which the isopleth diagram is being

 applied.  EKMA is a scaling procedure; it is not needed for  isopleth diagrams

 which agree with the observed precursors and 63 values.  In  simple models or

 models with gross assumptions, the absolute agreement is generally poor and

 EKMA provides a means of estimating control  requirements for  these cases.

      This study has shown that assumptions made by modelers can have an effect

 on the calculated control requirements.  The validity of EKMA and the effect

 of incomplete and assumed data has been considered by the EPA (1980):

      "...Sensitivity studies have shown that differences resulting
       from incomplete input data or gross assumptions employed by
       the model  tend to exhibit proportional impacts on the  positions
       of various ozone isopleths on an isopleth diagram.  Thus,  the
       model  should perform well when applied in a relative sense..."

      Some results from this study tend to disagree with the  above conclusion.

 Figure B.I  shows the results of two applications of the Level  II  approach

 to D275 October 1, 1976 (RAPS) with the Demerjian mechanism   and  the char-

 acteristic curve mixing height.  The carbon mole fraction for  aldehydes

 needs to be determined for both the emissions into the air parcel along the

 trajectory and the initial  conditions of NMHC.  Each row of  Figure B.I  shows

 the simulation and observed profiles for NO, N02> and 03 for the  day, the iso-

oleth  generated,  and the final  %Q^ reduction-%HC reduction plot obtained with

 the EKMA procedure (the horizontal  dashed line shows the % Oo  reduction

 needed to obtain the 0.12 ppm level).   Row A used the emission inventory

 value  of 2.02% carbon mole fraction for aldehydes for both  emissions and

                                    337

-------
      0,.  ppm
iO   «-  OJ
                                                         1      !
                                                                          0,   pom
                                                                                                             E
                                                                                                             o
                                                                                                            o  o  ••-
                                                                                                            rr: -i-  oj
                                                                                                            s: r-  -c
                                                                                                            z:  Q.
                                                                                                                 O_ CD
                                                                                                            i—  (O  C
                                                                                                             (O     T-
                                                                                                            •1- >-H  X
                                                                                                            4-> •—I  •!-
                                                                                                            •'-     E

                                                                                                            •r-    O  +•>
                                                                                                             O 4—  oo

                                                                                                             S_  C  i-
                                                                                                            4-  O  CU
                                                                                                                •i-  +->
                                                                                                             O) •*->  O

                                                                                                            'o 3  £
                                                                                                             E  O    fD
                                                                                                             u  c:
                                                                                                                 O  E
                                                                                                               C
J^:      ro
-!->  C -r-
     o ••->
 en      S-
 C r	•  CU
••- CO  B
 cn»—  0)
 C     {^

-C C3
 O   • ^O
    ur> r~^
4-     CTl
 O  O r—
    +J
 1 <       ^

 O '—-i—
 QJ eX
4	i.
4-      CU
 CD ^S -Q
    CM  O
 O) O 4->
-C   •  O
I— CO O
                                                                                                            CQ
                                                                                                             Q)
                                                                                                             s_
                                                                                                             0>
                                                                             CO
                                                    338

-------
o
C\J
c
o
O LD
r— 1 	
1 
"~ "0
« O)
"~~ 0
S-
O)
-Q O
O 3T

O
0
•r—
cn -^t- -a
cn ID r^ a)
00 cn r- j_
Q-
i — CO CM
c:
O
s- •<-
O -I-1
M- • fO
• — -• » —
co E
+- ' t/)
C •!••• C^
•i- c: i —
O fO S-
o r~
O
Z3
CM E
1 LT) -^J- ••-
i  o
<: E
2: CD
X^ Q
D- 3
^r co «*• . i —
^ UD CO
CM i — CM
• • •
CD CD O

1 i 1 «. 	

, —
•
CO

O)
, —
-Q
(O C
\— 0
•1 —
•J-)
•1 —
-a
cr
o
o

CO
a> c: c
3 O O

n3 4-> 4->
> fD fO
! 	 , 	
4-> rs r:
£= E E
CD -i- T-
•c— CO to
_Q
E ^5; ^V
rd CM IT)




(O
4-)
to c
•4-J CD
C -Q
^3 E
(O
CO
0 c,-
O
T3
c o
rt3 •' —
^J
^ rt3
X S-
O
2: cu
<~
" 4-*
C_)
3: co
•i —
^H , 	
fO &—
ns XI




CO
cu
(O
>

4-)
CZ
CU
•r~
E
n3
0
+-)

CO
c
o
*r-
4->
•i —
"O
C
0
o

, —
to
•i —
4->
•i —
C

T3
O)
4->
fO
r —
23
(_j
i —
fO
O

2:
UJ

(4-
o

0

4-}
ro
S-

ai
f~
+j

CO
••-
CM
S-
o
339

-------
initial NMHC.  Row B used the emission inventory value for the emissions,



but used a value of 5.0% for the initial  conditions of NMHC.   It can be



seen that not only the simulation profiles and isopleth diagrams changed



considerably, but that the EKMA calculated control  requirements changed



considerably as well .   The change in scale on the isoplel.i diagram was



necessary to obtain an EKMA solution for the upper  set of conditions.  The



numeral values are shown in Table B.I. The solution point moved from 15%



more initial and emitted HC and NO  to 110% more initial  and  emitted HC and
                                  X


NO , and the control requirement went from 15% to 22%, a ratio of 1 .47.



     Figure B.2 shows  again two approaches to estimating  control requirements



to D275 October 1, 1976.  Row A is a Lavel  II application and Row B  is a



Level  III application.  The different assumptions for each Level has been



described in this report.  Approximately  the same  mechanism  (CB2),  emissions,



initial conditions, 0^ aloft, and final mixing height were used on both



applications for this  day.  The basic difference lies in  the  initial mixing



height assumed (Level  II 109 m, Level III 250 m) and the  NMHC carbon mole



fractions assumed for  emissions and  initial conditions (see report for dis-



cussion).  Not only are  the simulation profiles and isopleth diagrams



drastically different, but while an  EKMA solution was obtained from the



Level  III approach, no solution at all was possible with  the  Level  II



approach.
                                   340

-------
  o
  LO
                                                                                                              (J
                                                                                                          
                                                                                                         -c:  to
                                                                                                          O  (O

                                                                                                         4->  O
                                                                                                          (O
                                                                                                          o -o
                                                                                                         ••-  E
                                                                                                         r—   CO
                                                                                                          O)
                                                                                                             us
                                                                                                            ^l--»
                                                                                                             CTl
                                                                                                         QJ
                                                                                                         >  i-
                                                                                                         O)  O)
                                                                                                         _J  -Q
                                                                                                             o
                                                                                                         (O  -4->
                                                                                                             O
                                                                                                         E  O
                                                                                                         O
                                                                                                         i.  5-
                                                                                                         M-  O
                                                                                                             <4-
                                                                                                         CT1
                                                                                                         C  C   •
                                                                                                         •r-  o *~~*>
                                                                                                         CD-I-  +->
                                                                                                         C  +->  JC
                                                                                                         ra  3  OI
                                                                                                         JT  i —  -r-
                                                                                                         O  O  O)
                                                                                                             in  -C
                                                                                                             o  c:
                                                                                                         +j  i, -i-
                                                                                                         o +->  x
                                                                                                         at  s= •<-
                                                                                                         M-  o  s
                                                                                                         t-  o
                                                                                                         OJ      O)
                                                                                                            •a:  >
                                                                                                         
-------
                       APPENDIX  C








 SIMULATIONS OF INDIVIDUAL  BUREAU  OF  MINES  EXPERIMENTS



USING CARBON BOND,  DEMERJIAN,  DODGE,  AND  CIT  MECHANISMS
                          342

-------
                                                             NO. NO,.  PflN,  ppm

                                                        l/y   o  tj~)   o  10  °  LO
                                                        ^-   «*-  co   co  ex  CM  •-*
                    'ON  'ON
                                                                   i, ppm
   pr,
  r-  10  U7  «/-   o^  CJ   --*  T>
  c^  r>  (—.  o   O  cj   <~i  c^ri

-rj-. rr-rn-n-'-T <  '
_!.Li J
                   O  O  CD  O  O


                   ':ON 'ON
                                                            'NOd ':OfJ  'ON
                                                  343

-------
             0-.  ppm
                                                                   0,.  ppm
L1__U_l_i_l_uLj_L1_LL_UJ_i_L
               'NUd  ''ON 'ON
uldd 'NUd ''CH
                                                                               'OM
                                                                   0,.  ppn
                                                                O  CJ  O  O  O  O  CJ  CIO
               'Nttd <:ON  'ON
                                                                               'ON
                                                   344

-------
 L
 u                       -        y  .
 U.: -i :_4 L4-!_i-L.4J_i-U i i.i_i._
                                                                               ''ON  'ON
               3. ppm
                                                                        ),  ppm
                   in  et-  Of.or3i/>c3i/>c»lOO
                                                                CD  O   O
C3OOOC>C»C;>OC3O

            """ 'NUd  ''ON 'ON
                                                                         'Nb'.)  ''ON 'ON
                                                       345

-------
.I ..J_. Li\>L,l_,L-
                                                    i, ppm
  ujdd  'f
                                       346

-------
                                                                ''ON  'ON
0,,  |,| m
   'NUd  ''CM'ON
                                         347

-------
   '.NUd  '!OM 'ON
                                                             'NBJ ''ON 'DM
 0,. ppm
lucid  -f
                                           348

-------
  0«.  PFitl
                                                                           ppm
  Oj,  ppi
UJdd  'N'l'd  '

-------
                                                             Oi. ppm
'Jdd  'NUd ''ON  'ON
                                                                'fjyj -ION 'ON
                                                              Oi.  ppm
 >^dd  'NUd ''ON'ON

-------
   0,. PPM
                                                              pptn
Wdd 'NUd  ''ON 'ON
                                                          0,.  PPM
             CO  CO
          'WON
                                                        WJd 'NUd '!ON 'ON
                                          351

-------
                                                      Oi.  ppm
                                                   oooo
  'NBd ''ON  'CM
                                                         'NUd  ''ON 'ON
Oi.  ppm
                                                         'NUJ ''OM  'ON
                                       352

-------
1_4_L_1.J 4_J	L.J	1—J	|_J_4—1--1—.1	I
                                                                       0,.  ppm
             'N:Ud ''ON 'ON
                                                                     ujdd -Nud  ''OH  'ON
                                                                       Os. ppm
                   !CN 'CN
                                                     353

-------
  ~rn~rrrr
                                                             Q», p pm
LoJ_l-l-I_i_lJ_lJ_l-l J.l-l-i.4-1
                                                                       —•  *—  o
                                                                       -Ofj
          u-"'d 'NUd ''ON 'ON
                                              354

-------
  Oj,  ppm
                                                            Oi.  ppm
          '!DM  'ON
UJdd  'l!Ud  '"ON  'ON
      ppm
""I"  'NBd  ''ON  'ON
                                                             'NUJ  ''ON 'ON
                                          355

-------
_LLj_J  ,  I  •  LJ  :  •.  U__.x.:
                                                                          Oi.  ppm
                                                          O  IO  O
                                                          QOOcnOOOClOO  C3O
                                                           OC3OOOOOC3CDCDO
                                                                                 -1OM 'OM
                                                                       "^d  'NUd ''ON 'ON
                                                        OCj

-------
                                                                 >.  ppm
                                                                        e>  to  o  1/7  o
                                                                                O  O  00
                                                                                   .  I  .'
                                                                            uO  «*-  OJ  C3
                                                                            O  CD  O  O
                                                                    CD  Q  CT  O  O
                                                                  'NUd  '"ON  'ON
                    1 ^   ,


                    t +   U
j        '.        J  ;  j:
                                                OOC3C3C?C_3C3O
                                                                  'NUd  ''ON  'ON
                                              357

-------
                                                                         0,, pptn

C*   CJ   Ci
                      -'o,M 'OM
                                                                            'NUd ''ON  'ON
             uidd 'f,y
                                                                                  ''ON  'ON
                                                          358

-------
                                                        0..  ppm
                                                          'Ndd  ''ON 'OH
0,.  p[,rn
                                                        Oj. ppm
       ''CN  'ON
                                                                    'ON
                                        359

-------
                                                                ppm
                                                        CO  CO   CM   OJ
  'NUJ
            'ON
                                                            03. ppm
CT,   LQ   
-------
                                                            Oi,  ppm
   'NUJ  ''CM  'OH
                                                         t"
-------
                                                             ppm
O   CJ  O  O



  'NUd  ''ON  'ON
                                                       ujdd
                                                                ''ON 'ON
0>,  ppm
       ''ON 'OM
                                                            'NUJ  '!OM  'ON
                                          362

-------
  Oi.  ppn
                                                             Oi,  ppm
                                                         CO  CO  OJ   CNJ   *-t  f-*
                                                          uidd, '
                                                                    "ON 'ON
  0>,  ppm
  O  C3   O  C_  O  Q   i^> CD
-r-prpr rrj ,- p"t i
        cS>
  OOOOC3OO
 d  'NbM "ON  'ON
                                                                    -'ON  'ON
                                           363

-------
              0,.  pf.ii
                                                                         ppm
,-T-r
                                                            I  .  I  . I  .  I  .  I .  I  .  I  ,  I
                NHd  ''ON  'ON
                                                                        'NUd ''ON  'ON
                 ppm
            «ii  'NUd  '-DM 'ON
'NUd ''ON  'ON
                                                     364

-------
                                                              Oi.  ppm
CO   O  C J  OJ
               l/j   0>  LO
                                                                     1JON 'ON
   0>.  ppm
                                                              Oi. ppm
           '»CfJ 'ON
                                                                'NUd  ''ON  'ON
                                            365

-------
                                                 0%. ppm
 ''ON 'ON
                                                   'NUd ''ON  'ON
                                                    ppm
''ON'  'ON
'NUd ''ON  'ON
                                 36fi

-------
                                                            0..  ppm
                                                         CO  CO  CNJ  OJ
uiad  'Ndd  ''ON  'CM
                                                              'NHd  ''ON 'ON
    'NtlJ  ''ON 'ON
                                                                   "ON  'ON
                                           3P7

-------
0).  ppm
                                                              ».  ppm
CV  O  O3
                                               _  o
                                                        '  I  '  1 '
  'Ndd  ''ON  'ON
''ON  -CN
                                                             Os.
                                              CO  C\J  CM  CM
        '!C.M  'CN
                                                               'NUd  ''ON  'ON

-------
s.  ppm
                                                        Ot,  ppm
 'NUJ  ''ON 'UN
                                                     ujdd
''ON  'ON
                                                        Os.  ppm
                                                               CD  C^>   O  O  O
                                                          'NUd ''ON  'DM
                                       3*9

-------
              0,.  PPM
                                                                              CM  —•   —•
o   o  o  o  o
  i hh I oil 1.1.1 ill h I i h  iLJ
                                                         •*- CO OJ •-• O  C7) CO |— vO \J~)  *f CO
-JOQOOOC3OOOO
 Wdd  'Ndd  "ON 'ON
                                                                         'Ndd ''ON  'ON
                                                                       0,.  PPM
            Wdd  'NUd 'BON  'ON
                                                          LoJ-d-l LLJ J.ljJ_Ll-LLLLd_l_L J-lJ-L.1
                                                                     Wdd  'KUd  '!ON  'ON
                                                      370

-------
      0,. PPM
                                                     CO  ti  'ON
                                              371

-------
  0,.  PPM
 0,,  ppm
Wdd  'NUd  '"ON 'ON
                                                                  ''ON  'ON
   0,. PPM
  0,.  PPM
                                              -..J-L.L.L.I  I I |J_I  I  I
VJdd  'NUd MON  'ON
Wdd  'NUd  ''ON  'ON
                                            372

-------
                                                     Ot, ppm
   a  '!ON 'ON
'NUd '"ON  'ON
                                                        ppm
'Ndd ''ON  'ON
                                                            -SON 'ON
                                    373

-------
                                   TECHNICAL REPORT DATA
                           (Please read Instructions on the "'verse before completing!
1. REPORT NO.
  EPA-450/4-81-034
                             2.
                                                           3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
  Effects of Chemistry and  Meteorology on Ozone Control
  Calculations Using Simple Trajectory Models and the
  EKMA Procedure
                                                          5. REPOR r DATE
                                                             November  1981
                                                          6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
                                                           8 PERFORMING ORGANIZATION REPORT NO.
     E. Jeffries, K. G.  Sexton,  Cc  N.  Salmi
9. PERFORMING ORGANIZATION NAME AND ADDRESS
  University of North Carolina
  Department cf Environmental Sciences and Engineering
  School of Public Health
  Chapel Hill, North Carolina   27514
12 SPONSORING AGENCY NAMF AND ADDRESS

  U.S. Environmental Protection  Agency
  Office of Air Quality  Planning and Standards
  Research Triangle Park, North  Carolina  27711
                                                           10. PROGRAM ELEMENT NO
                                                           11 CONTRACT/GRAN"1" NO.
                                                              68-02-3523
                                                           13. TYPi. OF BFPOHT AND PERIOD COVERED
                                                          !14 faPOMSORlNG AGENCV CODE
15 SUPPLLMT-NT ARY NO") t£
  EPA Project Officer:   Edwin L. Meyer
16 ASS r
       Three chemical kinetics  mechanisms (Carbon Bond II,  lumped  species mechanism used
  in Demerjian and Schere's  photochemical box model and the Cal  Tech  mechanism developed
  by McFae) were used in a modified  version of the OZIPP model  to  replicate smog chamber
  data in which automotive exhaust was  irradiated (Bureau of Mines data).  Two of the
  mechanisms (CBII and Demerjian) agreed with the data, as well  as the existing mechanism]
  in OZIPP (Dodge propylene/butane).   The OZIPP model was next  used to simulate several
  days of observations made  during the  St.  Louis RAPS.  In several  cases, it was found
  that the meteorological input and/or assumptions did not allow  an  adequate basis to
  compare the mechanisms.  Three days  in which the meteorological  input appeard appro-
  priate were simulated using OZIPP  with the Dodge, Demerjian and  CBII mechanisms.
  Differing peak ozone and control estimates were obtained, depending  on the day simulated
  and choice of mechanism.   Sensitivity tests were employed to  explore observed dif-
  ferences among the mechanisms  tested.
17.

a
                  DESCRIPTORS
                               KEY WORDS AND DOCUMENT ANALYSIS

                                              I-1 IDENTIFIERS/OPEN ENDED TtHMS
C. COSATI 1 Icld/Utoup
  Ozone
  Photochemical models
  Chemical  kinetics mechanisms
  OZIPP
  EKMA
  Sensitivity studies
18 DISTRIBUTION STATEMENT
                                              19 SECURITY CLASS (This h'epori.
  Unlimited
                                                                         21 NO. OF PAGES
                                                                              393
                                              20 SECURITY CLASS (This page)
                                                                         22. PRICE
EPA Form 2220-1 (Rev. 4-77)   PREVIOUS EDIT'
                                          F-   nvironmental rroiecnon   gency.
                                           Region V, Library               *    '
                                           230  South Dearborn  Street
                                           Chicago,  Illinois  60604

-------