EVALUATION OF THE RADM GAS-PHASE CHEMICAL MECHANISM



                                                by

                                       William P.  L.  Carter

                             Statewide  Air  Pollution Research Center
                                     University  of California
                                    Riverside, California 92521


                                               and

                                        Fredrick W. Lurmann

                                        Lurmann  Associates
                                      1819 State St., Suite D
                                 Santa  Barbara, California 93103



                             Cooperative Agreement No.  CR-81U558-01-0
                                          Project Officer

                                          Marcia C. Dodge

                         Chemical Processes and Characterization Division
                      Atmospheric  Research and  Exposure  Assessment Laboratory
       ARCHIVE            Research Triangle Park,  North Carolina 27711
       EPA
       600-
       3-
                      ATMOSPHERIC  RESEARCH AND  EXPOSURE  ASSESSMENT LABORATORY
                                OFFICE OF RESEARCH AND DEVELOPMENT
                               U.  S.  ENVIRONMENTAL  PROTECTION  AGENCY
                                 RESEARCH  TRIANGLE  PARK, NC  27711
iPA-89:radmreport

-------
                       EVALUATION OF THE RADM GAS-PHASE CHEMICAL MECHANISM



                                               by

                                      William  P. L. Carter

                             Statewide Air Pollution Research Center
                                     University of California
                                   Riverside, California 92521


                                               and

                                       Fredrick W. Lurmann

                                        Lurmann Associates
                                     1819 State St., Suite D
                                 Santa Barbara, California 93103
        IP
        L,                   Cooperative Agreement No.  CR-814558-01-0
                                         Project Officer

                                         Marcia C. Dodge

                         Chemical Processes and Characterization Divi-sion
                     Atmospheric  Research  and  Exposure Assessment Laboratory
                           Research Triangle Park,  North Carolina 27711
                     ATMOSPHERIC  RESEARCH  AND  EXPOSURE  ASSESSMENT  LABORATORY
                                OFFICE OF RESEARCH AND DEVELOPMENT
                              U.  S.  ENVIRONMENTAL  PROTECTION  AGENCY
                                RESEARCH TRIANGLE  PARK,  NC  27711
CPA-89:radmreport

-------
                                  NOTICE
     The information in this document has been funded by the United States
Environmental Protection Agency under Cooperative  Agreement No. CR-81M558-
01-0 to the University of  California at  Riverside.   It has been subject  to
the Agency's  peer  and  administrative  review,  and it has been approved for
publication as  an  EPA  document.   Mention of  trade names  or commercial
products does not constitute endorsement or recommendation for use.
                                     ii

-------
                                  PREFACE
     This  final  report  describes  an  independent  evaluation  of the  gas-
phase mechanism developed by  Dr. William R. Stockwell of  the  State  Univer-
sity of  New  York  at Albany for use  in the Regional Acid Deposition  Model
(RADM).   Although the authors are  solely  responsible for the  preparation
of this  report  and the development of its conclusions,  the work described
herein  includes  significant  contributions by  Dr. Stockwell  and  several
other scientists.   Dr. Stockwell  participated in this effort  by providing
data files necessary  to  implement  the several  versions  of the  mechanism
that were  evaluated.   He also evaluated  the reviews and recommendations
made by  the authors  throughout  the program,  incorporating  many of  these
recommendations  into  the mechanism   and  generally  keeping  the  authors
informed on  the  continuing  evolution of his mechanism and its  implementa-
tion into  RADM.   In addition, Stockwell  assisted our  evaluation  of  the
effects  of peroxy radical approximations  by  providing  us with  a detailed
mechanism  to use  as the standard.   Without his  cooperation and  participa-
tion, the  utility of  this program  to the overall  assessment of the  RADM
model would have been highly limited.

     In  addition  to  Dr.  Stockwell,  several  other scientists provided
valuable  assistance  in  this program.    Dr. Paulette  Middleton  of  the
National Center of Atmospheric Research solicited our advice  and partici-
pation in  the  development of  the RADM emissions processing system and  is
responsible  for  many  of the  ideas  discussed  in Section  3.2  of  this
report.  Dr. Harvey E. Jeffries  and co-workers at the University of  North
Carolina,  as subcontractors  to this program,  prepared the data  files  used
to implement  their new  light  characterization  model  for the  UNC chamber,
reviewed the input files we used when modeling  the UNC chamber  runs,  and
provided us  with  data for  UNC runs  we  had  not modeled  previously.   Dr.
Roger Atkinson  of the Statewide Air  Pollution  Research Center  (SAPRC)  at
U. C.  Riverside provided valuable  assistance in the initial  review of the
RADM gas-phase  mechanism for  consistency  with  current  chemical data  and
the  results  of  the latest data evaluations.   Finally, the authors  wish  to
acknowledge  the  role  played by  Dr.  Marcia  C.  Dodge  of  EPA/AREAL  in
conceiving of the  need for an independent  evaluation of  the  RADM gas-phase
mechanism,  and  providing  leadership  in  its  overall  direction,  and  her
active interest in  its progress.

     The  authors   also wish  to  acknowledge  the SAPRC   staff members  who
provided major  assistance in  the  preparation  of this report.   Ms.  Minn  P.
Poe  is  responsible for  the preparation  of many of  the figures, and  also
provided  computer  programming  assistance throughout  this  program.    Ms.
Christy  J.  LaClaire  is  responsible for most of  the word  processing  for
this report.
                                    iii

-------
                                 ABSTRACT
     Results are  presented  of a program to independently evaluate the gas
phase mechanism  developed for use  in  the second  version  of the Regional
Acid Deposition  Model (RADM-II).   This work  consisted  of the  following:
(1)  An   initial  review  of the  1987 version  of  the RADM  mechanism was
carried  out  and  several  modifications  were made to the mechanism prior to
the major  evaluation  effort described  in this report.  (2) Assistance was
provided to  the  RADM team  in developing methods for processing emissions
input  for  the  RADM  model  and  for representing  aggregated VOCs  in the
model.   (3)  The  RADM  mechanism was  tested  by  comparing predictions against
results  of over  550  environmental  chamber experiments  carried  out at the
University  of California at  Riverside and  at  the University  of  North
Carolina  (UNC).    The testing using  the  UNC  experiments  incorporated an
improved chamber  light  characterization model  developed  at UNC.  Statis-
tical summaries  of the  overall results and results for individual experi-
ments are  presented and  discussed.   (4)  A series of 90 test problems for
use  in  sensitivity calculations to  assess effects of alternative assump-
tions for  gas-phase mechanisms used in  regional models were  developed, and
these  were  used  to  test  several  condensation  approaches for  the  RADM
mechanism.   (5)  As a result  of this evaluation against the chamber  data
and  the  sensitivity  test calculations,  recommendations  were  made for
modifications  to the  RADM  mechanism concerning the  representation of the
gas-phase  chemistry of the aromatics and the  condensation  of the represen-
tation of  the  higher  alkanes.  Two  modified RADM mechanisms, one involving
primarily  changes  in  parameter values,   and   one  where  more  extensive
changes  and  use  of  fewer  model   species  are  involved,  were  developed.
These modified mechanisms and their performance in  simulating  the  chamber
data and the sensitivity test problems are discussed.  The  version of the
mechanism  involving  primarily parameter changes  has  been adopted by the
RADM team  for  implementation  into the model.

     It  is concluded  that  the RADM mechanism,  as modified as  a  result  of
this  program,  represents  the state-of-the-art  in  gas-phase  atmospheric
reaction mechanisms and  is suitable for use  in the  regional acid  deposi-
tion  model.   However,  there  are still  major uncertainties in  our  under-
standing of atmospheric  chemistry.   Also, available chamber data are not
sufficient for evaluating  the ability  of mechanisms to represent  the low
NOX  conditions characteristic of much  of the  regional modeling  domain, nor
for  evaluating their  prediction  of  peroxide species or organic  acids. The
major areas  of uncertainty  and needs for future research are discussed.
                                     iv

-------
                                 CONTENTS

                                                                        Page

     Tables	vii

     Figures	xi

1.    Introduction	1

2.    The March 1988 RADM and Recommended Modified RADM Gas-Phase
     Chemical Mechanisms	5

     2.1  The March 1988 RADM Mechanism	6
     2.2  Recommended Modifications to the RADM Mechanism	29
          2.2.1  Recommended Modifications to the Aromatic Mechanism	29
          2.2.2  Condensed Higher Alkane Representation	37
          2.2.3  Modifications to the XN02 Reactions	38
          2.2.4  Omitted OLN Reactions	40
          2.2.5  Parameter Modifications Resulting from Updated
                 Emissions Assignments	-	40
     2.3  The Recommended Modified Mechanism (RADM-M)	41
     2.4  The Modified Parameter Mechanism (RADM-P)	42

 3.   Aggregation and Representation of Organic Emissions in  RADM	57

     3.1   Introduction	57
     3.2   Aggregation of VOC Emissions Input for RADM	58
            3.2.1   Development of a Mechanism-Independent VOC
                   Emissions Classification System	59
            3.2.2   Recommended Extensions to the VOC  Emissions
                   Classification System	73
            3.2.3   Aggregation of Emissions into the  RADM Mechanism	76
     3.3   Derivation of Kinetic and Mechanistic Parameters  for
           Lumped Model Species  in RADM	87
            3.3.1   Mechanistic Parameters for Lumped  Alkanes	93
            3-3-2   Mechanistic Parameters for Lumped  Aromatics	97
            3.3.3   Mechanistic Parameters for Lumped  Alkenes	99

 4.   Evaluation  Against Environmental Chamber Data:
     Characterization  and  Lumping Procedures	100

     4.1   Introduction	100
     4.2   Characterization of Chamber Conditions	102
            4.2.1   Light Characterization	103
            4.2.2   Representation  of Other  Chamber-Dependent
                   Effects	112
     4.3   Representation of Organics  in Chamber  Experiments	138

-------
                                   CONTENTS
                                 (continued)
5.   Evaluation Against Environmental Chamber Data:  Results	142

     5.1  Organization of the Evaluation	143
     5.2  Measures of Model Performance	146
     5.3  Performance on Characterization Runs	149
     5.4  Performance of the RADM Mechanism on Single Organic Runs	155
           5.4.1  Carbonyl-NO  Runs	155
           5.4.2  Alkene-N0x Runs	162
           5.4.3  Alkane-NO  Runs	169
           5.4.4  Aromatic-N0x Runs	175
           5.4.5  Natural Hydrocarbon Runs	184
           5.4.6  Performance in Simulations of Formaldehyde Yields	190
           5.4.7  Performance in Simulations of PAN  Yields	192
     5.5  Performance of the RADM Mechanism on Runs  with
          Organic Mixtures	195
     5.6  Tests of the RADM VOC Aggregation Approach	218
     5.7  Performance of the Recommended Modified Mechanisms	219

6.   Evaluation of Alternative Chemical  Assumptions	242

     6.1  Test  Problems for Sensitivity  Analysis	242
     6.2  Presentation of Sensitivity Test Results	256
     6.3  Treatment  of Peroxy Radical Reactions	258
     6.4  Lumping of Higher Alkanes	290
     6.5  The Recommended vs. the March  1988 RADM Mechanisms	292

7.   Summary and Conclusions	310

     7.1  Evaluation of Kinetic and Mechanistic  Parameters	312
     7.2  Development of the Linkage Between the VOC Emissions
          Inventory  and the RADM  Mechanism	314
     7.3  Mechanism  Evaluation Against  Chamber Data	315
     7.4  Sensitivity Testing of  Alternate Mechanistic  Assumption	320
     7.5  Recommendations	322
            7.5.1  Longer Term Research  Needs	323
            7.5.2  Shorter Term Needs	325

     References	328

  APPENDICES

     A   Selected Results  of Simulations of  All  Experiments
          Modeled Using  the  RADM Mechanism	333

     B   Selected Results  of Simulations of  All  Organic-N0x
          Experiments Modeled Using  the RADM-M  Mechanism	356

     C   Selected Results  of Simulations of  All  Organic-N0x
          Experiments Modeled Using  the RADM-P  Mechanism	373
                                     vi

-------
                                   TABLES

Table
Number

  1.  Model Species Employed in the March 1988 Version of the RADM
      Gas-Phase Chemical Mechanism	8

  2.  Listing of Reactions in the March 1988 Version of the RADM
      Gas-Phase Chemical Mechanism	11

  3.  Absorption Coefficients and Quantum Yields Used in Versions
      of the RADM Gas-Phase Chemical Mechanism Evaluated in this
      Study	19

  4.  Organic Model Species Employed in the Recommended Modified
      Version of the RADM Gas-Phase Chemical Mechanism (RADM-M)	43

  5.  Listing of the Organic Reactions in the Recommended Modified
      Version of the RADM Gas-Phase Chemical Mechanism (RADM-M)	45

  6.  Listing of the Organic Reactions in the Modified Parameter
      Version of the RADM Gas-Phase Chemical Mechanism (RADM-P)	51

  7.  Summary of the Emissions Groupings Using the  32-Class Scheme,
      Amounts of Each  Emitted  in  the NAPAP Total U.S. Inventory,  and
      Estimated Amounts  Reacted in Conditions Representative  of RADM
      Appl ications	71

  8.  Assignments of Emissions Groups to RADM Model Species	77

  9.  Averages of  Parameters for  Emissions Groups Lumped as C^
      Alkanes and  for  RADM  Lumped Alkane Model Species, Derived Using
      the  NAPAP Total  U.S.  Emissions Inventory	89

  10.  Averages of  Parameters for  Emissions Groups Lumped as  Aromatics
      and  for  RADM  Lumped Aromatic Model Species, Derived  Using the
      NAPAP Total  U.S. Emissions  Inventory	90

  11.  Averages of  Parameters for  Emissions Groups Lumped as  Non-Ethene
      Alkenes  and  for  RADM  Lumped Non-Ethene  Alkene Model  Species,
      Derived Using the NAPAP  Total U.S. Emissions  Inventory	91

  12.  Averages of  Parameters for  Emissions Represented  by  the Combined
      Higher Alkane (HC5 +  HC8) Model  Species Recommended  for the
      RADM-M Mechanism	98

  13.  ^ummary  of Chamber-Dependent  Parameters Used  in the  Model
      Simulat ions  of the SAPRC EC Chamber  Runs	117

  14.   Summary  of Chamber-Dependent  Parameters Used  in the  Model
       Simulations  of the SAPRC ITC  Chamber Runs	119
                                     vii

-------
                                    TABLES
                                 (continued)

Table
Number                                                                  Page

 15.  Summary of Chamber-Dependent  Parameters Used in the Model
      Simulations of  the SAPRC OTC  Chamber Runs .......................... 121

 16.  Values of the Continuous Chamber Radical Input Parameter, kRg,
      used  in Modeling SAPRC ITC and OTC Chamber Runs, and Averages
      of Measured Values Used to Derive Them ............................. 124

 17.  Summary of Chamber-Dependent  Parameters Used in the Model
      Simulations of  the UNC Chamber Runs ................................ 125

 18.  Values of Adjustable Chamber-Dependent Parameters Employed
      in the Simulations of UNC Chamber Runs ............................. 127

 19.  Selected Performance Statistics in the RADM Simulations of
      the N0x-Air, C0-N0x-Air, Pure Air, and Carbonyl-Air Runs ........... 150

 20.  Performance Statistics on Maximum Ozone and on NO Oxidation
      and Ozone Formation Rates in  the Simulations of the
      Organic-N0x-Air Runs Using the RADM Mechanism ...................... 156

 21.  Performance Statistics on Maximum Formaldehyde in the
      Simulations of  the Organic-N0x-Air Runs Using the RADM
      Mechanism [[[ 191

 22.  Performance Statistics on Maximum PAN in the Simulations of
      the Organic-N0x-Air Runs Using the RADM Mechanism .................. 193

 23.  Comparison of Average Normalized Biases and Errors  in
      Simulations of  Chamber Experiments using the RADM Mechanism
      with  Run-to-Run Adjustment of Rate Constants with Simulations
      using the Standard RADM VOC  Aggregation Procedure .................. 220

 21.  Performance Statistics on Maximum Ozone and NO Oxidation
      and Ozone Formation Rates in the Simulations of
      Organic-N0x-Air Runs Using the RADM-M Mechanism
  25.   Performance Statistics on  Maximum Ozone  and  NO Oxidation
       and Ozone Formation Rates  in  the  Simulations of
       Organic-N0x-Air Runs Using the RADM-P Mechanism .................... 226

  26.   Performance Statistics on  Maximum Formaldehyde in the
       Simulations of Organic-N0x-Air Runs Using the RADM-M
       Mechanism [[[ 228

  27.   Performance Statistics on  Maximum Formaldehyde in the
       Simulations of Organic-N0x-Air Runs Using the RADM-P

-------
                                   TABLES
                                 (continued)

Table
Number                                                                   Page

 28.  Performance Statistics on Maximum PAN in the Simulations of
      Organic-N0x-Air Runs Using the RADM-M Mechanism	230

 29.  Performance Statistics on Maximum PAN in the Simulations of
      Organic-N0x-Air Runs Using the RADM-P Mechanism	231

 30.  Comparisons of Average Normalized Biases and Errors  in Maximum
      Ozone and NOX Oxidation Rates in the Simulations of  Organic-
      NO -Air Runs Using the RADM, RADM-M, and RADM-P Mechanisms	232
        X

 31.  Environmental Conditions Used in Test Problems for Sensitivity
      Calculations	244

 32.  Compositions of ROG Surrogate Mixtures  Used  in Test
      Calculations	'	249

 33.  Compositions of ROG Surrogates Used  in  the Test Calculations
      for  the RADM Mechanism	252

 34.  Summary of  Conditions of Test Calculations Used for
      Sensitivity Studies	253

 35.  Organic Peroxy +  Organic Peroxy  Reactions Used  in  Stockwell's
      "Detailed Radical"  Version  of the  RADM  Mechanism	262

 36.  Listing of the Organic  Reactions in  the Modified  RADM
      Mechanism Incorporating  the 1986 SAPRC  Approach for
      Representing  Reactions  of  Organic Peroxy Radicals	270

 37.  Results of Sensitivity  Test Calculations Using  the Evaluated
      RADM Mechanism as the Test Mechanism,  Relative  to  the
       "Detailed Radical"  Mechanism as the  Standard	277

  38.   Results of Sensitivity Test Calculations for the  Version of
       the RADM  Mechanism Using the 1986 SAPRC Peroxy  Radical
       Representation as the Test Mechanism, Relative  to the
       "Detailed Radical"  Mechanism as the  Standard	282

  39.   Results  of Sensitivity Test Calculations Using  the Version of
       the RADM  Mechanism with the Higher Alkane Species HC5 and HC8
       Lumped Together,  Relative to the Standard RADM Mechanism	293

  40.   Summary  of Results of Sensitivity Test Calculations Using the
       Recommended Modified (RADM-M) and the  Recommended Modified
       Parameter (RADM-P) Mechanisms,  Relative to the March 1988
       RADM Mechanism as the Standard	298
                                      IX

-------
                                   TABLES
                                 (continued)
Table
Number
 41.  Results of Sensitivity Test Calculations of Urban Ozone
      Formation Using the Recommended Modified (RADM-M) and the
      Recommended Modified Parameter (RADM-P) Mechanisms, Relative
      to the March  1988 RADM Mechanism as the Standard	301

 42.  Results of Sensitivity Test Calculations of Urban H202
      Formation Using the Recommended Modified {RADM-M) and the
      Recommended Modified Parameter (RADM-P) Mechanisms, Relative
      to the March  1988 RADM Mechanism as the Standard	304

-------
                                   FIGURES

Figure
Number                                                                   Page

  1.   Comparison of Fits of RADM and Modified RADM Simulations
      to Results of Toluene and Xylene-N0x-Air Chamber Experiments	36

  2.   Experimental and Calculated Concentration-Time Plots for
      Selected Species in the UNC Formaldehyde-N0v-Air Run AU0279B	113
                                                 A

  3.   Experimental and Calculated Concentration-Time Plots for
      Selected Species in the UNC Complex Mixture-NO -Air Run
      ST1682B	114

  4.   Experimental and Calculated Concentration-Time Plots for
      Selected Species in the UNC Complex Mixture-NO -Air Run
      ST2981R	115

  5.  Comparison of Results of UNC Characterization Simulations
      Using  Various Mechanisms and Chamber and Light-Character-
      ization Models	129

   6.  Hierarchy of Species for Mechanism Testing	144

   7.  Calculated vs Experimental  Changes  in  [NO]  in NO -Air  and
      C0-N0v-Air Runs	151
            A

   8.  Calculated vs Experimental  Changes  in  [N02]  in NO  -Air and
      C0-N0v-Air Runs	151
            A

   9.  Calculated vs Experimental  Maximum  Ozone  in  the  Pure  Air  and
      Carbonyl-Air Runs	153

  10.  Calculated vs Experimental  Rate  of  Ozone  Formation in  the
      Pure Air  and Carbonyl-Air  Runs	153

  11.  Histogram of Normalized  Biases  in Maximum Ozone  in Pure Air and
      Carbonyl-Air Runs	.-	154

  12.  Normalized Biases  in Maximum Ozone  vs  HC/NOX Ratios in
      Carbonyl-Air Runs	154

  13.  Calculated vs Experimental  Maximum  Ozone  in Carbonyl-N0x
      Runs	159

  1U.  Calculated vs Experimental  Rates of NO Oxidation and  Ozone
      Formation in  the Carbonyl-N0v Runs	159
                                   X

  15.  Histogram of Normalized  Biases  in Maximum Ozone  in
      Carbonyl-N0x  Runs	160

  16.  Normalized  Bias in Maximum Ozone vs HC/NOX Ratio in
      Carbonyl-N0x-Runs	160


                                     xi

-------
                                   FIGURES
                                 (continued)

Figure
Number                                                                  Page

 17.  Experimental and Calculated Concentration-Time Plots for
      Selected Species in Simulations of UNC Acetaldehyde-N0x Run
      AU2482B	161

 18.  Calculated vs Experimental Maximum Ozone  in the Alkene-N0x
      Runs for the RADM Mechanism	163

 19.  Calculated vs Experimental Rates of NO Oxidation and Ozone
      Formation in the Alkene-N0v Runs for  the  RADM Mechanism	163
                                A

 20.  Histogram of Normalized Biases in Maximum Ozone in Alkene
      Runs for the RADM Mechanism	164

 21.  Normalized Bias in Maximum Ozone vs HC/NOX  in Alkene Runs
      for the RADM Mechanism	164

 22.  Experimental and Calculated Concentration-Time Plots for
      Selected Species in Simulations of UNC Ethene-N0x Run
      OC058MB	165

 23.  Experimental and Calculated Concentration-Time Plots for
      Selected Species in Simulations of the Propene-N0x Run
      EC-216	166

 2M.  Experimental and Calculated Concentration-Time Plots for
      Selected Species in Simulations of UNC Propene-N0x Run
      JL2983B	167

 25.  Experimental and Calculated Concentration-Time  Plots for
      Selected Species in Simulations of UNC  1-Butene-NOx  Run
      ST2583B	168

 26.  Calculated  vs  Experimental Maximum Ozone in the  Alkane-N0x
      Runs  for  the  RADM  Mechanism	170

 27.  Calculated  vs  Experimental  Rates  of  NO Oxidation  and Ozone
      Formation  in  the  Alkane-N0x  Runs  for the RADM Mechanism	170

 28.  Histogram of  Normalized Biases in Maximum Ozone in  Alkane
       Runs  for the  RADM  Mechanism	171

  29.  Normalized Bias in Maximum Ozone  vs  HC/NOX in Alkane Runs
       for the RADM Mechanism	171

  30.   Experimental  and Calculated Concentration-Time Plots for
       Selected Species in Simulations of n-Butane-NOx Run EC-130	172
                                     xii

-------
                                   FIGURES
                                 (continued)

Figure
Number                                                                  Page

 31.  Experimental and Calculated Concentration-Time Plots for
      Selected Species in Simulations of n-Octane-NOx Run ITC-552	173

 32.  Experimental and Calculated Concentration-Time Plots for
      Selected Species in Simulations of UNC 2,3-Dimethylbutane
      -NOX Run OC1789R	174

 33.  Calculated vs Experimental Maximum Ozone in the Aromatic-
      NOX Runs for the RADM Mechanism	176

 3^.  Calculated ys Experimental Rates of NO Oxidation and Ozone
      Formation in the Aromatic-N0x Runs for the RADM Mechanism	176

 35.  Histogram of Normalized Biases in Maximum Ozone in Aromatic
      Runs for the RADM Mechanism	177

 36.  Normalized  Bias in Maximum Ozone ys HC/NO  in  Aromatic  Runs
      for the RADM Mechanism	177

 37.  Experimental and Calculated  Concentration-Time Plots for
      Selected Species in Simulations of  the Benzene-NO  Run
       ITC-562	178
  38.   Experimental  and  Calculated  Concentration-Time  Plots  for
       Selected  Species  in  Simulations of  the  Toluene-NO  Run
       EC-327	179

  39.   Experimental  and  Calculated  Concentration-Time  Plots  for
       Selected  Species  in  Simulations of  the  UNC  Toluene-NO  Run
       JL3080R	180
  40.   Experimental  and Calculated Concentration-Time Plots for
       Selected Species in Simulations of the  m-Xylene-NO  Run
       ITC-702	181

  41.   Experimental  and Calculated Concentration-Time Plots for
       Selected Species in Simulations of the  UNC o-Xylene-NO  Run
       JL3080B	182

  42.   Experimental  and Calculated Concentration-Time Plots for
       Selected Species in Simulations of the  Mesitylene-NO  Run
       EC-901	V	183

  43.   Calculated ys Experimental Maximum Ozone in the Natural
       Hydrocarbon Runs for the RADM Mechanism	185
                                    xiii

-------
                                   FIGURES
                                 (continued)

Figure
Number                                                                  Page

 44.  Calculated vs Experimental Rates of NO Oxidation and Ozone
      Formation in the Natural Hydrocarbon Runs for the RADM
      Mechanism	185

 45.  Histogram of Normalized Bias in Maximum Ozone in the Natural
      Hydrocarbon Runs for the RADM Mechanism	186

 46.  Normalized Bias in Maximum Ozone vs HC/NOX in the Natural
      Hydrocarbon for the RADM Mechanism	186

 47.  Experimental and Calculated Concentration-Time Plots for
      Selected Species in Simulations of the Isoprene-N0x Run
      ITC-811	187

 48.  Experimental and Calculated Concentration-Time Plots for
      Selected Species in Simulations of the UNC Isoprene-NO  Run
      JL1680B	188
 49.  Experimental and Calculated Concentration-Time Plots for
      Selected  Species in Simulations of the UNC a-Pinene-NO  Run
      JL1580B	189
 50.  Calculated  vs  Experimental Maximum Formaldehyde in the
      Single  Organic-NOK Runs for the RADM Mechanism	192

 51.  Calculated  vs  Experimental Maximum PAN in the Single
      Organic-N0x Runs  for  the RADM Mechanism	194

 52.  Calculated  vs  Experimental Maximum Ozone in  the Simple
      Mixture Runs for  the  RADM Mechanism	196

 53.  Calculated  vs  Experimental Rates of NO Oxidation and Ozone
      Formation in the  Simple Mixture Runs for the RADM Mechanism	196

 54.  Histogram of Normalized Biases in Maximum Ozone in the Simple
      Mixture Runs for  the  RADM Mechanism	197

 55.  Normalized  Bias  in Maximum Ozone vs HC/NOX in the Simple
      Mixture Runs for  the  RADM Mechanism	197

 56.  Calculated  vs  Experimental Maximum Ozone in  the Surrogate
      Mixture Runs for  the  RADM Mechanism	198

 57.  Calculated  vs  Experimental Rates of NO Oxidation and Ozone
      Formation in the  Surrogate Mixture Runs for  the RADM
      Mechanism	198
                                    xiv

-------
                                   FIGURES
                                 (continued)
Figure
Number
 58.  Histogram of Normalized Biases in Maximum Ozone in the
      Surrogate Mixture Runs for the RADM Mechanism	199

 59.  Normalized Bias in Maximum Ozone vs HC/NOX in the Surrogate
      Mixture Runs for the RADM Mechanism	199

 60.  Normalized Bias in Maximum Ozone vs Initial NOX in the
      Surrogate Mixture Runs for the RADM Mechanism	200

 61.  Normalized Bias in Maximum Ozone vs Initial HC in the
      Surrogate Mixture Runs for the RADM Mechanism	200

 62.  Experimental and Calculated Concentration-Time Plots for
      Selected Species in Simulations of the UNC Mixed Alkene Run
      OC1278R	,	201

 63.  Experimental and Calculated Concentration-Time Plots for
      Selected Species in Simulations of the 7 Hydrocarbon
      Surrogate Run EC-237	202

 64.  Experimental and Calculated Concentration-Time Plots for
      Selected Species in Simulations of the 8 Hydrocarbon
      Surrogate, Multi-Day Run ITC-630	203

 65.  Experimental and Calculated Concentration-Time Plots for
      Selected Species in Simulations of the 8 Hydrocarbon
      Surrogate, Multi-Day Run ITC-633	204

 66.  Experimental and Calculated Concentration-Time Plots for
      Selected Species in Simulations of the 8 Hydrocarbon
      Surrogate, Multi-Day Run ITC-637	205

 67.  Experimental and Calculated Concentration-Time Plots for
      Selected Species in Simulations of the 8 Hydrocarbon
      Surrogate, Multi-Day Run OTC-192B	206

 68.  Experimental and Calculated Concentration-Time Plots for
      Selected Species in Simulations of the 8 Hydrocarbon
      Surrogate, Multi-Day Run OTC-195A	207

 69.  Experimental and Calculated Concentration-Time Plots for
      Selected Species in Simulations of the 8 Hydrocarbon
      Surrogate, Multi-Day Run OTC-202B	208

 70.  Experimental and Calculated Concentration-Time Plots for
      Selected Species in Simulations of the 8 Hydrocarbon
      Surrogate, Multi-Day Run OTC-224B	209
                                    xv

-------
                                   FIGURES
                                 (continued)

Figure
Number                                                                   Page

 71.  Experimental and Calculated Concentration-Time Plots for
      Selected Species in Simulations of the UNC 3 Hydrocarbon
      Surrogate Run JN1483R	210

 72.  Experimental and Calculated Concentration-Time Plots for
      Selected Species in Simulations of the UNC Miscellaneous
      Surrogate Run AU2681B	211

 73.  Experimental and Calculated Concentration-Time Plots for
      Selected Species in Simulations of the UNC Synurban
      Surrogate Run JN2685R	212

 7M.  Experimental and Calculated Concentration-Time Plots for
      Selected Species in Simulations of the UNC Synauto
      Surrogate Run AU0684R	213

 75.  Experimental and Calculated Concentration-Time Plots for
      Selected Species in Simulations of the UNC Auto Exhaust
      Run AU1183R	214

 76.  Calculated vs Experimental Maximum Formaldehyde in the
      Simple Mixture Runs for the RADM Mechanism	216

 77.  Calculated vs Experimental Maximum Formaldehyde in the
      Surrogate Mixture Runs for the RADM Mechanism	216

 78.  Calculated vs Experimental Maximum PAN in  the Simple
      Mixture Runs for the  RADM Mechanism	217

 79.  Calculated vs Experimental Maximum PAN in  the Surrogate
      Mixture Runs for the  RADM Mechanism	217

 80.  Plots of Maximum Ozone Yields  in Simulations of Organic
      Mixture Experiments using the  RADM Mechanism with Run-to-Run
      Adjustment of Rate Constants Against  Yields Calculated Using
      the Standard RADM VOC Aggregation Procedure	222

 81.  Calculated vs Experimental Maximum Ozone  in the Surrogate
      Mixture Runs for the  RADM-M Mechanism	235

 82.  Calculated vs Experimental Maximum Ozone  in the Surrogate
      Mixture Runs for the  RADM-P Mechanism	235

 83.  Calculated vs Experimental Maximum Ozone  in the Alkane-NO
      Runs  for the RADM-M Mechanism	236

 8M.  Calculated vs Experimental Maximum Ozone  in the Alkane-NO
      Runs  for the RADM-P Mechanism	236
                                    xv i

-------
                                   FIGURES
                                 (continued)

Figure
Number                                                                  Page

 85.  Calculated vs Experimental Maximum Ozone in the Aromatic-
      NOY Runs for the RADM-M Mechanism .................................. 237
        A

 86.  Calculated vs Experimental Maximum Ozone in the Aromatic-
      NOX Runs for the RADM-P Mechanism .................................. 237

 87.  Comparisons of Maximum Ozone Calculated using the RADM-M
      and the Unmodified RADM Mechanisms for Simple and Surrogate
      Mixture Runs [[[ 240

 88.  Comparisons of Maximum Ozone Calculated using the RADM-P
      and the Unmodified RADM Mechanisms for Simple and Surrogate
      Mixture Runs [[[ 240

 89.  Concentration-Time Plots for Selected Species Calculated
      Using the RADM, RADM-M, and RADM-P Mechanisms for the Summer
      Surface Static Urban Test Problem with Initial NOX of 67 ppb
      and a ROG/NOV Ratio of 10 .......................................... 307
                  A

 90.  Concentration-Time Plots  for Selected Species Calculated
      Using the RADM, RADM-M, and RADM-P Mechanisms for the Winter

-------
                              1.  INTRODUCTION

     The Regional Acid Deposition Model  (RADM) plays a  central  role  in  the
National Acid  Precipitation Assessment  Program's  (NAPAP)  plan to develop
an understanding  of source-receptor  relationships for acidifying pollut-
ants in  the  United  States.   This model  is needed  to formulate  appropriate
and cost-effective  emission control  policies  to  mitigate the  potentially
damaging effects  of acid  deposition.   While transport,  deposition,  and
chemistry all  play  important roles in the acid deposition phenomenon,  the
chemistry plays a  particularly  important role in  determining  the response
of the  acid  deposition system to major  changes  in NO  and  SOV emissions.
                                                      A        A
The  chemistry  submodel  is  the  only component  of the  overall  modeling
system  that  is capable  of  predicting  nonlinearly  in  the  relationship
between  source emissions  and  acid deposition.    Thus,  it  is  critically
important that any  chemical mechanism incorporated into the RADM model be
fully evaluated using  the  best  available data and that  it  be recognized by
the  scientific community as  representing the current  state-of-the-art in
our understanding of atmospheric chemistry.
     Although  the  full  RADM  model  must  incorporate   both  gas  phase  and
aqueous  phase  chemistry,  the study described in  this  report  addresses only
the  gas phase portion of the chemistry  module.   As presently  implemented,
only  the gas  phase species  are  transported within the  model,  and their
concentrations at  each  time  step are  used  to   drive the  aqueous phase
chemistry,  based  on stationary  state approximations.   Therefore,  the gas
phase  chemistry module  contributes  substantially to  any  non-linearities
predicted  by the model.   The gas phase  chemistry module  is  also important
because,  as implemented in  the  first version of  the RADM model (RADM-I),
approximately  75J  of  the computer  time  is  consumed in  calculating the gas
phase  transformations.  This is with a smaller  version of the gas phase
chemical mechanism (Stockwell, 1986)  than  the one proposed  for use in the
second  version of  this  model.   Thus,  clearly  the efficiency  and  appro-
priateness  of the  level of  detail  in the gas phase transformation mechan-
 ism,  as  well  as  its  accuracy  and  scientific validity,  is  an important
 issue which needs  to  be  addressed.
     Because of  computer  constraints,  the gas-phase   chemistry  module in
 the  first version of  the RADM model  (Stockwell,   1986)   was highly condensed

-------
and  contained  a  number of  approximations  that  were  chemically unreal-
istic.  With the  expanded memory capacities  of  the newer computers, a new,
more  detailed  gas-phase chemical  mechanism was  proposed  for  use  in the
second version  of the RADM model  (designated  RADM-II)  which is now under
development  (NCAR,   1987).    Although  in  terms  of  chemical  detail  and
consistency with  current  knowledge, this  new  RADM gas-phase mechanism  is
an  improvement  over  the   previous version,   it  had  not  undergone  the
external review and  comprehensive  evaluation that  is  necessary  for such  an
important  component  of a  model  which  will   be  used  in  making  policy
decisions.   For  example,  this  new mechanism  had  undergone only limited
testing  against  the  results  of a small  number  of environmental chamber
experiments.   In  contrast,  the mechanisms most recently developed for use
in oxidant and  EKMA  models  have been tested against  results  of  hundreds  of
experiments  carried  out  in  several  different  chambers  (Carter  et  al.,
1986;  Gery  et al.,  1988).   The December  1986 EPA workshop  on evaluation
and documentation of chemical mechanisms for use  in air quality simulation
models concluded  that an  extensive testing program against chamber data  is
an  important  part  of the overall  evaluation  process for such mechanisms
(EPA,  1987).    In  addition,  the   new   RADM mechanism  had  not  undergone
external review for  consistency with available laboratory  data  and  current
chemical  knowledge.    Such  a review and evaluation would be particularly
useful  prior  to  the  implementation  of  the mechanism  in  RADM, when  any
problems  can  be  more readily  corrected.    To  address  this need, the  EPA
contracted  the Statewide  Air Pollution Center (SAPRC) at the University  of
California  at Riverside to carry  out an independent  review  and evaluation
of this new  proposed gas-phase chemical  mechanism  for RADM,  and to  make
suggestions and recommendations concerning  areas where the mechanism might
be improved.
      This   evaluation  program  was a   cooperative  effort  involving  the
authors  of this  report at SAPRC  and  Lurmann  and Associates,  Dr.  Harvey
Jeffries and  his group at  the University of North Carolina  (UNC),  Dr.
William Stockwell  at the  State University of New  York  (SUNY), and Dr.
 Paulette  Middleton  at  the  National  Center  for  Atmospheric  Research
 (NCAR).   The  evaluation  itself was carried out at  SAPRC  with  the assist-
 ance of  Lurmann  and  Associates.   Dr  Jeffries and his group  assisted  in
 this effort  by reviewing the  input data used when  testing the mechanism

-------
against  results  of  UNC  chamber  experiments  and preparing  input data  to
implement a new model  for  light  characterization  in  that chamber  (Jeffries
et al.f  1989a).    Dr.  Stockwell,  the  principal developer  of the  RADM-II
mechanism and  the member of the  RADM  team ultimately responsible  for  the
gas-phase mechanism  which will  be used in RADM-II,  provided the  initial
and  modified  versions  of the  mechanism  which were  evaluated  in  this
program,  and  served  as  an  overall  consultant  on this  project.    Dr.
Middleton is  responsible  for  preparing the emissions  input  for the  RADM
model  and  as part  of  this program,  SAPRC worked with  her  in developing
appropriate procedures for processing  organic emissions data for use  with
this  mechanism.    This  report  describes  primarily  the  work  on  this
evaluation effort carried out by SAPRC and Lurmann and Associates.
     This  evaluation  effort  involved  several  separate,  though related,
tasks:   (1)  SAPRC carried out an  initial  review of a preliminary  version
of  the RADM-II mechanism  (NCAR, 1987) to  determine its consistency  with
current  data  and  data evaluations  and made recommendations  for  modifica-
tions.   As  a result of these recommendations, Stockwell made some  modifi-
cations  in   the  RADM-II   mechanism prior  to  its  implementation  on SAPRC
computers  for further evaluation.    (2)  SAPRC  also made  recommendations
regarding organic  emissions  processing procedures and methods used  in  the
mechanism to  represent organic  emissions.   Most of  these  recommendations
were  subsequently implemented in  the  RADM-II  mechanism  and in  emissions
processing  procedures  developed  for it.   (3)  With the assistance  of  UNC,
modifications  were  made  in the  light  characterization and  chamber  effects
models  used when  testing  mechanisms  against UNC chamber  data.    (4)  The
predictions of the  RADM-II mechanism,  as modified by Stockwell as a result
of the  initial review, was tested against  the results of approximately  550
UNC  and  SAPRC chamber  runs,  and  a statistical analysis of  the results  was
carried  out.   Other modified  versions of  the  RADM  mechanism were  tested
against  these  chamber  data as  well.   (5)  A number of alternative  mechanis-
tic  assumptions  and condensation  approaches  for  the RADM mechanism  were
tested using  a series  of 90 test  problems representing a  range  of  condi-
tions  expected in  regional model applications.   (6)  Additional recommenda-
tions  for modifications  of the RADM mechanism were made as  a  result  of  the
simulations   of   the  chamber  experiments  and  the   test  calculations.
Stockwell incorporated some  of these  recommendations in a  "final"  version

-------
of the  RADM mechanism, which,  as  of this writing,  is  expected to be  the
version which will  be  implemented  into the RADM-II model.  This mechanism
is included among those  which were simulated  in  the chamber  experiments
and test calculations.
     This report  is  organized into seven sections.  Following  this  intro-
ductory section,  Section  2 describes and lists the version of  the RADM-II
mechanism which was  evaluated  in the model simulations  carried  out in  this
study,  and  two modified  versions  which  were  recommended as  a result  of
this program.   Section 3  describes the work  carried  out in this program
concerning  aggregation and  representation  of organic  emissions  data  for
use with  the RADM-II mechanism.   Section ^  describes the procedures  used
when  evaluating   the mechanisms against  the  chamber  data,  with emphasis
particularly  on  the  changes  in   the   procedures  and  input  data  used,
compared  with  previous studies.   Section 5  presents  the  results  of  the
simulations  of  the  evaluated  and  recommended modified  RADM mechanisms
against the chamber  data, including statistical summaries of  the perform-
ance  of  the  mechanisms  and  representative  concentration-time   plots.
Section 6 describes  the test calculations carried out  to  test  alternative
mechanistic  assumptions   and   degrees  of   condensation,   including   a
discussion  of  the  90  simulations  employed,  the mechanistic  alternatives
tested, and the  results obtained.    Finally,  Section 7 gives a summary of
the results  from  this  program and  the conclusions.

-------
           2.  THE  MARCH  1988  RADM  AND  RECOMMENDED MODIFIED RADM
                       GAS-PHASE  CHEMICAL MECHANISMS
     In this  section,  we summarize and  list the versions  of the  RADM-II
chemical mechanisms  which  were evaluated during  this study, and describe
the modifications  to the mechanism which  are recommended  as a result  of
this program.  At  the start of this program,  the current RADM-II mechanism
was  the  version   documented  in  the  NCAR  (1987)  report.    However,   as
indicated  earlier,  Stockwell modified  this mechanism  shortly after this
program  began as  a  result of an  initial  review  we  carried out  on this
mechanism.   This  modified mechanism,  which was dated  March 30, 1988,  is
the version  that was implemented at SAPRC  for evaluation purposes.   There-
fore,  the  March  1988 version, and not the  1987 mechanism,   is  the  version
of  the mechanism whose  evaluation  is  discussed  in this report.  A  listing
of  this  mechanism  is   presented   in  Section  2.2.    In  the  subsequent
discussion,  this version of  the mechanism  will  be designated simply as  the
"RADM"  or  the "March 1988 RADM" mechanism.
     The  results  of our  evaluation efforts, described in  more detail  in
Sections  3,  5 and  6  of  this  report,  indicated several areas where  improve-
ments  to  the RADM  mechanism  were warranted.   These  included changes in  the
representation  of  the   unknown  aspects  of  the  aromatic  photooxidation
mechanisms to improve  its performance  in  simulations  of the chamber data
while   reducing  the  number  of species  involved,  a   condensation   in  the
representation  of  the  alkanes, some  changes in rate constant  and  product
yield   parameters  for   lumped  species  to   better  correspond  to our  most
recent analysis  of emissions data, correction  of an  error  in  the  methyl-
glyoxal quantum yields, addition of a few  reactions omitted from the NO^ •*•
alkene mechanism,  and minor  improvements in the chemistry  of the condensed
representations  of the reactions  of nitrophenols.   The modifications  we
recommended  to  the  RADM  mechanism are described  in  Section 2.2,  and  the
resulting modified mechanism is listed in  Section  2.3-  In the subsequent
discussion,  this  modified version  of the  RADM-II  mechanism is designated
as  the "recommended  modified," or  simply as the "RADM-M," mechanism.
     By the  time  we completed  the evaluation  effort and  formulated  our
recommendations  for modifying  the mechanism,  much of the  work  in imple-
menting  the  RADM mechanism  in  the  overall RADM-II  model software  had

-------
already  been carried  out.    Although Stockwell  and  the RADM  team were
willing   to   implement  our   recommendations,   we   were  informed  that
modifications involving  more  than Just changes  in kinetic or mechanistic
parameters  (i.e.,  rate  constant values,  product yield  coefficients,   or
absorption  coefficient  and  quantum  yield  data)  would  be  difficult   to
implement  without  adversely  impacting  the  RADM-II  development schedule.
Therefore, we developed  a  set of recommendations for modifications  of  the
RADM  mechanism   that  involved  only  changes  in  parameter values.    These
changes  included  revising  the aroraatics  mechanism,  updating  the rate
constants  and  product yield  parameters  for lumped  species  based  on  the
most  recent  analysis of emissions data,  and correcting  the  methylglyoxal
quantum  yields.    In  addition,  the RADM  team was willing  to add the  few
reactions  which were  omitted  from the  NO?  +  alkene  mechanism,  so this
change  was also  incorporated  into  this  version  of  the  mechanism.   This
second  modified  version  of  the RADM   mechanism  is  designated  as  the
"modified  parameter,"  or simply the  "RADM-P," mechanism.  It  is  described
and listed in Section 2.4.
     The  RADM  team has accepted the  recommendations  for  modifications
which  were incorporated in the RADM-P mechanism,  and this version  of  the
mechanism  is now  being  implemented  into  the RADM-II  model  (Stockwell,
private  communication, 1989).

2.1  The March  1988  RADM Mechanism
     The  RADM  gas-phase chemical mechanism that  was used as  the  starting
point  in this evaluation effort is  the  March  1988 update of  the  RADM-II
mechanism  described in the NCAR (1987) report.  This is  a condensed mech-
anism which  updated and  extends the  earlier mechanism of Stockwell  (1986),
with  the representation  of  the reactions  of the aroraatics based on that of
Lurmann et  al.  (1986).   However,   this mechanism  has  a  more  detailed
representation  of  the  reactions of  the  alkanes,  organic acids,  organic
hydroperoxides,  and the  radical-radical  reactions occurring in the absence
of NOX.  The increased  level  of detail  for  the  acids,  organic peroxides,
and radical-radical reactions  is appropriate  for  a mechanism to  be used
for acid deposition modeling.
      The modifications  made  to  the  NCAR  (1987)  mechanism  included some
relatively minor  changes  in  some of the rate constants  to  be consistent

-------
with the latest evaluations of NASA (1987) and IUPAC (1988).   Changes were
also made  in  the  representation  of glycolaldehyde, an acetaldehyde  oxida-
tion product,  and in  the  representation of  lumped higher  alkanes  to  be
more consistent with  observed organic  nitrate  yields  from such  compounds
(Carter and  Atkinson,  1985).   Some changes  in  rate constants  and product
yields  for lumped species based  on  results of  a  preliminary  analysis  of
emissions  data  (see Section  3)  were also  made.   The absorption  coeffi-
cients  and quantum yields for the  photolysis  reactions were  not changed.
Most of the  changes made  to the mechanism were  relatively  minor;  the basic
characteristics, approaches,  and  types  of chemical approximations employed
in  this version are  the  same  as  the version documented in  the NCAR  (1987)
report.
     The model  species employed  in the RADM mechanism are  listed in Table
1.   There  are a total  of  18  inorganic species,  14  species  representing the
various  types  of  organic oxidation   products,   11  species  representing
emitted hydrocarbons,  and 15 species representing reactive organic  inter-
mediates.   The set of inorganic species is  the same as  that employed in
most other  state-of-the-art  mechanisms.   Like  the Lurmann et al.  (1986)
mechanism  and  the more  recent  mechanisms  developed  at  SAPRC  (Carter et
al.,   1986,  1987; Lurmann  et al.,  1987;  Carter,  1988),  but  unlike  the
Carbon Bond mechanisms (Gery et al.,  1988), the reactions of the organics
are represented on the molecular level, with some  species being  represent-
ed explicitly, and  others being  represented  using either the  "surrogate
species"  or the "lumped molecule" lumping approaches.   The organic  species
 represented explicitly in RADM are methane,  ethane, ethene,  formaldehyde,
 formic acid, methyl  hydroperoxide,  and  isoprene.  The  surrogate  species
 approach   involves using  a   single  representative  compound  to represent
 reactions  of other compounds assumed  to have  similar mechanisms or reac-
 tivity.   For example, the surrogate  acetaldehyde  (denoted as ALD)  is  used
 to represent all  C2+  aldehydes.   The  lumped molecule   approach  involves
 representing  mixtures   of   compounds  by  pseudo-molecules  whose   rate
 constants and perhaps other  mechanistic parameters are determined based on
 the mixture they  are  used to represent.  This lumping technique  is  used to
 represent  Co+  alkanes and  alkenes and  (to some  extent)   the aromatic
 hydrocarbons.  No attempt is made to conserve carbon  in this  mechanism, as
 is the case for most other current molecularly-based mechanisms.

-------
Table 1.  Model Species Employed in the March  1988 Version of the RADM
          Gas-Phase Chemical Mechanism


 Name      Description
  Transported Inorganic Species

  S02      Sulfur Dioxide
  03       Ozone
  H202     Hydrogen Peroxide
  NO       Nitric Oxide
  N02      Nitrogen Dioxide
  N205     Nitrogen Pentoxide
  HONO     Nitrous Acid
  HN03     Nitric Acid
  HN04     Peroxy Nitric Acid
  CO       Carbon Monoxide

  Inorganic Radical and Atomic Intermediates

  HO       Hydroxyl (OH) Radicals
  H02      Hydroperoxyl Radicals
  N03      Nitrate  Radicals
  03P      0(3P) Atoms
  01D      0('D) Atoms

  Constant Inorganic Species

  02       Oxygen
  N2       Nitrogen
  M        Air
  H20      Water

  Non-Reacting Inorganic Species

  H2       Hydrogen
  C02      Carbon Dioxide
  SULF     Sulfuric Acid (or sulfates)

  Organic Hydrocarbons

  CHI      Methane
  ETH      Ethane
  HC3      C3+ Alkanes with kOH < 5 x 103 (a)
  HC5      Alkanes with kOH Between 0.5 and 1.0 x 101*
  HC8      Alkanes with kOH > 1.0 x 1CT
  OL2      Ethene
  OLT      Terminal Monoalkenes
  OLI      Internal Alkenes, Terpenes, Conjugated Dialkenes
  ISO      Isoprene
  TOL      Aromatic Hydrocarbons with kOH < 2 x 1o!j
  XYL      Aromatic Hydrocarbons with kOH > 2 x 10^
                                                              (continued)
                                     8

-------
Table 1 (continued) - 2
 Name      Description
  Organic Oxidation Products

  HCHO     Formaldehyde
  ALD      Lumped Higher Aldehydes
  KET      Ketones
  GLY      Glyoxal
  MGLY     Methyl Glyoxal
  DCB      Uncharacterized Reactive  Aromatic  Fragmentation  Products
  CSL      Phenols and Cresols

  PAN      Acyl  Peroxy Nitrates  (other  than those  formed  from DCB)
  TPAN     Acyl  Peroxy Nitrates  Formed  from DCB
  ONIT     Other Organic Nitrates  (primarily  alkyl nitrates)

  OP1      Methyl Hydroperoxide
  OP2      Higher Organic Hydroperoxides
  ORA1     Formic Acid
  ORA2     Higher Organic Acids  (other  than peroxy acetic acid)
  PAA      Peroxy Acetic Acid

  Organic Radicals

  M02      Methyl  Peroxy Radicals
  ETHP     Ethyl Peroxy  Radicals
  HC3P     Peroxy  Radicals  from  OH + HC3
  HC5P     Peroxy  Radicals  from  OH + HC5
  HC8P     Peroxy  Radicals  from  OH + HC8
  OL2P     Peroxy  Radicals  from  OH + Ethene
  OLTP     Peroxy  Radicals  from  OH + Isoprene and Terminal Alkenes
  OLIP     Peroxy  Radicals  from  OH + Internal Alkenes
  KETP     Peroxy  Radicals  from  OH + Ketones
  TOLP      Peroxy  Radicals  from  OH + Benzene  and Monoalkyl Benzenes
   XYLP      Peroxy  Radicals  from  OH + Xylenes  and More Reactive Aromatics
  OLN      Peroxy  Radicals  from  NO? + Alkenes
   X02     Chemical "Operator" Used to Account for NO-to-N02 Conversions
           due to Secondarily Formed Organic  Peroxy Radicals

   AC03      Acyl Peroxy Radicals  (other than those formed from DCB)
   TC03      Acyl Peroxy Radicals  Formed from DCB

   XN02      Chemical Operator Used to Account for NOX Loss Due  to
            Reactions of NO? with Phenols and Cresols.


 (a) kOH  is the OH radical rate constant at 300 K  in ppm~1 min   .

-------
     The  RADM  mechanism  includes  a  relatively  detailed,   though  still
approximate,  representation  of  the  reactions  of peroxy  radical  inter-
mediates.   This representation involves significantly more reactions than
used for  these  processes  in most other  current  lumped mechanisms.   It also
uses more model species  to represent the products of  these reactions  in
the  absence of  NOX,  with  separate  representations of  methyl  and  higher
hydroperoxides,  formic and  higher  acids, and  separate  representation  of
peroxyacetic  acid.   Most  of the reacting  primary organics and  organic
products  have separate representations of  their peroxy  radical  products,
and  the  mechanism includes  explicit  representation  of  the  reactions  of
these  radicals with  NO,  with NC^ (for acyl peroxy  radicals),  with HC^,
with methyl  peroxy  (M02),  and with acetylperoxy  (AC03) radicals.   In
representing  explicitly  almost all of the possible peroxy  +  M02  and peroxy
+  AC03  reactions,  this mechanism  is more  detailed  in  this  regard than most
other  current  mechanisms,  which  either  ignore  most of  these  processes
(Gery  et al.f 1988),  or uses approximations  involving  chemical  "operators"
to represent  them (Carter  et al.,   1986;  Lurmann et al.,   1987;  Carter,
1988).   However,  this mechanism omits the reactions of  the  peroxy species
OLN (formed in the NO? + alkene reactions) with M02 and AC03, and ignores
peroxy  +  peroxy  reactions  not  involving M02  and AC03.   The  effects  of
alternative representations of peroxy radical  reactions  in  this mechanism
are discussed in Section 6.3.
     Table 2  gives  a  listing  of   the  reactions and  the  thermal  rate
constants  used  in the March  1988  RADM mechanism, and  Table 3  gives the
absorption coefficients  and  quantum  yields  used  for its  photolysis reac-
tions.     The  reactions  and  thermal rate  constants  were  obtained from
Stockwell  in  computer  readable  format, along with input  data and results
of a test  calculation employing constant photolysis rates.  This mechanism
was implemented   on  the  computers  at SAPRC and  Lurmann and  Associates
(which each use different  model calculation  software),  and  the  results of
test problem  were duplicated.   The  absorption coefficients and quantum
yields used in this mechanism for atmospheric simulations were transmitted
to us  separately, and  we  used these  to calculate photolysis  rates as a
function of zenith angle for the  solar spectrum used  in the test problems
described  in  Section 6.1  of this report.   Computer-readable  data  files
 implementing  this mechanism  are available  from the authors  of this  report
upon request  (with update information available from  Stockwell at SUNY).
                                     10

-------
Table 2.  Listing of Reactions in the March 1988 Version of the RADM
          Gas-Phase Chemical Mechanism
Rxn.
Label 	
(a)    k(300)
Kinetic Parameters (b)

       A        Ea    B
Reactions (c)
                                       Inorganic Reactions (d)
001
002
003
004
005
006
007
008
009
022
023
024
025
026
027
028
029
030
031



032

033A
033B
034A
034B
035
036



037
038
039
040
041
042


(Phot. Set = N02R )
(Phot. Set = 030 1DR )
(Phot. Set = 0303PR )
(Phot. Set = HONOR )
(Phot. Set = HN03R )
(Phot. Set = HN04R )
(Phot. Set = N03NOR )
(Phot. Set = N03N02R )
(Phot. Set = H202R )
2
1
3
5
3
2
1
3
1
2



. 16E-05
.42E+04
.82E+04
.93E+04
.23E+05
.76E+01
. 02E+02
. 05E+00
.21E+04
. OOE+03



2
9
2
4
3
2
2
1
5

6
6

. 16E-05
.54E+03
. 64E+04
.70E+04
.23E-05
.94E+03
•35E+03
.61E+01
.43E+03
(Falloff
.46E-03
. 90E+03
F= 0.60
0
-0
-0
-0
0
2
1
0
-0
.00
.24
.22
.14
.00
.78
.87
.99
.48
-4.30
-1.00
-1.00
-1.00
-1.00
-1.00
-1.00
-1.00
-1.00
Kinetics)
0
0
i
(Equilibrium
3
2
1
1
9
2
7



6
4
4
6
3
1


.24E-03
.55E+03
.79E-03
.34E-01
. 38E-02
.49E+03
.05E+03



.93E-10
.94E-02
. 11E+04
.08E-01
.67E+03
.85E+03


1
3
6
1
2
4

2
2

1
2
2
3
3
.95E+13
.23E+02
.82E-05
.11E-05
. 34E-06
.84E+03
(Falloff
.51E-02
.20E+04
F= 0.60
. 19E-10
.06E+02
.50E+04
.67E+01
. 67E+03
21
-1
-1
-5
-6
0
.00
.00
n =
-5.20
-2.40
1.00
with 031)
.66
.23
.95
.60
.32
.40
1.00
-1.00
-2.00
-2.00
-2.00
-1.00
Kinetics)
0
0
.00
.00
n=
-1
4
-0
2
0
.05
.97
.30
.44
.00
-4.60
-1.50
1.00
-2.00
-1.00
-1.00
-1.00
-1.00
(Falloff Kinetics)
7
2
.90E-02
.20E-1-03
0
0
F= 0.60
.00
.00
n=
-6.30
-1.50
1.00
N02 *
03 +
03 +
HONO
HN03
HN04
N03 +
N03 +
H202
03P +
03P +
01D +
01D +
01D +
03 +
03 +
03 +
H02 +
H02 +
(kO)
(kinf)

HN04
(Keq)
H02 +
H02 +
H02 +
H02 +
H202
NO +
(kO)
(kinf)

NO +
03 +
N03 +
N03 +
N03 +
N03 +
(kO)
(kinf)

HV =
HV =
HV =
+ HV
+ HV
+ HV
HV =
HV =
+ HV
02 +
N02
N2 =
02 =
H20
NO =
HO =
H02 =
NO =
N02



= H02

H02
H02
H02
H02
+ HO
HO =



NO +
N02 =
NO =
N02
H02
N02



03P +
01D +
03P f
= HO *
= HO +
= H02
NO +
N02 +
= HO +
NO
02
02
NO
N02
+ N02
02
03P
HO
M = 03
= NO t-
03P +
03P +
= HO +
N02 +
H02 +
HO +
N02 +
= HN04



+ N02

= H202
+ M =
+ H20
+ H20
= H02
HONO



02
N2
02
HO
02
02
12 02
HO







H202 +
= H202
= H202
* H20




02 = *2 N02
N03
N02 +
= NO +
= HN03
= N205




N02
N02 +
+ 02




                                                              M
                                                              02
                                                              (continued)
                                     11

-------
Table 2 (continued) - 2
Rxn.
1 aha 1
Ltd Dei.
(a)
043
044
045

046
047
048
049



050A
050B

010
011
012
013
014
015
016
017
018

019
020

021

Kinetic Parameters (b)
k(300) A Ea B
(Equilibrium with 042)
2.27E-03 3.72E+13 22.26 1.00
2.94E-06 2.94E-06 0.00 -1.00
1.66E+04 (Falloff Kinetics)
9.34E-02 0.00 -5.20
3.52E+04 0.00 -2.30
F= 0.60 n= 1.00
2.16E+02 2.16E+02 0.00 -1.00
6.77E+03 1.91E+03 -0.75 -1.00
1.45E+05 6.75E+04 -0.46 -1.00
1.29E+03 (Falloff Kinetics)
1.08E-02 0.00 -5.30
2.20E+03 0.00 -1.00
F= 0.60 n= 1.00
2.20E+02 2.20E+02 0.00 -1.00
1.31E-04 1.31E-04 0.00 -2.00

(Phot. Set = HCHOMR )
(Phot. Set = HCHORR )
(Phot. Set = ALDR )
(Phot. Set = OPR )
(Phot. Set = OPR )
(Phot. Set = PAAR )
(Phot. Set = KETR )
(Phot. Set = GLY2R )
(Phot. Set = GLY1R )

(Phot. Set = MGLYR )
(Phot. Set = DCBR )

(Phot. Set = ONJTR )

Rossfif1 1 finQ ( r* j
I\ wu w WXwllv \ w /
N205 = N02 + N03
(Keq)
N205 + H20 = #2 HN03
HO + N02 = HN03
(kO)
(kinf)

HO + HN03 = N03 + H20
HO + HN04 = N02 + H20
HO * H02 = H20 + 02
HO + S02 = SULF + H02
(kO)
(kinf)

CO + HO = H02 + C02
CO + HO + M = H02 + C02
Organic Photolysis Reactions
HCHO + HV = H2 + CO
HCHO + HV = H02 + H02 + CO
ALD * HV = M02 + H02 + CO
OP1 f HV = HCHO + H02 + HO
OP2 + HV = ALD + H02 + HO
PAA + HV = M02 * C02 * HO
KET + HV = AC03 + ETHP
GLY + HV = #.13 HCHO + #1.87 CO
GLY * HV = #.45 HCHO + 11.55 CO +
#.80 H02
MGLY + HV = AC03 + H02 + CO
DCB +'HV = #.98 H02 + #.02 AC03 +
TC03
ONIT + HV = #.20 ALD + #.80 KET *
H02 + N02
051
052
053


054
1
4
3


7
.29Ef
.11E+
.76E-h


.86E+
01
02
03


03
9
1
2


2
.23E+
.81E+
.28Ef


.79E+
02
03
04


04
2
0
1


0
.54
.88
.07


.75
1
1
-1


-1
.00
.00
.00


.00
CH4
ETH
HC3
#
#
HC5
                                       Organic + OH Reactions
                                           + HO = M02 + H20
                                           + HO = ETHP + H20
                                           + HO = #.90 HC3P + #.10 H02 +
                                           ,014 HCHO + 1.069 ALD +
                                           .026 KET + H20
                                           + HO = HC5P + #.27 X02 + H20

                                                              (continued)
                                     12

-------
Table 2 (continued) - 3
Rxn.
Label
(a)
055
056
057
058
074
059

060

063
064
065
066
067
061

062
068
069

070

071
072
073
Kinetic Parameters (b)
k(300)
1.
1.
3.
9.
1.
9.

66E+04
24E+04
82E+04
24E+04
46E+05
02E+03

4.54E+04


1 . 32E+04
2,
1,
1
2
5

5
4
1

1

1
2
3
.37E+04
.47E+03
. 69E+04
.50E+04
.87E+04

.28E+04
. 11E+04
.47E+04

.47E+04

.47E+04
.06E+02
.76E+03
5.
3.
7.
1.
3.
3.

3.

1,
1,
1,
1,
2
5

5
4
1

1

1
9
2
A
87E+04
16E+03
12E+03
48E+04
74E+04
08E+03

08E-.-04

.32E+04
.01E+04
.76E+04
.69E+04
.50E+04
.87E+04

.28E+04
. 11E+04
. 47E+04

.47E+04

.47E+04
. 06E+02
.28E+04
Ea
0.
-0.
-1.
-1.
-0.
-0.

-0.

0.
-0.
1.
0.
0.
0.

0.
75
82
00
09
81
64

23

00
51
,48
,00
,00
,00

.00
0.00
0.00

0,


.00

0.00
0
1
.88
.07
B
-1.
-1.
-1.
-1.
-1.
-1.

-1.

-1.
-1.
-1.
-1.
-1.
-1.

-1.
-1.
-1.

-1.

-1.
1,
-1,

00
00
00
00
00
00

00

00
00
00
00
00
00

00
00
,00

,00

.00
.00
.00
Reactions (
HC8 +
OL2 +
OLT H-
OLI +
ISO +
TOL +
# . 16
XYL +
#.17
HCHO +
ALD ••-
KET *
GLY *
MGLY -»
CSL +
#.9
CSL *
DCB +
OP1 +
#.5
OP2 +
#.5
PAA +
PAN +
ONIT <
HO =
HO =
HO =
HO =
HO =
HO =
H02
HO =
H02
0
HC8P * #.78 X02 + H20
OL2P
OLTP
OLIP
OLTP
#.84 TOLP + #.16 CSL +

#.83 XYLP -•- #.17 CSL +

HO = H02 -i- CO + H20
HO =
HO =
HO =
AC03 + H20
KETP * H20
H02 + #2 CO * H20
- HO = AC03 + CO 4- H20
HO =
TC03
HO =
HO =
HO =
HO
HO =
HO
HO =
HO =
* HO
Organic +
093
094
095
096
097
098
099
100
101
102
9
3
9
3
3
3
1
2
1
8
.23E-01
. 65E+00
.23E-01
.65E+00
. 65E+00
.23E-I-04
.72E-01
.33E+01
.84E+03
.53E+02
8
2
8
2
2
3
2
1
4
8
.81E+02
.06E+03
.81E+02
.06E+03
.06E+03
.23E+04
.94E+03
.47E-.-04
.74E+04
.53E+02
4
3
4
3
3
0
5
3
1
0
.09
.78
.09
.78
.78
.00
.81
.84
.94
.00
-1
-1
-1
-1
-1
-1
-1
-1
-1
-1
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
HCHO •
ALD +
GLY +
MGLY -
DCB +
CSL f
OL2 +
OLT >
OLI +
ISO +
«• N03
N03
N03
«• N03
N03
N03
N03
N03
N03
N03
#.1 H02 * #.9 X02 +

CSL
TC03 * H20
#.5 M02 + #.5 HCHO +

#.5 HC3P * #.5 ALD +

AC03 -*• H20
HCHO + N03 + X02
= HC3P + N02
N03 Reactions
= H02 + HN03 + CO
= AC03 f HN03
= HN03 + H02 + #2 CO
= HN03 * AC03 + CO
= HN03 -t- TC03
= HN03 + XN02 -t- #.5 CSL
= OLN
= OLN
= OLN
= OLN
 103   2.72E-03  1.76E+01  5.23 -1.00
Organic + 03 Reactions

OL2 + 03 = HCHO -H #.42 CO +
  #.4 ORA1 * #.12 H02
                                                               (continued)
                                      13

-------
Table 2 (continued) - 4
Rxn. Kinetic Parameters (b)
(a) k(300) A Ea B
104 1.74E-02 1.94E+01 4.18-1.00 OLT


106 3.02E-01 1.33E+01 2.26-1.00 OL


r «•
.33
.20
.22
+
.10
.29
#.26
106 2.20E-02 1.81E+01 4.00 -1.00 ISO +
#.33
03 =
CO +
ORA2
M02
03 =
KET
ORA2
H02
03 =
CO +


#

+
#
+

+
#

^
/






.53 HCHO
#.20 ORA1
+
*
#
.
.18
#
+
#
.
#
m
.53
#.
#.20 ORA2 +
#.22 M02
+
1


+


#


.50


ALD +
.23 H02 +
10 HO
HCHO
23 CO



+
•*-
-t-
#
#
#
.06
.72
.06
CH4
ALD +
ORA1 +
.09 CH4 +
14 HO
HCHO


+
+
#
#
.31
.50
M02
ALD *
20 ORA1 +
#
.
.23
10 HO
H02 +





 075    7.52E+03   4.11E+03 -0.36 -1.00
 076    2.90E-02   1.17E+18 26.91  0.00
 077    6.90E+03   6.90E+03  0.00 -1.00
 078    2.90E-02   1.17E+18 26.91  0.00
 086    1.12E+04  6.17E+03 -0.36 -1.00
 087    1.12E+04  6.17E+03 -0.36 -1.00
 079   1.12E+04  6.17E+03 -0.36 -1.00
 080   1.12E+04  6.17E+03 -0.36 -1.00
 081    1.12E+04  6.17E+03 -0.36 -1.00

 082    1.12E+OM  6.17E+03 -0.36 -1.00


 083    1.12E+04  6.17E+03 -0.36 -1.00

 084    1.12E+04  6.17E+03 -0.36 -1.00
 085    1.12E+04  6.17E-I-03 -0.36 -1.00

 088    1.12E+04  6.17E+03 -0.36 -1.00


 089   1.12E+04  6.17E+03 -0.36 -1.00
Organic Peroxy + N02 Reactions

AC03 + N02 = PAN
PAN = AC03 f N02
TC03 f N02 = TPAN
TPAN = TC03 + N02

Organic Peroxy + NO Reactions

AC03 * NO = M02 + N02
TC03 * NO = N02 i- #.92 H02 +
  #.89  GLY + #.11 MGLY +
  #.05 AC03 + #.95 CO + #2 X02
M02 + NO = HCHO + H02 * N02
HC3P + NO = #.69 ALD + #.26 KET +
  #.14 HCHO -t- #.03 ONIT +
  #.97 N02 + #.97 H02
HC5P + NO = #.36 ALD * #.76 KET +
  #.10 ONIT + #.90 N02 + #.90 H02
HC8P + NO = #.44 ALD * #1.05 KET +
  #.05 HCHO + #.24 ONIT +
  #.76 N02 + #.76 H02
OL2P + NO = #1.6 HCHO + H02 +
  N02 + #.2 ALD
OLTP * NO = ALD + HCHO i- H02 + N02
OLIP + NO = H02
  #.28  HCHO + #
TOLP + NO = N02
  #.724 MGLY + #
  #.905 DCB
XYLP * NO  = N02
 HCHO i-  H02
 #1.45 ALD +
10 KET + N02
 H02 +
181  GLY  +
 H02 + MGLY + DCB
                                                               (continued)
                                      14

-------
Table 2 (continued) - 5
Rxn.
Label 	
(a)    k(300)
     Kinetic Parameters  (b)

            A        Ea    B
Reactions (c)
090
091
092
150
 107
 108
 109
 110
 111
 112
 113
 111*
 115
 117
 118
 120
 146
 116
 119
 142
 145
 130

 134

 135

 136


 137

 138

 139
1.12E+04  6.17E+03 -0.36  -1.00
1.12E+04  6.17E+03 -0.36  -1.00
1.12E+04  6.17E+03 -0.36  -1.00
1.12E+04  6.17E+03 -0.36  -1.00
8.61E+03
ETHP + NO = ALD +
KETP + NO = MGLY
OLN + NO = HCHO +
X02 + NO = N02
                 H02 + N02
                + N02 +  H02
                 ALD + #2 N02
                                       Organic Peroxy + H02  Reactions
1.13E+02 -2.58 -1.00 H02 +
(Same
(Same
(Same
(Same
(Same
(Same
(Same
(Same
(Same
(Same
(Same
(Same
(Same
(Same
k
k
k
k
k
k
k
k
k
k
k
k
k
k
as
as
as
as
as
as
as
as
as
as
as
as
as
as
107 )
107 )
107 )
107 )
107 )
107 )
107 )
107 )
107 )
107 )
107 )
107 )
107 )
107 )
H02 +
H02 +
H02 +
H02 +
H02 +
H02 +
H02 +
H02 +
H02 +
H02 +
H02 +
H02 +
H02 +
H02 +
M02 :
ETHP
HC3P
HC5P
HC8P
OL2P
OLTP
OL1P
KETP
TOLP
XYLP
OLN
X02
AC03
TC03
: OP1
= OP2
= OP2
= OP2
= OP2
= OP2
= OP2
= OP2
= OP2
= OP2
= OP2
= ON IT
= OP2
= PAA
= OP2
                                        Organic Peroxy + AC03 Reactions
3.64E+03   1.75Ef03 -0.44 -1.00
3.64E+03   1.75E+03 -0.44 -1.00
2.93E+03   1.41E+03 -0.44 -1.00

 1.04E+03   4.99E+02 -0.44 -1.00

3.06E+02   1.47E+02 -0.44 -1.00

2.57Ef02   1.23E+02 -0.44 -1.00


2.20E+02   1.06E+02 -0.44 -1.00

 1.04E+03   4.99E+02 -0.44 -1.00

 1.04E+03   4.99E+02 -0.44 -1.00
AC03 * AC03
AC03 + TC03
  *.890 GLY
  *.05 AC03
M02 + AC03 =
  #.5 M02 +
ETHP + AC03
  1.5 M02 -K
HC3P t- AC03
  1.5 H02 +
HC5P + AC03
  *.86 KET +
  1.5 ORA2
HC8P + AC03
  #.5 H02 +
OL2P + AC03
  1.5 H02 +
OLTP + AC03
  #.5 H02 +
             =  #2 M02
             =  M02 + #.92 H02 -H
             +  1.11 MGLY +
             *  1.950 CO + *2 X02
             HCHO + #.5 H02 +
             1.5 ORA2
             =  ALD + #.5 H02 *
             #.5 ORA2
             =  1.2 ALD + 1.8 KET +
             1.5 M02 - 1.5 ORA2
             =  1.11 ALD +
              1.5 H02 +1.5 M02 +

             =  1.10 ALD + #.9 KET +
             #.5 M02 +1.5 ORA2
             =1.8 HCHO + #.6 ALD +
             1.5 M02 + 1.5 ORA2
             =  ALD + 1.5 HCHO +
             1.5 M02 +1.5 ORA2

                        (continued)
                                      15

-------
Table 2 (continued) - 6
Rxn.
Label 	
(a)    k(300)
Kinetic Parameters (b)

       A        Ea    B
Reactions (c)
140   1.28E+02  6.17E+01 -0.44 -1.00
141
143
144
148
121
122
123
124
125
126
127
128
129
131
132
133
1.28E+02
1.28E+02
1 . 28E+02
1.28E+02
5.81E+02
4.28E+02
1.28E+02
1.04E+02
8.86E+01
4.28E+02
4.28E+02
5.19E+01
5.19Ei-01
5.19E+01
5.19E+01
2.93E+03
                2.79E+02 -0.44 -1.00
                2.06E+02 -0.44 -1.00

                6.17E+01 -0.44 -1.00

                4.99E+01 -0.44 -1.00

                4.26E+01 -0.44 -1.00

                2.06E+02 -0.44 -1.00

                2.06E+02 -0.44 -1.00

                2.50E+01 -0.44 -1.00

                2.50E+01 -0.44 -1.00

                2.50Ef01 -0.44 -1.00
      2.93E+03  1.41E+03 -0.44 -1.00
147   5.19E+01  2.50Ef01 -0.44 -1.00
                            OLIP + AC03 = #.725 ALD +
                              #.55 KET + #.14 HCHO + #.5 H02 +
                              #.5 M02 + t.5 ORA2
                            KETP + AC03 = MGLY + #.5 H02 +
                              #.5 M02 + #.5 ORA2
                            AC03 * TOLP = M02 + #.8 MGLY *
                              8.2 GLY + DCB + H02
                            AC03 * XYLP = M02 + MGLY + DCB +
                              H02
                            X02 + AC03 = M02

                            Organic Peroxy + M02 Reactions

                            M02 + M02 = #1.5 HCHO + H02
                            M02 -t- ETHP = #.75 HCHO + H02 +
                              #.75 ALD
                            M02 + HC3P = #.75 HCHO * H02 *
                              #.15 ALD * #.60 KET
                            M02 + HC5P = #.75 HCHO * H02 *
                              #.105 ALD + #.645 KET
                            M02 + HC8P = #.75 HCHO + H02 *
                              #.075 ALD * #.675 KET
                            M02 + OL2P = #1.55 HCHO * H02 +
                              #.35 ALD
                            M02 + OLTP = #1.25 HCHO i- H02 +
                              #.75  ALD
M02
                         H02
                         H02
                                  OLIP  =  1.89  HCHO
                              #.725  ALD + #.55 KET
                            M02  + KETP  =  #.75  HCHO
                              #.75 MGLY
                            M02  + TOLP  =  #.976 HCHO +
                              #1.9 H02  +  #.724 MGLY +
                              #.181  GLY + #.905 DCB
                            M02  1- XYLP  =  HCHO  + #2 H02 +
                              MGLY + DCB
                            M02  + TC03  =  #.5 HCHO + #.5 ORA2 +
                              #.46 H02  *  #.445 GLY +
                              #.055  MGLY  + #.025 AC03 +
                              #.475  CO  +  X02
                            M02  + X02 = HCHO + H02

                                                   (continued)
                                    16

-------
Table 2 (continued) - 7
           Kinetic Parameters (b)
Rxn.
Label 	
(a)    k(300)     A        Ea    B
                                       Reactions (c)
                                       Other Organic Peroxy + Peroxy
                                       Reactions
 149    1.10E+00  5.28E-01 -0.44 -1.00   X02 + X02 =
 151    1.12E+04  6.17E+03 -0.36 -1.00
 152               (Same k as  107  )
 153    5.19E+01  2.50E+01 -0.44 -1.00
 154    1.28E+02  6.17E+01 -0.44 -1.00
 155    1.10E+00  5.28E-01 -0.44 -1.00
                                       XN02 Reactions

                                       XN02 + N02 = ONIT
                                       XN02 + H02 = OP2
                                       XN02 + M02 = HCHO
                                       XN02 + AC03 = M02
                                       XN02 + XN02 =
H02
 Notes:

 (a)   Reaction  labels  in  this  listing  are  the  reaction sequence numbers
      used by Stockwell to  implement this  mechanism.   Some reactions were
      re-arranged  for  the purpose  of this  listing.   Some reactions were
      split into two reactions for implementation on SAPRC modeling
      software.

 (b)   Except as indicated otherwise, rate  constants for reactions in this
      table are given  by  the expression

           k =  A (T/300)8 e~Ea/RT

      where k and  A are  in  ppm, minute units,  T is the temperature in
      degrees K, and R is 0.0019872 kcal deg   mole   .  If the
      notation  "Phot.  Set"  is used, then this is a photolysis reaction
      whose rate constant is calculated using the indicated set of
      absorption coefficients and  quantum yields given in Table 3.  If
      the notation "Falloff" is used,  then this reaction is both
      temperature  and  pressure dependent,  and its rate constant is given
      by
           k = t(k0 M) / (1 + [k0 M / kinf] ) ]
      where
           g = 1 / [1 + (Iog10 [k0 M / kinf] / n)2 ],

      M is the pressure in ppm, k« is the low pressure limiting rate
      constant calculated as a function of temperature from its
                                                               (continued)
                                      17

-------
Table 2 (continued) - 8
     corresponding A, Ea and B parameters given on the next line,
     is the high pressure limiting rate constant calculated from its
     parameters given on the line below that, and the parameters F
     and n are given as indicated in the line below that.  If the
     notation "Equilibrium with" is used, the rate constant is calculated
     by multiplying the rate constant for the indicated reaction by the
     equilibrium constant, whose T=300 K value and temperature dependence
     parameters, analogous to A, Ea and B, are given on the following
     line.

(c)  If a number preceded by a "K" appears on the list of reaction
     products, then the number is a product yield coefficient for the
     species whose name follows it.  Otherwise the product yield is 1.0.

(d)  These reactions are also used in the modified (RADM-M and RADM-P)
     versions of this mechanism.
                                     18

-------
Table 3.  Absorption Coefficients and Quantum Yields Used in Versions
          of the RADM Gas-Phase Chemical Mechanism Evaluated in this
          Study
WL
(nm)
Abs
(cm2)
QY
WL
(nm)
Abs
(cm2)
QY
WL
(nm)
Abs
(cm2)
QY
Data Used in the March 1988 and the Recommended Versions of the
RADM Mechanism

Photolysis Set = N02R
186.0 2.59E-19 1.000
191.4 2.73E-19 1.000
197.0 2.46E-19 1.000
203.0 2.92E-19 1.000
209.4 4.10E-19 1-000
216.2 4.60E-19 1.000
223.5 3.70E-19 1.000
231.2 2.76E-19 1.000
239.5 7.85E-20 1.000
248.4 2.13E-20 1.000
258.1 1.60E-20 1.000
268.5 2.72E-20 1.000
279.7 5.26E-20 1.000
292.0 9.39E-20 1.000
303.0 1.51E-19 1-000
306.0 1.53E-19 1.000
309.0 1.81E-19 ^.000
312.0 2.03E-19 1.000
316.0 2.20E-19 1.000
330.0 3.04E-19 1.000
345.0 4.05E-19 1.000
360.0 4.94E-19 1.000
375.0 5.63E-19 1.000
390.0 6.00E-19 0.981
405.0 5.49E-19 0.356
420.0 O.OOE-01 0.018
187.8 2.72E-19 1.000
193.2 2.51E-19 1.000
199.0 2.46E-19 1.000
205.1 3.53E-19 1.000
211.6 4.36E-19 1.000
218.6 4.17E-19 1.000
226.0 3.89E-19 1.000
233-9 1.71E-19 1.000
242.4 6.12E-20 1.000
251.6 1.40E-20 1.000
261.4 1.80E-20 1.000
272.1 3-27E-20 1.000
283.7 6.27E-20 1.000
296.3 1.09E-19 1.000
304.0 1.56E-19 1.000
307.0 1.61E-19 1.000
310.0 1.83E-19 1.000
313.0 2.01E-19 1.000
320.0 2.50E-19 1.000
335.0 3.40E-19 1.000
350.0 4.46E-19 1.000
365.0 5.29E-19 1.000
380.0 5.61E-19 1.000
395.0 5.69E-19 0.922
410.0 5.98E-19 0.135

189.6 2.85E-19 1.000
195.1 2.44E-19 1.000
201.0 2.82E-19 1.000
207.3 3-74E-19 1.000
213.9 4.45E-19 1.000
221.0 4.54E-19 1.000
228.6 2.69E-19 1.000
236.7 1.62E-19 1.000
245.4 2.88E-20 1.000
254.8 1.49E-20 1.000
264.9 2.18E-20 1.000
275.9 4.10E-20 1.000
287.8 7.62E-20 1.000
300.5 1.26E-19 1.000
305.0 1.55E-19 1.000
308.0 1.65E-19 1.000
311.0 1.91E-19 1.000
314.0 1.99E-19 1-000
325-0 2.81E-19 1.000
340.0 3.84E-19 1.000
355.0 4.88E-19 1-000
370.0 5.40E-19 1.000
385.0 5.72E-19 0.997
400.0 6.18E-19 0.693
415.0 5.77E-19 0.060

 Photolysis Set =  0301DR
186.0 6.23E-19 0.900
191.4 4.75E-19 0.900
197.0 3.46E-19 0.900
203.0 3.27E-19 0.900
209.4 5.46E-19 0.900
216.2 1.20E-18 0.900
223.5 2.57E-18 0.900
231.2 4.94E-18 0.900
239.5 8.11E-18 0.900
187.8 5.76E-19 0.900
193.2 4.27E-19 0.900
199.0 3-22E-19 0.900
205.1 3.65E-19 0.900
211.6 7.05E-19 0.900
218.6 1.56E-18 0.900
226.0 3-24E-18 0.900
233.9 5.93E-18 0.900
242.4 9.20E-18 0.900
189.6 5.26E-19 0.900
195.1 3.82E-19 0.900
201.0 3.14E-19 0.900
207.3 4.37E-19 0.900
213.9 9.20E-19 0.900
221.0 2.00E-18 0.900
228.6 4.04E-18 0.900
236.7 7.01E-18 0.900
245.4 1.02E-17 0.900
                                                               (continued)
                                     19

-------
Table 3 (continued) - 2
WL
(nm)
Abs
(cm )
QY
WL
(nm)
Abs
(cm2)
QY
WL
(nm)
Abs
(cm2)
QY
248.4  1.10E-17 0.900
258.1  1.14E-17 0.900
268.5  8.35E-18 0.900
279.7  4.03E-18 0.900
292.0  1.11E-18 0.900
303-0  2.64E-19 0.900
306.0  1.79E-19 0.837
309.0  1.19E-19 0.639
312.0  8.01E-20 0.304
316.0  4.69E-20 0.065

Photolysis Set = 0303PR

186.0  6.23E-19 0.100
191.4  4.75E-19 0.100
197.0  3.46E-19 0.100
203.0  3.27E-19 0.100
209.4  5.46E-19 0.100
216.2  1.20E-18 0.100
223.5  2.57E-18 0.100
231.2  4.94E-18 0.100
239.5  8.11E-18 0.100
248.4  1.10E-17 0.100
258.1  1.14E-17 0.100
268.5  8.35E-18 0.100
279.7  4.03E-18 0.100
292.0  1.11E-18 0.100
303.0  2.64E-19 0.100
306.0  1.79E-19 0.163
309.0  1.19E-19 0.361
312.0  8.01E-20 0.696
316.0  4.69E-20 0.935
330.0  6.69E-21  1.000
345.0  7.21E-22  1.000
360.0  5.49E-23  1.000
415.0  3.14E-23  1.000
430.0  6.83E-23  1.000
445.0  1.49E-22  1.000
460.0  3-57E-22  1.000
475.0  4.89E-22  1.000
490.0  8.28E-22  1.000
505.0  1.62E-21  1.000
520.0  1.78E-21  1.000
535.0  2.74E-21  1.000
550.0  3.17E-21  1.000
251.6 1.14E-17 0.900
261.4 1.08E-17 0.900
272.1 6.93E-18 0.900
283.7 2.79E-18 0.900
296.3 6.37E-19 0.900
304.0 2.33E-19 0.899
307.0 1.57E-19 0.789
310.0 1.03E-19 0.531
313.0 6.90E-20 0.216
320.0 2.65E-20 0.000
187.8 5.76E-19 0.100
193.2 4.27E-19 0.100
199.0 3.22E-19 0.100
205.1 3.65E-19 0.100
211.6 7.05E-19 0.100
218.6 1.56E-18 0.100
226.0 3.24E-18 0.100
233.9 5.93E-18 0.100
242.4 9.20E-18 0.100
251.6 1.14E-17 0.100
261.4 1.08E-17 0.100
272.1 6.93E-18 0.100
283.7 2.79E-18 0.100
296.3 6.37E-19 0.100
304.0 2.33E-19 0.101
307.0 1.57E-19 0.211
310.0 1.03E-19 0.469
313.0 6.90E-20 0.784
320.0 2.65E-20 1.000
335.0 3-06E-21 1.000
350.0 2.66E-22 1.000
365.0 O.OOE-01 1.000
420.0 3.99E-23 1.000
435.0 8.66E-23 1.000
450.0 1.71E-22 1.000
465.0 3.68E-22 1.000
480.0 7.11E-22 1.000
495.0 9.09E-22 1.000
510.0 1.58E-21 1.000
525.0 2.07E-21 1.000
540.0 2.88E-21 1.000
555.0 3.36E-21 1.000
254.8 1.16E-17 0.900
264.9 9.64E-18 0.900
275.9 5-43E-18 0.900
287.8 1.81E-18 0.900
300.5 3.63E-19 0.900
305.0 1.99E-19 0.872
308.0 1.36E-19 0.725
311.0 9.13E-20 0.413
314.0 6.33E-20 0.150
189.6 5.26E-19 0.100
195.1 3.82E-19 0.100
201.0 3.14E-19 0.100
207.3 4.37E-19 0.100
213.9 9.20E-19 0.100
221.0 2.00E-18 0.100
228.6 4.04E-18 0.100
236.7 7.01E-18 0.100
245.4 1.02E-17 0.100
254.8 1.16E-17 0.100
264.9 9.64E-18 0.100
275.9 5.43E-18 0.100
287.8 1.81E-18 0.100
300.5 3.63E-19 0.100
305.0 1.99E-19 0.128
308.0 1.36E-19 0.275
311.0 9.13E-20 0.587
314.0 6.33E-20 0.850
325.0 1.30E-20 1.000
340.0 1.37E-21 1.000
355.0 1.09E-22 1.000
410.0 2.91E-23 1.000
425.0 6.54E-23 1.000
440.0 1.25E-22 1.000
455.0 2.12E-22 1.000
470.0 4.06E-22 1.000
485.0 8.43E-22 1.000
500.0 1.22E-21 1.000
515.0 1.60E-21 1.000
530.0 2.55E-21 1.000
545.0 3.07E-21 1.000
560.0 3.88E-21 1.000
                                                               (continued)
                                     20

-------
Table 3 (continued) - 3
WL Abs QY
( nm ) ( cm )
565.0 4.31E-21 1.000
580.0 4.55E-21 1.000
595.0 4.61E-21 1.000
610.0 4.54E-21 1.000
625.0 3.60E-21 1.000
640.0 2.74E-21 1.000
660.0 2.07E-21 1.000
690.0 1.11E-21 1.000
720.0 6.40E-22 1.000
Photolysis Set = HONOR
311.0 2.50E-22 1.000
314.0 4.50E-21 1.000
325.0 3-93E-20 1.000
340.0 1.69E-19 1.000
355.0 2.58E-19 1.000
370.0 2.06E-19 1.000
385-0 1.46E-19 1.000
Photolysis Set = HN03R
191.4 1.37E-17 1.000
197.0 8.54E-18 1.000
203.0 3.60E-18 1.000
209.4 9.93E-19 1.000
216.2 2.57E-19 1.000
223.5 9.10E-20 1.000
231.2 4.66E-20 1.000
239.5 2.48E-20 1.000
248.4 1.92E-20 1.000
258.1 1.87E-20 1.000
268.5 1.57E-20 1.000
279.7 1.01E-20 1.000
292.0 4.44E-21 1.000
303.0 1.74E-21 1.000
306.0 1.24E-21 1.000
309.0 7.85E-22 1.000
312.0 5.15E-22 1.000
316.0 2.00E-22 1.000
330.0 O.OOE-01 1.000
WL Abs QY
( nm ) ( cm )
570.0 4.67E-21 1.000
585.0 4.35E-21 1.000
600.0 4.89E-21 1.000
615.0 4.24E-21 1.000
630.0 3-43E-21 1.000
644.8 2.61E-21 1.000
670.0 1.72E-21 1.000
700.0 9-13E-22 1.000
730.0 5.14E-22 1.000

312.0 2.02E-21 1.000
316.0 5.26E-21 1.000
330.0 8.51E-20 1.000
345.0 1.14E-19 1.000
360.0 7.31E-20 1.000
375.0 3.71E-20 1.000
390.0 1.42E-20 1.000

193.2 1.22E-17 1.000
199.0 6.66E-18 1.000
205.1 2.48E-18 1.000
211.6 6.64E-19 1.000
218.6 1.65E-19 1.000
226.0 7.26E-20 1.000
233.9 3.68E-20 1.000
242.4 2.20E-20 1.000
251.6 1.90E-20 1.000
261.4 1.77E-20 1.000
272.1 1.40E-20 1.000
283.7 8.13E-21 1.000
296.3 2.95E-21 1.000
304.0 1.55E-21 1.000
307.0 1.09E-21 1.000
310.0 6.71E-22 1.000
313.0 4.37E-22 1.000
320.0 8.50E-23 1.000

WL Abs QY
( nm ) ( cm )
575.0 4.75E-21 1.000
590.0 4.42E-21 1.000
605.0 4.84E-21 1.000
620.0 3.90E-21 1.000
635.0 3.17E-21 1.000
651.1 2.40E-21 1.000
680.0 1.37E-21 1.000
710.0 7.93E-22 1.000


313.0 3.97E-21 1.000
320.0 3-93E-20 1.000
335.0 6.31E-20 1.000
350.0 1.07E-19 1.000
365.0 1.84E-19 1.000
380.0 8.25E-20 1.000
395.0 O.OOE-01 1.000

195.1 1.04E-17 1.000
201.0 5.12E-18 1.000
207.3 1.67E-18 1.000
213-9 3.78E-19 1.000
221.0 1.22E-19 1.000
228.6 5.77E-20 1.000
236.7 3.00E-20 1.000
245.4 2.03E-20 1.000
254.8 1.89E-20 1.000
264.9 1.67E-20 1.000
275.9 1.21E-20 1.000
287.8 6.17E-21 1.000
300.5 2.08E-21 1.000
305.0 1.39E-21 1.000
308.0 9.35E-22 1.000
311.0 5.93E-22 1.000
314.0 3.30E-22 1.000
325.0 3.50E-23 1.000

                                                               (continued)
                                     21

-------
Table 3 (continued) - 4
WL Abs QY
(nm) (cm )
Photolysis Set = HN04R
189.6 1.00E-17 1.000
195.1 6.60E-18 1.000
201.0 4.00E-18 1.000
207.3 2.40E-18 1.000
213.9 1.50E-18 1.000
221.0 1.00E-18 1.000
228.6 7.50E-19 1.000
236.7 6.00E-19 1.000
245.4 4.40E-19 1.000
254.8 3.00E-19 1.000
264.9 2.00E-19 1.000
275.9 1.25E-19 1.000
287.8 4.50E-20 1.000
300.5 1.35E-20 1.000
305.0 9.00E-21 1.000
308.0 7.00E-21 1.000
311.0 5.50E-21 1.000
314.0 4.20E-21 1.000
325.0 1.50E-21 1.000
Photolysis Set = N03NOR
585.0 2.66E-18 0.032
600.0 2.69E-18 0.311
615.0 1.84E-18 0.194
630.0 4.89E-18 0.078
WL Abs QY
(nm) (cm2)

191.4 9.00E-18 1.000
197.0 5.60E-18 1.000
203-0 3.40E-18 1.000
209.4 2.10E-18 1.000
216.2 1.30E-18 1.000
223.5 9.00E-19 1.000
231.2 6.80E-19 1.000
239.5 5.50E-19 1.000
248.4 4.00E-19 1.000
258.1 2.70E-19 1.000
268.5 1.80E-19 1.000
279.7 9.00E-20 1.000
292.0 3.10E-20 1.000
303.0 1.10E-20 1.000
306.0 8.50E-21 1.000
309.0 6.60E-21 1.000
312.0 5.10E-21 1.000
316.0 3.30E-21 1.000
330.0 1.00E-21 1.000

590.0 4.74E-18 0.191
605.0 3.05E-18 0.272
620.0 4.61E-18 0.156
635.0 1.43E-18 0.039
WL Abs QY
(nra) (cm )

193-2 7.60E-18 1.000
199.0 4.80E-18 1.000
205-1 2.90E-18 1.000
211.6 1.70E-18 1.000
218.6 1.10E-18 1.000
226.0 8.20E-19 1.000
233.9 6.50E-19 1.000
242.4 5.00E-19 1.000
251.6 3.50E-19 1.000
261.4 2.35E-19 1.000
272.1 1.55E-19 1.000
283.7 6.50E-20 1.000
296.3 2.10E-20 1.000
304.0 1.00E-20 1.000
307.0 7.70E-21 1.000
310.0 6.00E-21 1.000
313.0 4.50E-21 1.000
320.0 2.40E-21 1.000
335.0 O.OOE-01 1.000

595.0 3.74E-18 0.350
610.0 1.51E-18 0.233
625.0 8.17E-18 0.117
640.0 1.08E-18 0.000
Photolysis Set = N03N02R
405.0 2.80E-20 1.000
420.0 8.20E-20 1.000
435.0 1.84E-19 1.000
450.0 2.83E-19 1.000
465.0 4.34E-19 1.000
480.0 6.44E-19 1.000
495.0 9.67E-19 1.000
510.0 1.32E-18 1.000
525.0 1.49E-18 1.000
540.0 1.83E-18 1.000
555.0 2.68E-18 1.000
570.0 2.49E-18 1.000
585.0 2.66E-18 0.968
600.0 2.69E-18 0.578
410.0 4.45E-20 1.000
425.0 1.05E-19 1.000
440.0 1.94E-19 1.000
455.0 3.35E-19 1.000
470.0 5.10E-19 1.000
485.0 6.86E-19 1.000
500.0 9.90E-19
515.0 1.40E-18
530.0 1.93E-18
545.0 1.82E-18
560.0 3.07E-18
575.0 2.61E-18
.000
.000
.000
.000
.000
.000
590.0 4.74E-18 0.809
605.0 3.05E-18 0.506
415.0 5.35E-20 1.000
430.0 1.29E-19 1.000
445.0 2.23E-19 1-000
460.0 3.72E-19 1.000
475.0 6.03E-19 1.000
490.0 8.80E-19 1.000
505.0 1.09E-18 1.000
520.0 1.45E-18 1.000
535.0 2.04E-18 1.000
550.0 2.35E-18 1.000
565.0 2.54E-18 1.000
580.0 2.91E-18 1.000
595.0 3.74E-18 0.650
610.0 1.51E-18 0.433
                                                              (continued)
                                    22

-------
Table 3 (continued) - 5
WL
(nm)
Abs
(cm2)
QY
WL
(nm)
Abs
(cm )
QY
WL
(nm)
Abs
(cm2)
QY
615.0 1.84E-18 0.361
630.0 4.89E-18 0.114

Photolysis Set = H202R
620.0 4.61E-18 0.289
635.0 1.43E-18 0.072
625.0 8.17E-18 0.217
640.0 1.08E-18 0.000
191.4
197.0
203.0
209.4
216.2
223.5
231.2
239.5
248.4
258.1
268.5
279.7
292.0
303.0
306.0
309.0
312.0
316.0
330.0
345.0
3.21E-19
5.27E-19
4.34E-19
3.56E-19
2.88E-19
2.26E-19
1.71E-19
1.25E-19
8.81E-20
5.80E-20
3.54E-20
2.02E-20
1.03E-20
5.59E-21
4.69E-21
4.00E-21
3.38E-21
2.65E-21
1.19E-21
5.03E-22
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
.000
.000
.000
.000
.000
.000
.000
.000
.000
.000
.000
.000
.000
.000
.000
.000
.000
.000
.000
.000
193.
199-
205.
211.
218.
226.
233.
242.
251-
261.
272.
283.
296.
304.
307.
310.
313.
320.
335.
350.
2 3
0 4
1 4
6 3
6 2
0 2
9 1
4 1
6 7
4 4
1 2
7 1
3 8
0 5
0 4
0 3
0 3
0 2
0 9
0 3
.OOE-19
.92E-19
.07E-19
.33E-19
.67E-19
.07E-19
.55E-19
. 12E-19
.76E-20
.92E-20
.96E-20
.61E-20
. 16E-21
.26E-21
.46E-21
.77E-21
. 19E-21
.07E-21
.36E-22
.25E-22
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
195
201
207
213
221
228
236
245
254
264
275
287
300
305
308
311
314
325
340
355
.1 2.
.0 4.
.3 3.
.9 3.
.0 2.
.6 1.
.7 1.
.4 9-
.8 6.
.9 4.
.9 2.
.8 1.
.5 6.
.0 H.
.0 4.
.0 3.
.0 2.
.0 1.
.0 6.
.0 0.
80E-19
61E-19
82E-19
10E-19
46E-19
89E-19
39E-19
93E-20
78E-20
21E-20
46E-20
29E-20
47E-21
93E-21
23E-21
58E-21
99E-21
53E-21
86E-22
OOE-01
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
 Photolysis Set = HCHOMR
251.6 3-91E-22 0.480
261.4 5.06E-21 0.498
272.1 1.16E-20 0.457
283.7 2.27E-20 0.325
296.3 2.99E-20 0.270
304.0 6.33E-20 0.247
307.0 2.04E-20 0.251
310.0 1.75E-20 0.260
313.0 1.25E-20 0.273
320.0 1.53E-20 0.394
335.0 2.13E-21 0.754
350.0 1.52E-21 0.338
365.0 0. OOE-01 0.005
254.8 1.18E-21 0.489
264.9 5.55E-21 0.495
275.9 1.60E-20 0.386
287.8 2.13E-20 0.306
300.5 1.52E-20 0.255
305.0 4.67E-20 0.246
308.0 1.41E-20 0.254
311.0 7.28E-21 0.264
314.0 3.92E-20 0.279
325.0 2.11E-20 0.508
340.0 1.63E-20 0.616
355.0 7.17E-21 0.198
258.1 2.81E-21 0.494
268.5 9.35E-21 0.484
279.7 1.58E-20 0.342
292.0 2.26E-20 0.287
303.0 2.28E-20 0.249
306.0 4.50E-20 0.248
309.0 2.96E-20 0.257
312.0 1.34E-20 0.269
316.0 3-95E-20 0.310
330.0 1.92E-20 0.676
345.0 6.67E-21 0.471
360.0 9.07E-23 0.085
                                                               (continued)
                                     23

-------
Table 3 (continued) - 6
WL
(nm)
Abs
(cm2)
QY
WL
(nm)
Abs
(cm2)
QY
WL
(nm)
Abs
(cm2)
QY
Photolysis Set  = HCHORR

251.6 3.91E-22  0.340
261.1 5.06E-21  0.319
272.1 1.16E-20  0.405
283.7 2.27E-20  0.606
296.3 2.99E-20  0.742
304.0 6.33E-20  0.754
307.0 2.04E-20  0.753
310.0 1.75E-20  0.749
313.0 1.25E-20  0.732
320.0 1.53E-20  0.593
335.0 2.13E-21  0.122

Photolysis Set  = ALDR

264.9 2.82E-20  0.340
275.9 4.01E-20  0.492
287.8 4.55E-20  0.531
300.5 3.90E-20  0.420
305.0 3.34E-20  0.351
308.0 3-11E-20  0.304
311.0 2.66E-20  0.255
314.0 2.25E-20  0.204
325.0 1.09E-20  0.050

Photolysis Set  = OPR

211.6 3.49E-19  1.000
218.6 2.42E-19  1.000
226.0 1.71E-19  1.000
233.9 1.19E-19  1.000
242.4 8.07E-20  1.000
251.6 5.48E-20  1.000
261.4 3.61E-20  1.000
272.1 2.29E-20  1.000
283.7 1.28E-20  1.000
296.3 6.98E-21  1.000
304.0 4.84E-21  1.000
307.0 4.12E-21  1.000
310.0 3.41E-21  1.000
313.0 2.95E-21  1.000
320.0 1.94E-21  1.000
335.0 8.50E-22  1.000
350.0 3.87E-22  1.000
254.8 1.18E-21 0.323
264.9 5.55E-21 0.333
275.9 1.60E-20 0.456
287.8 2.13E-20 0.680
300.5 1.52E-20 0.750
305.0 4.67E-20 0.755
308.0 1.41E-20 0.752
311.0 7.28E-21 0.745
314.0 3.92E-20 0.723
325.0 2.11E-20 0.458
340.0 1.63E-20 0.003
268.5 3.25E-20 0.369
279.7 4.34E-20 0.557
292.0 4.43E-20 0.502
303.0 3.54E-20 0.383
306.0 3.27E-20 0.335
309.0 2.97E-20 0.288
312.0 2.50E-20 0.238
316.0 2.05E-20 0.169
330.0 6.43E-21 0.006
213.9 3.14E-19 1.000
221.0 2.12E-19 1.000
228.6 1.50E-19 1.000
236.7 1.05E-19 1.000
245.4 7.18E-20 1.000
254.8 4.84E-20 1.000
264.9 3.16E-20 1.000
275.9 1.91E-20 1.000
287.8 1.03E-20 1.000
300.5 5.70E-21 1.000
305.0 4.60E-21 1.000
308.0 3.88E-21 1.000
311.0 3.25E-21 1.000
314.0 2.80E-21 1.000
325.0 1.50E-21 1.000
340.0 6.19E-22 1.000
355.0 2.00E-22 1.000
258.1 2.81E-21 0.317
268.5 9.35E-21 0.365
279.7 1.58E-20 0.522
292.0 2.26E-20 0.724
303.0 2.28E-20 0.753
306.0 4.50E-20 0.754
309.0 2.96E-20 0.751
312.0 1.34E-20 0.740
316.0 3.95E-20 0.690
330.0 1.92E-20 0.305
345.0 6.67E-21 0.000
272.1 3.68E-20 0.421
283.7 4.50E-20 0.552
296.3 4.22E-20 0.463
304.0 3.41E-20 0.367
307.0 3.19E-20 0.319
310.0 2.81E-20 0.272
313.0 2.36E-20 0.221
320.0 1.57E-20 0.106
335.0 2.79E-21 0.000
216.2 2.79E-19 1.000
223.5 1.92E-19 1.000
231.2 1.33E-19 1.000
239-5 9.07E-20 1.000
248.4 6.26E-20 1.000
258.1 4.19E-20 1.000
268.5 2.70E-20 1.000
279.7 1.54E-20 1.000
292.0 8.36E-21 1.000
303.0 5.08E-21 1.000
306.0 4.36E-21 1.000
309.0 3.64E-21 1.000
312.0 3.10E-21 1.000
316.0 2.50E-21 1.000
330.0 1.12E-21 1.000
345.0 5.00E-22 1.000
360.0 O.OOE-01 1.000
                                                              (continued)
                                    24

-------
Table 3 (continued) - 7
WL Abs QY
(nm) (cm)
Photolysis Set = PAAR
191.4 8.98E-20 1.000
197.0 1.48E-19 1-000
203.0 1.21E-19 1.000
209-4 9.98E-20 1.000
216.2 8.07E-20 1.000
223.5 6.33E-20 1.000
231.2 4.80E-20 1.000
239.5 3.49E-20 1.000
248.4 2.47E-20 1.000
258.1 1.62E-20 1.000
268.5 9.91E-21 1.000
279-7 5.66E-21 1.000
292.0 2.88E-21 1.000
303.0 1.57E-21 1.000
306.0 1.31E-21 1.000
309.0 1.12E-21 1.000
312.0 9.47E-22 1.000
316.0 7.43E-22 1.000
330.0 3.32E-22 1.000
345.0 1.U1E-22 1.000
Photolysis Set = KETR
279.7 6.00E-20 0.077
292.0 5.10E-20 0.077
303.0 3.00E-20 0.077
306.0 2.30E-20 0.077
309.0 1.70E-20 0.077
312.0 1.20E-20 0.077
316.0 7.80E-21 0.077
330.0 O.OOE-01 0.077
Photolysis Set = GLY1R
233.9 3.40E-21 1.000
242.4 5.73E-21 1.000
251.6 9.05E-21 1.000
261.4 1.62E-20 1.000
272.1 2.48E-20 1.000
283.7 3.18E-20 1.000
296.3 3.37E-20 1.000
304.0 2.90E-20 1.000
307.0 2.72E-20 1.000
WL Abs QY
( nm ) ( cm )

193.2 8.41E-20 1.000
199.0 1.38E-19 1.000
205.1 1.14E-19 1.000
211.6 9.32E-20 1.000
218.6 7.48E-20 1.000
226.0 5.79E-20 1.000
233.9 4.33E-20 1.000
242.4 3.13E-20 1.000
251.6 2.17E-20 1.000
261.4 1.38E-20 1.000
272.1 8.29E-21 1.000
283.7 4.52E-21 1.000
296.3 2.29E-21 1.000
304.0 1.47E-21 1.000
307.0 1.25E-21 1.000
310.0 1.06E-21 1.000
313.0 8.93E-22 1.000
320.0 5.81E-22 1.000
335.0 2.62E-22 1.000
350.0 9.10E-23 1-000

283.7 5.80E-20 0.077
296.3 4.10E-20 0.077
304.0 2.80E-20 0.077
307.0 2.10E-20 0.077
310.0 1.50E-20 0.077
313.0 1.10E-20 0.077
320.0 3.90E-21 0.077


236.7 4.01E-21 1.000
245.4 6.62E-21 1.000
254.8 1.12E-20 1.000
264.9 2.02E-20 1.000
275.9 2.76E-20 1.000
287.8 3.22E-20 1.000
300.5 3.43E-20 1.000
305.0 2.75E-20 1.000
308.0 2.72E-20 1.000
WL Abs QY
(nm) (cm2)

195.1 7.84E-20 1.000
201.0 1.29E-19 1-000
207.3 1.07E-19 1-000
213.9 8.68E-20 1.000
221.0 6.90E-20 1.000
228.6 5.29E-20 1.000
236.7 3.90E-20 1.000
245.4 2.78E-20 1.000
254.8 1.90E-20 1.000
264.9 1.18E-20 1.000
275.9 6.90E-21 1.000
287.8 3.61E-21 1.000
300.5 1.81E-21 1.000
305.0 1.38E-21 1.000
308.0 1.18E-21 1.000
311.0 1.00E-21 1.000
314.0 8.38E-22 1.000
325.0 4.28E-22 1.000
340.0 1.92E-22 1.000
355.0 O.OOE-01 1.000

287.8 5.50E-20 0.077
300.5 3.20E-20 0.077
305.0 2.50E-20 0.077
308.0 1.90E-20 0.077
311.0 1.40E-20 0.077
314.0 1.00E-20 0.077
325.0 1.80E-21 0.077


239.5 4.86E-21 1.000
248.4 7.54E-21 1.000
258.1 1.33E-20 1.000
268.5 2.24E-20 1.000
279.7 2.89E-20 1.000
292.0 3.21E-20 1.000
303.0 3.07E-20 1.000
306.0 2.72E-20 1.000
309.0 2.72E-20 1.000
                                                               (continued)
                                      25

-------
Table 3 (continued) - 8
WL Abs QY
(nm) (cnr)
310.0 2.73E-20 1.000
313.0 2.68E-20 1.000
320.0 1.51E-20 1.000
335.0 2.29E-20 1.000
Photolysis Set = GLY2R
355.0 2.86E-22 0.024
370.0 7.64E-21 0.024
385.0 1.66E-20 0.024
400.0 3-36E-20 0.024
415.0 6.43E-20 0.024
430.0 6.77E-20 0.024
445.0 7.15E-20 0.024
460.0 O.OOE-01 0.024
Photolysis Set = ONITR
264.9 2.91E-20 1.000
275.9 2.01E-20 1.000
287.8 9-92E-21 1.000
300.5 4.07E-21 1.000
305.0 2.68E-21 1.000
308.0 1.99E-21 1.000
311.0 1.38E-21 1.000
314.0 9.14E-22 1.000
325.0 1.90E-22 1.000
WL Abs QY
( nm ) ( cm )
311.0 2.79E-20 1.000
314.0 2.49E-20 1.000
325.0 1.27E-20 1.000
340.0 3.58E-21 1.000

360.0 2.08E-21 0.024
375.0 1.07E-20 0.024
390.0 3.03E-20 0.024
405.0 3.66E-20 0.024
420.0 5.46E-20 0.024
435.0 5.99E-20 0.024
450.0 7.30E-20 0.024


268.5 2.66E-20 1.000
279.7 1.64E-20 1.000
292.0 7.49E-21 1.000
303.0 3.28E-21 1.000
306.0 2.44E-21 1.000
309.0 1.76E-21 1.000
312.0 1.22E-21 1.000
316.0 6.87E-22 1.000
330.0 O.OOE-01 1.000
WL Abs QY
(nm) (cm )
312.0 2.83E-20 1.000
316.0 2.12E-20 1.000
330.0 1.42E-20 1.000
345.0 O.OOE-01 1.000

365.0 3.44E-21 0.024
380.0 1.59E-20 0.024
395.0 2.63E-20 0.024
410.0 4.56E-20 0.024
425.0 9.22E-20 0.024
440.0 1.17E-19 0.024
455.0 2.01E-19 0.024


272.1 2.37E-20 1.000
283.7 1.29E-20 1.000
296.3 5.62E-21 1.000
304.0 2.98E-21 1.000
307.0 2.21E-21 1.000
310.0 1.54E-21 1.000
313.0 1.07E-21 1.000
320.0 4.04E-22 1.000

Data Used  in the March 1988 Version of the RADM Mechanism,  but Not
in the Recommended Versions

Photolysis Set = MGLYR
233.9 1.23E-20 1.000
242.4 1.70E-20 1.000
251.6 1.80E-20 1.000
261.4 1.99E-20 1.000
272.1 2.61E-20 1.000
283.7 2.93E-20 1.000
296.3 2.32E-20 1.000
304.0 1.84E-20 1.000
307.0 1.55E-20 1.000
310.0 1.26E-20 1.000
313.0 1.19E-20 1.000
320.0 6.44E-21 1.000
236.7 1.47E-20 1.000
245.4 1.67E-20 1.000
254.8 1.81E-20 1.000
264.9 2.12E-20 1.000
275.9 2.79E-20 1.000
287.8 2.93E-20 1.000
300.5 2.14E-20 1.000
305.0 1.75E-20 1.000
308.0 1.45E-20 1.000
311.0 1.23E-20 1.000
314.0 1.17E-20 1.000
325.0 4.00E-21 1.000
239.5 1.69E-20
248.4 1.75E-20
258.1 1.89E-20
268.5 2.38E-20
279.7 2.96E-20
292.0 2.71E-20
303.0 1.93E-20
306.0 1.65E-20
309.0 1.35E-20
312.0 1.21E-20
316.0 1.04E-20
330.0 2.00E-21
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
                                                              (continued)
                                    26

-------
Table 3 (continued) - 9
WL
(nm)
Abs
(cm2)
QY
WL
(nm)
Abs
(cm2)
QY
WL
(nm)
Abs
(cm2)
QY
335.0 2.50E-22 1.000    340.0 O.OOE-01 1.000    355.0 7.23E-22 0.153
360.0 2.46E-21 0.153    365.0 4.00E-21 0.153    370.0 5.55E-21 0.153
375.0 8.00E-21 0.153    380.0 1.08E-20 0.153    385.0 1.51E-20 0.153
390.0 1.99E-20 0.153    395.0 2.40E-20 0.153    400.0 2.87E-20 0.153
405.0 3.75E-20 0.153    410.0 4.42E-20 0.153    415.0 4.82E-20 0.153
420.0 4.96E-20 0.153    425.0 4.97E-20 0.153    430.0 5.18E-20 0.153
435.0 5.87E-20 0.153    440.0 6.01E-20 0.153    445.0 6.06E-20 0.153
450.0 5.40E-20 0.153    455.0 2.97E-20 0.153    460.0 6.96E-21 0.153
465.0 2.00E-21 0.153    470.0 O.OOE-01 0.153

Photolysis Set = DCBR

186.0 2.59E-21 1.000    187.8 2.72E-21 1.000    189.6 2.85E-21 1.000
191.4 2.73E-21 1.000    193-2 2.51E-21 1.000    195.1 2.44E-21 1.000
197.0 2.46E-21 1.000    199.0 2.46E-21 1.000    201.0 2.82E-21 1.000
203.0 2.92E-21 1.000    205.1 3.53E-21 1.000    207.3 3.74E-21 1.000
209.4 4.10E-21 1.000    211.6 4.36E-21 1.000    213.9 4.45E-21 1.000
216.2 4.60E-21 1.000    218.6 4.17E-21 1.000    221.0 4.54E-21 1.000
223.5 3.70E-21 1.000    226.0 3.89E-21 1.000    228.6 2.69E-21 1.000
231.2 2.76E-21 1.000    233-9 1.71E-21 1.000    236.7 1.62E-21 1.000
239-5 7.85E-22 1.000    242.4 6.12E-22 1.000    245.4 2.88E-22 1.000
248.4 2.13E-22 1.000    251.6 1.40E-22 1.000    254.8 1.49E-22 1.000
258.1  1.60E-22 1.000    261.4 1.79E-22 1.000    264.9 2.18E-22 1.000
268.5 2.72E-22 1.000    272.1 3.27E-22 1.000    275.9 4.10E-22 1.000
279.7 5.26E-22 1.000    283.7 6.27E-22 1.000    287.8 7.62E-22 1.000
292.0 9.39E-22 1.000    296.3 1.09E-21 1.000    300.5 1.26E-21 1.000
303.0  1.51E-21 1.000    304.0 1.56E-21 1.000    305.0 1.55E-21  1.000
306.0  1.53E-21 1.000    307.0 1.61E-21 1.000    308.0 1.65E-21  1.000
 309.0  1.81E-21  1.000    310.0 1.83E-21 1.000    311.0 1.91E-21  1.000
 312.0 2.03E-21  1.000    313.0 2.01E-21 1.000    314.0 1.99E-21  1.000
 316.0 2.20E-21  1.000    320.0 2.50E-21 1.000    325.0 2.81E-21  1.000
 330.0 3.04E-21  1.000    335.0 3.40E-21 1.000    340.0 3.84E-21  1.000
 345.0 4.05E-21  1.000    350.0 4.47E-21 1.000    355.0 4.88E-21  1.000
 360.0 4.94E-21  1.000    365.0 5.29E-21 1.000    370.0 5.40E-21  1.000
 375.0  5.63E-21  1.000    380.0 5.61E-21  1.000    385.0 5.72E-21 0.997
 390.0  6.00E-21 0.981    395.0 5.69E-21 0.922    400.0 6.18E-21 0.693
 405.0  5.49E-21 0.356    410.0 5.98E-21 0.135    415.0 5.77E-21 0.060
 420.0  O.OOE-01  0.018

 Data Used in the Recommended Versions of the  RADM Mechanism, but Not
 in the  March 1988 Version

 Photolysis Set = MGLYRM

 231.2  O.OOE-01 1.000    233-9 1.23E-20 1.000     236.7  1.47E-20 1.000
 239.5  1.69E-20 1.000    242.4 1.70E-20 1.000     245.4  1.67E-20 1.000
                                                               (continued)
                                     27

-------
Table 3 (continued) - 10
WL
(nm)
Abs
(cm2)
QY
WL
(nm)
Abs
(cm2)
QY
WL
(nm)
Abs
(cm2)
QY
248.4  1.75E-20  1.000    251.6 1.80E-20 1.000    254.8 1.81E-20 1.000
258.1  1.89E-20  1.000    261.4 1.99E-20 1.000    264.9 2.12E-20 1.000
268.5  2.38E-20  1.000    272.1 2.61E-20 1.000    275.9 2.79E-20 1.000
279.7  2.96E-20  1.000    283-7 2.93E-20 1.000    287.8 2.93E-20 1.000
292.0  2.71E-20  1.000    296.3 2.32E-20 1.000    300.5 2.14E-20 1.000
303.0  1.93E-20  1.000    304.0 1.84E-20 1.000    305.0 1.75E-20 1.000
306.0  1.65E-20  1.000    307.0 1.55E-20 1.000    308.0 1.45E-20 1.000
309.0  1.35E-20  1.000    310.0 1.26E-20 1.000    311.0 1.23E-20 1.000
312.0  1.21E-20  1.000    313.0 1.19E-20 1.000    314.0 1.17E-20 1.000
316.0  1.04E-20  1.000    320.0 6.44E-21 1.000    325.0 4.00E-21 1.000
330.0  2.00E-21  1.000    335.0 2.50E-22 1.000    340.0 O.OOE-01 1.000
345.0  O.OOE-01  0.103    350.0 O.OOE-01 0.103    355.0 7.23E-22 0.103
360.0  2.46E-21  0.103    365.0 4.00E-21 0.103    370.0 5.55E-21 0.103
375.0  8.00E-21  0.103    380.0 1.08E-20 0.103    385.0 1.51E-20 0.103
390.0  1.99E-20  0.103    395.0 2.40E-20 0.103    400.0 2.87E-20 0.103
405.0  3.75E-20  0.103    410.0 4.42E-20 0.103    415.0 4.82E-20 0.103
420.0  4.96E-20  0.103    425.0 4.97E-20 0.103    430.0 5.18E-20 0.103
435.0  5.87E-20  0.103    440.0 6.01E-20 0.103    445.0 6.06E-20 0.103
450.0  5.40E-20  0.103    455.0 2.97E-20 0.103    460.0 6.96E-21 0.103
465.0  2.00E-21  0.103    470.0 O.OOE-01 0.103

Photolysis Set  = DCBRM

 186.0  7.90E-20  1.000     187.8 7.90E-20 1.000    189.6 7.90E-20 1.000
 191.4  7.90E-20  1.000     193.2 7.90E-20 1.000    195.1 7.90E-20 1.000
 197.0  7.90E-20  1.000     199-0 7.90E-20 1.000    201.0 7.90E-20 1.000
203.0  7.90E-20  1.000    205.1 7.90E-20 1.000    207.3 7.90E-20 1.000
209.4  7.90E-20  1.000    211.6 7.90E-20 1.000    213-9 7.90E-20 1.000
216.2  7.90E-20  1.000    218.6 7.90E-20 1.000    221.0 7.90E-20 1.000
223-5  7.90E-20  1.000    226.0 7.90E-20 1.000    228.6 7.90E-20 1.000
231.2  7.90E-20  1.000    233.9 7.90E-20 1.000    236.7 7.90E-20 1.000
239.5  7.90E-20  1.000    242.4 7.90E-20 1.000    245.4 7.90E-20 1.000
248.4  7.90E-20  1.000     251.6 7.90E-20 1.000    254.8 7.90E-20 1.000
258.1  7.90E-20  1.000     261.4 7.90E-20 1.000    264.9 7.90E-20 1.000
268.5  7.90E-20  1.000     272.1 7.90E-20 1.000    275.9 7.90E-20 1.000
279.7  7.90E-20  1.000     283-7 7.90E-20 1.000    287.8 7.90E-20 1.000
292.0  7-90E-20  1.000     296.3 7.90E-20 1.000    300.5 7-90E-20 1.000
 303.0  7.90E-20 1.000     304.0 7.90E-20 1.000    305.0 7.90E-20 1.000
 306.0  7.90E-20 1.000     307.0 7.90E-20 1.000    308.0 7.90E-20 1.000
 309-0  7.90E-20 1.000     310.0 7.90E-20  1.000    311-0 7.90E-20 1.000
 312.0  7.90E-20 1.000     313.0 7.90E-20  1.000    314.0 7.90E-20 1.000
 316.0  7.90E-20 1.000     320.0 7.90E-20  1.000    325.0 7.90E-20 1.000
 330.0  7.90E-20 1.000     335.0 7.90E-20  1.000    340.0 7.90E-20 1.000
 345.0 7.90E-20 1.000     350.0 7.90E-20  1.000    355.0 7.90E-20 0.500
 360.0 7.90E-20 0.000     365.0 O.OOE-01 0.000
                                     28

-------
2.2  Recommended Modifications to the RADM Mechanism
     The results of this evaluation study indicated that for the most  part
the  March  1988  version  of  the  RADM  gas-phase  chemical mechanism  is
chemically reasonable and  performs as well  as  can reasonably be  expected
in simulating  most of  the  available chamber  data.    However,  there  were
aspects of the mechanism which we  recommended be  modified  prior  to its use
in the  RADM-II  model.   The results  of  the chamber simulations, discussed
in Section 5, show that the performance of the  RADM gas-phase mechanism in
fitting results  of aromatic-NOx-air chamber  experiments  was not  quite as
good  as should be expected  given  the  performance of  other current  gas-
phase chemical mechanisms.  The  test calculations described in  Section 6.3
show  that the  representation of  the higher  alkanes  could be  condensed
without  significantly  affecting  model predictions.    In  addition,  there
were  several other areas where  it  could be argued that minor modifications
to  the  current mechanism might  improve  its  chemical  reasonableness.   The
recommended  modifications  to  the  RADM  gas-phase  chemical mechanism are
described and justified below.
      2.2.1  Recommended Modifications to the Aromatic  Mechanism
            There  are  major gaps  in  our  understanding of the  atmospheric
reactions  of  aromatics, and  the  available laboratory data  are  not suffi-
cient to  serve as a basis  for  deriving  models for the  reactions  of these
compounds.   Thus, we  have no  choice  but to use simulations  of  environ-
mental  chamber experiments as  essentially  the only means to  Judge other-
wise chemically reasonable alternative assumptions concerning  the unknown
portions  of the aromatics mechanisms.   The  performance of the  RADM mech-
anism in  simulating  results of toluene-NOx-air and xylene-NOx-air environ-
mental  chamber runs,  discussed in Section 5, indicated that there was  room
for  improvement in this mechanism.  The  RADM mechanism was found to  have
considerably  more scatter  in  the simulations of maximum  ozone yields and
rates of  NO  oxidation  and  0^  formation than   the  SAPRC mechanism.   By
comparison,  Jeffries (1989b) found  that  the Carbon  Bond-IV mechanism  — a
more condensed  mechanism  than RADM —  performed comparably to the  SAPRC
mechanism  in   simulating  the  aromatic  runs,  being  somewhat  better in
simulating the  UNC toluene  runs,  and  somewhat worse in  simulating UNC
 xylene runs.    Thus  by  the criterion  of  performance in  simulating the
 chamber  data,  the  model   used  in  the  RADM mechanism   to represent the
                                     29

-------
unknown  portions  of the toluene and  xylene mechanisms —  which is based
largely  on  an  earlier  published aromatics mechanism  (Lurmann et  al., 1986)
— must  be  judged  to be  less  reliable than  the models used in the  current
SAPRC  or Carbon Bond  mechanisms.   Since  the mechanisms  for  toluene and
xylene  serve  as  the basis for  the  representation  of all  the aromatics
hydrocarbons,  it  is clearly important  that  the RADM-II model incorporate
mechanisms  for these compounds that are as reliable  as  possible.  For  this
reason,  a  major  recommendation  resulting  from  this  study was that the
aromatics  photooxidation mechanism  be modified to improve  its performance
in simulations of  the toluene and xylene chamber data.
     To  address this concern, we investigated how the  RADM  mechanism might
be modified to improve its performance in simulating the results of these
chamber  experiments.   The  modifications examined  involved making the  RADM
aromatics  mechanism more like the  aromatics mechanism used in the  current
SAPRC  mechanism (Carter  et al.,  1988).  This  approach  was  employed  because
(1)  independent development  of  an entirely   new  aromatics mechanism was
beyond the scope  of this project;  (2) we  are  familiar  with the performance
and  implementation of the  SAPRC  aromatics model and know how to adjust  it
to  optimize  its  ability  to  fit the  chamber data; and  (3)  although  the
aromatics   model  in  Carbon  Bond  IV  also  performs as  satisfactorily  in
simulating chamber data, its use of lumped structure groups makes  it  less
straightforward to  implement  in   a   lumped   molecule  mechanism  such  as
RADM.    In addition, we are less familiar with how to  optimize the  ability
of  the  Carbon Bond mechanism to  fit chamber data.   However,   it  must  be
recognized that the SAPRC aromatics mechanism uses  a  highly  idealized and
simplified representation  of  the  unknown  aspects of  aromatic  reactions,
and  no claim  is made that it is  necessarily any more chemically  reasonable
than the  aromatics representation used  in  any  other current  mechanism,
including  that in  the RADM mechanism  itself.
     Two levels of modifications  of  the aromatic mechanism  were  examined
in  the  course of  this  study:   one   in  which  the  only  changes  involved
parameter   values,  and the other where more   extensive changes  were  made.
The  latter  is the  version  of  the modified  aromatics mechanism which  is
designated as  RADM-M and  is listed in Section 2.3-  The former is referred
to  as  the  "modified parameter"  aromatics mechanism, which was recommended
for  implementation if more  extensive changes  in the mechanism were not
                                     30

-------
feasible.   It  is designated  as  the  RADM-P mechanism  and  is  listed  in
Section  2.4.   In  the  subsequent  discussion,  the modifications  discussed
for toluene refer  to the  reactions of the RADM model species TOL, and  the
modifications  discussed  for  xylenes  refer  to  reactions  of  the   model
species  XYL.    In  both  cases,  the  following  parameter  changes  were
recommended:
     (1)  It was  recommended that the yields of methyl glyoxal  (MGLY)  and
(for the modified  parameter  mechanism)  glyoxal  (GLY)  be  reduced  to be more
consistent with experimental yields of these products from  toluene  and m-
xylene  (Atkinson,  1989,  and references  therein).   In addition,  the yield
of cresols  (CSL)  from  toluene  was increased  slightly  to  be more  consistent
with  the results  of Gery  et al.  (1985)  and  Atkinson et  al.  (1989); this
also  resulted  in  somewhat  improved fits  of  model simulations  of maximum
ozone  yields  in   toluene-NOx-air  experiments.  The  specific  changes  in
yields  of these products are as  follows:

                                    RADM           RADM-M and  RADM-P
   MGLY yield from TOLP             0.72*4               0.17
   MGLY yield from XYLP             1.0                 0.45
   GLY  yield form  TOLP              0.181               0.16
   CSL  yield form  TOL               0.16                0.25

      (2) Although the   available  documentation  of  the  derivation  of
 photolysis  rates  to be used in the RADM mechanism (NCAR,  1987) states  that
 the  methylglyoxal quantum yields in the higher wavelength  band are based
 on the  data of Plum et al. (1983), the RADM mechanism actually uses  higher
 quantum yield  than indicated  by  those  data.   In particular,  as shown in
 Table 3 (see  photolysis file  "MGLYR"),  the RADM mechanism uses a wave-
 length-independent quantum  yield of 0.153  for wavelengths  above 350 run,
 while  the  data  of  Plum  et al.  (1983)  indicate  that  the  quantum yield
 should  be  0.107.   We  therefore recommended  that the methylglyoxal  quantum
 yields  for  these wavelengths be  decreased to the  lower value which  is  more
 consistent  with  the  data  of Plum  et  al. (1983).   The  modified sets of
 quantum yields  recommended  for  methylglyoxal  are listed in Table  3
 (photolysis file  "MGLYRM").
      (3) The  RADM mechanism uses the model species DCB  to represent the
 uncharacterized  aromatic  ring  fragmentation  products.    The   photolysis

                                     31

-------
rates of  these compounds, and  the  wavelengths where  they  photolyze,  are
unknown.   In  the RADM mechanism, the  species DCB  photolyzes  at the same
wavelengths as NC^, but at a rate which  is  10 times slower.  The ratio to
NCs  is  presumably  based on  results  of  simulations  of  chamber  runs.
However,  in  the development  of the  SAPRC  mechanism,  it  was  found that
better performance in simulating the chamber data,  especially runs carried
out using  a blacklight  light  source,  could  be obtained  if  it was assumed
that  this  species photolyzes  more  rapidly,  and photolyzes  primarily at
much  lower wavelengths  than  those responsible for  the photolysis of NO^.
Therefore, we  recommended that the absorption coefficient, quantum yield
products used  in the SAPRC mechanism  for the unknown aromatic  fragmenta-
tion  species  be used  in  the  RADM  mechanism for  DCB.   This  consists of
assuming the absorption coefficient x  quantum yield product is constant at
       ?o   ?           i
7.9x10     cm   molecule  , base e,  at wavelengths of  350 nm or below, and
decreases  linearly to zero at  360 nm.   This is essentially arbitrary, but
allows  good   fits  to  aromatic-NO -air  experiments  carried  out  using   a
                                  A
variety of light sources. The modified absorption coefficient x quantum
yield products are listed in Table 3 (photolysis set "DCBRM").
      (H)  The  yields of  the  DCB species  from  toluene and  xylene  were
adjusted to optimize  the  fits  of model simulations to results of  selected
SAPRC  toluene-NOx-air  and   m-xylene-NOx-air  experiments    The  approach
employed,  and  the  specific  experiments whose results  were  utilized  in the
optimizations  is  the  same  as those  used  to derive the  uncharacterized
aromatic  fragmentation  product yields  in  the  current SAPRC  mechanism
(Carter,  1988).   The  best  fit  DCB  yields  depended  on  the other  changes
made  to  the  aromatic mechanism,  which  differed  depending  on  whether
changes other  than just modifying parameter  values were made.   Thus, the
optimum  DCB  yields  for  the  modified  parameter   (RADM-P)  mechanism are
slightly different from those  for the  recommended  (RADM-M)  mechanism where
other changes  were made.  These are indicated below:

                             RADM           RADM-P            RADM-M
    DCB yield  from TOLP      0.905           0.70               0.68
    DCB yield  from XYLP      1.0             0.806              0.79
                                     32

-------
     (5)  The  RADM  mechanism uses  an OH  + XYL  rate constant  which was
derived  based  on  an  analysis  of  total U.  S.  emissions  data  which  we
carried  out  during  the  initial  review  of the  1987 version  of the  RADM
mechanism.  However, since  the  time this initial  analyses was  carried  out,
the emissions assignments and rate constant estimates have  been  updated  as
a  result  of  the work  carried  out for the  California ARB  (Carter,  1988),
and  there have  also been  some changes to  methods  used   for aggregating
emissions input  into RADM.   (See  Section 3 of  this  report for  a  discussion
of  our analyses  for aggregating  emissions  into RADM.)   Therefore,  an up-
dated  analysis  of the emissions  data was  carried  out  for  the  purpose  of
developing our  final recommendations  in this  regard,  and  the results are
documented  in Section 3.3.    This  updated  analysis resulted  in  slight
changes  in  the recommended  OH  radical rate  constants  for  XYL and  some  of
the other lumped species.   In  the recommended  mechanisms, the A   factor
for  the  OH +  XYL  reaction was changed from  2.1x10~11  to 1.89x10"11 cm^
molecule"  sec,  to  correspond  better to  the revised emissions analysis.
This  value  was  used  in  calculating  the  optimum  DCB  yields  from XYLP,
discussed above.
      The modified  product yields recommended  for toluene and  xylene given
above actually refer to the yields of these products formed  in the reac-
tion  of their corresponding peroxy radicals  (TOLP  and  XYLP,  respectively)
with  NO.  The RADM mechanism also has these  radical species  reacting with
H02,   methyl   peroxy  (M02),  and acetyl  peroxy  (AC03)   radicals.    The
recommendations  do  not affect the species  representing the products formed
 in the  HOp  reactions  (which  is the  lumped higher  hydroperoxide   species
OP2), but do  affect the  species formed in  the  other two  reactions.   For
 these reactions,  the recommended  product  yields  were  derived based  on
assuming that the set of organic products  formed  from TOLP  and  XYLP are
 the same as those formed  in the NO  reaction,  with the  radical  and M02 or
 AC03 product  yields being  determined based  on assuming  these reactions
occur 50^ via RO + RO + 02 formation, and 50% via  ROH formation.  This may
 be somewhat  arbitrary,  but represents a  systematic and  self-consistent
method for deriving product yields for  the  many peroxy radical species in
 this mechanism  and,  in view of  our  lack  of  knowledge  of  what  the actual
 reactions and  species are,  appears   to  be as  good as  any   other  set of
 assumptions.
                                     33

-------
     The  above  are the modifications recommended for the aromatic mechan-
ism  that  involve only  changes  of parameter  values,   and  thus  are  the
modifications   implemented   in   the  RADM-P  version  of  the  mechanism.
However,  if  more  extensive modifications  could  be made,  then  it  was
recommended  that  the  representation  of  the  aromatic fragmentation  products
be  simplified.    Chamber simulations carried  out  in  this  study  indicate
that  a  simplified DCB  mechanism,  where  the  species  TPAN and  TC03  are
replaced  by  PAN  and AC03,  respectively,  and  where GLY  is  eliminated,
performs  equally  well, or even slightly better,  in  simulating  the  results
of  the chamber experiments, yet requires three fewer species.  In  view of
the  fact that the more complex "DCB" mechanism used  in  the evaluated RADM
mechanism is based  on theories of  aromatic  reactions  (e.g.,   Atkinson et
al.  1980) which are now known to be incorrect  or at best incomplete (e.g.,
Atkinson, 1988 and  references therein),  that mechanism has  no claim to
greater  chemical  reasonableness  than   the  more  simplified representation
that  is  recommended.   The  recommended  simplifications are implemented in
the  RADM-M  mechanism, and  the  specific  changes  involved  are  discussed
below.    (All these simplifications were implemented  in  the RADM-M mechan-
ism  before the  optimum DCB  yields were  calculated.)
      (1)   Glyoxal (GLY)  makes a relatively small contribution  to  the reac-
tivity of  aromatics  compared  to the  much more reactive products  methyl
glyoxal  (MGLY)  and DCB.  Thus, it is recommended that GLY  be  removed from
the  RADM mechanism, where it is formed  only from TOL, and its  contribution
to  reactivity  be represented  by  having it lumped  with  the other  aromatic
fragmentation products which are represented  by DCB.  (The  optimization of
the  yield of DCB from TOL  in a mechanism where  GLY is  removed will auto-
matically  take  its   contribution  to  reactivity  into  account).    Since
glyoxal  in  itself is not  (to our  knowledge)  of  special  significance to
acid deposition,  there is no reason to represent it explicitly in regional
acid deposition  models  if  its  contribution to reactivity can  be satisfac-
torily represented in other  ways.   If  it  is  important,  its formation  from
xylenes  and other aromatics  should  also be included.   We assume that  this
is  not the case.
      (2)  The  products  of  the reactions of DCB  are  assumed  to be similar
to  those for  the reactions  of methyl  glyoxal,  and  are represented as
follows:

-------
      DCBX * HV = H02 + CO * AC03
      DCBX + HO = AC03

(DCB is  renamed  DCBX in  the  RADM-M mechanism  to  indicate that  it  reacts
differently.)  The photolysis  reaction occurs with  the  same  rate  and spec-
tral response  as used in the  modified parameter  mechanism,  which in  turn
is  the  same  as   that  used in the SAPRC  mechanism  for  unknown  aromatic
fragmentation  products  (as  discussed above).   The  OH radical  reaction  is
assumed  to  have  the same  rate constant  as  that  for methyl  glyoxal.   The
NO? radical  reaction  is unlikely to be a  significant loss route  for  this
species  (or  for  NOo), and  is  ignored.   Although this  simpler  representa-
tion of  DCB reactions  is not  necessarily  any  more  chemically reasonable
than that used in RADM or RADM-P,  it has  the significant  advantage that it
results  in  the removal of TC03  and TPAN  from the  mechanism since the DCB
reactions  are  the  only  source  for these  species   in  the evaluated  RADM
mechanism.
     The  removal of  TPAN  from the mechanism  means that PAN  is used  to
represent all  types  of PAN analogues  that might be  formed in the reactions
of  unidentified  aromatic fragmentation products.   This is not  inconsistent
with the general concept  of  the level of  detail  intended  for this mech-
anism, since PAN is not  represented explicitly in  any case.   In the RADM
mechanism,  the model  species  PAN represnts not only PAN,  but PPN and other
higher   PAN  analogues   formed   in  non-aromatic  systems.   Thus,   in  the
recommended  modification  for  RADM-M,  the  higher  PAN analogues  (if any)
formed  from unknown aromatic  fragmentation products  are in effect lumped
with the other higher PAN analogues.  We see no  a-priori reason why it is
necessary to  have separate  representations of  the higher  PAN analogues
formed  from aroraatics  when  this   is not the  case for  those  formed from
alkanes,  alkenes, and other species.
     Comparisons of  the performance of  the RADM,  the modified parameter
RADM-P  and  the   recommended  RADM-M mechanisms  in   simulating  the  maximum
ozone  yields  and  rates  of   NO   oxidation  and  ozone  formation  in  the
aromatic-NO-air experiments  is shown in Figure 1.   (See Section 4 for a
            A
discussion of how the  simulations  of the environmental chamber experiments
were  carried out, and  see  Section  5 for a more extensive discussion of the
 performance  of  the  RADM  and  the RADM-M mechanisms in simulating these
                                     35

-------
                    RADM
  RADM-P
         RADM-M
                             Fits to Maximum Ozone
                        [Calculated - Experimental
                (ppm)]
Model
<
0.33 -
•0.27 -
•0.21 -
•0.15 -
•0.09 -
•0.03 -
0.03 -
0.09 -
0.15 -
0.21 -
0.27 -
>
Model
Low
-0.33
-0.27
-0.21
-0.15
-0.09
-0.03
0.03
0.09
0.15
0.21
0.27
0.33
0.33
High
                   :5
                   :2
                    : 14446
                    :333
                    :3
                    :3
 :24
 : 111111111446
 :116
 : 133466
 :33
 :3
        :1246
       •:11111
        :11366
        :13334
                          Fits to Average d([Oo]-[NO])/dt
                     [(Calculated - Experimental)/Calculated]
    Model Slow
<
-1.00 -
-0.82 -
-0.64 -
-0.45 -
-0.27 -
-0.09 -
0.09 -
0.27 -
0.45 -
0.64 -
0.82 -
>
Model
-1.00
-0.82
-0.64
-0.45
-0.27
-0.09
0.09
0.27
0.45
0.64
0.82
1.00
1.00
Fast
; ;
J J
* J
:125 :
• * 1 1
* • I I
:1144 :133
--: 13336 -: 112334456
:1 1111 13446 :1 11 11366
:1366 :1446
: :1
:1 :
• • 1
* 1 *





11
133
111112334456
11366
1446
1

1


Figure  1.  Comparison of Fits of RADM and Modified RADM Simulations
           to Results of Toluene and Xylene-N0v-Air Chamber Experiments.
                                              A
           Symbols:
           1: EC Toluene
           2: ITC Toluene
           3: UNC Toluene
           4: EC m-Xylene
           5: ITC m-Xylene
           6: UNC o-Xylene
Mechanisms:

RADM
RADM-P

RADM-M
March 1988 Mechanism
Recommended Parameter-only
Modifications
Recommended Modifications
                                     36

-------
experiments).   Both  modified versions  of RADM  perform  better  than  the
current  version  in simulating  these results,  with fewer  cases  of  large
discrepancies and  more cases of  good fits  being obtained.   Despite  its
more  condensed  DCB  mechanism,   the  more   extensively   modified   RADM-M
mechanism  seems  to  be  slightly   better  in  its  overall  distribution  of
discrepancies than the  version  where only   the  parameters are  modified,
though the  differences are probably not  significant.   It  is  significant,
however, that the  version of the  mechanism with  the more  detailed  (though
not necessarily more accurate)  representation of  DCB mechanism performs no
better  in  simulating  the results  of  these  experiments  than  does  the
version with three fewer species.
     2.2.2  Condensed Higher Alkane Representation
            One  area  where the  RADM mechanism is  more detailed  than the
current  ADOM  (Lurmann and Karamchandani,  1987),  Carbon Bond (Gery  et al.,
1988),  and  other  lumped . mechanisms  used   in  regional  models  is  its
representation of  the  reactions  of  the higher  (C^)  alkanes.   RADM uses
three  lumped species  to  represent  the C?+  alkanes (plus  one  additional
species  each  to  represent their  peroxy  radical  intermediates),  while both
ADOM  and Carbon  Bond  use  only  one  lumped  species for  these  compounds.
Test  calculations  were carried  out  as  part of this  study  to  determine
whether  a  moderate  level of condensation of the   RADM  alkane  mechanism,
where  the  two  higher  alkane model species HC5 and HC8 were lumped together
and represented  by a single species, would  have a significant  effect on
model  predictions  relative to  the evaluated  version.   The results,  given
 in detail  in  Section  6.4,  indicate  that  using  this   condensation has
 insignificant  effects  on  simulations of  species of  interest in regional
modeling simulations  for  the wide range  of  chemical  conditions examined.
 Because  of this, we implemented this condensation in the RADM-M mechanism.
      The OH radical rate constant and product yield parameters recommended
 for  the  new  lumped   higher  alkane  species  were  derived  in  a   manner
 analogous to the  derivation  of  the recommended  parameters for the  lumped
 higher  alkane  species  in  the  current  RADM mechanism,  as  discussed in
 Section 3.3.1 of  this  report.  Specifically, values of average OH  radical
 rate  constants  and  product  coefficients calculated  for  the  aggregated
 emissions  groups  represented  by HC5 and  HC8 in  the  evaluated mechanism
 were  further  averaged,  using  reactivity  weighing   and  relative  molar
                                     37

-------
emissions  in  the NAPAP  total  U.S.  anthropogenic  emissions inventory, to
determine  values  for  these  parameters which  are  appropriate  for  this
lumped HC5+HC8  alkane  model  species.   The  averages of the parameters, and
the  contributions  of the various emissions  groups to the values derived,
are  included  with  the tabulations given for  the  other RADM model species
in Section  3.3.1-   Since the average number of carbons calculated for  this
lumped species  based on  the NAPAP inventory  is approximately 6.3, this new
lumped  model  species  is designated  "HC6," to   be  consistent  with the
nomenclature  used   in  the  RADM mechanism.    The  listing  of  the  RADM-M
mechanism  shows the reactions  recommended for this  model  species,  which
replaces the model  species HC5 and HC8 in the RADM and RADM-P mechanisms.
      (Note  that the specific kinetic and mechanistic  parameters recommend-
ed for  the new  lumped  "HC6"  species based  in  the  analysis in Section 3.3.1
are  slightly  different from those used in the sensitivity  tests discussed
in Section 6.4.   This is because the  parameters used in  the  sensitivity
tests were derived based on  those used  for HC5 and HC8 in the  RADM  mechan-
ism,  for more straightforward  comparison with that mechanism.  As discussed
below in  Section 2.2.5,  the mechanistic parameters for HC5 and HC8  in the
evaluated  mechanism are slightly different from  those which are currently
recommended and implemented  in RADM-P.)
      It  is possible that it may not be  much  worse an  approximation  to lump
HC3  with  the  higher alkane classes, as  is  done  in the ADOM  and (in  effect)
the  Carbon Bond mechanisms.  Test calculations carried out using the most
recent  SAPRC  mechanism indicate that much  more  severe condensations of the
alkane  mechanisms  can be employed without  significantly affecting model
predictions,  provided the parameters for  the lumped  species  are adjusted
appropriately  based  on  the mixture  being  represented.    However,  more
severe  approximations for alkane representation in RADM were  not examined,
and  are not currently  recommended.
      2.2.3 Modifications to the XN02 Reactions
            The species "XN02"  is used in the  RADM mechanism  to represent
the  additional  NOX consumption caused by  reactions  of NO^ radicals with
dinitrophenols.   As such,  it  actually represents a phenoxy-radical-like
species.    However,  in the RADM mechanism, this  species  is represented as
forming  organic nitrates  (ONIT) when it reacts  with NOg,  forming organic
hydroperoxides   (OP2)  when  it reacts  with H02,  and reacting  with  M02 and
                                     38

-------
AC03 as if it were a peroxy radical.  We believe that  it  is more  chemical-
ly reasonable not to have  it  form ONIT  or OP2 in its  respective  reactions
with NC>2  and  H02,  and  that  it is equally  chemically  reasonable to  use  a
simpler representation  of  its loss in the  absence  of  N02 or HOp.   There-
fore, modifications  in this  aspect  of the mechanism  are recommended  for
both  improved  chemical reasonableness  and  to  simplify  the  mechanism.
These modifications are implemented  in the  RADM-M mechanism,  but  could  not
be implemented into RADM-P.
     The  reaction  of  XN02 with  N02  actually represents  the reaction  of
nitro-substituted phenoxy  radicals with N02 to form dinitrophenols.   These
species probably condense  into the aerosol  phase and  do not  react;  but if
they react at  all,  it  is  estimated  that  the  most  likely  process would be
reaction  with NO^  to  consume even more NOX.   However, the RADM  mechanism
in effect represents dinitrophenols  as alkyl  nitrates  (ONIT), whose subse-
quent  reactions tend   to  re-introduce NOV  into   the  system.    For  this
                                           A
reason, it is considered to  be more  chemically reasonable for the nitrogen
in  the  XN02  + N02 reaction to  be "lost",  rather than for it to  re-appear
as  ONIT.   Therefore,  we   recommend  that  the  XN02  + N02   reaction  have no
products, or,  if there is a need to keep  track of N-balance in the RADM
simulations,  its  products  be  represented  by  some  unreactive N-containing
"counter" species.
     The  reaction  of  XN02  with H02 represents a reaction  of  H02  with some
type of phenoxy  radical.   The most likely organic  product of such a reac-
tion would  be a phenol or nitrophenol, and  not a  hydroperoxide.   There-
fore,  the use of OP2  as   a product  in  this  reaction  is  probably inappro-
priate.   Since hydroperoxides  are assumed  to be involved in formation of
acid deposition, we recommend that this  source  of OP2 be  removed, and that
the  XN02  + H02 reaction also  have no  reactive products.
     The  reactions used to  represent the  removal  of  XN02 in  those cases
where  it  is formed  in the absence  of  significant amounts of H02 and N02
are  usually  not  important, since usually  removal by N02 dominates, and the
H02  reaction is probably  adequate  to represent its removal  in  most other
cases.    The  use  of peroxy-like   reactions   for this  purpose   is  not
necessarily  inappropriate,  but  is essentially  arbitrary.   However, this
representation   introduces  unnecessary   complexity   into  the  mechanism,
probably  making  it more  difficult  to  implement  in  "solvers"  where  the
                                     39

-------
steady  state approximation is used for  XN02.   To  simplify  the mechanism
and make  it easier to implement, we recommend  using  a simple first-order
loss process  to  represent  the consumption of XN02 in cases where levels of
both H02  and  NOp are  low.   The  rate constant we chose for this process in
RADM-M  (2.0x10"^ sec"1)  is essentially  arbitrary,  and corresponds to the
loss  rate by reaction with  1  ppt of NC^, or approximately  2 ppt of HC^.
Test calculations  show that  this  simpler representation has essentially no
effects on model predictions representative of conditions likely to occur
in regional model  simulations.
     2.2,4  Omitted OLN Reactions
             The  RADM  mechanism  includes  the  reactions  of M02  and AC03 with
all  the organic  peroxy radical  species  in  the mechanism  except for OLN,
the  peroxy radical which  is formed when  NO^ reacts  with  alkenes.   This
probably  was an oversight,  since  there is no  reason why  OLN should be
treated  differently  than  the other organic peroxy  radical species in this
regard.    Because under  certain  conditions at nighttime  OLN  can  be  the
dominant  peroxy radical,   it  is  advisable that the OLN  + OLN  reaction be
included   in  the mechanism,  since  that  reaction  might be  its major sink
under  those  conditions.   Since  the  RADM  development team  were  able to
readily add  these reactions  to the mechanism being implemented into RADM
(Stockwell,  private  communication,  1989),  these  changes are  incorporated
in  the RADM-P as well as  the RADM-M recommended modified mechanisms.   The
rate   constants   and  mechanisms  used  for  these  reactions  are  those of
Stockwell (private communication, 1988)  derived for the  "detailed  radical"
version of the  RADM mechanism,  given in  Section 6.2.
      2.2.5  Parameter Modifications Resulting from  Updated Emissions
             Assignments
             As   indicated  above and discussed  in  more  detail  in  Section
3.3,  kinetic, and in some cases  product, yield parameters for a number of
lumped species  in the RADM  mechanism  art derived based on  an analysis of
the  species  they represent  in  the NAPAP total U.S.  emissions inventory.
The  parameters  used  in the  evaluated RADM  mechanism  for  a  number  of these
species are based on recommendations we made in this regard as part of our
initial review  of the 1987  version of the  mechanism.   However, an updated
analysis of  these data, discussed  in Section  3.3  of this report,  resulted
in  slight  changes being  recommended for several of  these  parameters.  The
                                     40

-------
specific  changes which  were  recommended are  given  in  the  appropriate
tabulations in Section 3.3, and are incorporated  into  both  the recommended
modified RADM-M and RADM-P mechanisms.
     The  recommended  product yields  given  in  the tables  in  Section  3.3
refer to  yields under conditions where the  reactions of  peroxy  radicals
with NO  dominate.   The  RADM  mechanism also  has  peroxy radicals  reacting
with H02>  M02,  and  AC03.  The modified  recommended product yields  do  not
affect the  products of the HC^ reactions, since  these are  represented by
separate  hydroperoxide species.   However,  they  do  affect  the  products
formed in  the M02 or  AC03  reactions.   The  products in  those reactions were
derived based on assuming  that the products  from  M02 and  AC03 are the same
as  in the  RADM  mechanism,  and  that the oxygenated products formed from the
lumped peroxy radicals (e.g.,  HC3P,  etc.) are the  same as  those formed in
the  NO  reaction, excluding ONIT.   In all cases, the radical yields were
derived  based  on  assuming  the  reactions  occur 50%  via  RO +  RO + Oo
formation,  and  50% via  ROH formation.   This   is  the  same  approach as
discussed  above  for TOLP  and  XYLP in  the modified aromatics mechanisms and
is  in  line with the  representation  for most  of  these reactions  in other
current mechanisms.

2.3 The  Recommended Modified Mechanism  (RADM-M)
     The  RADM-M mechanism  is  the  version   of   the  RADM  mechanism which
incorporates  all the modifications recommended as  a result of this  study.
It   has  the  same  inorganic  reaction  set  and   (except  for  the  aromatic
product  species DCB and  its  PAN analogue and  the rate  of methyl glyoxal
photolysis)  the same  mechanisms  for  reactive product species as employed
 in  the  RADM mechanism.  Except for the  lumping together  of the two  higher
alkane  classes  and slight simplifications in mechanisms  for  some aromatic
products,  this  mechanism represents  the  same level  of  condensation and
chemical detail as  employed  in  the  RADM  mechanism.   However, results of
test calculations  representing the range  of chemical  conditions likely to
be  encountered  in  regional model simulations, given in Section 6.5 of  this
 report,   indicate  that  there  are cases where these  mechanisms  can  give
significantly  different  predictions.   These  differences  are attributed
primarily to changes  in  the  aromatics mechanism  that  were made to  improve
 its  performance  in   simulating  results  of  chamber  data.    Other  test

-------
calculations suggest  that the other modifications,  which are recommended
primarily  to yield  a mechanism that is somewhat more  compact and (in  the
case  of reactions  involving  XN02)  somewhat more  chemically reasonable,
should have relatively little effect on predictions under most conditions.
     The organic  model  species used in the RADM-M mechanism are  listed  in
Table  4,  and  their  reactions and  thermal  rate parameters  are  listed  in
Table  5.   The  inorganic  species  and their  reactions  and rate parameters
are not  included  in  these tables  because they are  the  same as  those  listed
for  the  RADM mechanism in  Tables 1 and 2,  respectively.  The absorption
coefficients and  quantum  yields used for the photolysis  reactions in  this
mechanism  are  given  in Table 3-

2.4  The Modified Parameter Mechanism (RADM-P)
     The  RADM-P  mechanism is  the  version  of  the RADM mechanism  which
incorporates  all  the modifications  recommended  as a result of this study
that   can  be   implemented  by  changing  only  kinetic  or  product   yield
values.    Because of practical difficulties involved  in making  the  more
extensive  changes required to implement RADM-M,  this is the  version  of the
mechanism  which  is  being implemented  into  the RADM-II  model  (Stockwell,
private  communication,  1989).   This  mechanism has  the same  number  of
species  and  (except for the three OLN reactions which  were added) the same
number of reactions as the March  1988  RADM  mechanism,  and thus  represents
almost exactly the  same  level of chemical  detail  and condensation.   Its
major  modifications are the changes in parameters involved in the aromatic
mechanisms which were  made to improve  its  performance  in  simulating the
chamber data.  As  with the RADM-M mechanism,  test calculations  discussed
in Section  6.5  indicate  that these  changes  result  in  cases where  this
mechanism can yield  significantly  different  predictions  in  simulations
representing conditions of regional model simulations.   The other modifi-
cations concern  the addition  of the  usually  relatively unimportant OLN
reactions and  the minor  changes  in  parameters  for lumped species based on
the updated  analysis of the emissions data.

-------
Table M.  Organic Model Species Employed in the Recommended Modified
          Version of the RADM Gas-Phase Chemical Mechanism  (RADM-M)


 Name      Description
  Organic Hydrocarbons

  CH4      Methane
  ETH      Ethane
  HC3      C3+ Alkanes with kOH  < 5  x  103  (a)
  HC6      Alkanes with kOH > 0.5 x  1 2 x 10

  Organic Oxidation Products

  HCHO    Formaldehyde
  ALD      Lumped Higher  Aldehydes
  KET      Ketones
  MGLY    Methyl Glyoxal
  DCBX    Uncharacterized  Reactive  Aromatic Fragmentation Products
  CSL      Phenols and Cresols

  PAN      Acyl  Peroxy Nitrates
  ONIT    Other Organic  Nitrates  (primarily alkyl nitrates)

  OP1      Methyl Hydroperoxide
  OP2      Higher Organic Hydroperoxides
  ORA1    Formic Acid
  ORA2    Higher Organic Acids  (other than peroxy acetic acid)
   PAA      Peroxy Acetic  Acid

  Organic Radicals

  M02      Methyl Peroxy  Radicals
   ETHP    Ethyl Peroxy  Radicals
   HC3P     Peroxy  Radicals from OH + HC3
   HC6P     Peroxy  Radicals from OH + HC6
   OL2P     Peroxy  Radicals from OH + Ethene
   OLTP     Peroxy  Radicals from OH +  Isoprene  and Terminal  Alkenes
   OLIP     Peroxy Radicals from OH •*•  Internal  Alkenes
   KETP     Peroxy Radicals from OH + Ketones
   TOLP     Peroxy Radicals from OH + Benzene  and Monoalkyl  Benzenes
   XYLP     Peroxy Radicals from OH -»- Xylenes  and More Reactive  Aromatics
   OLN      Peroxy Radicals from NOo + Alkenes

                                                                (continued)

-------
Table l» (continued) - 2
 Name      Description
  X02      Chemical Operator Used to Account for NO-to-NC^ conversions
           due to Secondarily Formed Organic Peroxy Radicals

  AC03     Acyl Peroxy Radicals

  XN02     Chemical Operator Used to Account for NOX Loss Due  to
           Reactions of NO? with Phenols and Cresols.


(a)  kOH is the OH radical rate constant at 300 K  in ppm   min  .

-------
Table 5.  Listing of the Organic Reactions in the Recommended Modified
          Version of the RADM Gas-Phase Chemical Mechanism (RADM-M)
Rxn.
Label
(a)
010
011
012
013
014
015
016
019
M01
021
Kinetic


k(300)
(Phot.
(Phot.
(Phot.
(Phot.
( Phot ,
(Phot,
(Phot,
(Phot,
(Phot
(Phot

Parameters (b)
A
Ea

B
, Set = HCHOMR )
. Set = HCHORR )
. Set = ALDR )
. Set = OPR )
. Set = OPR )
. Set = PAAR )
. Set = KETR )
. Set = MGLYRM )
. Set = DCBRM )
. Set = ONITR )
Reactions (b)
Organic Photolysis Reactions
HCHO + HV = H2 * CO
HCHO + HV = H02 * H02 + CO
ALD + HV = M02 + H02 * CO
OP1 * HV = HCHO + H02 + HO
OP2 + HV = ALD t- H02 * HO
PAA * HV = M02 + C02 + HO
KET -i- HV = AC03 f ETHP
MGLY + HV = AC03 + H02 + CO
DCBX + HV = H02 + CO + AC03
ONIT + HV = #.20 ALD * #.80 KET +
H02 + N02
Organic + OH Reactions
051
052
053


M02
056
057
058
074
059

060

063
064
065
067
M03
061

062
069A
069B
070A
070B
1
4
3


1
1
4
9
1
9

4

1
2
1
2
2
5

5
7
7
7
7
.29E+01
.11E+02
.86Ef03


.09E+04
.24E+04
. 19E+04
.79E+04
.46E+05
.02E+03

.09E+04

. 32E+04
.37E+04
.47E+03
. 50E+04
.50E+04
.87E+04

.28E+04
.34E+03
.34E+03
.34E+03
.34E+03
9.
1.
2.


3.
3.
7.
1.
3.
23E+02
81E+03
33E+04


86E+04
16E+03
81E+03
57E+04
74E+04
3.08E+03


2.77E+04

1.

.32E+04
1.01E+04
1.76E+04
2.50E+04
2.50E+04
5.87E+04

5
7
7
7
7

.28E+04
.34E+03
.34E+03
.34E+03
.34E+03
2.54
0.88
1.07


0.75
-0.82
-1.00
-1.09
-0.81
-0.64

-0.23

0.00
-0.51
1.48
0.00
0.00
0.00

0.00
0.00
0.00
0.00
0.00
1
1
-1


-1
-1
-1
-1
-1
-1

-1

-1
-1
-1
-1
-1
-1

-1
-1
-1
-1
-1
.00
.00
.00


.00
.00
.00
.00
.00
.00

.00

.00
.00
.00
.00
.00
.00

.00
.00
.00
.00
.00
CH4
ETH
HC3
#,
#
HC6
OL2
OLT
OLI
ISO
TOL
#
XYL
#
+ HO =
+ HO =
+ HO =
M02 H
ETHP
#.83
i- H20
+ H20
HC3P -i- #.17 H02 +
.009 HCHO + #.075 ALD +
.025 KET + H20
+ HO =
+ HO =
+ HO =
+ HO =
+ HO =
+ HO =
.24 H02
+ HO =
.17 H02
HCHO + HO
ALD
KET
+ HO =
+ HO =
MGLY + HO
DCBX + HO
CSL
#
CSL
OP1
HO
OP2
HO
+ HO =
.9 AC03
+ HO =
+ HO =
+ OP1 =
+ HO =
+ OP2 =
HC6P
OL2P
OLTP
OLIP
OLTP
#.76

#.83

= H02
AC03
KETP
+ #.51 X02 + H20




TOLP + #.24 CSL +

XYLP + #.17 CSL +

+ CO + H20
+ H20
+ H20
= AC03 + CO + H20
= AC03
#.1

CSL
M02
HO +
HC3P
HO *
H02 +1.9 X02 +



HCHO

ALD
                                                               (continued)
                                     45

-------
Table 5  (continued)  - 2
Rxn.
Label
(a)
071
072
073
Kinetic Parameters (b)

1.
2.
3.
k(300)
47E+04
06E-t-02
76E+03

1
9
2
A
.47E+04
.06E+02
.28E+04

0
0
1
Ea
.00
.88
.07

-1
1
-1
B
.00
.00
.00
Reactions (b)
PAA
PAN
ONIT
+ HO = AC03 + H20
+ HO = HCHO + N03 + X02
* HO = HC3P + N02



Organic + N03 Reactions
093
094
096
098
099
100
101
102
9.
3.
3.
3.
1.
2.
1.
8.
23E-01
65E+00
65E+00
23E+04
72E-01
65E+01
84E+03
53E+02
8
2
2
3
2
1
4
8
.81E+02
.06E+03
.06E+03
.23E+04
.94E+03
.47E+04
.74E+04
.53E+02
4
3
3
0
5
3
1
0
.09
.78
.78
.00
.81
.77
.94
.00
-1
-1
-1
-1
-1
-1
-1
-1
.00
.00
.00
.00
.00
.00
.00
.00
HCHO
ALD
MGLY
CSL
OL2
OLT
OLI
ISO
+ N03 = H02 + HN03 * CO
+ N03 = AC03 + HN03


+ N03 = HN03 + AC03 * CO
+ N03 = HN03 + XN02 * 1.
+ N03 = OLN
-»- N03 = OLN
+ N03 = OLN
* N03 = OLN
5 CSL




Organic ••• 03 Reactions
103

104


2.

1.


72E-03

74E-02


1

1


.76E+01

.94E+01


5

4


.23

.18


-1

-1


.00

.00


OL2
#.
OLT
#.
#.
+ 03 = HCHO + #.42 CO +
4 ORA1 +1.12 H02
+ 03 = #.53 HCHO + #.50
33 CO + #.20 ORA1 +
20 ORA2 + #.23 H02 +
#.22 M02 + #.10 HO + #.06
105



106



2.43E-01






2.20E-02






1



1



.07E+01



.81E+01



2



4



.26



.00



-1



-1



.00



.00



OLI
.
.
^
ISO
.
•
m
+ 03 = 1.18 HCHO + #.72
10 KET + #.23 CO + #.06
29 ORA2 + #.09 CH4 +
26 H02 + #.14 HO + #.31
+ 03 = #.53 HCHO + #.50
33 CO + #.20 ORA1 +
20 ORA2 + #.23 H02 +
22 M02 + #.10 HO


ALD +


CH4
ALD +
ORA1

M02
ALD +



 075   7.52E+03  4.11E+03 -0.36 -1.00
 076   2.90E-02  1.17E+18 26.91  0.00
 086   1.12E+04  6.17E+03 -0.36 -1.00
 079   1.12E+04  6.17E+03 -0.36 -1.00
 080   1.12E+04  6.17E+03 -0.36 -1.00
Organic Peroxy + N02 Reactions

AC03 + N02 = PAN
PAN = AC03 * N02

Organic Peroxy + NO Reactions

AC03 + NO = M02 + N02
M02 + NO = HCHO + H02 + N02
HC3P + NO = #.75 ALD + #.25 KET +
  #.09 HCHO + #.036 ONIT +
  #.964 N02 + #.964 H02

                       (continued)
                                     46

-------
Table 5 (continued) - 3
Rxn.
Label
(a)
M04


083

084
085

088

089

090
091
092
150
Kinetic
k(300)
1.12E+04


1.12E+04

1.12E+04
1.12E+04

1.12E+04

1.12E+04

1.12E+04
1.12E+04
1.12E+04
1.12E+04

6.


6.

6.
6.

6.

6.

6.
6.
6.
6.
Parameters (b)
A
17E+03


17E+03

17E+03
17E+03

17E+03

17E+03

17E+03
17E+03
17E+03
17E+03
Ea
-0.36 -1,


B
,00


-0.36 -1.00


-0.36 -1.00
-0.36 -1.00

-0.36 -1

-0.36 -1

-0.36 -1
-0.36 -1
-0.36 -1
-0.36 -1

.00

.00

.00
.00
.00
.00
Reactions (b)
HC6P +
#.15
#.85
OL2P +
N02
OLTP +
OLIP +
#.28
TOLP +
#.17
XYLP +
#.45
ETHP +
KETP +
OLN +
X02 +
NO = #.38 ALD + #.87 KET +
ONIT + #.03 HCHO +
N02 + #.85 H02
NO = #1.6 HCHO + H02 +
+ #.2 ALD
NO = ALD + HCHO + H02 + N02
NO = H02 + #1.45 ALD +
HCHO + #.10 KET + N02
NO = N02 + H02 +
MGLY + #.68 DCBX
NO = N02 + H02 +
MGLY + #.79 DCBX
NO = ALD + H02 -t- N02
NO = MGLY + N02 + H02
NO = HCHO + ALD + #2 N02
NO = N02
Organic Peroxy + H02 Reactions
107
108
109
M05
112
113
114
115
117
118
120
146
116
8.61E+03












1.












13E+02
(Same
(Same
(Same
(Same
(Same
(Same
(Same
(Same
(Same
(Same
(Same
(Same
-2.58 -1
k as 107
k as 107
k as 107
k as 107
k as 107
k as 107
k as 107
k as 107
k as 107
k as 107
k as 107
k as 107
.00
)
)
)
)
)
)
)
)
)
)
)
)
H02 +
H02 +
H02 +
H02 +
H02 +
H02 +
H02 +
H02 +
H02 +
H02 +
H02 +
H02 +
H02 +
M02 = OP1
ETHP = OP2
HC3P = OP2
HC6P = OP2
OL2P = OP2
OLTP = OP2
OLIP = OP2
KETP = OP2
TOLP = OP2
XYLP = OP2
OLN = ONIT
X02 = OP2
AC03 = PAA
Organic Peroxy + AC03 Reactions
142
130

134

135
3.64E-t-03
2.93E+03

1.04E+03

3.06E+02
1
1

4

1
.75E+03 -0.44 -1
.41E+03 -0.44 -1


.99E+02 -0.44 -1


.47E+02 -0.44 -1
.00
.00

.00

.00
AC03 + AC03 = #2 M02
M02 +
#.5
AC03 = HCHO + #.5 H02 +
M02 + #.5 ORA2
ETHP + AC03 = ALD + #.5 H02 +
#.5
HC3P H
M02 + #.5 ORA2
* AC03 = #.77 ALD +
#.26 KET + #.09 HCHO + #.5 H02 -t






#.5
M02 + #.5 ORA2
                                                              (continued)
                                     47

-------
Table 5 (continued)  - 4
Rxn.
Label
(a)
M06

138

139

140


141

143


144


M07

148
Kinetic Parameters (b)

k(300)
2.57E+02

1

1

1


1

1


1


1

1

.04E+03

.04E+03

.28E+02


.28E+02

.28E+02


.28E+02


.28E+02

.28E+02

1.

4.

4.

6,


6.

A
23E+02

99E+02

99E+02

17E+01


17E+01

6.17E+01


6


6

6


. 17E+01


.17E+01

.17E+01

-0.

-0.

-0.

-0.


-0.

-0.


Ea
44

44

44

44


44

44


-0.44




-0.44


-0.44

-1.

-1,

-1

-1


-1

-1


-1


-1

-1
B
00

00

00

.00


.00

.00


.00


.00

.00
Reactions (b)
HC6P
11
1.
OL2P
1.
OLTP
1.
OLIP
#.
1.
KETP
1.
+ AC03 =
.02 KET +
5 H02 + 1
+ AC03 =
5 H02 + 1
+ AC03 =
5 H02 + 1
+ AC03 =
#.45 ALD +
#.04 HCHO +
.5 M02 + #.5 ORA2
#.8 HCHO + #.6 ALD -
.5 M02 + #.5 ORA2
ALD + #.5 HCHO +
.5 M02 + #.5 ORA2
#.725 ALD +
55 KET + #.14 HCHO +1.5 H02
5 M02 + #
+ AC03 =
5 M02 + #
AC03 + TOLP =
1.
*.
68 DCBX +
5 ORA2
AC03 + XYLP =
i.
i.
OLN
1.
X02
79 DCBX +
5 ORA2
+ AC03 =
5 ORA2 +
+ AC03 =
.5 ORA2
MGLY + #.5 H02 +
.5 ORA2
#.17 MGLY +
#.5 H02 + #.5 M02 +

#.45 MGLY +
#.5 H02 + #.5 M02 +

HCHO + ALD +
N02 + #.5 M02
M02
Organic Peroxy + M02 Reactions
121
122

123

M08

126

127

128

129

131

132

5
4

1

1

4

4

5

5

5

5

.81E+02
.28E+02

.28E+02

.04E+02

.28E+02

.28E+02

.19E+01

.19E+01

.19E+01

.19E+01

2
2

6

4

2

2

2

2

2

2

.79E+02
.06E+02

.17E+01

.99E+01

.06E+02

.06E+02

.50E+01

.50E+01

.50E+01

.50E+01

-0.44
-0.44


-0.44


-0.44


-0.44


-0.44

-0

-0

-0

-0


.44

.44

.44

.44

-1
-1

-1

-1

-1

-1

-1

-1

-1

-1

.00
.00

.00

.00

.00

.00

.00

.00

.00

.00

M02
M02
1.
M02
t.
M02
#.
M02
1.
M02
*
M02
1
M02
1
M02
t
M02
*
+ M02 = #
+ ETHP =
75 ALD
+ HC3P =
77 ALD +
+ HC6P =
45 ALD +
+ OL2P =
35 ALD
+ OLTP =
75 ALD
+ OLIP =
725 ALD ••
+ KETP =
.75 MGLY
+ TOLP =
.17 MGLY H
+ XYLP =
1.5 HCHO + H02
#.75 HCHO + H02 +

.84 HCHO + H02 +
.26 KET
.79 HCHO + H02 +
1.02 KET
1.55 HCHO + H02 +

#1.25 HCHO + H02 +

#.89 HCHO + H02 +
• #.55 KET
#.75 HCHO + H02 +

#.75 HCHO + H02 +
- #.68 DCBX
#.75 HCHO + H02 +
.45 MGLY + #.79 DCBX
                                                              (continued)
                                     48

-------
Table 5 (continued) - 5
Rxn.
Label
(a)
           Kinetic Parameters (b)

        k(300)     A        Ea    B
Reactions (b)
M09   5.19E+01  2.50E+01 -0.44 -1.00   M02 + OLN = 11.75 HCHO
                                         ALD + N02
147   5.19E+01  2.50E+01 -0.44 -1.00   M02 + X02 = HCHO + H02
                                                                #.5 H02
 149
M10
M11
M12
M13
                                       Other Organic Peroxy + Peroxy
                                       Reactions

      1.10E+00  5.28E-01 -0.44 -1.00   X02 + X02 =
      1.10E+00  5.28E-01 -0.44 -1.00   OLN + OLN = 12 HCHO + #2 ALD
                                         12 N02

                                       XN02 Reactions
      1.12E+04  6.17EH-03 -0.36 -1.00
                  (Same k as 107 )
      1.20E-02   (No T Dependence)
XN02 + N02 = (Loss-of-NOx)
XN02 + H02 =
XN02 =
 Notes:

 (a)  Reactions which correspond directly  to  reactions  in  the  March 1988
     RADM mechanism in Table 2 are  given  the same  reaction  labels as used
     for them in Table 2.  Reactions  that were added or significantly
     modified relative to the mechanism  in Table 2 are indicated by the
     label  "Mxx".

 (b)  See Footnotes (b) and (c) in Table  2 for a description of the
     formats used in the kinetic parameter and the reaction listings.
                                     49

-------
     The model species used  in the  RADM-P  mechanism are the same as those
listed for the RADM mechanism in Table 1.  The organic reactions and their
thermal rate  parameters  for the RADM-P  mechanism are listed  in Table 6.
The inorganic reactions and  rate parameters  are  also the same as those of
the mechanism listed in Table 2 and are thus not  included in Table  6.  The
absorption coefficients and  quantum yields used  for the photolysis reac-
tions in this mechanism are  the same  as  used for the RADM-M mechanism and
are given in Table 3.  This  mechanism was  implemented by Stockwell on his
modeling software at  SUNY,  and test  calculations indicate  that the mech-
anism implemented at SUNY is consistent with its implementation at SAPRC.
                                     50

-------
Table 6.  Listing of the Organic Reactions in the Modified Parameter
          Version of the RADM Gas-Phase Chemical Mechanism (RADM-P)
Rxn.
Label 	
(a)    k(300)
 Kinetic  Parameters  (b)

       A        Ea    B
Reactions (b)
010
011
012
013
014
015
016
017
018
019
020
( Phot .
(Phot.
( Phot .
(Phot.
(Phot.
(Phot.
(Phot.
(Phot.
(Phot.
(Phot.
(Phot.
Set
Set
Set
Set
Set
Set
Set
Set
Set
Set
Set
= HCHOMR )
= HCHORR )
= ALDR )
= OPR )
= OPR )
= PAAR )
= KETR )
= GLY2R )
= GLY1R )
= MGLYRM )
= DCBRM )
 021
                                       Organic Photolysis Reactions
                                       HCHO + HV  = H2
                                       HCHO + HV  = H02
                                              CO
                                             t- H02
                             ALD + HV = M02 + H02
                             OP1 + HV = HCHO + H02
                             OP2 * HV = ALD + H02
                             PAA
                             KET
(Phot. Set = ONITR   )
                        CO
                       CO
                        HO
                       HO
    + HV = M02 + C02 -i- HO
    + HV = AC03 + ETHP
GLY f HV = #.13 HCHO + #1.87 CO
GLY 1- HV = #.45 HCHO + #1.55 CO
  #.80 H02
MGLY + HV = AC03 + H02 + CO
DCB + HV = #.98 H02 + #.02 AC03
  TC03
ONIT + HV = #.20 ALD + #.80 KET
  H02 •*• N02

Organic + OH  Reactions
051
052
053


054
055
056
057
058
074
059

060

063
064
065
066
067
061

1,
4
.29E+01
.11E+02
3.86E+03


7
1
1
4
9
1
9

4

1
2
1
1
2
5



.16E+03
.51E+04
.24E+04
.19E+04
.79E+04
.46E+05
.02E+03

.09E+04

.32E+04
.37E+04
.47E+03
.69E+04
.50E+04
.87E+04

9.23E+02
1.81E+03
2.33E+04


2.54E+04
5.34E+04
3.16E+03
7.81E+03
1.57E+04
3.74E+04
3.08E-1-03

2 . 77E+04

1.32E+04
1.01E+04
1.76E+04
1.69E+04
2.50E+04
5.87E+04

2.54
0.88
1.07


0.75
0.75
-0.82
-1.00
-1.09
-0.81
-0.64

-0.23

0.00
-0.51
1.48
0.00
0.00
0.00

1
1
-1


-1
-1
-1
-1
-1
-1
-1

-1

-1
-1
-1
-1
-1
-1

.00
.00
.00


.00
.00
.00
.00
.00
.00
.00

.00

.00
.00
.00
.00
.00
.00

CH4
ETH
HC3
#.
#.
HC5
HC8
OL2
OLT
OLI
ISO
TOL
#
XYL
*•
+ HO =
+ HO =
+ HO =
M02 ,
ETHP
#.83
i- H20
+ H20
HC3P *


#.17 H02 +
009 HCHO + #.075 ALD +
025 KET + H20
+ HO =
+ HO =
+ HO =
+ HO =
+ HO =
+ HO =
+ HO =
.25 H02
+ HO =
.17 H02
HCHO + HO
ALD
KET
GLY
+ HO =
+ HO =
+ HO =
MGLY + HO
CSL
#
+ HO =
.9 TC03
HC5P
HC8P
OL2P
OLTP
OLIP
OLTP
#.75

#.83

= H02
AC03
KETP
H02
+ #.25
+ #.75




TOLP t-

XYLP +

+ CO +
+ H20
+ H20
+ #2 CO
= AC03 + CO
#.1

H02 + #

X02 + H20
X02 + H20




#.25 CSL *

#.17 CSL +

H20


I- H20
+ H20
.9 X02 +

                                                                (continued)
                                      51

-------
Table 6 (continued) - 2
Rxn.
Label
(a)
062
068
069A
069B
070A
070B
071
072
073
Kinetic Parameters (b)
k(300)
5
4
7
7
7
7
1
2
3
.28E+04
. 11E+04
. 34E+03
. 34E+03
. 34E+03
.34E+03
.47E+04
.06E+02
. 76E+03
5
4
7
7
7
7
1
9
2
A
.28E+04
.11E+04
.34E-I-03
.34E+03
.34E+03
.34E+03
.47E+04
.06E+02
.28E-I-04

0
0
0
0
0
0
0
0
1
Ea
.00
.00
.00
.00
.00
.00
.00
.88
.07
B
-1.00
-1.00
-1.00
-1.00
-1.00
-1.00
-1.00
1.00
-1.00
Reactions (
CSL
DCB
OP1
HO +
OP2
HO +
PAA
PAN
ONIT
+ HO =
+ HO =
+ HO =
OP1 =
••- HO =
OP2 =
-i- HO =
+ HO =
•»• HO =
b)
CSL
TC03 + H20
M02
HO + HCHO
HC3P
HO + ALD
AC03 + H20
HCHO + N03 + X02
HC3P + N02
Organic + N03 Reactions
093
094
095
096
097
098
099
100
101
102
9
3
9
3
3
3
1
2
1
8
.23E-01
.65E+00
.23E-01
. 65E+00
.65E+00
. 23E+04
.72E-01
.65E+01
.84E+03
.53E-I-02
8
2
8
2
2
3
2
1
14
8
.81E+02
.06E+03
.81E+02
.06E+03
.06E+03
.23E+04
.94E+03
.47E+04
.74E+04
.53E+02
4
3
4
3
3
0
5
3
1
0
.09
.78
.09
.78
.78
.00
.81
.77
.94
.00
-1.00
-1.00
-1.00
-1.00
-1.00
-1.00
-1.00
-1.00
-1.00
-1.00
HCHO
ALD
GLY
MGLY
DCB
CSL
OL2
OLT
OLI
ISO
+ N03
+ N03 =
+ N03 =
* N03
+ N03 =
+ N03 =
* N03 =
+ N03 =
+ N03 =
+ N03 =
= H02 + HN03 > CO
AC03 + HN03
HN03 -•• H02 > #2 CO
= HN03 -i- AC03 + CO
HN03 + TC03
HN03 -*• XN02 + #.5 CSL
OLN
OLN
OLN
OLN
Organic + 03 Reactions
103

104



105



106



2

1



2



2



. 72E-03

.74E-02



.43E-01



.20E-02



1

1



1



1



.76E+01

.94E+01



.07E+01



.81E+01



5

4



2



4



.23

.18



.26



.00



-1.00

-1.00



-1.00



-1.00



OL2
#.
OLT
.
•
^
OL
•
.
•
ISO
•
.
.
+ 03 =
4 ORA1
+ 03 =
33 CO +
20 ORA2
22 M02
+ 03 =
10 KET
29 ORA2
26 H02
+ 03 =
33 CO +
20 ORA2
22 M02
HCHO + #.42 CO t-
f #.12 H02
#.53 HCHO + #.50 ALD +
#.20 ORA1 +
+ #.23 H02 +
+ #.10 HO -•- .06 CH4
#.18 HCHO + .72 ALD +
* #.23 CO + .06 ORA1
+ #.09 CH4
+ #.14 HO + .31 M02
#.53 HCHO + .50 ALD +
#.20 ORA1 +
+ #.23 H02 +
-t- #.10 HO
                                       Organic Peroxy + N02 Reactions
075   7.52E+03  4.11E+03 -0.36 -1.00
076   2.90E-02  1.17E+18 26.91  0.00
AC03 + N02 = PAN
PAN = AC03 + N02
                                                              (continued)
                                    52

-------
Table 6 (continued) - 3
Rxn.
Label
(a)
077
078
Kinetic
k(300)
6.90E+03
2.90E-02

6.
1.
Parameters (b)
A
90E+03
17E+18
Ea B
0.00 -1.
26.91 0.
00
00
Reactions (b)
TC03 +
TPAN =
N02 =
TC03
TPAM
+ N02


Organic Peroxy + NO Reactions
086
087


079
080

1.12E+OM
1.12E+04


1.12E+04
1.12E+04

6.
6.


6.
6.

17E+03
17E+03


.17E+03
, 17E+03

-0.36 -1.
-0.36 -1.


-0.36 -1,
-0.36 -1,

.00
,00


.00
.00

AC03 +
TC03 +
1.89
#.05
NO =
NO =
GLY
AC03
M02 + N02

N02 + #.92 H02 +
+ #.11 MGLY
+ #.95 CO +
M02 + NO = HCHO + H02 +
HC3P +
1.09
NO =
HCHO
1.964 M02
081

082


083

084
085

088

089

090
091
092
153
1.12E+04

1.12E+04


1.12E+04

1.12E+04
1.12E+04

1.12E+04

1.12E+04

1.12E+OM
1.12E+04
1 . 12E+04
1.12E+04
6.

6


6

6
6

6

6

6
6
6
6
. 17E+03

. 17E+03


. 17E+03

. 17E+03
. 17E+03

. 17E+03

. 17E+03

.17E+03
.17E+03
.17E+03
.17E+03
-0.36 -1.

-0.36 -1


-0.36 -1

-0.36 -1
-0.36 -1

-0.36 -1

-0.36 -1

-0.36 -1
-0.36 -1
-0.36 -1
-0.36 -1
.00

.00


.00

.00
.00

.00

.00

.00
.00
.00
.00
HC5P +
1.08
HC8P +
1.04
#.76
OL2P *
N02 -
OLTP +
OLIP +
#.28
TOLP +
#.17
XYLP +
#.45
ETHP +
KETP +
OLN +
X02 +
NO =
ONIT
NO =
HCHO
#.75 ALD + #
+ #.036 ONIT
+ #.964 H02
#.38 ALD + #
+ #.92 N02 +
#.35 ALD + #
+ #.24 ONIT
+
#2 X02
N02
.25 KET +
+

.69 KET +
#.92 H02
1.06 KET +
-»-
N02 + #.76 H02
NO =
«• #.2
NO =
NO =
#1.6 HCHO +
ALD
H02 +

ALD + HCHO + H02 + N02
H02 + #1.45
HCHO +1.10 KET
NO =
MGLY
NO =
MGLY
NO =
NO =
NO =
NO =
N02 + H02 +
+ #.16 GLY -•
N02 + H02 +
+ #.806 DCB
ALD + H02 +
MGLY + N02 H
HCHO + ALD +
N02
ALD +
+ N02

- #.70 DCB


N02
K H02
#2 N02

Organic Peroxy + H02 Reactions
107
108
109
110
111
112
113
114
115
117
8.61E+03









1









.13E+02
(Same
(Same
(Same
(Same
(Same
(Same
(Same
(Same
(Same
-2.58 -1
k as 107
k as 107
k as 107
k as 107
k as 107
k as 107
k as 107
k as 107
k as 107
.00
)
)
)
)
)
)
)
)
)
H02 +
H02 -•-
H02 +
H02 «•
H02 +
H02 +
H02 +
H02 +
H02 +
H02 *
M02 =
ETHP
HC3P
HC5P
HC8P
OL2P
OLTP
OLIP
KETP
TOLP
OP1
= OP2
= OP2
= OP2
= OP2
= OP2
= OP2
= OP2
= OP2
= OP2










                                                               (continued)
                                      53

-------
Table 6  (continued) - 4
Rxn.
Label 	
(a)    k(300)
Kinetic Parameters (b)

       A        Ea    B
Reactions (b)
118
120
149
116
119
(Same k as 107 )
(Same k as 107 )
(Same k as 107 )
(Same k as 107 )
(Same k as 107 )
H02
H02
H02
H02
H02
+ XYLP = OP2
+ OLN = ONIT
+ X02 = OP2
+ AC03 = PAA
+ TC03 = OP2
 143    3.64E+03  1.75E+03 -0.44  -1.00
 146    3.64E+03  1.75E+03 -0.44  -1.00
 130   2.93E+03  1.41E+03 -0.44 -1.00

 133   2.93E+03  1.41E+03 -0.44 -1.00



 135   1.04E+03  4.99E-I-02 -0.44 -1.00

 136   3.06E+02  1.47E+02 -0.44 -1.00


 137   2.57E+02  1.23E+02 -0.44 -1.00


 138   2.20E+02  1.06E+02 -0.44 -1.00


 139   1.04E+03  4.99E+02 -0.44 -1.00

 140   1.04E+03  4.99E+02 -0.44 -1.00

 141   1.28E+02  6.17E+01 -0.44 -1.00


 142   1.28E+02  6.17E+01 -0.44 -1.00

 144   1.28E+02  6.17E+01 -0.44 -1.00

 145   1.28E+02  6.17E+01 -0.44 -1.00

 147   1.28E+02  6.17E+01 -0.44 -1.00
                            Organic Peroxy + AC03 Reactions

                            AC03 + AC03 = 12 M02
                            AC03 + TC03 = M02 + #.92 H02 +
                              1.890 GLY + 1.11 MGLY +
                              1.05 AC03 f 1.950 CO + 12 X02
                            M02 + AC03 = HCHO + #.5 H02 +
                              4.5 M02 + 4.5 ORA2
                            M02 * TC03 =1.5 HCHO + 4.5 ORA2 +
                              1.46 H02 + 1.445 GLY +
                              1.055 MGLY + 1.025 AC03 +
                              #.475 CO + X02
                            ETHP + AC03 = ALD + *.5 H02 +
                              1.5 M02 +1.5 ORA2
                            HC3P + AC03 = #.77 ALD +
                              1.26 KET -K 1.5 H02 * 1.5 M02 +
                              1.5 ORA2
                            HC5P + AC03 = 1.41 ALD -»-
                              1.75 KET + 4.5 H02 + #.5 M02 +
                              1.5 ORA2
                            HC8P -K AC03 = #.46 ALD +
                              #1.39 KET + #.5 H02 + #.5 M02 +
                              #.5 ORA2
                            OL2P + AC03 = #.8 HCHO + #.6  ALD +
                              #.5 H02 + #.5 M02 + #.5 ORA2
                            OLTP + AC03 = ALD + #.5 HCHO  +
                              #.5 H02 + #.5 M02 + #.5 ORA2
                            OLIP + AC03 = #.725 ALD +
                              #.55 KET * #.14 HCHO * #.5  H02 +
                              #.5 M02 + #.5 ORA2
                            KETP + AC03 = MGLY + #.5 H02  +
                              #.5 M02 + #.5 ORA2
                            AC03 + TOLP = M02 + #.17 MGLY +
                              #,16 GLY * #.70 DCB + H02
                            AC03 + XYLP = M02 + #.45 MGLY +
                              #.806 DCB + H02
                            AC03 + OLN = HCHO + ALD +
                              #.5 ORA2 + N02 + 4.5 M02

                                                    (continued)
                                      54

-------
Table 6 (continued) - 5
Rxn,
Label
(a)
151
Kinetic
k(300)
1.28E+02

6.
Parameters (b)
A
17E+01
Ea
-0.44
B
-1.00
Reactions (b)
X02 + AC03 =
M02

Organic Peroxy + M02 Reactions
121
122

123

124

125

126

127

128

129

131

132

134

150
5.81E+02
4.28E+02

1.28E+02

1.04E+02

8.86E+01

4.28E+02

4.28E+02

5.19E+01

5.19E+01

5.19E+01

5.19E+01

5.19E+01

5.19E+01
2.
2.

6.

4.

4.

2.

2.

2.

2.

79E+02
06E+02

17E+01

99E+01

26E+01

06E+02

06E+02

50E+01

50E+01

2.50E-I-01


2.50E+01


2.50E+01


2.50E+01
-0.
-0.

-0.

-0.

-0.

-0.

-0.

-0.

-0.

44
44

44

44

44

44

44

44

44

-0.44


-0.44


-0.44


-0.44
-1.00
-1.00

-1.00

-1.00

-1.00

-1.00

-1.00

-1.00

-1.00

-1.00

-1.00

-1.00

-1.00
M02 + M02 = #1.5 HCHO •»
M02 •»• ETHP =
#.75 ALD
M02 -•- HC3P =
#.77 ALD +
M02 -t- HC5P =
#.41 ALD -t-
M02 1- HC8P =
#.46 ALD +
M02 + OL2P =
#.35 ALD
M02 + OLTP =
#.75 ALD
M02 -H OLIP =
#.725 ALD
M02 + KETP =
#.75 MGLY
M02 + TOLP =
#.17 MGLY
M02 + XYLP =
#.45 MGLY
M02 + OLN =
ALD + N02
X02 + M02 =
#.75 HCHO

.84 HCHO
.26 KET
.77 HCHO
.75 KET
.80 HCHO
#1.39 KET
- H02
+ H02 +

+ H02 +

+ H02 +

+ H02 +

#1.55 HCHO -i- H02 +


#1.25 HCHO * H02 +

#.89 HCHO
+ #.55 KET
#.75 HCHO

HCHO + #2
+ #.16 GLY
HCHO + #2

-.- H02 +

+ H02 -.-

H02 +
+ #.70 DCB
H02 +
+ #.806 DCB
#1.75 HCHO

HCHO + H02
Other Organic Peroxy +

148

152

1.10E+00

1.10E+00

5

5

.28E-01

.28E-01

-0

-0

.44

.44

-1.00

-1.00
Reactions
OLN -f OLN =
#2 N02
X02 + X02 =

#2 HCHO +


I- #.5 H02


Peroxy

#2 ALD +


XN02 Reactions
154
155
156
157
158
1.12E+04

5.19E+01
1.28E+02
1.10E+00
6

2
6
5
.17E-I-03
(Same
.50E+01
.17E+01
.28E-01
-0
.36
-1.00
k as 107 )
-0
-0
-0
.44
.44
.44
-1.00
-1.00
-1.00
XN02 + N02 =
XN02 + H02 =
XN02 * M02 :
XN02 + AC03
XN02 * XN02
: ONIT
: OP2


: HCHO + H02
= M02
=


                                                               (continued)
                                     55

-------
Table 6  (continued)  -  6
Notes:

(a)   Reaction  labels  shown  in  this table are the reaction sequence
      number used  by Stockwell  to  implement this mechanism at SUNY.
      Some  reactions were  re-arranged for the purpose of this listing.
      Some  reactions were  split into two reactions for implementation
      on  SAPRC  modeling  software.

(b)   See Footnotes (b)  and  (c) in Table 2 for a description of the
      formats used in  the  kinetic  parameter and the reaction listings,
                                     56

-------
      3.   AGGREGATION  AND  REPRESENTATION OF ORGANIC  EMISSIONS  IN RADM

3.1   Introduction
     The  gas-phase  reactions  of  volatile  organic compounds  (VOCs)  are
important in  regional oxidant  and acid deposition  models for a number  of
reasons.  They play a major role  in the formation of oxidants on the local
and  regional  scale,   contribute  to the generation  (or  inhibition)  of  the
radicals  responsible  for  converting sulphur  oxides to sulphuric  acid  and
nitrogen  oxides  to  nitric acid,  and are  involved  in  the  formation  of
peroxides.  VOCs differ significantly  in  their  effects  on these processes,
and  the model must be  able  to represent these differences  in  reactivity
with  a  reasonable  degree of accuracy.   This  is  a difficult problem,  not
only  because  there remain significant uncertainties  in  our knowledge  of
the  atmospheric  reactions of most types of VOCs,  but also because  the many
hundreds  (if  not thousands)  of different VOCs  emitted  into  the atmosphere
must  be  represented  by  only  a  limited number  of  model species.   For
example,  the  1985  NAPAP  inventory  for anthropogenic emissions of VOCs in
the  United States  (Wagner et  al.f  1986; Shareef et  al.,   1988) include
emissions for over 600 different chemical  compounds,  while  the  gas-phase
mechanism proposed for  RADM-II  has only  17 model  species  which are used to
represent them.  The  methods used to  represent these hundreds of compounds
with the limited  number  of  model   species  is   therefore  an   important
component of  the gas-phase mechanism in RADM.
      As part of the  most recent update  of  the SAPRC  gas-phase  mechanism
for  oxidant models,   we had  previously,  under  funding  from  the California
Air  Resources Board,  developed new VOC emissions processing procedures for
representing  in  such  models  the  large numbers   of  chemical species  in
detailed emissions  inventories (Carter et al., 1988a,b).   These procedures
were designed to  utilize, as much as possible,  the  chemical information
available in the detailed VOC  emissions  inventories to derive the  kinetic
and  mechanistic  parameters for the model species.  As  part of  this  effort,
kinetic and mechanistic  parameters  were assigned  to  the major categories
of compounds used in current  detailed emissions inventories.  In  addition,
emissions processing software  was developed to  utilize  these assignments
 to derive the parameters for  the lumped model species which best  corres-
pond to  those  of the  emissions  profile  being represented.   Although  the
                                     57

-------
complexity  of this new emissions processing system is such that it is not
practical   to  utilize  it  in  its  current  form  in   the  RADM  model,  the
procedures, data  bases,  and  software  developed  in   this  effort  could be
utilized  to recommend  appropriate  methods  for representing VOC emissions
in RADM.
     Therefore,  as part of this gas-phase mechanism evaluation  program for
RADM,  we  assisted  the RADM  team  in  the  development of  procedures for
processing  and  representing  VOC emissions for RADM-II.   This effort has
involved:    (1)   working with  Dr.  Paulette  Middleton of  the RADM team in
the  development of  a new emissions processing and aggregation system for
regional  models which  can  be used  with a variety of  chemical  mechanisms;
(2)  working  with  Stockwell  and  Middleton  to  determine  how   best  to
distribute   the  various VOC  compounds  in  emissions  inventories among the
species used  in  the RADM mechanism;  and  (3) working with Stockwell in
deriving the kinetic and mechanistic parameters for  the lumped  VOC species
in  the RADM mechanism which  best  represent  those of the sets of emitted
species they  are  intended  to  represent.   The results  of  the first two
efforts are discussed  below  in Section 3.2,  and  the results of the  third
are  discussed in Section 3-3-

3.2   Aggregation of VOC Emissions Input for  RADM
      The current  NAPAP  inventory  of VOC emissions  in  the  RADM  modeling
region gives  mass emissions  of over  550 types of chemicals  from several
 thousand different  source  categories  throughout  the  United  States.   These
detailed emissions  data are  not input  into  the  RADM model  directly,  but
 instead must  be aggregated into the  much smaller  number  of VOC categories
 which are  utilized  in the model.  Because of the extremely large size  of
 the data bases  involved,  this is a very expensive procedure.  In the past,
 the procedure has been  to aggregate  the detailed VOC emissions  data into
 the species used by the model  in a single  step.   Since  the number  and
 classification  of  VOC  species in  a  model  in  general  depends on  the
 chemical mechanism  used,  the  aggregation procedure  must be  carried  out
 separately for  each chemical  mechanism, and  once this procedure  has been
 carried  out,  the representation of  VOCs used in the mechanism  cannot  be
 readily modified.

                                     58

-------
     Because there are currently available a number of chemical mechanisms
which can  be  utilized  in regional  models,  each in general using differing
methods for aggregating  VOC emissions,  it would be advantageous to develop
a procedure  to  aggregate the detailed  emissions  data into a more manage-
able data  base  which still  contains  all the  information  required by  the
different  mechanisms,  including  mechanisms which may be developed in  the
future.    To  address  this,   Middleton  proposed  developing  a   two-step
emissions  aggregation process  for  the RADM model, where the  detailed  data
are  first  aggregated  into a  manageable  number  of  emissions categories
which can  be  utilized  by a variety of chemical mechanisms, and then  these
emission categories  are  further aggregated  into the smaller  set of  VOC
species  which are  directly  input  into  the  RADM  mechanism.   Since,  in
principle,  only  the second,  much  less  expensive,  aggregation  step  is
necessarily  mechanism-dependent, this  means  that  this  data base can  be
more  readily  utilized  by  alternative  chemical  mechanisms  either  for
different   types  of  applications,   or  for  mechanism   intercomparison
studies.    The  development of  the  mechanism-independent   VOC  emissions
classification  scheme  as  it  will  be  used  for  RADM,  and  recommended
extensions to this  scheme which  are not yet implemented,  are  discussed
below.   Following that,  the  procedures developed  for further aggregating
these emissions  into RADM  model  species are  discussed.
     3.2.1 Development  of a Mechanism-Independent  VOC Emissions
            Classification System
            The  primary  objective  in  the   development  of  a  mechanism-
 independent   VOC  emissions  aggregation  scheme   is  to  utilize a  minimum
number  of  chemical  groupings while  retaining,   as  much  as  possible,  the
 types  of  chemical   information  needed to represent  the  various  types of
 relevant  impacts that VOCs might  have on model  simulations.  These types
of   impacts,  and  their magnitudes  on  model  outputs  of   interest,  are
 referred  to  under   the  general   term  of  "reactivity,"  where  relevant
measures of reactivity depend on the application of  the model.  Therefore,
 the objective   in  developing a general  emissions  classification  scheme
 would  be  to  organize the VOCs  according to their  reactivity  where each
 class   has  compounds  which   are   similar  in  all   aspects   of  reactivity
 important  for  the  model  application.   Since  the  aspects  of reactivity
 which are of greatest interest  depend  on  the  application  of  the model, the
                                     59

-------
most appropriate aggregation scheme will also be dependent on the intended
application of the data.
     In  practice,   if  the  number  of  categories  which  can  be used   is
limited, compounds with different reactivity characteristics would have  to
be  lumped  together  in  the  same  category.   Given a set number  of  cate-
gories,  the extent  to which  reactivities should differ within  a  given
class  would  depend on how  important  the compounds  are expected  to  be  in
affecting the model,  which  in turn depends both on how much is emitted and
on how  reactive they  are  (i.e., on the magnitude of  the effects of a  given
amount  of the VOC  on  the  relevant predictions of the model).  For example,
reactive compounds  which  are emitted in very large  amounts  (such as, for
example,  ethylene) should  be  in  classes by  themselves,  and  categories
representing groups of  relatively reactive types of  compounds with signif-
icant  amounts  of  emissions  should  represent relatively  narrow ranges  of
reactivity.   On  the other  hand,  it is  not worthwhile  to  have separate
classes for  reactive  compounds  of very  low emissions; they can be grouped
into classes which are closest  to them  in reactivity, even if the classes
are not particularly  close.  Furthermore, compounds which are essentially
unreactive  for  the  purposes of  the  particular model  application  can  be
lumped  together into  a single "inert" category, regardless of how much may
be  emitted.   However, a compound which  is  "inert" in one application may
be  reactive in  another,  so  if the categorization scheme  is to be used  for
several different  applications,  there may be more than one category  which
is  treated  as "inert"  in a  given type of application.
     In the case  of aggregation of emissions for models  such as RADM,  the
relevant reactivity  criteria concern reactivity  with  respect  to  oxidant
formation  and  reactivity with  respect to  formation of acidic  species.
Reactivity  with respect to oxidant formation is influenced by how  rapidly
the VOC reacts in the atmosphere, and,  if it reacts sufficiently  rapidly,
by  the nature  of  its reaction mechanism.  Therefore, compounds  which react
sufficiently  rapidly  that they  participate  in  the  oxidant  formation
process, yet react at  significantly  different  rates, should be placed  in
different  reactivity categories, while  compounds  which do not react to a
significant extent in the  time  frame of the model  application can  all  be
lumped together and be classified as "inert."  If a  VOC reacts,  the nature
of  its  reaction mechanism  is also  important, since  different aspects of a
                                     60

-------
VOCs reaction can affect  the system in different ways.   For example, the
reactions of a VOC  can  have  a relatively  direct effect on ozone  formation
by forming  radicals which  convert  NO to NOo (which causes ozone  formation
in the presence of sunlight), and more  indirectly (but often more signifi-
cantly)  by  enhancing  or  inhibiting levels of  the  radicals which  control
the overall rate of the oxidant formation  process, and by  the formation  of
reactive organic products.   For this reason, the emissions  categorization
scheme  must not  group  together  types of  compounds  with significantly
different types of  reaction  mechanisms, even if they may  react at similar
rates.   At a minimum,  the  classification  scheme  should  include  separate
categories  of compounds  which  are  represented  with  different  sets  of
reactions  in  current oxidant formation mechanisms.    Finally,   since  the
model  is  concerned  with formation of acidic species, VOCs which  are  acids
themselves  should  not be  lumped  into the same categories as those  which
are not.
     Ideally, it would  be desirable that  the emissions  aggregation scheme
used for  RADM and  other oxidant and acid  deposition models also be  suit-
able  for use with  models concerned  with  global effects  and the strato-
sphere.   This would  require the  use of  additional  VOC categories.   For
global models, reactions  occur on a much  longer time scale, and  thus many
compounds classified  as "inert" for regional model  applications  would have
to  be placed  in  separate  "reactive"  categories  for global  models.   In
addition,   for models  where the  stratosphere  is  of  interest,  separate
categories  would have to  be  created for halogen-containing compounds which
are  transported to  the  stratosphere.
     Unfortunately,  current  limitations  in the  software  utilized by  the
EPA  and its contractors  for processing emissions input for  RADM restrict
the  number  of categories  to 32.   Our analysis indicates  that  this may be
sufficient  to convey  the  necessary  reactivity  information required  for
local  and  regional  oxidant and  acid deposition  models  using  currently
available  chemical  mechanisms,  but  is not sufficient to convey the necess-
ary  additional  information   required  by  global and  stratospheric models.
Therefore,  it was  determined that  the  initial emissions aggregation scheme
to be developed for  RADM would not contain  the  additional  VOC   categories
required by such models.
                                     61

-------
     An analysis of  the  total  emissions  into  the  contiguous  United States
of the  approximately -550 different  types of VOCs in  the  NAPAP  inventory
was  carried  out  to  determine  the  most  appropriate  aggregation  scheme,
involving no more than 32 categories, to use for regional oxidant and acid
rain  models.    This  was  carried out  jointly  by  SAPRC as  part  of this
program  and  by  Middleton  and  co-workers  at  NCAR as  part  of  the RADM
development effort.   NCAR determined  the major overall characteristics of
the  classification  scheme  and  the  types of organics  that had  to  be
represented  in  separate  classes  because of their   importance  in  the
emissions inventory; they also had the final responsibility for the  system
which  was developed.  SAPRC provided  guidance to  NCAR  in  this  task by
analyzing the  reactivity characteristics of the  various  types of  VOCs in
the  inventory, and  making recommendations  as to  how species  should be
classified based  on  their reactivity.   This analysis utilized the kinetic
and  mechanistic assignments which were made for emitted VOC categories as
part of its development  of  the new emissions processing system for the new
California ARB model (Carter,  1988),  together with new assignments which
had  to  be made as part of this program for a number of major VOC  categor-
ies  which are  on  the NAPAP, but not the California inventories.
      A   number of   emissions  categories  were  assigned  without   direct
consideration  of  reactivity,  but based on considerations of  importance of
individual compounds in  the total mass emissions, or on considerations of
special chemical  characteristics.   In addition,  special  categories  were
created for  emissions where there is  information missing  concerning their
chemical characteristics.   These are as follows:
      (1) On  the basis  of  total amounts of mass emitted  into the  United
States, combined  with considerations of  their  special  chemical  character-
istics,  eight individual  compounds  were  determined  to  be sufficiently
important that  they  should  not  be  aggregated  with  other  species,  but
should  be  in categories  by  themselves.    These  were  methane,   ethane,
propane, ethene,  propene,  acetylene,  formaldehyde, and  acetone.   Initial-
 ly,   it  was   also   proposed  that benzene,  methanol,  ethanol,   isopropyl
alcohol, perchloroethylene, and  methyl  chloroform be  in their own cate-
 gories  because of  large  current or  (in the  case  of methanol)  possible
 future  emissions; but  it  was decided that given the  32-class  limitation
 there  were an insufficient number of classes  available  for  this  purpose.
                                     62

-------
These  compounds  were  lumped  with other  species either  because they  are
relatively  unreactive,  or  because  their reactivities  were  judged to  be
fairly  similar  to those  of other species.   Note, however,  that  many  of
these compounds dominate the categories in which they were placed.
     (2)  Organic  acids  were   lumped  into  their  own  separate  category
because  of their  obvious role  in  acidification.  Their contribution  to
reactivity  with   respect  to oxidant  formation  is minor,  so  it  was  not
necessary  to differentiate  them further  on this basis.   Although the  RADM
mechanism  represents formic acid  separately  from the  higher  organic acids,
the  emissions  of formic  acid  in the NAPAP  inventory are insufficient to
justify  utilization  of  a separate  emissions  category  grouping  just  for
this compound.
     (3)   Two aggregation categories were  created for "chemical compounds"
listed  in  the  current  detailed NAPAP (and California)  emissions inventor-
ies  which  are  not   true  compounds at  all,  but  represent  unspecified
mixtures  of  compounds  of  ambiguous  reactivity.    The  category  "Alkane/
Aromatic  Mix"  was  used  to represent emissions which  are   designated as
"Naphtha,"  "Mineral  Spirits,"  or "Lactol  Spirits."    These  are solvents
believed  to consist  primarily  of Cg^  alkanes and  aromatics,  but  whose
exact   chemical   compositions   are  unspecified.    The  category   "Alkenes
(Primary/Internal  Mix)"  was used to represent  emissions  designated as "Cn
alkenes"  (where  n>^ 4),  without  reference  to the  location   of  the double
bond.   The  location  of  the  double bond  significantly affects  the reac-
tivity of  the  alkene,  and is a necessary basis for the reactivity  classif-
ications of these compounds.
      (4)   An "unreactive"  category  was  included  for compounds  which are
estimated   to  react  too  slowly  to  be  important  in  affecting  chemical
transformations  on the regional scale.  Note, however, that many compounds
in  this   category  are  not  "unreactive"  on  a  global  scale  or  in  the
stratosphere.
      (5)  An "unidentified" category  was included for  unidentified VOCs.
These  constitute somewhat less than 6J of the mass in the NAPAP  total U.S.
anthropogenic  emissions inventory.
      (6)  An "unassigned" category was  included for  those VOCs which  were
 identified, but  for  which no reactivity  assignments  or classifications
were made.   As  indicated above,  as part  of  this  and our previous ARB
                                     63

-------
program  (Carter,   1988)  we  made  mechanistic  classifications  and  (where
applicable)  reactivity  assignments  for  most  of  the  detailed emissions
categories in the  California and  NAPAP  inventories.   However, a number  of
compounds  in  the  NAPAP  inventory  were  not categorized  because they  both
were very difficult to classify and had small total U.S. emissions.   These
"unassigned" compounds constitute  less  than 1/t  of the mass  in the  NAPAP
total U.S. anthropogenic  emissions inventory.
     The above amounts to  a  total  of 1U categories, leaving  18 categories
available for all other types of compounds  in a  32-category system.   Based
on considerations of differences in chemical reaction mechanisms,  together
with  considerations of  total  amounts  of  different  types  of chemicals
emitted,  it  was  determined  that  the  remaining  categories  should,  at  a
minimum,  differentiate  between the  following  general  types  of chemicals
(listed in order of mass  emitted):

      •  Alkanes (other than  methane, ethane,  and propane)
      •  Aromatics
      •  "Others" (alcohols,  ethers, alcohol ethers, esters, and other
         types of compounds which are generally represented as reacting
         analogously to alkanes in current mechanisms)
      •  Alkenes (other than ethene and  groups of chemicals classified as
         "primary/internal mix")
      •  Haloalkenes
      •  Aldehydes  (other than formaldehyde)
      •  Styrenes
      •  Phenols and Cresols

There are a number of other chemical  classes emitted,  but  it was  judged
not  to  be worthwhile  to  assign them to separate groups because either the
amounts  emitted  were too  small  or  they were  judged   to  be  of  minor
reactivity  with respect  to  oxidant  or  acid formation,  and thus  could  be
classified as "inert."
      In   the  case of   the   first  four   groups  listed   above  (alkanes,
aromatics,  "others,"  and alkenes),  the  variety of rates of reaction  of the

-------
individual species are  such  that  these groups needed to be further  broken
down on  the  basis of  differences in  reaction  rates.   At  a minimum,  the
numbers of reactivity  categories  for each type of chemical should be  such
that there  are at  least as  many as  the  number of  lumped species  which
might be  used  for them  in  the models, and  the  criteria used to make the
reactivity differentiations  should also be  compatible  with those used  in
the models.  Based on  these  considerations,  the  reactivity  classifications
were derived as follows:
     Alkenes:  Most current  mechanisms differentiate  between  ethene, other
terminal alkenes, and  internal  alkenes.   This differentiation is therefore
also used in the  emissions categorization.   As  indicated above,  ethene and
propene are  given emissions  categories by themselves,  so  the differentia-
tion of  the  mechanisms between ethene and  other alkenes is  already taken
care of.   The terminal  alkenes   other  than propene are  lumped into the
group designated  "alkenes (primary),"  and  the more  reactive  alkenes with
internal double bonds are lumped  into  "alkenes (internal)."
     However,  certain  "terminal"  alkenes, such  as  1,3-butadiene and beta-
pinene  have atmospheric reaction rates  which  are  more  typical   of the
internal  alkenes,  and  lumping  these into  the "primary  alkenes"  group
seemed  to us to  be  inappropriate.   To provide  an objective criterion to
determine  the  classification of  these more  reactive alkenes,  all  alkenes
(other  than ethene  or  propene)  which  react with OH  radicals  with rate
                              it      1      1
constants  less than 7.5 x 10  ppm   min   were assigned  to  the "primary"
group,  and  alkenes  with  higher  rate  constants  were assigned   to  the
"internal" group.   This rate constant lies well between the typical ranges
for "ordinary" primary and  internal  alkenes,  but  results in the unusually
reactive  "primary"  alkenes  such as beta-pinene being grouped with the more
reactive  internal alkenes,  which have more similar atmospheric reaction
rates.
     Note that  this  classification   for  the alkenes  is  appropriate for
inventories  which  contain  anthropogenic  emissions only.    If biogenic
emissions are  to be merged  into  the same scheme,  then  separate  categories
should  be reserved for  them.   This  is particularly important for regional
modeling  applications, since biogenics tend  to dominate the.mass emissions
for large regions of  the modeling domain.   The biogenic emissions  inven-
 tory includes species  such  as isoprene, alpha-pinene,  other terpenes, and
                                     65

-------
unidentified biogenics (Middleton, private communication, 1988), and, at a
minimum, separate  categories  for  each of these should  be  included in any
merged  emissions   grouping  system.    However,  in  this  discussion  it  is
assumed  that  biogenic emissions  are  processed separately  from anthropo-
genic emissions, and  thus separate categories  for  biogenic emissions are
not included in this scheme.
     Cnt  Alkanes:    The  RADM  mechanism uses  three different  species  to
represent  the  C^+  alkanes.    Other  mechanisms use  fewer species for this
purpose,  and  unlike  the  case for  the  alkenes,  the- mechanisms generally
differ on  how  the  lumped  groups  of alkanes  are  defined in the model.  To
allow  for  maximum  flexibility   in  use  with  mechanisms   with differing
aggregation schemes for the  alkanes,  four different reactivity groups are
used  for  the  emitted C^+  alkanes,  with  (as  indicated  above) methane,
ethane, and propane being  in groups by themselves.
     The criterion used in  this  aggregation scheme to determine the  reac-
tivity classifications for the alkanes  is based solely  on how  rapidly they
react  in  the  atmosphere.   This is an important  factor in measuring  reac-
tivity, since  it determines how much of the  alkane  which is emitted under-
goes reaction  and  thus participates in  chemical transformations.   However,
the  rate  of reaction is  not  the only  factor  affecting the reactivity of
alkanes; aspects such as  the alkyl nitrate  yield in  the reaction is also
important  (Carter  and Atkinson,  1985,  1989).  Nevertheless,  at least to
some extent the nitrate  yields and other aspects are correlated with  the
size  of the  molecule,  which  in  turn  is  correlated with  how rapidly it
reacts.
     Alkanes are consumed  in the atmosphere primarily by reaction with OH
radicals,  but  the  rate  at which they react  can vary significantly depend-
ing  on the size of  the  alkane,  and to a  lesser extent on its structure.
For  example,  the  rate  constants  for  n-butane  and  n-pentadecane,  both
important   C4+ alkanes  in  the   NAPAP  inventory,  have  OH  radical rate
constants  which differ  by  approximately  an order  of magnitude.   2,2-
Dimethylbutane and cyclohexane have  the same number  of carbons, yet  the
latter  reacts more than  twice as rapidly as  the former.   Most other  C^
alkanes  with  significant  emissions  in  the NAPAP  inventory  have  rate
constants  between  those of n-butane and n-pentadecane.
                                     66

-------
     Based on an analysis of the amounts of alkanes with various  ranges of
rate constants  in  the  available NAPAP total  U.S.  inventory,  the four C^
alkane  reactivity  classes are  defined as  follows (where  kOH  is  the OH
radical rate constant at T=300 K in units of ppm   min"1):
     (1)  The group  designated  "alkanes (<0.5  react)," consists  of all
alkanes  where  kOH  is   less  than  5 x  1CT  ppm"   min  .   This  group  is
dominated by n-butane and isobutane, but also contains 2,2-dimethylbutane.
     (2)  The group  designated "alkanes  (0.5-1  react)," contains  alkanes
with kOH between 0.5 and  1 x  10  ppm   min   .   It  is  dominated  by pentanes
and  hexanes.   This  is  the second  largest  group  of  alkanes  in  the  NAPAP
inventory  in terms of  estimated amount  reacted in regional model  simula-
tions (see below).
     (3)  The group  designated "alkanes  (1-2  react)" consists of  alkanes
                              U     1     1
with kOH between  1 and  2  x 10  ppm   min"1.  It is dominated  by n-heptane
and  other  C^-C^Q  isomers.  This is the  largest group of alkanes in  terms
of estimated amount reacted in regional model simulations.
     (4)  The group  designated  "alkanes (>2 react)" consists of  alkanes
with kOH greater  than 2 x 10   ppm    min.   Almost HQ% of  the  emissions of
this  group is  accounted for  by n-pentadecane, with the  remainder  being
various  C11+ acyclic or CQ+ cyclic alkanes.   Although this  is  the smallest
group of Cjj+ alkanes in terms of estimated amount reacted,  it  still  makes
a  non-negligible  contribution to  the  emissions,  consisting of  over  3£ of
the  mass of VOC emitted.
     Aromatics:    The   aromatics,  like  the  alkanes,   are  consumed in  the
atmosphere  primarily  by reaction with OH radicals,  and it is  the  rate of
this reaction which  was used  to define their reactivity classes.  However,
in  terms  of  distribution of  reaction  rates  and  fractions  of  emitted
aromatics  estimated  to react  in regional model simulations,  the aromatics
tend to fall  into three  distinct groups.    These groups  also  correspond
reasonably well  with  how they  are aggregated in current  lumped kinetic
mechanisms.  These reactivity  classifications are  as  follows:
      (1)  The  lowest reactivity  group  is  dominated   by benzene, but also
 includes several  halobenzenes which are  also in  the  NAPAP inventory.  The
halobenzenes react  somewhat  slower than benzene, but not so  slowly that
 they should be considered unreactive.   Since  there  are not  enough halo-
benzenes emitted   to  Justify  placing them  in  their own aggregation  group,
 they are lumped with benzene.

                                     67

-------
     (2)  The middle  reactivity  aromatics group  is  designated "aromatics
(<2  react)," and  is  defined as   those  aromatics  which  react  with  OH
radicals more rapidly than benzene, but with rate constants of less than 2
x  10  ppm    min   .   The compounds  in this group  have a relatively narrow
range of kOH values,  being  around  0.8-1.2 x 10   ppm   min  ,  and consist
entirely of  toluene and other monoalkyl benzenes.  Toluene is the dominant
compound  in  this  group,  accounting  for over  80%  of the  mass of  such
compounds emitted in the NAPAP inventory.
     (3)  The most  reactive group  of  aromatics  is  designated "aromatics
(>2 react)," and is defined as those aromatics which have  kOH greater than
2  x  10  ppm" min   .   This  group consists primarily of di- and poly-alkyl
benzenes, but it also includes naphthalenes  and indans, which account for
approximately 5J  of  the  mass emissions into  this group.   This  is the
largest group of aromatics,  both in terms of mass  emissions and in  terms
of estimated amounts  reacting in a  model simulation.  Almost 10/t of the
estimated number of moles of VOC reacted  in a model  simulation corresponds
to compounds in this group.
     It should be noted that had it been  practical  to  include more than 32
groups  in this total  aggregation scheme,  the most reactive aromatics  group
would have  been  broken down  further with dialkylbenzenes, tri- and  poly-
alky Iben2enes, and naphthalenes and  indans each being  in  their own groups.
This would  permit  the scheme  to  be used with the most detailed version of
the  SAPRC/ERT mechanism  described  by  Lurmann et  al. (1987),  which has
separate  model  species  for xylenes and  trialkylbenzenes.   In  addition,
smog chamber studies  (Carter et  al., 1987) indicate that  the  naphthalenes
are  much less  reactive in  N0v-air irradiations  than the  di-  and  poly-
                              A
alkylbenzenes,  despite their high OH  radical  rate  constants.   However,
this  grouping corresponds  to the  lumping  schemes  used   in  the  RADM and
other chemical mechanisms in  current regional models,  which  generally lump
all  polyalkylbenzenes with  the xylenes.   The emissions of the  naphthalenes
and  the indans  are not sufficient  to justify creating a separate  group for
them in this 32-class scheme.
     "Others":   Compounds  such  as alcohols,  ethers,  alcohol ethers, and
other  compounds (primarily 0-containing  species  which are  not  aldehydes,
 ketones,  organic  acids, and which  do not contain  aromatic rings  or  carbon-
 carbon  double bonds)  are lumped  into  the  miscellaneous  group  designated
                                     68

-------
"others."   All these  compounds  are believed  to react  in  the atmosphere
primarily  with OH  radicals,  and  current  kinetic  mechanisms (including
RADM)  generally  lump  them  with  the  alkanes.   There  is  an  insufficient
number of available groups  in  the  32-class scheme to  represent  the  variety
of chemical  types of  this  group separately,  so  they are  aggregated  into
four reactivity classes  based  on their OH radical rate constants,  using  a
method  similar to  that  used  for  the  alkanes.   These  categories  are  as
follows:
     (1)  Compounds with OH radical  rate constants  less  than  2.5 x  10^
ppm    rain"'  are  lumped  into the group designated "others (<0.25 react)."
The major  compounds in this group are  methyl  and ethyl acetates (-51J  of
the emitted  mass  in this group) and methanol  (-40% of the  emitted mass).
Emissions  of this  class are  relatively minor  in  the current NAPAP  total
U.S. inventory (consisting  of  less than 0.4/1 of the  mass emitted and  less
than Q.2%  of  the  estimated moles  reacted),  but these  compounds  are  too
reactive to  be treated  as  "inert,"  and too  unreactive to  be  lumped  into
the next highest  reactivity "others" group.   In addition,   in  view of the
interest  in  using  methanol as  an alternative  fuel,  emissions  into  this
class  may become much  more  important in  the future.
     (2)  Compounds with kOH  values  between 2.5  and approximately 5.0 x
10^ ppm"1 min"1 are lumped  into the group "others (0.25-0.5  react)."  This
group  is dominated by ethanol, which accounts for over 85?  of the  mass of
this group  in the  NAPAP total  U.S.  inventory.   It  is  also  non-negligible
in  terms of  total emissions,  accounting  for almost  4%  of  the  estimated
moles  of  VOC  reacted  in   the  model  simulations.    [Subsequent  to  the
development  of this  classification system,  we  revised our assignment of
the  OH radical rate parameters  for ethanol to  be  consistent with the most
recent recommendation  of  Atkinson  (1989).    This  yields  a  300   K  rate
constant which is slightly  higher (by less than }%) than the nominal upper
limit  rate  constant for this group.  However, ethanol was  not re-assigned,
since  otherwise there would be no significant  emissions for  this  group.]
      (3)   Compounds with kOH  values between  approximately  0.5  and  1.0 x
 10   ppm   min   are  lumped  into the  group  "others (0.5-1   react)."   The
major  compounds  in this group are isopropyl alcohol  (-60J of the mass) and
propyl and  higher  acetates (-25%).   This group accounts  for an estimated
4-5/t of the  VOC mass  reacted in model  simulations.
                                     69

-------
     (4)  Compounds with kOH values higher than 1.0 x  1CT ppm'1 min~   are
lumped  into  the group "others (>1  react)."   The  major compounds  in  this
group consist  of glycols and  cellosolves and  other  alcohol ethers.   [It
also has  very  minor contributions by alkynes, allenes, and  amines,  which
react within this rate constant range.  In our opinion, these are  not well
represented  by  any  of  the other  chemical  classes  in  this  aggregation
scheme  (for example,  the  alkynes  and  allenes are  not  lumped with  the
alkenes  since  they do  not  react significantly with  ozone),  and  are  not
emitted  in sufficient  quantities  to justify  being  in their  own  group. ]
The  emissions  of  compounds in this  group  are small  but  non-negligible,
accounting  for  approximately  1*  of the  mass emitted  and  1-2* of  the
estimated moles  reacted in  the total U.S.  inventory.
     Summary of Emissions  Groups:    Table  7 gives a  summary of  all  the
groups  in this 32-class aggregation  scheme, and the  total  mass and molar
emissions  of compounds in  each  group of the  current  NAPAP  inventory  for
total anthropogenic  emissions  in the contiguous United States.  Note that
the  detailed  emissions data  bases  give  mass emissions,  but these  are
aggregated  into the 32-class scheme on a molar basis, with  the number  of
moles  of  each  type  of emitted  compound  being  calculated  based  on   its
molecular  weight to determine  its  contribution to the  aggregated emissions
group.   The table  therefore also shows the  percent  contribution of each
group  to the total mass and molar emissions.  Finally, to give an indica-
tion of the relative importance  of the  emissions of each category  on model
simulations,  Table  7  also gives  estimates  of  the  number  of  moles  of
compounds in each group that  is estimated to undergo  chemical reaction  in
a  2-3  day RADM simulation.  These  latter estimates were made as  indicated
in footnotes to the table,  and as discussed  in the following section.
     It can  be  seen that the  most  important  group  in  terms  of moles
emitted is methane, accounting for almost kQl of the moles emitted  in  the
inventory,  but  it  accounts for  less than 0.2% of  the moles reacted.   In
terms   of  moles reacted,   the  most  important  single  group  is  ethene,
accounting for  almost  21/1  of the  moles reacted, followed  by the  middle
reactivity C^+  alkane  group and the high  reactivity group of aromatics.
Alkanes, aromatics,  and  alkenes account for  -80J,  aldehydes and  ketones
for -1%, and all other types  of compounds for -11J of the  total  estimated
number of moles of VOC reacted.   As indicated  in Table 7, approximately  1%
                                     70

-------
Table 7.   Summary of the Emissions Groupings Using the 32-Class Scheme,
          Amounts of Each Emitted in the NAPAP Total U.S. Inventory, and
          Estimated Amounts Reacted in Conditions Representative of RADM
          Applications
• •

C1-C2 Alkanes
Methane
Ethane
Total
Higher Alkanes
Propane
Alkanes (<0.5 react)
Alkanes (0.5-1 react)
Alkanes (1-2 react)
Alkanes ( >2 react)
Alkane/ Aromatic Mix (b)
Total
Ethene
Higher Alkenes
Propene
Alkenes (primary)
Alkenes (internal)
Alkenes (mix)
Total
Aromatics
Benzene, Halobenzenes
Aromatics (<2 react)
Aromatics (>2 react)
Phenols and Cresols
Styrenes
Total
Aldehydes and Ketones
Formaldehyde
Higher Aldehydes
Acetone
Higher Ketones
Total
Miscellaneous
Organic Acids (c)
Acetylene
Haloalkenes
Total

M.Kg

2.83
0.46
3.29

0.12
1.40
2.03
2.49
0.60
0.25
6.89
1.27

0.44
0.33
0.76
0.20
1.73

0.60
1.56
2.03
0.05
0.08
4.31

0.32
0.17
0.21
0.20
0.89

0.04
0.36
0.52
0.92
A rn/*"M i n V c
rulivJlln Ui

12.18
1.98
14.16

0.53
6.01
8.72
10.71
2.60
1.06
29.63
5.45

1.91
1.42
3.28
0.85
7.46

2.59
6.69
8.72
0.19
0.36
18.55

1.37
0.71
0.90
0.85
3.83

0.17
1.56
2.24
3.97
iPm i f ^ ^H
dlu A. w v\?U
K.Mole

176.7
15.3
192.0

2.8
23.9
25.5
21.9
3.1
1.8
78.9
45.2

10.6
4.6
11.0
1.9
28.0

7.3
16.1
16.3
0.3
1.5
41.5

10.7
2.5
3.6
2.2
19.0

0.3
14.0
2.1
16.4

*

38.42
3.33
41.75

0.61
5.20
5.53
4.76
0.67
0.38
17.16
9.82

2.29
1.00
2.38
0.41
6.09

1.58
3.51
3.54
0.06
0.33
9.02

2.32
0.54
0.79
0.48
4.12

0.07
3.04
0.46
3.57
PAO /%f- Arl
rvcaC wCU
K.Mole

0.2
0.7
0.9

0.5
8.0
13.2
16.7
2.8
1.6
42.9
33.9

10.5
4.6
11.0
1.9
27.8

1.4
10.2
16.1
0.3
1.0
29.0

8.2
2.3
0.1
0.3
11.0

0.0
1.7
0.1
1.8
( a ">
V a )
n

0.15
0.41
0.55

0.31
4.86
8.06
10.19
1.73
0.98
26.13
20.65

6.37
2.78
6.68
1.13
16.96

0.84
6.23
9.81
0.17
0.58
17.64

4.98
1.41
0.08
0.20
6.68

0.00
1.01
0.04
1.05
                                                                (continued)
                                      71

-------
Table 7 (continued) - 2
  Emissions Group
	  Amounts Emitted
 M.Kg    %      K.Mole
                         Reacted (a)
                          K.Mole    %
"Others"
  Others (<0.25 react)
  Others (0.25-0.5 react)
  Others (0.5-1 react)
  Others (>1 react)
Total

Unreactive

Unknown/Unassigned
  Unidentified
  Unassigned
Total
0.10
0.80
0.88
0.30
2.08
0.35
0.42
3.44
3.76
1.28
8.90
1.50
2.0
16.7
12.5
3.3
34.7
4.3
0.43
3.63
2.73
0.72
7.53
0.94
0.3
6.3
7.7
2.7
16.9
0.0
0. 17
3.85
4.67
1.66
10.34
0.00
1.36
0.16
1.52
5.83
0.72
6.55
 (a)  Estimated assuming reaction with OH radicals is the main
      consumption process for compounds in the groups, and assuming that
      the ratio of reacted to emitted compounds can be estimated by
      1 - exp (-kOH x INTOH), where kOH is the average OH radical rate
      constant for compounds in the group, and INTOH is a parameter
      reflecting the effective integrated OH radical levels in the model
      simulation (Carter and Atkinson, 1989).  The tabulated values are
      calculated assuming INTOH =110 ppt-min, based on analyses of
      results of previous RADM model simulations (Stockwell, private
      communication, 1988), and the kOH values are weighed averages of
      those assigned for compounds in the groups (Carter, 1988).  See
      Section 3.2.3.

 (b)  These are estimated to consist of approximately 90? alkanes and 10J
      aromatics (on a molar basis), so for the purpose of this summary
      they are counted as alkanes.

 (c)  Amount reacted using kOH of acetic acid, the major emitted organic
      acid in the NAPAP inventory.
                                     72

-------
of the mass in the current NAPAP total U.S.  inventory  is either  unknown  or
unassigned.
     3.2.2  Recommended Extensions to the VQC Emissions Classification
            System
            The  32-class  VOC  classification   system described  in  the
previous  section  is  the version which is being  implemented  for use with
the  RADM-II model.   Although  we  consider this  system to be  satisfactory
for  this  purpose,  we  have  several  recommendations,  discussed  below,  for
extending  its  utility  for  use with other  mechanisms besides  that  which
will  be  used  in  RADM-II,  and  for  other  types  of models.   All of  these
recommendations  require updating  the  software  currently  used  for  aggre-
gating detailed emissions data.
      Use  with the Carbon Bond  Mechanism.   The emissions aggregation system
discussed above  was designed  primarily for  use with mechanisms,  such  as
that used in  RADM,  which represent reactions of  emitted VOCs  on a molecu-
lar  level,  where,  in general,  only one model species  is used  to represent
the  reaction  of any type  of emitted  molecule.   However, the  Carbon Bond
mechanism (Whitten  et  al.,  I960;  Gery et  al.,  1988),  which  is widely used
 in urban  airshed modeling applications, represents  portions  of some types
of  molecules  reacting  independently, and  thus  requires at   least some
portions  of the emissions to  be  aggregated in  terms  of  reactive portions
of molecules,  rather than  molecules as a whole.  For that reason, this  32-
class system,  as described above,  cannot  be used in aggregating emissions
 input for models using  the  Carbon Bond mechanism.   However,  as indicated
 below,  this limitation can be  readily removed  by adding two new emissions
 categories.
      Although  as  originally   conceived   the Carbon   Bond  mechanism   was
 radically different from molecularly-based  mechanisms in  its treatment of
 parts of  molecules as reacting independently, in recent years the need  for
 improved  performance  in  simulating  results of chamber experiments  has
 required  it to evolve  more  towards the molecularly-based mechanisms, with
 the latest version  (Gery  et  al.,   1988) using almost  the  same  numbers  and
 types of species to represent alkenes,  aroraatics, and oxygenates as other
 state-of-the  art mechanisms.   The  principal  remaining difference between
 the current version of Carbon Bond and the molecular-based mechanisms,  is
 the use  of a  single-carbon model  species called "paraffin  bond" (PAR)  to
                                     73

-------
represent both the alkanes and the "extra" carbons on molecules which have
more carbons  than  the species in  the  model  representing them.   In addi-
tion, the double bonds  in  some diolefins  and other bifunctional compounds
are  represented  as reacting  independently.   The  present version  of the
emissions aggregation system does not contain the information necessary  to
determine the  levels of  the  PAR  model  species  or  the  contributions  of
bifunctional compounds to levels of alkene model  species.
     The  deficiency  with  respect  to aggregation  of the  "paraffin bond"
species can  be  addressed by adding  a  separate  category designated "reac-
tive carbon,"  and using  it to  tabulate  the total  number of  carbons  in
reactive emitted species.   (The specific  criterion used for "reactive"  in
this context would  be  determined  by  the  criterion presently  used  when
assigning  contributions to  PAR  when  aggregating  emissions   for  Carbon
Bond.)    The levels of  PAR emissions  to  use in  the model  could then  be
determined  by  subtracting  from this number  the number of  carbons  in the
other model  species  used  to represent  the emissions.   For  example,  if a
simplified emissions  inventory contained  1 mole each of n-butane, ethanol,
and  butyl  benzene,  then  the  corresponding emissions  aggregation would
include  1  mole each  of "alkanes  (0.25-0.5),"  "others (0.25-0.5 react),"
"Aromatic (<2 react),"  and 16  moles  of the special "reactive carbon" group
(4 moles  from n-butane, 2 from ethanol,  and 10  from butyl benzene).   In
models using the  Carbon Bond mechanism,  the emissions group "Aromatic  (<2
react)"  would be  represented  by the seven-carbon species  "TOL," and  the
"PAR" levels would be calculated by subtracting these 7 moles  of reactive
carbons already  represented  in TOL from the total of 16 moles  of reactive
carbon,  yielding  9  moles  of  "PAR."   Since  the "alkanes"  and "others"
emissions in this  inventory  are already represented  by  "PAR,"  the aggrega-
tions  in  these groups are not used  in  determining input for models  using
this mechanism.
     An  analogous method  could  be used  to  account  for  the extra double
bonds  in poly functional alkenes.    In  this  case, it  is recommended  that
total  terminal  double bonds and total internal double  bonds be aggregated
into separate  groups,   since  these  are   represented  differently  in  the
Carbon Bond mechanism.
     In  view of  the extensive  use of the  Carbon  Bond  mechanism,  it  is
perhaps  surprising that this emissions aggregation system was  not extended

-------
so  it  could be  used for this mechanism.   The principle  reason for  this
appears  to be  that  the software  presently  used  to process  the  NAPAP
emissions  input  data cannot be  readily  adapted to  be able to  distribute
more  than  one  detailed  emissions  class  into more  than  one  aggregated
emissions  category.   Aggregation for Carbon  Bond obviously requires this
capability.  This artificial limitation  would  presumably be addressed when
the emissions processing software is upgraded.
     Aggregation  of  Aromatics.   The  present  32-class scheme  lumps  the
xylenes  and  other  dialkylbenzenes,   tri-  and  polyalkyl  benzenes,  and
naphthalenes into  a single group ("Aromatics  >2  react").   This  is satis-
factory for use  with the RADM  and most other  lumped mechanisms,  which only
uses  a single  model  species  (generally  designated  XYL)  for  these  com-
pounds.    However,  the most detailed  version  of  the  SAPRC/ERT  mechanism
(Lurmann  et al.,  1987)  has  a separate model  species for the tri- and poly-
alkylbenzenes,  and  thus this  grouping  is not  adequate for use  with this
mechanism.  Splitting the  "Aromatics  (>2 react)"  group into "Aromatics (2-
4 react)" and  "Aromatics (>4 react)," with the trialkylbenzenes and other
aromatics which react faster  than  4  x  10  ppm   min   being lumped into
the.latter group,  would address this problem.  Both of these  groups would
be significant   in  terms of amounts  of  emitted  species,  and  thus aggre-
gating them separately would not be inappropriate  on that  basis.
      It would  also  be desirable to aggregate the  naphthalenes  and other
 tri-  and  poly-cyclic aromatics  in separate  groups from the alkylbenzenes.
 Environmental  chamber and  modeling  studies  carried  out at SAPRC  indicate
 that   these  species  are  much   less  reactive  than  alkylbenzenes,  despite
 similar or higher OH radical rate constants (Carter et al., 1987).  There-
 fore,  lumping  them  with alkylbenzenes will  result in models  significantly
 overpredicting  their reactivity in model  simulations.  On the other hand,
 the naphthalenes are  relatively unimportant  in  terms of  amounts  emitted,
 contributing  only  approximately 0.4jt  of  the  mass  emissions  in the NAPAP
 total  U.S. anthropogenic  inventory.   However, there may  be special urban
 pollution cases where their contributions may be  more important.
      Use  with Global and Stratospheric Models.   As indicated above, this
 aggregation scheme  was  developed  for use with local and  regional oxidant
 and acid  deposition models, and additional  classes would  have to  be added
 for it to  be used  for other types  of modeling applications.  For  example,
                                     75

-------
global models may  require  more  distinction among less reactive compounds,
and stratospheric  models  would  require halogenated  compounds  to be split
out from  the "unreactive" category, or perhaps  separate categories tabu-
lating  the   total  numbers of Cl  and  Br  atoms  which reach  the stratos-
phere.  A discussion  of the input requirements  for  such models is beyond
the scope of this report.
     3.2.3  Aggregation of Emissions into  the RADM Mechanism
            The  gas-phase  RADM  mechanism  contains  17  different  model
species which can  be  used to represent the  reactions of various types  of
emitted reactive  VOCs.   These include five  species  for  alkanes, four for
alkenes,  two for aromatics, two  for  aldehydes,  one  for  ketones,  one for
phenols,  and two  for  organic  acids.    As  indicated above,  the system
developed for processing  anthropogenic emissions input for the RADM model
involves  a   two-step  procedure  where  first  the  detailed  emissions data,
given  in  terms of  mass emissions of  over 550  compounds,  are  aggregated
into molar  emissions  of 32 compounds  judged  to  be similar in  reactivity,
and  then  these  are  further  aggregated  into molar  emissions  of  the  16
species used in RADM.  As part of this program,  SAPRC  was asked to  make
recommendations for how to aggregate emissions organized  into  the  32-class
system  into  the species  in the  RADM mechanism.    The system  that was
recommended,  which we anticipate will  be  used  in RADM,  is summarized  in
this section.
     The  allocation of  the 32 emissions  groups  to species  in the  RADM
mechanism is summarized in Table 8.   In  many cases   these allocations are
reasonably  obvious;  cases  that are less obvious  are discussed below for
the  various types of  chemicals.   As  indicated  in the table, most of the
emissions groups  are  represented  by  the RADM species on  a  simple mole-per-
mole  basis, with  one mole  of  RADM species  being used for  each mole  of
compound  emitted.    However,  in some  cases,  particularly  those  where
emissions groups of significantly different reactivities have to be lumped
together, "reactivity weighing" is  employed.   This involves use of a  reac-
tivity  weighing factor in calculating the  number of  moles of model species
from the moles  of emitted species.   The  general principle of reactivity
weighing  will  be discussed  first,  followed by  a discussion  of  specific
considerations  involved for the various types of chemicals.
                                     76

-------
Table 8.  Assignments of Emissions Groups to RADM Model Species
Emissions Class Mix
____ Fao
No.
1
2
3
4
5
6
7
8

9
10
11
12
13

14
15
16
17
18

19
20
21
22
23
24
25
26
27
28
29
30
Description (a)
Methane
Ethane
Propane
Alkanes (0.25-0.5 react)
Alkanes (0.5-1 react)
Alkanes (1-2 react)
Alkanes (>2 react
Alkane/Aromatic Mix 0.91
0.09
Ethene
Propene
Alkenes (Primary)
Alkenes (Internal)
Alkenes (Prim/Inter Mix) 0.50
0.50
Benzene, Halobenzenes
Aromatics (<2 react)
Aromatics ( >2 react)
Phenols and Cresols
Styrenes 1.0
1.0
Formaldehyde
Higher Aldehydes
Acetone
Higher Ketones
Organic Acids
Acetylene
Haloalkenes
Unreactive
Others (<0.25 react)
Others (0.25-0.5 react)
Others (0.5-1. react)
Others (>1 react)
RADM
Class
CH4
ETH
HC3
HC3
HC5
HC8
HC8
HC8
XYL
OL2
OLT
OLT
OLI
OLT
OLI
TOL
TOL
XYL
CSL
TOL
OLT
HCHO
ALD
KET
KET
ORA2
HC3
HC3
_
HC3
HC3
HC5
HC8
Weighing Factor (b,c)
Factor
1.0
1.0
0.519
0.964
0.956
0.945
1.141
1.002
0.090
1.0
1.0
1.0
1.0
0.500
0.500
0.293
1.0
1.0
1.0
1.0
1.0
1.0
1.0
0.253
1.0
1.0
0.343
0.078
1.0
0.404
1.215
1.075
1.011
k(emit)


0.180
0.369
0.668
1.315
2.339
2.018







0.185







0.035


0.115
0.025

0.137
0.496
0.801 '
1.552
k(RADM)


0.386
0.386
0.715
1.507
1.507
1.507







0.902







0.147


0.386
0.386

0.386
0.386
0.715
1.507
  (a)  Molar distribution of RADM species used for groups represented by
      more than one RADM species.

  (b)  "Factor" is number of moles of model species used to  represent
      the emissions group.  If  reactivity weighing is not employed, the
      factor  is 1.0 or  "Mix Fac.,"  if  applicable.  If reactivity weighing
      is employed, these are  given  by

                                                               (continued)
                                     77

-------
Table 8 (continued) - 2
 (c)
Factor = Mix Fac. (if applicable)
                x (1-exp[-k(emit)xINTOH])/(1-exp[-k(RADM)xINTOH])

where k(emit) and k(RADM) are respectively the 300 K  OH  rate
constants for the emissions group and the RADM species  (given in
the table in units of 10  ppm"1 rain"1), and INTOH is  assigned a
value of 110 ppt-min.  See text.

The tabulated k(RADM) values for the lumped alkane model species
are for the RADM-P mechanism.  The March 1988 RADM mechanism used
k(RADM) values for HC3, HC5, and HC8 of 0.376, 0.786, and 1.655,
respectively.
                                     78

-------
     Reactivity  Weighing.    Reactivity  weighing is  used  as  a  means to
represent groups  of compounds with similar  reaction  mechanisms, but  with
significantly differing  reaction  rates,  with a  single  model species.  It
is based on the  assumption that, at least for compounds with similar  reac-
tion  mechanisms,  the  effects  of  emissions  of  different compounds on
results of a  model  simulation is  approximately proportional to  the amount
of  compound  which  undergoes chemical  reaction  in  the  simulation.    For
example, if compound  "A" reacts rapidly  and  essentially is  all  consumed in
a model simulation, while compound "B" reacts more slowly and only half is
consumed, then,  under this  approximation, emissions  of  1 mole of "B"  has
the  same  effect as emissions  of  0.5 moles  of "A."    If  this   is assumed,
then  the appropriate  amount  of model species  to use  to  represent a  given
amount of emitted species would be such  that the estimated  amounts of each
which  react  in  the simulation are  the  same.   Thus, for  example,   if an
emissions group E  is to be represented  in a  model  simulation by  model
species M, then  the reactivity weighing  factor  for E  would  be derived such
that:
                Moles of M Reacting
                in Simulation
                Moles of E Reacting
                in Simulation
 or
Moles of M
used to
Represent E
Fraction
x of M which =
Reacts
Moles of
E Emitted
X
Fraction
of E which
Reacts
 Therefore,
       Reactivity
       Weighing
       Factor  for  E
Moles of M
used to
Represent E
Fraction
of E which
Reacts
                                               (I)
Moles of E
Emitted
Fraction
of M which
Reacts
 Thus,  for example,  if  half of  the  emitted species is  estimated to react
 under  the conditions of  a model simulation, and  these  species are repre-
 sented by a more rapidly  reacting model  species  where 75? is estimated to
 react  under the same conditions,  then the appropriate reactivity weighing
 factor would be 0.5/0.75  = 0.667.
                                     79

-------
     The use  of  this reactivity weighing approach requires  an ability to
estimate fractions of emitted  species  which undergo  reaction in represen-
tative model  simulations.   Such estimates  are  relatively straightforward
for  compounds which react  in  the  atmosphere primarily  with  hydroxyl
radicals, as  is  the case for  most  of the  types of  compounds where reac-
tivity weighing  is  employed.   For such compounds,  we have  shown  that the
fraction of  compound which reacts  in  an  airshed model simulation  can be
fit reasonably well by the following (Carter and Atkinson, 1989):
       Moles Compound
       Reacted                     - kOH x INTOH
      	  =  (  1 - e               )                     (II)
       Moles Compound
       Emitted
where kOH  is the OH  radical  rate constant and INTOH  is  a  measure of the
effective  integrated  OH  radical levels under  the  conditions  of the model
simulation.  (The above equation  is exact  if all the VOC  is present at the
beginning  of  the  simulation,  but is  approximate otherwise.)    If this
approximation  is employed, and  if  an emissions group E is to be represent-
ed by the model species M, then combining Equations (I) and  (II) yields:
                                              - kOH(e) x INTOH
    Reactivity     Moles of M         ( 1 - e                  )
    Weighing    =  Representing   =  	    (III)
    Factor         one mole of E              - kOH(m) x INTOH
    for E                             ( 1 - e                  )
where  kOH(e)  is the  average  OH radical  rate  constant  for  species  in
emissions group  E,  and  kOH(m)  is the OH radical rate constant used  in  the
gas-phase mechanism for model  species  M.   Note that if E  and M are  both
rapidly reacting, then  the  reactivity weighing  factor is always unity  (and
thus  in effect  reactivity  weighing  need not be  employed);  while if  they
both react  slowly,  then the reactivity  factor is  proportional  to  the ratio
of the  rate  constants,  and  independent of the estimated INTOH parameter.
     The  appropriate value  to  use for  INTOH  in derivation  of the  reac-
tivity  weighing  factors will depend  on  the  conditions of the  model simula-
tion,  since it  is dependent on both  the overall OH radical  levels  in  the
simulation,  and the  length of  time  the simulation is carried  out.   For
                                     80

-------
RADM simulations  lasting 2-3  days,  Stockwell  and co-workers  (Stockwell,
private communication,  1988)  estimate that an  INTOH value of  110 ppt-min
is appropriate.  Therefore, that value is  used  in  deriving  our  recommended
weighing factors.   Note  that  as indicated above the estimates of INTOH do
not affect  weighing for  very rapidly or  very  slowly reacting compounds,
but they  do affect  weighing  factors for  compounds of intermediate  reac-
tivity.   A  limited number  of test calculations we have carried out with
one-cell box models suggests  that  in  practice the  results  of  model simula-
tions will not be highly  sensitive to estimates of INTOH provided they are
within reasonable limits.   However,  the  recommended factors in  Table  8 can
be  readily  re-calculated for use in  processing  VOC  emissions input for
model simulations where an  INTOH of  110  ppt-min is inappropriate.  Smaller
values may  be  more appropriate for  1-day urban simulations, while larger
values may be appropriate for simulations of extended time  periods.
     The major parameters  affecting  the  reactivity  weighing factors are
the OH radical rate constants assigned  to the  emissions groups and to the
model species.   The average  T=300 K  OH  radical rate constants calculated
for the emissions groups  where  reactivity  weighing is  recommended, and for
the model species  used  to represent  them in the RADM mechanism, are  shown
in  Table 8.    The  former  were  calculated  using the  distributions  of
emissions in   the  current NAPAP total  U.S. inventory,  together  with our
assignments of OH  radical rate constants for the  various  types of emitted
compounds (Carter,  1988).  The  rate constants  for several of the lumped
model  species  in   the  RADM  gas-phase  mechanism  vary  depending  on the
version of  the mechanism,  since these  were modified  as  a results of the
recommendations  made  during   this  program.    The  table   gives  the  rate
constants for  the  version  employing  our final  recommended values for the
lumped model  species which are discussed  in  the  following section,  since
it  is  our understanding  that our  recommendations  in  this regard will be
implemented into RADM.
     This  reactivity weighing  approach  is  clearly not   appropriate for
lumping  together  species with  different  types  of reaction  mechanisms,
since different  types  of mechanisms have been  shown  to have  significantly
different effects  on results  of model simulations,  even if differences in
reaction  rates  are  taken  into  account  (Carter  and  Atkinson,  1989).
Furthermore, reactivity weighing is  an approximation even  for species with
                                    81

-------
very similar  reaction mechanisms,  since differences  in  rates  of reaction
affect  the distribution  of where  and when  the  compound  reacts  in the
simulation.   Rapidly  reacting  compounds react  closer to  the source areas,
and slower reacting compounds have greater effects on remote regions.  For
this reason,  it  is not a good  approximation  to lump together species with
very large differences in reaction  rates.   In addition, since  the  frac-
tions of emitted  compounds or model  species which react will depend on the
conditions of the model simulation,  the  weighing  factors appropriate for
one set of conditions may not necessarily be  appropriate  for others.
     Clearly,  the use of separate  model species for compounds  with appro-
priate  reaction  rates would be  preferable to  using reactivity  weighing.
However, when the  number  of species  which  can be  used  in the  model  is
limited, as  is the  case for regional  modeling applications,  the  use  of
reactivity weighing  is preferable  to options  which are even more approxi-
mate.   In  particular, it is preferable to the  options of either ignoring
the effects of less reactive compounds for which there are no  correspond-
ing model  species,  or of representing emitted compounds  with equal number
of moles of model species of significantly different reactivity.
     Representation of Alkanes  and  "Others"  in  RADM.  The  RADM mechanism
employs  five  model   species  to represent reactions of  the  alkanes and
species which  are assumed to have alkane-like reaction mechanisms.  Two  of
these species (CHM and ETH) are used  exclusively  for methane  and  ethane,
and the  remaining three (HC3,  HC5,  and HC8)  are lumped species represent-
ing the  higher alkanes and  the various types of species  which are judged
to be better  represented by alkane model species than other species in the
RADM mechanism.   The  nomenclature  for HC3-HC8 is derived from  the  average
number  of  carbons  associated  with  each,  as  discussed   in  the  following
section.
     As shown  in  Table 8, the miscellaneous groups of compounds aggregated
into  the  "others" emissions  groups are  represented in RADM  by  lumped
alkane  model  species.   These  "others"  consist  primarily of  alcohols,
ethers,  and  esters,  though  a  variety of other  reactive  compounds,  which
are  not   considered  appropriate  for  the other   emissions   groups, are
included.  Although except for the simple alcohols  the reaction mechanisms
of these compounds are uncertain,  the major  compounds in these groups are

                                     82

-------
believed  to  react  analogously  to  alkanes,  reacting  primarily  with OH
radicals,  and probably forming similar types of oxygenated products.
     Alkane model  species are  also  used  to represent  acetylene and  the
haloalkenes.   Although  these  are unsaturated compounds, the use  of  alkene
model species to  represent them  is  considered inappropriate because  they
react much slower  with OH radicals  than any of  the alkene model  species
(Atkinson, 1986),  and because their  reaction with ozone is  relatively  much
less  important  than is  the  case  for most alkenes  (Atkinson and  Carter,
1984).  They are  thus lumped as  alkanes by default because no other  RADM
model species is  appropriate for  them.   The specific  reaction mechanisms
for these compounds are highly uncertain.
     The  allocation  of lumped higher alkane model  species to the  various
alkane  and "others"  reactivity  groupings is  shown  in  Table 8.   These
allocations were  made such  that each alkane  emissions group (other  than
the  "alkane/aromatic mix" group) is represented  by  only  one RADM model
species.    In  addition,  the  allocations were  determined,  as  much  as
possible  given  the emissions  groupings,  (a)  to  minimize  the  range  of
reactivities of emissions groups  represented  by  RADM species, and  (b)  to
have  the  three higher  lumped  alkane  RADM groups similar in their  estimated
numbers  of moles  reacted when  representing the NAPAP U.S.  inventory  in
RADM  simulations.    Since all   three of  the  higher  alkane  RADM  species
represent two or more  emissions groupings of differing reactivities, reac-
tivity  weighing  is  used  in  determining numbers of moles of RADM  species
representing these  groups.
      The  group "alkane/aromatic mix" is used to represent the emissions in
the  NAPAP inventory which are  characterized as mixtures with designations
such  as  "mineral   spirits,"  "naphtha,"  etc.    Based  on  an analysis  of
detailed  composition data for  one  type of mineral spirits  (Weir  et  al.,
1988),  we estimate  that  these  consist of approximately 91 %  Cy and higher
alkanes,  with  the remainder being aromatics.   The  average  OH radical  rate
constant  for the alkanes in  this group is within the  range used  for the
RADM species HC8,  and thus these 91/1 of the molar  emissions  lumped in the
"alkane/aromatic mix"  group  are represented by HC8, with reactivity weigh-
ing  being employed.  However,  the extent to which  the  analysis  used  as a
basis  for this  assignment  is  appropriate  in  representing  the  actual

                                     83

-------
emissions lumped  into  this group  is  highly uncertain.   This is an  area
where the emissions inventory  needs to  be  improved.
     Representation of Alkenes  in  RADM.   The RADM mechanism  employs  four
species to represent alkenes.   Ethene  (ETH) and  isoprene  (ISO) are  repre-
sented  explicitly,  OLT  represents propene and  those  alkenes  which  are
lumped  into  the  "alkenes (terminal)"  emissions group,  and  OLI  represents
those lumped into  "alkenes  (internal)."   As indicated  above,  the "alkenes
(terminal)"  group  is  defined as those alkenes,  other  than ethene,  which
                                                                  U     -1
react   with   OH  radicals  with  rate constants less  than  7.5 x  10   ppm
min"1 and  includes most higher  alkenes  with only terminal double  bonds.
The  "alkenes  (internal)"  group  consists of alkenes, other  than  isoprene,
which react with higher OH  radical rate constants and consists of internal
alkenes, dialkenes, and terpenes.
     Other than  isoprene and  the "alkene (mixed)" emissions groups,  which
are  special  cases discussed below, the allocation of  alkenes in the RADM
mechanism  corresponds almost exactly  with that  used  in  the  emissions
grouping.    The  only  other  exception  is that,   because   of  its  high
emissions, propene  is  treated separately in the emissions grouping,  while
in RADM it  is  lumped  with  the other "primary"  alkenes.  These correspond-
ences are  indicated in Table  8.  As indicated  there, no reactivity weigh-
ing  is  employed  for alkene species.   This  is  because  the only case where
more  than  one  emissions  group is aggregated into a RADM species are those
represented  by  OLT, and these  compounds react so rapidly  that  the reac-
tivity  weighing factor would be near unity (see Table 7).
     The  "alkene  (mixed)"  group  consists  of those emissions  which  are
identified  in the  emissions  inventories with  terms such  as "isomers  of
pentene,"  "C-iQ  olefins,"   etc.    It  is  not  known  to what  extent   these
emissions  consist  of  terminal  alkenes  (represented  by  OLT in  RADM)  or
internal alkenes  (represented by OLI), and  we arbitrarily assume that  they
consist of equal amounts of  each.  Note,  however,  that in making alloca-
tions  for  the Carbon  Bond and the CAL (SAPRC/ERT) mechanisms, Jeffries et
al.  (1989c) arbitrarily  assumed  that  they were  all   terminal.   This  is
another area where work is needed to  improve the quality of  the emissions
 inventories.
      Because of  the importance  of isoprene in  biogenic emissions, combined
with the importance of  biogenic emissions in regional model simulations,

-------
isoprene  is  treated  explicitly  in  the  RADM mechanism.    However,   its
emissions are essentially negligible  in  the NAPAP  anthropogenic inventory,
and thus the  RADM  isoprene species (ISO)  is not  used in representing  any
of  the  anthropogenic emissions.    (The  small  amount of  isoprene in  the
anthropogenic inventory,  consisting of only 0.001% of the  mass emissions,
is represented by OLI.)   Thus,  the  RADM  model  species ISO is used only  for
biogenically emitted isoprene.
     On  the other  hand,  there is  no  RADM model  species for  terpenes,
despite  the fact that  these  are comparable  in importance  as  isoprene in
the  biogenic inventory  (Middleton, private  communication,  1989).   Thus,
the  biogenically  emitted  terpenes  are  represented by  the  same  model
species, OLI, which  is  used to  represent anthropogenically  emitted alkenes
of  similar  reactivity.    However,  despite  the   potential  importance  of
terpenes  in regional model simulations,  it is not clear whether there is
any  significant  advantage  to representing  them  separately in the model.
Relatively  little is  known  about  their  reaction mechanisms,  other  than
their  initial rates of reaction,  which are in the  range  appropriate  for
OLI.     Note  that  the  model   simulations  of  the  chamber  experiments
(discussed  in Section 5) indicate  that  representing  at  least one terpene,
alpha-pinene,  by OLI may  not be inappropriate.   As far as  we are aware,
there  are no chamber data  available  to  indicate how  appropriate OLI is in
representing the other  terpenes.
     Representation  of Aromatics  in  RADM.  The RADM mechanism uses three
species to represent aromatic  compounds.   The most important of these are
TOL,  representing  toluene, other  monoalkylbenzenes, and  lesser reactive
aromatics,  and XYL,  representing xylenes, other di- and polyalkylbenzenes,
and other aromatics which  react  more rapidly  than the  xylenes.   In addi-
 tion,   the  RADM mechanism  includes  the model species  CSL,  primarily to
 represent the phenol and cresol products formed in the  reaction  of TOL and
 XYL, but which can also be used to represent  phenols and cresols which are
 emitted directly.
      The  allocations   of   the  aromatic  emissions groups  to  these  three
 aromatic model  species in RADM are  shown in Table  8.   Since there is no
 separate model  species  for  benzene, the  "benzene/halobenzene" emissions
 group  is  represented  by  the  RADM species TOL,   with  reactivity  weighing
 being  employed  to account  for  the differences in reaction rates for these
                                     85

-------
species.   The small amount  of styrene emissions  are represented  by  two
RADM groups,  with TOL being  used  to represent  reaction at  the  aromatic
group,  and OLT being used to represent reaction at the double bond.  (This
is the  only  case  where  a version of  the "lumped structure"  approach  is
used with  the  RADM mechanism.)   Most of the aromatics  emitted are in the
groups  "aromatics  (<2  react)," and "aromatics  (>2  react)," which  are
represented by TOL and  XYL,  respectively.   Since  the  mechanisms  used for
TOL and  XYL  are  based primarily on  those for the major  species  in these
two groups (see  following  section),  no  reactivity weighing is employed  in
these cases.
     As  indicated  above,  the  group  "alkane/aromatic  mix"   is   used   to
represent  the  emissions  in the NAPAP inventory  that  are solvent mixtures
which  we  estimate to  consist of  approximately  9%  aromatics,  with  the
remainder  being  alkanes.   The  average OH  radical rate  constant  for the
aromatics  estimated to  be  in these mixtures is  within  the range  used for
the RADM  species XYL,  and thus these 9% of  the  molar emissions lumped  in
the  "alkane/aromatic  mix"  group  are  represented  by  this  species.    No
reactivity weighing is  employed  because these  higher  aromatic  species
react sufficiently rapidly that the reactivity  weighing factors  would  be
near unity.
     Representation of Aldehydes. Ketones, and Organic Acids  in RADM.  The
RADM mechanism includes  the  species HCHO for formaldehyde, ALD for higher
aldehydes, KET  for ketones,  ORA1   for  formic acid,  and  ORA2  for acetic
acid.   The aldehyde species  corresponds  directly  to the aldehyde  alloca-
tion in  the  emissions  group, as shown in Table  8.  The reactions  used  in
the  RADM mechanism for KET  are based on estimated  mechanisms for methyl
ethyl  ketone and  higher ketones,  and  therefore  KET  is  used  to  represent
the "higher  ketone" group  directly,  without reactivity weighing.   There  is
no species in  the RADM mechanism corresponding to  acetone, so  it  is repre-
sented  by the higher  ketone model  species, KET,  using reactivity weigh-
ing.  The  organic  acids are represented by ORA2,  as discussed below.
     Note  that the use of reactivity weighing factors calculated  using  OH
radical  rate constants  (as is the  case  for  all those shown in Table 8)  is
not  strictly valid when  lumping acetone  with  the higher  ketones,  because
these  compounds  (especially  acetone) are consumed to a significant extent
by  photolysis.   However,  reactivity simulations we  carried out  using  the
                                     86

-------
SAPRC mechanism (Carter, unpublished  results,  1988) indicate that acetone
is about  1/3 as  reactive  as  the  higher  ketones  in terms  of  effects  on
ozone formation,  which is  reasonably close to  the reactivity difference
estimated using OH  radical  rate constants.  Since  the  reactivity weighing
factor obtained using  OH radical  rate constants is reasonably consistent
with the reactivity differences  for the  ketones  indicated  by modeling,  the
former is retained  in  our recommendations for  consistency  with  the deriva-
tions of the factors for the other types  of emitted species.
     The representation  of  organic acids is the one  case where the  RADM
mechanism is more detailed  than the  32-class emissions  aggregations.   The
mechanism has  a separate species  for formic acid  and for  higher organic
acids, while the emissions  scheme aggregates all  organic acids into  one
group.   However,  the  emissions  of  formic  acid in the NAPAP total  U.S.
inventory is minor, accounting  for  only  2.5% of  the  mass  of  the  organic
acid  emissions.   For  that reason,  all  the  organic  acid  emissions  are
represented  by  ORA2 (higher organic acids) in the mechanism.

3.3  Derivation of  Kinetic and Mechanistic Parameters  for  Lumped Model
     Species in RADM
     As  discussed in  the previous section, the RADM mechanism includes a
number of model species which are intended to represent aggregate mixtures
of  many  different  emitted  VOCs.   These  include the lumped alkane  species
HC3, HC5, and  HC8,  the lumped aromatic species TOL and XYL (which,  despite
their names,  actually  represent a mixture of compounds), the lumped alkene
species  OLT and OLI,  and the  lumped  oxygenate species  ALD,  KET,  CSL,  and
ORA2,  which are  also  used to  represent various  types of photooxidation
products.    As  part of the  RADM mechanism evaluation  study,  we1 evaluated
 the appropriateness of  the  reaction mechanisms used  for  these species to
 the specific groups of emitted compound  each represents  and made several
 recommendations  in this  regard.   These  recommendations,  which  have been
 adopted  for the  latest version of  the RADM mechanism, are summarized in
 this section.
      Lumped model species  can  be  represented in molecularly-based mechan-
 isms such  as  RADM  either  using  the  "surrogate  species"  or  the  "lumped
 molecule"  approach.   In the "surrogate  species" approach,  the mechanism of
 a typical or representative compound in  the group  is used  to represent  the
                                     87

-------
reactions of all members of the group.  For example, in the RADM mechanism
(as in many others),  the mechanism used  for  the  species  ALD,  which repre-
sents the  higher aldehydes, is  based on  that  for acetaldehyde.   In the
"lumped molecule"  approach, the  reactions of a  mixture of  compounds is
represented by a pseudo-molecule whose mechanisms are derived to represent
an aggregate  average  of those of the sets of compounds they  are  used to
represent.   The former approach is most useful when one compound dominates
in the  lumped group,  when  there is  only one compound  in the  group for
which there is  enough kinetic  and mechanistic data available to derive or
estimate a  mechanism, or  when the distribution of compounds  in the group
is unknown.   The latter approach has  the  greatest potential for chemical
accuracy, provided that there  is sufficient  information available concern-
ing  the  distribution of  compounds  represented  by  the  lumped  group and
relevant aspects of the reaction mechanisms of those compounds.
     In  the  current  RADM  mechanism,  the reactions  of  the  lumped model
species  ALD,  KET, CSL,  ORA2, TOL,  and  (in part) XYL  are based  on the
surrogate  species  approach,  primarily  because  there  is  only   limited
information  available  concerning the mechanisms for  more  than  a few
examples of these types of compounds.   On  the  other hand, the reactions
used  for  the lumped  alkane species  HC3,  HC5, and HC8,  the lumped  alkene
species OLI and OLT,  and  (in part)  the lumped higher aromatic species XYL
are derived based on  the  lumped molecule approach.  The derivation  of the
kinetic and mechanistic parameters  for these lumped molecules was  carried
out at SAPRC  as part of this RADM evaluation study, and were  incorporated
into the final recommended  version of the RADM mechanism.
     Tables 9 through 11  give  the averages of the  OH  radical  rate  constant
and  other  mechanistic parameters of relevance  to RADM  for  the  emitted
species  in the  various emissions  groups which are  lumped as C3+  alkanes
(Table 9),  aromatic  hydrocarbons (Table 10)  and non-ethene alkenes (Table
11).   The averages for the  emissions groups were derived using the NAPAP
total  contiguous U.S. emissions inventory, and were  computed  based on the
relative number of moles  of the various types of species in each emissions
group.   No reactivity weighing of any kind was employed  in computing these
averages,  because reactivity  weighing  is  not employed when  the emissions
are   aggregated  into these groups.   The tables also  show  the  total number

-------
Table 9.
Averages of Parameters for Emissions Groups Lumped as CU  Alkanes and
for RADM Lumped Alkane Model Species, Derived Using the NAPAP Total
U.S. Emissions Inventory
K Moles
Emit
kOH
(a)
% Ret
(b)
nf
(c)

ON IT X02 HCHO ALD KET
Lumped as HC3
23.9
16.7
14.0
2.8
2.0
2.1


0.
0.
0.
0.
0.
0.
0.
0.
369
496
115
180
137
025
386
376
45.
40.
9.
2.
1.
0.


6
1
5
9
6
3


4
2
2
3
1
2
2

.02
.10
.00
.00
.88
.00
.99

0.07
0.00
0.00
0.04
0.00
0.00
0.03
0.03
0.48
-0.87
-0.30
0.00
-0.63
0.00
-0.17
-0.10
0.16
0.02
0.00
0.00
0.69
0.00
0.09
0.14
0.61
0.98
0.70
0.30
0.12
1.00
0.75
0.69
0.50
0.04
0.00
0.33
0. 19
0.00
0.25
0.26
                                                             Alkanes  (0.25-0.50)
                                                             Others (0.25-0.50)
                                                             Acetylene
                                                             Propane
                                                             Others (<0.25)
                                                             Haloalkenes

                                                             Recommended HC3  (d)
                                                             Initial  RADM HC3  (e)
Lumped as HC8
                64.3   5.52   0.11  0.74  0.01  0.51  0.80
                35.7   3.43   0.02 -0.63  0.03  0.15  0.50

                       4.77   0.08  0.25  0.02  0.38  0.69
                              0.10  0.27  0.00  0.36  0.76
                                                    Alkanes  (0.50-1.00)
                                                    Others  (0.50-1.00)

                                                    Recommended  HC5  (d)
                                                    Initial  RADM HC5 (e)
21.9
3.1
3.3
1.6


1
2
1
2
1
1
.32
.34
.55
.02
.51
.66
70.5
12.0
11.5
6.0


7.
12.
2.
10.
7.

74
38
87
04
87

0.25
0.34
0.02
0.30
0.24
0.24
0.88
1.10
-0.76
1.34
0.75
0.78
0.04
0.06
0.01
0.12
0.04
0.05
0.32
0.20
0.74
0.37
0.35
0.44
1.11
1.39
0.30
1.31
1.06
1.05
Alkanes (1.00-2.00)
Alkanes O2.00)
Others (>2.00)
Alkane/Aromatic Mix
Recommended HC8 (d)
Initial RADM HC8 (e)
 (a)  OH Radical rate constant  in  units of 10  ppm    min    at T =  300 K.

 (b)  Percent of estimated moles reacted for each lumped  RADM group.   Estimated  by
     moles  reacted  = moles  emitted  x [ 1 - exp(-kOH  x  INTOH) ], where INTOH =  110
     ppt-min.

 (c)  Average number of  carbons.

 (d)  Derived from averages  for the  emissions groups, as  described in the text.

 (e)  Used  in the March  1988 version of the RADM mechanism.
                                        89

-------
Table 10.  Averages of Parameters for Emissions Groups Lumped as Aromatics
           and for RADM Lumped Aromatic Model Species, Derived Using the
           NAPAP Total U.S.  Emissions Inventory
K. Moles
Emit

16.1
7.3
0.8




16.3
0.2


kOH
(a)

0.919
0.185
0.902
0.918
0.902
0.902

4.09
3.73
4.09
4.54
% Ret. nC
(b) (c)
Lumped as TOL
84.9 7.26
11.1 6.00
4.0 7.36
7.13


Lumped as XYL
99.0 8.88
1.0 8.26
8.87

Description

Aromatics (<2.00)
Benzene, Halobenzenes
Styrenes
Average for RADM TOL (d)
Recommended RADM TOL (e)
Initial RADM TOL (f)

Aromatics (>2.00)
Alkane/Aromatic Mix
Recommended for RADM XYL (d)
Initial RADM XYL (f)
(a)  OH Radical rate constant in units of 10  ppm"1 min"' for T = 300  K.

(b)  Percent contribution to the RADM group Reacted, estimated using
     Moles Reacted  = Moles Emitted x [ 1 - exp(-kOH x INTOH) ], where
     INTOH = 110 ppt min"1.

(c)  Average number of carbons.

(d)  Derived from averages for the emissions groups.

(e)  The OH radical rate constant for TOL used in the initial RADM
     mechanism  is based on that for toluene.  This is sufficiently
     close to the average for the emissions represented by TOL that
     no change  is recommended.

(f)  Used in the March 1988 version of the RADM mechanism.
                                     90

-------
Table 11.  Averages of Parameters for Emissions Groups Lumped as Non-Etnene Alkenes and
           for RADM Lumped Non-Ethene Alkene Model Species, Derived Using the NAPAP Total
           U.S. Emissions Inventory
K
.Moles
Emit
*
Ret.
(a)
nC
(b)
kOH
(c)
k03
(d)
kM03
(e)
Structural
=CH2 =CHR
Groups
=CR2
Description
Lumped as OLT
10.6
 1.6
 0.9
 0.8
62.6
27.3
 5.5
 1.5
                 3.00
                 1.71
                 5.89
                 3.50
3.82
1.72
5.11
1.99

1.19
1.19
3.82
1.71
1.67
1.70
  78
                                1
                               1.73
                               1.71
                               1.71
 11.1    1.00  1.00  0.00
 21.1    1.00  0.99  0.01
 68.1    1.00  0.89  0.11
159-1    1.00  0.68  0.32

 26.5    1.00  0.98  0.02
 26.5    1.00  1.00  0.00
 23.3    1.00  1.00  0.00
Propene
Primary Alkenes
Mixed Alkenes
Styrenes

Average for OLT (f)
Recommended for OLT (g)
Initial RADM OLT (h)
Lumped as OLI
11.
0.



0
9



92
7



.2
.8



1.85
5.89
1.93
1.93

9.88
9.11
9.82
9.82
9.21
23.9
29.0
21.3
21.3
30.2
1911.
578.
1837.
1837.
1837.
0.30
0.00
0.28
0.28
0.28
1.10
2.00
1.11
1.15
1.15
0.11
0.00
0.10
0.10
0.10
Internal Alkenes
Mixed Alkenes
Average for OLI (f)
Recommended for OLI (g)
Initial RADM OLI (h)
(a)  Percent contribution to the RADM group reacted,  estimated assuming all  emitted  alkenes
     are consumed through reaction.
(b)  Average number of carbons.
(c)  OH radical rate .constant in units of 10^  pom"1 min~'  for T =  300 K.
(d)  Ozone rate constant in units of ppm   min"^  for  T  =  300  K.
(e)  NO^ radical rate  constant in units of ppm"   min    for T  = 300 K.
(f)  Derived from averages for the emissions groups.
(g)  The average values for the  structural group  parameters are so close  to  the  initial  RADM
     values that no change is recommended.
(h)  Used in the March 1988 version  of the RADM mechanism.

-------
of moles of each  emissions  group which were derived from  the  NAPAP total
emissions inventory.
     The estimated percent  contributions  of the various  emissions groups
to the reactivities of  the  lumped RADM groups used to  represent  them are
also  shown  in  Tables  9  through  11.   In  the  case  of  the  alkanes  and
aromatics,   these  contributions  were  computed  based  on   the  approximate
amounts of  each group  expected to react in a  model  simulation,  estimated
as  described   above   in  the  discussion  of  reactivity  weighing  [i.e.,
estimated using Equation  (II), with  INTOH  =  110  ppt-min].   For non-ethene
alkenes, these  were  estimated based  on  assuming all  the  emitted alkenes
react in the  model  simulation as expected based on  their relatively high
OH radical  rate constants and the fact that  they  are  consumed non-negli-
gibly by reaction with ozone as well.   These percent contributions are
then  used  to  compute  the  recommended  averages  for  the  OH  radical  rate
constants and  for the  other mechanistic  parameters  for   the  lumped  RADM
species.
     It should  be noted that as  indicated  in  Section  2 the parameters for
the evaluated RADM mechanism incorporated  earlier  recommendations we made
for  kinetic and  mechanistic parameters  for  these  species,   based  on  an
analysis similar to that  described here.  However,  the emissions parameter
data  base  at  SAPRC  was  still  undergoing development  at the time these
initial recommendations were made, and thus the final recommendations are
slightly different.  The  parameter values  used in  the initially evaluated
mechanism,  derived based  on this  earlier analysis,  are shown in the tables
as "Initial RADM."   Since the final  recommendations have been implemented
into  the new  version  of  the RADM-II  mechanism,   the derivation  of the
earlier recommendations is not discussed here.
     The derivations of the values for the kinetic and mechanistic param-
eters shown in  Tables 9 through  11, and the ways they are  used in  the RADM
mechanism for the various types  of lumped species are discussed below.   In
most  cases,  the  parameters are  based on  the kinetic   and  mechanistic
assignments  we  previously  made  for  representing  complex  mixtures   of
emitted  species  in  the  SAPRC  mechanism,   which  is  documented   elsewhere
(Carter, 1988).   However, in some cases,  the product yield parameters had
to  be modified to be  consistent with the set of  product species used  in
the  RADM mechanism.
                                     92

-------
     3.3.1   Mechanistic  Parameters for Lumped Alkanes
            Alkanes  react in the  atmosphere primarily with OH radicals  to
form organic peroxy  radicals.   The  peroxy radicals  can  react with NO  to
form N02,  H02,  and a variety of oxygenated  products  or they can  react  with
NO to form organic nitrates.  Techniques exist  for estimating  the  rates  of
initial OH  radical  attack  at various  positions on the alkanes  (Atkinson,
1986,  1987; Carter  and  Atkinson,  1985),  and the  branching  ratios of the
various  reactions  of   the   radicals  subsequently  formed   (Carter  and
Atkinson,   1985).    We  have  used  these estimates  to derive  kinetic and
mechanistic parameters for the major  alkane species  in the NAPAP inventory
(Carter,  1988a).   If the distribution of  alkanes  in the available  NAPAP
total U.S. emissions inventory can be  considered to  be generally represen-
tative, then  these  estimates  can be used  to   derive average  values for
these parameters for the three alkane model species  in RADM.
     The mechanistic  parameters  for  the alkanes, the ways  in  which  they
are  estimated,  and  the  ways  in  which  they  are implemented into  the  RADM
mechanism, are as follows:
     The  OH Radical  Rate  Constants.   Estimates  of the  OH   radical  rate
constants  at  300 K were made  for  the various  types  of emitted  alkanes
using  the  OH  radical   rate  constant estimation  technique  of  Atkinson
(1987).  These  values were  then  used to derive  rate  constants  at 300 K for
the  alkane  species  used in  the model.  No  recommendation  is made concern-
ing  the temperature dependence  for   these  reactions, and the recommended
modified  RADM  mechanism uses the same  activation  energies  for these reac-
tions  as employed in the original RADM-II  mechanism  (NCAR  1987).
     Organic Nitrate Yields.   Organic nitrates  are formed in  alkane photo-
oxidations  by  the reaction  of peroxy  radicals  with  NO,  competing with the
alternative reaction route  involving  conversion of  NO to  N02  and ultimate
regeneration of radicals.   The nitrate yield is an  important  parameter  in
alkane mechanisms since its  formation represents a  radical and NOX sink.
Note that by conservation of  radicals and NOX,  the  overall  radical yield
in the alkane  photooxidation reaction is 1 minus the nitrate yield.  These
are estimated  for  the  various  types of alkanes  in   the   inventory  as
described  by  Carter and Atkinson  (1985),  with  parameters  updated as given
by  Atkinson  (1988) and  Carter  and  Atkinson  (1989b).    Note  that these
estimates are based on  assuming that  alkyl nitrate formation occurs only
                                     93

-------
during  reaction of  NO with  unsubstituted  peroxy radicals  such as  those
formed  immediately  following  the initial alkane + OH reaction,  and not  in
the reactions of  the hydroxy- or oxy-substituted  radicals  formed following
the alkoxy  radical  isomerizations  involved  in the photooxidations of  the
long-chain alkanes  (Carter  and Atkinson,  1985).   This  is highly uncertain,
and  thus the  simulated  nitrate yields  in  the  reactions  of  the higher
alkanes may be underestimated.
     The  RADM  mechanism  represents  the formation  of  alkyl  nitrates  in
alkane  systems  by a two-step  process.   The first is the  initial  reaction
with  OH to form  a peroxy  radical.   This radical then  reacts  with  NO  to
form  the  lumped alkyl nitrate species (ONIT) in variable yields and H02 in
yields  of 1 minus  the alkyl  nitrate yield.   For HC5 and HC8,  the peroxy
radical yield  in  the  first reaction  is  unity, and thus the  nitrate  yield
in  the second  reaction is the  same  as the  nitrate  yield derived in  the
emissions  analysis.   However,  to represent the appropriate  numbers  of NO
to  N02 conversions  as discussed below,  the peroxy radical  in the initial
reaction  of the lumped alkane HC3 has a less-than-unit yield, and thus the
nitrate yield  in  the reaction of HCS's peroxy radical with NO is increased
accordingly,  to  obtain  the   overall  yield  indicated   in  Table  9.    The
specifics of  how  nitrate formation and NO to N02 conversion is  represented
in  the RADM mechanism is  discussed  below.
      NO-to-NOo  Conversions.   The alkane photooxidation reactions involve
the formation of organic peroxy radicals which convert  varying amounts of
NO  to N02-   The total  amount  of NO converted to  N02  by  these reactions is
determined  by a number of  factors, including the alkyl  nitrate yields and
the overall number  of isomerizations and decompositions of radicals  which
are involved in  the photooxidation process.  The number  of these conver-
sions have been estimated  for the  major types of alkanes  in the emissions
inventory (Carter,  1988) by using a computer program employing  the peroxy-
and  alkoxy-radical  branching  ratio estimates  of  Carter  and  Atkinson
(1985), as updated  by Atkinson  (1988).   Analogous methods can  be used for
estimating NO-to-NO^  conversions in  species such as alcohols,  esters, and
others aggregated in  the "others"  categories which are lumped with alkanes
in  RADM,  since generally  they  are  predicted to form  the  same  types of
radical intermediates.   Note, however,  that the photooxidations of  simple
alcohols  involve  fewer  NO-to-N02  conversions  than do  alkanes,   since,

-------
unlike alkanes,  photooxidations of  alcohols do  not involve formation  of
peroxy radicals.
     In the RADM mechanism,  the  representation  of NO to N02 conversions  by
peroxy radicals  in  the lumped alkane  mechanism depends on  the  net  number
of  conversions  that  are  estimated   to  occur.    If  the  emissions  data
indicate  that  an  average  of more  than  one  NO  is  consumed  per  alkane
reacted, then the reactions can be represented by

      HCn + OH  — >  HCnP + a X02
     HCnP + NO  -->  b ONIT + (1-b) H02 +  (1-b) N02  + oxygenated  products
      X02 * NO  —>  N02

where  HCn  is the lumped  model  species  representing alkanes, HCnP  is the
lumped peroxy  radical  species associated with HCn,  and X02 is  a "chemical
operator" used  to represent  excess  NO-to-N02 conversions.   The parameter
"a"  in  the  above reaction  is shown in Table  9 as  the  X02  yield,  and the
parameter "b"  indicates  the  organic nitrate (ONIT) yield.  However, if the
emitted species represented  by the lumped  alkane species "HCn" includes a
large  fraction  of  alcohols,  whose   reactions  generate  H02 and  organic
products without  producing  peroxy radicals and NO-tc-N02 conversions, then
there  may be on the average  less than one  NO  converted to N02 per alkane
reacted.   This  is  the  case  for the  lumped species "HC3"  in  the present
mechanism,  and  is   indicated  in  Table 9  by  a   negative  number  for the
parameter "a,"  the  X02 yield.   In this  case the  reactions are represented
by

       HCn +  OH   —>  (Ua) HCnP            (where  a < 0)
      HCnP +  NO   —>  b/(Ua)  ONIT +  (1-[b/( Ua)]) (H02 + N02)
                     + oxygenated products.

Note that  in  both  cases  the  parameter  "b"  reflects  the total organic
nitrate  (ONIT)  yield,  and the total number  of molecules of  NO reacting  per
molecule of  HCn consumed is  a+1.
      The  above discussion concerns  the  case where reaction with NO is  the
major fate  for organic peroxy  radicals.   The RADM mechanism also  includes
 reactions  of peroxy radicals with each other and with H02.  The  products
                                     95

-------
formed in these reactions  are  not  affected by the X02 and the ONIT yields,
though they are affected by the yields of the oxygenated organic products
as discussed below.
     Oxygenated Organic Products.   The  atmospheric  photooxidation  reac-
tions of  the alkanes  in  the  presence of NO  involve the formation  of a
variety of  oxygenated  products, including complex bi- and poly-functional
species as  well  as simple  aldehydes and ketones.   The yields  of  these
oxygenates  can be estimated using the estimates of the relative rates for
alkoxy radical decomposition,  isomerization, and  reaction with Oo given by
Carter and  Atkinson (1985), and updated by Atkinson (1988).   The computer
program used  to  derive  these  estimates expresses the yields  in  terms of
the oxygenated products used  in the  most detailed  version  of the current
SAPRC mechanism.   These products  are  HCHO (formaldehyde), CCHO (acetalde-
hyde), RCHO (C^  and higher  aldehydes),  ACET  (acetone), and  MEK (methyl
ethyl and higher  ketones)  (Carter,  1988).   The RADM mechanism represents
alkane oxygenated  products  using  the  model  species HCHO (formaldehyde),
ALD  (higher aldehydes), and  KET  (ketones).   The SAPRC representation is
converted to  RADM model species  by using ALD  to represent  both CCHO and
RCHO, using KET to  represent MEK,  and  using  1/2 KET to represent ACET.
     The yields shown  in Table 9  for the products HCHO,  ALD, and KET  refer
to  their  overall yields under conditions  where reaction with NO  is  the
major  fate  for organic peroxy radicals.   As  indicated above,  the  RADM
mechanism also includes reactions of peroxy  radicals with  H02  and  with
other  peroxy  radicals.   The  RADM mechanism represents all  peroxy  +  H02
reactions as  forming lumped hydroperoxide species, which do not depend on
the  set  of  oxygenated  products  assumed  when  the  reaction  with NO
dominates.  When  the peroxy radicals react with other peroxy  radicals,  the
set of oxygenated organic  products will in general  be different from  those
formed in the reaction  with NO,  and organic nitrates will not be formed.
However,  no separate  estimates were made concerning  the  set of organic
products  formed  from peroxy  + peroxy reactions; instead they were repre-
sented by  the same set of products  formed in  the  NO reaction,  with  the
same  relative yields,  but their   overall yields  increased  by a factor of
b/(1-b) (where "b"  is the ONIT yield) to account for the fact that organic
nitrates  are  not  formed.  This is consistent  with the way  these  reactions
are represented in  the  SAPRC mechanism (Carter,  1988).
                                     96

-------
     Mechanistic  Parameters  for  Lumped  Alkanes in  the  RADM-M Mechanism.
As discussed  in  Section 2, one of  the  modifications we  recommend for  the
RADM mechanism  is that the two higher lumped alkane classes, HC5 and  HC8,
be combined into  a  single model species.   This is  because,  as shown  in
Section  6.4,  this  condensation  in  the  representation of  the  alkanes  is
unlikely  to have  significant  impacts on  predictions  of species  of  interest
in  RADM  applications  and  reduces  the  size  of   the  mechanism  by  two
species.    The  kinetic  and  mechanistic parameters  for  this  new  lumped
species  were derived  in  the  same   manner  as  discussed above.   Table  12
gives  the averages for the parameters for the  emissions groups  represented
by  this  new  lumped  higher  alkane  species  and  the  resulting  averages
calculated  for  the model species.   As indicated in  the table,  the  average
number of carbons of the compounds  it represents  in  the NAPAP inventory is
6.3,  and thus the model species  is  designated  "HC6."  The  parameters used
for  HC6  in the  recommended  RADM-M  mechanism are  based  on  the  results of
this  analysis.
      3.3.2   Mechanistic  Parameters  for  Lumped Aromatics
             The  aromatic hydrocarbons are  also consumed  in the atmosphere
primarily by reaction with  OH  radicals,  and  the  rate  constants  for the
reactions  of most of  the emitted  aromatic species  are  known or  can be
estimated  based  on  those  for  chemically  similar  species  (Atkinson,
1986).  OH radical rate constants  for T=300 K  were  assigned for the major
aromatic hydrocarbon  species  in  the NAPAP total  U.S. emissions inventory,
and  these  values, combined with  the relative  molar  emissions of aromatics
in that  inventory, were used  to  derive  the weighed  average rate constants
shown in Table  10 for the various aromatic emissions groups  and  for the
two  aromatic model  species  in RADM.   Note that the  weighed average rate
constant for the RADM species TOL  is so close to that for toluene that use
of the  toluene  rate  constant  is not inappropriate.  However,  because of
polyalkyl  benzenes,  naphthalenes,  and  other  higher aromatics  represented
by  the  RADM  species  XYL,  the  average  rate constant  derived  for  that
species  is higher than that for the xylenes.
      We  could   not   derive  average  values   for   the  other  mechanistic
parameters for  the  aromatics in a manner analogous  to that done for the
alkanes  because  the  mechanisms   for   aromatic   reactions  are  highly
uncertain.  We  must rely on chamber data,  as  discussed  in Section 2.2.1,
 to derive  these  parameters.
                                     97

-------
Table 12.   Averages of Parameters for Emissions Represented by the Combined Higher
           Alkane (HC5 + HC8) Model Species Recommended for the RADM-M Mechanism
K. Moles kOH
Emit (a)
Lumped
21.9
25.5
12.5
3.3
3.1
1.6

as
1
0
0
1
2
2
1
HC3
.315
.668
.801
.552
.339
.018
.088
'„ Ret.
(b)

37.8
29.9
16.6
6.1
6.4
3.2

nC
(c)

7.74
5.52
3.43
2.87
12.38
10.04
6.26

ON IT

0.25
0.11
0.02
0.02
0.34
0.30
0.15

X02

0.88
0.74
-0.63
-0.76
1.10
1.34
0.51
roduct
HCHO

0.04
0.01
0.03
0.01
0.06
0.12
0.03

ALD

0.32
0.51
0.15
0.74
0.20
0.37
0.38

KET

1.11
0.80
0.50
0.30
1.39
1.31
0.87
Description

Alkanes ( 1 . 00-2
Alkanes (0.50-1
Others (0.50-1.
Others (>2.00)
Alkanes O2.00)
Alkane/Aromatic
Recommended HC6


.00)
.00)
00)


Mix
(d)
(a)  OH radical rate constant in units of 10^ ppm"1 min"1 at T = 300 K.

(b)  Percent of estimated moles reacted for each lumped RADM group.  Estimated by
     moles reacted = moles emitted x [ 1 - exp(-kOH x INTOH) ], where INTOH = 110
     ppt-min.

(c)  Average number of carbons.

(d)  Derived from averages for the emissions groups, as described in the text.
                                       98

-------
     3.3-3  Mechanistic Parameters for Lumped Alkenes
            The RADM  mechanism includes  the  reaction of  alkenes  with OH
radicals, ozone,  and  NO^ radicals.   Rate  constants  were  assigned  for the
major alkene  species  in the NAPAP total U. S.  inventory  based on experi-
mental or estimated  values  (Carter,  1988,  and references therein).  Using
these and  the  relative molar emissions  in the  NAPAP  inventory,  average
rate  constant values for  the various lumped alkene  emissions groups and
for  the  RADM  lumped  alkene  species OLT and OLI  were derived.  The results
are  shown  in  Table 11.   The  contributions of the emissions groups to the
weighed  averages  for  the lumped RADM  species  were calculated assuming  that
all  the  emitted alkenes represented  in these groups are consumed during  a
model simulation.
     The  appropriate  sets  of organic products to use  in alkene  reaction
mechanisms depend  on  the nature of the substituents  around  the double  bond
(Carter,  1988).   For  example, reactions of OH or  NOo  radicals  with alkenes
result  in formation  of formaldehyde  from =CH2  groups,  higher aldehydes
from =CHR groups  (where  R =  alkyl), and  ketones  from  =CR2  groups,  with
analogous considerations being involved in determining the  products  in the
ozone  reactions  (Carter,  1988).    Table  11  shows the  average  number  of
substituents  about  the  double  bonds  for  the  alkenes  in  the  various
emissions  groups  and  for  those  represented by  OLT and OLI  in the  RADM
mechanism.    (The  number  of  substituents  does  not sum up  to  two for  the
"internal  alkenes"  emissions group  because cycloalkenes  are counted  as
having  only  one   substituent  about   the  double  bond for  the purpose  of
determining  the organic product yields.)  The recommended sets of  products
for  the lumped internal alkene  species (OLI)  is based  on  the  results  of
this analysis.    Thus,  for  example,   the  reaction of OLI with OH  radicals
yields  0.28  HCHO + 1.M5 ALD -•• 0.10 KET as the organic products,  reflecting
the   weighed   average  numbers  of  sCHj,  =CHR,   and  =CR2  substituents,
respectively.  In  the  case  of the  terminal  alkenes (OLT), this  analysis
indicates  that the  continued use of  propene  (with  one  =CH2  and  one  =CHR
substituent)   as   the  surrogate   species  for  the  terminal   alkenes  is
appropriate.
                                     99

-------
            M.  EVALUATION AGAINST ENVIRONMENTAL CHAMBER DATA:
                 CHARACTERIZATION AND LUMPING PROCEDURES
4.1  Introduction
     An  important  step  in  the  model  design and  evaluation  process   is
evaluation  of  the  gas   phase   chemistry  against  environmental  chamber
data.   The gas phase  chemistry  module is one  of the  few  modules  in the
atmospheric  modeling  system that  can  be  independently  tested  against
experimental data.   The testing  against chamber data  is essential for, but
not necessarily sufficient  to,  fully  establishing the  scientific validity
of a mechanism  for use in regional acid deposition  modeling.   The reason
for  this  is that  while  a large  data  base of experiments  exists  for the
ROG/NOx/Oo system  under urban-like  conditions,  virtually no reliable data
exist to test mechanisms'  performance  for  other  important species, such  as
sulfuric  acid,  nitric  acid,  hydrogen peroxide,  organic  peroxides,  and
organic acids,  or  for other  important concentration  regimes,  such as the
rural regime which is dominant in regional model applications.  Neverthe-
less, the  present  chamber data  base  can be  used to  test the rate of NOX
oxidation  and concentrations of ozone, peroxyacetylnitrate, and formalde-
hyde  predicted  by  gas phase chemical  mechanisms  for urban  conditions.
This is important  because  these  species are  important in  long-range trans-
port  and  the regional model must simulate  the  chemistry of emissions  as
they  are  transported from  urban  areas  to  remote  areas  with sensitive
receptors.
     It  is also important to recognize  that even though the present  data
base  does  not  allow  for  direct  evaluation  of  model  predictions  for
sulfuric acid and  nitric  acid,  the  testing protocol  provides some  indirect
evidence  for evaluation of performance for these species.  First,  accurate
prediction of  the  rate  of  NOX  oxidation  and ozone  formation  necessitate
accurate   predictions  of  OH radical  concentrations.    OH,  according  to
present  theory,   is  the  sole  gas-phase  oxidant  that  converts  SOp  to
sulfuric  acid and  NOp to  nitric acid  in the  daylight hours..  Second,  since
NOV  is primarily  oxidized  to nitric  acid or PAN,  the evaluation  results
   A
for  PAN  provide some information on whether  the mechanism is oxidizing NOV
                                                                          A
to the correct  species.
                                    100

-------
     An  evaluation  of the  March  1988 RADM,  recommended  modified  RADM
(RADM-M),  and  the  recommended  modified   parameter   (RADM-P)  mechanisms
against  environmental  chamber  data  was  performed  in  this study.   The
objective  of  the evaluation  was  to assess the  accuracy and precision of
the mechanisms'  predictions  for ozone, PAN, formaldehyde, and the rate of
NOX  oxidation.    The  approach  for  the evaluation  involved simulating  a
large  number  of experiments  (over 500) from  four different chambers and
statistically evaluating  the  mechanisms' performance.
     Experimental  data collected  in  the  following  chambers was used  for
the evaluation:

     1)  The SAPRC evacuable  indoor chamber (EC),
     2)  The SAPRC indoor Teflon  chamber (ITC),
     3)  The SAPRC outdoor Teflon chamber  (OTC), and
     4)  The University of North  Carolina  outdoor  chamber  (UNC).

Only   experiments  with  relatively  complete  data were  employed  in  the
evaluation.   The specific sets of experiments used  were those  recommended
for  use  in model  evaluation  by the  experimentalists  (Jeffries et  al.,
 1985;  Carter et al.,   1986),  with additional  experiments  added  which were
employed  in more  recent  evaluations of  the  SAPRC  mechanisms  (Carter  et
al.,  1987, 1988).
     The  results  of the  chamber simulations  are  described  in  Section  5.
Before presenting  these   results,  it  is  first  necessary to  indicate  the
assumptions made  and  procedures  that  were followed.    Carrying  out  model
 simulations of  environmental chamber experiments  involves  more  than just
 implementing the  chemical mechanism  on  a  computer  with  the  appropriate
 software,  specifying the  initial  reactant concentrations, and then running
 the calculation.    Before  such   simulations  can  be  carried  out,   it  is
 necessary to specify  in  the  model input the appropriate conditions of the
 chamber experiments,  including appropriate  representation of such factors
 as  temperature,  light intensity and  spectral  characteristics,  humidity,
 and chamber  wall  and  contamination  effects.    In addition, for  a  lumped
 mechanism such  as RADM,  which does not contain explicit  representation of
 all of  the various  types of  organic  compounds which  are  present  in the
 experiments,   appropriate   procedures   need  to   be   established  for
                                     101

-------
representing the organics used in the chamber experiments with the species
incorporated in the mechanism.   These are discussed in this section.

U.2  Characterization of Chamber Conditions
     In order  to  test  chemical  mechanisms  against  environmental chamber
results, it  is necessary to include  in  the model appropriate representa-
tions for chamber-dependent effects  such as wall reactions and character-
istics  of  the  light  source used  during  the  experiments.   The specific
chamber-dependent aspects represented in  the process  of testing mechanisms
are the following:

     1)  Light characterization;
     2)  Chamber-dependent radical sources;
     3)  Offgassing of NOX from the walls;
     4)  Heterogeneous hydrolyses of N20c and NC^;
     5)  Wall  loss of ozone;
     6)  Excess NO oxidation caused by background organics;
     7)  Dilution rates; and
     8)  Characterization of temperature  and humidity.

There  is  significant  uncertainty regarding  the most appropriate  way  to
represent some of  these  aspects and different  modeling  groups have  made
different assumptions.   As  in our  previous mechanism evaluation studies
(Carter  et  al.,  1986,   1987,  1988), the goal  in  this  study  was  not  to
represent  these  effects  exactly  (since this  is not possible  given  the
current  state of  knowledge),  but to represent  these effects in the  four
chambers  in a consistent manner  and with the minimum number of  empirical
parameters  required to be  consistent with  the  available  characterization
data.
     We  had developed  a comprehensive  set  of  procedures and  assumptions
for  representing  these  chamber-  and   run-dependent processes during the
course  of our previous mechanism evaluation studies  (Carter  et  al.,  1986,
1987).   However, prior to conducting the evaluation,  these procedures and
assumptions were  reviewed and  updated  where  new  data  were  available.
Procedures  for  characterizing  the  SAPRC  chambers  were  almost  entirely
adopted from those used in our previous evaluation studies (Carter et al.,
                                     102

-------
1986,  1987)  and  therefore  are  only briefly  summarized  here.    However,
significant improvements  were made  in  the  procedures to  characterize  the
light intensity and spectral  distribution in the UNC  chamber,  based  on  the
recent work carried out  by Jeffries et  al.  (1989a).  These changes  in  the
light characterization  for the UNC  chamber necessitated other  refinements
in the characterization procedures  for that chamber.
     4.2.1  Light Characterization
            The rates of  the photolysis reactions  in  environmental chamber
simulations  are  obviously  important   since  these  reactions  supply  the
energy to  drive  the chemical  transformations that  the model  is  attempting
to simulate.   For both indoor and  outdoor  chamber  simulations,  photolysis
rate constants were calculated using the following  equation,

                  kPnot(t)  =  I [ J(x,t)  c(x) *(X)  ]                  (I)
                              X

where  kPhot(t)  is  a photolysis  rate  constant  at time  t,  J(X,t)  is  the
intensity  at  time t of the light source in  a wavelength region centered at
wavelength X  and o(X)  and 4>(X)  are  the  absorption coefficient  for  the
photolyzing  species and the quantum yield  for the photolysis reaction over
that wavelength  region.   For each photolysis reaction, the  values  of the
absorption coefficients  and  quantum yields are  considered  to  be part of
the  homogeneous  reaction mechanism (given  in  Section 2).   On  the other
hand,  the values  J(X,t)  are determined  by  the  intensity and  spectral
distribution  of  the  light source.  Thus,  specification  of  this  parameter
 is part of  the  chamber-dependent  input in the  model simulations  of the
chamber  experiments.   The  procedure used  for deriving J(X,t)  depends on
whether  the  chamber is an indoor or an  outdoor  chamber.
      Light Characterization  for  Indoor Chambers.    The  light  intensities
and   spectral  distributions   used   for  the SAPRC  EC  and  ITC are those
 discussed  by  Carter  et al.  (1986).   These are believed  to be reasonably
 well characterized and are assumed to  be  constant throughout each  experi-
 ment.   For both  chambers,  the light intensity is monitored by carrying out
 periodic NOp actinometry  experiments,  and  the results of  these  are  used to
 assign k1  values for  various groups  of  runs.   For most experiments the
 values of the NC>2 photolysis  rates (k^)  were in  the  range  of 0.2 to 0.4
                                     103

-------
min  ,  with the values changing relatively slowly over time.  Based on  the
observed  variability  of  the  k^  measurements,  the  values  assigned  for
individual runs are estimated to have an uncertainty of -10/1 or less.
     The  ratios of rates of  the  other photolysis  reactions  to those  for
N02 are  determined by  the spectral distribution of the  light source.  For
the experiments carried out  in the SAPRC EC, the relative spectral distri-
bution was measured during  the course of each run,  as described by  Pitts
et al.  (1986).   For runs carried out  around  the  same  time  and with  the
same lamp, essentially the same spectral distribution  was observed, though
with some run-to-run variability; but over longer periods of  time, system-
atic changes  in  the  spectral  distribution occurred, with  the relative
intensity  in  the  UV  region  becoming  progressively  less.    In  the  model
testing,  consecutive  runs  with  similar  measured  spectral  distributions
were  grouped  together,   and  the  averages of   their  measured spectral
distributions were used  for  all  the runs in  the  group.   The  sets  of  EC
spectral distributions which  were employed are summarized by  Carter et  al.
(1986).
     The  relative  spectral  distribution  for the blacklights used in  the
ITC was  measured   much less  frequently, but the  few  measurements  carried
out  indicate  that it  can  be assumed  to  be  essentially  constant.    The
measured spectral  distribution for  this chamber is  also  given by Carter et
al. (1986).   Compared  to the solar  or  EC spectral distributions,  the  ITC
spectral  distribution is  relatively more intense in the ~330 -  370 micron
region,  but  is  much  less   intense  for  wavelengths  greater  than   -380
microns.
     Light  Characterization  for Outdoor Chambers.    The  intensity  and
spectral  distribution of  the light source  for the  UNC and SAPRC OTC out-
door  chambers,  which  consists  of  natural  sunlight  passing through  the
chamber  walls,  is  more difficult to characterize.  Both  the  intensity  and
spectral  distribution of  the light source  depend  on the time of day,  the
time of year, and  the local meteorological conditions.  Jeffries  (1989a)
recently  completed an  extensive experimental and theoretical  study of the
light  characteristics  inside and outside  the  UNC  chamber  and  developed a
model  and  a data  base  for  deriving  light  intensities  and  spectral
distributions as  a function  of  time for each of the UNC  chamber  experi-
ments  whose  data  are available for  model  evaluation.   This  new UNC  light
                                    104

-------
characterization data  base  was employed in this study.  In principle,  the
procedures employed  to develop the UNC  chamber  light characterization data
base could  be employed  to  characterize the light  in  the  SAPRC OTC runs,
but there was  insufficient  time and  resources available  in  this  program to
carry  this  out.    Therefore,  the  methods  used  to   represent  OTC light
characteristics in our previous mechanism  evaluation studies,  described by
Carter et al.  (1986), were also employed in this study.
     The overall  features of  the methods  used for light  characterization
for  the OTC and  the  UNC  chambers  are summarized below.   However,   the
reports of Jeffries  et al.  (1989a,d)  (for  the UNC  chamber)  or  of Carter ec
al. (1986) (for the  OTC)  should be consulted for details.
     Light Characterization  of the  SAPRC OTC.  The light  characterization
data available for the OTC experiments consist of  continual UV  radiometer
measurements of sunlight outside  the  chamber  and, for many  experiments,
continual  N02 actinometry (k^)  measurements  made  underneath  the  reaction
bag, with  the light  passing in and  out of  the  chamber before  reaching  the
actinometer.  Such data  can be used to estimate how total  light intensity
is  varying  with time outside the chambers,  but  they give no direct indica-
tion of how  the relative spectral distribution  is  varying,  or  (except  to a
limited  extent  for  the k1  measurements   under  the  OTC)  how  the light
intensity  and  spectral  distribution inside  the chamber  may  differ  from
that   outside.     Therefore,   much  of  the  light  characterization  input
required  for   the  OTC  was  derived based  on  theoretical  calculations,
assumptions  and approximations which are discussed  below.
     The   light  intensities  used   in   calculating  the  photolysis  rate
constants  for  a given OTC  chamber  experiment  are  given by J(X,t), which
was calculated as  a  function of wavelength (X)  and time  of day (t) accord-
ing to the  equation:
                                UV(t)
         J(X,t)  =  J  (z[t'J)   	   fU.t)                       (II)
                  0          UV (z[t'j)
 where z[t']  is  the  solar  zenith  angle  at  time  t1, which  is our  best
 estimate of  the true  solar time  t  of the experiment,  J0(z[t'])  is  the
 theoretically-calculated  clear-sky   light   intensity   for  the   region
                                     105

-------
associated  with  wavelength  g,  UV(t)  is the  experimental  UV  radiometer
reading taken  at  time t during the  experiment,  UV0(z[t'])  is the theore-
tically calculated  clear-sky  UV radiometer reading  for z[t'],  and  f(X,t)
is a correction factor  to account for the effects of transmission of light
through the chamber walls.
     The  clear-sky  light  intensities given  by  Peterson  (1976)  for  his
"best  estimate"  surface albedos were used  to  derive  J0(x,z[t'])  for  the
OTC experiments.  The values  of JQ  used  for a given wavelength and  zenith
angle  were  calculated using an empirical  function fit  to  the  data  tabu-
lated  by  Peterson  (1976) as  described by Carter et al.  (1986).   The solar
zenith angle  z(t')  was  calculated  from the  true solar  time,  the date,  and
the latitude  of  the SAPRC OTC  (34.06 degrees)  using the standard  formula
as employed by Carter et al. (1986).   Because  the clocks used during  the
outdoor  chamber  experiments  were  not  always   accurate (Carter  et  al.,
1986), the  difference between the  clock  time t and the  true  solar  time t'
were  derived   based  on  adjusting  t-t'  so  that the  experimental  and  the
theoretically  calculated  UV curve  shapes agreed.   As expected,  the time
offsets did not vary  randomly from  run-to-run,  but were generally  the same
for runs  carried out  around the  same time  period.    For  a few runs  the
light  quality  was so  poor  that  it  was not possible to determine  t-t1 using
the light intensity  data,  and  those runs  were  assumed to have the same
time offset as observed in runs carried  out at  around the same time.
     The  experimentally measured  UV data  are  used  to take into  account
local  meteorological  factors  which  affect the light intensity entering the
chamber as  a  function  of  time  during the  individual  chamber experiments.
As indicated  in  Equation  (II),  this is done by  assuming that for  any wave-
length the  ratio of the actual light  intensity  outside the  chamber  to JQ,
the  theoretical  "clear  sky" light  intensity  value,  is the same  as  the
ratio  of  the  measured UV  radiation  intensity  to  UVQ,  the  calculated
estimated "clear  sky" UV  values,  which,  like JQ, are assumed to be a func-
tion   only  of the  solar  zenith  angle.    These latter  quantities  are
calculated  from  an  empirical  formula  (Carter  et al.,  1986),  which is
expressed as:
                                       expt        calc
           UV  = f     (z)   lim (UV/k )    (0)    k     (0)             (III)
             0    shape     z->0     1             1
                                     106

-------
where fsnape  is  an empirical function, normalized to unity for z=0,  which
was derived  based on  fits  to the shapes of  experimental  UV vs  time  data
obtained  during  SAPRC  OTC  and  UNC  outdoor  chamber  experiments  on  clear
days; lim(z->0)  (UV/k1)exPfc(0)  is the estimated z=0 ratio of UV  intensity
to the  N02 photolysis rate  (k^  derived  based  on the empirical  relation-
ship between  simultaneous experimental UV  radiometer  and k1 measurements
given  as  a  function of zenith  angle  by  Zafonte  et  al.  (1977);  and
k1ca c(0)  is  the calculated  N02  photolysis rate using  the  JQ values  for
z=0.
     At  both  the SAPRC  and  the  UNC  outdoor chamber  laboratories the  UV
instrument  is located  outside  the  chamber, and  thus  does  not  directly
measure  the light  intensity inside the chamber.  In the case of  the  SAPRC
OTC, which  consists  of a flexible Teflon  bag held on a  rope netting  above
a  green indoor-outdoor  carpet,  the  major  cause  for  differences  in  light
intensity  between  the inside and the outside of the chamber is assumed  to
be  the  attenuation  of  the  light  as  it passes  through  the Teflon walls.
The k1  data taken during the OTC  runs are  used to obtain a rough  indica-
tion of the magnitude of this attenuation,  since  the tube used  to obtain
these data  is  located underneath the  reaction bag,  and thus  light reaching
it passes  through the chamber walls  twice.    A  comparison of observed  with
theoretically  calculated k1  readings  for these OTC  runs  (Carter et  al.,
1986)  indicates   that the  light reaching  the  k1  tube  under  the OTC  is
attenuated  by  a  factor of -1.44.   Since the light has to pass  through  two
Teflon  surfaces  (the top and the bottom of  the chamber) before  reaching
the NC>2 actinometer, this corresponds to a suppressipn  factor  of 1.2  when
passing  through  each wall of the  chamber.   Thus in modeling the OTC  runs
for  this study,  we  use  a wall transmission  correction  factor of 0.83  in
equation  (II) when calculating the light intensity.
     This  estimation  that   a simple  0.83   correction   factor  takes   into
account  all  the  differences between  the   light  intensity  and  spectral
distribution  inside  the  OTC and  the theoretically  calculated  values  is
obviously  subject to  a  number of uncertainties,  since it  does not  take
into account,  for  example,  possibilities that  the walls, or  reflections
from the surfaces  and structures around the OTC may be changing the zenith
angle  dependences  of  the  light  intensity,  or  that  attenuation and/or
scattering  of  light as it passes  through the walls may cause  a  change  in
                                    107

-------
the spectral  distribution.    These uncertainties  are  expected to  be the
greatest at  the higher  zenith angles,  but for  most  SAPRC OTC  runs the
chamber is covered before 0900 PST and after  1500 or  1600 PST, so uncer-
tainties associated with light  characteristics at high zenith angles will
not affect  model  simulations of  most SAPRC  OTC  runs.   Based  on  these
considerations, we  estimate  that  the light  intensity inside  the  OTC  is
probably uncertain  by  -15J;  though  this  is  a subjective estimate.   The
uncertainty is  probably  somewhat  less at  the wavelength region affecting
the photolysis  of  MOo,  and probably  somewhat greater  at  the  lower  wave-
length region  which affects the photolysis of aldehydes and the formation
of 0(1D) from 0^.
     It should  be  noted that  both the k1calc(0)  values  used  in Equation
(III), and the  comparison of  observed  and  calculated k1 values  used  in  the
derivation of  the wall transmission correction factor  depend on the  set  of
absorption  coefficients  and  quantum  yields  assumed  for  N02-   However,
these  two  factors  have  opposing effects on the calculation of  the  overall
in-chamber light  intensity  in Equation  (II),  and  thus the light  intensi-
ties used  in  the model  simulations  are not  sensitive to  the sets  of  N02
absorption coefficients  and quantum yields used  in this analysis.   At  the
time of the  derivation  of the data  base used for  the  modeling of  the  OTC
runs,  we used  the  N02 absorption  coefficient  and quantum yields recommend-
ed by  NASA  (1985),  which are  somewhat different (yielding,  for example,  1%
lower  calculated  N02 photolysis   rates  for  z=0)   than those  used  in  the
current RADM  (see Section 2)  and  SAPRC (Carter  et al.,  1988)  mechanisms.
However,  since as  indicated  above these  differences  will tend to  cancel
out  in the derivation  of the overall  light intensity,  and  since  there are
a  number  of much  more  significant uncertainties  involved  in  this  deriva-
tion,  it  was  determined not  to  be productive  to expend  the  significant
effort required  to re-derive  the OTC  light  characterization data base
using  the  updated N02 absorption  coefficients and quantum yields.   If the
OTC  light  characterization  data base is to be re-derived,  it would be more
appropriate  to use  the updated  procedures developed  by  Jeffries  (1989a)
for  re-deriving the UNC chamber  light characterization data,  as discussed
below.
                                     108

-------
     Updated Light  Characterization of the  UNC  Chamber.   In our  previous
mechanism  evaluation  studies  (Carter et   al.,  1986,  1987),   the  light
characterization input  data used in the model simulations of the  UNC  out-
door chamber were  derived in a manner similar to that described above for
the SAPRC  OTC.   However, this previous light characterization  for the UNC
chamber  was considered  to  be highly  uncertain  for  a  number of  reasons.
The particular  uncertainties  in UNC  light  characterization concerned  the
attenuation and  reflection of light  inside  the  chamber due to  the Teflon
walls  of  that  chamber,  the  effects   of  enhancement of  light  due  to  the
reflective chamber floor,  the  fact  that (unlike  the case  for  the  OTC)
irradiations in  UNC chamber experiments begin with  the sun very low in the
sky, and evidence  that at certain times  in  the  past  the  UNC  UV instrument
was  significantly  out  of calibration.   Because of  lack of data,  highly
uncertain  and  simplified  assumptions had   to  be  made   in  our  previous
mechanism  evaluation  studies concerning  these effects  (see Carter et al.,
1986),  such as, for  example,  assuming that  effects  of attenuation due to
wall  absorption and  enhancement  due  to  floor reflection exactly  canceled
out,  and thus  using a  net  chamber  transmission  factor  of 1.   In addition,
correction factors for  UV  data  when  the instrument  was  apparently out of
calibration were roughly estimated based on incomplete  information.   For
these  reasons,  we  determined that the light characterization model for the
UNC  chamber should be  updated and  improved  as  part  of  the  effort carried
out  in this program.
      For a number  of years, Dr.  Harvey Jeffries and co-workers at UNC have
been  working on developing new models for  solar radiation and  for light
characteristics inside  the UNC  chamber  (Jeffries  et  al.,  1989a)  and on
developing correction factors for questionable  light characterization data
from past UNC  chamber  runs.   Therefore, as a  part of  this  program, the
results of this effort have been utilized   to develop  a  new  data base for
light characterization  for use  in modeling UNC chamber  runs.   The major
part of the effort at UNC in developing and testing new light character-
 ization models  for solar  radiation  and  radiation  inside the  UNC  chamber
were  carried   out  under  separate  funding and   is described   elsewhere
 (Jeffries  et  al.,  1989a).   For  this  program,  Jeffries and co-workers
 performed quality assurance and applied  their best  estimate corrections to
 the  past  UNC   light   characterization  data, and,   for  each  UNC  chamber
                                     109

-------
experiment whose  results  are  currently  available for  modeling,  derived
sets of wavelength- and  time-resolved  light  flux data for use in modeling
the run.   These  consist  of in-chamber  actinic  flux values  for  91  wave-
length intervals for each  16 minute interval in  the runs.  These data were
then delivered  to  SAPRC  in computer readable  form,  and the SAPRC chamber
model  simulation   software  was  adopted  to  incorporate  these data when
simulating the UNC chamber runs.
     A description  of the  models,  data base,  and procedures utilized  to
develop  these  run-specific  light  characterization   data   is  given   by
Jeffries et al. (1989a)  and is beyond  the scope of this  report.  Briefly,
the procedure involves the following:
     (1)  A  spectral  radiation  transfer model  is  utilized  to calculate
clear-sky solar radiation  fluxes outside the chamber.   Its  input includes:
solar  zenith  angle;  air pressure; total  atmosphere  ozone  column;  total
atmosphere  water  vapor   column;   specification  of  atmospheric  aerosol
density,  type,  and  optical properties; relative  humidity;  and spectral
albedo of the surrounding surface.
     (2)  A  broadband radiation  transfer model  is  utilized  to calculate
clear-sky  UV and  TSR predictions,  which  can  be  directly  compared  with
measured  values taken during  an experiment.   Its input requirements  are
the same as the spectral radiation transfer model.
     (3)  A  relatively  simple   cloud  transmission  model  is utilized  to
account for  the effects  of clouds on  surface level radiation  fluxes  and UV
and TSR  readings.   Its  input consists of ratios of cloudy-to-clear-sky UV
and TSR values; these are derived by  fitting calculated to observed  UV and
TSR  readings for  various  time  periods  in  the  runs  as  indicated  below.
These  ratios  then account for  the effects of clouds on the near UV  and the
higher wavelength spectral  regions.   No  attempt  is  made to  convert  the
broadband  transmissions  derived  from  the UV and TSR data into wavelength-
dependent  spectral transmissions  in  the current  version of  the UNC cloud
model.   However,   Jeffries  et  al.  (1989a) indicate  that such a refinement
 is unlikely  to have a significant effect if the clouds attenuate less than
 ~HQ%  of the  light intensity.
      (4)   Models  for spectral transmission, reflection,  and  absorption of
 the FEP Teflon film  used  in the UNC  chamber, for reflection of light off
 the floor of the chamber, and for other chamber-specific light effects are
                                     110

-------
utilized  to  calculate  the light  fluxes  inside the chamber.   Its  inputs,
besides  those  specifying characteristics  of  the  chamber  which  remain
constant,  include  the  actinic fluxes  outside the  chamber  for  both  the
direct  and  diffuse components, the  solar zenith  angle,  and transmission
factors  for  the  direct  and   the  diffuse  radiation.    This  model  was
developed and validated  utilizing an extensive series  of spectral flux  and
UV and  broadband light intensity  measurements made inside  and outside  the
UNC chamber  (Jeffries et al.,  1989a).
     For  each  run,  the procedure  for   developing  the  run-specific  in-
chamber  light fluxes involved  the following:   (1) Values  of run-specific
parameters  such as humidity,  total  column  ozone,  etc., which are  derived
from meteorological  observations,  measurements made during  the run,  etc.,
and other available  data bases are entered into the data base used by  the
models.   (2) Other  parameters required  for calculation of out-of-chamber
light  fluxes, such as  atmospheric aerosol  parameters  and  broadband  cloud
transmission factors,  are estimated  based on  time of year, meteorology,
and past experience with the model and then are adjusted  to yield  accept-
able fits of observed-to-calculated  UV and  TSR readings for the  individual
run.    Cloud adjustment factors  are  derived  for  approximately  three-hour
intervals,  which is  a  similar time resolution used to derive the  analogous
factors in the  previous development  of the UNC and OTC light characteriza-
tion data base  by Carter et al. (1986).  The  types  of fits of  calculated-
to-observed UV  and TSR  data obtained after the application of  this proce-
dure  are illustrated in  sample plots  given  by Jeffries (1989d).   (3)  The
UNC  chamber transmission, absorption,  and reflection model is then used to
calculate the  in-chamber  light fluxes for each 16-minute  interval of the
run.   These  fluxes are  incorporated into computer data sets (one for each
run-day) which  were  then transmitted  to SAPRC  for  use in modeling.
      The updated UNC  light  characterization data  were found to  be differ-
ent  in  a number of respects  from  the UNC light  characterization  data of
Carter et  al.  (1986)  which  was employed in our previous  modeling studies
of this  chamber.   The  major differences result  from the use of  the new
 inside-vs-outside  chamber  model,  which predicts   enhancement  of  light
 intensity at longer wavelengths due to floor reflection and  suppression at
 shorter  wavelengths due to  light absorption by the  walls.  [As indicated
 above,  because of lack  of  data, the Carter  et  al.  (1986) model assumed
                                     111

-------
that these  effects  canceled and thus  the  inside and  outside fluxes were
the same.]   The magnitude of  these  chamber light  effects  depends on the
time of  day and on  the relative amount  of direct  and diffuse radiation
calculated by the outside radiation  transfer model.
     A  comparison  of  the  effects  of  changing  from  the  previous  light
characterization of Carter  et  al.  (1986) to  this new light  characteriza-
tion is  shown  for several  UNC  chamber runs  in  Figures 2 through 4,  where
the former  shows results for  a formaldehyde-NOv-air  run,  and the  latter
                                                A
two show runs employing complex surrogate  mixtures.   [All  these  simula-
tions used  the  Carter  et al.  (1986) set of assumptions for  other  chamber
effects, though as discussed below  some of the adjustable  chamber  effects
parameters were subsequently modified  in  this study to be  more consistent
with simulations of characterization  runs  using the new light model.)   As
illustrated  in  Figure  2,  the  new  characterization model  tends  to  yield
significantly better fits of formaldehyde decay  rates  in  the  formaldehyde-
NOX runs.   However, in  the  case of this run the  new model does not  improve
the fits to the initiation time for  NO-to-NOp  conversion,  though  the new
model causes an overprediction of  the initiation  time,  while in  the old
model  it  is underpredicted.   These  effects  can  be attributed  to the
suppression  of UV intensity by the  chamber walls  at low  zenith  angles,
which is represented  in the new but not the old  model.  On  the other hand,
in many  other  types of runs, such as  that shown  in  Figure 3,  the  new light
characterization  model  results in  higher  predictions  of maximum  ozone
yields,  presumably  due to  the effects of assumed  enhancement  of  light
intensity  caused  by  the chamber's  reflective floor.   This  tendency of the
new characterization model  to  increase predicted initiation times but also
increase maximum  ozone yields  was  the case to varying degrees for  most  of
the UNC  chamber runs  modeled.   However, for some runs,  such  as that shown
in  Figure  M, the effect of using  different light  characterization models
was relatively  small.
     4.2.2   Representation  of Other Chamber-Dependent  Effects
             Assumptions regarding  chamber-dependent   reactions   and  the
presence of reactive  background species have been  recognized for a number
of  years as being important in affecting  results of  model simulations  of
chamber  experiments.     For  example,   it  has been  recognized  for  several
years   that  overall   rates of initiation  in  environmental chambers are
                                    112

-------
            OZONE
 E
 a
cr
UJ
CJ
z
o
o
        300
  420   540   660   780   900



NITROGEN DIOXIDE
                                           0.16-1
                                           0.12-
                                           0.08-
                                                 NITRIC OXIDE
                               0.04-
300   420   540   660   780   900



   FORMALDEHYDE
        300
  420   540   660   780   900


                    ELAPSED  TIME (minutes)
1.0-

0.8-
0.6-
-i
0.4-
-
0.2-
_
0.0-
-^•B ^^ ~"
*^<7 r-
Y\.
\*
\ •
v. \

\"'"

^s> • •

                                             300   420   540   660   780   900
 Figure  2.   Experimental and Calculated Concentration-Time  Plots for

             Selected Species in the UNC Formaldehyde-NO  -Air  Run AU0279B,
             [All  calculations used the March  1988  RADM mechanism and the'
             Carter et al.  (1986) wall characterization model.]


     	   Calculated Using the Carter et al.  (1986) Light Model
     - » -   Calculated Using the New UNC Light  Model
       *    Experimental Data
                                      113

-------
           OZONE
 E
 Q.
 Q.
g
*
LJ
O
2
O
O
                                           0.32-,
                                           0.24-
                                           0.1 6-*
                            0.08-
                                                 NITRIC  OXIDE
        300  420  540  660  780  900  1020
           NITROGEN DIOXIDE
                                           0.00
                               300  420   540  660  780  900  1020
                                                 PAN
                      "i     i	1	r
        300   420  540  660  780  900 1020
                               300  420  540  660  780   900  1020
                                ELAPSED  TIME (minutes)
   Figure  3.
Experimental and Calculated  Concentration-Time Plots  for
Selected Species in the UNC  Complex Mixture-NO -Air Run
ST1682B.  [All calculations  used  the March 1988 RADM
mechanism with adjusted OH rate constants for lumped
hydrocarbon species and the  Carter et al. (1986) wall
characterization model.]

Calculated Using the Carter  et al. (1986) Light Model
Calculated Using the New  UNC Light Model
Experimental Data

-------
            OZONE
                                                 NITRIC OXIDE
 E
 a
 a
O

<
a:
LU
o
z
o
a
300  420   540   660  780  900  1020



   NITROGEN DIOXIDE
300  420  540   660   780   900  1020
                                                 PAN
              ,	,	j	f

        300  420  540  660  780   900  1020
                                             300  420   540   660   760   90C  1020


                               ELAPSED  TIME (minutes)
  Figure 4.   Experimental  and  Calculated Concentration-Time Plots  for

              Selected  Species  in the UNC Complex Mixture-NO -Air Run

              ST2981R.   [All  calculations used the March  1980* RADM  mechanism

              with adjusted OH  rate constants for lumped hydrocarbon species

              and the Carter  et al.  (1986)  wall characterization model.]



        	 Calculated  Using the Carter et al. (1986) Light Model

        - •-  - Calculated  Using the New UNC Light Model

               Experimental  Data
                                      115

-------
higher than can be accounted for by known homogeneous processes  (Carter et
al.f 1979a), and  successful  model  simulations of chamber experiments have
had  to  include  provisions  for "chamber  radical  sources,"  whose exact
physical cause remains  unknown  or  speculative.   In addition,  formation of
PAN and ozone in acetaldehyde-air  runs with no added NOX has been observed
in  all  four chambers whose  data  are  modeled in  this  program  (e.g., see
Carter et al.,  1986),  indicating that offgassing of NOX must  be occurring
from the walls of these chambers.   Excess  rates of NO oxidation observed
in  N0v-air  irradiations,  and   ozone  formation  in pure  air  irradiations
      A
indicate that  unmeasured  reactive organic species are also  either off-
gassed from the chamber  walls or present in the  background  air in environ-
mental chamber experiments.   Heterogeneous hydrolyses of NOp  and NpOc and
wall losses of  ozone are  also  known to occur.   All of  these need to be
represented in model simulations of chamber experiments.
     The methods  we  used  in representing  these chamber-dependent  effects
in  the model simulations carried out in this study are summarized  in  this
section.  Methods  used  to represent other chamber- or  run-specific condi-
tions  such  as water  concentration,  temperature,  or  dilution  in modeling
runs from the various chambers  are also  summarized.  The general approach-
es  and  types  of  assumptions  employed  in representing these  effects  is
essentially the  same as those described  in  Carter et  al. (1986).   That
report should  be consulted  for detailed discussions  of these effects  and
for  the  reasons  why  we chose to use the specific  types of  representations
employed.  Since  the  time of the Carter  et al.  (1986)  report,  we have made
some  minor  modifications  in  the magnitude of  the  continuous,   light-
dependent portion of the  chamber  radical  source used when modeling SAPRC
ITC and  OTC experiments (Carter  et  al.,  1987) and  in  this  study  had  to
make more extensive modifications  in the representation of chamber  effects
in  the UNC  chamber  as a result of the changes  in  the UNC  light character-
ization  model.    These  changes  to  the  UNC  characterization model  are
discussed  below,  following  the  summary   of  the  representation  of  the
chamber-dependent  parameters for the SAPRC  chambers.
      Representation  of Chamber- or  Run-Dependent Parameters  in the SAPRC
Chambers.   Tables 13 through 15 summarize  the  methods  and  parameters used
to   represent  chamber-dependent   reactions,  effects  of offgasses   or
background  species,  and  other  chamber-  or  run-specific conditions  for
                                     116

-------
Table 13.  Summary of Chamber-Dependent Parameters Used in the Model
           Simulations of the SAPRC EC Chamber Runs
 Parameter(s)
                        Value and  Derivation
General Run Conditions

 Water Concentration
 Temperature


 Dilution
                         Based on  measured  temperature and
                         humidity  during  the  final  pure air
                         fill of  the  chamber.  Typically
                         2.0 x  104 ppm.

                         303 K.   This is  the  typical  average
                         temperature  in EC  chamber  runs.

                         3.0 x  10*4 min'1  (Pitts  et al.,
                         1979).
Chamber  Radical  Sources

  Initial Nitrous Acid  (MONO)
      wall
  NO-
   0.5 HONO +
     0.5 loss of N02
  NO-
   wall
hv  ->  0.
        0.
HONO +
loss of NO-
      wall
  hv   ->  OH
             Initial HONO = 1% of the initial N0?,
             based on measured or estimated initial
             HONO concentrations in EC tracer - NOX
             air irradiations (Carter et al., 1982;
             Pitts et al., 1983).

             Included to represent the dark
             formation of HONO from N02 observed in
             this chamber.  Measured rate constant in
             the EC is 2.8 x 10~4 min"1 (Pitts et
             al., 1984).
                                                                       -1
Included to represent the N02-dependent
component of the continuous,  light-
dependent radical source in this chamber
Rate constant given by 2.16 x 10"-^ ppm
x k1f where k« is the N02 photolysis
rate.  Derived based on results of
tracer-NOx-air irradiations (Carter et
al., 1982; Pitts et al., 1983).
                         Included to represent the N02-independent
                         component of the continuous, light-
                         dependent radical source.  Rate given by
                         0.39 ppb x k1? where k1 is the N02
                         photolysis rate.  Derived based on
                         results of tracer-NOx-air irradiations
                         (Carter et al.,  1982; Pitts et al., 1983)

                                                       (continued)
                                     117

-------
Table 13 (continued) - 2
 Parameter(s)
                                Value and  Derivation
NOX Offgassing

 wall + hv ->  N02




Background Organics

        (none)
Other Wall Reactions

     wall
  0->  -> lOSS Of Oo
    wall, H20
N205   ->  Loss of N205
                                Rate given by 0.5 ppb x k1t the N02
                                photolysis rate.   Based on model
                                simulations of the acetaldehyde-air run
                                EC-253 (Carter et al., 1986).
                                No representation of effects of
                                background organics is used in model
                                simulations of EC runs. (Carter et al.,
                                1986).
                                First order loss rate of 1.1 x 10"
                                min   is used, based on results of
                                ozone dark decay experiments (Pitts et
                                al., 1979; Carter et al., 1986).
                                Rate constant given by
                                M.65 x 1CT3 + 7.21 x 10'7 [H20] min'1,
                                based on measurements of N20c loss rates
                                in the EC by Tuazon et al. (1983).
                                    118

-------
Table 14.  Summary of Chamber-Dependent Parameters Used in the Model
           Simulations of  the SAPRC ITC Chamber Runs
 Parameter(s)
Value and Derivation
General Run Conditions
 Water Concentration
 Temperature


 Dilution
Based on measured temperature and
humidity during the final pure air
flush of the chamber.  Typically
2.0 x 1CT ppm.

303 K.  This is the typical average
temperature in ITC chamber runs.

No dilution is assumed to occur, due to
flexible nature of chamber walls.
Chamber  Radical  Sources
  Initial  Nitrous Acid (HONO)
Initial nitrous acid  is assumed  to be
negligible  in Teflon  film chambers
(Carter et  al., 1986).
      wall
  N02   ->   0.2 HONO
              0.8 loss of N02
 Included to  represent the dark  formation
 of HONO from N02 observed in  this
 chamber.  The  pseudo-first-order rate
 constant used  for  the ITC is  1.4 x  10
 min   .  This rate  constant, and the
 assumed HONO/N02 stoichiometry, is  based
 on studies of  tnis process  in a smaller,
 1300-liter Teflon  film  chamber  (Pitts  et
 al.,  1984),  corrected for differences
 in chamber size  (Carter et  al.,  1986).
      wall
  hv   ->  OH
 This  is  the only  process  used  to
 represent  the continuous  light-dependent
 component  of the  radical  source in Teflon
 film  chambers (Carter et  al.,  1986).   The
 rate  used  is given  by kp<> x k-j, where
 k,  is the  NOo photolysis  rate,  and
 kps is determined by averaging, for
 each  reaction bag,  the radical  input
 rates measured using tracer-NO -air
 irradiations (Carter et al., 1982).  The
 kRS values used in modeling ITC and OTC
 runs  are given in Table 16.

                               (continued)
                                     119

-------
Table 14 (continued) - 2
 Parameters}
                               Value and Derivation
NOH Offgassing

 wall + hv ->  N02
                                Rate  given  by 0.15  ppb  x  k1?  the N02
                                photolysis  rate.  Based on model
                                simulations of  acetaldehyde-air runs  in
                                this  chamber.
Background Organics

     wall
 OH   ->  H02
                                This "reaction"  is used to simulate the
                                effect of trace  contamination  by reactive
                                organics which convert NO to NQ2.   The
                                rate constant used is 250 min   , based
                                on model simulations of pure air
                                irradiations (Carter et al., 1986).
Other Wall Reactions

     wall
  0->  -> loss of 0-3
    wall, H20
N20c   ->  Coss of
                                First order loss rate of  1.3 x  10~
                                min"' is used, based on results of
                                ozone dark decay experiments (Carter et
                                al.,  1986).
                                 Rate  constant given by
                                 2.5 x 10'3  + 5.0 x  10~°  [H20] min'1,
                                 based on  measurements of N20c loss  rates
                                 in a  smaller Teflon bag  reactor  by  Tuazon
                                 et al.  (1983),  adjusted  for  differences
                                 in chamber  sizes as discussed by Carter
                                 et al.  (1986).
                                     120

-------
Table 15.  Summary of Chamber-Dependent Parameters Used in the Model
           Simulations of the SAPRC OTC Chamber Runs
 Parameter(s)
Value and Derivation
General Run Conditions
 Water Concentration
 Temperature
Derived based on continuous measurements
of dew point and temperature data for
runs where such data are available, as
discussed by Carter et al. (1986).  If
such data are not available, a default
water concentration of 5 x 1CH ppm was
used.  (OTC runs use unhumidified pure
air, so the typical water concentration
is relatively low.)

Plots of measured temperatures against
time were approximated by straight lines,
which were used to derive the temperature
as a function of time in the model
simulations.
  Dilution
No dilution is assumed to occur, due to
flexible nature of the chamber walls.
 Chamber Radical Sources
  Initial Nitrous Acid (HONO)
      wall
  N02  ->  0.2 HONO
            +0.8 loss of NO-
      wall
  hv   ->  OH
Initial nitrous acid  is assumed to be
negligible  in Teflon  film chambers
(Carter et  al., 1986).
 Included  to  represent  the dark  formation
 of  HONO from NOo observed in Teflon
 bag chambers.  The  pseudo-first-order  rate
 constant  used  for the  OTC is 9.0  x  10~5
 min  .  This rate constant, and the
 assumed HONO/NOp stoichiometry, is  based
 on  studies of  this  process  in a smaller,
 4300-liter Teflon film chamber  (Pitts  et
 al.,  1984),  corrected  for differences  in
 chamber size (Carter et al.,  1986).
 This is  the only  process  used to
 represent the continuous  light-dependent
 component of the  radical  source in Teflon
 film chambers (Carter et  al., 1986).   The

                               (continued)
                                     121

-------
Table 15 (continued) - 2
 Parameter(s)
                                Value and Derivation
                                                          x k^,  where
NOX Offgassing

 wall + hv ->  N02




Background Organics

     wall
 OH   ->  H02
Other Wall Reactions
     wall
  0,   -> loss of
     wall,
    -   ->   Loss  of
                                rate used is given by k
                                k-| is the N02 photolysis rate, and
                                kpg is determined by averaging, for
                                each reaction bag, the radical input
                                rates measured using tracer-NCL-air
                                irradiations (Carter et al.,  1982).  The
                                kRS values used in modeling ITC and OTC
                                runs are given in Table 16.
                                Rate given by 0.1 ppb x k1? the N02
                                photolysis rate.  Based on model
                                simulations of acetaldehyde-air runs in
                                this chamber.
                                This "reaction" is used to simulate the
                                effect of trace contamination by reactive
                                organics which convert NO to NOp.  The
                                rate constant is assumed to be the same.
                                                                      -1
                                as used for  ITC runs, which  is 250 min
                                First order loss rate of  1.67 x  10"
                                min"1 is used, based on results  of
                                ozone dark decay experiments (Carter et
                                al., 1986).
                                Rate constant given by  1.6  x  10"^ + 3.0 x
                                10   [H20] min   , based on  measurements of
                                N20c  loss rates  in  a smaller Teflon  film
                                reactor  by Tuazon et al.  (1983),  adjusted
                                for   differences  in   chamber  sizes   as
                                discussed  by Garter et al.  (1986).
                                     122

-------
experiments carried out  in  the  SAPRC  EC,  ITC,  or OTC.  In the case of  the
ITC and  the  OTC,  separate values of  k^,  the  parameter used to  determine
the rate  of light-induced continuous  production of radicals from unknown
sources,  were  employed   for  experiments  using different  reaction   bags.
These  values  were determined by  averaging the results of radical tracer-
NO -air  experiments  carried out with  the  different bags.   [See  Carter  et
  A
al. (1982)  for  a  discussion of  tracer-NOx-air  experiments  and measurements
of chamber  radical sources.]  The averages of  the  measured values of  these
parameters  for  the various  reaction  bags and  the  k^ values used in  the
model  simulations are given in Table  16.   Except  for  the use of separate
values  of kRS for different Teflon reaction  bags  for  ITC and OTC experi-
ments,  the representation of these  chamber-dependent  parameters and run-
specific  conditions  for  the SAPRC chambers is the same as  chat  discussed
by Carter et  al.  (1986).
     The  parameters  shown  in  these  tables are  also  the  same as employed
when  evaluating  the more recent  version of the SAPRC gas-phase  mechanism
(Carter et al.,  1988).
     Representation  of  Chamber-  and  Run-Dependent Parameters  in the  UNC
Chamber.   Tables 17 and  18 summarize  the methods used  to  represent  the
chamber- and  run-dependent effects  when  modeling  the  UNC chamber runs
during the course of  this  program.   Table 17  indicates  the  general  forms
used   for the  representation  of  these  effects   and  the  values  of  the
parameters which were not  varied during this  program [being  in  all  cases
the  same as used by Carter et al. (1986)], while Table 18 gives  the  values
of the  parameters which were varied  during  the course.of  this  study, as
discussed below.
      As  discussed  by  Carter  et al.  (1986),  values of several of  the
 chamber-dependent parameters   for  the  UNC and other  chambers   could  be
 derived  only as a  result of  model  simulations  of  characterization runs
 which are particularly  sensitive to  these parameters.  These are referred
 to as the  "adjustable" chamber-dependent parameters.   Since   run-by-run
 optimization  of  uncharacterized  parameters   is  incompatible  with   the
 objectives  of  using chamber data for  mechanism  testing, this  adjustment
 must  be done  on a  global basis, with the same  sets of parameter  values
 being used  in  modeling  all  of  the runs from a given  chamber.  In the  case
 of  the  UNC  chamber, adjustable chamber-dependent  parameters included k^g,
                                     123

-------
Table 16.  Values of the Continuous Chamber Radical Input Parameter, kRg>
           used in Modeling SAPRC ITC and OTC Chamber Runs, and Averages
           of Measured Values Used to Derive Them
Chamber
ITC







OTC

Inclusive
Run Numbers
450
532
621
692
737
793
859
924
183
209
- 503
- 563
- 638
- 734
- 789
- 832
- 893
- 970
- 206
- 253

Average Measured (a) Used in Model
0.16
0.44
0.21
0.61
0.49
0.36
0.28
0.25
0.25
0.20
± 0.03
± 0.13
± 0.03
± 0.20
± 0.20
t 0.12
± 0.09
± 0.25
± 0.19
± 0.09
0.15
0.4
0.2
0.6
0.5
0.35
0.3
0.25
0.25
0.2
 (a)  Runs carried out with new, unconditioned  reaction  bags  or  runs
     with anomalous  results  were not  counted  in  the  average.
                                     124

-------
Table 17.  Summary of Chamber-Dependent Parameters Used in the Model
           Simulations of the UNC Chamber Runs
 Parameter(s)
Value and Derivation
General Run Conditions

 Water Concentration
 Temperature
  Dilution
 Chamber Radical Sources
  Initial Nitrous Acid (HONO)
      wall
  N02  ->  0.2 HONO
              0.8 loss of N02
      wall
  hv   ->  OH
Derived based on continuous measurements
of dew point and temperature data for
runs where such data are available, as
discussed by Carter et al. (1986).  If
such data are not available, a default
water concentration of 2 x 10  ppm was
used.
Plots of measured temperatures against
time were approximated by straight lines,
which were used to derive the temperature
as a function of time in the model
simulations.

Dilution rates of 2 x 10   min   were
assumed, as discussed by Carter et al.
(1986).  No "dynamic dilution"
experiments were modeled in  this study.
 Initial nitrous acid  is  assumed  to  be
 negligible  in  Teflon  film chambers
 (Carter et  al.,  1986).
 Included  to  represent  the  dark formation
 of  HONO from NOp  observed  in  Teflon
 bag chambers.   The  pseudo-first-order  rate
 constant  used for the  UNC  chamber  is 5.0  x
 10"-* min   .   This rate constant, and the
 assumed HONO/NOo  stoichiometry,  is based
 on  studies of this  process in a smaller,
 4300-liter Teflon film chamber (Pitts  et
 al., 1984),  corrected  for  differences  in
 chamber size (Carter et al.,  1986).
 This is the only process used to
 represent the continuous light-dependent
 component of the radical source in Teflon
 film chambers (Carter et al., 1986).   The

                               (continued)
                                     125

-------
Table 17 (continued) - 2
 Parameter(s)
                                Value  and  Derivation
                                rate used is given by kRo x k,,  where
                                   is the
                                              photolysis  rate,  and
                                kRS was adJusted based on fits of model
                                simulations to alkane-NOx-air experiments.
                                The values used for this parameter are
                                are given in Table 18.
NOH Offgassing

 wall + hv ->  N02
                                Rate given by kEN x k-j,  the N02
                                photolysis rate,  where k™ was adjusted
                                based on fits of  model simulations to
                                acetaldehyde-air  experiments.   The
                                values used for this parameter are given
                                in Table 18.
Background Organics

     wall
 OH   ->  H00
                                This "reaction" is used to simulate the
                                effect of trace contamination by reactive
                                organics which convert NO to N02.  The
                                rate constant for this process, ^XSHC'
                                was adjusted based primarily on
                                fits of model simulations of pure air
                                irradiations.  The values used for this
                                parameter are given in Table 18.
Other Wall Reactions
     wall
      -> loss of
    wall,  H20
        ->   Loss of
                                First order loss rate of 1.M3 x 10
                                min   is used, based on results of two
                                ozone dark decay experiments (Carter et
                                al., 1986).
                                Rate constant given by 9 x 10"  +
                                2 x 10~tt [H20] min"1, based on
                                measurements of N20e loss rates in a
                                smaller Teflon film reactor by Tuazon et
                                al. (1983), adjusted for differences in
                                chamber sizes as discussed by Carter
                                et al. (1986).
                                    126

-------
Table 18.  Values of Adjustable Chamber-Dependent Parameters Employed
           in the Simulations of UNC Chamber Runs
Characterization Model

Original model
(Carter et al., 1986)
Re-adjusted. No
temperature dependence
Re-adjusted, with
temperature dependence



Temperature
(K)
All

All

T <
290 < T <
T =
300 < T <
T >





290
300
300
310
310
kRS
(ppb)
0.30-

0.30

0.05
(a)
0.10
(b)
0.60
kEN
(ppb)
0.30

0.11

0.025
(a)
0.050
(b)
0.20
kXSHC
(min"1)
500.

250.

750.
750.
750.
750.
750.
(a)  Obtained by  linear  interpolation between values given  for T  = 290 K
     and T  = 300  K.

(b)  Obtained by  linear  interpolation between values given  for T  = 300 K
     and T  = 310  K.
                                     127

-------
which reflects the magnitude of the continuous radical source and which  is
adjusted so model simulations fit observed rates of NO oxidation and ozone
formation in  alkane-NOx-air  runs;  kEN, which  reflects  the NOX offgassing
rates and which is adjusted so simulations fit observed rates of ozone and
PAN  formation in  acetaldehyde-air  runs,  and kx<>Hc»  which  reflects  the
extent  of  excess NO-to-N02  conversions caused  by reactive  organic con-
taminants and  which  is determined based on  fits  to ozone formation rates
observed in pure air photolyses.  Since simulations of these characteriza-
tion runs are  affected at  least  to  some extent by the full set of  charac-
terization  parameters,   this  optimization  has to  be carried  out  in   an
iterative manner.
     This  use of  model  simulations  to determine  rates  of some  of the
chamber-dependent processes means  that the values derived will depend,  at
least  to some  extent,  on  the  light  characterization  model used  in the
model simulations.   Therefore,  since the light characterization model for
the UNC  chamber  was  modified during this  study,  we also had to re-derive
the adjustable chamber-dependent parameters  used  for  modeling data  in this
chamber.
     Figure  5  shows distribution  plots  of  discrepancies  between model
predictions and experimental results  for the chamber  characterization runs
used  to derive  the  adjustable  chamber-dependent parameters  for   the UNC
chamber.   These include  distributions of fits to (1) maximum ozone yields
in pure  air  irradiations,  (2)  changes  in NO and N02 concentrations  in NOX-
air  and  CO-NOx-air  experiments,   (3)  maximum  ozone and  PAN  yields  in
acetaldehyde-air  experiments,  and  (4)  NO oxidation  and ozone  formation
rates  observed in alkane-NOx-air  runs.   Also shown are fits to  maximum
ozone yields  and  NO  oxidation  and  ozone formation rates  in propene-NOx-air
runs;  although modeling these runs was not used to  derive  the  adjustable
chamber-dependent  parameters,  the  results of modeling  these runs  give a
good  indication  of  the overall performance and variability of the  model  in
simulating  a  relatively  large number  of  similar experiments carried  out
over  a  number of years.
      Distributions of fits for five sets of model simulations are shown in
Figure  5.   The first four sets were all carried out using the latest SAPRC
mechanism   (Carter  et  al.,  1988),  but  with  varying  light  and  chamber
characterization  models.    The  first  set  (designated "original") employed
                                     128

-------
  Set   Mechanism
                  Light Model
                                 Other Chamber Effects
1
2
3
14
5
SAPRC
SAPRC
SAPRC
SAPRC
RADM
Carter et al. (1986)
New UNC Model
New
New
New

Experimental
Set 1
UNC Model
UNC Model
UNC Model

Set 2
Carter et al. (1986)
Carter et al. (1986)
This
This
This
work,
work,
work,
Air Ri in* •
Set 3
T- independent kRS
T-dependent k^,
T-dependent k^,
.' k£N
''EN
kEN



Set H
Set
5
 -0
  33 -
•0.27 -
•0.21 -
•0.15 -
-0.09 -
-0.03 -
 0.03 -
 0.09 -
 0.15 -
 0.21 -
 0.27 -
-0.33
-0.27
-0.21
-0.15
-0.09
-0.03
 0.03
 0.09
 0.15
 0.21
 0.27
 0.33
 0.33
                   111
1
11
111
11 i
Figure 5.   Comparison  of Results  of UNC Characterization Simulations Using
            Various Mechanisms  and Chamber and Light Characterization
            Models.
                                                                    (continued)
                                        129

-------
Set   Mechanism     Light Model
Other Chamber Effects
1 SAPRC
2 SAPRC
3 SAPRC
4 SAPRC
5 RADM
Calculated-
Experimental
< -0.33
-0.33 - -0.27
-0.27 - -0.21
-0.21 - -0.15
-0.15 - -0.09
-0.09 - -0.03
-0.03 - 0.03
0.03 - 0.09
0.09 - 0.15
0.15 - 0.21
0.21 - 0.27
0.27 - 0.33
> 0.33
. .
Experimental)
/Experimental
< -1.00
-1.00 - -0.82
-0.82 - -0.64
-0.64 - -0.45
-0.45 - -0.27
-0.27 - -0.09
-0.09 - 0.09
0.09 - 0.27
0.27 - 0.45
0.45 - 0.64
0.64 - 0.82
0.82 - 1.00
> 1.00
Carter et al. (1986) Carter et
New UNC Model Carter et
New UNC Model This work
New UNC Model This work
New UNC Model This work
- (1) Butane-NOx-Air and (2) Branched
Set 1 Set 2 Set 3
1 1 1

1
1
1 1 11
1
— 22 - 122 - 12
2


1
12 1 1
2 122 122
_.-.-.-. f 1 \ Ru fa no UfW Air anH O\
Average Initial d((
Set 1 Set 2 Set 3
.




111 111 11
-- 122 - 122 - 1122

12 12
1122 12 1
2

•
al. (1986)
al. (1986)
, T-independent k^, kEN
, T-dependent k^, kEN
, T-dependent k,^, kEN
Alkane-NOx-Air: Maximum Ozone -
Set 4 Set 5
.
:1 1
;
: 2
: 12
:112 1
-:11 - 112
:2
:1 12
:2
:2
:
•

1031 - [N0]))/dt
Set 4 Set 5




2
1112 1112
- 1112 - 11122
22





                            Figure  5  (continued) - 2
                                        130

-------
Set   Mechanism
Light Model
Other  Chamber Effects
1 SAPRC
2 SAPRC
3 SAPRC
1 SAPRC
5 RADM

Experimental
< -0.33
-0.33 - -0.27
-0.27 - -0.21
-0.21 - -0.15
-0.15 - -0.09
-0.09 - -0.03
-0.03 - 0.03
0.03 - 0.09
0.09 - 0.15
0.15 - 0.21
0.21 - 0.27
0.27 - 0.33
> 0.33
( Pa 1 r*i 1 1 af Art —
Experimental)
/Experimental
< -0.50
-0.50 - -0.11
-0.11 - -0.32
-0.32 - -0.23
-0.23 - -0.11
-0.1*4 - -0.05
-0.05 - 0.05
0.05 - 0.11
0.11 - 0.23
0.23 - 0.32
0.32 - 0.11
0.11 - 0.50
> 0.50
Carter et al. (1986) Carter et
New UNC Model Carter et
New UNC Model This work
New UNC Model This work
New UNC Model This work

Set 1 Set 2 Set 3
111
— 1 - - 1
111 11
11 1111
11

Set 1 Set 2 Set 3 '
11
1
1
1 1
1
1 1
1111 11111
al. (1986)
al. (1986)
, T- independent k^, k^
, T-dependent kR£, kg^
, T-dependent k^, kEN
.
Set 1 Set 5
11
1 11
- 11111 - 11
MS- ₯ay
-------
Set   Mechanism
Light  Model
Other Chamber Effects
1 SAPRC
2 SAPRC
3 SAPRC
U SAPRC
5 RADM
Calculated-
Experimental
< -0.20
-0.20 - -0.16
-0.16 - -0.13
-0.13 - -0.09
-0.09 - -0.05
-0.05 - -0.02
-0.02 - 0.02
0.02 - 0.05
0.05 - 0.09
0.09 - 0.13
0.13 - 0.16
0.16 - 0.20
> 0.20
Calculated-
Experimental
< -0.20
-0.20 - -0.16
-0.16 - -0.13
-0.13 - -0.09
-0.09 - -0.05
-0.05 - -0.02
-0.02 - 0.02
0.02 - 0.05
0.05 - 0.09
0.09 - 0.13
0.13 - 0.16
0.16 - 0.20
> 0.20
Carter et al. (1986)
New UNC Model
New UNC Model
New UNC Model
New UNC Model
----- (1) NOx-Air and (2)

Set 1 Set 2
22 22



122 122
1 12
— 22 - 22
11111112 111111112
1112 11




„ 	 	 	 NOx-Air and CO-N

Set 1 Set 2
:22 :22
: :
:
i ;
:
: 1 :111
--:11122 -:1112
:12 -.111222
: 11 12 :112
: 11 1122 :12
: ;
* *
: :
Carter et
Carter et
This work
This work
This work
CO-NOx-Air:
Set 3
-.22
J
;
:2
:12
:122
-:22
al. (1986)
al. (1986)
, T- independent ko<
, T-dependent kj^,
, T-dependent k^,
Change in NO/lni
Set <4
:22
*
J
:2
:2
: 11122
-: 11111122


I' EN
it
EN
I*
kEN
Hal UfW .... .

Set 5
:22
*
;
:2
:2
: 11 122
-: 1111 1122
-.111111111 : 1 1 1 : 1 1 1
:1
:
;
:
•
j
:
:
*
:
Ox-Air: Change in N02/lnitial
Set 3
22


111
112
11122
- 1112
12
2




Set H
22



112
1111122
- 1122
11
12




*
;
j
*
:
UQv 	

Set 5
22



112
1111122
- 1122
11
12




                             Figure 5  (continued)  - 4
                                         132

-------
Set   Mechanism
Light  Model
Other Chamber Effects
1 SAPRC
2 SAPRC
3 SAPRC
14 SAPRC
5 RADM
Calculated-
Experimental
< -0.33
-0.33 - -0.27
-0.27 - -0.21
-0.21 - -0.15
-0.15 - -0.09
-0.09 - -0.03
-0.03 - 0.03
0.03 - 0.09
0.09 - 0.15
0.15 - 0.21
0.21 - 0.27
0.27 - 0.33
> 0.33
(Calculated-
Experimental)
/Experimental
< - 1 . 00
-1.00 - -0.82
-0.82 - -0.611
-0.64 - -0.45
-0.45 - -0.27
-0.27 - -0.09
-0.09 - 0.09
0.09 - 0.27
0.27 - 0.45
0.45 - 0.64
0.64 - 0.82
0.82 - 1.00
> 1.00
Carter et al . (1986) Carter et
New UNC Model Carter et
New UNC Model This work
New UNC Model This work
New UNC Model This work


Set 1 Set 2 Set 3
. f
; ;
: ;
:111 :
:11 :11 111
:11 :1111 11
— :1111111 -:111 - 111111
:1111 :11111 11111
:11 :111 111
:1 :1111 11
:1 : 1
: :1
• •
al. (1986)
al. (1986)
, T-independent kRS, k£N
, T-dependent k^, k^N
, T-dependent kR^» ^£y
Mil Y *! flfll 1 fTi f*i*^ rtn A

Set 4 Set 5



11
11111 111
1111 11111
- 11111 - 111
1 111
1111 11
11 1
1 111





Set 1 Set 2 Set 3





1111 1111 1111
— 11111111 - 11111111111 - 1111111
1111111 111 1111
11 111 111
1 1 1





Set 4 Set 5

;
;


111111 111111
111 - 111111111111- 11 m 1 1 1 1 • i 1 1
11 i
1 11




                          Figure 5 (concluded) -  5
                                       133

-------
the Carter et al. (1986) model for both light characterization and chamber
effects, and  reflected our  best estimates  prior to  this program.   The
second  set employed  the  new UNC  light  characterization  model  but the
original (Carter et al.,  1986)  model for the other  chamber effects.   The
third and  fourth sets  employ the new UNC light  model  and two new sets of
values  for the adjustable chamber-dependent  parameters,  discussed below.
The  fifth  set employed the  same  light and characterization  input as the
fourth set, but was carried out using the RADM mechanism.
     Figure 5 shows that none of  these chamber models yields  close fits to
experimental  results for all of the characterization experiments modeled;
there were always runs  where significant discrepancies are  obtained.  This
can  be  attributed  to  run-to-run  variation  of chamber  effects  due  to
unknown or uncontrolled factors.  Improved  fits  could obviously  be obtain-
ed by  carrying out run-to-run optimization of the  adjustable parameters,
but  such a procedure  would not be  consistent  with  our objectives.   Until
all  significant chamber effects are  identified and quantified so that they
can  be  predictively modeled, the best that can  be  hoped for is that the
biases  and errors  in  the model simulations of these characterization runs
is as small as possible.
     A  comparison of the first  two  sets of  distributions  in Figure  5  shows
that updating the light characterization model without  changing  the  repre-
sentation  of  the other  chamber  effects tends  to  cause  a slight  increase  in
the  tendency  of the model to  overpredict  maximum ozone yields.   Although
this effect does not appear  to  be large,  it has  a non-negligible effect  on
the  model  performance  statistics when simulating ozone yields  in the full
set  of UNC  chamber experiments.   The  updated  light model  with the old
values  of the adjustable  parameters  also  results in  significant overpre-
dictions  of  PAN yields  in acetaldehyde-air irradiations,  indicating that
the  assumed  value  for  the NOX offgassing  rate  in  the UNC chamber may  be
too  high  if the new light  model is  correct.  The use of an apparently high
NOX  offgassing rate in the chamber  model  could contribute to a  tendency  of
the  model  to  overpredict maximum ozone yields.  On  the other hand,  use  of
the  new light  characterization  model does not  necessarily  indicate that
the  assumed  magnitudes of  the  chamber  radical  source  need to be  modi-
fied.   The fits of model  simulations  to NO oxidation  rates in  alkane-NO -
                                                                         A
air  experiments are sensitive to the representation of the chamber radical

-------
source, but  use  of the new  light  model  without changing other  parameters
was  found  not to  significantly degrade  the  performance of  the model  in
simulating these results.
     Better  fits  of  model  simulation  to  results  of acetaldehyde-air
experiments were obtained  if the magnitude  of the  NOX offgassing parameter
kg-jj  was  reduced  by approximately  a factor of  three.   This adjustment  of
kEN  required  reducing  by a factor  of two the  value of kXSHC,  the parameter
measuring  the amount  of excess NO-to-NOp conversion caused by  background
or contaminant reactive  organics,  so the model would still  fit  the  maximum
ozone  yields observed  in  the  pure air  irradiations.   The  third set  of
distributions  shows   the  performance  of  the  model   after  these  minor
modifications  are  made.   It can  be seen that  these changes result  is  a
significant  improvement in  terms  of the overall  biases in fits to  ozone
and  PAN yields  in  the  acetaldehyde-air runs and also a  slight reduction  of
the  tendency of the model to overpredict ozone  yields  in the  propene-NOx-
air  experiments.
     The  original  chamber characterization model  of Carter et  al.  (1986)
neglected  any  effects of  temperature  on  the  magnitudes  of the  chamber
radical  source or  the NOX  offgassing  rate parameters, and  these  effects
are  also  ignored  in  the re-adjusted model  discussed above.  However,  in
the  process of re-assessing  this  chamber characterization  model, we noted
that there was  a  consistent tendency for  the model to overpredict  reac-
tivity in  runs  with   relatively low average  temperatures  (below -290-300
K).   This suggests that the  chamoer  radical  source in  the  UNC chamber may
be temperature dependent.    In addition,  the modeling of the lower tempera-
ture acetaldehyde-air  runs  suggests a  temperature dependence  in  the NOX
offgassing  rates  in   that  chamber.   Therefore,  we investigated  whether
using  a modified  chamber  effects  model with  temperature  dependences for
the parameters k™ and  k^  might  result in reduced discrepancies in model
simulations of the UNC chamber  experiments.
     The number of UNC  characterization runs carried  out  at  a  variety  of
 temperatures  is  not  sufficient   to   unambiguously   derive   temperature
dependencies for the  radical source or  for NOX offgassing.  Therefore,  we
 tried  a  variety of  assumed temperature  dependencies   to  determine which
would  minimize  the overall  average error in  the  distribution  of the fits
 to  the  runs used  to  characterize  these  effects.  The model  simulations
                                     135

-------
suggested that runs with average temperatures  below  -300  had  significantly
lower radical  sources and  NOX  offgassing  rates  than runs  with the more
typical average  temperatures  of -300-305, but that  runs  with still  higher
average temperatures did not  appear  to have correspondingly  higher  radical
sources.   This  suggests  that  these chamber  effects increase  relatively
rapidly with  temperature at temperatures below around 300 K, but  increase
slower at higher  temperatures.  To represent this, we assumed that  k^  and
knij  increased  linearly with  temperature  both above  and  below  300  K,  but
that  the  rate of change is  less  in  the higher temperature  range.   The
dependences which we  found  gave the  best fits of those we tried are given
in Table  18.   To obtain satisfactory  fits  to maximum ozone  yields  in  the
pure  air  irradiations,  it  was  also necessary  to  use a larger value  of
kXSHC'   The *ast set  of< Pl°ts in figure 5 show the  distribution  of fits
obtained  for  the UNC characterization runs  using this set of temperature-
dependent chamber characterization parameters.
     Figure 5  shows  that using  the temperature-dependent  model  for k™  and
k™  yields  somewhat  better fits to  ozone yields observed in the acetalde-
hyde-air  experiments,  and  significantly  better fits to NO oxidation rates
and  ozone yields  in  alkane-NOx-air  runs,  compared  to using the  adjusted
model  where temperature dependences of  these parameters are ignored.   It
also gives better distributions of  fits to NO oxidation rates and ozone
yields  in  the  simulations of  the  propene runs.    However, it should  be
noted  that  the number of runs useful for deriving temperature dependencies
for  the radical  source and the NOV  offgassing  rates in the  UNC chamber is
                                  A
extremely limited,  and thus  the temperature dependencies derived  for them
in  this  study  must  be  considered  to be highly  uncertain.    In addition,
there is  uncertainty  in  the  temperature   dependencies  of   the  reactions
involved  in  PAN  formation  (Gery   et  al.,  1988),  and  use of  different
assumptions in the mechanism in this regard, such as  those employed  in the
latest  Carbon Bond mechanism (Gery et al.,  1988) may  result in differences
in   the  apparent temperature dependencies  of  k™  being  derived  by model
simulations of the acetaldehyde-air  experiments. . (On the other hand, this
uncertainty in the temperature dependence  of PAN formation reactions will
have less  of an impact on the values of  kRg which  are derived  from the
model  simulations  of  the  alkane-NO-air experiments.)    Nevertheless,
because  of  this significantly  better  performance  in  the   temperature-
                                     136

-------
dependent  chamber  effects  model in our simulations of these data,  this  is
the model  which  was  used in this study in simulating the UNC chamber  runs
for the evaluation of the RADM mechanism.
     It should  be emphasized that  this  derivation of adjustable  chamber-
dependent  parameters by  model simulation  means  that  the  parameters derived
will  depend on  the  chemical mechanism  used for  this  purpose.   In  this
study  these parameters  were derived primarily based on  calculations using
the latest SAPRC  mechanism  (Carter et al.,  1988).    This  is because  the
SAPRC mechanism  was  used in  our previous  chamber characterization  modeling
work,   and  because the SAPRC mechanism has a more explicit  representation
of  the  reactions of the alkanes than does  the  RADM mechanism.    In addi-
tion,   since the RADM mechanism is being evaluated  in  this study, it  was
not considered  to be in its final  form  at the  time these  chamber  charac-
terization  calculations were  being  carried  out.    Ideally,  the  chamber
characterization runs used  to derive  the  adjustable  parameters should  only
involve the "known"  reactions,  so differences in  current mechanisms would
not affect modeling  of such runs.  However,  this  is not yet entirely  the
case.
     The differences  between the  RADM and  the  current SAPRC mechanism  in
modeling  these  UNC  runs used  for chamber  characterization is  shown  in
Figure  5,  where the distributions of  the fits  calculated using these  two
mechanisms  with  the  new UNC light and chamber characterization models  can
be  compared.   It can be seen that, as expected,   there  are  essentially  no
differences  between  the  mechanisms  in the  simulations of  the  pure  air,
N0x-air,  and CO-NOx-air runs,   but  there are  slight differences in  the
simulations  of  the propene-NOx  runs,  and relatively larger  differences  in
the simulations  of the alkane-NOx-air and the acetaldehyde-air runs.   The
differences  in  the  simulations  of  the  alkane-NOx-air  runs  is  expected
since  the  RADM  mechanism uses  a more  lumped representation  of the  alkanes
than the SAPRC mechanism, and strictly speaking is not designed  to be  used
for modeling  such  runs.    The  differences in   the  simulations  of   the
acetaldehyde-air runs are due to differences between the mechanisms in  the
rate constants  used  for the decomposition and formation of  PAN, which  are
such  that  PAN   is  predicted by  the  RADM  mechanism  to  be more stable
relative to N02  + acetyl peroxy radicals.  This results in  the RADM mech-
anism predicting  lower ozone yields and higher  PAN yields in acetaldehyde-
                                    137

-------
air  irradiations  than  predicted  by  the SAPRC  mechanism.    With  the  kRS
parameters  adjusted  to give optimum  fits  using the SAPRC mechanism,  this
results  in  the  RADM  mechanism underpredicting  ozone  and  overpredicting
PAN, relative  to  the experimental results.  However, either  increasing or
decreasing  kRS  from  the  optimum  for  the SAPRC  mechanism  would  cause
greater discrepancies  for  either  PAN  or  ozone,  respectively,  with  the RADM
model.   Because of  this,  it is considered  to  be unlikely that using  the
RADM mechanism  to derive the adjustable  chamber-dependent parameters would
have yielded significantly different  values  than  those used in this  study.

M.3  Representation  of Organics in Chamber Experiments
     Many  of  the  environmental  chamber experiments  used  in  this  study
include  runs containing organic  compounds,  either singly or  in mixtures,
which  are not  represented  explicitly in the RADM mechanism.  The  appro-
priate procedures  for representing such compounds  when  simulating chamber
experiments  depends  on  the  intended application  of  the mechanism being
evaluated.   If the mechanism is intended to be  a fully  detailed mechanism
to  test  our ability  to understand  and predictively  model the  chemical
system being simulated,  or if the mechanism is  to be used as a standard or
"master   mechanism"   against  which   more   condensed  mechanisms   will  be
compared,  or if the mechanism  is  to  be  used to  assess  relative  reactivi-
ties  of  individual  organic compounds,  then  the  approach  should  be  to
adjust  the mechanism so that it  represents  the  chemical conditions of the
mixture  being simulated as  closely   as  possible.   For  example, the SAPRC
detailed  mechanism  is  designed to serve  both  as  a basis  for  developing
more  condensed mechanisms  and for use  in  reactivity  assessment  studies,
and  thus  its  evaluation  involved  adjusting the kinetic and  mechanistic
parameters for lumped model species  based on the compounds  in the experi-
ments  being simulated (Carter, 1988).   If,  on  the other hand,  the mechan-
ism  being evaluated is a condensed mechanism which is  intended  to be used
as  a  component  of a  larger  model  such  as  RADM,  then the  appropriate
procedure is to represent the  compounds in  the-chamber experiments in the
same  way  they  would be represented  when  the mechanism  is  implemented in
the  larger model.  Otherwise, the mechanism being evaluated in the  chamber
simulations is not  the same  as  the mechanism  which  will be used  in its
intended application.   Since the mechanism being  evaluated  in  this study
                                     138

-------
is  intended  to  be  a condensed  mechanism for use  as a  component  of  the
RADM-II model,  and  since as far as we  are aware it  is not anticipated  it
will be  used as a  basis for developing  more  condensed  mechanisms or  for
assessing relative  reactivities  of different organics, it  is obviously  the
latter approach that is appropriate for this study.
     Therefore,  the approach we used when simulating chamber experiments
to evaluate  the RADM mechanism  was  to  process  the  organic input data  in
the  chamber  simulations  in  the same way that we  anticipate  the organic
emissions data  will be  processed  for  input into  the RADM-II  model.   The
RADM emissions  input processing  procedure we anticipate  will  be used  is
the  two-step  procedure  described  in   Section  3  of  this report.   This
involves no  adjustment  of rate constants or mechanistic parameters  in  the
mechanism  — these  are  the  same  regardless  of  the  mixture  being  simu-
lated.  On  the  other hand,  in some cases model species concentrations  are
adjusted to  account for  differences in  reactivities between them and  the
input  species  they represent.   The  application of  this  procedure to  the
simulations of the  chamber runs is briefly summarized below.
     The first  step in  processing the  organic input data  involves  aggre-
gating, on  a mole-per-mole  basis,  the individual  chemicals  into the  32-
class  emissions grouping  discussed  in Section  3.2.1.   Table  7 in that
section  indicates  how  various  types  of  chemicals  are aggregated in this
step.   As  indicated there, some of the  groupings  are determined based  on
the  OH radical  rate  constants  assumed  for  the  individual  organic com-
pounds, and  the specific sets of OH  radical rate constants used  in  making
these  assignments  in  this study  are those listed by  Carter  (1988a).   Other
than  the  UNC  auto  exhaust  experiments,   no  experiments  modeled in this
study  contained unknown  or  unidentified  compounds.   The  initial organic
concentrations  used in  the UNC auto  exhaust experiments in our  1986  mech-
anism  evaluation study  (Carter  et al.,   1986) were  also  employed in this
study.   [The alkanes contained in these experiments are given in terms  of
the  two lumped  alkane classes,  C^-C^ and  C^.   The lumped  class  C1|-C5  was
aggregated into  emissions class  "Alkanes  (0.25-0.5  react)," and  the  lumped
class  C6+  was  aggregated into "Alkanes  (0.5-1  react)",  based  on  the  OH
radical rate constants used  for  them  in  the SAPRC/ERT mechanism (Lurmann
et  al.,  1987).]    Therefore,  there  are  no  cases  where  chemicals  in  the
chamber  experiments  were  aggregated  to the  "unknown"  or   "unassigned"
categories.
                                    139

-------
     The organics  aggregated  into the emissions  classes  were then aggre-
gated into the species in the RADM mechanism as discussed in Section 3.2.3
and as  indicated  in  Table  8 in that  section.   Some emissions classes are
aggregated on  a  mole-per-mole  basis, while  others are  aggregated using
"reactivity weighing,"  based on  estimates of  how much  of  the emissions
class and  the model  species  will react  in the  model  simulation.   Note,
however,  that  the aggregation  of classes  based  on "reactivity weighing"
require use  of the  parameter  "INTOH" in  the derivation of  the weighing
factors.   Since  INTOH reflects  the effective integrated  radical levels  in
the simulation,  it is a model  input variable  and not  a  parameter in the
mechanism.   As such,  the  weighing  factors shown  in Table  8,  calculated
using   the  INTOH value  of  110 ppt-min,   which was   estimated  to   be
appropriate  for  use  in  three-day  RADM  simulations  (Stockwell,   private
communication,  1988),  are  not  necessarily  appropriate  for  use  in the
chamber simulations.   Therefore,  separate  estimates for  INTOH  values were
made for each chamber.
     The  INTOH  values used for calculating reactivity weighing  factors  in
the simulations  of the chamber  experiments carried out in this  study were
as follows:

                   Chamber             INTOH  (ppt-min)
                   SAPRC EC                    35
                   SAPRC ITC                   40
                   SAPRC OTC                   MO
                   UNC                         70

These  were derived  based on  a  rough  analysis  of average  integrated  OH
radical levels  in model simulations  of one-day chamber  experiments employ-
ing  various  complex  mixtures of organics.   (The model  simulations  employed
either  the SAPRC  detailed mechanism,  or  the RADM  mechanism  with  the rate
constants of  the lumped species adjusted  based on the  compounds present in
the  experiments — the results were similar in either case.)   The higher
INTOH  value  for  the  UNC  chamber  can be attributed to  the  longer  irradia-
tion  time used  in that  chamber, where runs  start at  sunrise  and end  at
sunset, and  are  usually  carried  out in  the  summer.   The SAPRC  runs  are
usually limited to six hours.
                                     140

-------
     Strictly  speaking,  the  integrated  OH  radical  levels, and  thus  the
appropriate  INTOH  values,  will depend  on  the  type of  experiment, with
lower  values  for runs with  high levels of  radical  inhibitors,  etc.,  but
this type  of  run-by-run  analysis was not carried out.  The simulations of
only a relatively  small  number of  runs are affected  to any significant
extent  by  reactivity weighing,  since, of  the experiments modeled  in this
study,  reactivity  weighing  is  of   importance  only   for  those containing
significant amounts of methanol or  benzene.  For the  other compounds in
these  experiments,  the  weighing factors are either unity or very  close to
unity.  For this reason,  we  did not  consider a  comprehensive  investigation
of  the most  appropriate INTOH  value  to  use  in  the chamber simulations to
be worthwhile.
     To provide an  indication  of  the level  of  inaccuracy  introduced  into
the  model  simulations  by using  this  aggregation   approach,  a  separate
series of  simulations of the chamber experiments was carried  out  where  the
rate constants for  the  lumped model  species  were adjusted for  each experi-
ment  based  on  the  compounds  present.    The  rate   constants  which  were
subject to adjustment when this procedure was employed were  those for  the
reactions  of  OH radicals  with  HC3,  HC5, HC8, TOL, XYL,  OLI, and  OLT,  and
for  the reactions  of ozone  and NO?  radicals with OLI  ana OLT.   The  rate
constants  for  the  individual organics  used  when  this  procedure  was
employed  are  those  assigned  to the  detailed model  species in the latest
SAPRC  mechanism, as  given by Carter  (1988).   In most cases,  the results
were  not  significantly  different  than when  the  standard RADM  aggregation
approach  was  employed,  tending to   indicate  that  the aggregation  approach
proposed  for  RADM is probably satisfactory.   The results of these compari-
sons are  given in more  detail in Section 5.6.

-------
        5.   EVALUATION AGAINST ENVIRONMENTAL  CHAMBER DATA:   RESULTS

     The results  of the simulations of  the  chamber experiments using the
March  1988  and the  recommended  modified versions  of  the  RADM mechanisms
are  described  in  this  section.    The  organization  of  this evaluation
against  the  chamber data is  discussed  first,  then  the  measures  of model
performance are described,  and finally  the results of the evaluations are
summarized and  discussed.   The results  of the simulations using the March
1988 version  of the RADM mechanism (designated simply as the  "RADM" mech-
anism) are  described first  and in the most  detail.  Then the  differences
in the  results  of the simulations using the recommended modified (RADM-M)
and the  recommended modified parameter   (RADM-P) versions of the mechanism
are discussed.
     Because  of the  large  number (over  550)  of chamber  runs which  were
used  in  this evaluation, the discussion of the  results of these  simula-
tions  focuses primarily on  statistical  measures  of model performance  for
the various  groups of experiments.  However,  concentration-time plots  for
selected  experiments  are  also  shown  for  illustrative  purposes.    For  a
listing  of  all  the individual experiments used in this evaluation,  and for
summaries of  the  performances of the mechanisms on  a run-by-run basis,  the
reader  is  referred to Appendices  A  through  C  of this report.  Appendix  A
lists  all  of  the chamber  experiments  which were  modeled  in  this  study,
together with their initial NOX  and organic concentrations,  their  average
temperatures  and  water concentrations,  and  their maximum ozone yields and
average  initial rates of NO oxidation and ozone  formation.   (An "*"  after
the  temperature  or  water  concentration  indicates that  it  was  was  not
measured or  that  the data was unreliable, and  the default  for that  chamber
was  used  in  the  model calculations.)   Appendix  A  also  gives  for  each
experiment  the performance of the March  1988  RADM  mechanism  in simulating
the maximum ozone yields and  the average rates of NO oxidation and  ozone
formation.    Appendices  B and C  give  similar  listings  of the run-by-run
performance  of the  RADM-M  and RADM-P mechanisms, respectively, for all of
the  experiments whose results  differed  from those obtained with  the  RADM
mechanism.

-------
5.1  Organization of the Evaluation
     The mechanism  testing  was  organized in accordance with the hierarchy
of  species shown  in Figure  6.   This  schematic shows  the hierarchical
nature and  interdependence  of  the major components  of atmospheric photo-
chemical mechanisms.   This hierarchy was  recommended  by the EPA Workshop
on Evaluation and Documentation of Chemical Mechanisms  (EPA,  1987) and  has
been  used  successfully  in  past  evaluations.    The   mechanism  testing
followed  this  hierarchy,   beginning  with  the lowest  levels  (i.e.,   the
inorganic  species)  and  proceeding to the  highest level by stepwise addi-
tion of  species  with increasingly  complex  chemistry.  The  stepwise feature
of  this  approach is designed  to facilitate identification  of  the uncertain
components  of   chemical   mechanisms,   and  minimize   the  occurrence   of
compensating  errors  in  the  mechanisms and  fortuitous agreement between
models  and  chamber  data.    In  concept,   IT  a  substantial  disagreement
between  a  model and data occurs at  a  high level in the hierarchy and  the
model has  shown to  perform well for  experiments  involving  species  lower in
the  hierarchy,  then it  is most likely  that the new  higher  level  chemistry
is  in error and  not the whole mechanism.
     The  chamber data base was subdivided into six  major  groups  for  test-
ing  purposes.   The  first  group included -20 irradiations of pure  air,  NOX-
air.  and  carbon monoxide-MOv-air  mixtures.    The  purpose of  simulating
                             A
these  experiments  was  to  test the  accuracy and   precision  of both  the
inorganic  reactions in  the mechanism and the assumptions used to represent
chamber-dependent  parameters.   Since the  level of organics is  very  low in
these  experiments,  the rate of NOV oxidation  is  very  sensitive to chamber
                                  A
characteristics, such  as  the intensity  and  spectral  distribution  of
radiation,  the  rates of  radical  offgassing  from  chamber walls, and  the
levels  of background organic  contaminants.
     The  second group  of  runs  consisted   of  formaldehyde  -  air  and
acetaldehyde  -  air  mixtures.    These  runs   were  simulated primarily to
further test  the  submodels  for  the   treatment  of chamber  effects.   As
discussed earlier,   the acetaldehyde-air runs are particularly  useful  for
testing how well NOV offgassing  from chamber walls is represented  in the
                     A
model.

-------
Ve'.-'C'e
 £ inane
            >C2 Ai^c^es !  ' >C2 A^er-es
                        n-gr.e-
          Ccrbor
       I NO, N02, N03, N205
       !     HN05. HCNC
       i    OH, H02,  -202
Figure 6.  Hierarchy of Species for  Mechanism Testing.
                       144

-------
     The  third group  consisted of  -20 single  carbonyl -  NO  mixtures.
These  included  runs  with  formaldehyde,   acetaldehyde,  propionaldehyde,
acetone,  or  methyl ethyl  ketone.   These were simulated  next to test the
carbonyl  chemistry, since  the  chemistry  of  the remaining  compounds depends
on these  species.
     The  fourth groups  consisted  of the approximately  200 single  hydro-
carbon  -  NOX  experiments.   The  91  alkene  data  sets  included  runs for
ethene,  propene,  1-butene, trans-2-butene,  isobutene, 1-hexene,  isoprene,
and  alpha-pinene.   The  45 single  alkane-NOx data  base  included runs for
ethane,  n-butane,  three branched alkanes (2,3-dimethylbutane,  isopentane,
and  iso-octane),  six long  chain alkanes (n-pentane, n-hexane,  n-heptane,
n-octane,  n-nonane,  and n-pentadecane), and cyclohexane.   The  aromatic
experiments   included   runs  for   benzene,   toluene,   m-xylene,   o-xylene,
mesitylene,  tetralin,  naphthalene, and  dimethyl  naphthalene.  These  were
simulated to  test  the  performance of  the  subgroups of  hydrocarbon reac-
tions  incorporated into  the mechanisms.
     The fifth group consisted of  -40 runs  for  simple mixtures of organics
and  NOV.   By "simple mixtures" we mean mixtures  of two or  more  organic
       A
species either  from  a  given  class  (alkenes,  alkanes,  and  aromatics)  or
from two classes,  but not from all three classes.   The simple mixture runs
were simulated  to  test  for  compensating   errors  that  might occur  when
compounds  from  different  hydrocarbon  classes  in  the  mechanism  were
present.
      The sixth group consisted of over 200 runs involving complex mixtures
of organic  species and NOV,  where one or more hydrocarbon  species  from
                            A
each major  class  was included in  the  mixtures.   These  runs contained  at
least one alkene,  alkane,  and aromatic —  i.e., at  least  one representa-
tive from each of  the  major  classes  of   organics.   The majority  of the
mixtures included 3 to  8 organic species.   However, some mixtures,  such as
 those  employing  actual  automobile exhaust,  included a  large  number  of
compounds.  The complex mixtures  used  in  these experiments are surrogates
 for typical atmospheric mixtures  of  organics.   (For this reason, they are
 also designated as "surrogate" runs.)   The  purpose  of  simulating a large
 number of runs of  this  type was to assess  model performance  as robustly as
 possible  for  the  type  of mixtures expected in atmospheric applications of
 the mechanism.
                                     145

-------
     As indicated above,  the  mechanisms  which were evaluated included the
March  1988  version  of the RADM mechanism,  the recommended modified mech-
anism  (RADM-M) and  the  recommended  modified parameter mechanism (RADM-P).
The evaluation of the March  1988 RADM mechanism  is  discussed  in the most
detail since  evaluation  of this mechanism  was the  primary  focus  of this
study.   The  recommended modifications  involve no  changes in  the lower
portions of the  hierarchy where  only the inorganic reactions or the reac-
tions  of  the  carbonyls  (other than  the carbonyls  formed  only  from the
aromatics)  are  involved; therefore,  the results  discussed  below  for the
RADM mechanism  concerning the characterization and  the carbonyl runs are
applicable  to  the modified versions of the  mechanism.  The  results of the
evaluation  for  the  runs  containing  compounds  higher in the hierarchy are
discussed separately for the three versions of the mechanism.

5.2  Measures of Model Performance
     The  rates of decay  of  precursor species and the  rates of formation
and the yields of key product species are important  for assessing chemical
mechanism performance.   The  selection of appropriate performance measures
depends on  the type of  experiment.   The most important distinction among
experiments  in this  data base is  whether  or not significant amounts  of
ozone  were  observed.   In simulations of the N0x-air and  CO-NOx-air runs,
which  generally  do  not  generate  much ozone,  the  total changes in NO and
N02  concentrations  over  the  course of the experiments  are the  most useful
measures  of performance.  These  provide a  coarse measure of  the  rate  of
precursor  (NO) oxidation and product (N02)  formation  in  these generally
"slow" experiments.
      In  the organic-NOx-air   runs,  where significant ozone  is  formed,  the
relative  and absolute  errors in the  maximum ozone concentration and  the
average  rate of  change  of  [0^]  minus  [NO] during  the first half of  the
time  period  to  reach maximum experimental ozone are  the  most  important
measures  of   performance.    The  latter  parameter  (designated  "dQOgl-
[N0])/dt"  in the tabulations, and referred  to variously as the "NO oxida-
tion  and  ozone formation  rate" or  simply  as the  "NO  oxidation  rate")
provides  a  relatively direct indication of the rate of NO oxidation by the
peroxy radicals (H02,  R02»  and  RCOq) that  are produced  by the organics.
This is  illustrated  using  the  kinetic rate  equations for  the important

-------
reactions of NO and ozone, as shown below:
      N02 + hv  ---- >   NO +  03
NO
                       N0
                         2  +
      NO + R02  ---- >  N02 + H02 + Carbonyls
                                                                       (1)

                                                                       (2)

                                                                       (3)
where:
    d([03])/dt  = ^  [N02] - k2 [N0][03]

    d([NO])/dt  = k1  [N02] - k2 [N0][03] - k^ [MO][R02]
By subtraction, one obtains the rate of NO oxidation by peroxy radicals:
d([03]-[NO])/dt =
                          [NO][R0],
We  have found  this  to be a  robust  measure of  the  overall timing  of  the
simulations.   When the rate  of  NO  oxidation is slow, the  time  to  the  NO-
N02  cross-over,  the  time to the N02  maximum,  and  the  rates  of  organic
precursor  decay are almost always slow  and  vise-versa.
      This  measure is also a useful measure  of the rate of  ozone formation
in  experiments  containing  no  initial  NOV,   such   as   the  pure  air  or
                                            X
carbonyl-air  runs.  However,  in this case,  it is referred  to  as d[0-a]/dt,
or  the average initial  rate of ozone  formation.
      For experiments  where  formaldehyde and PAN are  formed, the errors for
the maximum predicted concentrations are also  reported.   These species are
important   intermediates  and  products  in  the  photo-oxidation   cycle.
However, when evaluating a condensed mechanism  such  as RADM,  it should be
recognized that  the  model  species called  "PAN"  actually refers to  PAN +
higher  acyl  peroxynitrates,   and thus calculated  "PAN"  yields are  not
always directly comparable with  experimental PAN measurements.
                                     147

-------
     We have adopted an  evaluation  approach were we examine the errors in
both the timing parameter and the maximum concentrations unpaired in time,
rather than  examining  Just the error  in  the predicted maximum concentra-
tions at the time  of the observed maximum  (i.e.,  paired  in time).  While
our approach is less stringent than  the latter,  it measures the aspects of
performance  which  we  believe are  most important,  i.e.,  the  yields and
timing of  the  simulations.   The focus on maxima unpaired  in time is also
often used  in  evaluation of Eulerian photochemical models against ambient
data (Burton, 1988).
     In this  study, the  maximum  ozone statistics  were  only compiled for
those  experiments  where  the  maximum experimental   ozone  concentration
exceeded 40 ppb.   In experiments with  lower ozone yields,  relatively  small
errors  and  interferences  in  instrumental  readings  would  constitute   a
relatively high percentage  of the  total yield and would not be a meaning-
ful measurement of  model  performance.   In addition, performance statistics
on  ozone  yields  for runs  with  less  than  approximately  40  ppb  of  ozone
(approximately the  global background)  are not  meaningful in  the context of
environmental chamber modeling  where maximum ozone concentrations usually
range from  100 to 1000 ppb.
     The statistical  parameters reported for  each  group  and subgroups of
runs are the mean experimental value,  the mean  calculated value, the mean
bias, the  mean  normalized bias, the mean error  (or gross error), and  the
mean  normalized  error.   -The  latter are defined below in accordance with
the  American  Meteorological   Society's  recommendations   for  air quality
model evaluation (Fox, 1981).
                               1
Mean Absolute Bias  (ppb)  =  —   £   (C* - E*
                               N   i=1,N
                               1
Mean Absolute Error  (ppb)  =  —   £     \C= - Ez|
                               N   l=l,N
                              100            (Ci - E^
 Mean Normalized Bias  (1)  =   	   £    [	]
                               N   1=1,N   0.5  (Ci +  Ei)
                                    1118

-------
                             100           \Ci -
Mean Normalized Error (%) =  —   I   [
                              N  i=1,N   0.5 (Ci + Et)

where  C-  =  Calculated  concentration,  Ei  =  Experimental  concentration,  and
the  summations  are over all N  experiments  in  the group or subgroup.   The
standard deviations of  the mean normalized  bias is also  reported.
      In terms of  our  philosophy of evaluating  models, we believe the test-
ing  should cover  a wide range of conditions to "stress" the  mechanism and
to  facilitate  identification  of  situations  where  the  mean bias  and/or
error are  so large that one would have to reject  the mechanism.   A mechan-
ism  can be  Judged acceptable  if  it  doesn't fail any  of the tests.   The
performance  levels at  which one decides to  accept  or  reject a mechanism
are  subjective.   Nevertheless,  it  is  useful to define  such  a criteria so
that mechanism performance  can be judged  consistently.   Our criteria for
rejection  is a mean normalized bias in excess of 20% and a mean  normalized
error in  excess  of  50%.   These  are  referred  to  below at the minimum
desired performance  criteria.    In  addition,  because the  results for any
one  experiment  are  not nearly  as meaningful  as  those  for  a  group of
 experiments, one  needs  to determine the minimum number  of  experiments  that
 must  be simulated to  determine  whether  or not a mechanism  is  performing
 satisfactorily.   We believe a  minimum of 10 to 15 experiments is  needed to
 reasonably assess model performance for a particular type of run.

 5.3   Performance  on Characterization  Runs
       The  performance  of  the  RADM mechanism,  combined  with  our  current
 chamber effects  submodels,  in  simulating  the N0x-air  and CO-NOX  air  runs
 is  summarized in Table 19.    Run-by-run  summaries of these  results are
 given in  Appendix A.   (These results  are also applicable to  the recommend-
 ed  modified versions  of this  mechanism.)   This  table  (and  Appendix  A)
 shows the performance statistics for the  total changes in NO and N02 con-
 centrations during  the runs.   Figures 7 and  8  graphically  compare the
 change  in [NO] and [N02] for each run, respectively.   In the N0x-air  runs,
 the model  slightly underpredicts  the absolute changes  in NO and N02 con-
 centrations,   but in   the  CO-NOX  air  runs,  the  model  overpredicts the
 absolute  changes in these  values.   On the average  for all  of  these  runs,
  the model  overpredicts the changes  in  NO by 9% with  an average error of

-------
    and underpredicts  the  changes  in NC^ by  10/1  with  an average error of
45J.  The graphical comparisons illustrate that the model performs well on
the majority  of these characterization runs.   This  suggests the combina-
tion  of  the   chamber  characterization   procedures  and  the  inorganic
chemistry in  the RADM mechanism provides  moderately accurate predictions
for these types  of experiments.  This  is despite  the fact that the chamber
characterization  model  was  derived  using  the  SAPRC and  not the  RADM
chemical mechanism.  This  level of performance  is as good (or better) than
reported elsewhere (Carter et al., 1986) and  meets our desired performance
criteria.
Table 19.  Selected Performance Statistics in the RADM Simulations of the
           N0v-Air, CO-NO-Air, Pure Air, and Carbonyl-Air Runs
             A           A
Run Type
 No.   Experimental    -Absolute-
Runs   Avg.  (sdev)    Bias   Err.
                                — Normalized  —
                               Bias  (%)   Err.  (%)
N0x-Air
C0-N0v-Air
     A
 23
 23
 13
 13
                             1)   Change in NO  (ppm).
                             2)   Change in N02  (ppm)
 0.06  (0.49)
-0.14  (0.56)
-0.11  (0.10)
 0.06  (0.09)
-0.004  0.038
-0.005  0.025
-0.016  0.019
 0.005  0.018
Pure  Air
Formaldehyde-Air
Acetaldehyde-Air
       1)   Maximum Ozone (ppm)
       2)   Maximum Pan (ppb)
  6    0.11 (0.05)   -0.005  0.008    -4 (12)     9
  8    0.18 (0.06)    0.042  0.042    20 (13)    20
  18    0.16 (0.16)   -0.014  0.039   -12 (30)    26
  18      19   (13)       3      6      19 (40)    34
                                    150

-------
                               RADM Mechanism
p
5.
3; 0.1
^ 0
c
X" NOx Air
; *
/' CO-NOx Air
^
t/^m
& -o.m ./•"" :
X ft* !
- /. • ;
^ I- .«•' •
—0-2
-Q
D
1 -0.3
"5
0 _nA
A x' 1
-
. / \
/
/ m
/ ,.,...,. , . i
                   -0.4   -0.3  -0.2   -c.*    : -   o/    :
                      Experimental Change in [NO]  (ppmj
Figure 7.   Calculated vs Experimental Changes  in  [NO]  in N0x-Air and
            C0-N0x-Air Runs.
                               RADM Mechanism
                                                          NO A>
                  0.2 r-
                   -0.2    -0.1     0      0.1      0.2     0.3
                     Experimental Change in  [N02] (ppm)
 Figure 8.  Calculated vs Experimental Changes  in  [N02]  in N0x-Air and
            C0-N0x-Air Runs.
                                      151

-------
     The performance  of  the RADM mechanism  and  our chamber effects model
in simulating ozone  formation  in pure air and carbonyl-air experiments  is
also summarized  in Table  19-    Run-by-run  listings of  these  results are
given  in  Appendix A.   (Again,  these  results are  also  applicable to the
recommended modified mechanisms.)  As discussed  in  Section M, the  pure air
runs primarily  test  the  levels of  background  reactive  organics  and the
rate  of NOX  off gassing  used  in  the  chamber  effects  submodels.   The
carbonyl-air  runs,  especially  the  acetaldehyde-air runs,  are  good tests
for  the  NOX offgassing  rates  because  the  ozone and  PAN predictions are
very sensitive to NOX offgassing.  The  performance  statistic indicate that
on the average  there is  relatively little bias  and error (9»  in  maximum
ozone  predictions for pure-air  runs.    Ozone concentrations are  somewhat
overpredicted in  the formaldehyde-air and underpredicted in the acetalde-
hyde-air runs.  The  maximum PAN concentrations tend to be overpredicted  in
the acetaldehyde-air runs.
     Comparisons  of  the  calculated and  experimental maximum ozone yields
and ozone  formation  rates in the pure air and carbonyl-air runs are shown
in  Figures 9  and  10,  respectively,  and a  histogram  of  the normalized
biases in  maximum ozone  is  shown in  Figure 11.  There is a fair amount  of
scatter  in  these  fits   with  biases  for  ozone that are  fairly evenly
distributed between  -40?  and +50?.  A  plot  of the  error in maximum ozone
vs.  the  initial  HC/NOX  ratio  is shown  in Figure 12.  There is no system-
atic dependence  of  the  ozone prediction errors on this important ratio;
this is  a  positive result  since it  probably means   there is no systematic
source of  error  in  the mechanism.  We suspect that  the scatter  is  probably
primarily  due   to   uncontrolled  or  incompletely   characterized   chamber
effects (Carter et al., 1986).
     Overall,  the   RADM  mechanism's   performance   on  the pure-air  and
carbonyl-air  is  considered to  be  reasonably acceptable, given the uncer-
tainties in the chamber characterization  model.
                                    152

-------
                              RADM Mechanism
                600
                                                     /I  HCHO Air
                    C     100   200    300    4-00  ' 500   600
                     Experimental Ozone Maximum  (ppb)
Figure 9.   Calculated vs Experimental Maximum Ozone  in  the  Pure Air and
            Carbonyl-Air Runs.
                               RADM  Mechanism
                 100 r
                                                        HCHC-Air
                                                      100
                    Experimental d([03]-[NO])/dt (ppb/hr)
 Figure  10.   Calculated vs Experimental Rate of Ozone  Formation in the
              Pure Air and Carbonyl-Air Runs.
                                     153

-------
                               RADM Mechanism
/
6
2 5
3
(T
•4- 4
O

-------
5.M  Performance of the RADM Mechanism on Single Organic Runs
     Model performance  statistics  for maximum  ozone  and the NO oxidation
rate for  the March  1988  RADM  mechanism are shown  in  Table 20 for all of
the organic-NOx  runs.   Run-by-run listings of  these  results are given in
Appendix  A.   In general,  these  results are applicable only to the  March
1988 RADM mechanism.  However, except  for  the  aromatics-NOx and (for  the
RADM-M  mechanism)   the   C<-+   akane-NOx   runs,   the  performance  for  the
recommended  modified mechanisms  was  found to be the same or  very similar.
The  cases where  the  recommended modified mechanisms provide different
results are  discussed in  Section 5.7.
     In  the  following  discussion,  concentration-time  plots for selected
representative  experiments  will  be  presented.    In  all  cases,  the  "*"
symbols are  used to designate experimental data, and  solid lines are used
to designate model calculations  using the March 1988  RADM mechanism.   (For
experiments  where  the predictions of the recommended  modified  mechanisms
are noticeably  different, the plots  also  include results of model  simula-
tions  using  those mechanisms.  These are indicated by  dashed  lines, with
the  longer  dashes  being  for the RADM-P  mechanism  and  the  shorter  dashes
being  for RADM-M.)
     5.4.1   Carbonyl-NO^  Runs
             Tabulations   of  the  performance   of   the   RADM  mechanism  in
simulations  of  maximum ozone and NO  oxidation  rates  for carbonyl-NOx runs
are  given in Table  20, and  graphical displays of  these results are shown
in  Figures  13  through  16.  The majority of runs  in  these groups  are for
formaldehyde.   On the average, there is only +2% bias  in the maximum ozone
and  the  NO  oxidation  rate  predictions  for  the formaldehyde  runs.   The
error  in  the maximum  ozone  and timing parameter predictions are  +21? and
+ 13J,  respectively, on the  average.   The performance  of the RADM mechan-
ism's  formaldehyde chemistry  is  considered  acceptable  by  the  criteria
discussed in Section 5.2.
     The results  for  acetaldehyde runs  show  underprediction  of  both the
maximum  ozone  (-37J mean bias) and  the  rate  of NOX  oxidation (-18J mean
bias).   An  example of  one  of the  UNC Acetaldehyde-N0x runs  is  shown in
Figure 17.   The  ozone  results  for  propionaldehyde,  acetone,  and  methyl
ethyl   ketone  show biases of -31t,  +52J, and  -19>.  There  does not appear
                                     155

-------
Table 20.  Performance Statistics on Maximum Ozone and on NO Oxidation
           and Ozone Formation Rates in the Simulations of the Organic-
           N0x-Air Runs Using the RADM Mechanism
Organic
or
Mixture
Formaldehyde

Acetaldehyde

Propionaldehyde

Acetone

Methy lethy 1 ketone

Ethene

Propene

1-Butene

trans-2-Butene

Isobutene

1-Hexene

Isoprene

a-Pinene

1)
2)
No.
Runs
15
15
5
5
2
2
1
1
2
2
14
14
49
49
11
11
4
4
1
1
4
4
13
13
4
4
Statistics for Maximum Ozone (ppra)
Statistics for d( [03]-[NO])/dt (ppb/min)
Experimental
Avg. (sdev)
0.47
12.67
0.60
1.73
0.84
1.79
0.23
0.48
0.61
1.19
0.91
5.23
0.61
3.62
0.42
2.86
0.28
6.17
0.90
8.84
0.34
1.57
0.72
5.66
0.35
0.97
( 0.58)
(27.92)
( 0.40)
( 0.29)
( 0.15)
( 0.05)
( 0.00)
( 0.00)
( 0.06)
( 0.09)
( 0.20)
( 4.10)
( 0.24)
( 2.60)
( 0.26)
( 2.18)
( 0.17)
( 2.80)
( 0.00)
( 0.00)
( 0.25)
( 1.04)
( 0.25)
( 4.06)
( 0.11)
( 0.43)
Absolute
Bias Err.
0.03
0.89
-0.17
-0.29
-0.23
-0.41
0.16
0.04
-0.11
0.27
-0.09
-1.21
0.03
0.26
0.04
0.20
-0.08
-1.50
-0.19
-3.25
0.26
2.95
-0.23
0.86
-0.03
0.09
0.09
1.71
0.17
0.29
0.23
0.41
0.16
0.04
0.11
0.27
0.16
1.21
0.09
0.57
0.13
0.81
0.09
1.68
0.19
3.25
0.26
2.95
0.24
1.81
0.08
0.17
— Normalized —
Bias (%) Err. (%)
2 (
9 (
-37 (
-18 (
-31 (
-26 (
52 (
9 (
-19 (
19 (
-16 (
-27 (
5 (
6 <
-2 (
0 <
-52 I
-25 I
-23 I
-45 I
57 (
82 i
-42 i
-2
-8
4
30)
22)
17)
11)
20)
1)
0)
0)
1)
17)
30)
: 20)
; 20)
: 19)
; 56)
: 32)
: 38)
; 26)
[ 0)
[ 0)
[ 25)
( 25)
( 3D
( 36)
( 35)
( 19)
21
13
37
18
31
26
52
9
19
19
23
27
15
14
45
25
54
31
23
45
57
82
43
26
28
14
                                                               (continued)
                                     156

-------
Table 20 (continued) - 2
Organic
or
Mixture
Ethane
n-Butane
C4+ Br. Alkane
C5+ n-Alkane
Cycloalkane
Benzene
Toluene
Xylene
Mesitylene
Tetralin
Naphthalene
2,3-Dimethylnaph.
Simple Mixtures
Simple Surrogates
1)
2)
No.
Runs
1
1
27
27
7
7
14
14
4
4
6
6
20
20
10
10
8
6
5
5
5
5
4
4
40
39
36
36
Statistics for Maximum Ozone (ppm)
Statistics for d( [Oo]-[NO] )/dt (ppb/min)
Experimental
Avg. (sdev)
0.24 (
1.32 (
0.31 (
1.53 (
0.28 (
0.93 (
0.20 (
1.15 (
0.05 (
0.92 (
0.29 (
3.39 (
0.34 <
3.07 (
0.44 (
5.39 <
0.52 I
9.50 I
0.29 i
1.89 i
0.20 i
1.59
0.32
2.48
0.49
4.10
0.37
2.81
0.00)
0.00)
0.27)
0.98)
0.18)
0.46)
0.15)
0.75)
0.00)
0.12)
: 0.14)
: 2.30)
; 0.15)
: 1.87)
: 0.19)
: 4.14)
[ 0.26)
( 4.97)
I 0.23)
( 0.88)
( 0.08)
( 0.41)
( 0.04)
( 0.51)
( 0.27)
( 4.06)
( 0.15)
( 1.61)
Absolute
Bias Err.
-0.04
0.08
0.02
0.33
-0.05
-0.18
0.10
0.22
0.01
0.06
0.08
1.22
0.09
0.04
-0.02
-0.41
-0.18
-6.64
0.11
12.58
0.28
7.04
-0.01
0.13
-0.05
-0.87
0.02
0.33
0.04
0.08
0.13
0.81
0.10
0.20
0.10
0.35
0.02
0.19
0.08
1.22
0.12
0.83
0.08
1.28
0.22
6.64
0.13
12.73
0.28
7.04
0.05
0.47
0.13
1.09
0.05
0.90
— Normalized —
Bias (%} Err. (J)
-19 (
6 (
0 (
19 (
-17 (
-22 (
42 (
24 (
33 (
3 (
22 (
36 (
24 <
7 (
-4 (
0 (
-39 I
-93 i
-6 i
106 i
84
134
-6
1
-12
-13
7
8
0)
0)
59)
50)
67)
24)
28)
37)
0)
29)
15)
19)
: 43)
: 34)
: 25)
: 35)
: 50)
[ 32)
( 62)
( 85)
( 40)
( 13)
( 24)
( 24)
( 45)
( 26)
( 24)
( 38)
19
6
46
43
53
26
42
34
33
19
22
36
33
27
18
25
48
93
43
119
84
134
17
19
34
22
15
27
                                                               (continued)
                                     157

-------
Table 20 (continued) - 3
Organic
or
Mixture
SAPRC
SAPRC
7-HC Surgs.
8-HC Surgs.
UNC Misc. Surgs.
UNC "
UNC "
Synurban"
Synauto"
UNC Auto Exhaust
Syn.
Syn.
Jet Fuel
Jet Exhaust
1)
2)
No.
Runs
11
11
82
82
21
21
11
11
15
15
28
28
8
8
4
4
Statistics for Maximum Ozone (ppm)
Statistics for d( [03]-[NO])/dt (ppb/min)
Experimental
Avg. (sdev)
0.60 (
7.74 (
0.45 (
2.81 (
0.50 (
1.32 <
0.37 <
1.09 (
0.62 <
1.81 I
0.57 I
1.68 1
0.75 i
3-47 <
0.78 i
6.66 i
: 0.17)
: 5.16)
: 0.27)
: LSI)
: 0.17)
: o.4o)
; o.28)
: 0.47)
; 0.18)
; 0.63)
[ 0.27)
[ 0.61)
[ 0.10)
[ 0.74)
( 0.13)
( 3.48)
Absolute
Bias Err.
-0.01
-0.38
0.04
-0.27
0.06
0.00
0.13
0.07
0.13
0.02
0.06
-0.08
0.11
0.38
-0.09
-0.58
0.04
0.65
0.09
0.54
0.12
0.13
0.13
0.09
0.14
0.28
0.09
0.19
0.11
0.61
0.09
0.58
— Normalized —
Bias (%) Err. (%)
0 (
-5 (
9 (
-10 (
17 (
2 (
39 (
7 (
23 <
2 l
8 1
-6 1
14 l
11 i
-11 i
-8 i
: n)
: 12)
: 26)
: 23)
: 40)
: 15)
; 3D
: 8)
: 17)
; 18)
; 29)
: 15)
: 12)
; 16)
( 8)
[ 4)
8
10
20
19
27
11
39
8
23
14
19
13
14
16
11
8
                                    158

-------
                               RADM Mechanism
                •200
              .=  80C i-
              X     i

              I     I
              a,  50C -
                    0    200   IOC   6CC   500  1000   12
                    Experimental Ozone  Maximum (ppb
Figure  13.   Calculated ys Experimental Maximum Ozone in Carbonyl-N0x
             Runs.
                               R.ADM Mechanism
               f;
               G
               2  30
                  20
                    20   30     50       100      200  300
                   Experimental d([03]-[NO])/dt (ppb/hr)
 Figure  14.   Calculated vs Experimental Rates of  NO  Oxidation and Ozone
              Formation in the Carbonyl-N0x  Runs.
                                      159

-------
                               RADM  Mechanism
                  4 •
                OT
                c
                3
                a:
                
-------
     1.0-1
           OZONE
                                           0.25-,
                                                  NITRIC OXIDE
                                      1
        300   420  540  660  780  900  1020
           NITROGEN  DIOXIDE
     T	1	1	1	1	1	1	~]	1	"]

  300  420   540   660  780  900  1020



     PAN
Q.
CL
o

<
tr
UJ
o
~z.
o
o
            1  1   '  [  '   I  '  I   '  I   '  I  '
        300  420  540   660  780  900  1020
      1.0-1
           ACETALDEHYDE
0.3-,
                                             0.2-
                                             0.1-
  300  420  540  660   780  900  1020


      FORMALDEHYDE
            1   I  '—I   '  I  '   I  '   I  '  I   '
         300  420  540  660   780  900  1020
                                             0.0
         I  '   I    I     I  '   I     l
   300   420  540  660  780   900  1020
                                 EUPSED TIME  (minutes)
   Figure  17.   Experimental  and Calculated  Concentration-Time  Plots for

                Selected Species in Simulations  of UNC Acetaldehyde-NO  Run
                AU2482B.
                                           161

-------
to  be any  systematic  dependence of  the  errors on  the  initial  HC/NOX
ratio.  While  the simulation  results for the higher aldehydes and acetone
do  not  meet our  criteria  for acceptable  performance,  there are  too few
runs  with  these  compounds  to draw  meaningful conclusions  regarding the
adequacy of  the  chemistry.   Nevertheless,  we are concerned with the  large
biases for  these simulations  and recommend additional  testing when more
data become available.
     5.4.2  Alkene-NCL Runs
            The result for tests  of  the  RADM mechanism against  the alkene-
NO  runs  are shown  in  Table  20  and  in  Figures  18  through  21.  Selected
  A
concentrations  versus time plots for  one  of  the  best  simulations  of the
UNC ethene-NOx experiments are shown in Figure 22.
     The statistics  for ethene runs show  a tendency for the mechanism  to
underpredict maximum  ozone.   On the  average,  it underpredicts the ozone  by
16? with 23? error.   Also, it underpredicts  the rate of NOV oxidation  in
                                                            A
ethene runs  by 27% on the average.   The RADM  mechanism's underpredictions
are primarily  a result of its not including ethene's reaction  with atomic
oxygen,  which  we have  found is important  in  these  high concentration
experiments.   Test  calculations  where  this  reaction is  included  in the
mechanism  significantly  improves  its  performance  in   simulating   these
results;  with  its  performance   being  comparable  to that  of  the  SAPRC
mechanisms  (Carter et  al.,   1986;  Carter, 1988).   However, calculations
using  the  set of  test problems discussed in  Section  6  of  this  report
indicate  that neglecting  these   reactions  has a  negligible  effect  under
ambient conditions.   For  that reason,  this reaction  is not  included  in the
recommended  modified mechanisms.    However,  its omission  does limit the
range  of  validity of this mechanism to  a concentration regime  below that
of  these chamber  experiments.
     There  are a large number of  propene-NOv  runs in the data base  and the
                                            A
RADM mechanism performs very  well on these runs.   Results from  two  typical
propene-NOx  simulations are  shown  in Figures 23 and 2U.   The  mechanism
predicts  the maximum ozone and rate of  NOX oxidation with a +5? bias and
 15? error  on the average.  The results  for 1-butene simulations also show
a  small mean bias (-2?),  but the mean error  (45?) is much  higher than for
the propene runs.  Figure 25 shows  a fairly typical  simulation  of a UNC  1-
butene-NOx   experiment  where the ozone  is overpredicted by -30?.   (PAN  is
                                    162

-------
                               RADM Mechanism
                1400
              p 100C
              X
              2  800
                    i

              O  600 -

              o     i
     •°...vx
                          T-2-5utene
   **

./"••
                          ^r

                          A »
                    0   200  4CO  SCO   300   1CCC  '2CO 14QC
                     Experimental  Ozone Moxinaum (ppb)
Figure  18.  Calculated vs Experimental Maximum Ozone in the  Alkene-N0x
            Runs for the RADM Mechanism.
                               RADM Mecnanism
                    20      50    100   200     500   1000
                    Experimental d([03]-NO])/dt (ppb/hr)
 Figure 19.  Calculated ys Experimental Rates  of NO Oxidation and Ozone
             Formation  in the Alkene-N0x Runs  for the RADM Mechanism.

                                      163

-------
                              RADM Mechanism
               OT
               C

               tr
                  5 -

-
•
__


.


,*








\ •'-•'• 1
IP •









*.



I — 1





T-° ,

—







vO


. —






O

—







vO









^0








'/,
^








n
IP







//t
/^
*p
{: | Ethene
1^1 Bu tones
\/\ Hexene








                        Bias in Maximum Ozone (%)
Figure 20.  Histogram of Normalized Biases in Maximum Ozone  in  Alkene  Runs
            for the RADM Mechanism.
                              RADM Mechanism


g 100
1)
c
0
N 50
E
3
C n
	 U
X
0
2
c
". -50
a
5
— inn
0
A
a

A • •
• A
*^ ,
•sr^'V * * *
. V ":*"•"-
*
1 *
A *
-
Ethene
A
Propene
*
Butenes
A
Hexene
o






                     0.2    0.5   1    2     5

                                HC/NOx Ratio
10
                                                    20
Figure 21.   Normalized  Bias  in Maximum Ozone vs HC/NOX in Alkene Runs
             for  the  RADM Mechanism.

-------
            OZONE
                                              NITRIC OXIDE
 Q.
 a

z:
o

I
LJ
O
-z.
o
o
0.30 -,



0.24-



0.18-



0.12-



0.06-
        300  420   540   660   780  900  1020



            NITROGEN DIOXIDE
     0.00
                                             0.8-1
0.6-
0.4-
0.2-
                                           300   420  540  660   780   900   1020



                                              FORMALDEHYDE
                                             0.0-
                    I  '   :   •  •   •  !   '   |    —''    '   I

        300  420   540   660   780  900  1020    300   420   540   660  780   900  1020
      1.0-1
            ETHENE
      0.0
               I     IT     I     I      I
        300  420   540   660   780  900  1020
                                 ELAPSED TIME  (minutes)
  Figure  22.   Experimental  and Calculated  Concentration-Time  Plots for

               Selected Species in Simulations  of UNC Ethene-NO  Run

               OC0584B.
                                          165

-------
           OZONE
                                                 NITRIC  OXIDE
 E
 Q.
 CL
g

or:
o
               120    240     360


           NITROGEN DIOXIDE
     0.0
               120     240     360     480
                               ELAPSED  TIME (minutes)
  Figure 23.  Experimental  and  Calculated  Concentration-Time Plots for
              Selected Species  in  Simulations  of the  Propene-N0v Run
              EC-216.
                                        166

-------
     O.B-i
     0.6-
     0.4 H
      0.2-
           OZONE
                                           NITRIC  OXIDE
      0.0
           1
                r
300  420  540   660   780   900  1020



   NITROGEN DIOXIDE
E
Q.
Q_
Ld
O
•z.
o
o
0.04-

    -
 —/
     °-QO
I
                I
                                       I
300  420  540  660   780   900  1020



    PROPENE
      0.4
         j


      0.2 J




      0.1 -
      0.0
                              0.15-1
                                   i


                              0.12-1



                              0.09-



                              0.06-



                              0.03-

                                   I


                              0.00-i-
                                                300  420  540  660   780   900  1020



                                                    PAN
                                      0.2
                                                300  420  540  660  780  900  1020



                                                    ACETALDEHYDE
                                      0.2J



                                      0.1-

                                         4

                                  8.0E-02-



                                  4.0E-02-
        300  420  540  660  780  900  1020
                                      0.0
                                             ^\^ •  i   :  :   ;  :   •       i   •
                                        300  420  540  660  780  900  1020
                                  ELAPSED TIME (minutes)
  Figure 24.  Experimental and  Calculated  Concentration-Time Plots  for

               Selected Species  in Simulations of UNC  Propene-NO   Run

               JL2983B.
                                          167

-------
 a
 a
Ld
o
z
o
o
      0.8-
      o.e-
            OZONE
         300   420  540  660  780  900  1020
            NITROGEN  DIOXIDE
      0.0
                      ' — I — ' — 1 — •  i  •   i  •
        300  420  540  660  780  900  1020
      0.8-n
      0.6-
      0.4-
            1 -BUTENE
                                              0.4-,
                                              0.3-
                                                    NITRIC OXIDE
                                             0.24-.
                                             0.18-
                                             0.12-
0.06-
    300   420   540  660  780  900  1020



       PAN
                                             °-0°
                                                                      //
                     11 • i—i—r—i—]—i

   300  420  540  660  780  900  1020
        300  420  540  660  780   900  1020


                                  ELAPSED  TIME (minutes)
  Figure 25.   Experimental  and Calculated Concentration-Time Plots for
                *?& 1 ,d *•* «• n rf O «*«*.*•[.>„. i 	 r* •   t  ,
                                                  of  UNC 1-Butene-NOx  Run
                                          168

-------
greatly  overpredicted  in  this  simulation because  the model  species PAN
represents  peroxypropionyl  nitrate as  well  as  PAN.)   Figure  21 clearly
shows  that  there  is  a lack of systematic  dependence  of the biases on the
initial  HC/NOX ratios  in  these runs.  There are a small number of runs for
other  butenes  and  1-hexene.   The mechanism's  performance on these runs  is
poor.   The ozone yields  and  rate of NOV  oxidation in the  trans-2-butene
                                        A
and  isobutene  runs  are  underpredicted.  Conversely,  they  are  both  over-
predicted by almost 60% on the average in the  1-hexene runs.
     Clearly,  the propene  portion  of  the  mechanism  perform  well and easily
meet the desired performance  criteria.  The 1-butene  results also meet the
criteria,  although they  are  not  as accurate.   The  internal  alkene portion
of  the RADM mechanism do not perform well, but  once again, this is  based
only on  five experiments.  The mechanism  does not perform as well as  other
mechanisms  in  the simulations of the  ethene, runs because  the concentration
regime in  these runs  is  beyond  the range of  validity of  this mechanism.
Test calculations show that  if the 0(^P)  + ethene reaction  were added,  the
mechanism would perform better in  simulating the ethene runs.
     5.4.3  Alkane-NOH Runs
            The model  performance  results for alkane-NOv  runs  are shown  in
                                                        A
Table  20 and in Figures  26  through  29.   Typical simulations of  n-butane,
n-octane,  and  2,3-dimethylbutane  experiments  are shown   in  Figures  30
through  32, respectively.    The  simulation  of the one  ethane-NOx  system
underpredicts  the  maximum ozone by  18%.   For the  butane experiments,  the
mechanism  shows  no bias  but  has  a fairly large error  of  46/1.   For  the >C4
branched alkanes,  the mean  bias and error are  -17% and 53%,  respective-
ly.  However,  for  the >C4 long chain alkanes, the.  mean bias and  error are
both 42%.   For the  cyclohexane  runs,  the mechanism predicts ozone with a
bias  and error  of  33% on the  average.   Overall,   for  all of the  alkane
experiments,  the RADM mechanism is able  to predict the maximum ozone with
a  relatively  low bias but a relatively  high error.   The statistics for the
rate  of NOX  oxidation are  only  slightly better.   Figures 26 through  29
confirm that there  is a large amount of scatter in  the results.  Figure 29
shows  that there does not appear  to be  any  systematic  dependence  of the
bias on initial  HC/NOX ratios, which  is reassuring.
                                     169

-------
                                RADM Mechanism
                 1000
              .a
              a.
              a.
              x
              O
              c
              O
              N

              O
              O

              O
                 800 -
                  600 •
   400 •
                  200
                     0     200    400    600    800    1000


                     Experimental Ozone Maximum (ppb)
Figure 26.   Calculated vs  Experimental Maximum Ozone  in the  Alkane-N0x

             Runs for the RADM Mechanism.
                                RADM Mechanism
£  500

\
.C
Q.
vCX 200



^ 200



cT
z

^ 100
I—"1


^


T5   50

T3


°   30
               O
               o
                                                          Butane

                                                           A



                                                       >C- 3rcnct-.ec

                                                           •


                                                       >C4 Long Chain

                                                           •


                                                       Cyclic Alkanes

                                                           A
    20
      20   30    50    100    200  300   500


    Experimental d([03]-[NO])/dt (ppb/hr)
Figure 27.  Calculated vs Experimental Rates of NO Oxidation and Ozone

             Formation in the Alkane-N0x Runs for  the RADM Mechanism.
                                      170

-------
                              RADM Mechanism
1^
12
£ 10
rr
*- B
O

-------
            OZONE
      0.0
                                            0.06
                                            0.04
                                            0.02-
                                            0.00
                 NITRIC OXIDE
         0     120    240    360




            NITROGEN DIOXIDE
480   600
                                                          ••*•••••••••••_•
                                                                     •*••••••••••
120    240   360    480    600
                 PAN
E
a
a


z
a:
CJ
2:
o
o
     0.00
                                                   1   I—'—I—'—I—r
                                                     120    240    360    480
                                             600
            n-BUTANE
                 FORMALDEHYDE
                                                     120    240    360    480   600
                      -      1	'	1	r   .
         0     120    240    360   480   600
                                 ELAPSED  TIME (minutes)
  Figure 30.  Experimental and Calculated  Concentration-Time  Plots for

               Selected  Species in Simulations of n-Butane-NOx Run EC-130

-------
           OZONE
E
Q.
Q.
or
u
o
z
o
o
                                              NITRIC OXIDE
                 "I	1	1	1	1	1
         0   60  120  180  240  300  360  420
                                           0   60  120  180  240 300  360  420
     0.12-
0.09 -I
    0.06-
    0.03-
    0.00-
       NITROGEN DIOXIDE

             • _"•••••••,
            •   	=	
           '•
       ~1    1     I	1	1	1	1
        60  120  180  240  300  360  420
                                           0.003-,
                                           0.002-
                                          0.001 -
                                          O.OOO-i
                                                  PAN
                                                    I
                                                    60
                                                   120  180  240  300  360  420
                                 EUVPSED TIME (minutes)
  Figure 31.  Experimental and Calculated  Concentration-Time  Plots for

               Selected  Species in Simulations of n-Octane-NO   Run ITC-552.
                                                               A
                                        173

-------
           OZONE
Q_

Q.
UJ
 .
o

-------
     We do not believe  the  large error in the results indicates that there
is necessarily  a  problem in the  mechanism  because model  simulations of
alkane-NOv runs are very  sensitive to the chamber effects parameters.  The
         A
variability  in  the chamber  effects could account for a significant amount
of the variability in  the  mechanism's  predictions  for alkane  runs.   As
noted  in  previous evaluations  (e.g.,  Carter  et  al.,  1986),  it  is not
possible to  unambiguously evaluate the alkane portion of the  mechanism on
chamber  data because  of the  uncertainties  in chamber  effects and  vari-
ability.   Therefore, within  the limitations  of  the testing  methodology,
the  alkane reactions appear  to  give acceptable predictions for ozone and
the  NOV  oxidation  rate.
       A
     5.4.4  Aromatic-N
-------
                            RADM Mechanism
         O
                0      200     400     600     800     1000

                  Experimental Ozone Maximum (ppb)
Figure 33.  Calculated ys  Experimental  Maximum Ozone in the Aromatic-
            NCL Runs for  the  RADM Mechanism.
              A
                            RADM Mechanism
             000
          .O
          CL
          _  1000

          •V
          -O

          ^  500
          r—t
          o
          z


          •-U  200

          C^

          T^  100


          •«-
          ^   50
 Benzene
   *

 Toluene
   •

 Xylenes


Mesitylene
   A

  Other
          O          50    100    200      500   1000   2000

                Experimental  d([03]-[NO])/dt (ppb/hr)

 Figure 31*.   Calculated vs Experimental Rates of NO Oxidation  and  Ozone
             Formation in the Aromatic-N0.x  Runs for the RADM Mechanism.
                                     176

-------
                               RADM Mechanism
                  12
               cr
               *-
               o
^
                              I
                                    "1
                                                       ;-;-;-:';l Mesitylene

                                                         Xylenes

                                                         Toluene

                                                         Benzene
                        Bias in Maximum Ozone (%)
Figure  35.   Histogram of Normalized  Biases in Maximum  Ozone in Aromatic
             Runs for the RADM Mechanism.
                               RADM Mechanism
TL/U
g

-------
           OZONE
 E
 Q.
 Q.
g
i
LJ
O
2
O
O
                                            0.5-,
                                                 NITRIC  OXIDE
0    60   120   180   240   300   360


  NITROGEN  DIOXIDE
                                            0.0
60   120   180
               240   300  360
                                                 PAN
        0    60   120   180   240   300  360
                                                   60   120   180   240   300  360
                                ELAPSED TIME  (minutes)
   Figure 31.   Experimental and Calculated Concentration-Time Plots for
                Selected Species in Simulations of the  Benzene-NO  Run
                ITC-562 (	 RADM, 	 RADM-P,  	  RADM-M^.
                                        178

-------
            OZONE
 CL
 Q.
v-x

z
o


I


uj
o
o
                                             0.4-,
                                             0.3 -I
                                             0.2-
                                                   NITRIC OXIDE
      0.0
                      "I     I

0   60  120  180  240  300  360  420
      0.3-
            NITROGEN DIOXIDE
                                                      ] - 1 - r

                                                 0   60  120  180  240  300  360  420
                                                   PAN
      0.0-
     0.6-



     0.5 -,



     0.4-



     0.3-




     0.2-



     0.1-



     0.0
                  i     i	1	r
             60  120  180  240  300  360  420
           TOLUENE
                                   0.06 -,



                                   0.05-



                                   0.04-




                                   0.03-



                                   0.02-



                                   0.01 -
                                       6    60   120  180  240  300




                                          FORMALDEHYDE
                                                                           360
            ~~1I     I	1	1	1	1
             60   120  180  240  300  360  420
                                            0.00
                                       6   60   120  180  240  300  360  420
                                 ELAPSED TIME  (minutes)
  Figure 38.  Experimental  and Calculated  Concentration-Time Plots  for

              Selected Species in Simulations of the Toluene-NO  Run
                                         179

-------
       0.5 -,
             OZONE
                                             0.16-
                                             0.12-
                                             0.08-
                                             0.04-
                                                    NITRIC OXIDE
                            . "'  I—'—I—'—I
         300 420  540  660  780  900 1020 1140
                                             0.00-
                                                         • i
 300  420  540  660  780  900 1020 1140
            NITROGEN DIOXIDE
                                                   PAN
 a
 a
or
UJ
o
z
o
o
            1  I  '  I—'   I  '  I  '  I   '  I  '—1
        300  420  540 660  780  900 1020 1140


            TOLUENE
              1  '—I—'—I—'—1—'
300  420  540  660  780  900 1020 1140
                                                  FORMALDEHYDE
              I  '  I  '   I  '  I  '  I   '  |  '  ,
        300  420  540 660  780  900 1020 1140
                 '—I—'—I—'—I—'
300  420  540  660  780  900 1020 1140
                                 ELAPSED  TIME (minutes)
  Figure  39.   Experimental  and  Calculated Concentration-Time  Plots  for

               Selected Species  in Simulations  of the UNC Toluene-NO  Run
               JL3080R.                                                ^
                                        130

-------
           OZONE
                                       NITRIC OXIDE
 E
 Q.
 a
g
i—
LU
a

o
o
                                                         120    180   240   300
0    60    120



  M-XYLENE
     0.0
                               ELAPSED TIME  (minutes)
 Figure 40.  Experimental and Calculated Concentration-Time Plots for

             Selected Species in Simulations of the m-Xylene-NO  Run

             ITC-702.
                                       181

-------
     0.6-



     0.5-



     0.4-



     0.3-



     0.2-



     0.1 -



     0.0 -<•
           OZONE
                 NITRIC OXIDE
          0.12-
          0.08-
          0.04-
        300  420  540  660  780  900  1020  1140
          0.00
              300  420  540 660  780 900  1020 1140
ex
Q.
g

*
UJ
o
z
o
o
     0.15-.
            NITROGEN DIOXIDE
     0.00
          0.10-1




          0.08-




          0.06-



          0.04-




          0.02-
                 PAN
              :    I  ;^  ;  !  :  •   ; T •  :
        300  420  540  660  780  900 1020 1140
          0.00--•• i',-.
              300  420  540  660  780  900  1020 1140
      0.3-1
      0.2-
            0-XYLENE
          0.20-,
                 FORMALDEHYDE
              I    T    I  '  '  '   :  'I  f   :

         300  420  540  660  780 900  1020 1140
              300  420  540  660 780  900 1020 1140


ELAPSED  TIME (minutes)
   Figure  41.   Experimental and Calculated Concentration-Time  Plots for

                Selected Species in  Simulations of  the UNC o-Xylene-NO  Run
                JL3080B.
                                          182

-------
      0.5-,
            OZONE
                                             0.5-,
                                             0.4
                                                   NITRIC OXIDE
                                             0.0-

              60   120   180   240  300  360      6    60   120   180   240  300  360
            NITROGEN DIOXIDE
Q.

CL
g

S
UJ
o
~z.
o
o
                         :


                      V • • v« • ••..,...
      0.0
                                             0.3-,
                                             0.2-
              60   120   180   240  300  360
                                                   PAN
                                             0.1-
                                                     60   120   180   240  300  360
      0.3-»
      0.2-
      0.1 -
            MESITYLENE
      0.0-
                                            0.08 -,
                                            0.06-
                                            0.04-
                                            0.02-
                                                   FORMALDEHYDE
             n	1	1
              60   120   180   240  300  360
                             -f	T	T   o.oo-
                                                     60    120   180   240   300  360
                                 ELAPSED TIME (minutes)
  Figure U2.  Experimental  and Calculated  Concentration-Time  Plots for

               Selected  Species in Simulations of the Mesitylene-NO  Run
               EC-901.                                                x
                                          183

-------
     Considering the  uncertainties  in the yields and subsequent  reactions
of the  aromatic  ring-opening products, this level of performance might  be
considered  acceptable.    However,  unlike  the alkanes,  the  modeling  of
aromatic runs should  not  be highly  sensitive  to uncertain  chamber effects,
and  cases  of generally poor  performance can indicate  problems with  the
mechanism.   As  discussed in Section 2, other mechanisms perform  better  in
simulating  results of  these experiments,  and there are  several  changes
that can be  made  to the RADM aromatic mechanism which can  reduce the error
considerably  and  reduce  the  systematic  bias   in  ozone  yields  between
benzene  and  mesitylene.    The  modified  RADM  mechanisms,  discussed  in
Section 5.7, perform  better in  simulating the results of these  runs.
     5.4.5   Natural Hydrocarbon Runs
             The  result of tests on  experiments  with  two natural  hydro-
carbon  species,  isoprene  and  alpha-pinene, are  shown  in  Table 20  and  in
Figures  43 through 46.   Comparisons of  predicted and  observed concentra-
tion  profiles  for two  isoprene  runs and  one  alpha-pinene  run  are shown in
Figures  47 through 49.  Although there are some  minor  differences  between
the  recommended  modified  and  the March  1988  versions of  the  RADM  mechan-
isms  in their  reactions  of the natural hydrocarbons, these  differences
have  no significant  effect  on  the  predictions for  isoprene and only minor
effects on predictions  for alpha-pinene.  (The  most important difference
is  a slight change  in  the rate constant  used  for  the reaction  of ozone
with  OLI,  the  lumped  model  species which  is   used  to  represent  alpha-
pinene.)  Therefore,  the  results discussed in this  section are applicable
to  all  three versions of the mechanism.
      In the case of  isoprene,  the  mechanism  has  a  significant tendency to
underpredict ozone yields,  with an average  bias of -42/L   This level of
performance does not meet the  desired performance  criteria.   On the other
hand,  it has very  little bias  (-2%)  in  simulating  the NO oxidation rate,
and the error (26/t)  is only slightly  higher  than in the simulations of  the
propene  runs.    This  mechanism represents  isoprene   explicitly,  so  in
principle one should expect  it to perform  reasonably  well  in simulating
 this compound.   However,  although the RADM mechanism uses the appropriate
 rate  constants  for  the  initial  atmospheric  reactions  of  isoprene,  it
 represents  the  products  of  these reactions with the  same set of species
                                     184

-------
                           RADM Mechanism
                                                      Isoprene

                                                         A


                                                    Alpho-Pinene
                    200  400   600   800   1000  1200  1400

                 Experimental Ozone Maximum (ppb)
Figure 43.  Calculated vs Experimental Maximum Ozone in the Natural

            Hydrocarbon Runs for the RADM Mechanism.



                            RADM Mechanism
          g- 1000
         \  500


         o1
         2^

         JL,  200

         c^

         *T3  100



         "5   50
          c?
          o
                                                       Isoprene

                                                         A
    ,50    100    200      500   1000

Experimental d([03]-[NO])/dt (ppb/hr)
Figure ^4.  Calculated  vs Experimental  Rates  of NO Oxidation and Ozone

            Formation in the Natural  Hydrocarbon Runs for  the  RADM
            Mechanism.
                                    185

-------
                               RADM Mechanism
                c
                o:
                •4-
                o

                
-------
E
Q.
Q.
ce
h-

LJ
o
z
o
u
     1.0-1
           OZONE
     0.5-.
         0      60      120


           NITROGEN  DIOXIDE
        180
       240
                    0.4-,
                                          0.3-
                                          0.2-
                                          0.1-
                                          0.0-
                                         0.15-,
                         NITRIC OXIDE
                                     120
                                     180
                                     240
                         PAN
     0.0-
               60
           ISOPRENE
120
180
240
0      60     120     180


  FORMALDEHYDE
               60
120
180     240      0       60      120


 ELAPSED  TIME (minutes)
                                                                   180
                                                    240
  Figure 47.   Experimental  and Calculated Concentration-Time Plots for

              Selected  Species in Simulations of the Isoprene-NO  Run

              ITC-811.
                                       187

-------
     1.0-,
           OZONE
                                            0.15-
                                              NITRIC  OXIDE
        300   420  540   660   780   900  1020      300   420  540  660   780   900  1020
Q.
a

•z.
or
I—

bJ
O
Z
O
O
    0.16-,
0.12-
     0.08-
0.04-
           NITROGEN DIOXIDE
     0.00
                                            0.12-1
                                        0.08-
                                        0.04-
                                               PAN
                                        0.00-

    300   420  540  660   780   900   1020      300   420   540  660   780   900   1020
            ISOPRENE
                                              0.4-,
                                              0.3-
                                              0.2-
                                              0.1-
                                               FORMALDEHYDE
                                              0.0
               -i—i—I—i—r"~i—I—'—I—'—T    "'" "1  '   T  T" 1   '  I   '   I  '   I   '  I
         300  420  54-0  660   780   900  1020      300   420  540  660   780   900  1020
                                  ELAPSED TIME  (minutes)
   Figure 48.   Experimental  and Calculated  Concentration-Time Plots  for
                Selected Species in Simulations of the  UNC Isoprene-N0x Run

                JL1680B.
                                         188

-------
           OZONE
E
Q_
Q.
tr
 .
UJ
o
-z.
o
u
                                            0.16
                                            0.12-
                                            0.08-
                                       0.04-
                                              NITRIC OXIDE
                                            0.00

        300  420  540   660   780   900  1020     300   420   540   660  780   900  1020
     0.16-,
0.12-
     0.08-
     0.04-
            NITROGEN DIOXIDE
     0.00
                                       0.12-,
                                            0.08^
                                            0.04-
                                        0.00
                                              PAN
        300   420   540   660   780  900  1020      300   420   540  660  780   900   1020


                                 ELAPSED TIME (minutes)
  Figure 49.  Experimental and Calculated Concentration-Time Plots  for

               Selected  Species in Simulations of  the  UNC a-Pinene-NO  Run

               JL1580B.
                                         189

-------
used  to represent  the  terminal  alkenes.   Preliminary results  from  our
laboratories indicate that using mechanisms with more explicit  representa-
tions  of  the  reactions  of  isoprene's  major  products  will  result  in
somewhat better  model performance, although  there  are  still  inconsisten-
cies  between the  model  and the  data  (Carter and  Atkinson,  unpublished
SAPRC  results).   [In addition,  approximately 40* of the  products  in  the
isoprene-OH   reaction  system  have  not  been  unidentified   (Atkinson,
unpublished  SAPRC  results).]  Better performance of the RADM mechanism in
simulating  isoprene  runs would probably require  adding  species  to  the
mechanism to represent  the  reactions  of isoprene's  oxidation products,  but
the mechanism would still be uncertain.
     The  performance  of  the  RADM  mechanism  in  simulating  the  limited
number  of alpha-pinene  runs, all carried out  in the UNC chamber in  July of
1980,  is surprisingly good.   The mean bias and error in the ozone  maximum
are -8% and 28%, respectively, while the bias and error  in the  NO oxida-
tion  rate are only  +H% and 1MJ.   (These four  statistics are  -8/t,  27%,  -»-2%,
and  13%,  respectively,  for the RADM-P and RADM-M mechanisms.)   This level
of  performance  is perhaps  fortuitous,  since  the atmospheric chemistry of
alpha-pinene is  not well understood and  is represented by the  same lumped
model   species  (OLD  which  is  used  to  represent  the internal  olefins.
Indeed, as  discussed in Section 3.3, the rate constants and product yield
parameters   for  this  lumped  species are  derived  based  on analysis  of
anthropogenic emissions  data,  where  the   contribution  of  terpenes  is
negligible.   Nevertheless,  the  alpha-pinene  results  for  these  four  runs
meet  the desired performance  criteria.    Of  course,  these  results  do not
necessarily indicate that the performance would be  equally  good for beta-
pinene or the other  terpenes.
      5.4.6   Performance in Simulations  of Formaldehyde  Yields
             Model  performance statistics  for  the  maximum  formaldehyde
concentrations   are  shown  in Table  21.   Comparisons  of  the individual
calculated  and  experimental values for  all of the  single  organic runs are
shown  in   Figure  50.     The  maximum  formaldehyde  concentrations  are
underpredicted   on   the  average   in  carbonyl,   aromatic,  and  natural
hydrocarbon  runs  and  overpredicted  by  small amounts  on  the  average in
alkene and  alkane runs.  The  mean errors in  the  prediction  are large:  on
 the  order   of  30 to 80% for most  types of runs.   Thus, there clearly is a
                                     190

-------
Table 21.  Performance Statistics on Maximum Formaldehyde in the
           Simulations of the Organic-N0x-Air Runs Using the RADM
           Mechanism
Compound or
Mixture
Acetaldehyde
Propionaldehyde
Ethene
Propene
1-Butene
trans-2-Butene
Isobutene
1-Hexene
Isoprene
a-Pinene
n-Butane
C4-t- Br. Alkane
C5+ n-Alkane
Cycloalkane
Toluene
Xylene
Mesitylene
Naphthalene
Simple Mixtures
Simple Surrogates
SAPRC 7-HC Surgs.
SAPRC 8-HC Surgs.
UNC "Synurban"
UNC "Synauto"
UNC Auto Exhaust
Syn. Jet Fuel
No.
Runs
5
2
14
39
8
4
1
4
13
2
21
5
10
4
20
10
8
5
34
7
11
82
11
13
28
8
Experimental(a)
Avg. (sdev)
0.17 (
0.10 (
0.86 (
0.36 (
0.32 (
0.08 (
0.64
0.33 (
0.30 <
0.08 (
0.04 (
0.03 (
0.01 I
0.07 I
0.09 i
0.09 i
0.09 i
0.01 -
0.25 '
0.13
0.38
0.18
0.12
0.16
0.18
0.05
0.
0.
0.
0.
0.
0.

: o.
! o.
: o.
; o.
; o.
: o.
[ o.
( o.
15)
01)
41)
20)
27)
06)

27)
16)
04)
03)
02)
.01)
.12)
.09)
( 0.05)
[ 0.04)
( 0.01)
( 0.20)
( 0.05)
( 0
( 0
( o
( o
( o
( o
.23)
.12)
.04)
.08)
.10)
.03)
Absolute(a)
Bias Err.
-0.
0.
-0.
0.
' 0.
0.
-0.
0.
0.
-0.
0,
-0.
-0
-0
-0
-0
-0
0
0
-0
0
-0
-0
0
-0
-0
05
01
02
00
10
04
,02
30
,01
,01
.01
.02
.01
.06
.04
.05
.05
.02
.02
.05
.09
.02
.05
.02
.10
.03
0.
0.
0.
0.
0.
0.
0.
0.
07
01
23
10
10
06
02
30
0.12
0.01
0.03
0.02
0.
0,
0
0
0
0
0
0
0
0
0
0
0
0
.01
.06
.04
.05
.05
.02
.09
.06
.09
.05
.05
.03
.10
.03
Normalized
Bias (%) Err. (%}
-5 (
10 (
-9 (
4 (
21 (
60 (
-3
76 (
-6 <
-24 1
35 <
-75 i
-26 i
-32
-48
-71
-75
100
-14
-45
18
-1
-55
13
-68
-58
56)
17)
41)
45)
; 18)
: 58)

: 33)
: 59)
; 12)
( 73)
( 68)
(110)
(133)
( 56)
( 51)
( 45)
( 64)
( 56)
( 45)
( 18)
( 47)
( 18)
( 3D
( 3D
( 62)
44
12
34
33
22
71
3
76
46
24
70
91
95
101
54
76
75
100
47
56
21
35
55
24
70
67
 (a)  Concentrations in ppm.
                                     191

-------
                             RADM  Mechanism
o
2_
Si
c
5
E
X
C
2
g
c
|
~
-

1

0.3


0.1

0.03

0.01

3.003

4 '•' '
. *.'i< *
• -IxV"***
* *Si 9^/1} * ^
• 5»» *
*1 /'*"""
a A"»,^tf 44 *
o o ^ 2 ° 4^/^:° °
* ^Vi * • = 0
"x *•*** a
/

RCHC Air
A
Aldehydes

Alkenes
*
Alkanes
A
Aromotics

Biogenics


                 OOC1  0.003  0.01   003   O.1   C.3    1
                    Experimental HCHO Maximum (pom)
Figure 50.  Calculated vs Experimental Maximum Formaldehyde  in  the
            Single Organic-N0x Runs for  the RADM Mechanism.
large  amount  of  scatter  in  the  formaldehyde  predictions.    While  the
accuracy of  the  measurements is  questionable, especially at the low levels
observed  in  some runs,  it  is unlikely  that  this is the  primary  cause of
the  poor  comparisons in  all  cases.   Formaldehyde is difficult  to predict
because  it is both formed and destroyed  in the  photooxidation cycle.  The
RADM mechanism and other mechanisms (e.g.,  Carter  et  al., 1986)  are not
able to reproduce  the  observed  formaldehyde levels  well.   If  the same
desired  performance criteria  that  were used for ozone  and  rate  of NCL
oxidation  were  used  for  formaldehyde,  the  mechanism  would clearly  be
rejected.    Fortunately,  accurate prediction of  formaldehyde is  not the
primary purpose  of  the mechanism.
      5.4.7  Performance  in  Simulations  of PAN Yields
             Model  performance statistics for  the maximum  PAN concentra-
tions are shown in Table 22,  and  comparisons of the individual calculated
and  experimental values  for all of  the single  organic  runs  are shown in
Figure 51.    These results show that the RADM mechanism predicts PAN well
                                     192

-------
Table 22.  Performance Statistics on Maximum PAN in the Simulations of
           the Organic-N0x-Air Runs Using the RADM Mechanism
Compound or
Mixture
Acetaldehyde
Propionaldehyde
Methylethylketone
Propene
1-Butene
trans-2-Butene
Isobutene
Isoprene
n-Butane
C4+ Br. Alkane
C5+ n-Alkane
Toluene
Xylene
Mesitylene
2,3-Dimethylnaph.
Simple Mixtures
Simple Surrogates
SAPRC 7-HC Surgs.
SAPRC 8-HC Surgs.
UNC Misc. Surgs.
UNC "Synurban"
UNC "Synauto"
UNC Auto Exhaust
Syn. Jet Exhaust
No.
Runs
5
2
1
46
7
4
1
12
21
6
11
20
10
8
4
38
32
11
68
12
10
15
28
4
Experimental(a)
Avg. (sdev)
0.141 (0.088)
0.052 (0.036)
0.070
0.142 (0.084)
0.039 (0.016)
0.080 (0.044)
0.137
0.133 (0.082)
0.296 (1.012)
0.226 (0.506)
0.247 (0.814)
0.051 (0.036)
0.125 (0.102)
0.407 (0.189)
0.046 (0.019)
0.081 (0.059)
0.048 (0.020)
0.095 (0.047)
0.044 (0.030)
0.055 (0.046)
0.024 (0.024)
0.054 (0.039)
0.047 (0.029)
0.080 (0.03D
Absolute(a)
Bias Err.
-0.002 0.020
0.129 0.129
-0.001 0.001
0.044 0.051
0.054 0.055
-0.004 0.014
0.141 0.141
-0.029 0.080
-0.257 0.294
-0.207 0.224
-0.243 0.247
0.019 0.037
-0.041 0.043
-0.317 0.317
-0.001 0.015
0.018 0.033
0.004 0.018
0.035 0.035
0.039 0.040
0.007 0.024
0.012 0.012
0.014 0.029
-0.003 0.009
0.098 0.098
Normalized
Bias (%) Err. (%)
-1 ( 18)
110 ( 63)
-1
28 ( 39)
64 ( 54)
-14 ( 25)
68
-19 ( 85)
17 (104)
-44 (123)
7 (126)
39 ( 65)
-24 ( 47)
-124 ( 40)
-2 ( 44)
14 ( 56)
12 ( 57)
33 ( 13)
66 ( 34)
33 ( 69)
48 ( 30)
30 ( 43)
-14 ( 55)
78 ( 13)
14
110
1
34
69
21
68
61
89
104
98
61
26
124
33
48
39
33
67
55
49
44
36
78
 (a)  Concentrations in ppm.
                                     193

-------
                           RADM Mechanism
,-v t /I RCHO Air
E 1 • . ,-•"
g; 0.3 f . **\"JL
"-' 1- • • _ •* ,**+^ °
1 o.
E
X
•i 0.03

Z
0 » >^ " o °
0 ^i^°0^*ft 0 0
° &* °0*^«* ^ °° *
" . * *>s*/ ' •

... .*/• °*
<— ^— ^ ^,
& A. V
CL 0.01
•n
\"'X** *
/ A
A
Aldehydes
.
Alkenes
•
Aikones
^

Arornatics
o
Biogenics
           a>
           5 0.003
           o     i
           O O.OOi t
                 O.OC1  0.003   0.01   0.03    0.1   0.3
                    Experimental  PAN Maximum (ppm)
Figure 51.  Calculated vs Experimental Maximum PAN  in the Single
            Organic-N0x Runs for the RADM Mechanism.
in  acetaldehyde runs,  but poorly  in most  other runs.   The  model over-
predicts  the PAN  concentrations  in  most  but not  all  types of  runs,  and
there  is also  a  large  amount of  scatter  in the  PAN  predictions.   The
graphical  comparisons confirm this.   Cases  of positive bias are at least
partially  explained by  the fact that PAN is used  in  the mechanism as a
surrogate   for   PAN  plus  many   analogous   organic   peroxyacyl  nitrate
species.   However, this  is not expected  to be a factor  for  the propene,
isobutene  and  toluene  runs.   On  the  other hand,  PAN  is  significantly
underpredicted  in the mesitylene  runs,  and  is  slightly underpredicted  in
the trans-2-butene.  xylene, isoprene, and C4+ branched alkane  runs  (though
with large  mean errors  in the latter  two  cases.)   If  the  same  desired
performance criteria that  were used  for  ozone  and rate  of  NOX oxidation
were used  for  PAN, the  mechanism  would  clearly  be rejected.   Given  the
partial mismatch  between  the  species that mechanism predicts  and what  is
measured,  this  criteria  is probably too  strict.
                                     194

-------
5.5  Performance of the RADM Mechanism on Runs with Organic Mixtures
     Model performance  statistics  for maximum  ozone  and the NO  oxidation
rate for the March  1988  RADM  mechanism are shown in Table 20 for  the  runs
containing mixtures of organics.   Run-by-run  listings  of these  results are
given  in  Appendix A.   Graphical  displays of  these  results are  shown  in
Figure  52 through  55  for the  simple  mixture runs,  and  in  Figures  56
through  61  for  the  complex  and  surrogate  mixture  runs.    Concentration
versus  time profiles  for a typical simple mixture run  are shown  in  Figure
62, and similar  plots are shown  in Figures 63 through  75 for a variety of
representative  complex  mixture  simulations,   including several  multi-day
simulations of  ITC  and OTC  experiments.    [These plots  employ  the  same
notation  as  those  in the previous  section  —  the  "*" symbols  designate
experimental data,  solid lines  designate  data calculated using  the March
1988 RADM mechanism,  longer dashed lines  designate results  for the  RADM-P
mechanism,  and  shorter  dashed   lines  are   results  for   the  RADM-M
mechanism.]  The results discussed in this section are applicable only to
the  March 1988 RADM mechanism,  though  the  performance  of the  modified
mechanisms in  simulating  experiments not containing aromatics are similar.
     The  results for the  simple mixture runs show the mechanism tends to
underpredict the maximum ozone by  12£ on  the average  with a mean error of
34J.   The NOX  oxidation rates  are underpredicted by  a  similar amount on
the  average  in  these runs.   The  results  for the surrogate mixture runs
show  overprediction  of  the  maximum  ozone  by approximately  10J  on the
average with  a mean  error of approximately 2Q%.   The  mean bias and errors
in the NO oxidation  rates for these runs  are  slightly better than this on
the average.
     Since  this data  set  has  over  200  experiments  for  mixtures   that
include a wide range of organic compounds and range of hydrocarbon and NOX
concentrations,  the  evaluation  statistics are  fairly robust.   The  model
performance  statistics  for  ozone  and  NOX  oxidation  rate  are   good,
especially for the surrogate mixture runs.  The mean error  in  ozone  in the
surrogate mixture  simulations  is  less  than  that  for most of the single
organic compound  simulations —  particularly the aromatics  and alkanes.
This  suggests  that  there are some  compensating  errors  in  the mechanism,
which  is true of most photochemical reaction mechanisms.  (Also,  for runs
with   mixtures  of organics,  any error in  the mechanism for any particular
                                     195

-------
                              RADM Mechanism
                1200
                    0     200    400   600    800   1000   1200

                      Experimental Ozone Maximum (ppb)
Figure 52.  Calculated  vs  Experimental Maximum Ozone in the Simple
            Mixture  Runs for the RADM Mechanism.
                               RADM  Mechanism
                       20      50   100    200     500   1000

                      Experimental d([03j-[NO])/dt (ppb/hr)
 Figure 53   Calculated ys Experimental Rates of NO Oxidation  and  Ozone
              Formation in the Simple Mixture Runs  for the  RADM Mechanism.
                                       196

-------
                              RADM Mechanism
                 12
                 10
               in

               §  8
              a:
                          Bias in Maximum Ozone (%)
Figure 5^.  Histogram of Normalized Biases in Maximum Ozone  in  the  Simple

            Mixture Runs for the RADM Mechanism.
                              RADM Mechanism
                 100
                  50
V
c

a
o

E
D

E
X
O
             .£   -so

              
-------
                              RADM Mechanism
                    0    200    400   600   800   1000   1200
                      Experimental Ozone Maximum (ppb)
Figure 56.   Calculated vs Experimental Maximum Ozone in the Surrogate
             Mixture Runs for the RADM Mechanism.
                               RADM Mechanism
              _u
               O
              (J
20      50    100   200     500   1000
Experimental d([03]-[NO])/dt (ppb/min)
 Figure 57.  Calculated vs Experimental Rates of NO Oxidation and Ozone
             Formation in the Surrogate Mixture Runs for the RADM
             Mechanism.
                                      198

-------
                               RADM Mechanism
~HJ
40
OT
C
K 30
"o
i_


z
10


n







.


i } j










j£i






r~
H


*v>-
















—



/,
^



-_


x>
^
//

^







_^.
^
/ /


—



\^
^
\\
//
//
^







	

v\
A


















«
^
^\
^




i .
J"T

s\
\^
\
^->
[:• • j 3 Comp.
jivx::] 7 Comp.
JH:KKJ 8 Comp
^ UNC Mixes
p^Auto

|y^] Syn Fuels





                        Bias in  Maximum Ozone (%)
Figure 58.  Histogram  of Normali2ed Biases in Maximum Ozone  in  the
            Surrogate  Mixture Runs for the RADM Mechanism.
                               RADM  Mechanism
              
-------
                               RADM Mechanism



£S
0)
c
o
N
O
E
1
"x
O
2
C
•~~
C/)
a
in

&JU


150


100

50




-50

—inn
&
a
A ,

a
A
^° °
Q
' D*°.

*** * "*» £i^*^4if °
* ^. o "^ a?a * * ° °
4 o o o

•

. , . . 	 ,
3 Comp.
A

7 Comp.

8 Comp.
a
UNC Mixes
4
Auto
*
Syn. Fuel
•





                      0.1        0.2          0.5        1
                        Initial NOx Concentration (ppm)
Figure 60.  Normalized Bias in Maximum Ozone vs  Initial  NO  in the
            Surrogate Mixture Runs for the RADM  Mechanism.
                               RADM  Mechanism

•^- 150
0)
0
N 100
E
P 50
X
O
C
S -50
CD
—inn
a
* *
a
* a *
o
a
° *
A 4* A*«^° *
*A. *• f * 4* "I*lM^Vt • ;. v
O O Q
-

3 Corn p.
A
7 Comp.
8 Cornp.
o
UNC Mixes
Auto
*
Syn. Fuel
•


                    0.1   0.3     1     3     10    30    100
                        Initial ROG Concentration  (pprnC)
Figure 61.   Normalized Bias in Maximum Ozone vs Initial HC in the
             Surrogate Mixture Runs for the  RADM Mechanism.
                                     200

-------
            OZONE
                                             0.4-1
                                             0.3-
                                             0.2-
                                             0.1 -
                                                   NITRIC OXIDE
        240  360  480   600  720  840  960
                                             0.0-
                                                240  360  480   600  720  840  960
Q.
CL
o
£
UJ
o
z
o
CJ
      0.4-,
      0.3-
      0.2-
      0.1 -
            NITROGEN DIOXIDE
      0.0
        240  360  480  600   720  840  960
0.10-,



0.08-



0.06-



0.04-



0.02^
                                                    PAN
                                             0.00
                                                240  360  480  600   720  840  960
     0.10-,
     0.08-
     0.06-J
     0.04-J
     0.02-
            ETHENE
     0.00
            '   I    I   !  II     I1^
         240  360  4«0  600  720  840  960
                                  EU\PSED TIME  (minutes)
   Figure 62.   Experimental  and Calculated Concentration-Time  Plots for

                 Selected  Species in Simulations of  the  UNC Mixed  Alkene Run
                 OC1278R.
                                           201

-------
           OZONE
E
CL
O_
O
UJ
O
-z.
O
O
              60   120  180  240   300   360
                                            NITRIC OXIDE
                                                         120   180   240   300   360
     0.4-,
0.3-
           NITROGEN DIOXIDE
     0.0-
                                     0.15-,
                                            PAN
             60
120   180  240  300  360      0

              EUPSED  TIME (minutes)
                                                         r~  —
                                              60   120  180  240  300   360
  Figure 63.  Experimental and Calculated  Concentration-Time Plots  for
              Selected Species in Simulations  of the 7 Hydrocarbon
              Surrogate Run EC-237.
                                      20:

-------
E
Q.
Q.
o
LU
O

o
o
     0.5 -i
           OZONE
                                      NITRIC  OXIDE
                                       /    0.2
      0.3-,
      0.2-
   720    1440   2160   2880   3600


NITROGEN DIOXIDE
                                                     720    1440   2160   2880   3600
                                                  PAN
      0.0-
          I I I I I | I IT IT | I I I I I ]  I

              720    1440   2160   2880  3600
                                       i i i  | i  i i i i | i i i i i | i i i i i | i i i i i
                                         720   1440  2160  2880  3600
                                 ELAPSED TIME  (minutes)
  Figure  64.   Experimental and Calculated Concentration-Time  Plots for
               Selected Species in Simulations of the 8  Hydrocarbon
               Surrogate, Multi-Day  Run  ITC-630.

-------
Q.
CL
O
z
O
     0.5 -i
           OZONE
               720    1440    2160   2880
     0.0
           NITROGEN DIOXIDE
I I I I I  I I I IT i Ti i i i

    720    1440   2160   2880
                                           0.5
                                       NITRIC OXIDE
                                            0.0
                                     0     720    1440   2160   2880



                                       PAN
                                0.08-
                                           0.04-
                                           o.oo
                                                       i  i i i i I i i  i i i I i i i i i I  i i i
                                                     720     1440   2160   2880
                                 ELAPSED TIME  (minutes)
  Figure 65.  Experimental  and  Calculated Concentration-Time Plots for
               Selected  Species  in Simulations of the 8  Hydrocarbon

               Surrogate,  Multi-Day Run ITC-633.

-------
           OZONE
E
a
a.
g
t—
<
LU
a
•z.
o
o
                                           0.3 -,
                                            0.2-
                                            0.1 -
                                            NITRIC OXIDE
     0.4-,
0.3-
     0.2- !
         0       1440      2880



           NITROGEN DIOXIDE
                            4320
     0.0-
                                            0.0-
                                     0.15-,
1440     2880     4320
                                            PAN
         0      1440      2880
                                          0       1440


                           ELAPSED TIME (minutes)
                                                               2880
                 4320
 Figure  66.   Experimental and  Calculated Concentration-Time  Plots for

              Selected Species  in  Simulations of the 8 Hydrocarbon

              Surrogate, Multi-Day Run ITC-637.
                                       205

-------
           OZONE
E
Q_
Q_
LJ
O
-z.
O
o
                                           0.3-,
                        NITRIC OXIDE
     0.0
                480
960
                                1440           0       480

                                ELAPSED TIME (minutes)
  Figure 67.   Experimental and Calculated Concentration-Time Plots  for
              Selected Species in Simulations of the 8 Hydrocarbon
              Surrogate,  Multi-Day Run OTC-192B
                                       206

-------
            OZONE
Q.
Q.
o
<
Ld
CJ
-z.
o
0
                                            0.25-,



                                            0.20-



                                            0.15-



                                            0.10-



                                            0.05-
                                                   NITRIC OXIDE
                                            0.00-
         0   480  960  1440  1920  2400  2880


            NITROGEN DIOXIDE
                                            0.05 -,
480   960  1440  1920 2400 2880
                                                   PAN
                i i i j  I I I | !  I I I1

         0   480  950  1440  1920 2400 2880
                                                    i |  i i i |  i i i |  i i i | i  i i | i  i

                                                    480  960  1440  1920  2400  2880
                                 ELAPSED TIME (minutes)
  Figure 68.  Experimental and Calculated  Concentration-Time Plots for
               Selected  Species in Simulations of the 8  Hydrocarbon
               Surrogate,  Multi-Day Run  OTC-195A.

-------
           OZONE
Q.
a
g
<
UJ
O
-2L
O
O
                                                  NITRIC OXIDE
             480   960   1440  1920 2400 2880
           NITROGEN  DIOXIDE
                   :    j-pr-i-i-y-r Ty-r T-p
             480   960   1440 1920 2400 2880
                                            0.1 -
                                            0.0
                                           0.02-
                                           0.00
  480   960  1440 1920 2400 2880


PAN
                                                     | i i i  | i i i  | i i i  | i i i  | i i i
                                               0   480  960  1440  1920  2400  2880
                                 ELAPSED TIME  (minutes)
  Figure 69.  Experimental and Calculated  Concentration-Time Plots for
              Selected Species in Simulations of the 8 Hydrocarbon
              Surrogate, Multi-Day  Run  OTC-202B.
                                       208

-------
E
Q.
Q.
g

£
UJ
o
2:
o
o
     1.0-1
           OZONE
                            0.25-
                                   NITRIC  OXIDE
                                                 1080   1440   1800
 T
360
720    1080  1440  1800
               0     360    720


ELAPSED TIME (minutes)
                                                               1080   1440   1800
  Figure  70.   Experimental and Calculated Concentration-Time  Plots for

               Selected Species in Simulations  of  the  8  Hydrocarbon

               Surrogate, Multi-Day Run OTC-224B.
                                       209

-------
E
Q.
CL


Z
o
o:

z
LU
o

o
o
0.7 -,



0.6-



0.5-



0.4-



0.3-



0.2-



0.1 -



0.0
            OZONE
                                             NITRIC OXIDE
        300   420   540   660  780  900  1020
            NITROGEN DIOXIDE
                                            0.08^
                                            0.06-
                                            0.04-
                                            0.02-
                                          300   420  540  660   780   900  1020




                                             PAN
               I     I                .
        300   420   540   660  780  900  1020
                                            0.00»
                                                           .  1	1	r  .
                                          300   420  540   660   780   900  1020
                                 ELAPSED TIME (minutes)
  Figure 71.   Experimental and  Calculated Concentration-Time Plots  for

                Selected Species  in  Simulations of  the  UNC 3 Hydrocarbon

                Surrogate Run JN1483R.
                                        210

-------
           OZONE
                                                 NITRIC OXIDE
E
CL
Q.
g

<
UJ
o
z
o
o
300   420   540   660   780   900  1020



   NITROGEN DIOXIDE
                                             300   420  540  660   780  900  1020



                                                 PAN
              I     I     I    T
        300   420   540   660   780   900  1020
                                                         i     [^   i^   i     r
                                              300   420   540   660   780   900  1020
                                ELAPSED TIME  (minutes)
  Figure 72.   Experimental and Calculated Concentration-Time  Plots for

               Selected Species in Simulations of  the  UNC  Miscellaneous
               Surrogate Run AU2681B.
                                        211

-------
Q.
a
O
LU
O
z
O
O
     O.B-i
     0.6-
     0.4-
     0.2-
           OZONE
   NITRIC OXIDE
     0.0
        300   420  540  660  780  900  1020
           NITROGEN DIOXIDE
           I    ....
300  420  540  660  780  900  1020
   PAN
                  I     1     ;
        300   420  540  660  780  900  1020
               I     I     I
300  420   540  660  780  900  1020
                                ELAPSED TIME  (minutes)
  Figure 73.   Experimental and Calculated Concentration-Time  Plots for
               Selected Species in Simulations of the UNC Synurban
               Surrogate Run JN2685R.
                                       112

-------
           OZONE
                                          NITRIC OXIDE
E
Q.
Q.
g
LJ
O
•z.
O
O
300  420   540   660  780  900  1020
           NITROGEN DIOXIDE
                                              300  420  540   660  780  900  1020
                                          PAN
        300  420   540   660  780  900   1020      300  420  540   660  7SO  900  1020

                                ELAPSED TIME (minutes)
  Figure  74.   Experimental and Calculated  Concentration-Time Plots  for
               Selected Species in Simulations  of the UNC Synauto
               Surrogate Run AU0684R.
                                       213

-------
E
Q_
Q.
g

S
UJ
o
z
o
o
     1.0-1
           OZONE
          0.20-1
                 NITRIC OXIDE
        300   420  540  660   780  900  1020
           NITROGEN DIOXIDE
              300  420  540   660  780  900  1020
                 PAN
                  T    |^
        300  420  540  660  780   900  1020
              300   420  540  660   780  900  1020


EUPSED  TIME (minutes)
  Figure  75.   Experimental  and  Calculated Concentration-Time  Plots for

               Selected Species  in Simulations of the UNC  Auto Exhaust

               Run AU1183R.
                                        214

-------
component would have a  relatively  smaller  effect  than would be  the  case  in
a run  if that component  were  the only  reactive  organic present.)   There
does not  appear  to be any systematic  bias in the results with respect  to
the  initial  HC to  NOV  ratios,  initial  NOV  concentration, or  initial  HC
                      A                     X
concentrations.   This  is  reassuring,  but  is not unexpected,  considering
that  such a  systematic bias was  not  observed for  any of the groups  of
individual organics where one might expect such biases to be more evident.
     The  mechanism's  performance on multi-day  runs  is comparable  to that
for  single  day runs, especially when  good performance is achieved on  the
first  day of the simulation.   (See, for example, Figures 64 through 70.)
If  poor  performance  is  achieved  on  the  first  day,  then  obviously  the
performance  on the  subsequent  days would be even  worse,  since the  starting
conditions would be different.
     Figures  76  and 77  compare the calculated and  experimental  formalde-
hyde  maxima  for  the simple  and  complex  mixture runs.   The  performance
statistics  are  shown  in  Table 21.    These results  show that  the  RADM
mechanism underpredicts  the  maximum formaldehyde on  the average,  though
for  most types of runs  the  mean biases are  smaller  in  magnitude  than  the
mean  errors,  which are typically on the order of 40-50/t.  The  performance
on  formaldehyde  in mixture  runs   is  poorer  than  that  found  for  single
carbonyl  and most  types of alkene  runs,  but is better than the  results for
alkane and  aromatic runs, on  the  average.  The  figures clearly  show that
for  the runs with  mixtures as well as  those for the  single compounds there
is a large  amount  of scatter in the results  and  the mechanism  is  not able
to  reproduce the  formaldehyde  maxima  very accurately.  As already noted,
however,  the accuracy of the formaldehyde  measurements  is  questionable.
      Figures 78  and 79  compare the  calculated  and experimental PAN maxima
for  the  simple and  complex  mixture runs.  The performance statistics are
shown  in Table 22.   Except  for the UNC auto exhaust  runs (whose chemical
compositions are the most uncertain of all  the  runs),  the RADM mechanism
consistently overpredicts PAN  in  the mixture runs,  with average biases
ranging  from  12  to over  66^.  The mean  error  of  these predictions  are
typically on the order of HO-605&  or greater.   The positive bias and large
error  is consistent with the  results  for  single  organic compound runs.   A
positive  bias  is expected  since  the  PAN species in the mechanism is used
                                     215

-------
                                    Mechanism
c
,^
X
i

0.5
0.2
0.1
A * /
4 £/
A y/
xi *
\ s*>
              -  0.05 (•
              u
                0.01
                    0.01  0.02     0.05   0.1    0.2     0.5
                      Experimental  HCHO Maximum  (ppm)


Figure 76.  Calculated vs  Experimental Maximum Formaldehyde in the
            Simple Mixture Runs for the RADM Mechanism.
                               RADM Mechanism
— ' f _ *,••"' : ~" '"^
C »• • , ! A
5- f X i , ,
i <•"- i • -;'-
§ 0.3
w
X.
O
5 0.1
o
5
| 0.03
_o
3
Q 0.01
. :^-* . •
• . • «J*j4?*t* °£
9 . ••• 7V» •* "°
x°»° 8*°8 *
« / 0*0 »
/. * * « *
./ •
X i . 	 i . 	 i
8 Comp.
UNC Mixes
a
Auto
o
Syn. Fuel
*



Figure 77.
       0.01     0.03      0.1      0.3        1
         Experimental HCHO Maximum (ppm)

Calculated vs Experimental  Maximum Formaldehyde in the
Surrogate Mixture Runs  for  the RADM Mechanism.
                                      216

-------
                                 RADM Mechanism
              ,-,   0.3

               E
               Q_
               Q.

              ^   01

               E
               D

               E

               2  0.03
                  CO1
               —' 0 003
               O
              o
                 0.001
Figure  78.
        0.001    0.003    0.01     0.03     01     0.3

            Experimental PAN Maximum (ppm)



Calculated vs Experimental Maximum  PAN in the  Simple
Mixture Runs for  the  RADM Mechanism.
                   0.3
                                RADM Mechanism
                   C.i
              '    °-°3
               D
                  0,01
              CL
               0)
                 0.003
               D
              ±j
                 o.ooi -
                         A
                                               8 Comp.
                                                 «

                                              UNC Mixes
                                                          I
                                                            Syn. Fue!
                      0.001   0.003    0.01    003    0.1     0.3

                        Experimental PAN Maximum (ppm)
Figure  79.   Calculated vs Experimental Maximum  PAN in the  Surrogate
             Mixture Runs for the  RADM Mechanism.
                                      217

-------
to represent  PAN  plus numerous analogous  organic  nitrates.   However, the
magnitude of  the  bias,  especially in the surrogate mixture runs, suggests
the mechanism  overpredicts  the PAN  and  analogous  species concentrations.
This is of concern for  regional acid deposition modeling because the model
may underpredict  nitric  acid  in the first  few  100 km downwind of sources
(i.e., if NOX  is  oxidized  disproportionately to PAN rather than HNO^) and
overestimate  the  distances  for  which  nitrogen   species  are transported
(since PAN is the long-range transport product of NOX emissions).

5.6  Tests of the RADM VOC Aggregation Approach
     As  discussed in Section  U.3,  the  evaluation of  the RADM mechanism
involved processing  organic  input data  in the  simulations of the chamber
experiments in a  manner  analogous to the expected  procedure for  processing
of emissions  input  data for the  RADM  model.   This involves,  for example,
representing some compounds by  model species on a  mole-per-mole  basis, and
others using  "reactivity weighing," and not adjusting any parameters  in
the mechanism on a  run-to-run  basis.    This  is  the appropriate procedure
for evaluating the mechanism  to be used  in RADM,  but in general it  is not
the one  which would be expected to give  the  optimum fits  of the  model
simulations to the experimental  results.   An alternative approach,  which
is appropriate for  evaluations of detailed mechanisms  or  mechanisms  to  be
used  for assessment of  relative  reactivities,  is  to adjust  parameters  in
the mechanism based on  the mixtures being simulated.  This  latter  proce-
dure  should  in general   result  in  the  optimum performance of the model  in
simulating  the experimental  data;  provided, of  course,  that  errors  in
other  aspects of the mechanism  do not compensate  for errors  introduced
using a  more  approximate representation.
      Because  of  this, a comparison  of the  performance  of  the RADM mechan-
ism using  run-by-run adjustment  of parameters with  its performance  using
the standard  evaluation procedure  can give an indication of the level  of
inaccuracy  introduced  by  using  the RADM   VOC aggregation procedure.   To
evaluate this, a  separate  set of  simulations  of the  chamber  experiments
employing  organics not  explicitly  represented in the RADM  mechanism  was
carried  out  where  the  rate  constants  of the lumped  model  species  were
adjusted on  a run-to-run  basis  to  correspond to those  of  the  compounds
present  in  each  experiment.   The averages  of the  percent  errors and  biases
                                    218

-------
in predictions  of  ozone yields and NO oxidation  rates  for groups of runs
calculated  for  the RADM mechanism using  this  procedure are compared with
the  corresponding  results  for  the  standard  evaluation  in  Table  23.
Differences  in  ozone predictions for  individual  organic mixture runs are
shown  in  Figure 80, where the maximum  ozone yields calculated using this
modified  procedure are  plotted  against  the yields  calculated  using the
standard evaluation procedure.
     The  results shown  in  Table 23  and  Figure  80  indicate that the two
methods give very similar results  for  all  types  of mixture runs and for
most single component  runs.  The differences are  obviously  greater for the
single  component runs;  but,  except  for  the one  isobutene run  and (to  a
lesser  extent)  the   trimethylbenzene   runs,   the  statistics  for  the
simulations using  the  RADM VOC  aggregation method  are  either  about the
same or  somewhat better than the simulations where  the  rate constants are
adjusted.   Improvement of fits using the more  approximate  RADM VOC  aggre-
gation  method relative  to  those obtained with adjusted rate constants  is
clearly a case of partial cancellation of  errors.   These  results indicate
that  for   realistic  mixtures of organics  likely  to  be  encountered  in
regional  modeling applications,  any  inaccuracies introduced by  using  the
RADM  VOC aggregation  procedure  discussed  in  Section 3 are likely to  be
small  compared  to  inaccuracies  introduced  by  other uncertainties  in  the
mechanism.

5.7  Performance of the Recommended  Modified Mechanisms
     The  recommended modified versions of the RADM mechanism, discussed in
Section  2  of this report, were  also evaluated against the chamber data.
These  mechanisms are  the  RADM-M, which  incorporates all  of the modifica-
tions  we recommend as  a result  of this  evaluation  study,  and  the RADM-P,
which  primarily incorporates those  recommendations  involving only changes
in parameter values.   These mechanisms are  the same as the March 1988 RADM
mechanism  in their inorganic and  simple carbonyl  reaction  set,  and thus
the evaluation of the  RADM mechanism for the characterization and carbonyl
runs is  also applicable  to  these mechanisms.   However,   both  RADM-M and
RADM-P have  significant differences  in their  representation of aromatics,
and in addition the RADM-M has a more condensed  representation of the more
 rapidly  reacting  (generally  Ce+) alkanes.  The two mechanisms also have
                                     219

-------
Table 23.  Comparison of Average Normalized Biases and Errors in
           Simulations of Chamber Experiments using the RADM Mechanism
           with Run-to-Run Adjustment of Rate Constants with Simulations
           using the Standard RADM VOC Aggregation Procedure
Compound
or
Mixture
Propene

1-Butene

trans-2-Butene

Isobutene

1-Hexene

a-Pinene

Ethane

n-Butane

C4+ Br. Alkane

C5+ n-Alkane

Cycloalkane

Benzene

Toluene

1) Adjusted Rate Constants
2) Standard VOC Aggregation
Maximum Ozone (?)
Bias (sdev) Err.
4
5
5
-2
-52
-52
-12
-23
64
57
-12
-8
-17
-19
-1
0
-12
-17
49
42
34
33
22
22
24
24
( 19)
( 20)
( 54)
( 56)
( 38)
( 38)
( 0)
( 0)
( 26)
( 25)
( 36)
( 35)
( 0)
( 0)
( 59)
( 59)
( 58)
( 67)
( 34)
( 28)
( 0)
( 0)
( 16)
( 15)
( 43)
( 43)
15
15
45
45
54
54
12
23
64
57
29
28
17
19
46
46
44
53
49
42
34
33
22
22
33
33
d([03]-[NO])/dt (%)
Bias (sdev) Err.
6
6
9
0
-26
-25
10
-45
101
82
-10
4
7
6
18
19
-16
-22
31
24
5
3
28
36
6
7
( 18)
( 19)
( 32)
( 32)
( 26)
( 26)
( 0)
( 0)
{ 30)
( 25)
( 10)
( 19)
( 0)
( 0)
( 50)
( 50)
( 20)
( 24)
( 36)
( 37)
( 28)
( 29)
( 21)
( 19)
( 34)
( 34)
14
14
29
25
31
31
10
45
101
82
10
14
7
6
43
43
19
26
35
34
17
19
31
36
26
27
                                                               (continued)
                                    220

-------
Table 23 (continued) - 2
Compound
or
Mixture
Xylene

Mesitylene

Tetralin

Naphthalene

2,3-Dimethylnaph.

Simple Mixtures

Simple Surrogates

SAPRC 7-HC Surgs.

SAPRC 8-HC Surgs.

UNC Misc. Surgs.

UNC "Synurban"

UNC "Synauto"

UNC Auto Exhaust

Syn. Jet Fuel

Syn. Jet Exhaust

1 ) Adjusted Rate Constants
2) Standard VOC Aggregation
Maximum Ozone (%)
Bias (sdev) Err.
-9 i
-4 i
-21 i
-39 '
-2
-6
82
84
18
-6
-12
-12
10
7
1
0
12
9
14
17
36
39
22
23
10
8
10
14
-10
-11
( 30)
( 25)
( 30)
( 50)
( 56)
( 62)
( 44)
( 40)
( 15)
( 24)
( 45)
( 45)
( 28)
( 24)
( 11)
( 11)
( 25)
( 26)
( 39)
( 40)
( 30)
( 3D
( 17)
( 17)
( 29)
( 29)
( 11)
( 12)
( 8)
( 8)
19
18
30
48
41
43
82
84
18
17
34
34
17
15
8
8
21
20
27
27
36
39
23
23
19
19
11
14
10
11
d([03]-[MO])/dt (%)
Bias (sdev) Err.
-19 <
0 i
-60
-93
108
106
118
134
58
1
-13
-13
7
8
-9
-5
-1
-10
1
2
4
7
0
2
-3
-6
35
11
-15
-8
( 3D
( 35)
( 39)
( 32)
( 85)
( 85)
( 16)
( 13)
( 28)
( 24)
( 25)
( 26)
( 43)
( 38)
( 12)
( 12)
( 24)
( 23)
( 16)
( 15)
( 9)
( 8)
( 18)
( 18)
( 15)
( 15)
( 16)
( 16)
( 8)
( 4)
20
25
60
93
121
119
118
134
58
19
22
22
30
27
12
10
17
19
12
11
8
8
14
14
12
13
35
16
15
8
                                    221

-------
               C
               o
               u
               cr
               TJ
               O>
               C
                                          ccrd Methoc
Figure 80.
Plots of Maximum Ozone Yields in Simulations of Organic
Mixture Experiments Using the RADM Mechanism with Run-to-Run
Adjustment of Rate Constants Against Yields Calculated Using
the Standard RADM VOC Aggregation Procedure
                                    222

-------
small differences  in  their reactions for the akenes and the less reactive
Co+ alkanes.   The evaluation  calculations  for  these mechanisms were  thus
carried out  for  all  the single and mixed organic-NOx-air runs  except  those
involving only simple carbonyls.
     The  summary  statistics  for the  performance  of  the  RADM-M  and  the
RADM-P  mechanisms  in  simulating  maximum  ozone  yields  and  rates  of  NO
oxidation  and ozone  formation  are  given in Tables  24 and  25;  statistics
for predictions  of maximum formaldehyde are given in Tables 26 and  27; and
statistics  for predictions of maximum  PAN  are  given in Tables  28  and 29.
Table  30 shows  comparisons of selected summary  statistics for all  three
versions  of the mechanism which were evaluated in  this study.   Figures 81
and  82  show plots  of  calculated vs  experimental  maximum ozone  for the
RADM-M  and  RADM-P mechanisms for the  surrogate  mixture runs;  Figures 83
through 86  show  similar plots  for the alkane-NOx and the  aromatic-NOx
runs,  respectively.   Finally, Figures  23-25,  30-32,  37-42,  49 and  62-75
(presented  earlier)  give concentration-time  profiles  illustrating  the
performance of  these mechanisms  in  simulating  results of selected experi-
ments where  their predictions  differed from  the  evaluated  RADM  mechan-
 ism.   (Solid lines are  for the RADM mechanism, longer dashed lines are for
RADM-P, and shorter dashed lines are for RADM-M.)   Run-by-run summaries of
 the performances  of these mechanisms  in  simulating  maximum  ozone and NO
oxidation rates are  found in Appendix B for RADM-M and  in Appendix  C  for
 RADM-P.
      The  differences between  the   predictions of the mechanisms  for  the
 alkene-NOv  and  alkane-NOv runs  are relatively minor, as  expected due to
          X               A
 the relatively  small changes in the mechanisms  for these compounds.   The
 RADM-M mechanism  performs slightly  worse than  the other two for the  higher
 alkane-NO   runs  because of  its more  condensed  representation  of  these
 compounds.    However,  the  test calculations  discussed   in  Section  6.3
 indicate  that this  should have relatively  little  effect  in simulations of
 more realistic  mixtures of organics.
                                      223

-------
Table 24.  Performance Statistics on Maximum Ozone and NO Oxidation and
           Ozone Formation Rates in the Simulations of Organic-N0x-Air
           Runs Using the RADM-M Mechanism
Organic
or
Mixture
Ethene
Propene
1-Butene
trans-2-Butene
Isobutene
1-Hexene
Isoprene
a-Pinene
Ethane
n-Butane
C4+ Br. Alkanes
C5+ n- Alkanes
Cycloalkanes
Benzene
1)
2)
No.
Runs
14
14
49
49
11
11
4
4
1
1
4
4
13
13
4
4
1
1
27
27
7
7
14
14
4
4
6
6
Statistics for Maximum Ozone (ppb)
Statistics for d( [03l-[NO])/dt (ppb/min)
Experimental
Avg. (sdev)
0.91 (
5.23 (
0.61 (
3.62 {
0.42 (
2.86 (
0.28 {
6.17 (
0.90 (
8.84 (
0.34 (
1.57 (
0.72 (
5.66 (
0.35 (
0.97 <
0.24 1
1.32 !
0.31 <
1.53 I
0.28 I
0.93 <
0.20 i
1.15 i
0.05
0.92 *
0.29
3-39
0.20)
4.10)
0.24)
2.60)
0.26)
2.18)
0.17)
2.80)
0.00)
0.00)
: 0.25)
: 1.04)
; o
; 4
; o
; o
; o
; o
[ o
I o
[ o
[ o
( o
( o
( o
( o
( o
( 2
.25)
.06)
.11)
.43)
.00)
.00)
.27)
.98)
.18)
.46)
.15)
.75)
.00)
.12)
.14)
.30)
Absolute
Bias Err.
-0.09
-1.21
0.04
0.53
0.05
0.41
-0
-1
-0
-2
0
3
-0
0
-0
0
-0
0
-0
0
-0
-0
0
0
.08
.68
.17
.81
.28
.32
.23
.86
.03
.06
.04
.09
.01
.18
.05
.21
.16
.42
0.05
0.36
0.07
6.11
0.16
1.21
0.
0.
0.
0.
0.
1.
0.
2.
0.
3.
0.
1.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
09
76
14
96
09
76
17
81
28
32
24
81
08
15
04
09
12
72
10
23
16
56
0.05
0.41
0.07
6.11
— Normalized —
Bias (») Err. (%)
-16 (
-27 (
8 (
11 (
2 (
5 (
-53 (
-29 (
-21 <
-38 (
59 (
88 (
-42 <
-2 (
-8 {
2 (
-18 <
7 I
-10 <
12 I
-17 I
-25 i
58 i
38 i
85
25
20
92
30)
20)
20)
19)
55)
32)
38)
24)
: o)
0)
: 26)
: 26)
: 3D
; 36)
: 3*0
: 18)
: o)
; o)
[ 59)
! 49)
( 73)
( 25)
( 39)
( 43)
( 0)
( 41)
( 13)
( 23)
23
27
16
17
45
27
54
32
21
38
59
88
43
26
27
1.3
18
7
47
41
58
29
58
46
85
31
20
92
                                                                 (continued)
                                    224

-------
Table 24 (continued) - 2
Organic
or
Mixture
Toluene
Xylene
Mesitylene
Tetralin
Naphthalene
2 , 3-Dimethylnaph .
Simple Mixtures
Simple Surrogates
SAPRC 7-HC Surgs.
SAPRC 8-HC Surgs.
UNC Misc. Surgs.
UNC "Synurban"
UNC "Synauto"
UNC Auto Exhaust
Syn. Jet Fuel
Syn. Jet Exhaust
1)
2)
No.
Runs
20
20
10
10
8
6
5
5
5
5
4
4
40
39
36
36
11
11
82
82
21
21
11
11
15
15
28
28
8
8
4
Statistics for Maximum Ozone (ppb)
Statistics for d( [O^l-fNO] )/dt (ppb/min)
Experimental
Avg. (sdev)
0.34 (
3.07 (
0.44 (
5.39 (
0.52 (
9.50 (
0.29 (
1.89 (
0.20 (
1.59 (
0.32 (
2.48 (
0.49 (
4.10 (
0.37 (
2.81 (
0.60 (
7.74 <
0.45 I
2.81 I
0.50 I
1.32 I
0.37 i
1.09 i
0.62 i
1.81 <
0.57
1.68
0.75
3.47
0.78
6.66
0.15)
1.87)
0.19)
4.14)
0.26)
4.97)
0.23)
0.88)
0.08)
0.41)
0.04)
0.5D
0.27)
: 4.06)
: 0.15)
: 1.61)
; 0.17)
: 5.16)
; 0.27)
: LSD
[ 0.17)
( 0.40)
( 0.28)
( 0.47)
( 0.18)
( 0.63)
( 0.27)
( 0.61)
( 0.10)
( 0.74)
( 0.13)
( 3.48)
Absolute
Bias Err.
0.03
0.12
-0.01
0.67
-0.13
-4.66
0.23
25.41
0.33
19.56
0.08
3.90
-0.06
-0.89
0.01
1.08
0.00
-0.55
0.02
-0.16
0.02
-0.06
0.06
0.01
0.09
-0.05
0.04
-0.14
0.13
8.22
-0.07
0.06
0.06
0.62
0.05
0.96
0.14
4.66
0.23
25.56
0.33
19.56
0.08
3.90
0.12
1.11
0.05
1.32
0.04
0.72
0.08
0.54
0.10
0.15
0.07
0.06
0.11
0.30
0.10
0.21
0.13
8.22
0.08
0.33
— Normalized —
Bias (?) Err. (?)
13 (
7 (
0 (
17 (
-26 (
-57 (
57 (
125 (
93 (
170 (
21 (
79 (
-15 (
-15 <
8 (
28 (
0 <
-8 I
3 I
-5 i
11 \
-3 <
22 i
1 i
16
-2
1
-10
16
107
-9
3
3D
3D
16)
20)
19)
22)
50)
9D
37)
8)
17)
; 34)
: 43)
: 25)
: 24)
: 36)
: 12)
; 12)
; 26)
[ 24)
( 40)
( 15)
( 25)
( 9)
( 17)
( 18)
( 34)
( 15)
( 1D
( 14)
( 8)
( 5)
21
21
12
20
29
57
57
137
93
170
21
79
32
23
16
36
8
11
19
19
25
12
24
6
18
16
23
14
16
107
10
5
                                    225

-------
Table 25.  Performance Statistics on Maximum Ozone and NO Oxidation and
           Ozone Formation Rates in the Simulations of Organic-N0x-Air
           Runs Using the RADM-P Mechanism
Organic
or
Mixture
Ethene

Propene

1-Butene

trans-2-Butene

Isobutene

1-Hexene

Isoprene

a-Pinene

Ethane

n-Butane

C4+ Br. Alkane

C5+ n-Alkane

Cycloalkanes

Benzene

1)
2)
No.
Runs
14
14
49
49
11
11
4
4
V
1
4
4
13
13
4
4
1
1
27
27
7
7
14
14
4
4
6
6
Statistics for Maximum Ozone (ppb)
Statistics for d( [03]-[NO])/dt (ppb/min)
Experimental
Avg. (sdev)
0.91 (
5.23 (
0.61 (
3.62 (
0.42 (
2.86 (
0.28 (
6.17 (
0.90 (
8.84 (
0.34 (
1.57 (
0.72 (
5.66 <
0.35 I
0,97 <
0.24 I
1.32 I
0.31 i
1.53 i
0.28
0.93
0.20
1.15
0.05
0.92
0.29
3.39
0.
4.
0.
2.
0.
2.
20)
10)
24)
60)
26)
18)
0.17)
2.80)
0.00)
0.00)
: 0.25)
: i.
.04)
: 0.25)
[ 4.06)
: 0.11)
; 0.43)
[ 0
[ o
( 0
( 0
( 0
( 0
( 0
( o
( o
( o
( o
( 2
.00)
.00)
.27)
.98)
.18)
.46)
.15)
.75)
.00)
.12)
.14)
.30)
Absolute
Bias Err.
-0.
-1.
0.
09
21
04
0.53
0.05
0.41
-0.08
-1,
.68
-0.17
-2.80
0.28
3.33
-0
0
-0
0
-0
0
-0
0
-0
-0
0
0
0
0
0
6
.23
.86
.03
.06
.04
.10
.01
.21
.06
.18
.10
.27
.00
.03
.07
.20
0.16
1.21
0.09
0.76
0.14
0.96
0.09
1.76
0.17
2.80
0.28
3.33
0.24
1.81
0.08
0.15
0.04
0.10
0.12
0.74
0.09
0.20
0.10
0.40
0.02
0.21
0.07
6.20
Bias
-16
-27
8
11
2
5
-53
-29
-21
-38
59
88
-42
-2
-8
2
-18
7
-8
13
-19
-22
41
25
27
-1
19
93
Normalized —
(») Err. (J)
(
(
(
(
(
(
(
(
(
(
(
(
(
(
(
(
(
(
(
(
(
(
(
(
(
(
(
(
30)
20)
20)
19)
55)
32)
38)
24)
0)
0)
26)
26)
3D
36)
34)
18)
0)
0)
59)
49)
62)
24)
27)
36)
0)
3D
14)
24)
23
27
16
17
45
27
54
31
21
38
59
88
43
26
27
13
18
7
47
41
50
25
41
35
27
22
19
93
                                                                 (continued)
                                     226

-------
Table 25 (continued) - 2
Organic
or
Mixture
Toluene

Xylene

Mesitylene

Tetralin

Naphthalene

2,3-Diraethylnaph.

Simple Mixtures

Simple Surrogates

SAPRC 7-HC Surgs.

SAPRC 8-HC Surgs.

UNC Misc. Surgs.

UNC "Synurban"

UNC "Synauto"

UNC Auto Exhaust

Syn. Jet Fuel

Syn. Jet Exhaust

D
2)
No.
Runs
20
20
10
10
8
6
5
5
5
5
4
4
40
39
36
36
11
11
82
82
21
21
11
11
15
15
28
28
8
8
4
4
Statistics for Maximum Ozone (ppb)
Statistics for d( [03]-[NO])/dt (ppb/min)
Experimental
Avg. (sdev)
0.34 (
3.07 (
0.44 (
5.39 (
0.52 (
9.50 (
0.29 (
1.89 (
0.20 (
1.59 (
0.32 (
2.48 (
0.49 <
4.10 (
0.37 I
2.81 <
0.60 I
7.74 I
0.45 I
2.81 I
0.50 l
1.32 I
0.37 i
1.09 i
0.62 i
1.81 i
0.57
1.68
0.75
3.47
0.78
6.66
0.15)
1.87)
0.19)
4.14)
0.26)
4.97)
0.23)
; 0.88)
; 0.08)
: o.4D
: 0.04)
: o.5D
; 0.27)
: 4.06)
; 0.15)
: 1.6D
; 0.17)
: 5.16)
[ 0.27)
( 1.51)
( 0.17)
( 0.40)
( 0.28)
( 0.47)
( 0.18)
( 0.63)
( 0.27)
( 0.61)
( 0.10)
( 0.74)
( 0.13)
( 3.48)
Absolute
Bias Err.
0.05
0.19
0.02
0.61
-0.11
-4.63
0.23
25.16
0.34
19.27
0.10
3.90
-0.06
-0.87
0.02
1.05
-0.01
-0.46
0.03
-0.10
0.02
-0.06
0.07
0.02
0.10
-0.04
0.05
-0.13
0.12
5.54
-0.07
0.00
0.07
0.68
0.06
0.93
0.13
4.63
0.23
25.30
0.34
19.27
0.10
3.90
0.12
1.10
0.05
1.28
0.04
0.77
0.08
0.54
0.10
0.14
0.08'
0.06
0.12
0.29
0.10
0.21
0.12
5.54
0.08
0.34
— Normalized --
Bias (?) Err. (%)
17 (
10 (
5 (
17 (
-21 (
-55 (
56 (
125 (
93 (
170 (
26 (
79 (
-14 (
-14 (
9 (
28 (
-1 (
-6 (
5 (
-3 (
10 (
-3 (
24 (
2 (
18 (
-1 (
3 (
-9 (
15 (
87 (
-9 (
2 (
3D
3D
16)
21)
19)
22)
49)
9D
38)
8)
17)
32)
44)
25)
23)
36)
12)
13)
25)
25)
37)
15)
25)
9)
16)
18)
34)
15)
10)
17)
8)
5)
22
23
12
21
25
55
56
137
93
170
26
79
32
22
16
35
8
12
19
19
24
12
25
6
19
15
23
14
15
87
10
5
                                    227

-------
Table 26.  Performance Statistics on Maximum Formaldehyde in the
           Simulations of Organic-N0x-Air Runs Using the RADM-M
           Mechanism
Compound or
Mixture
Ethene
Propene
1-Butene
trans-2-Butene
Isobutene
1 -Hexene
Isoprene
a-Pinene
n-Butane
C4+ Br. Alkane
C5+ n-Alkane
Cycloalkane
Toluene
Xylene
Mesitylene
Naphthalene
Simple Mixtures
Simple Surrogates
SAPRC 7-HC Surgs.
SAPRC 8-HC Surgs.
UNC "Synurban"
UNC "Synauto"
UNC Auto Exhaust
Syn. Jet Fuel
No.
Runs
14
39
8
4
1
4
13
2
21
5
10
4
20
10
8
5
34
7
11
82
11
13
28
8
Experimental(a) Absolute(a)
Avg. (sdev) Bias Err.
0.86 (
0.36 (
0.32 (
0.08 (
0.64
0.33 (
0.30 (
0.08 (
0.04 <
0.03 (
0.01 (
0.07 (
0.09 <
0.09 I
0.09 <
0.01 I
0.25 I
0.13 I
0.38 I
0.18 I
0.12 I
0.16 i
0.18 i
0.05 i
0.41)
: 0.20)
0.27)
; 0.06)

: 0.27)
: 0.16)
: o.o4)
: 0.03)
: 0.02)
: O.OD
: 0.12)
; 0.09)
; 0.05)
; 0.04)
! O.OD
[ 0.20)
[ 0.05)
[ 0.23)
( 0.12)
( 0.04)
( 0.08)
( 0.10)
( 0.03)
-0.02
0.01
0.12
0.04
0.00
0.32
0.01
-0.02
0.01
-0.02
-0.01
-0.06
-0.03
-0.04
-0.03
0.05
0.02
-0.05
0.10
-0.02
-0.05
0.02
-0.10
-0.01
0.23
0.11
0.12
0.06
0.00
0.32
0.12
0.02
0.03
0.02
0.01
0.07
0.04
0.05
0.04
0.05
0.09
0.06
0.10
0.05
0.05
0.04
0.10
0.03
Normalized
Bias (%) Err. (%)
-9
7
24
59
-1
79
-6
-25
14
-57
-19
-31
-26
-45
-30
134
-14
-45
18
1
-54
14
-68
-13
( 41)
( 45)
( 19)
( 58)

( 34)
( 60)
( 12)
( 76)
( 78)
(101)
(135)
( 62)
( 68)
( 49)
( 43)
( 53)
( 45)
( 19)
( 48)
{ 18)
( 3D
( 3D
( 67)
34
34
25
70
1
79
46
25
65
87
89
105
48
67
44
134
45
55
22
35
54
25
70
53
 (a)  Concentrations  in ppm.
                                    228

-------
Table 27.  Performance Statistics on Maximum Formaldehyde in the Simulations
           of Organic-NOx-Air Runs Using the RADM-P Mechanism
Compound or
Mixture
Ethene
Propene
1 -Butene
trans-2-Butene
Isobutene
1-Hexene
Isoprene
a-Pinene
n-Butane
C4+ Br. Alkane
C5+ n-Alkane
Cycloalkane
Toluene
Xylene
Mesitylene
Naphthalene
Simple Mixtures
Simple Surrogates
SAPRC 7-HC Surgs.
SAPRC 8-HC Surgs.
UNC "Synurban"
UNC "Synauto"
UNC Auto Exhaust
Syn. Jet Fuel
No.
Runs
14
39
8
4
1
4
13
2
21
5
10
4
20
10
8
5
7
11
82
11
13
28
8
Experimental(a)
Avg. (sdev)
0.86 (
0.36 (
0.32 (
0.08 (
0.64
0.33 (
0.30 (
0.08 (
0.04 (
0.03 (
0.01 (
0.07 <
0.09 I
0.09 I
0.09 <
0.01 '
0.25
0.13
0.38
0.18
0.12
0.16
0.18
0.05
0.41)
0.20)
0.27)
0.06)

0.27)
0.16)
0.04)
0.03)
; 0.02)
; O.OD
; 0.12)
( 0.09)
( 0.05)
( 0.04)
( 0.01)
( 0.20)
( 0.05)
( 0.23)
( 0.12)
( 0.04)
( 0.08)
( 0.10)
( 0.03)
Absolute(a)
Bias Err.
-0.02
0.01
0.12
0.04
0.00
0.32
0.01
-0.02
0.01
-0.02
-0.01
-0.06
-0.05
-0.07
-0.06
0.01
0.02
-0.06
0.09
-0.03
-0.05
0.01
-0.10
-0.03
0.23
0.11
0.12
0.06
0.00
0.32
0.12
0.02
0.03
0.02
0.01
0.06
0.05
0.07
0.06
0.01
0.09
0.06
0.10
0.06
.0.05
0.03
0.10
0.03
Normalized
Bias (*) Err. (%)
-9 i
7
24
59
-1
79
-6
-25
14
-74
-35
-41
-90
-109
-101
64
-21
-50
17
-4
-61
9
-74
-75
( 41)
( 45)
( 19)
( 58)

( 34)
( 60)
( 12)
( 76)
( 70)
(106)
(128)
( 58)
( 56)
( 44)
( 75)
( 62)
( 44)
( 19)
( 48)
( 19)
( 3D
( 3D
( 52)
34
34
25
70
1
79
46
25
65
92
95
98
91
114
101
71
52
59
21
36
/• 4
61
24
75
75
  (a)   Concentrations in ppm.
                                      229

-------
Table 28.  Performance Statistics on Maximum PAN in the Simulations of
           Organic-N0x-Air Runs Using the RADM-M Mechanism
Compound or
Mixture
Propene
1-Butene
trans-2-Butene
Isobutene
1-Hexene
Isoprene
a-Pinene
n-Butane
C4+ Br. Alkane
C5+ n-Alkane
Cycloalkanes
Toluene
Xylene
Mesitylene
2 , 3-Dimethylnaph .
Simple Mixtures
Simple Surrogates
SAPRC 7-HC Surgs.
SAPRC 8-HC Surgs.
UNC Misc. Surgs.
UNC "Synurban"
UNC "Synauto"
UNC Auto Exhaust
Syn. Jet Exhaust
No.
Runs
46
7
14
1
3
12
2
21
6
11
4
20
10
8
4
38
32
11
68
12
10
15
28
4
Experimental(a)
Avg. (sdev)
0.112 (0.084)
0.039 (0.016)
0.080 (0.044)
0.137
0.003 (0.003)
0.133 (0.082)
0.022 (0.016)
0.296 (1.012)
0.226 (0.506)
0.247 (0.814)
0.002 (0.003)
0.051 (0.036)
0.125 (0.102)
0.407 (0.189)
0.046 (0.019)
0.081 (0.059)
0.048 (0.020)
0.095 (0.047)
0.044 (0.030)
0.055 (0.046)
0.024 (0.024)
0.054 (0.039)
0.047 (0.029)
0.080 (0.03D
Absolute(a)
Bias Err.
0.049 0.055
0.059 0.060
-0.004 0.014
0.140 0.140
0.319 0.319
-0.029 0.080
0.048 0.048
-0.259 0.293
-0.208 0.223
-0.243 0.247
-0.001 0.002
0.045 0.051
0 025 0.059
-0.230 0.230
0.095 0.095
0.023 0.037
0.017 0.020
0.054 0.054
0.042 0.042
0.011 0.019
0.010 0.010
0.018 0.030
0.000 0.011
0.117 0.117
Normalized
Bias (%) Err. (J)
31 ( 39)
70 ( 50)
-14 ( 24)
68
195 ( 4)
-19 ( 85)
100 ( 12)
10 (105)
-46 (119)
21 (129)
-112 ( 14)
66 ( 43)
35 ( 41)
-69 ( 27)
98 ( 28)
18 ( 58)
36 ( 51)
43 ( 12)
66 ( 35)
35 ( 64)
31 ( 28)
29 ( 45)
-13 ( 60)
87 ( 13)
36
72
21
68
195
61
100
90
99
101
112
71
46
69
98
51
41
43
66
48
37
41
41
87
 (a)   Concentrations in ppm.
                                     230

-------
Table 29.  Performance Statistics on Maximum PAN in the Simulations of
           Organic-NOx-Air Runs Using the RADM-P Mechanism
Compound or
Mixture
Propene
1-Butene
trans-2-Butene
Isobutene
1-Hexene
Isoprene
a-Pinene
n-Butane
C4+ Br. Alkane
C5+ n-Alkane
Cycloalkane
Toluene
Xylene
Mesitylene
2,3-Dimethylnaph.
Simple Mixtures
Simple Surrogates
SAPRC 7-HC Surgs.
SAPRC 8-HC Surgs.
UNC Misc. Surgs.
UNC "Synurban"
UNC "Synauto"
UNC Auto Exhaust
Syn. Jet Exhaust
No.
Runs
46
7
4
1
3
12
2
21
6
11
4
20
10
8
4
38
32
11
68
12
10
15
28
4
Experimental (a
Avg. (sdev)
0.142
0.039
0.080
0.137
0.003
0.133
0.022
0.296
0.226
0.247
0.002
0.051
0.125
0.407
0.046
0.081
0.048
0.095
0.044
0.055
0.024
0.054
0.047
0.080
(0.084)
(0.016)
(0.044)


(0.003)
(0,
(0,
(1
(0
(0
(0
(0
(0
(0
(0
(0
(0
(0
(0
(0
(0
(0
(0
(0
.082)
.016)
.012)
.506)
.814)
.003)
.036)
.102)
.189)
.019)
.059)
.020)
.047)
.030)
.046)
.024)
.039)
.029)
.031)
) Absolute(a)
Bias Err.
0.
0.
-0.
0.
0.
-0.
0.
-0.
-0.
-0.
-0.
-0.
-0.
049
059
004
140
319
029
048
259
206
243
002
027
073
-0.353
-0.007
0.008
0.002
0.009
0.032
-0.005
0.005
0.003
-0.015
0.097
0.055
0.060
0.014
0.140
0.319
0.080
0.048
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.293
.224
.248
.002
.035
.073
.353
.009
.027
.017
.019
.033
.023
,005
.027
.016
.097
Normalized
Bias (%) Err. (%)
31
70
-14
68
195
-19
100
11
-46
8
-165
-64
-64
-147
-16
0
9
13
54
5
15
10
-44
78
( 39)
( 50)
( 24)

( 4)
( 85)
( 12)
(105)
(124)
(132)
( 8)
( 62)
( 40)
( 19)
( 34)
( 55)
( 57)
( 21)
( 39)
( 66)
( 29)
( 47)
( 56)
( 14)
36
72
21
68
195
61
100
90
105
105
165
82
64
147
24
44
38
20
57
54
25
41
56
78
 (a)  Concentrations in ppm.
                                     231

-------
Table 30.  Comparisons of Average Normalized Biases and Errors in Maximum
           Ozone and NO Oxidation Rates in the Simulations of Organic-
           N0v-Air Runs Using the RADM, RADM-M,  and RADM-P Mechanisms.
             A
Organic or Mixture Mechanism
Ethene


Propene


1-Butene


trans-2-Butene


Isobutene


1-Hexene


Isoprene


a-Pinene


Ethane


n-Butane


C4+ Br. Alkane


RADM
RADM-M
RADM-P
RADM
RADM-M
RADM-P
RADM
RADM-M
RADM-P
RADM
RADM-M
RADM-P
RADM
RADM-M
RADM-P
RADM
RADM-M
RADM-P
RADM
RADM-M
RADM-P
RADM
RADM-M
RADM-P
RADM
RADM-M
RADM-P
RADM
RADM-M
RADM-P
RADM
RADM-M
RADM-P
Maximum Ozone (%)
Bias (sdev) Err.
-16 ( 30)
-16 ( 30)
-16 ( 30)
5 ( 20)
8 ( 20)
8 ( 20)
-2 ( 56)
2 ( 55)
2 ( 55)
-52 ( 38)
-53 ( 38)
-53 ( 38)
-23 ( 0)
-21 ( 0)
-21 ( 0)
57 ( 25)
59 ( 26)
59 ( 26)
-42 ( 3D
-42 ( 3D
-42 ( 3D
-8 ( 35)
-8 ( 34)
-8 ( 34)
-19 ( 0)
-18 ( 0)
-18 ( 0)
0 ( 59)
-10 ( 59)
-8 ( 59)
-17 ( 67)
-17 ( 73)
-19 ( 62)
23
23
23
15
16
16
45
45
45
54
54
54
23
21
21
57
59
59
43
43
43
28
27
27
19
18
18
46
47
47
53
58
50
d([03]-[NO])/dt (?)
Bias (sdev) Err.
-27 ( 20)
-27 ( 20)
-27 ( 20)
6 ( 19)
11 ( 19)
11 ( 19)
0 ( 32)
5 ( 32)
5 ( 32)
-25 ( 26)
-29 ( 24)
-29 ( 24)
-45 ( 0)
-38 ( 0)
-38 ( 0)
82 ( 25)
88 ( 26)
88 ( 26)
-2 ( 36)
-2 ( 36)
-2 ( 36)
4 ( 19)
2 ( 18)
2 ( 18)
6 ( 0)
7 ( 0)
7 ( 0)
19 ( 50)
12 ( 49)
13 ( 49)
-22 ( 24)
-25 ( 25)
-22 ( 24)
27
27
27
14
17
17
25
27
27
31
32
31
45
38
38
82
88
88
26
26
26
14
13
13
6
7
7
43
41
41
26
29
25
                                                               (continued)
                                    232

-------
Table 30 (continued) - 2
Organic or Mixture
C5+ n-Alkane


Cycloalkane


Benzene


Toluene


Xylene


Mesitylene


Tetralin


Naphthalene


2 , 3-D imethy Inaph .


Simple Mixtures


Simple Surrogates


Mechanism
RADM
RADM-M
RADM-P
RADM
RADM-M
RADM-P
RADM
RADM-M
RADM-P
RADM
RADM-M
RADM-P
RADM
RADM-M
RADM-P
RADM
RADM-M
RADM-P
RADM
RADM-M
RADM-P
RADM
RADM-M
RADM-P
RADM
RADM-M
RADM-P
RADM
RADM-M
RADM-P
RADM
RADM-M
RADM-P
Maximum Ozone
Bias (sdev)
42 ( 28)
58 ( 39)
41 ( 27)
33 ( 0)
85 ( 0)
27 ( 0)
22 ( 15)
20 ( 13)
19 ( 14)
24 ( 43)
13 ( 3D
17 ( 3D
-4 ( 25)
0 ( 16)
5 ( 16)
-39 ( 50)
-26 ( 19)
-21 ( 19)
-6 ( 62)
57 ( 50)
56 ( 49)
84 ( 40)
93 ( 37)
93 ( 38)
-6 ( 24)
21 ( 17)
26 ( 17)
-12 ( 45)
-15 ( 43)
-14 ( 44)
7 ( 24)
8 ( 24)
9 ( 23)
(I) d([03]-[NO])/dt (%)
Err. Bias (sdev) Err.
42
58
41
33
85
27
22
20
19
33
21
22
18
12
12
48
29
25
43
57
56
84
93
93
17
21
26
34
32
32
15
16
16
24 ( 37)
38 ( 43)
25 ( 36)
3 ( 29)
25 ( 41)
-1 ( 3D
36 ( 19)
92 ( 23)
93 ( 24)
7 ( 34)
7 ( 3D
10 ( 3D
0 ( 35)
17 ( 20)
17 ( 21)
-93 ( 32)
-57 ( 22)
-55 ( 22)
106 ( 85)
125 ( 9D
125 ( 9D
134 ( 13)
170 ( 8)
170 ( 8)
1 ( 24)
79 ( 34)
79 ( 32)
-13 ( 26)
-15 ( 25)
-14 ( 25)
8 ( 38)
28 ( 36)
28 ( 36)
34
46
35
19
31
22
36
92
93
27
21
23
25
20
21
93
57
55
119
137
137
134
170
170
19
79
79
22
23
22
27
36
35
                                                               (continued)
                                     233

-------
Table 30 (continued) - 3
Organic or Mixture
SAPRC 7-HC Surgs.
SAPRC 8-HC Surgs.
UNC Misc. Surgs.
UNC "Synurban"
UNC "Synauto"
UNC Auto Exhaust
Syn. Jet Fuel
Syn. Jet Exhaust
Mechanism
RADM
RADM-M
RADM-P
RADM
RADM-M
RADM-P
RADM
RADM-M
RADM-P
RADM
RADM-M
RADM-P
RADM
RADM-M
RADM-P
RADM
RADM-M
RADM-P
RADM
RADM-M
RADM-P
RADM
RADM-M
RADM-P
Maximum Ozone (%)
Bias (sdev) Err.
0 ( 11)
0 ( 12)
-1 ( 12)
9 ( 26)
3 ( 26)
5 ( 25)
17 ( 40)
11 ( 40)
10 ( 37)
39 ( 3D
22 ( 25)
24 ( 25)
23 ( 17)
16 ( 17)
18 ( 16)
8 ( 29)
1 ( 34)
3 ( 34)
14 ( 12)
16 ( 11)
15 ( 10)
-11 ( 8)
-9 ( 8)
-9 ( 8)
8
8
8
20
19
19
27
25
24
39
24
25
23
18
19
19
23
23
14
16
15
11
10
10
d([03]-[NO])/dt (%)
Bias (sdev) Err.
-5 ( 12)
-8 ( 12)
-6 ( 13)
-10 ( 23)
-5 ( 24)
-3 ( 25)
2 ( 15)
-3 ( 15)
-3 ( 15)
7 ( 8)
1 ( 9)
2 ( 9)
2 ( 18)
-2 ( 18)
-1 ( 18)
-6 ( 15)
-10 ( 15)
-9 ( 15)
11 ( 16)
107 ( 14)
87 ( 17)
-8 ( 4)
3 ( 5)
2 ( 5)
10
11
12
19
19
19
11
12
12
8
6
6
14
16
15
13
14
14
16
107
87
8
5
5
                                    234

-------
                              RADM—M Mechanism
              _o
              CL
              Q.
              X
              0
              0)
              c
              o
              N
                    0     200   400   600   800    1000   1200

                      Experimental Ozone Maximum (ppb)
Figure 81.   Calculated vs Experimental  Maximum Ozone in the  Surrogate
             Mixture Runs for  the  RADM-M Mechanism.
                              RADM— P  Mechanism
                                                          UNC Mixes
                                                          Syn. Fuel
                     0    200   4QO    600    800   1000   1200

                       Experimental Ozone Maximum  (ppb)
 Figure 82.  Calculated vs Experimental Maximum Ozone in the Surrogate
             Mixture Runs for the RADM-P Mechanism.
                                     235

-------
                             RADM-M Mechanism
                1000
              c  300 *-
              X
              o
              o
              NJ
              o
                 600 -
400 -
                    0     200    400    600    800   'COO

                    Experimental Ozone Maximum (ppo)
Figure 83.   Calculated vs Experimental Maximum Ozone in the Alkane-N0>
             Runs for the RADM-M Mechanism.
                              RADM-P Mechanism
                 1000
              c.
              Q.
              -  SOO
              x
              O
              0)
              c
              o
              N
              o
               o
              o
                  600 •
400 -
                  200
                     0     200    400    600    800    1000

                     Experimental Ozone Maximum (ppb)
Figure 84.  Calculated vs Experimental Maximum Ozone in the  Alkane-N0x
             Runs for the RADM-P  Mechanism.
                                     236

-------
                             RADM — M  Mechamism
                1000
             _Q
              CL
              CL
              X
              0
              0)
              c
              o
              N
              O
                 800
                 60C -
400 r
                 200 -
                    0      200     400     600  '   800     1000

                      Experimental Ozone Maximum (ppb)
Figure  85.   Calculated vs Experimental Maximum Ozone in the  Aromatic-
             NOV Runs for the RADM-M Mechanism.
               A
                              RADM-P  Mechanism
                                                          Mesitylene
                                                             A
                     0     200    400     600     800    1000

                       Experimental Ozone  Maximum (ppb)
 Figure 86.  Calculated vs Experimental Maximum Ozone in the Aromatic-
             NOV  Runs for the RADM-P Mechanism.
                A
                                      237

-------
     The greatest  differences between the March  1988 and the recommended
modified mechanisms concern the aromatic chemistry.   The  RADM-M and RADM-P
have significantly less scatter in the simulations of the toluene, xylene,
and  mesitylene-NOx  runs,  and  their  biases  in  the  simulations  of  the
toluene  and  the mesitylene  runs are also  less.    The  apparently better
biases of  the  RADM mechanism relative to  RADM-M  and RADM-P (Table 30)  in
the simulations  of ozone  yields and  NO oxidation rates in the xylene  runs
are due  in  large  part to cancellation of  large and opposite  errors  in
simulations  of  xylene runs  carried  out  in  the  ITC compared  to  those
carried  out  in other chambers.   (See  Figure 40 for the  relative  perform-
ances of these mechanisms in simulating  an ITC xylene run, and Appendix A
for a  run-by-run  listing  showing the poor  performance  of the unmodified
mechanism  in  simulating the  xylene  runs.)   It is interesting to note  that
the differences between  the mechanisms are the least  in simulations  of
runs  in  the UNC chamber   and  the greatest in simulations  of  runs in  the
SAPRC ITC  (see Figures 39  through 41 for examples).
     Although  the recommended  modified mechanisms  perform significantly
better  in the  simulations of runs  employing alkylbenzenes,  they  perform
significantly  worse  in  simulations  of   runs  employing  benzene  or  the
naphthalenes,  where  they   consistently overpredict  reactivity.   The  over-
prediction of  reactivities of these  compounds is  due to  the fact  that they
are  represented in  the  RADM mechanisms  by  model   species  (TOL  and  XYL)
which are  optimized  to simulate toluene and xylenes.  Toluene and xylenes
are mechanistically  much  more reactive than benzene and  naphthalenes (see
Carter  et al.,  1987 for  a  discussion of  this).    The  fact  that  benzene
reacts slower  than TOL is corrected  for using  "reactivity weighing," but,
this  does not  correct  for   differences  in  mechanistic  aspects  of  reac-
tivity.    The  naphthalenes react more  rapidly than  XYL, and yet  produce
less ozone (Carter et al., 1987).  Note that all of the  runs with benzene
and the  naphthalenes were carried out in the SAPRC  ITC  (see  Appendices A-
C),   where  the  unmodified  mechanism  significantly   underpredicts  the
reactivities  of  alkylbenzene-NOx runs.   (Figure  40 is typical in  this
regard.)   This  underprediction  for  alkylbenzenes  in  the  ITC  tends  to
"correct"  for  the inappropriate use of TOL to represent  benzene  and XYL to
represent  naphthalenes, as shown, for example, in Figure 37 in  the case of
benzene.    Modifying   the   TOL  and  XYL   mechanism   so  they   perform
                                    238

-------
satisfactorily in  simulations  of ITC toluene  and  xylene  runs thus causes
the mechanism  to  be too reactive in the  simulations  of the ITC runs with
the mechanistically  less reactive aromatics.   This modification is still
appropriate, however,  because  alkylbenzenes are relatively more important
than benzene or the naphthalenes in the ambient atmosphere.
     With  two  exceptions,  the  RADM-M   and   RADM-P   mechanisms   perform
comparably or  better  than the unmodified RADM mechanism in simulations  of
runs employing organic mixtures.  The major exception  concerns  the  simula-
tions  of the  "synthetic Jet fuel"  runs,  which were  ITC  runs  containing
relatively  large  amounts of naphthalenes.  The  other exception concerns
the "simple  surrogate" runs.  This  series is dominated by  the  large number
of "mini-surrogate" runs carried  out in the SAPRC  ITC  in  1982-83 (Atkinson
et  al.,  1983),  and  all versions  of  the  RADM mechanism, and the SAPRC
mechanism  as well (Carter,   1988), tend to .consistently overpredict the  NO
oxidation  rate.    However,  in  most cases  the statistical  measures  for
performance  on the runs employing mixtures  is  not  significantly different
for the three  versions of the RADM mechanism.
     The RADM-P and  RADM-M mechanisms  are comparable  to the  unmodified
version in the predictions  of formaldehyde.  The  mechanisms  have  greater
differences  in  their predictions  of  PAN, with  the  RADM-M  predicting
slightly higher,  and  RADM-P  predicting  slightly  lower,   yields   of  this
product for most, but not all, types of  runs.  The higher PAN predictions
for  RADM-M are probably mostly  due  to  the  lumping of  the RADM's  aromatic
PAN-analog (TPAN) with  PAN  in the  RADM-M  mechanism.   The causes  for  the
different PAN predictions  for RADM-P  relative  to RADM are  less  obvious,
but  must be  due  primarily  to changes  in the aromatics mechanisms,  since
that  is the area  where the  two mechanisms differ the most.
      Comparison   of  the ozone   predictions of  the  modified  mechanisms
relative to the unmodified  version  is  shown in Figures 87 and 88.  These
give  plots of ozone yields  calculated using RADM-M and RADM-P, respective-
ly,  against  those calculated  using  the  RADM  mechanism   for  all of  the
chamber runs employing  organic mixtures.   It  can  be  seen  that while there
are  many experiments  where  the mechanisms give the same ozone predictions,
there  are  cases  where the modified  mechanisms  predict  less  ozone,  and
practically no  cases where  they   tend  to predict  more.    The   tendency
towards predicting less ozone in simulating these  chamber  experiments  with
                                     239

-------
                               RADM-M vs. RADM
                         0.2     C.
-------
organic  mixtures  is slightly  greater  with RADM-M  than  RADM-P,   though
except  for  one  run this  difference  is slight.   This  suggests  that  the
recommended  modifications  on  the  average  result  in  a  slightly  less
reactive mechanism with respect to ozone  formation.    Differences  between
these mechanisms in simulations more  directly  representative of regional
modeling applications are discussed in Section 6.
                                     211

-------
            6.  EVALUATION OF ALTERNATIVE CHEMICAL ASSUMPTIONS

     An important  task in the overall  evaluation  of a chemical mechanism
is investigation of alternative versions of  the mechanism that  incorporate
differing  chemical  approximations  or  degrees of  chemical  fidelity.    In
this study, we  developed  a series of 90 test problems representing a wide
range  of  chemical conditions  likely to be  encountered  in regional model
simulations and used  them to assess the effects of  selected modifications
of  the mechanism on  model predictions using  the RADM  mechanism.   These
included  calculations  to  examine  effects  of  using  varying  levels   of
details   in   representing  peroxy   radical   reactions   in   the   mechanism,
calculations  to determine the effects of using a more condensed represen-
tation of the reactions of the higher alkanes, and  calculations to deter-
mine  the  differences  in  the predictions  of  the  recommended modified
versions  of  the RADM  mechanism  relative  to the  March  1988 version.   The
results of these tests are  discussed in this  section.   Tests  of  several
other  mechanistic options were carried  out during the  course of this  study
and  are  referred to  where appropriate in other  portions  of this  report.
The  test  problems  developed  in this study  can  serve as  a basis  for  further
studies  of the  effects  of  alternative chemical  assumptions   on  regional
model  predictions and  the future development of more condensed or  effi-
cient  mechanisms for  use  in such models.

6.1  Test Problems for Sensitivity Analysis
     A set of  test  problems  that  span  the  conditions  of importance in
regional   acid  deposition modeling is  an  essential  component  to  any
investigation of  sensitivities  of mechanism  predictions  to  alternative
formulations.  If the set of conditions is not sufficiently comprehensive,
then conditions where a   set  of  alternative assumptions may  be important
may not  be represented, resulting in the possibility of misleading conclu-
sions  regarding  the  sensitivity  of model  predictions  to these  assump-
tions.   Several  aspects   of  regional modeling must be considered  in  the
selection of test problems.   First, regional  modeling  involves transport
of pollutants  over  long   distances,  so it  is  essential to use multi-day
test problems.   Second, the vertical domain of the  regional models extends
from  the earth's surface to  the  tropopause.    Atmospheric   temperature,
                                     242

-------
pressure, solar radiation  intensity,  and the radiation's spectral distri-
bution  vary  significantly  from the  lower  to the  upper  troposphere.   The
test conditions must address  these variations.  Third, the  regional models
are designed  to simulate the chemical and physical evolution of emissions
as they  are  transported  from source  areas to remote areas.  Therefore,  it
is important  for  the  test conditions to span  the range  of pollutant  con-
centrations  encountered  in  both  urban  and  rural  regions.   This  can  be
accomplished  by varying  the VOC and NOV concentrations, VOC-to-NOv ratio,
                                       A                          A
and VOC composition.
     The  ideal  framework in which to test alternative chemical mechanisms
is  in  a  three-dimensional   transport  model  such as  RADM  using  actual
meteorological  conditions.    However,  this  approach  is  not  practical
because  of the large  computing resources needed  for the  transport  model
and the large effort  required to both install an  alternative mechanism  in
a  transport  model and develop  mechanism specific  speciated VOC emissions
inventories.   Even  if  this  were practical,  the  fact  that the meteoro-
logical  inputs have only  been developed  for  a  small  number of episodes
limits  the utility of  this approach.
     An  alternative approach adopted here is to  use a simple photochemical
box model and exercise the model  for a large number of cases that span the
expected  range  of conditions encountered in  actual regional modeling.   The
box modeling approach can sometimes  produce unrealistic  pollutant  concen-
trations  because  of its simplistic  treatment  of  transport.   Nevertheless,
it is  well suited to  investigating the relative  differences in  alternative
chemical   mechanisms.   In   fact,   most  mechanism  comparisons   have   been
performed with photochemical  box models  (Jeffries et al.,  1981;  Dunker,
1984;  Leone  and  Seinfeld,  1985;  Schafer and  Seinfeld,   1985;  Stockwell,
1986;  Hough,  1987;  Milford,  1988;   Stockwell  and  Lurmann,  1989;  Dodge,
1989).
     The matrix of the  test  problems selected for the analysis  involves a
relatively small  number of  environmental  conditions  and  a large number of
chemical conditions.   The  rationale  for this  approach is that  alternative
mechanisms are expected to  respond  similarly to changes  in environmental
conditions and differently  to variations in  the  chemical mixtures.   The
three  sets  of  environmental conditions considered are shown in Table 31.
These   constant  temperature  cases   include  summer conditions at  surface
                                     243

-------
Table 31.  Environmental Conditions Used in Test Problems for Sensitivity
           Calculations
Parameter
Simulation Duration
Start Time (LDT) (a)
End Time on Last Day (LDT)
Summer
Surface

0500
2000
Number of Days in Simulation (b) 2-5
Light Conditions (c)
Date
Latitude (deg)
Elevation (km)
Meteorological Conditions
Temperature (K)
Pressure (atm)
Rel. Humidity (%)
Water Vapor (ppm)

Inversion Height Schedule


Background Pollutant
Concentrations (ppb) (e)
Oo
CO
CH4
Daytime Dry Deposition
Velocities (cm/sec):
Q
so2
Sulfate
HpOp
N02

June 21
40
0

298
1
50
15,600
LDT (m)
(d) 0539 800
1116 1116
1501 1200


20-40 (f)
200
1800


0.6
0.8
0.4
1.6
0.4
PAN and Organic Nitrates (ONIT) 0.4
Organic peroxides (OP1
Organic Acids (ORA1 and
HNO,
and OP2) 0.8
ORA2) 0.8
3.0
Winter
Surface

0500
2000
3-5

December 21
40
0

273
1
50
3,000
LDT (m)
0805 800
1133 958
1352 1000


20
200
1800


0.3
0.4
0.2
0.8
0.2
0.2
0.4
0.4
1.5
Summer
Aloft

0500
2000
3-5

June 21
40
6

248
0.5
10
160

(Constant)




0
200
1800


0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
                                                              (continued)
                                    244

-------
Table 31 (continued) - 2
 (a)  LDT = Local Daylight Time.

 (b)  Number of days of simulation depended on chemical conditions.  See
      Table 34.

 (c)  Solar flux data output by the RADM radiation module was provided
      by Stockwell (private communication, 1988).  The data were given in
      terms of actinic fluxes for 93 wavelength  intervals between 280 and
      735 nm for zenith angles of 0, 10, 20, 30, 40, 50, 60, 70, 78, and
      86.  These fluxes are available  in computer readable form from the
      authors upon request.  The zenith angle dependencies of these
      actinic fluxes for  the various wavelength  intervals were fit using a
      parameterized function similar to that used by Carter et al. (1986),
      which was then used to calculate the actinic flux at any given
      zenith angle in the simulation.

 (d)  Inversion height is constant at  initial value shown at beginning
      of the day, then changed linearly with time to the values at the
      times indicated.  The inversion  height was constant at the final
      value for the remainder of the day.  The  inversion height schedule
      is based on a simplified approximation of  the "characteristic
      curve" used in the  OZIPM software (Gipson, 1984).  The final
      inversion heights were determined to obtain the desired amount of
      dilution for each simulated day.

 (e)  "Background" species concentrations were  added to the initial
      pollutant concentrations in the  simulations, and  (for the ground
      level simulations)  are entrained into  the  air mass as the inversion
      height increases.

 (f)  20 ppb background 0^ was used  in the "rural" summer surface
      simulations with initial NCL levels of 0.5 ppb.   40 ppb background
      ozone was used  in all the other  summer surface simulations.
                                    245

-------
elevation, winter conditions  at surface elevation,  and  summer conditions
at 6 km  above  the  surface.  The solar  actinic  fluxes used in the simula-
tions  of these conditions  were obtained  from  the  RADM  radiation module
(Stockwell,  personal  communication,   1988).    Solar radiation   for  the
simulations is based on the diurnal  cycle at MO  degrees  latitude in the
summer and winter solstices.   A small amount of  dilution  (25% per day  in
the winter case and  50?  per day in the summer case) with relatively clean
background air  is  included  along  with very  approximate  dry deposition
velocities to  inhibit  build-up of unrealistically  high  concentrations.
Only daytime  dry deposition  is included  in  the  simulations.   Nighttime
deposition is assumed to be negligible.
     All  the  winter  surface  and  summer  aloft simulations  were  "static"
with respect  to  pollutant injection.   All pollutants were  assumed to  be
present  initially,  with no subsequent  injection  of pollutants other  than
background  species   entrained  from   aloft   as   the  inversion  height
increased.   "Static" and  "dynamic" simulations were  carried  out for the
summer surface  conditions.   In the  "dynamic"  simulations,  no pollutants
were present initially  (other  than  the  background  species), and pollutants
were  injected  at a  constant  flux (moles  per  unit  area)  throughout the
simulation.   Thus,   four  different types  of  environmental conditions are
used in  the simulations:

    (1)  Summer Surface - Static (designated "Summer Surface")
    (2)  Summer Surface - Dynamic (designated "Summer Dynamic")
    (3)  Winter Surface - Static (designated "Winter Surface")
    (4)  Sunnier Aloft   - Static (designated "Summer  Aloft")

     The chemical conditions  for these runs can be described  as  "rural"  or
"urban," with  the  former having the  relatively  low  pollutant  levels
characteristic  of   rural  conditions  and   the  latter  having  the  higher
pollutant  levels more typical of urban areas.   For each type of  chemical
condition,  the total  NOX and  reactive organic  (ROG)  levels were  varied
over  relatively  wide ranges,  and  several  different  mixtures  representing
ROG pollutants were employed.  The ROG mixtures employed  in  these simula-
tions  were  chosen   to  represent  typical  ambient  mixtures  of  varying

                                    2U6

-------
composition and  reactivity,  with  three different compositions used in the
urban simulations and two used in the rural runs.
     The three  urban  ROG composition profiles are based on 6-9 AM ambient
urban samples  collected  by EPA/AREAL (under Bill Lonneman's direction)  in
41  U.S.  cities  during  1984-1986.    These  hydrocarbon  data  have  been
thoroughly  analyzed  by Jeffries et  al.  (1989c)  to derive the recommended
carbon  fractions for  the "Carbon Bond"  and  the "CAL"  mechanisms.   The
specific  profiles used  in  these  simulations  are the average of the 773
samples  collected  in  all  of the  cities  (referred  to  as  the  "allcity"
average), the  average  of the 8 samples collected in Washington,  DC, during
1986  and the  average of  9  samples  collected  in  Beaumont,  Texas during
1984.   The latter  two were  selected because  their  composition and  reac-
tivity  are  most  different  from the  average.   The Beaumont profiles have
unusually  high  alkane fractions  and low  aromatic  and  C3+  alkene  frac-
tions.   The Washington DC profiles have  unusually high levels  of aromatics
and  low  levels  of  alkanes.   This  data  base  does  not contain  aldehyde
measurements,  and for all three profiles we  followed  the standard  assump-
tion  employed  by the EPA and  its contractors  (e.g., EPA,  1984;  Lurmann  et
al.,  1987;  Jeffries et al.,  1989c)  that  urban  ROG profiles consist  of  5%
aldehydes  (on  a  carbon basis).   This is based  on an  upper  limit estimate
derived from  data collected in Los  Angeles  (Grosjean,  1983).   It  was
assumed that  these urban  aldehydes consist of 60%  formaldehyde and  40?
acetaldehyde  (on a carbon  basis  - as did Lurmann et  al.,  1987),  which  is
slightly different than  the 40£ formaldehyde-60/& acetaldehyde  assumption
employed by Jeffries et al.  (1989c).  Both assumptions are consistent with
the  available  data.  These three urban ROG surrogates are similar to those
used in  the  mechanism  comparison  study  recently  carried  out by  Dodge
(1989).
      The ROG  profiles  employed in  the  rural  simulations are based  on  53
morning samples collected above the mixed layer by aircraft in four cities
(Westburg  and  MacGregor,  1987).    These measurements  include  aldehydes.
These data have  also  been  extensively analyzed  by Jeffries et al. (1989c)
for  determination  of typical  ROG  concentrations aloft.   One of  the   two
rural profiles is based on  the average composition of these data.  This is
referred to  as  the  "aloft" surrogate.    The   other  rural  surrogate  is
designed to  represent  air  masses  with significant  biogenic hydrocarbon
                                     247

-------
input.  It has 46%  isoprene  (on  a carbon basis), the same aldehyde  levels
as the aircraft data,  and the remaining hydrocarbons apportioned based on
the aircraft data.   This is referred to as the "aloft+isoprene" surrogate.
     The detailed compositions used  for these five surrogate mixtures are
given in Table 32.  These were  taken from the detailed data sets provided
by  Jeffries  (private  communication,  1988).    As  is  the case  with most
ambient  ROG  samples,   numerous   assumptions  were  needed  to  assign the
unidentified and ambiguously  classified carbon into known species.   These
detailed profiles  were  converted to  profiles  of specific  compounds (or
groups of compounds assumed  to  have the same reactivity) based on  (1) the
assumptions  employed   by  Jeffries   et  al.  (1989c)   concerning  ambiguous
classes of  olefins and aromatics  and concerning the  classifications of
species such as alkynes and  styrenes, and (2) assuming the  "unknown" data
were  hydrocarbons with the same distribution of species as  the  identified
species in  the profiles.    These individual species  were then  converted
into  species  in  the RADM mechanism  using  the same set of procedures used
in processing organic  emissions  input into the  RADM-II  model,  discussed in
Section 3.   The  distribution of RADM-II species for  these five  surrogates
is given in Table 33.   Note  that in many cases  the molar  concentrations of
the  lumped RADM  species  do not  add  up  exactly to the totals for  the
species they represent because  of the use of "reactivity  weighting" in  the
RADM  emissions processing  system  (see  Section 3).
      Table  34 summarizes the  conditions of  all the  test  calculations,
including  the environmental  conditions,  the duration  of the  simulation,
the  ROG  surrogate mixture,  and  the ROG and  NOX levels employed  for each
run.   The  minimum duration of  the  simulations was two days.   The maximum
duration was  either five days  or  one day  more than needed  for  complete
consumption of  the  NOV, whichever was less.   With the exception  of a  few
                      A
simulations  at  very low  ROG/NOX ratios,  the simulations were  carried  out
for  a sufficient amount of  time that the effects of  alternative  assump-
tions concerning  the  radical-radical reactions occurring in  the absence of
NOX   could  be adequately tested.    For  the  rural  simulations,   the  ROG
concentrations ranged  from 10 to 100 ppbC, the NOX concentrations from 0.5
to  5 ppb,  and the  initial ROG-to-NOx ratio from  2 to  200.   For the urban
simulations,  the  ROG  concentration  ranged from 200 to 2000  ppbC,  the  NOX
concentration  from  20  to 200 ppb, and the initial ROG-to-NOx ratio from  2
                                    248

-------
Table 32.  Compositions of ROG Surrogate Mixtures Used in Test
           Calculations
Compound and i
Emissions Group (a)
Ethane
Propane
Alkanes (0.25-0.50) (b)
n- Butane
Isobutane
Neopentane
Unspeciated C4-C5 Alkanes
Alkanes (0.50-1.00)
n-Pentane
Iso-Pentane
Cyclopentane
n-Hexane
2-Methyl Pentane
3-Methylpentane
2 , 3-Dimethyl Butane
Branched C6 Alkanes
Alkanes (1.00-2.00)
Cyclohexane
C6 Cycloalkanes
n-Heptane
Branched C7 Alkanes
C7 Cycloalkanes
n-Octane
Branched C8 Alkanes
C8 Cycloalkanes
n-Nonane
Branched C9 Alkanes
n-Decane
Branched C10 Alkanes
Alkanes (>2.00)
Branched C11 alkanes
Branched C12 Alkanes
Branched C13 Alkanes
Composition (ppb Species/ppmC ROG Surrogate)
Allcity
Average
20.78
15.47

16.39
6.91
0.39
0.01

6.81
14.09
0.63
2.94
3.82
2.74
1.01
0.89

0.84
1.66
1.12
5.17
0.84
0.59
4.89
0.10
0.46
1.92
0.63
1.38

0.55
0.13
0.00
Washing-
ton DC
24.33
11.35

19.92
5.78
0.20
0.00

5.37
16.31
0.50
1.98
3.51
3.03
1.17
1.84

0.27
1.43
0.89
4.46
0.58
0.58
4.69
0.05
0.42
1.82
0.90
0.96

1.09
0.18
0.00
Beaumont
TX
38.15
27.40

24.62
12.49
0.57
0.00

8.50
14.48
1.11
4.03
4.36
3.00
1.08
0.72

1.38
2.15
0.95
3.65
1.17
0.39
3.54
0.10
0.33
1.71
0.33
1.26

0.20
0.11
0.00
Aloft
111.06
48.47

23.64
12.69
0.00
0.00

7.14
11.59
2.29
2.02
2.44
1.32
0.57
0.00

0.45
0.86
0.00
2.57
0.00
0.35
0.90
0.00
0.00
0.00
0.00
0.00

0.00
0.00
0.00
Aloft
+ Isop.
54.42
23.91

11.70
6.28
0.00
0.00

3.54
5.75
1.14
1.00
1.21
0.66
0.28
0.00

0.22
0.43
0.00
1.28
0.00
0.17
0.45
0.00
0.00
0.00
0.00
0.00

0.00
0.00
0.00
                                                                (continued)
                                     249

-------
Table 32 (continued) - 2
Compound and
Emissions Group (a)
Ethene
Propene
Other Terminal Monoalkenes
1-Butene
1-Pentene
1-Hexene
2-Methyl-1-Butene
C4 Terminal Alkanes
C5 Terminal Alkanes
C6 Terminal Alkanes
C7 Terminal Alkanes
C8 Terminal Alkanes
C9 Terminal Alkanes
C10 Terminal Alkanes
C11 Terminal Alkanes
Internal Alkenes, Dienes,
cis-2-Butene
trans-2-Butene
1,3-Butadiene
2-Methyl-2-Butene
C5 Terminal Alkenes
C6 Terminal Alkenes
C7 Terminal Alkenes
C8 Terminal Alkenes
a-Pinene
Benzene
Aromatics (<2.00)
Toluene
Ethyl Benzene
Isopropyl Benzene
n-Propyl Benzene
Monoalkyl Benzenes
Composition (ppb Species/ppmC ROG Surrogate)
Allcity Washing-
Average ton DC
19.18
4.93

2.49
0.54
0.21
0.82
1.72
1.30
0.51
1.14
0.40
0.71
0.03
0.40
and Pinenes
0.79
1.02
0.67
0.11
1.99
0.93
0.10
0.02
0.61
3.43

8.19
1.23
0.14
0.28
1.67
16.33
3.25

2.56
0.43
0.13
0.81
0.25
1.42
0.31
0.82
0.40
0.19
0.15
0.01

0.64
0.74
0.26
0.00
1.25
0.61
0.05
0.00
0.18
3.58

11.18
1.61
0.07
0.39
2.05
Beaumont
TX
23.27
5.16

1.78
0.53
0.09
0.72
1.33
0.98
0.47
0.71
0.20
0.41
0.01
0.76

0.81
1.09
0.44
0.03
1.70
0.72
0.04
0.03
0.28
2.31

4.09
0.71
0.05
0.18
0.93
Aloft
14.98
2.57

0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00

0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
4.07

4.87
0.88
0.00
0.00
0.00
Aloft
+ Isop.
7.34
1.26

0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00

0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
2.02

2.42
0.44
0.00
0.00
0.00
                                                              (continued)
                                    250

-------
Table 32 (continued) - 3
Compound and
Emissions Group (a)
Composition (ppb Species/ppmC ROG Surrogate)

 Allcity  Washing- Beaumont  Aloft    Aloft
 Average  ton DC      TX             + Isop.
Aromatics (>2.00)
m-Xylene
o-Xylene
p-Xylene
Dialkyl Benzenes
1 ,2,3-Trimethyl Benzene
Trialkyl Benzenes
Tetraalkyl Benzenes
Acetylene
Formaldehyde
Acetaldehyde
Isoprene
1.93
1.53
1.93
4.80
2.88
1.92
0.18
11.43
30.00
10.00
0.67
2.54
2.37
2.54
5.22
4.03
1.89
0.11
10.55
30.00
10.00
0.61
1.08
0.72
1.08
2.98
1.41
1.44
0.10
4.82
30.00
10.00
0.78
0.72
2.03
0.72
0.00
0.00
0.00
0.00
8.99
74.86
10.75
3.42
0.36
1.01
0.36
0.00
0.00
0.00
0.00
4.41
74.86
10.75
92.75
  (a)  "Emissions  Group"  is  the  group  the  individual  compounds are lumped
      into during processing  of emissions data  (see  Section 3).  If a
      compound is given  by  itself  (e.g.,  ethane),  it is  in an emissions
      group by itself.   If  more than  one  compound  (or group of compounds)
      is lumped into an  emissions  group,  [e.g.,  n-butane,  isobutane, etc.,
      in group "Alkanes  (0.25-0.50)"],  the individual compounds or groups
      are listed  below  the  emissions  group name.  .

  (b)  Quantities  in parentheses refer to  range  of  OH radical rate constants
      (in units of 10  ppm    min  )  for compounds  in the emissions group.
                                     251

-------
Table 33.  Compositions of ROG Surrogates Used in the Test Calculations
           for the RADM Mechanism
RADM Description
Species (a)
Name
ETH
HC3
HC5
HC8
OL2
OLT
OLI
ISO
TOL
XYL
HCHO
ALD
Ethane
Alkanes (0.25-0.50)
Alkanes (0.50-1.00)
Alkanes (>1.00)
Ethene
Terminal Alkenes
Internal Alkenes
Isoprene
Aromatics (<2.00)
Aromatics (>2.00)
Formaldehyde
Higher Aldehydes
Composition (ppb RADM Species/ppmC ROG)
Allcity
Average
20.78
35.54
29.60
18.64
19.18
15.21
6.24
0.67
12.54
15.16
30.00
10.00
Washing-
ton DC
24.33
35.22
30.31
16.95
16.33
10.73
3-72
0.61
16.37
18.71
30.00
10.00
Beaumont Aloft
TX
38.15
53-33
33.53
15.81
23.27
13.15
5.13
0.78
6.65
8.81
30.00
10.00
111.1
64.63
24.61
4.68
14.98
2.65
0.0
3.42
6.97
3.48
74.86
10.75
Aloft
+ I sop.
54.42
31.93
12.21
2.33
7.34
1.27
0.0
92.75
3.47
1.73
74.86
10.75
 (a)  Quantities  in parentheses for lumped higher alkanes and. aromatics
      refer to OH radical rate constant ranges  in units of  10  ppm
      min"1.
                                    252

-------
Table 34.  Summary of Conditions of Test Calculations Used for
           Sensitivity Studies
No.    Environmental       ROG
         Condition      Surrogate
Reactant Levels (a)

ROG
                                                 NOX   ROG/NOX
 No.
Days
                              Rural Simulations
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Summer
Summer
Summer
Summer
Summer
Summer
Summer
Summer
Summer
Winter
Winter
Winter
Summer
Summer
Summer
Summer
Summer
Summer
Summer
Summer
Summer
Summer
Summer
Summer
Summer
Summer
Summer
Winter
Winter
Winter
Summer
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Aloft
Aloft
Aloft
Dynamic
Dynamic
Dynamic
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Aloft
Summer Aloft
Summer Aloft
Aloft
Aloft
Aloft
Aloft
Aloft
Aloft
Aloft
Aloft
Aloft
Aloft
Aloft
Aloft
Aloft
Aloft
Aloft
Aloft
Aloft
Aloft
Aloft-i-lsoprene
Aloft+Isoprene
Aloft+Isoprene
Aloft+Isoprene
Aloft+Isoprene
Aloft+Isoprene
Aloft-i-Isoprene
Aloft+Isoprene
Aloft+Isoprene
Aloft+Isoprene
Aloft+Isoprene
Aloft+Isoprene
Aloft+Isoprene
Aloft+Isoprene
Aloft+Isoprene
10
33
100
10
33
100
10
33
100
10
33
100
10
33
100
10
33
100
10
33
100
10
33
100
10
33
100
10
33
100
10
33
100
0.5
0.5
0.5
1.7
1.7
1.7
5.0
5.0
5.0
1.7
1.7
1.7
1.7
1.7
1.7
1.7
1.7
1.7
0.5
0.5
0.5
1.7
1.7
1.7
5.0
5.0
5.0
1.7
1.7
1.7
1.7
1.7
1.7
20
66
200
6
20
60
2
7
20
6
20
60
6
20
60
6
20
60
20
66
200
6
20
60
2
7
20
6
20
60
6
20
60
3
3
3
2
2
2
3
3
3
5
4
3
5
4
3
3
4
4
5
5
5
2
2
2
3
3
3
5
4
3
5
4
3
                                                                (continued)
                                     253

-------
Table 34 (continued) - 2
No.

34
35
36
Environmental
Condition

Summer
Summer
Summer

Dynamic
Dynamic
Dynamic
ROG
Surrogate

Aloft+Isoprene
Aloft+Isoprene
Aloft+Isoprene
Reactant Levels (a)

ROG
10
33
100

NOX
1.7
1.7
1.7

ROG/ NO v
A
6
20
60
No.
Days

3
4
4
                              Urban Simulations
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
Summer
Summer
Summer
Summer
Summer
Summer
Summer
Summer
Summer
Winter
Winter
Winter
Summer
Summer
Summer
Summer
Summer
Summer
Summer
Summer
Summer
Summer
Summer
Summer
Summer
Summer
Summer
Winter
Winter
Winter
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Aloft
Aloft
Aloft
Dynamic
Dynamic
Dynamic
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Allcity Avg
Allcity Avg
Allcity Avg
Allcity Avg
Allcity Avg
Allcity Avg
Allcity Avg
Allcity Avg
Allcity Avg
Allcity Avg
Allcity Avg
Allcity Avg
Allcity Avg
Allcity Avg
Allcity Avg
Allcity Avg
Allcity Avg
Allcity Avg
Washington
Washington
Washington
Washington
Washington
Washington
Washington
Washington
Washington
Washington
Washington
Washington
,
.
m
.
•
^
»
•
•

.
•
•
.
•

.
•
DC
DC
DC
DC
DC
DC
DC
DC
DC
DC
DC
DC
200
667
2000
200
667
2000
200
667
2000
200
667
2000
200
667
2000
200
667
2000
200
667
2000
200
667
2000
200
667
2000
200
667
2000
67.
67.
67.
20.
20.
20.
100.
200.
200.
67.
67.
67.
67.
67.
67.
67.
67.
67.
67.
67.
67.
20.
20.
20.
100.
200.
200.
67.
67.
67.
3
10
30
10
33
100
2
3
10
3
10
30
3
10
30
3
10
30
3
10
30
10
33
100
2
3
10
3
10
30
                                                                  3
                                                                  3
                                                                  2
                                                                  3
                                                                  2
                                                                  2
                                                                  5
                                                                  3
                                                                  3

                                                                  5
                                                                  5
                                                                  4

                                                                  5
                                                                  4
                                                                  3

                                                                  5
                                                                  5
                                                                  5

                                                                  3
                                                                  3
                                                                  2
                                                                  3
                                                                  2
                                                                  2
                                                                  5
                                                                  3
                                                                  3

                                                                  5
                                                                  5
                                                                  4

                                                              (continued)
                                   254

-------
Table 34 (continued) - 3
No.
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
Environmental
Condition
Summer Aloft
Summer Aloft
Summer Aloft
Summer Dynamic
Summer Dynamic
Summer Dynamic
Summer Surface
Summer Surface
Summer Surface
Summer Surface
Summer Surface
Summer Surface
Summer Surface
Summer Surface
Summer Surface
Winter Surface
Winter Surface
Winter Surface
Summer Aloft
Summer Aloft
Summer Aloft
Summer Dynamic
Summer Dynamic
Summer Dynamic
ROG
Surrogate
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Washington DC
Beaumont TX
Beaumont TX
Beaumont TX
Beaumont TX
Beaumont TX
Beaumont TX
Beaumont TX
Beaumont TX
Beaumont TX
Beaumont TX
Beaumont TX
Beaumont TX
Beaumont TX
Beaumont TX
Beaumont TX
Beaumont TX
Beaumont TX
Beaumont TX
Reactant Levels (a)
ROG
200
667
2000
200
667
2000
200
667
2000
200
667
2000
200
667
2000
200
667
2000
200
667
2000
200
667
2000
NOX
67.
67.
67.
67.
67.
67.
67.
67.
67.
20.
20.
20.
100.
200.
200.
67.
67.
67.
67.
67.
67.
67.
67.
67.
ROG/NOX
3
10
30
3
10
30
3
10
30
10
33
100
2
3
10
3
10
30
3
10
30
3
10
30
No.
Days
5
4
3
5
5
5
3
3
2
3
2
2
5
3
3
5
5
4
5
4
3
5
5
5
  (a)  For the static calculations ("summer surface," "winter surface,"
       and "summer aloft"), the reactant levels reTer to initial
       concentrations in ppbC (for ROG) or ppb (for NOX), excluding
       background species.  For the "summer dynamic" concentrations, the
       reactant levels refer to the increase in concentrations of ROG (in
       ppbC) or NOX  (in ppb) that would result following one day of
       continuous emissions if there were no chemical reaction and the
       inversion height remained constant at its initial level of 800
       meters.
                                     255

-------
to  100.   These  combinations  of ROG  and NOV adequately  bracket probable
                                            A
atmospheric  concentrations in  most  rural  and  urban  areas.    This set
consists of  a total  of 90 simulations, 36  representing  rural conditions
and 54 representing urban pollution.

6.2  Presentation of Sensitivity Test Results
     The results  of  test calculations using alternative mechanisms can be
presented in  a  number of ways.   Many  sensitivity studies have used  plots
to compare concentrations  of  species  of interest.  While this is a useful
approach when a  relatively small number of  test  problems are used,  it is
less practical  when  the number  of  test cases  is relatively large,  as is
the case in  this  study.  This is analogous  in some  ways to  the problems of
summarizing  the  performance of  a mechanism in simulating the results of  a
large number  of  chamber  experiments.   Therefore,  in presenting the  results
of the  test  calculations,  we  use an  approach  similar to that used in the
evaluation against the  chamber  data  — the results  are  presented in the
form  of tabulations  of statistical  measures  of the comparisons  of the
mechanisms,  with  examples  given for results  of individual calculations
which are of  interest.
     Statistics  are  tabulated  for  the following  species,  which  are of
known or potential interest  in regional acid deposition  model simulations
(RADM  model   species name  —  used  in  the tabulations  —  are  given in
parentheses):

     1.  Ozone (03);
     2.  Hydrogen Peroxide (H202);
     3.  Sulfuric Acid or Sulfates  (SULF);
     4.  Nitric  Acid  (HN03);
     5.  Formic  Acid  (ORA1);
     6.  Higher  Organic  Acids (ORA2);
     7.  Methyl  Hydroperoxide (OP1);
     8.  Higher  Organic  Hydroperoxides  (OP2);
     9.  Peroxy  Acetic  Acid (PAA);  and
     10.  PAN and Higher  Peroxy  Nitrates  (PAN).
                                    256

-------
Given  the results of  model calculations using two  different models  (one
designated the  "standard"  calculation, and the other — generally having  a
modification to  the  "standard"  mechanism whose effect  is being examined —
designated the  "test"  calculation),  the  following statistics  are tabulated
for each  of these species  for each of the test problems:

     1.   Maximum concentration  for the standard calculation;
     2.   Percent change  in the  maximum concentration in the test
          calculation relative to the standard calculation;
     3.   Average concentration  for the "standard" calculation;
     4.   Percent change  in the  average concentration in the test
          calculation relative to the standard calculation;
     5.   Percent change  in the  average concentration,  relative to the
          maximum average concentration for,the set (see below); and
     6.   The "Percent  fit" statistic, calculated as discussed below.

     The  statistics  are calculated  for  the  entire time duration of these
multi-day calculations.    The   first  four  of the  above  statistics  are
derived   in  the usual manner.    A  positive number  means  that  the  test
calculation  predicts  a  larger  maximum  or  average  than  the  standard
calculation, while  a negative number means it predicts a  lower value.  The
tabulations  indicate which of the mechanisms is  used as the "standard" and
which  is  used  as  the "test."
     The   fifth  statistic,  designated   "%  /Max"  on  the  tabulations,  is
calculated for a given species  in a  given  test calculation as follows:

                           (Average Cone.,      (Average Cone.
                             Test  Calc.)    -    Standard Calc.)
      "%  /Max"   =  100 x
                          (Maximum of the Averages  for  the  Group)

 The purpose of this statistic is to obtain an indication of the changes in
 absolute  concentrations  of  the  species for  an  entire  set  of  calcula-
 tions.   These are normalized to the maximum value for the set to allow for
 meaningful comparisons with  predictions obtained for other  species whose
 concentration may be  quite different.   Thus,  if there is  a large positive
 or negative magnitude in  this statistic,  it  means that  the two models are
                                     257

-------
significantly different  in  their calculations of  this  species,  and it  is
not  a  case  of  a large  relative  difference in  the  calculation  of  an
insignificant  concentration.     For   the  purpose  of  calculating   this
statistic,  the   calculations   are  grouped   into   "rural"  and   "urban"
problems.  This  is to  account  for the fact that concentration regimes  for
at  least some of the species  of  interest  might  be different  for these
types of problems.
     The  sixth  statistic,   designated  as  "%  Fit"  on  the   tabulations,
measures  the  overall  differences  in  the  simulations of  the  species on  a
point-by-point basis,  paired in time.   It is calculated by
                    i |T, - s,i
      '% Fit"  =  	
                 0.5 x Z (Ti + S^

where T* and S. are  the  concentrations  of  the  species  for  the  test  and  the
standard calculation,  respectively, at  time  "i,"  and the summations  are
over all the times where data  are saved in the calculations, which  is once
every hour.  The  purpose of this statistic  is  to make evident differences
in curve shapes which might still have similar  average or maximum values.

6.3  Treatment of Peroxy Radical  Reactions
     Alkyl  and acyl  peroxy radicals  are   important  intermediates   in  the
gas-phase atmospheric  reactions of organic  compounds.  In  the presence of
NOX  their major  fates are reaction with NO  and (for acyl  peroxy radicals)
with  NOo,   and these  reactions  are  important  in  affecting   ozone  forma-
tion.   When NOX  levels are low, their major fates are  reaction with H02 or
with other  peroxy  radicals,  and these reactions are important  in affecting
predictions of  H202  and  organic  hydroperoxy  species.    Therefore,  the
appropriate representation of the  reactions of these  species  is important
in  acid deposition models, where accurate  predictions of  both oxidant  and
hydroperoxide  formation are essential.   This  includes appropriate repre-
sentations  of  peroxy-peroxy radical reactions,  since  sensitivity calcula-
tions we have  carried out previously (Carter et al.,   1986) and more recent
calculations  carried  out  by  Dodge  (1989)  indicate  that  they  cannot be
ignored.    However,   because   of  the  relatively large number  of  organic
peroxy  species which  must be  included in  even  relatively condensed gas-

                                    258

-------
phase atmospheric representations, almost all current mechanisms represent
these peroxy-peroxy  radical  reactions in an approximate manner.  The RADM
mechanism is no exception in this regard.
     As  indicated  in  Section  2,  the  RADM mechanism  has  a relatively
detailed  representation  of  peroxy  radicals,   using  separate  species   to
represent peroxy  radicals formed from most  of  the organic model species,
and representing  explicitly  their reactions with  NO,  NOp  (for acylperoxy
radicals), HC^,  methyl peroxy radicals  (M02),  and acetyl  peroxy radicals
(AC03).   However,  this  mechanism  ignores  all  the reactions  of organic
peroxy  radicals  with  organic  peroxy radicals  other  than M02  or  AC03.
Stockwell  (private   communication)   conducted   test  calculations  which
indicated that  using  only  reactions  with M02  and AC03  should be  suffi-
cient.   However,  the description and  results of these tests have not been
included  with  the available documentation  of  the mechanism.   In view  of
the   importance   of  peroxide  and   hydroperoxide  predictions  in  acid
deposition models,  we  felt  that an  independent  test of  the validity  of
this  approximation  in  the RADM  mechanism would be a  useful component  of
this evaluation study.
     On  the  other hand,  it  may  well  be that  the  RADM  mechanism is more
detailed  in  its  representation  of  peroxy  radical   reactions   than   is
necessary.   The latest Carbon Bond (Gery et al.,  1988) and SAPRC  (Carter,
1989a) mechanisms,  and the chemical mechanism for the ADOM model  (Lurmann
and  Karamchandani,   1987)  employ  much  more condensed  representations  of
these reactions.   These mechanisms do not have separate reactions for  the
many  kinds of peroxy radical  species;  instead they use a  limited  number of
chemical  "operators"  to represent the net effects  of  their reactions.   The
representations  used in these other  current mechanisms are  too  condensed
to  permit the  separate  representation  of  methyl hydroperoxide  and  the
higher  hydroperoxides  which  are utilized  in  the  RADM  mechanism.   Since
Stockwell  (private  communication) considers  separate  representation  of
methyl  and  the higher  hydroperoxides to be important  for RADM,  the more
condensed approaches which require that these be  lumped  together were  not
examined  in  this  study.
      However,  an earlier  version of the  SAPRC mechanism  (Carter et al.,
1986)  uses  a  less   condensed  representation  which  allows  for   separate
representation  of the two types  of hydroperoxide  species,  and also  allows
                                    259

-------
for  all  the  possible organic  peroxy  +  organic  peroxy reactions  to be
represented without  using many more  reactions  than used for this purpose
in the RADM mechanism.   A detailed discussion  of this approach is  beyond
the  scope  of  this  report  (see  Carter  et  al.,   1986),  but  briefly it
involves using  chemical  operators to  account  for total  organic  peroxy
("R02")  and  acylperoxy  ("RC03")   radical  concentrations,  and  then  using
these  to calculate  rates of  reactions of the  individual  organic  peroxy
radicals with  themselves  and  all  other  organic  peroxy radicals.    This
allows for  all  the  possible radical combinations  to  be represented,  but
requires  using  the  same  rate constants   for  all  these  reactions.   By
contrast,  the  RADM  mechanism  is  more  approximate  in the  sense  that it
represents only a  subset  of  the possible peroxy +  peroxy  reactions,  but is
more detailed in  the sense that it uses separate  estimated  rate constants
for  the  various  individual  reactions which  it  does  include.  Therefore,
the  effect  of modifying  the RADM  mechanism  so  that  it incorporates  the
SAPRC approach was also evaluated in  this study.
     In  order to  test  effects of  alternative  approximation  methods of
representing the  organic  peroxy radical reactions,  it  is necessary to have
a  detailed  representation  against  which to   compare   them.    For  this
purpose,  Stockwell  (private communication,  1988)  provided  SAPRC  with  a
"detailed  radical" version of  the  RADM mechanism,  which included all the
possible combination  reactions  of  the  organic  peroxy  and  acylperoxy
species  in  the  RADM  mechanism.   Other than the  additional reactions  of the
organic  peroxy  and acylperoxy  radical  species  with peroxy  radical species
other  than  M02  and AC03  (and the inclusion of the OLN  + M02 and OLN + AC03
reactions),  this  mechanism was exactly the same as the  March 1988 version
of the  RADM  mechanism which  was  evaluated  in  this study.  The  approach
Stockwell  used  to  estimate the  rate constants  and  products for these
additional  reactions was  consistent  with  the approach used  for the reac-
tions  in the March  1988  mechanism.   Almost all of these additional reac-
tions  involve estimates  which  are  highly  uncertain,  and in  some  cases we
may  have made different  estimates were we required  to independently derive
such a mechanism.   However,  because  this  detailed mechanism and  the March
 1988 RADM mechanism are  consistent with each other in their approaches to
the  chemistry,  any  differences  between  their predictions  would  reflect
only the effects of the approximations in the RADM mechanism regarding the
                                    260

-------
peroxy  radical  reactions.    Therefore,  the  detailed  radical  mechanism
provided  by  Stockwell  was  used without  modification as  the  standard  in
this evaluation.
     The  organic  peroxy  +  organic   radical   reactions   in  Stockwell's
detailed  radical  mechanism,  which  was  used as  the  standard mechanism  in
the tests of  peroxy  radical  representations,  are listed in Table 35.   All
the other reactions  are the same as those  listed in Table 2 in Section  2
of  this  report,  excluding  those  indicated  as  "Organic   Peroxy  +  M02,"
"Organic  Peroxy  +  AC03,"  "Other  Organic  Peroxy  +  Peroxy"  and   "XN02"
reactions.
     The reactions of  organic species in  the version of  the RADM mechanism
employing the  approach of the  1986 SAPRC mechanism  (Carter et al.,  1986)
are given  in  Table  36.  All organic  reactions  have  to be shown  because
there  are modifications  made  throughout . this  portion  of the mechanism.
The inorganic  reactions are  the same as those used for all versions of the
RADM mechanism and are given  in Table 2.   Except for the reactions  involv-
ing organic  peroxy radicals,  this mechanism is the  same as  the March 1988
RADM mechanism.   The major difference between this and the March  1988 RADM
mechanism is  that the  chemical  operators "[R02]" and "[RC03]" were  added
to  represent  total  amounts of organic  peroxy  and acylperoxy  radicals,
respectively,  and instead of  reacting  with M02  and AC03,  the  individual
organic  peroxy radicals  are  represented  as reacting with  these species.
Reactions  forming  individual   peroxy  radicals  were  represented  as  also
forming  [R02].   Reactions  forming acetyl  peroxy radicals (AC03 or  TC03)
were represented  as  also forming [RC03].   Reactions  of [RC03]  with  N02 and
[R02]  and [RC03J  with NO,  H02,  [R02],  and [RCQ3]  were added.   The rate
constants used for the [R02]  +  [R02],   [R02] +  [RC03], and [RC03] -»• [RC03]
reactions were the same as those used  in the SAPRC mechanisms  (Carter et
al.,  1986;  Carter,  1988) —  the RADM  rate constants  could  not  be used
because  the   RADM  reactions  they  replace have several  different  rate
constants.   The  rate  constants for the  other  [R02] and  [RC03]  reactions
corresponded  to those  used in  the  RADM mechanism.  The reactions  for the
individual  peroxy radicals  in  this mechanism  with  NO, NOp  (where appli-
cable),  and  H02 are the  same as in the  RADM mechanism.   The  reactions of
the peroxy  radicals  with [R02]  and  [RC03] are  analogous to their reactions
with   M02  and  AC03,  respectively,  in  the RADM mechanism, except that (1)
                                    261

-------
Table 35.  Organic Peroxy + Organic Peroxy Reactions Used in Stockwell's
           "Detailed Radical" Version of the RADM Mechanism
Rxn.
Label
Kinetic Parameters (a)
       k(300)
                Ea
B
                                       Reactions  (a,b)
R001  5.81E+02  2.79E+02 -0.44 -1.00

R002  4.28E+02  2.06E+02 -0.44 -1.00

R003  1.25E+02  6.02E+01 -0.44 -1.00

R004  1.04E+02  4.99E+01 -0.44 -1.00

R005  8.86E+01  4.26E+01 -0.44 -1.00

R006  4.28E+02  2.06E+02 -0.44 -1.00

ROOT  4.28E+02  2.06E+02 -0.44 -1.00

R008  5.19E+01  2.50E+01 -0.44 -1.00

R009  5.19E+01  2.50E+01 -0.44 -1.00

R010  5.19E+01  2.50E+01 -0.44 -1.00



R011  5.19E+01  2.50E+01 -0.44 -1.00


R012  5.19E+01  2.50E+01 -0.44 -1.00

R013  5.19E+01  2.50E+01 -0.44 -1.00

R014  5.19E+01  2.50E+01 -0.44 -1.00

R015  2.93E+03  1.41E+03 -0.44 -1.00

R016  2.93E+03  1.41E+03 -0.44 -1.00



R017  7.64E+01  3.67E+01 -0.44 -1.00

R018  4.58E+01  2.20E+01 -0.44 -1.00

R019  3.67E+01  1.76E+01 -0.44 -1.00
                            M02 + M02 = 11.5 HCHO + H02 +
                              i.5 ROM
                            M02 + ETHP =4.75 HCHO + H02 +
                              1.75 ALD + t.5 ROH
                            M02 + HC3P = #.75 HCHO + H02 +
                              1.15 ALD * *.6 KET + #.5 ROH
                            M02 * HC5P = #.75 HCHO + H02 +
                              #.105 ALD + #.645 KET * #.5 ROH
                            M02 + HC8P = #.75 HCHO + H02 +
                              #.075 ALD + #.675 KET + #.5 ROH
                            M02 * OL2P = #1.550 HCHO * H02 +
                              #.350 ALD + #.5 ROH
                            M02 + OLTP = #1.25 HCHO + H02 +
                              1.75 ALD + #.5 ROH
                            M02 «• OLIP = #.890 HCHO + H02 *
                              #.725 ALD + #.3 KET + #.5 ROH
                            M02 t- KETP = #.75 HCHO + H02 +
                              #.75 MGLY +1.5 ROH
                            M02 + TOLP = #.75 HCHO +
                              #1.405 H02 + #.071 ALD +
                              #.543 MGLY + #.274 ROH +
                              #.181 GLY -i- #.905 DCB
                            M02 + XYLP = #.75 HCHO -»•
                              #1.5 H02 + #.75 MGLY +
                              #.25 ROH + DCB
                            M02 + OLN = #1.75 HCHO + #.5 H02 +
                              ALD + #.25 ROH + N02
                            M02 + XN02 = #.75 HCHO + #.5 H02 +
                              #.25 ROH
                            M02 + X02 = #.75 HCHO + #.5 H02 +
                              #.25 ROH
                            M02 f AC03 = HCHO + #.5 H02 +
                              #.5 ORA2 + #.5 M02
                            M02 + TC03 = HCHO + #.96 H02 +
                              #.055 MGLY + #.5 ORA2 +
                              #.445 GLY + #.025 AC03 + X02 +
                              #.475 CO
                            ETHP + ETHP = H02 + #1.5 ALD +
                              #.5 ROH
                            ETHP + HC3P = H02 + #.9 ALD H-
                              #.6 KET + #.5 ROH
                            ETHP + HC5P = H02 + #.855 ALD *
                              #.645 KET -i- #.5 ROH

                                                   (continued)
                                    262

-------
Table 35 (continued) - 2
Rxn.
Label
Kinetic Parameters (a)
       k(300)
                Ea    B
                            Reactions (a,b)
R020  3.05E+01   1.47E+01 -0.44 -1.00

R021  1.53E+02   7.34E+01 -0.44 -1.00

R022  1.53E+02   7.34E+01 -0.44 -1.00

R023  1.83E+01   8.81E+00 -0.44 -1.00

R024  1.83E+01   8.81E+00 -0.44 -1.00

R025  1.83E+01   8.81E+00 -0.44 -1.00



R026  1.83E+01   8.81E+00 -0.44 -1.00


R027  1.83E+01   8.81E+00 -0.44 -1.00

R028  1.83E+01   8.81E+00 -0.44 -1.00

R029  1.83E+01   8.81E+00 -0.44 -1.00

R030  1.04E+03   4.99E+02 -0.44 -1.00

R031  1.04E+03   4.99E+02 -0.44 -1.00



R032  7.64E+01   3.67E+01 -0.44 -1.00

R033  1.10E+01   5.28E+00 -0.44 -1.00

R034  9.47E+00   4.55E+00 -0.44 -1.00

R035  4.58E-I-01   2.20E+01 -0.44 -1.00

R036  4.58E+01   2.20E+01 -0.44 -1.00

R037  5.50E+00   2.64E+00 -0.44 -1.00

R038  5.50E+00   2.64E+00 -0.44 -1.00


R039  5.50E+00   2.64E+00 -0.44 -1.00
ETHP + HC8P =
  #.675 KET +
ETHP i- OL2P =
  11.1 ALD +
ETHP -K OLTP =
  #1.5 ALD +
ETHP + OLIP =
  #1.475 ALD
ETHP + KETP =
  #.75 MGLY +
ETHP + TOLP =
  #.821 ALD *
  #.274 ROH +
  #.905 DCB
ETHP * XYLP =
  #.75 ALD +
  #.25 ROH +
ETHP + OLN =
  #1.75 ALD +
ETHP + XN02 =
  #.25 ROH
ETHP + X02 =
  #.25 ROH
ETHP + AC03 =
  1.5 ORA2
ETHP i- TC03 =
  #.055 MGLY
  #.445 GLY +
  #.475 CO
HC3P + HC3P =
  #1.200 KET
HC3P + HC5P =
  # 1 . 245 KET
HC3P + HC8P =
  #1.275 KET
HC3P + OL2P
  #.5 ALD +1
HC3P + OLTP =
  #.9 ALD +1
HC3P + OLIP
  #.875 ALD +
HC3P + KETP =
  #.6 KET + #
                                          H02 + #.825 ALD +
                                          1.5 ROH
                                          1.8 HCHO + H02 +
                                         #.5 ROH
                                          1.5 HCHO + H02 +
                                         1.5 ROH
                                          #.14 HCHO + H02 +
                                         * #.3 KET + #.5 ROH
                                          H02 + #.75 ALD +
                                          #.5 ROH
                                          #1.405 H02 +
                                          #.543 MGLY +
                                          1.181 GLY *

                                          11.5 H02 +
                                         1.75 MGLY +
                                         DCB
                                         HCHO + #.5 H02 +
                                          1.25 ROH + N02
                                          1.5 H02 + #.75 ALD

                                         #.5 H02 + #.75 ALD

                                          #.5 H02 + ALD +
                                         *.5 M02
                                          #.96 H02 + ALD +
                                         + #.5 ORA2 +
                                          #.025 AC03 + X02 +

                                          H02 * 1.3 ALD +
                                         + #.5 ROH
                                          H02 1- #.255 ALD +
                                         + # . 5 ROH
                                          H02 + #.225 ALD +
                                         + #.5 ROH
                                        = #.8 HCHO + H02 +
                                        1.6  KET +1.5 ROH
                                        = #.5 HCHO + H02 +
                                          .6  KET + #.5 ROH
                                        =1.14 HCHO + H02 +
                                        + #.9 KET + #.5 ROH
                                        = H02 + 1.15 ALD +
                                          .75 MGLY + #.5 ROH
                            HC3P + TOLP  =  #1.405 H02 +
                              #.221 ALD  +  #.6 KET +
                                                   (continued)
                                    263

-------
Table 35 (continued) - 3
Bxn.
Label
Kinetic Parameters (a)
       k(300)
                Ea
                            Reactions (a,b)
R040  5.50E+00  2.64E+00 -0.44 -1.00


R041  5.50E+00  2.64E+00 -0.44 -1.00


R042  5.50E+00  2.64E+00 -0.44 -1.00

R043  5.50E+00  2.64E+00 -0.44 -1.00

R044  3.06E+02  1.47E+02 -0.44 -1.00

R045  3.06E+02  1.47E+02 -0.44 -1.00



R046  7.64E+01  3.67E+01 -0.44 -1.00

R047  7.64E+00  3-67E+00 -0.44 -1.00

R048  3.67E+01  1.76E+01 -0.44 -1.00

R049  3.67E+01  1.76E+01 -0.44 -1.00

R050  4.58E+00  2.20E+00 -0.44 -1.00

R051  4.58E+00  2.20E+00 -0.44 -1.00

R052  4.58E+00  2.20E+00 -0.44 -1.00



R053  4.58E+00  2.20E+00 -0.44 -1.00


R054  4.58E-.-00  2.20E+00 -0.44 -1.00


R055  4.58E+00  2.20E+00 -0.44 -1.00

R056  4.58E+00  2.20E-I-00 -0.44 -1.00
                              #.543 MGLY -t- #.274 ROH +
                              #.181 GLY + #.905 DCB
                            HC3P + XYLP = #1.5 H02 +
                              #.15 ALD + #.6 KET + #.75 MGLY *
                              #.25 ROH + DCB
                            HC3P + OLN = HCHO + #.5 H02 +
                              #1.150 ALD * #.6 KET +
                              #.25 ROH + N02
                            HC3P + XN02 = #.5 H02 + #.15 ALD *
                              #.6 KET H- #.25 ROH
                            HC3P + X02 = #.5 H02 * #.15 ALD +
                              #.6 KET •*• #.25 ROH
                            HC3P + AC03 = *.5 H02 + #.2 ALD +
                              #.8 KET + #.5 ORA2 + #.5 M02
                            HC3P + TC03 = #.96 H02 * #.2 ALD •»
                              #.8 KET + #.055 MGLY +
                              #.5 ORA2 + #.445 GLY +
                              #.025 AC03 + X02 + #.475 CO
                            HC5P + HC5P = H02 + #.21 ALD +
                              #1.290 KET -t- #.5 ROH
                            HC5P * HC8P = H02 + #.180 ALD +
                              #1.320 KET + #.5 ROH
                                           .8 HCHO + H02 +
                                           .645 KET + #.5 ROH
                                           .5 HCHO * H02 +
                                           .645 KET + #.5 ROH
                                           .14 HCHO + H02 >
                                           .945 KET + #.5 ROH
HC5P + OL2P =
  #.455 ALD +
HC5P + OLTP =
  #.855 ALD +
HC5P * OLIP =
  #.830 ALD
                            HC5P + KETP  =  02+1.105  ALD +
                              #.645 KET
                            HC5P + TOLP  =
                                .176 ALD  +
                                .543 MGLY ^
                                .181 GLY  +
                            HC  P + XYLP  =
                                .105 ALD  +
                                .75 MGLY  +
               .75 MGLY + #.5 ROH
               1.405 H02 +
               .645 KET +
               #.274 ROH +
               .905 DCB
               1.5 H02 *
               .645 KET +
               .25 ROH + DCB
                            HC P +  OLN  =  HCHO +  #.5  H02  +
                               1.105  ALD  +  #.645 KET +
                               .25  ROH  +  N02
                            HC5P +  XN02 = #.5 H02 +
                              #.105 ALD + #.645  KET  + #.25 ROH
                            HC5P +  X02  =  #.5 H02 + 1.105 ALD +
                              #.645 KET + #.25 ROH

                                                   (continued)
                                    264

-------
Table 35 (continued) - 4
Rxn.
Label
Kinetic Parameters (a)
       k(300)
                Ea
B
                            Reactions (a,b)
R057  2.57E+02   1.23E+02 -0.44 -1.00

R058  2.57E+02   1.23E+02 -0.44 -1.00




R059  7.64E+01   3-67E+01 -0.44 -1.00

R060  3.05E+01   1.47E+01 -0.44 -1.00

R061  3.05E+01   1.47E+01 -0.44 -1.00

R062  3.97E+00   1.91E+00 -0.44 -1.00

R063  3.97E-t-00   1.91E+00 -0.44 -1.00

R064  3.97E+00   1.91E+00 -0.44 -1.00



R065  3.97E+00   1.91E+00 -0.44 -1.00


R066  3.97E-I-00   1.91E+00 -0.44 -1.00


R067  3.97E+00   1.91E+00 -0.44 -1.00

R068  3-97E+00   1.91E+00 -0.44 -1.00

R069  2.20E+02   1.06E+02 -0.44 -1.00

R070  2.20E+02   1.06E+02 -0.44 -1.00



R071  7.64E+01   3.67E+01 -0.44 -1.00

R072   1.53E+02   7.34E+01 -0.44 -1.00

R073   1.83E+01   8.81E+00 -0.44 -1.00
                            HC5P + AC03 = #.5 H02 + #.14 ALD +
                              #.860 KET + #.5 ORA2 + #.5 M02
                            HC5P + TC03 = #.96 H02 +
                              #.14 ALD + #.860 KET +
                              #.055 MGLY + #.5 ORA2 +
                              #.445 GLY + #.025 AC03 + X02 +
                              #.475 CO
                            HC8P + HC8P = H02 + #.15 ALD +
                              #1.350 KET -t- #.5 ROM
                            HC8P + OL2P = #.8 HCHO + H02 +
                              #.425 ALD + #.675 KET + #.5 ROH
                            HC8P + OLTP = #.5 HCHO + H02 +
                              #,825 ALD i- #.675 KET + #.5 ROH
                            HC8P + OLIP = #.14 HCHO + H02 +
                              #.8 ALD + #.975 KET + #.5 ROH
                            HC8P -t- KETP = H02 + #.075 ALD +
                              #.675 KET + #.75 MGLY +1.5 ROH
                            HC8P + TOLP = #1.405 H02 +
                              #.146 ALD -i- #.675 KET +
                              #.543 MGLY + #.274 ROH +
                              1.181 GLY + #.905 DCB
                            HC8P + XYLP = #1.5 H02 +
                              #.075 ALD -i- #.675 KET +
                              #.75 MGLY + #.25 ROH + DCB
                            HC8P + OLN = HCHO + #.5 H02 +
                              #1.075 ALD + #.675 KET +
                              #.25 ROH + N02
                            HC8P + XN02 = #.5 H02 +
                              #.075 ALD + #.675 KET + #.25  ROH
                            HC8P + X02 = #.5 H02 + #.075 ALD +
                              #.675 KET + #.25 ROH
                            HC8P + AC03 = #.5 H02 + #.1 ALD +
                              #.9 KET + #.5 ORA2 + #.5  M02
                            HC8P + TC03 = #.96 H02 + #.1 ALD +
                              #.9 KET + #.055 MGLY +
                              #.5 ORA2 + #.445 GLY +
                              #.025 AC03 + X02 + #.475  CO
                            OL2P + OL2P  = #1.6 HCHO + H02  +
                              #.7 ALD + #.5  ROH
                            OL2P + OLTP  = #1.300 HCHO + H02 +
                              #1.1 ALD + #.5  ROH
                            OL2P + OLIP  = #.940  HCHO +  H02  +
                              #1.075  ALD + #.3 KET  -t- #.5  ROH

                                                    (continued)
                                     265

-------
Table 35 (continued) - 5
Rxn.       Kinetic Parameters (a)
Label	   Reactions (a,b)
       k(300)     A        Ea    B
R074  1.83E+01  8.81E+00 -0.44 -1.00

R075  1.83E+01  8.81E+00 -0.44 -1.00



R076  1.83E+01  8.81E+00 -0.44 -1.00


R077  1.83E+01  8.81E+00 -0.44 -1.00


R078  1.83E+01  8.81E+00 -0.44 -1.00

R079  1.83E+01  8.81E+00 -0.44 -1.00

R080  1.04E+03  4.99E+02 -0.44 -1.00

R081  1.04E+03  4.99E+02 -0.44 -1.00




R082  7.64E+01  3.67E+01 -0.44 -1.00

R083   1.83E+01  8.81E+00 -0.44 -1.00

R084   1.83E+01  8.81E+00 -0.44 -1.00

R085   1.83E+01  8.81E+00  -0.44 -1.00



 R086   1.83E+01   8.81E+00  -0.44  -1.00


 R087   1.83E+01   8.81E+00 -0.44  -1.00

 R088  1.83E+01   8.81E+00 -0.44 -1.00

 R089  1.83E+01   8.81E+00 -0.44 -1.00

 R090  1.04E+03  4.99E+02 -0.44 -1.00
OL2P + KETP = #.8 HCHO + H02 +
  #.350 ALD + 1.75 MGLY + 1.5 ROH
OL2P + TOLP =1.8 HCHO +
  11.405 H02 -t- 1.421 ALD +
  1.543 MGLY + 1.274 ROH +
  #.181 GLY + #.905 DCB
OL2P + XYLP = #.8 HCHO +
  11.5 H02 t- #.350 ALD +
  #.75 MGLY + #.25 ROH * DCB
OL2P + OLN = #1.800 HCHO +
  #.5 H02 + #1.350 ALD +
  #.25 ROH * N02
OL2P + XN02 = #.8 HCHO + #.5 H02  *
  #.350 ALD * #.25 ROH
OL2P + X02 = #.8 HCHO +1.5 H02 +
  #.350 ALD + #.25 ROH
OL2P + AC03 = #.8 HCHO + #.5 H02  +
  #.6 ALD + #.5 ORA2 + #.5 M02
OL2P + TC03 = #.8 HCHO +
  #.96 H02 + #.6 ALD +
  #.055 MGLY +1.5 ORA2 +
  #.445 GLY + #.025  AC03 + X02  +
  #.475 CO
OLTP + OLTP = HCHO + H02 +
  #1.5 ALD +  #.5  ROH
OLTP + OLIP  = #.640  HCHO +  H02  +
  #1.475  ALD  +  #.3  KET  + #.5  ROH
OLTP + KETP  = #.5  HCHO  + H02  +
  #.75 ALD +  #.75  MGLY  + #.5  ROH
OLTP + TOLP  =1.5  HCHO  +
  #1.405  H02  +  #.821  ALD +
  #.543  MGLY  +  #.274 ROH  +
  1.181  GLY  + #.905 DCB
 OLTP + XYLP  =1.5 HCHO +
  #1.5 H02  + #.75 ALD +
   #.75 MGLY + #.25 ROH +  DCB
 OLTP  + OLN =  #1.5 HCHO +  #.5 H02 +
   #1.75  ALD + 1.25 ROH + N02
 OLTP  + XN02 = #.5 HCHO + #.5 H02 +
   1.75 ALD + #.25 ROH
 OLTP + X02 = #.5 HCHO +4.5 H02  +
   #.75 ALD + #.25 ROH
 OLTP + AC03 = #.5 HCHO + #.5 H02 +
   ALD + #.5 ORA2 + #.5 M02

                        (continued)
                                     266

-------
Table 35 (continued) - 6
Rxn.
Label
Kinetic Parameters (a)
       k(300)
                Ea
                            Reactions (a,b)
R091  1.04E+03  4.99E+02 -0.44 -1.00



R092  1.10E+00  5.28E-01 -0.44 -1.00

R093  2.23E+00  1.07E+00 -0.44 -1.00


R094  2.23E+00  1.07E+00 -0.44 -1.00
R095  2.23E+00   1.07E+00  -0.44 -1.00


R096  2.23E+00   1.07E+00  -0.44 -1.00


R097  2.23E+00   1.07E+00  -0.44 -1.00


R098  2.23E+00   1.07E+00  -0.44 -1.00

R099  1.28E+02   6.17E+01  -0.44 -1.00


R100  1.28E+02   6.17E+01  -0.44 -1.00
 R101   1.10E+00  5.28E-01  -0.44  -1.00

 R102   2.23E+00  1.07E+00  -0.44  -1.00



 R103   2.23E+00  1.07E+00  -0.44  -1.00

 R104   2.23E+00  1.07E+00  -0.44  -1.00
                            OLTP t- TC03 = #.5 HCHO +
                              #.96 H02 + ALD + 1.055 MGLY +
                              #.5 ORA2 + #.445 GLY +
                              #.025 AC03 + X02 + #.475 CO
                            OLIP + OLIP = #.280 HCHO + H02 +
                              #1.450 ALD * #.6 KET + #.5 ROH
                            OLIP -i- KETP = #.14 HCHO + H02 +
                              #.725 ALD + #.3 KET +
                              #.75 MGLY * #.5 ROH
                            OLIP + TOLP = //.14 HCHO +
                              #1.405 H02 + #.796 ALD +
                              #.3 KET + #.543 MGLY +
                              #:274 ROH + #.181 GLY +
                              #.905 DCB
                            OLIP f XYLP = #.14 HCHO +
                              #1.5 H02 + #.725 ALD + #.3 KET +
                              #.75 MGLY -f #.25 ROH + DCB
                            OLIP + OLN = #1.140 HCHO +
                              #.5 H02 + #1.725 ALD + #.3 KET +
                              #.25 ROH + N02
                            OLIP + XN02 = #.14 HCHO +
                              #.5 H02 * #.725 ALD + #.3 KET +
                              #.25 ROH
                            OLIP + X02 = #.14 HCHO + #.5 H02 +
                              #.725 ALD + #.3 KET + #.25 ROH
                            OLIP > AC03 = #.14 HCHO +
                              #.5 H02 + #.725 ALD + #.55 KET +
                              #.5 ORA2 + #.5 M02
                            OLIP + TC03 = #.14 HCHO +
                              #.96 H02 + #.725 ALD +
                              #.55 KET-+ #.055 MGLY +
                              #.5 ORA2 + #.445 GLY +
                              #.025 AC03 + X02 + #.475  CO
                            KETP + KETP = H02 + #1.5 MGLY  +
                              #.5 ROH
                            KETP + TOLP = #1.405 H02 +
                              #.071 ALD + #1.293 MGLY +
                              #.274 ROH + 1.181 GLY +
                              #.905 DCB
                            KETP + XYLP  = 11.5 H02 +
                              #1.5 MGLY + #.25 ROH + DCB
                            KETP + OLN  = HCHO + #.5 H02 *
                              ALD + #.75 MGLY + #.25 ROH + N02

                                                    (continued)
                                     267

-------
Table 35 (continued) - 7
Rxn.
Label
Kinetic Parameters (a)
       k(300)
                Ea    B
                            Reactions (a,b)
R105  2.23E+00  1.07E+00 -0.44 -1.00

R106  2.23E+00  1.07E+00 -0.44 -1.00

R107  1.28E+02  6.17E+01 -0.44 -1.00

R108  1.28E+02  6.17E+01 -0.44 -1.00



R109  1.10E+00  5.28E-01 -0.44 -1.00



R110  2.23E+00  1.07E+00 -0.44 -1.00



R111  2.23E+00  1.07E+00 -0.44 -1.00



R112  2.23E+00  1.07E+00 -0.44 -1.00



R113  2.23E+00  1.07E+00 -0.44 -1.00



R114  1.28E+02  6.17E+01 -0.44 -1.00



R115  1.28E+02  6.17E+01 -0.44 -1.00
 R116   1.10E+00  5.28E-01  -0.44  -1.00

 R117   2.23E+00  1.07E-I-00  -0.44  -1.00
                            KETP + XN02 = #.5 H02 +
                              #.75 MGLY + #.25 ROM
                            KETP + X02 = 1.5 H02 + #.75 MGLY +
                              #.25 ROH
                            KETP + AC03 = #.5 H02 * MGLY +
                              #.5 ORA2 * #.5 M02
                            KETP + TC03 = #.96 H02 +
                              #1.055 MGLY +1.5 ORA2 +
                              #.445 GLY + #.025 AC03 + X02 +
                              #.475 CO
                            TOLP + TOLP = #1.810 H02 +
                              #.142 ALD + #1.448 MGLY +
                              #.047 ROH + #.362 GLY +
                              #1.810 DCB
                            TOLP + XYLP = #1.905 H02 +
                              #.071 ALD + #1.655 MGLY +
                              #.024 ROH + 1.181 GLY +
                              #1.905 DCB
                            TOLP + OLN = HCHO + #.905 H02 +
                              #1.071 ALD + #.905 MGLY +
                              #.024 ROH + 1.181 GLY +
                              #.905 DCB + N02
                            TOLP + XN02 = #.905 H02 +
                              #.071 ALD + #.905 MGLY +
                              #.024 ROH + 1.181 GLY +
                              #.905 DCB
                            TOLP + X02 = #.905 H02 +
                              #.071 ALD + #.905 MGLY +
                              #.024 ROH + #.181 GLY +
                              #.905 DCB
                            TOLP + AC03 = #.905 H02 +
                              #.095 ALD + #.724 MGLY +
                              #.5 ORA2 + #.181 GLY +
                              #.905 DCB + #.5 M02
                            TOLP + TC03 = #1.365 H02 +
                              #.095 ALD + #.779 MGLY +
                              #.5 ORA2 + #.626 GLY +
                              #.905 DCB + #.025 AC03 + X02 +
                              #.475 CO
                            XYLP + XYLP = #2 H02 + 12 MGLY +
                              #2 DCB
                            XYLP + OLN = HCHO + H02 + ALD +
                              #1.25 MGLY + DCB + N02

                                                   (continued)
                                    268

-------
Table 35 (continued) - 8
Rxn.
Label
Kinetic Parameters (a)
       k(300)
                Ea
B
                            Reactions (a,b)
R118  2.23E+00   1.07E+00  -0.44 -1.00

R119  2.23E+00   1.07E+00  -0.44 -1.00

R120  1.28E+02   6.17E+01  -0.44 -1.00

R121  1.28E+02   6.17E+01  -0.44 -1.00



R122  1.10E+00   5.28E-01  -0.44 -1.00

R123  2.23E+00   1.07E+00  -0.44 -1.00
R124  2.23E+00   1.07E+00  -0.44 -1.00
R125  1.28E+02   6.17E+01  -0.44 -1.00

R126  1.28E+02   6.17E+01  -0.44 -1.00
R127   1.10E+00  5.28E-01  -0.44  -1.00
R128   2.23E+00  1.07E+00  -0.44  -1.00
R129   1.28E+02  6.17E+01  -0.44  -1.00
R130   1.28E+02  6.17E+01  -0.44  -1.00
R131   1.10E+00   5.28E-01  -0.44  -1.00
R132   1.28E+02   6.17E+01  -0.44  -1.00
R133   1.28E+02   6.17E+01  -0.44  -1.00
R134   3.67E+03   1.76E+03  -0.44  -1.00
R135   7.33E+03   3.52E+03  -0.44  -1.00
 R136   3.67E+03   1.76E+03 -0.44  -1.00
                                              >• #1.25 MGLY

                                               #1.25 MGLY H

                                              »• MGLY +
                                               it.5 M02
                                                #.025 AC03
                                                     ALD
XYLP + XN02 = H02
  DCB
XYLP + X02 = H02 H
  DCB
XYLP + AC03 = H02
  1.5 ORA2 + DCB 1
XYLP -•- TC03 = #1.46 H02
  #1.055 MGLY + #.5 ORA2
  #.445 GLY + DCB
  X02 + #.475 CO
OLN + OLN = #2 HCHO
  #2 N02
OLN + XN02 = HCHO
OLN *• X02 = HCHO
OLN + AC03 = HCHO
  #.5 ORA2 -i- N02 1
OLN + TC03 = HCHO
  ALD + #.055 MGLY + #.5 ORA2
  #.445 GLY > #.025 AC03 + X02
  #.475 CO + N02
XN02 + XN02 =
XN02 + X02 =
XN02 1- AC03 = #.5 ORA2
XN02 + TC03 = #.46 H02
                                N02
                               N02
                                                ALD
                                               ALD
                                                ALD +
                                               #.5 M02
                                                #.46 H02
                                                     #.5 M02
                                         #,
                                         #
                                055 MGLY
                                445 GLY H
                    > #.5 ORA2 H
                    1.025 AC03
                                          #.475 CO
      X02
      X02
      X02
        #
        #
        #
    + X02 =
    + AC03 = #.5
    + TC03 = #,
    ,055 MGLY +
    .445 GLY
                           X02
                                                        M02
                 ORA2 + #.5
               46 H02 +
               #.5 ORA2 +
              #.025 AC03 + X02 4
    475 CO
AC03 + AC03 = #2 M02
AC03 + TC03 = #.920 H02 +
  #.110 MGLY + #.890 GLY +
  #.05 AC03 + #2 X02 + #.950 CO
TC03 + TC03 = #1.84 H02 +
  #.220 MGLY + #1.780 GLY +
  #.1 AC03 + #4 X02 + #1.900 CO
                                                                          M02
 (a)   See  Footnotes  (b)  and (c)  in  Table 2  for a description of the formats
      used in  the  kinetic parameter and  the reaction listings.

 (b)   Note that  ROH,  a species not  in  the standard RADM mechanism,  is shown
      as a product in some of these reactions.  However, no reactions for
      it are  included in this mechanism, so its formation has no effect.
                                    269

-------
Table 36.  Listing of the Organic Reactions in the Modified RADM
           Mechanism Incorporating the 1986 SAPRC Approach for
           Representing Reactions of Organic Peroxy Radicals
Rxn.
Label
(a)
010
011
012
013
014
015
016
017
018
019
020
021
Kinetic Parameters (b)
k(300)

(Phot.
(Phot.
(Phot.
(Phot,
( Phot ,
(Phot,
(Phot,
(Phot
(Phot
(Phot
(Phot
(Phot
A
Ea
B
, Set = HCHOMR )
, Set = HCHORR )
. Set = ALDR )
, Set = OPR )
. Set = OPR )
. Set = PAAR )
. Set = KETR )
. Set = GLY2R )
. Set = GLY1R )
. Set = MGLYR )
. Set = DCBR )
. Set = ONITR )
Reactions (b)
Organic Photolysis Reactions
HCHO * HV = H2 * CO
HCHO + HV = H02 + H02 + CO
ALD + HV = M02 + H02 + CO + [R02]
OP1 -i- HV = HCHO + H02 + HO
OP2 f HV = ALD + H02 + HO
PAA * HV = M02 + C02 -e HO + [R02]
KET * HV = AC03 + [RC03] + ETHP +
[R02]
GLY + HV = 1.13 HCHO + #1.87 CO
GLY + HV = #.45 HCHO + #1.55 CO +
#,80 H02
MGLY * HV = AC03 + [RC03] f H02 +
CO
DCB + HV = #.98 H02 * #.02 AC03 +
TC03 f #1.02 [RC03]
ONIT + HV = #.20 ALD + #.80 KET +
H02 + N02
Organic + OH Reactions
051
052
053


054

055

056
057
058
074
059

060

063
064
065
1
4
3


7

1

1
3
9
1
9

4

1
2
1
.29E+01
. 11E+02
.76E+03


.86E+03

.66E+04

.24E+04
.82E+04
.24E+04
.46E+05
.02E+03

.54E+04

.32E+04
.37E+04
.47E+03
9.23E+02
1.81E+03
2 . 28E+04


2.79E+04

5.87E+04

3.16E+03
7.12E+03
1.48E+04
3.74E+04
3.08E+03

3.08E+04

1.32E+04
1.01E+04
1.76E+04
2.54
0.88
1.07


0.75

0.75

-0.82
-1.00
-1.09
-0.81
-0.64

-0.23

0.00
-0.51
1.48
1
1
-1


-1

-1

-1
-1
-1
-1
-1

-1

-1
-1
-1
.00
.00
.00


.00

.00

.00
.00
.00
.00
.00

.00

.00
.00
.00
CH4
ETH
HC3
#
#
HC5
#
HC8
#
OL2
OLT
OLI
ISO
TOL
#
XYL
#
+ HO = M02 + H20 +
* HO = ETHP
* HO = #.90
+ [R02]
"HC3P *
.10 H02 + #.014 HCHO
.069 ALD + #
+ HO = HC5P
1.27 [R02]
+ HO = HC8P
1.78 [R02]
+ HO = OL2P
+ HO = OLTP
+ HO = OLIP
+ HO = OLTP
+ HO = #.84
. 16 CSL + #.
+ HO = #.83
.17 CSL + #.
HCHO + HO = H02
ALD
KET
+ HO = AC03
* HO = KETP
.026 KET
[R02]
+ H20
[R02]"
+
+ H20


f


+ #.27 X02 + H20 -i

+ #.78

+ [R02]
+ [R02]
+ [R02]
* [R02]
"TOLP +
16 H02
"XYLP +
17 H02
+ CO +


X02 -t- H20 -.





[RQ2]"

[R02]"

H20





+

+


+ [RC03] + H20
+ [R02]
+ H20

                                                               (continued)
                                     270

-------
Table 36 (continued) - 2
Rxn.
Label
(a)
066
067

068
061
Kinetic Parameters (b)
k(300)
1.
2.

4.
5.
.69E+04
.50E+04

, 1 1 E+04
.87E+04
1,
2

4
5
A
.69E+04
.50E+04

. 11 E+04
.87E+04
Ea
0.
0.

0.
0.
00
00

00
00
B
-1
-1

-1
-1
.00
.00

.00
.00
Reactions (b)
GLY + HO = H02 + #2 CO +
MGLY + HO = AC03 + [RC03]
H20
DCB + HO = TC03 + [RC03]
CSL + HO = #.1 H02 + #.9

H20
+



CO +

+ H20
"X02 +
[R02]" + #.9 "TC03 + [RC03
062
069

070

071
072

073

093
094
095
096

097
098
099
100
101
102

103

104




105




5.
1,

1.

1.
2

3

9
3
9
3

3
3
1
2
1
8

2

1




3




.28E+04
.47E+04

.47E+04

.47E+04
.06E+02

.76E+03

.23E-01
. 65E+00
.23E-01
. 65E+00

. 65E+00
.23E+04
.72E-01
.33E+01
.84E+03
.53E+02

.72E-03

.74E-02




.02E-01




5
1

1

1
9

2

8
2
8
2

2
3
2
1
4
8

1

1




1




.28E+04
.47E+04

.47E+04

.47E+04
.06E+02

.28E+04

.81E+02
.06E+03
.81E+02
.06E+03

.06E+03
.23E+04
.94E+03
.47E+04
0.
0.

0.

0.
0.

1.

4.
3.
4.
3.

3.
0.
00
00

00

00
88

07

09
78
09
78

78
00
5.81
3.
84
.74E+04 1.94
.53E+02

.76E+01

.94E+01




.33E+01




0.

,00

5.23

4,





.18




2.26








-1
-1

-1

-1
1

-1

-1
-1
-1
-1

-1
-1
-1
-1
-1
-1

-1

-1




.00
.00

.00

.00
.00

.00

.00
.00
.00
.00

.00
.00
.00
.00
.00
.00

.00

.00




-1.00








CSL + HO = CSL

1"

OP1 + HO = #.5 "M02 + [R02]" +
#.5 HCHO + #.5 HO

OP2 + HO = #.5 "HC3P + [R02]
#.5 ALD + #.5 HO
PAA + HO = AC03 + [RC03]


" +

+ H20
PAN + HO = HCHO + N03 + X02
[R02]
ONIT + HO = HC3P + [R02]
Organic + N03 Reactions
HCHO + N03 = H02 + HN03 n

+

+ N02


H CO
ALD + N03 = AC03 + [RC03] +
GLY + N03 = HN03 + H02 +
MGLY + N03 = HN03 + AC03
[RC03] + CO
DCB + N03 = HN03 + TC03 -
#2
+

HN03
CO


H [RC03]
CSL + N03 = HN03 + XN02 + #.
OL2 + N03 = OLN + [R02]
OLT + N03 = OLN + [R02]
OLI + N03 = OLN + [R02]
ISO + N03 = OLN + [R02]
Organic + 0-3 Reactions





.5 CSL





OL2 + 03 = HCHO + #.42 CO +
#.4 ORA1 + #.12 H02
OLT + 03 = #.53 HCHO + #
#.33 CO + #.20 ORA1 +
#.20 ORA2 + #.23 H02 +
#.22 "M02 + [R02]" + #
#.06 CH4
OLI + 03 = #.18 HCHO + #
#.10 KET + #.23 CO + #
#.29 ORA2 + #.09 CH4 +
#.26 H02 + #.14 HO + #
[R02]"

.50


.10

.72
.06

.31


ALD +


HO +

ALD +
ORA1 -

"M02 •

                                                               (continued)
                                    271

-------
Table 36 (continued) - 3
Rxn.       Kinetic Parameters (b)

(a)    k(300)     A        Ea    B
Reactions (b)
106   2.20E-02  1.81E1-01  4.00 -1.00
M01   7.52E+03  4.11E+03 -0.36 -1.00
075               (Same k as M01 )
077               (Same k as M01 )
076   2.90E-02  1.17E+18 26.91  0.00
078   2.90E-02  1.17E+18 26.91  0.00
M02   1.12E+04  6.17E+03 -0.36 -1.00
079               (Same k as M02 )
080               (Same k as M02 )
081                (Same k as M02 )

082                (Same k as M02 )


083                (Same k as M02 )

084                (Same k as H02 )
085                (Same k as M02 )

088                (Same k as M02 )


089                (Same k as M02 )
090                (Same k as M02 )
091                (Same k as M02 )
092                (Same k as M02 )
 150                (Same k as M02 )

M03                (Same k as M02 )
086                (Same k as M02 )
087                (Same k as M02 )
ISO + 03 = #.53 HCHO + 1.50 ALD
  #.33 CO f #.20 ORA1 +
  #.20 ORA2 + #.23  H02 +
  1.22 "M02 + [R02]" + if. 10 HO

Organic Peroxy + N02 Reactions
[RC03] * N02 = N02
AC03 + N02 = PAN
TC03 + N02 = TPAN
PAN = AC03 + [RC03]
TPAN = TC03 f [RC03]
                      N02
                       N02
Organic Peroxy + NO Reactions

[R02] + NO = NO
M02 + NO = HCHO + H02 + N02
HC3P * NO = 1.69 ALD + 1.26 KET
  1.14 HCHO + #.03 ONIT +
  *.97 N02 + 1.97 H02
HC5P + NO = #.36 ALD
  1.10 ONIT
                       1.76  KET *
              #.90 N02 + #.90  H02
HC8P + NO  = #.M4 ALD + #1.05 KET  +
  #.05 HCHO + #.24 ONIT +
  #.76 N02 + 1.76 H02
OL2P + NO  = #1.6 HCHO + H02 +
  N02 +  #.2 ALD
OLTP + NO  = ALD + HCHO * H02 +  N02
OLIP + NO  = H02 + #1.45 ALD +
  #.28   HCHO + #.10 KET +  N02
TOLP * NO  = N02 -K H02 +
  #.724  MGLY + #.181 GLY +
  #.905  DCB
XYLP + NO  = N02 + H02 + MGLY +  DCB
ETHP + NO  = ALD + H02 + N02
KETP + NO  = HGLY + N02 + H02
OLN + NO = HCHO + ALD + #2 N02
X02 + NO = N02

 [RC033 + NO = NO
AC03 + NO  = M02 + N02 +  [R02]
TC03 + NO  = N02 + #.92 H02 +
  #.89   GLY + 1.11 MGLY +

                        (continued)
                                    272

-------
Table 36 (continued) - 4
Rxn.
Label 	
(a)    k(300)
Kinetic Parameters (b)

       A        Ea    B
Reactions (b)
MOM
107
108
109
110
111
112
113
IIM
115
117
118
120
1M6
M05
116
119
 M06
 121

 122

 123
 125

 126

 127

 128

 129
                                         #.05 "AC03 +  [RC031" + #.95 CO
                                         #2 "X02 + [R02]"

                                       Organic Peroxy  + H02 Reactions
3.61E+03 1.13E+02 -2.58 -1.00
(Same k as MOM )
(Same k as MOM )
(Same k as MOM )
(Same k as MOM )
(Same k as MOM )
(Same k as MOM )
(Same k as MOM )
(Same k as MOM )
(Same k as MOM )
(Same k as MOM )
(Same k as MOM )
(Same k as MOM )
(Same k as MOM )
(Same k as MOM )
(Same k as MOM )
(Same k as MOM )

1.47E+00 1.47E+00 0.00 -1.00
(Same k as M06 )

(Same k as M06 )

(Same k as M06 )

(Same k as M06 )

(Same k as M06 )

(Same k as M06 )

(Same k as M06 )

(Same k as M06 )

(Same k as M06 )

H02 + [R02] = H02
H02 + M02 = OP1
H02 + ETHP = OP2
H02 + HC3P = OP2
H02 + HC5P = OP2
H02 + HC8P = OP2
H02 + OL2P = OP2
H02 ,+ OLTP = OP 2
H02 + OLIP = OP2
H02 + KETP = OP2
H02 + TOLP = OP2
H02 + XYLP = OP2
H02 + OLN = ON IT
X02 + H02 = OP2
H02 + [RC03] = H02
H02 + AC03 = PAA
H02 + TC03 = OP2
Organic Peroxy + R02 Reactions
[R02] + [R02] =
M02 + [R02] = [R02] + tf.75 HCHO +
#.5 H02
ETHP + [R02] = [R02] + #.75 ALD +
#.5 H02
HC3P + [R02] = [R02] + #.15 ALD +
#.6 KET + #.5 H02
HC5P + [R02] = [R02] + #.105 ALD +
#.645 KET + #.5 H02
HC8P + [R02] = [R02] + #.075 ALD +
#.675 KET -i- #.5 H02
OL2P + [R02] = [R02] + #.8 HCHO +
#.35 ALD + 1.5 H02
OLTP + [R021 = [R02] + #.5 HCHO +
#.75 ALD -i-f.5 H02
OLIP + [R02] = [R02] +1.14 HCHO +
#.725 ALD + #.55 KET + #.5 H02
KETP + [R02] = [R02] + #.75 MGLY +
#.5 H02
                                                                (continued)
                                     273

-------
Table 36 (continued) - 5
Rxn.
Label 	
(a)    k(300)
     Kinetic Parameters (b)

            A        Ea    B
131


132

M07
(Same


(Same

(Same
k


k

k
as


as

as
M06


M06

M06
)


)

)
TOLP
#1
#.
XYLP
#1
OLN
+ [R02]
.4 H02 +
181 GLY
+ [R02]
.5 H02 +
+ [R02]
= [R02
#.724
+ #.905
= [R02
MGLY +
= [R02]
] + #
MGLY
DCB
] + #
DCB
.23 HCHO
+

.25 HCHO

* HCHO + ALD



+

+
N02
147
(Same
k
as
M06
)
X02
+ [R02]
= [R02]



M08
M09
135

136

137

138

139

140


141

143



144


M10

148
130
2.93E+03  1.41E+03 -0.44 -1.00
            (Same k as M08 )

            (Same k as M08 )

            (Same k as M08 )

            (Same k as M08 )

            (Same k as M08 )

            (Same k as M08 )

            (Same k as M08 )

            (Same k as M08 )


            (Same k as M08 )

            (Same k as M08 )



            (Same k as M08 )


            (Same k as M08 )

            (Same k as M08 )
            (Same k as M08 )
Organic Peroxy + RC03 Reactions

[RC03] + [R02] =
M02 + [RC03] = [RC031 *
  #.75 HCHO + #.5 H02
ETHP * [RC03] = [RC03] +
  #.75 ALD + #.5 H02
HC3P + [RC03J = [RC03] +
  #.15 ALD + #.6 KET + #.5 H02
HC5P + [RC03] = [RC03] +
  #.105 ALD + #.645 KET + #.5 H02
HC8P + [RC03] = [RC031 +
  #.075 ALD + #.675 KET + #.5 H02
OL2P + [RC03] = [RC03] +
  #.8 HCHO * #.35 ALD + #.5 H02
OLTP + [RC031 = [RC03] «•
  #.5 HCHO + #.75 ALD + #.5 H02
OLIP + [RC03] = [RC03] +
  #.14 HCHO + #.725 ALD +
  #.55 KET + #.5 H02
KETP + [RC03] = [RC03] +
  #.75 MGLY + #.5 H02
TOLP + [RC03] = [RC03] +
  #.23 HCHO + #1.4 H02 +
  #.724 MGLY + 1.181 GLY +
  #.905 DCB
XYLP + [RC03] = [RC03] +
  #.25 HCHO + #1.5 H02 + MGLY +
  DCB
OLN +  [RC03] = [RC03] + HCHO +
  ALD 1- N02
X02 +  [RC03] = [RC03]
AC03 + [R02] = #1.5  [R02] +
  #.5 M02 + #.5 ORA2

                       (continued)
                                    274

-------
Table 36 (continued) - 6
Rxn.
Label 	
(a)    k(300)
     Kinetic Parameters (b)

            A        Ea    B
Reactions (b)
133
M11
142

145
 151
 152
 153
 154
 155
            (Same k as M08 )
3.64E+03  1.75E+03 -0.44 -1.00
            (Same k as M11 )

            (Same k as M11 )
1.12E+04  6.17E+03 -0.36 -1.00
            (Same k as M04 )
5.19E+01  2.50E+01 -0.44 -1.00
1.28E-t-02  6.17E+01 -0.44 -1.00
1.10E+00  5.28E-01 -0.44 -1.00
TC03 + [R02] = #2 [R02] -t-
  4.5 ORA2 + #.46 H02 +
  1.445 GLY + #.055 MGLY +
  1.025 "AC03 + [RC03]" +
  #.475 CO + X02

[RC03] + [RC03] =
AC03 f [RC031 = [RC03] + M02 +
  [R02]
TC03 + [RC03] = [RC03] +
  1.92 H02 + 1.890 GLY +
  1.11 MGLY + 1.05 "AC03 +
  [RC03]" + #.950 CO + #2 "X02
  [R02]"

XN02 Reactions

XN02 * N02 = ONIT
XN02 + H02 = OP2
XN02 + [R02] = [R02]
XN02 + [RC03] = [RC03]
XN02 + XN02 =
 (a)   Reactions which  correspond directly with  those  in  the  standard  RADM
      mechanism are  given  the  same  reaction  labels  as used for  them in
      Table  2.  Reactions  which are added or significantly modified
      relative to  that mechanism are indicated  by the label  "Mxx."

 (b)   See  Footnotes  (b) and (c) in  Table 2 for  a description of the formats
      used in the  kinetic  parameter and the  reaction  listings.
                                     275

-------
the products from M02  and  AC03 are not  included,  (2)  [R02]  or [RC03] was
included as  a product  as  well as  a reactant  so  the reactions  with the
individual  radicals would be not counted as contributing to their consump-
tion rate  (since this  is  already accounted  for in their  reactions with
[R02] or [RC03]),  and  (3)  the rate  constants  for  the  corresponding  reac-
tion of  [R02] or  [RC03]  are  employed.   Thus,  other than the approximation
approach used  in lumping  the  organic peroxy +  organic  peroxy reactions,
this mechanism was made  to represent, as closely as possible,  the chemical
approaches  used  in  the  March  1988  and the detailed  radical  RADM mechan-
isms.  Therefore, any  differences found  between  the predictions using this
and  the  other mechanisms  should  be due  primarily to differences  in the
approximation approach.
     Table  37  gives the results of the comparisons  of the  calculations
using  the  March  1988  RADM  mechanism  (the "test  set") relative  to the
calculations using the  detailed radical  mechanism  (the "standard  set") for
the  species  of  known  or  potential  interest  in RADM applications.   The
table  gives  the summary  statistics  for  all  the  rural  and  the  urban
scenarios aggregated together, and  the  results  for the individual calcula-
tions where  the  "% Fit" or some other  percent  change statistics  is  greater
than 5%.    (Thus,  if the  results  for an individual test  problem are not
shown,  then  the  two  mechanisms gave predictions to within 5J.)   With very-
few  exceptions,  the two mechanisms give almost exactly the  same  predic-
tions.   Even  in the  cases  where  they are  the most  different,  the  two
mechanisms are generally within 5%  of each  other in  predictions  of  maximum
and  average  concentrations of (X, H202, SULF, HNO^, ORA1, ORA2,  OP1, PAA,
and  PAN, and within  10% in predictions of OP2,  excluding some cases  where
insignificantly  low  concentrations are predicted.
     These  results  indicate that the  approximation  used in  the  evaluated
mechanism,  that  all  the organic peroxy + R02  reactions  can  be represented
by  reactions only with  M02,  and  that  all  the organic  peroxy -^RCO^  reac-
tions  can  be represented only by reactions with AC03, is justified.   This
is  consistent with  the  results  of previous unpublished tests carried  out
by  Stockwell (private  communication).
                                    276

-------
Table 37.  Results of Sensitivity Test Calculations Using the Evaluated
           RADM Mechanism as the Test Mechanism, Relative to the
           "Detailed Radical" Mechanism as the Standard
Test Problem (a)
Env Mix NOX C/N
	 Calculated Concentration (ppm) (b) 	
— Maximum —    	 Average 	
Standard  % Chg    Standard  % Chg  % /Max   % Fit
Rural O^
Maximum (c) 6.50E-02
Average (d)
Avg. Abs. Value (e)
Urban Oo
Maximum (c) 3.47E-01
Average (d)
Avg. Abs. Value
Rural H202
Maximum (c) 2.41E-03
Average (d)
Avg. Abs. Value (e)
Urban H202
Maximum (c) 1.48E-02
Average (d)
Avg. Abs. Value (e)
SA AC 67 3 2.58E-07
SA DC 67 3 2.73E-07
WS BT 67 3 2.05E-08
Rural SULF
Maximum (c) 2.20E-03
Average (d)
Avg. Abs. Value (e)
Urban SULF
Maximum (c) 6.25E-03
Average (d)
Avg. Abs. Value (e)

0.8
0.0
0.1

-1.6
-0.2
0.2

-2.2
0.1
0.7

-16.5
-1.4
1.5
-14.9
-5.3
-16.5

-1.4
0.0
0.5

-1.9
-0.3
0.4

4.96E-02 0.6
0.1
0.1

2.22E-01 -1.0
-0.1
0.2

1.85E-03 3.2
0.2
0.5

1.20E-02 -13.3
-1.2
1.4
1.06E-07 -13.3
1.15E-07 -11.2
6.64E-09 -5.1

1.53E-03 -1.3
0.1
0.4

4.26E-03 -1-6
-0.2
0.3

0.3
0.0
0.0

-0.6
0.0
0.1

-1.5
0.0
0.1

-2.3
-0.2
0.2
0.0
0.0
0.0

0.6
0.0
0.0

-0.3
0.0
0.0

0.6
0.1

1.1
0.1

3.1
0.3

14.3
1.0
14.3
12.0
12.3

1.3
0.2

1.6
0.2
                                                                (continued)
                                     277

-------
Table 37 (continued) - 2
Test Problem (a)
Env Mix NOV C/N
A
Rural HN03
Maximum (c)
Average (d)
Avg. Abs. Value
Urban HN03
Maximum (c)
Average (d)
Avg. Abs. Value
Rural ORA1
Maximum (c)
Average (d)
Avg. Abs. Value
SA AL -2 6
Urban ORA1
Maximum (c)
Average (d)
Avg. Abs. Value
Rural ORA2
Maximum (c)
Average (d)
Avg. Abs. Value
WS AL -2 20
SA AL -2 60
Urban ORA2
Maximum (c)
Average (d)
Avg. Abs. Value
SS DC 300 10


	 Maxim
Standard

1.96E-03
(e)

8.44E-02
(e)

1.85E-04
(e)
9.81E-08

4.52E-03
(e)

2.68E-04
(e)
2.55E-06
2.12E-05

6.46E-03
(e)
5.87E-03

alculate
urn ---
I Chg

3.1
0.2
0.8

-4.6
-0.5
0.6

5.1
0.1
0.7
5.1

0.6
0.1
0.2

-4.5
-0.7
1.2
0.1
-4.5

-5.3
-1.6
1.6
-5.3

id Concentration (ppm) (b)
Standard % Chg

9.86E-04 2.1
0.2
0.6

3.36E-02 -2.9
-0.5
0.5

1.38E-04 4.7
0.1
0.7
7.85E-08 4.7

3.00E-03 0.5
0.1
0.1

2.10E-04 -8.2
-0.9
1.2
2.12E-06 -8.2
1.02E-05 -5.6

3.52E-03 -3.9
-1.3
1.3
3.16E-03 -3.9

% /Max

0.4
0.0
0.0

-0.6
0.0
0.1

-0.7
0.0
0.0
0.0

0.2
0.0
0.0

-0.6
0.0
0.0
-0.1
-0.3

-3.5
-0.3
0.3
-3.5


% Fit

2.2
0.3

2.9
0.4

4.6
0.3
4.6

0.5
0.1

8.7
0.6
8.7
5.8

4.0
0.8
4.0
(continued)
                                    278

-------
Table 37 (continued) - 3
Test Problem (a)
Env Mix NOX C/N
	 Calculated Concentration  (ppm)  (b)  	
— Maximum 	    	 Average  	
Standard  % Chg    Standard  % Chg   %  /Max    % Fit
Rural OP1
Maximum (c)
Average (d)
Avg. Abs. Value
Urban OP1
Maximum (c)
Average (d)
Avg. Abs. Value
SA AC 67
SA DC 67
WS BT 67
Rural OP2
Maximum (c)
Average (d)
3
3
3



Avg. Abs. Value
WS AL -2
SA AL -2
WS AI -2
Urban OP2
Maximum (c)
Average (d)
6
6
6



Avg. Abs. Value
SS AC 67
SS AC 300
WS AC 67
WS AC 67
WS AC 67
SA AC 67
SS DC 67
SS DC 300
WS DC 67
WS DC 67
WS DC 67
SA DC 67
SS BT 67
SS BT 300

10
10
3
10
30
3
10
10
3
10
30
3
10
10

1.72E-03
(e)
1.38E-02
(e)
7.44E-07
5.29E-07
6.42E-07

1.59E-03

(e)
6.40E-06
5.32E-06
4.72E-06

3.42E-02

(e)
3.66E-03
1.39E-02
4.29E-07
1.74E-03
7.97E-03
7.20E-07
4.28E-03
1.62E-02
4.36E-07
2.07E-03
9.08E-03
6.50E-07
2.31E-03
7.86E-03

-3.1
-0.2
0.8
73.4
-0.6
3.6
-5.4
73.4
-54.1

-8.4
0.3
1.4
5.9
5.5
-8.4

2246.7
56.8
61.9
-9-9
-12.8
63.6
-5.4
-7.3
83.7
-11.1
-13-1
51.2
-6.7
-9.1
173.7
-9.3
-13.6

1.25E-03
8.38E-03

3.58E-07
2.55E-07
1.20E-07

1.28E-03


3.78E-06
2.12E-06
3.44E-06

2.63E-02


1.65E-03
6.38E-03
9.43E-08
1.03E-03
5.63E-03
2.41E-07
1.88E-03
7.33E-03
8.53E-08
1.24E-03
6.40E-03
2.34E-07
1.29E-03
4.38E-03

-3.3
-0.1
0.6
57.3
-0.6
2.8
-6.3
57.3
-33.1

-4.1
0.2
1.1
2.2
3.1
-4.1

1623.2
39.8
42.6
-3.6
-5.3
54.2
-4.5
-6.0
101.7
-4.4
-5.5
50.3
-5.7
-7.8
149.8
-2.9
-4.6

-0.4
0.0
0.0
-1.4
-0.2
0.2
0.0
0.0
0.0

1.1
0.0
0.0
0.0
0.0
0.0

-1.9
-0.2
0.2
-0.2
-1.3
0.0
-0.2
-1.3
0.0
-0.3
-1.5
0.0
-0.3
-1.9
0.0
-0.1
-0.8

3.3
0.3
58.3
1.9

6.5
44.5
58.3

9.3
0.7

2.3
3.4
9.3

178.7
6.9

3.7
5.5
44.9
4.7
6.2
67.2
4.5
5.7
40.9
5.9
8.1
86.8
3.0
4.7
(continued)
                                     279

-------
Table 37 (continued) - 4
Test Problem (a)
Env Mix NOV C/N
A
WS BT 67 3
SA BT 67 3
Rural PAA
Maximum (c)
Average (d)
Avg. Abs. Value
Urban PAA
Maximum (c)
Average (d)
Avg. Abs. Value
Rural PAN
Maximum (c)
Average (d)
Avg. Abs. Value
Urban PAN
Maximum (c)
Average ( d )
Avg. Abs. Value

— Maximum 	
Standard % Chg
1.10E-07 2246.7
7.64E-08 582.8

1.87E-03 -2.9
0.2
(e) 0.9

1.05E-02 3.4
0.5
(e) 0.9

1.87E-03 2.1
0.0
(e) 0.4

4.10E-02 0.9
0.1
(e) 0.2
;d Concentration (ppm) (b)
f» W W » U^V-
Standard % Chg
2.18E-08 1623.2
4.90E-08 241.5

1.24E-03 3.2
0.2
0.6

6.31E-03 3.5
0.4
0.8

1.26E-03 2.3
0.3
0.6

3.28E-02 1.3
0.1
0.3
I /Max
0.0
0.0

-0.4
0.0
0.0

1.1
0.1
0.1

-0.5
0.0
0.0

-0.4
0.0
0.0

% Fit
178.7
109.7

3.1
0.3

3.5
0.6

2.5
0.4

2.3
0.3
 (a)  The  test problems are as  indicated  in Table 34.  The following
     notation is employed:  (1) "Env"  =  environmental condition, where
     SS = Summer surface, WS = Winter  Surface, SA  = Summer Aloft, and
     SD = Summer Dynamic.  (2) "Mix" = ROG surrogate, where "AL" = aloft
     mixture, "AI" = aloft + isoprene, "AC" = allcity average,  "DC"  =
     Washington DC mixture, and "BT" = Beaumont, Texas, mixture.  (3)
     "NO" is the effective NOV level in  ppb.  (4)  C/N is the ROG/NOV ratio.
       Ji                     A                                     A

 (b)  See  Section 6.2 for definitions of  the statistical measures.

 (c)  The  values given in the "standard"  column are the highest  maximum
     or average concentrations in  the  set of standard calculations.  The
     values given in the "J Chg,"  "% /Max" and "%  Fit" columns  are the
     comparison statistics for the calculations where the two mechanisms
     are  the most different.

 (d)  Average of the comparison statistics for all  the calculations.

 (e)  Average of the absolute value for the comparison statistics for
     all  the calculations.
                                    280

-------
     The  results  of  the  test calculations  comparing the  version  of the
RADM mechanism employing the  1986 SAPRC peroxy radical representation with
the  detailed  radical  version  are summari2ed  in Table  38.   As  with the
previous  comparison,  the  detailed   radical  mechanism   is  used  as  the
standard, and  the  results  for individual  test problems are shown only for
those  cases where  differences are  greater  than 5%.    For  most  of the
species  whose  results  are  compared,  this  version of the RADM mechanism
also  gave  predictions  very   close  to  those  for  the   detailed  radical
version.    However,   this  approximation  did  not   perform  as  well  in
simulating  the  predictions of the detailed radical  model  for the higher
organic  peroxides  (OP2) or  the higher organic acids  (ORA2) in some of the
test problems, particularly for some  of the urban  cases.   In the case of
OP2, the  approximate  mechanism had a  slight tendency (generally less than
5%)  towards predicting  lower levels of  OP2  in  those  simulations  where
significant  quantities are predicted,  with  a few  cases  of much higher
relative  predictions  for  some  test  problems where very  low  (sub-ppt)
levels  of OP2 are  predicted.   Since the large  relative differences are
generally   restricted   to   cases  where  OP2  predictions  are   generally
negligible, this  is probably  not  a significant discrepancy.  On the  other
hand,  the disagreements for  ORA2 are  significant,  since  the  approximate
model predicted from  0-40J higher ORA2  levels  in  cases where  its  formation
is non-negligible.
     The  discrepancies  in the  higher organic acid and (to a lesser  extent)
higher organic peroxide predictions can be  attributed to the  fact  that the
SAPRC approximation requires  assuming  the same rate  constants  for  similar
types of  peroxy +  peroxy reactions while  the  detailed model uses  a  variety
of  rate  constants for these  reactions.   In particular,   the  most  signifi-
cant discrepancies  for ORA2 are  due  to the  fact that  the RADM  mechanism
uses different  rate constants for RCO^ + R02 reactions depending on the
nature of R02.   These differences can be  represented in the  detailed RADM
mechanism,  but not  in  the  version using the SAPRC approximation.   If  it  is
assumed  that the  rate  constants for all RCOo  + R02 reactions  are  the  same,
then the mechanism using the  SAPRC approximation performs much  better  in
simulating  the ORA2 and OP2 results than the detailed version.
                                    281

-------
Table 38.  Results of Sensitivity Test Calculations for the Version of
           the RADM Mechanism Using the 1986 SAPRC Peroxy Radical
           Representation as the Test Mechanism, Relative to the
           "Detailed Radical" Mechanism as the Standard
Test Problem (a)
Env Mix NOX C/N
	 Calculated Concentration (ppm) (b) 	
	 Maximum 	    	 Average 	
Standard  % Chg    Standard  % Chg  % /Max   % Fit
Rural Oo
Maximum (c)
Average (d)
Avg. Abs. Value
Urban 0^
Maximum (c)
Average (d)
Avg. Abs. Value
Rural H202
Maximum (c)
Average (d)
Avg. Abs. Value
Urban H202
Maximum (c)
Average (d)
Avg. Abs. Value
SA AC 67 3
WS DC 67 3
SA DC 67 3
WS BT 67 3
Rural SULF
Maximum (c)
Average (d)
Avg. Abs. Value
Urban SULF
Maximum (c)
Average (d)
Avg . Abs . Value

6.50E-02
(e)

3.47E-01
(e)

2.41E-03
(e)

1.48E-02
(e)
2.58E-07
3.12E-07
2.73E-07
2.05E-08

2.20E-03
(e)

6.25E-03
(e)

-1.0
0.0
0.2

-2.3
-0.7
0.8

-4.4
-0.2
0.8

-21.7
-1.1
2.1
-21.7
-15.8
-16.1
2.6

-2.4
-0.3
0.6

-2.2
-0.1
0.4

4.96E-02 -1.1
0.0
0.2

2.22E-01 -2.1
-0.7
0.7

1.85E-03 3.6
-0.3
0.7

1.20E-02 -18.7
-0.8
2.1
1.06E-07 -17.3
3.93E-08 -12.4
1.15E-07 -18.7
6.64E-09 12.9

1.53E-03 -1.5
-0.2
0.5

4.26E-03 1.3
0.1
0.5

-0.5
0.0
0.0

-1.6
-0.2
0.2

-1.6
-0.1
0.1

2.4
0.0
0.2
0.0
0.0
0.0
0.0

0.3
0.0
0.0

-0.7
0.0
0.1

1.1
0.1

2.1
0.5

3.5
0.4

20.5
1.4
19.0
13-7
20.5
12.3

1.5
0.2

1.4
0.3
                                                               (continued)
                                     282

-------
Table 38 (continued) - 2
Test Problem (a)
Env Mix NOX C/N
	 Calculated Concentration (ppm)  (b)  	
— Maximum —    	  Average 	
Standard  % Chg    Standard  % Chg  % /Max    %  Fit
Rural HNO^
Maximum (c)
Average (d)



Avg. Abs. Value
Urban HN03
Maximum (c)
Average (d)



Avg. Abs. Value
SD AC 67
SD DC 67
SD BT 67
Rural ORA1
Maximum (c)
Average (d)
30
30
30



Avg. Abs. Value
SA AL -2
SD AI -2
SD AI -2
Urban ORA1
Maximum (c)
Average (d)
20
6
20


1.

(e)

8.

(e)
5.
5.
5.

1.

(e)
5.
3.
2.


96E-03



44E-02


45E-03
59E-03
83E-03

85E-04


12E-07
58E-06
03E-05

4.52E-03

-4.
0.
1.

9.
0.
1.
9.
6.
8.

-6.
-0.
1.
-5.
-6.
-6.


1
2
1

3
2
0
3
6
0

6
3
1
1
1
,6

-1.4

9.



3.


2.
2.
2.

1.


3.
1,
1,


86E-04



36E-02


61E-03
68E-03
79E-03

38E-04


.79E-07
.99E-06
. 11E-05

3.00E-03
-0.4
Avg. Abs. Value
SA AL -2
SD AI ~2
SD AI -2
Rural ORA2
Maximum (c)
Average (d)
20
6
20

(e)
5,
0.4
. 12E-07
3.58E-06
2.03E-05


2.68E-04

Avg. Abs. Value
SS AL < 1
S3 AL < 1
20
66

(e)
3
1


.80E-06
.34E-05
-5
-6
-6

102
25
25
12
33
.1
.1
.6

.8
.5
.5
.3
.5
3
1
1

2


2
1
.79E-07
.99E-06
. 11E-05

. 10E-04


.92E-06
.04E-05

-2.
0.
0.

5.
0.
0.
5.
3.
5.

-5.
-0.
1.
-4,
-5,
-4

-1
-0
. 0
-4
-5
-4

116
24
24
13
34

9
0
8

6
0
7
6
8
0

,7
.2
.0
.1
.7
.1

.9
.3
.4
.1
.7
.1

.1
.8
.8
.2
.9

-0.5
0.0
0.1

-1.1
-0.1
0.1
0.4
0.3
0.4

1.3
0.0
0.0
0.0
-0.1
-0.3

-0.6
-0.1
0.1
0.0
-0.1
-0.3

20.0
1.5
1.5
0.2
1.7

3.3
0.4


5.5
0.5

5.5
3.8
4.9

5.9
0.4

4.2
5.9
4.3

1.1
0.2

4.2
5.9
4.3

73.4
8.2

12.3
29.7
                                                               (continued)
                                     283

-------
Table 38 (continued) - 3
Test Problem (a)
Env Mix NOX C/N
SS
SS
SS
SS
SS
SS
SS
us
ws
SA
SA
SA
SD
SD
SD
SS
SS
SS
SS
SS
SS
SS
WS
WS
us
SA
SA
SA
SD
SD
SD
AL
AL
AL
AL
AL
AL
AL
AL
AL
AL
AL
AL
AL
AL
AL
AI
AI
AI
AI
AI
AI
AI
AI
AI
AI
AI
AI
AI
AI
AI
AI
<1
-2
-2
-2
5
5
5
-2
_2
-2
-2
-2
~2
-2
-2
<1
<1
<1
-2
-2
5
5
~2
-2
~2
-2
-2
-2
-2
-2
-2
200
6
20
60
2
7
20
20
60
6
20
60
6
20
60
20
66
200
20
60
7
20
6
20
60
6
20
60
6
20
60

	 Maxin
Standard
4.14E-05
7.68E-06
2.47E-05
7.19E-05
6.21E-06
2.23E-05
7.44E-05
2.55E-06
1.18E-05
1.01E-07
1.33E-06
2.12E-05
2.44E-06
9.98E-06
3.04E-05
1.22E-05
5.39E-05
1.88E-04
6.62E-05
2.68E-04
4.92E-05
2.18E-04
1.06E-05
3.42E-05
1.80E-04
1.94E-07
6.55E-06
1.62E-04
6.37E-06
3.36E-05
1.24E-04
Calculate
mm —
* Chg
71.2
7.4
25.7
54.3
10.5
21.3
44.9
13.4
45.2
16.5
55.3
102.8
8.0
23.8
50.5
9.2
17.4
26.8
12.8
22.6
9.3
19.6
5.0
8.9
10.7
7.9
3L9
39.3
31.9
31.9
30.6
;d Concentra
	 	 a
Standard
3.15E-05
4.63E-06
1.58E-05
4.79E-05
4.24E-06
1.56E-05
5.31E-05
2.12E-06
8.46E-06
4.03E-08
6.82E-07
1.02E-05
1.13E-06
5.02E-06
1 .52E-05
6.37E-06
2.80E-05
1.06E-04
5.23E-05
2.10E-04
3.37E-05
1.37E-04
6.48E-06
2.50E-05
1.37E-04
1.59E-07
5.31E-06
7.78E-05
3.90E-06
1.94E-05
6.94E-05
ition (p
iverage
* Chg
71.7
6.3
28.5
59.0
10.8
20.3
42.3
11.1
33.3
24.9
63.7
116.1
7.6
22.0
46.1
8.5
17.1
27.8
10.7
20.0
7.8
16.9
4.9
7.9
9.8
10.8
31.3
38.4
16.9
21.1
24.5
pm) (b)
% /Max
10.8
0.1
2.1
13.4
0.2
1.5
10.7
0.1
1.3
0.0
0.2
5.6
0.0
0.5
3.3
0.3
2.3
14.1
2.7
20.0
1.2
11.0
0.2
0.9
6.4
0.0
0.8
14.2
0.3
1.9
8.1

% Fit
52.7
6.2
24.9
45.5
10.2
18.4
34.9
10.6
28.7
22.1
48.3
73.4
7.3
19.8
37.5
8.2
15.8
24.4
10.2
18.2
7.5
15.6
4.8
7.7
9.4
10.2
27.1
32.2
15.6
19.1
21.8
 Urban ORA2

 Maximum (c)
 Average (d)
 Avg. Abs. Value (e)

  SS  AC  20  10
  SS  AC  20  33
  SS  AC  20 100
  SS  AC  67   3
  SS  AC  67  10
  SS  AC  67  30
6.46E-03


4.22E-04
1 .44E-03
4.19E-03
2.05E-04
1.79E-03
4.94E-03
82.4
37.4
37.4
39.9
59.5
40.7
6.6
50.1
57.1
3.52E-03


2.53E-04
1.03E-03
3.22E-03
1.50E-04
1.00E-03
3.39E-03
76.9
31.5
31.5
33.6
54.4
38.5
14.8
47.1
52.8
60.7
7.3
7.3
2.4
15.8
35.2
0.6
13.4
50.8
55.6
15.5

28.8
42.9
32.4
13.9
38.2
41.9
                                                               (continued)
                                     284

-------
Table 38 (continued) - 4
Test Problem
Env Mix NOV
A
SS
ss
SS
ws
W2
SA
SA
SD
SD
SD
SS
SS
SS
SS
SS
SS
SS
SS
SS
ws
ws
SA
SA
SD
SD
SD
SS
SS
SS
SS
SS
SS
SS
SS
SS
ws
ws
SA
SA
SD
SD
SD

AC
AC
AC
AC
AC
AC
AC
AC
AC
AC
DC
DC
DC
DC
DC
DC
DC
DC
DC
DC
DC
DC
DC
DC
DC
DC
BT
BT
BT
BT
BT
BT
BT
BT
BT
BT
BT
BT
BT
BT
BT
BT

100
20Q
300
67
67
67
67
67
67
67
20
20
20
67
67
67
100
200
300
67
67
67
67
67
67
67
20
20
20
67
67
67
100
200
300
67
67
67
67
67
67
67

(a)
C/N
2
3
10
10
30
10
30
3
10
30
10
33
100
3
10
30
2
3
10
10
30
10
30
3
10
30
10
33
100
3
10
30
2
3
10
10
30
10
30
3
10
30


._ _ c
	 Maxim
Standard
1.
.94E-04
6.42E-04
6.46E-03
7
.27E-04
3.39E-03
4
4
9
4
2
3
1
3
1
1
4
1
5
5
4
2
3
3
6
3
1
3
1
3
1
1
4
1
6
5
5
2
3
.89E-04
.84E-03
.29E-05
.79E-04
.04E-03
.71E-04
.25E-03
.25E-03
.38E-04
.64E-03
.23E-03
.25E-04
.50E-04
.87E-03
.69E-04
.35E-03
.57E-04
.70E-03
.87E-05
.78E-04
.58E-03
.58E-04
.21E-03
.44E-03
.73E-04
.52E-03
.22E-03
.63E-04
.38E-04
-53E-03
.71E-04
.59E-03
.45E-04
3.50E-03
7
4
1

.94E-05
.02E-04
.68E-03

alculat<
urn 	
* Chg
0.6
46.9
54.7
1.9
8.0
37.7
47.5
26.4
68.4
66.6
53.0
76.4
58.6
43.4
60.8
75.2
27.1
59.1
67.4
18.8
17.4
60.6
64.4
31.0
78.2
82.4
34.5
51.2
39.0
10.7
40.9
51.5
0.1
34.0
45.3
0.2
4.9
22.4
44.1
24.9
65.7
65.6

3d Concentra
Standard
1.
4.
3.
5.
2.
3.
3.
5.
2.
1.
2.
8.
2.
1.
9.
2.
8.
3.
04E-04
66E-04
52E-03
02E-04
29E-03
79E-04
09E-03
28E-05
55E-04
09E-03
17E-04
51E-04
44E-03
07E-04
02E-04
78E-03
09E-05
45E-04
3.16E-03
3.46E-04
1.64E-03
2.34E-04
2.11E-03
3.84E-05
1.98E-04
8.33E-04
2.20E-04
8.41E-04
2.63E-03
1.31E-04
8.62E-04
2.78E-03
8,
. 17E-05
3.99E-04
3.00E-03
3
1
2
2
4
2
8

.81E-04
.75E-03
.98E-04
.24E-03
.49E-05
. 12E-04
.83E-04

tion (p
verage
% Chg
9.2
24.9
54.6
12.5
16.0
11.9
22.8
18.6
53.5
59.2
45.2
73.7
57.0
22.5
58.1
73.2
16.5
36.7
67.6
22.0
26.4
23.9
36.6
23.5
65.2
76.9
30.7
47.3
36.5
11.2
. 40.6
47.7
6.9
16.3
47.5
6.3
11.9
6.1
19.6
17.2
50.1
56.9

>pm) (b)
% /Max
0.3
3.3
54.6
1.8
10.4
1.3
20.0
0.3
3.9
18.3
2.8
17.8
39.5
0.7
14.9
57.8
0.4
3.6
60.7
2.2
12.3
1.6
21.9
0.3
3.7
18.2
1.9
11.3
27.3
0.4
9.9
37.5
0.2
1.8
40.4
0.7
5.9
0.5
12.5
0.2
3.0
14.3


% Fit
8.8
22.3
42.9
11.8
14.9
11.4
20.7
17.0
42.2
45.7
36.9
53.9
44.5
20.4
45.0
53.7
15.3
31.2
50.6
19.9
23.4
21.6
31.2
21.0
49.2
55.6
26.7
38.4
31.1
10.8
33.8
38.6
6.8
15.3
38.4
6.2
11.3
6.1
18.1
15.9
40.1
44.3
(continued)
                                    285

-------
Table 38 (continued) - 5
Test Problem (a)
Env Mix NOX C/N
	 Calculated Concentration (ppm) (b) 	
	 Maximum —    	 Average 	
Standard  % Chg    Standard  J Chg  % /Max   % Fit
Rural OP1
Maximum (c)
Average (d)
Avg.
SS
SS
SS
SS
US
SA
SS
SS
SS
WS
US
Abs.
AL
AL
AL
AL
AL
AL
AI
AI
AI
AI
AI
. Value <
<1
<1
5
5
~2
-2
<1
5
5
-2
_2
20
66
2
7
20
6
20
2
7
6
60
1.72E-03
:e)
4.
4.
6.
6.
4.
2.
4.
6.
6.
3.
2.

01E-04
03E-04
79E-04
61E-04
24E-05
73E-05
08E-04
97E-04
95E-04
43E-05
63E-04
7.9
2.9
3.5
5.8
5.0
7.8
7.5
4.6
-6.5
7.0
7.9
6.9
2.2
5.2
1.

2.
2.
3.
3.
2.
8.
3.
3.
4.
25E-03

58E-04
85E-04
62E-04
84E-04
48E-05
98E-06
12E-04
79E-04
06E-04
1.89E-05
1.60E-04
56.9
2.8
3.2
3.7
3.0
4.3
4.3
5.5
-5.5
4.3
4.1
3.4
6.9
15.0
1.9
0.2
0.2
0.8
0.7
1.3
1.3
0.1
0.0
1.1
1.3
1.1
0.1
1.9
13.8
1.3

3.6
3.0
4.5
4.2
5.4
6.6
4.5
4.1
3.4
6.7
13.8
Urban OP1
Maximum
Average
Avg.
SS
SS
WS
US
US
SA
SA
SS
SS
SS
us
SA
SA
SS
US
us
us
SA
SA
Abs
AC
AC
AC
AC
AC
AC
AC
DC
DC
DC
DC
DC
DC
BT
BT
BT
BT
BT
BT
(c)
(d)
1.38E-02

. Value
67
300
67
67
67
67
67
67
100
300
67
67
67
300
67
67
67
67
67
10
10
3
10
30
3
30
10
2
10
3
3
30
10
3
10
30
3
30

(e)


1.67E-03
4.82E-03
1
5
2
7
1
1
6
4
1
5
1
3
6
3
1
.72E-06
.24E-04
.21E-03
.44E-07
.38E-02
.66E-03
.50E-04
.74E-03
.34E-06
.29E-07
.27E-02
.68E-03
.42E-07
.68E-04
.78E-03
2.77E-07
1
.09E-02
404.3
29.7
30.0
7.5
12.5
78.5
6.3
4.0
286.3
9.0
8.6
6.1
14.7
197.9
237.4
9.1
6.0
258.6
8.7
3.1
404.3
8.5
8.38E-03


1
2
3
3
1
3
8
1
2
3
2
2
7
2
1
1
1
1


. 11E-03
.95E-03
.74E-07
.21E-04
•55E-03
.58E-07
.38E-03
. 12E-03
.88E-04
.OOE-03
.92E-07
.55E-07
.52E-03
.37E-03
.20E-07
.79E-04
.18E-03
.53E-07
6.42E-03
417.2
3L3
31.8
-1.1
-0.7
125.4
10.4
5.3
273.2
3.9
-1.4
2.3
-0.5
202.6
233.4
4.0
-0.6
417.2
17.9
5.4
367.6
3.6
3.9
0.1
0.2
-0.1
-0.3
0.0
0.4
1.0
0.0
3.9
-0.2
0.1
-0.2
0.0
0.0
3.6
-0.2
0.0
0.4
0.8
0.0
2.7
135.1
8.9

3.7
6.1
78.2
9.8
5.1
115.5
3.9
4.3
2.8
6.6
100.6
107.6
4.0
5.5
135.1
16.4
5.3
129.5
3.6
                                                                (continued)
                                     286

-------
Table 38 (continued) - 6
Test Problem (a)
Env Mix NOV C/N
          A
	 Calculated Concentration (ppra)  (b)  	

— Maximum —    	  Average 	

Standard  % Chg    Standard  f. Chg  % /Max    % Fit
Rural OP2
Maximum (c)
Average (d)
Avg.
SA
WS
Urban
Abs
AL
AI
. Value
-2
-2
6
6
1.
(e)
.59E-03

5.32E-06
4.72E-06
-7.5
-0.8
1.6
-7.5
6.2
1.

2.
3-
28E-03

12E-06
44E-06
17.5
-0.6
2.1
-7.7
17.5
-1.3
-0.1
0.1
0.0
0.0
16.0
0.9

8.3
16.0
OP2
Maximum
Average
Avg.
SS
SS
SS
SS
SS
SS
SS
WS
WS
WS
SA
SA
SA
SS
SS
SS
SS
SS
SS
SS
SS
WS
WS
WS
SA
SA
SS
SS
SS
Abs
AC
AC
AC
AC
AC
AC
AC
AC
AC
AC
AC
AC
AC
DC
DC
DC
DC
DC
DC
DC
DC
DC
DC
DC
DC
DC
BT
BT
BT
(c)
(d)


. Value
20
20
20
67
67
200
300
67
67
67
67
67
67
20
20
20
67
67
100
200
300
67
67
67
67
67
20
20
20
10
33
100
10
30
3
10
3
10
30
3
10
30
10
33
100
10
30
2
3
10
3
10
30
3
30
10
33
100
3

(e)
8
5
1
3
1
1
1
4
1
7
7
4
3
9
6
1
4
2
2
1
1
4
2
9
6
3
6
4
1
.42E-02


. 10E-04
.97E-03
.76E-02
.66E-03
.87E-02
.55E-03
.39E-02
.29E-07
.74E-03
.97E-03
.20E-07
.08E-03
.34E-02
.23E-04
.68E-03
.86E-02
.28E-03
.09E-02
.22E-04
.78E-03
.62E-02
.36E-07
.07E-03
.08E-03
.50E-07
.42E-02
.05E-04
.41E-03
.44E-02
408.0
7.4
24.7
-15.0
-8.3
-8.6
-27.4
-14.8
-3.1
-3^.6
32.5
-12.2
-13.8
99.8
-2.9
-6.3
-16.8
-10.2
-9.3
-28.5
-15.3
-4.2
-3.5
-3M.O
62.2
-13.2
-15.2
140.9
-6.3
-12.7
-6.4
-6.8
2.63E-02




4.26E-04
4.23E-03
1.
1 ,
1.
.34E-02
.65E-03
.30E-02
6.95E-04
6.38E-03
9.43E-08
1.
5
2
2
2
4
4
1
1
1
9
8
7
8
1
6
2
2
3
3
1
.03E-03
.63E-03
.41E-07
.43E-03
.53E-02
.67E-04
.69E-03
.42E-02
.88E-03
.45E-02
.35E-05
.07E-04
.33E-03
.53E-08
.24E-03
.40E-03
.34E-07
.63E-02
.81E-04
.17E-03
.10E-02
311.2
10.1
23.2
-7.5
-8.3
-7.9
-14.8
-11.5
-6.2
-21.3
36.5
-11.1
-12.2
125.5
-0.9
-3.8
-7.7
-9.0
. -8.5
-15.7
-12.0
-6.2
-6.5
-21.5
73.6
-12.1
-13.6
136.6
-4.1
-6.1
-7.3
-7.0
-6.6
-0.8
0.8
-0.1
-1.3
-4.0
-0.9
-5.7
-0.2
-5.2
0.0
-0.4
-2.6
0.0
-0.1
-3.7
-0.1
-1.6
-4.6
-1.1
-6.6
0.0
-0.2
-6.0
0.0
-0.6
-3.3
0.0
-4.1
-0.1
-0.9
-2.9
121.8
9.5

7.8
8.6
8.3
16.0
12.2
6.8
23.8
36.5
11.8
13.0
77.1
0.9
3.9
8.0
9.5
8.9
17.1
12.8
6.4
7.0
24.1
53.8
12.9
14.6
81.3
4.2
6.3
7.6
7.3
                                                               (continued)
                                     287

-------
Table 38 (continued)  - 7
Test Problem (a)
Env Mix NOX C/N
SS BT 67
SS BT 67
SS BT 200
SS BT 300
WS BT 67
WS BT 67
WS BT 67
SA BT 67
SA BT 67
Rural PAA
Maximum (c)
Average (d)
10
30
3
10
3
10
30
3
30



Avg. Abs. Value
WS AL -2
SA AL -2
Urban PAA
Maximum (c)
Average (d)
6
6


p
	 Maxim
Standard
2.
1.
8.
7.
1.
1.
5.
7.
2.

1.

(e)
31E-03
39E-02
26E-04
86E-03
10E-07
02E-03
41E-03
64E-08
76E-02

87E-03


3.29E-06
3.34E-06


1.05E-02

Avg. Abs. Value
WS AC 67
WS AC 67
SA AC 67
HS DC 67
WS DC 67
WS DC 67
SA DC 67
SS BT 300
WS BT 67
SA BT 67
Rural PAN
Maximum (c)
Average (d)
3
30
3
3
10
30
3
10
30
30



Avg. Abs. Value
SD AL -2
SD AI -2
60
6

(e)
3
3
6
5
7
2
7
4
2
7

1

(e)


.31E-08
.01E-03
.38E-08
.56E-08
. 17E-04
.96E-03
.26E-08
.07E-03
,51E-03
.94E-03

.87E-03


7.41E-05
2.04E-05
alculate
ium 	
% Chg
-28.1
-14.1
-6.8
-30.8
122.5
-8.5
-11.0
408.0
-5.5

-7.3
-0.2
1.8
-7.3
-7.0

-12.3
-1.0
3.0
-6.3
9.0
-6.3
-12.3
5.6
11.8
-9.7
-5.9
6.5
-6.2

-5.3
-0.8
1.0
-2.2
-5.3
id Concentra
Standard
1.
9.
3-
4.
2.
4.
3.
4.
2.

29E-03
68E-03
29E-04
38E-03
18E-08
80E-04
56E-03
90E-08
01E-02

1.24E-03




1.35E-06
1,

6


4
1
3
7
4
1
3
2
1
4

1


. 13E-06

.31E-03


.95E-09
.91E-03
. 11E-08
.11E-09
.32E-04
.93E-03
.28E-08
.54E-03
.52E-03
.45E-03

.26E-03


3.37E-05
7.42E-06
.tion (pi
iverage •
% Chg
-13.8
-10.3
-9.4
-21.1
215.1
-7.2
-10.2
311.2
-2.7

-5.0
-0.2
1.4
-5.0
-0.4

11.8
-1.0
2.4
-2.5
8.8
-2.8
-9-2
5.4
11.8
-4.4
-4.2
6.6
-2.2

-7.1
-0.8
2.2
-5.6
-6.7
3m)
(b)
% /Max
-0.
-3.
-0.
-3.
0.
-0.
-1.
0.
-2.

7
8
1
5
0
1
4
0
1

2.0
0.0


% Fit
14.
10.
9.
23.
103.
7.
10.
121.
2.

8
9
9
5
7
7
7
8
8

5.1
0.6
0.1
0.0
0.0

3
-0
0
0
2
0
0
0
3
0
-1
1
-1

-1

.6
.1
.3
.0
.6
.0
.0
.4
.6
.0
.7
.6
.6

.8
0.0
5.1
1.7

11
1

3
8
2
9
5
11
4
5
6
2


.1
.6

.8
.8
.9
.8
.4
.1
.5
.4
.5
.3

9.3
1
.0
0.1
-0.1
0.0
5.8
7.3
                                                               (continued)
                                     288

-------
Table 38 (continued) - 8
Test Problem
Env Mix NOV
A
SD AI
SD AI
Urban PAN
Maximum (
Average (
Avg. Abs.
SD AC
SD DC
SD BT
-2
-2

c)
d)
(a)
C/N
20
60



Value
67
67
67
30
30
30


— Maxin
Standard
6
1

4

(e)
5
4
5
•33E-05
.34E-04

.10E-02


. 18E-03
.97E-03
.32E-03
^alculat
num 	
% Chg
1.8
1.1

-6.6
-1.3
1.3
-6.6
-5.0
-5.1
ed i
Concentre

Standard
2
5

3


3
3
3
.57E-05
.91E-05

.28E-02


.67E-03
.51E-03
.74E-03
ition (f
Average
% Chg
-7.1
-8.5

-8.5
-1.0
1.4
-7.0
-5.5
-6.5
jpm) (b)

% /Max
-0.1
-0.4

-2.3
-0.1
0.2
-0.8
-0.6
-0.7


%
8
9

7
1

7
5
6


Fit
.5
.3

.2
.0

.2
.6
.7
(a)  See Footnote (a) in Table 37 for a description of the notation used
     to indicate the test problems.

(b)  See Section 6.2 for definitions of the statistical measures.

(c)  The values given in the "standard" column are the highest maximum
     or average concentrations in the set of standard calculations.  The
     values given in the "% Chg," "% /Max" and "% Fit" columns are the
     comparison statistics for the calculations where the two mechanisms
     are the most different.

(d)  Average of the comparison statistics for all the calculations.

(e)  Average of the absolute value for the comparison statistics for
     all the calculations.
                                    289

-------
     The significance of this  apparent failure of the SAPRC approximation
to  reproduce  detailed  model  predictions  of  the  higher  organic  acids
depends  upon the  extent  to  which  the  detailed  mechanism taken  as  the
standard is  assumed  to  represent reality.   Had  we developed the detailed
radical  mechanism  for use  in this  evaluation,  we  would  have  estimated,
based on the limited  laboratory  data available (Atkinson,  1989), that  the
rate constants  for all the  RC^ +  RCOg  reactions  were  approximately  the
same, and  thus  the SAPRC radical  approximation  would have  performed much
better in  reproducing the  detailed model.   It  is clear  that if there  are
variations in the  RCO-j  +  RC^ rate constants to  the extent  assumed in  the
RADM mechanism,  then  the  mechanisms  utilizing  the SAPRC  approximation to
represent  peroxy  radical  +  peroxy  radical reactions  will  not  perform
satisfactorily in  predicting yields  of the  higher organic  acids. However,
if the detailed model uses incorrect estimates for these rate constants —
which may  well be  the case since few have been measured  — the  predictions
of  the  detailed  model  may  be  even  less  reliable  than  the  approximate
one.   No suitable  data  are available to test model  predictions of  organic
acids.
     In  summary,  the test calculations  on  the representation  of peroxy +
peroxy reactions  tend  to  support the approach which  is  currently employed
in the evaluated  RADM mechanism.  This does not  mean,  however,  that it  may
not  be  possible  to employ more  efficient approaches  involving  fewer reac-
tions  and  possibly fewer  species.   However, at  the  present  time no alter-
natives  in this regard are recommended for  the RADM mechanism.

6.4  Lumping of Higher Alkanes
     One area where the  RADM mechanism is  more detailed  than most other
current  condensed mechanisms  is  its  representation  of the higher (C3-O
alkanes.   However,  test  calculations  previously carried  out using  the
latest   SAPRC  mechanism  (Carter,  1988)  indicate  that  relatively severe
condensations of  alkanes  have relatively  little effect  on model  predic-
tions  if  the parameters and  rate  constants  for  the  lumped  species  are
adjusted appropriately based  on the  mixtures  being  represented.   There-
fore,  as part of  this study we  carried  out test calculations   to determine
whether  a moderate level of  condensation of the RADM mechanism, employing
two  lumped  species to  represent the C3+ alkanes  rather than   three, would
                                    290

-------
have a  significant  effect  on model  predictions relative  to  the  current
version.
     The  condensation  examined was  to lump  the  species HC5  and HC8  and
represent them by a  single species, designated HC6.   For direct  comparison
with the March  1988  RADM  mechanism,  the  OH  radical  rate  constant  and
product yield parameters for  the new  lumped species  were determined by  the
averages  of those for  HC5 and HC8  in the March 1988 mechanism,  weighed by
the estimated relative numbers of  moles of each reacting in RADM applica-
tions.    These were  derived  using  the  total  detailed U.S.  emissions,  and
assuming  an integrated OH  radical  parameter  (INTOH) value of  110 ppt-min,
using the methods  and  data  bases discussed in Section 3.2.  The reaction
used for  HC6 was,

        HC6 + HO —> HC6P * 0.54 X02 + H20

where k  = 3.02  x 10~11 exp  (-380/T)  cm3 molecule'1 sec"1.  The reactions
used for  HC6P, its associated peroxy peroxy radical, were

       HC6P + NO —> 0.4 ALD + 0.91 KET +1.17 ONIT +  0.03 HCHO  +
                     0.83 N02 + 0.83 H02

      HC6P + H02 —> OP2

     HC6P -t- AC03 —> 0.12 ALD + 0.88 KET + 0.5 H02 +  0.5 M02 + 0.5  ORA2

      HC6P + M02 —> 0.75 HCHO + H02 +  0.089  ALD +0.66 KET

where  their  rate  constants  were   the same  as  those used for HC5P  and
HC8P.   The only  other modifications  made  to the  RADM mechanism  was  the
removal of the reactions of HC5, HC5P,  HC8, and HC8P.
     (Note that  the  kinetic and product yield parameters used  for HC6  and
HC6P in  the  modified mechanism for  the purpose  of  this test  are slightly
different from  those used for this  species  in the  RADM-M  mechanism.   The
parameters  in  the  latter mechanism are based on a  more  recent analysis of
the emissions  data than used  to derive the parameters for  the  March  1988
version  of the  RADM mechanism used  as the  standard in  this  test.   The
                                    291

-------
parameters used in the test were  derived to be consistent with those used
in the standard mechanism for  more direct comparison.)
     A comparison  of the  predictions  using this  "condensed alkane" RADM
mechanism  with  those of the  standard  mechanism in  simulations  of the 90
test problems  is presented in Table 39.   The  results for individual test
problems are shown only  for those  cases  where  differences are greater than
5%.   These results show that despite the  fact  that these scenarios used
five  different  ROG  surrogates  with  four  different  profiles  for   the
alkanes, this  level  of  condensation had an almost  insignificant effect on
the  simulations  of the  species of  interest  in regional  oxidant  or acid
deposition models.   There are  only  a  few cases where relative  discrepan-
cies greater than  3%  were  observed,  and  in essentially all  these cases  the
predicted  concentrations of the species  were very low.
     It  is on  the  basis of these  results that we  recommend that  the RADM
mechanism  be condensed  in  its representation of these higher alkanes.   It
is possible  that  it  may also be  appropriate  to lump HC3 with  the  higher
alkane  classes.   However,  more  significant  condensations of  the RADM
mechanism  was  considered  to  be  beyond  the  scope  of  this  evaluation
program.

6.5  The Recommended  vs. March  1988  RADM Mechanisms
     As  discussed  in Section  2,  as a result  of the  tests  carried  out in
this program,  we  have   recommended  modifications  in  the  RADM  mechanism,
relative  to the March  1988,   version  which was  evaluated in this  study.
The  most important modifications  in terms of  effects on model  performance
concern  changes  in the  representation of the unknown aspects of the aroma-
tic  photooxidation mechanism  to  improve the  performance  of the mechanism
in  simulating the results of  aromatic  -  NOX  - air  chamber  experiments.
Minor  changes  in rate constant and product yields are recommended for  some
of   the  lumped species  based  on  the results of  an updated analysis of
emissions  data as discussed  in Section  3.   Other minor  changes are recom-
mended for the representation of phenol and nitrophenol + NO^ reactions to
 improve chemical  reasonableness.   In addition,  as a result of the tests
discussed  in  the previous  section,  we  recommended  that the  two  higher
alkane  species  in   the  RADM  mechanism  be   lumped  together.   All  the
                                     292

-------
Table 39.  Results of Sensitivity Test Calculations Using the Version of
           the RADM Mechanism with the Higher Alkane Species HC5 and HC8
           Lumped Together, Relative to the Standard RADM Mechanism
Test Problem (a)
Env Mix NOX C/M
	 Calculated Concentration (ppm) (b) 	
— Maximum 	    	 Average 	
Standard  % Chg    Standard  % Chg  % /Max   % Fit
Rural O^
Maximum (c) 6.49E-02
Average (c)
Avg. Abs. Value (c)
Urban Cs
Maximum (c) 3.47E-01
Average (c)
Avg. Abs. Value (c)
Rural HoC^
Maximum (c) 2.41E-03
Average (c)
Avg. Abs. Value (c)
Urban H202
Maximum (c) 1.48E-02
Average (c)
Avg. Abs. Value (c)
SA DC 67 3 2.59E-07
WS BT 67 3 1.71E-08
SA BT 67 3 4.36E-08
Rural SULF
Maximum (c) 2.21E-03
Average (c)
Avg. Abs. Value (c)
Urban SULF
Maximum (c) 6.25E-03
Average (c)
Avg. Abs. Value (c)

-1.9
-0.1
0.2

1.6
0.0
0.2

2.8
-0.1
0.6

27.4
1.1
1.5
5.1
27.4
20.2

-3.6
-0.7
0.9

-1.1
-0.2
0.4

4.97E-02 -1.4
-0.2
0.2

2.23E-01 1.3
0.1
0.2

1.82E-03 3.2
-0.1
0.6

1.20E-02 11.1
0.5
0.9
1.02E-07 0.8
6.31E-09 11.1
3.47E-08 4.5

1.53E-03 -3-9
-0.7
0.8

4.26E-03 -1.1
-0.1
0.3

-0.6
0.0
0.0

0.2
0.0
0.0

-1.5
0.0
0.1

-0.5
0.0
0.1
0.0
0.0
0.0

-1.0
0.0
0.1

-0.5
0.0
0.1

1.4
0.1

1.3
0.1

3.3
0.3

13.2
0.6
2.3
13.2
4.6

3-9
0.4

1.1
0.2
                                                                (continued)
                                     293

-------
Table 39 (continued)  - 2
Toot- D*»r»K1 ARI ^2^ __ P
— Maxim
Env Mix NOX C/N Standard
Rural HN03
Maximum (c) 1.95E-03
Average (c)
Avg. Abs. Value (c)
SS AL <1 200 6.48E-05
Urban HN03
Maximum (c) 8.46E-02
Average (c)
Avg. Abs. Value (c)
Rural ORA1
Maximum (c) 1.84E-04
Average (c)
Avg. Abs. Value (c)
Urban ORA1
Maximum (c) 4.53E-03
Average (c)
Avg. Abs. Value (c)
Rural ORA2
Maximum (c) 2.68E-04
Average ( c )
Avg. Abs. Value (c)
SA AL -2 20 1.29E-06
SA AL -2 60 2.03E-05
Urban ORA2
Maximum (c) 6.19E-03
Average (c)
Avg. Abs. Value (c)
alculatc
urn —
% Chg

-3.8
-1.3
1.4
-3.6

-1.0
-0.1
0.3

3.2
0.4
0.7

1.2
0.3
0.4

5.2
0.9
1.5
4.8
5.2

2.7
0.3
0.5
;d Concentration (ppra) (b) -
Standard \ Chg

9.87E-04 -6.9
-1.3
1.4
2.68E-05 -6.9

3.36E-02 -2.3
-0.1
0.2

1.37E-04 2.7
0.4
0.7

3.00E-03 1.0
0.3
0.3

2.11E-04 5.2
1.0
1.3
6.60E-07 5.2
9.61E-06 5.0

3.42E-03 1.1
0.3
0.4
% /Max

-1.7
-0.1
0.1
-0.2

-0.6
0.0
0.0

0.4
0.0
0.0

0.7
0.1
0.1

1.1
0.1
0.1
0.0
0.2

0.7
0.1
0.1

* Fit

7.1
0.6
7.1

0.9
0.2

2.6
0.3

1.0
0.2

5.1
0.6
5.1
4.9

1.1
0.3
                                                               (continued)
                                     294

-------
Table 39 (continued) - 3
Test Problem (a)
Env Mix NOV C/N
          A
 	 Calculated Concentration (ppm) (b)  	

 	 Maximum 	    	 Average 	

 Standard  % Chg    Standard  % Chg  % /Max    % Fit
Rural OP1
Maximum (c) 1.70E-03
Average (c)
Avg. Abs. Value (c)

5.4
0.2
0.8

1.25E-03 5.0
0.5
0.7

0.9
0.0
0.0

4.9
0.4

 SA  AL  ~2  20



Urban OP1
8.48E-05    5.4    4.61E-05    5.0
0.2
Rural OP2
Urban OP2
4.9
Maximum
Average
Avg.
SA
SA
WS
SA
Abs
AC
DC
BT
BT
(c)
(c)

1
.37E-02
. Value (c)
67
67
67
67
3
3
3
10
7
9
2
9
.04E-07
.17E-07
.95E-07
.OOE-04
124.0
1.7
3.8
-5.4
-40.2
124.0
5.4
8

3
4
8
3
-35E-03

.36E-07
.01E-07
.04E-08
.05E-04
56.5
0.8
2.5
-2.5
-35.0
56.5
6.1
0.8
0.0
0.0
0.0
0.0
0.0
0.2
56.8
1.6

2.6
42.4
56.8
5.9
Maximum
Average
Avg.
SS
ss
SS
WS
SA
SA
SD
WS
SA
Abs
AL
AL
AL
AL
AL
AL
AL
AI
AI
(c)
(c)

1
.61E-03
. Value (c)
<1
5
5
-2
-2
-2
~2
-2
-2
200
2
7
20
6
60
6
6
6
6
2
9
3
5
7
2
4
2
.31E-04
.92E-05
. 38E-05
.87E-05
.61E-06
.45E-04
.33E-05
.33E-06
.71E-06
16.2
1.4
3.6
5.7
-5.0
-4.4
8.1
-7.8
6.5
2.1
16.2
-5.9
1

4
2
7
2
2
6
1
3
1
.30E-03

.39E-04
.07E-05
. 10E-05
.56E-05
.19E-06
.60E-04 -
.37E-05
.30E-06
.40E-06
13.7
1.4
3.2
5.9
-1.7
-2.6
7.8
-5.6
6.2
5.0
13.7
-7.5
3.2
0.1
0.2
2.0
0.0
-0.1
0.2
0.0
3.2
0.1
0.0
0.0
13.0
1.5

5.7
5.1
2.8
7.5
7.2
6.0
4.9
13.0
8.3
Maximum
Average
Avg.
SS
WS
SS
(c)
(c)

3
.40E-02
Abs. Value (c)
AC
AC
DC
100
67
100
2
3
2
2
7
2
. 17E-04
.03E-07
.20E-04
-85.6
-2.2
4.3
-5.8
9.3
-6.9
2

8
1
9
.61E-02

.62E-05
.45E-07
.25E-05
-90.6
-2.3
4.6
-7.0
18.7
-7.9
1.5
0.1
0.1
0.0
0.0
0.0
165.4
4.3

7.2
17.0
8.3
                                                              (continued)
                                    295

-------
Table 39 (continued) - 4
Test Problem (a]
Env Mix NOV C/N
A
WS DC 67 3
SA DC 67 3
SS BT 200 3
WS BT 67 3
SA BT 67 3
SA BT 67 10
Rural PAA
Maximum (c)
Average (c)
Avg. Abs. Value
SA AL -2 6
Urban PAA
Maximum (c)
Average (c)
Avg. Abs. Value
SA BT 67 3
SA BT 67 10
Rural PAN
Maximum (c)
Average (c)
Avg. Abs. Value
SS AL <1 200
Urban PAN
Maximum (c)
Average (c)
Avg. Abs. Value
I 	 (
	 Maxiu
Standard
6.60E-07
1.78E-06
8.21E-04
2.58E-06
5.22E-07
3.10E-03

1.86E-03

(c)
3.35E-06

1 . 05E-02

(c)
1.65E-08
6.45E-04

1.87E-03

(c)
1.19E-04

4.10E-02

(c)
'alculat<
num 	
* Chg
6.5
-17.1
-5.5
-85.6
-34.3
3.8

-5.9
0.6
1.3
-5.9

6.6
0.4
1.3
6.5
6.6

-3.5
0.0
0.7
-3.5

2.1
0.1
0.4
2d Concentre
Standard
1.28E-07
5.85E-07
3.24E-04
3.76E-07
1.67E-07
1.42E-03

1.24E-03


1.17E-06

6.33E-03


1.37E-08
2.02E-04

1.26E-03


2.89E-05

3.28E-02


it ion (p
Werage
% Chg
7.4
-13.7
-0.8
-90.6
-31.0
5.3

4.8
0.9
1.4
-2.6

6.9
0.3
0.9
2.1
6.9

-6.1
-0.2
0.9
-6.1

-1.7
0.0
0.3
pm) (b)
* /Max
0.0
0.0
0.0
0.0
0.0
0.3

1.1
0.0
0.1
0.0

0.6
0.0
0.1
0.0
0.2

-0.8
0.0
0.0
-0.1

0.4
0.0
0.0

% Fit
11.1
23.1
3.9
165.4
38.1
5.2

4.6
0.6

3.4

6.6
0.7

2.1
6.6

6.4
0.5

6.4

2.6
0.3

(a)  See Footnote (a) in Table 37 for a description of the notation used
     to indicate the test problems.

(b)  See Section 6.2 for a discussions of the specific measures of
     differences between the mechanisms.

(c)  See footnotes (c-e) in Table 37 or 38 for a description of these
     quantities.
                                   296

-------
recommended changes are  implemented in the RADM-M mechanism.  The  changes
involving  the  aromatics mechanism  and  the  updated  parameters  for  the
lumped species  were implemented in  the  RADM-P  mechanism,  a version  where
only parameter  values  (for  the most part) are changed.  The modifications
involved  in the RADM-P mechanism have been accepted by the RADM  team,  and
are being implemented  into the  RADM model (Stockwell, private communica-
tion, 1989).
     Table 40 gives the summary statistics of the differences between  the
two modified  RADM mechanisms and the  March  1988 version in their  predic-
tions of  the  species of interest in the  90 test  problems.   (The March 1988
mechanism  is  taken  as  the "standard" for  the purpose  of this  tabulation.)
Tables 11 and  42  give  the  differences  in the  calculations  of  ozone  and
1^02  for  each  of  the  individual  sets  of  urban  test  calculations,  and
Figures   89  and  90  show  comparisons , of  concentration-time  plots  for
selected  compounds  for  two such  calculations.   Figure  89 shows  results
calculated for  the  urban summer surface  static  simulation with initial  NOV
                                                                          A
of  67 ppb and ROG/NOX  ratio  of   10,  and  Figure  90 shows  the  results
calculated  for  winter  surface  conditions with  the same pollutant  levels.
As  with   the  plots  in  Section  5,   the  solid  line  designates results  of
calculations  using  the RADM mechanism,  the longer dashes  designate results
calculated using RADM-P, and the shorter  dashes  indicate RADM-M results.
     The  results  shown in Table 40  indicate that  the  recommended modified
mechanisms  can  in some   cases  give  quite   different   predictions  for
compounds of  potential interest in RADM applications,  especially  for urban
conditions.   The  greater  sensitivities  in  the  urban cases  are expected
since  the  most important  modifications  in  the  mechanism concerned  the
aromatics,  which are  relatively more important in urban conditions.   The
magnitudes  of  the  differences  depend  on the  conditions  of  the specific
test  problems,  as shown in Tables  41 and 42 for  the  predictions of ozone
and  ^2^2 *>or a1^  the  ur^an cases,  and  as shown in Figures 89 and 90 for
summer  vs. winter  environments for one  set  of pollution  conditions.   In
general,  the  modified  mechanisms had a tendency to predict lower levels of
ozone,  hydroperoxides  and  organic  acids than  the  March  1988  mechanism,
with  the effects  on  the peroxides and  the  acids being greater than  the
effects  on ozone.   In  general,  the test cases  in  Tables  41 and  42 that
show  the  greatest differences in  predictions  for  H202 are also  the cases
                                     297

-------
Table 40.  Summary of Results of Sensitivity Test Calculations Using the
           Recommended Modified (RADM-M) and the Recommended Modified
           Parameter (RADM-P) Mechanisms, Relative to the March  1988
           RADM Mechanism as the Standard
Test Cases
and
Compound
Rural 03
Maximum
Average
Avg. Abs. Value
Urban 0^
Maximum
Average
Avg. Abs. Value
Rural H202
Maximum
Average
Avg. Abs. Value
Urban H202
Maximum
Average
Avg . Abs . Value
Rural SULF
Maximum
Average
Avg. Abs. Value
Urban SULF
Maximum
Average
Avg. Abs. Value
Rural HN03
Maximum
Average
Avg. Abs. Value
Urban HN03
Maximum
Average
Avg. Abs. Value
	 RADM-M vs RADM (a) —
Max. - Average - %
*Chg. JChg. J/Max Fit

3.9
0.0
0.4

-3^.8
-3.0
6.7

-9.3
-1.7
1.9

-85.3
-17.5
17.6

5.9
0.7
1.0

-42.4
1.5
10.8

7.8
0.8
1.8

-32.0
-0.1
9.7

3.
-0.
0.

-32.
-5.

3
2
5

7
5
7.3


-13.9
-2.1
2.3


-72.9
-19-8
19.

6
0
1

-32
1
8

6
0
1

-31
0
7
.8

.1
.6
.0

.9
.0
.6

.8
.4
.3

.3
.8
.8

1.5
-0.1
0.1

-9.0
-0.9
1.5

-4.0
-0.2
0.2

-18.1
-1.8
1.8

2.0
0.1
0.1

-13.1
0.0
1.2

3.0
0.1
0.1

-17.5
-0.1
1.6

3.2
0.2


39.2
5.4


14.8
1.0


115.8
14.9


5.9
0.4


39.3
5.8


6.5
0.7


37.1
6.2

— RADM-P vs RADM (a) —
Max. - Average - %
JChg. *Chg. %/Max Fit

6.
0.
0.

-34.
-1.
6.

-10.
-1.
1.

-89.
-16.
17.

7,
1,
1,


2
2
5

6
4
0

7
5
9

3
,9
,0

.4
.2
.4

-42.8
0.4
9

10
2
2

27
0
8
.4

.0
.0
.1

.8
.9
.9

5.
0.
0.


3
1
4

-27.9
-3.4
6.1


-15.1
-2.0
2,

-73
-19.
19

7
1
1

-31
-0
7

9
1
1

-28
1
7
.2

.7
.1
.1

.7
.1
.3

.9
.1
.6

.0
.6
.7

.5
.6
.6

2.5
0.0
0.1

9.9
-0.3
1.2

-4.5
-0.2
0.2

-17.8
-1.8
1.8

2.5
0.1
0.1

-13.2
-0.2
1.1

3.9
0.1
0.2

15.9
0.1
1.5

5.2
0.2


32.5
4.5


16.2
1.0


117.9
14.2


7.4
0.5


38.0
5.2


8.6
0.8


33.2
5.7

                                                               (continued)
                                     298

-------
Table 40 (continued) - 2
Test Cases
and
Compound
Rural ORA1
Maximum
Average
Avg. Abs. Value
Urban ORA1
Maximum
Average
Avg. Abs. Value
Rural ORA2
Maximum
Average
Avg. Abs. Value
Urban ORA2
Maximum
Average
Avg. Abs. Value
Rural OP1
Maximum
Average
Avg. Abs. Value
Urban OP1
Maximum
Average
Avg. Abs. Value
Rural OP2
Maximum
Average
Avg. Abs. Value
Urban OP2
Maximum
Average
Avg. Abs. Value
Rural PAA
Maximum
Average
Avg. Abs. Value
— RADM-M vs RADM (a) —
Max. - Average - %
JChg. %Chg. */Max Fit

-6.
-1.
1.

-24.
-8.
8.


7
2
5

4
9
9

-28.5
-4.5
4.7


-40.7
-13.2
13.2


-10.2
-0.7
1

-79
-8
12

-28
-14
14

191
-36
43

-14
-1
2
.6

.9
.7
.8

.1
.1
.3

.3
.3
.4

.8
.5
.1

-6.5
-1.2
1.4

-21.6
-8.1
8.1

-42.5
-5.1
5.2

-29.3
-13.7
13.7

-15.3
-1.2
2.0

-75.7
-15.0
16.6

-23.1
-13.8
13.8

-84.8
-40.5
40.5

-22.1
-1.5
2.7

-0.
0.
0.

-8.
-1.
1.

-1.
-0.
0.


5
0
0

9
2
2

9
1
1

-17.8
-2.2

6.7
0.6


24.2
5.3


53.7
2.3
— RADM-P vs RADM (a) —
Max. - Average - %
JChg. JChg. */Max Fit

-7.
-1.
1.

-19.
-6.
6.


1
0
3

3
6
6

-20.4
-4.6
4.9

34.6
8.9
2.2

-1.

.6
0.0

16.5
0.9
0.1

-15
-0
0

-9
-0
0

-28
-3
3

-1
0
0

.3
.5
.9

.8
.6
.6

.6
.1
.1

.4
.0
.1

121.8
13.1


26.1
6.1


157.5
33.0


16.9
1.2



-39.1
-13.2
13.2

-12
-1
2

-73
-18
18

35
-5
7

900
-6
39

-17
-5
5

.8
.9
.1

.9
.7
.7

.7
.8
.8

.4
.2
.8

.8
.1
. 1

-6.6
-1.0
1.2

-17.0
-6.0
6.1

-21.1
-5.0
5.1

-26.4
-13.6
13.6

-13.8
-2.5
2.6

-69.0
-22.6
22.6

-17.3
-6.4
7.0

748.8
-10.0
37.7

-21.0
-6.1
6.1

-0.7
0.0
0.0

-6.4
-0.9
0.9

-1.6
-0.1
0.1

-16.4
-2.4
2.4

-3.2
-0.1
0.1

-19.8
-1.8
1.8

-7.6
-0.3
0.3

-19.5
-1.5
1.5

-4.3
-0.2
0.2

6.9
0.5


18.6
4.0


23.6
2.2


30.6
8.9


14.8
1.2


105.2
16.6


18.9
3.0


160.9
20.8


23.4
2.6

                                                               (continued)
                                     299

-------
Table 40 (continued) - 3
Test Cases
and
Compound
Urban PAA
Maximum
Average
Avg. Abs. Value
Rural PAN
Maximum
Average
Avg. Abs. Value
Urban PAN
Maximum
Average
Avg. Abs. Value
	 RADM-M vs RADM (a) —
Max. - Average - T>
JChg. JChg. J/Max Fit

-88
-13
15

9.

.2
.4
.8

2
2.4
2.

-61.
3.
15.
8

4
4
9

-73.4
-15.5
17.1

9-5
0.3
1.6

-39.3
-1.6
11.9

-20
-0
1

6.
0.
0.

30.
1.
1.

.3 117.2
.6 12.7
.3

8 9.1
1 1.2
2

7 61.4
1 9.4
6
Max.
JChg

-91
-25
25

-14.
-4.
4.

-64.
-17.
18.
RADM-P vs RADM (a
- Average -
. JChg. J/Max

.1
.1
.1

2
6
6

5
8
5

-82
-27
27

-14.
-4.
4.

-51.
-17.
17.

.0
.2
.2

2
6
6

6
8
8

-37.3
-3.3
3.3

-3.3
-0.2
0.2

-17.3
-1.7
1.7
) —
*
Fit

139. 1
20.!


15.3
1.9


69.7
13.0

(a)  See Section 6.2 for a discussions of the specific measures of
     differences between the mechanisms.  See footnotes (c-e) in
     Table 37 or 38 for a description of the quantities tabulated.
                                    300

-------
Table 41.  Results of Sensitivity Test Calculations of Urban Ozone Formation Using the Recommended
           Modified (RADM-M) and the Recommended Modified Parameter (RADM-P) Mechanisms,  Relative
           to the March 1988 Version of the RADM Mechanism as the Standard


Test Problem (a)   Standard Cone (ppm)    --- RADM-M vs RADM (b) ---    --- RADM-P vs RADM (b) -—
Env  Mix NOx C/N    Maximum    Average    Max.    - Average -     %      Max.   - Average  -     %
                                          %Chg.  flChg. £/Max    Fit     *Chg.  £Chg.  J/Max    Fit
ss
ss
ss
ss
ss
ss
ss
ss
ss
ws
ws
ws
SA
SA
SA
SD
SD
SD
SS
SS
SS
SS
ss
AC 20
AC 20
AC 20
AC 67
AC 67
AC 67
AC 100
AC 200
AC 200
AC 67
AC 67
AC 67
AC 67
AC 67
AC 67
AC 67
AC 67
AC 67
DC 20
DC 20
DC 20
DC 67
DC 67
10
33
100
3
10
30
2
3
10
3
10
30
3
10
30
3
10
30
10
33
100
3
10
1.00E-01
1.11E-01
1.04E-01
6.10E-02
1.86E-01
1.94E-01
6.62E-02
9.94E-02
3.44E-01
2.00E-02
5.48E-02
1.55E-01
7.93E-02
2.37E-01
2.05E-01
8.46E-02
1.11E-01
1.19E-01
1.00E-01
1.09E-01
1.03E-01
6.10E-02
1.85E-01
7.06E-02
9-32E-02
8.68E-02
4.89E-02
1.25E-01
1.59E-01
3.90E-02
7.00E-02
2.23E-01
8.12E-03
3.76E-02
9.46E-02
5.52E-02
2.13E-01
1.90E-01
6.72E-02
8.56E-02
9.29E-02
7.08E-02
9.18E-02
8.63E-02
4.97E-02
1 .25E-01
-3.8
0.1
3.6
-0.8
-4.2
0.4
-30.3
13.1
-3.1
0.0
9.6
2.7
-26.7
-2.4
7.8
-3.2
-4.2
-3.1
-4.4
0.0
3.4
-1.7
-5.1
-4.5
-0.7
2.4
-7.7
-6.3
-0.7
-19.2
-4.4
-6.6
-19.2
4.8
3.3
-26.8
-5.8
8.1
-2.6
-3.2
-1.6
-5.3
-1.0
2.1
-9.8
-7.7
-1.4
-0.3
0.9
-1.7
-3.5
-0.5
-3.4
-1.4
-6.6
-0.7
0.8
1.4
-6.7
-5.5
6.9
-0.8
-1.2
-0.7
-1.7
-0.4
0.8
-2.2
-4.3
4.6
1.5
3.2
7.9
6.5
2. 1
21.1
18.8
6.9
21.1
10.4
3.4
31.0
5.9
7.9
2.6
3-3
1.7
5.5
2.0
3.3
10.2
8.0
-2.1
2.0
4.5
0.7
-2.0
2.3
-30.1
13.8
-0.7
0.0
10.3
2.8
-22.5
-0.3
9.6
-1.7
-1.4
0.2
-2.4
2.3
4.8
0.1
-2.4
-2.6
1.4
3.7
-5.5
-3.2
1.6
-17.0
-1.6
-2.8
-17.0
5.8
3.5
-22.7
-3.7
9.9
-1.6
-1.3
0.6
-3.1
1.6
4.0
-7.1
-3.8
-0.8
0.6
1.5
-1.2
-1.8
1. 1
-3.0
-0.5
-2.8
-0.6
1.0
1.5
-5.6
-3.6
8.5
-0.5
-0.5
0.2
-1.0
0.6
1.5
-1.6
-2.1
2.7
1.7
3.6
6.1
3.3
2.3
18.4
17.6
2.8
18.4
10.6
3.7
25.6
3.8
9.5
1.6
1.3
0.6
3.1
2.1
3.9
7.3
3.9
                                                                                           (continued)

-------
             Table 41  (continued)  - 2
u>
o
KJ
             Test Problem (a)
             Env  Mix NOx C/N
Standard Cone (ppm)
 Maximum    Average
._. RADM-M VS RADM (b) ---
Max.   - Average -     %
JChg.  JChg. %/Max    Fit
— RADM-P vs RADM (b) —
Max.   - Average -     %
       JChg. %/Max    Fit
ss
ss
ss
ss
ws
ws
ws
SA
SA
SA
SD
SD
SD
SS
SS
SS
SS
SS
SS
SS
SS
ss
ws
ws
ws
DC 67
DC 100
DC 200
DC 200
DC 67
DC 67
DC 67
DC 67
DC 67
DC 67
DC 67
DC 67
DC 67
BT 20
BT 20
BT 20
BT 67
BT 67
BT 67
BT 100
BT 200
BT 200
BT 67
BT 67
BT 67
30
2
3
10
3
10
30
3
10
30
3
10
30
10
33
100
3
10
30
2
3
10
3
10
30
1.89E-01
7.07E-02
9.65E-02
3.38E-01
2 . OOE-02
5.26E-02
1.50E-01
8.08E-02
2.32E-01
1.98E-01
8.50E-02
1.11E-01
1.17E-01
9.63E-02
1.16E-01
1.11E-01
6.36E-02
1.81E-01
2.11E-01
1.90E-02
1.22E-01
3.17E-01
2. OOE-02
5.12E-02
1.16E-01
1.51E-01
1.06E-02
7.07E-02
2.21E-01
8.52E-03
3.65E-02
9.25E-02
5.65E-02
2.10E-01
1.85E-01
6.76E-02
8.55E-02
9.20E-02
6.77E-02
9.60E-02
9.31E-02
1.51E-02
1.19E-01
1.70E-01
3.19E-02
6.82E-02
2.16E-01
6.19E-03
3.18E-02
8.80E-02
0.1
-31.8
11.7
-1.1
0.0
11.1
3.8
-32.6
-2.3
9.7
-1.1
-5.1
-3.9
-2.7
-0.2
2.6
0.6
-3.6
0.1
6.5
-31.2
-3.5
0.0
-11.1
-2.0
-1.2
-22.8
-6.5
-8.2
-21.0
8.3
1.5
-32.7
-6.1
9.1
-3.3
-3.8
-2.2
-3.0
-0.7
1.6
-1.0
-1.7
-0.7
-3.2
-29.3
-5.5
-9.6
-8.9
-2.3
-0.8
-1.2
-2.1
-8.1
-0.9
1.1
1.9
-8.3
-5.7
7.8
-1.0
-1.5
-0.9
-0.9
-0.3
0.7
-0.8
-2.5
-0.5
-0.5
-9.0
-5.1
-0.3
-1.1
-0.9
2.6
25.5
22.3
8.6
27.1
11.7
1.5
39.2
6.2
9.1
3.3
3.9
2.3
3.0
1. 1
2.5
1.6
1.8
1.5
9.8
31.0
5.7
10.0
9.3
2.1
2.5
-31-6
15.1
-1.0
0.0
11.8
1.0
-27.8
0.5
12.1
-2.3
-1.8
0.2
-1.7
1.2
3.6
1.2
-2.5
1.5
6.5
-25.3
-2.3
0.0
-13.2
-2.5
1.7
-20.2
-3.2
-3.3
-21.3
9.5
1.8
-27.9
-3.1
11.9
-2.1
-1.5
0.6
-1.9
0.8
2.9
-2.9
-3.1
0.9
-2.3
-21.1
-3.5
-8.1
-8.2
-2.8
1.2
-3.7
-1.0
-3.3
-0.8
1.6
2.0
-7.1
-3.2
9.9
-0.6
-0.6
0.2
-0.6
0.3
1.2
-0.6
-1.7
0.7
-0.3
-7.1
-3.1
-0.2
-1.3
-1.1
2.5
22.3
20.6
3.1
23.7
11.8
1.9
32.5
3.8
11.3
2.1
1.5
0.6
1.9
1.1
2.8
1.3
3.2
1.5
8.9
27.1
3.5
8.7
8.5
2.9
                                                                                                       (continued)

-------
            Table  41  (continued)  -  3
o
OJ
            Test  Problem  (a)   Standard  Cone  (ppm)     —  RADM-M vs  RADM  (b)  —     — RADM-P vs RADM (b)
            Env   Mix NOx  C/N    Maximum     Average     Max.    -  Average  -      %       Max.    -  Average -     %
                                                       £Chg.   %Chg. £/Max     Fit      £Chg.   £Chg.  £/Max    Fit
SA
SA
SA
SD
SD
SD
BT
BT
BT
BT
BT
BT
67
67
67
67
67
67
3
10
30
3
10
30
5.95E-02
2.33E-01
2.23E-01
8.37E-02
1.10E-01
1.24E-01
4.00E-0?
1.98E-01
2.07E-01
6.62E-02
8.45E-02
9.56E-02
-17.0
-2.3
6.9
-2.2
-2.8
-2.3
-19.1
-5.8
5.8
-1.8
-2.2
-1.3
-3.4
-5.1
5.4
-0.5
-0.8
-0.6
21.1
5.9
5.8
1.8
2.2
1.4
-15.4
-0.9
7.9
-1.4
-1.1
0.0
-17.0
-4.6
6.8
-1.2
-1.0
0.2
-3.0
-4.1
6.3
-0.4
-0.4
0.1
18.6
4.7
6.7
1.2
1.0
0.2
             (a)  See Footnote  (a)  in Table 37  for a description of the notation used  to  indicate  the test
                 problems.
             (b)  See Section 6.2 for a discussions of  the specific measures of differences between  the
                 mechanisms.

-------
Table 12.  Results of Sensitivity Test Calculations of Urban HpOp Formation Using the Recommended
           Modified (RADM-M) and the Recommended Modified Parameter (RADM-P) Mechanisms,  Relative
           to the March 1988 RADM Mechanism as the Standard
Test Problem (a)
Env  Mix NOx C/N
Standard Cone (ppm)
 Maximum    Average
... RADM-M vs RADM (b) ---
Max.   - Average -     %
JChg.  £Chg. 2/Max    Fit
— RADM-P vs RADM (b)
Max.   - Average -     %
*Chg.  tChg. */Max    Fit
ss
ss
ss
ss
ss
ss
ss
ss
ss
ws
ws
ws
SA
SA
SA
SD
SD
SD
SS
SS
SS
SS
SS
AC 20
AC 20
AC 20
AC 67
AC 67
AC 67
AC 100
AC 200
AC 200
AC 67
AC 67
AC 67
AC 67
AC 67
AC 67
AC 67
AC 67
AC 67
DC 20
DC 20
DC 20
DC 67
DC 67
10
33
100
3
10
30
2
3
10
3
10
30
3
10
30
3
10
30
10
33
100
3
10
2.62E-03
5.15E-03
1.17E-02
1.91E-03
1.11E-03
1.22E-02
1.77E-03
3.12E-03
8.35E-03
1.80E-07
1.06E-03
1.26E-03
2.19E-07
1.56E-03
1.18E-02
1.71E-03
3.76E-03
7.11E-03
2.65E-03
5.37E-03
1.10E-02
1.98E-03
1.15E-03
1.86E-03
3.86E-03
8.16E-03
9.12E-01
3.12E-03
8.56E-03
7.61E-01
1.57E-03
5.90E-03
2.92E-08
6.27E-01
3.03E-03
9.15E-08
1.08E-03
1.20E-02
1.03E-03
2.29E-03
1.67E-03
1.89E-03
3.81E-03
7.68E-03
9.28E-01
3.19E-03
-6.1
-5.0
-2.3
-4.8
-11.3
-5.7
-12.6
-18.7
-15.0
-81.9
-27.7
-11.1
-77. 1
-32.8
-16.1
-7.5
-8.2
-6.9
-7.5
-5.5
-3.0
-7.1
-13.3
-8.1
-6.1
-3.8
-10.8
-13.7
-7.3
-13.9
-29.1
-16.9
-61.8
-33.0
-12.8
-58.9
-11.7
-15.3
-7.2
-8.0
-6.1
-9.1
-7.1
-1.7
-12.7
-15.6
-1.3
-2.1
-2.6
-0.8
-3.6
-5.2
-0.9
-3.8
-8.3
0.0
-1.7
-3.2
0.0
-1.0
-15.3
-0.6
-1.5
-2.5
-1.1
-2.3
-3.0
-1.0
-1.2
8.1
6.7
1.0
11.3
11.7
7.7
11.9
33.6
18.5
97.0
39.1
13.7
83.9
57.2
16.6
7.1
8.3
6.7
9.5
7.7
1.9
13.5
16.9
-5.5
-6.0
-3.8
-3.6
-9-8
-7.5
-12.5
-16.2
-12.3
-80.5
-29.2
-13.7
-75.6
-29.7
-16.2
-5.8
-6.5
-5.8
-7.5
-7.0
-1.7
-5.5
-11.1
-7.8
-6.7
-5.0
-9.3
-12.7
-8.1
-12.9
-23.0
-15.1
-65.2
-33.8
-15.5
-51.5
-12.0
-15.2
-6.5
-7.2
-6.1
-9.1
-7.6
-5.8
-11.1
-11.1
-1.2
-2.2
-3.1
-0.7
-3-3
-6.0
-0.8
-3.0
-7.6
0.0
-1.8
-3.9
0.0
-3.8
-15.2
-0.6
-1.1
-2.1
-1.5
-2.1
-3.7
-0.9
-3.8
8.0
6.9
5.1
9.6
13.6
8.8
13.7
25.6
16.6
97.6
10.6
16.8
75.3
52.9
16.1
6.8
7.1
6.3
9.8
7.9
6.0
11.7
15.5
                                                                                          (continued)

-------
            Table 12  (continued) - 2
U)
o
Ln
            Test Problem (a)
            Env  Mix NOx C/N
Standard Cone (ppm)
 Maximum    Average
___ RADM-M vs RADM (b) --•
Max.   - Average -     %
%Chg.  %Chg. %/Max    Fit
___ RADM-P vs RADM (b)
Max.   - Average -     %
%Chg.  %Chg. J/Max    Fit
ss
ss
ss
ss
ws
ws
ws
SA
SA
SA
SD
SD
SD
SS
SS
SS
SS
SS
SS
ss
ss
ss
ws
ws
ws
DC 67
DC 100
DC 200
DC 200
DC 67
DC 67
DC 67
DC 67
DC 67
DC 67
DC 67
DC 67
DC 67
BT 20
BT 20
BT 20
BT 67
BT 67
BT 67
BT 100
BT 200
BT 200
BT 67
BT 67
BT 67
30
2
3
10
3
10
30
3
10
30
3
10
30
10
33
100
3
10
30
2
3
10
3
10
30
1 . 18E-02
1.85E-03
3.17E-03
8.16E-03
3.08E-07
1 . 08E-03
1.09E-03
2.59E-07
1.67E-03
1.16E-02
1.77E-03
3.80E-03
7.36E-03
2.11E-03
1.96E-03
1.09E-02
1.90E-03
3.96E-03
1.09E-02
1.60E-03
2.83E-03
7.29E-03
1.71E-08
7.11E-01
3.10E-03
8.32E-03
8.11E-01
1.60E-03
6.08E-03
3.89E-08
6.55E-01
2.96E-03
1.02E-07
1.22E-03
1.18E-02
1.05E-03
2.32E-03
1.66E-03
1.65E-03
3.51E-03
7.61E-03
8.08E-01
2.51E-03
7.62E-03
6.67E-01
8.98E-01
1.17E-03
6.31E-09
3.57E-01
2.23E-03
-6.5
-16.8
-20.9
-17.9
-85.3
-28.8
-13.3
-81.9
-10.6
-19.7
-9.1
-9-5
-8.1
-4.5
-1.1
-2.1
-2.3
-8.3
-5.2
0.7
-25.1
-11.7
-12.6
-12.0
-1.1
-8.5
-19.5
-30.9
-19.1
-72.9
-33.5
-11.9
-67.2
-53.8
-18.5
-8.3
-9.1
-7.7
-6.2
-5.0
-3.6
-12.1
-11.8
-6.1
-8.3
-65.2
-16.2
-31.1
-16.5
-6.2
-5.9
-1.3
-1.1
-9.8
0.0
-1.8
-3.7
0.0
-5.5
-18.1
-0.7 .
-1.8
-3.0
-0.9
-1.5
-2.3
-0.8
-2.5
-1.1
-0.5
-1.9
-6.0
0.0
-0.5
-1.1
9.0
21.1
36.1
21.5
115.8
10.1
16.1
101.7
73.3
20.3
8.7
9.5
8.0
6.1
5.2
3.7
12.6
12.5
6.6
10.1
91.8
17.6
11.3
17.9
6.1
-8.8
-17.1
-18.3
-11.5
-89.3
-30.8
-16.9
-81.1
-36.9
-19.6
-7.5
-7.7
-7.0
-1.5
-1.1
-2.7
-2.5
-7.8
-5.9
0.9
-22.3
-10.1
-31.7
-12.7
-1.6
-9.8
-18.8
-23.6
-17.1
-73.7
-31.9
-18.6
-61.6
-51.1
-18.2
-7.8
-8.0
-7.2
-6.2
-1.8
-3.1
-11.3
-11.7
-6.1
-7.1
-63.1
-15.2
-29.0
-16.9
-7.0
-6.8
-1.3
-3.1
-8.8
0.0
-1.9
-1.6
0.0
-5.2
-17.8
-0.7
-1.6
-2.8
-0.9
-1.1
-2.2
-0.8
-2.5
-1.1
-0.1
-1.7
-5.7
0.0
-0.5
-1.3
10.3
20.6
26.5
19.0
117.9
12.1
20.1
96.0
68.3
19.9
8.1
8.1
7.1
6.1
1.9
3.5
11.8
12.1
6.7
8.9
90.8
16.1
31.6
18.1
7.3
                                                                                                       (continued)

-------
            Table  42  (continued)  -  3
U)
o
Test
Env

SA
SA
SA
SD
SD
SD
Problem (a)
Mix NOx C/N

BT
BT
BT
BT
BT
BT

67
67
67
67
67
67

3
10
30
3
10
30
Standard Cone (ppm)
Maximum Average

4.36E-08
1.14E-03
1.20E-02
1.67E-03
3.54E-03
6.96E-03

3.47E-08
6.21E-04
9.78E-03
9.66E-04
2.11E-03
4.31E-03
._. RADM-M vs RADM (b) ---
Max. - Average - %
*Chg.
-30.7
-17.5
-13.5
-6.0
-6-.0
-5.0
*Chg.
-37.4
-34.7
-13.5
-5.4
-5.6
-4.8
J/Max
0.0
-1.8
-11.0
-0.4
-1.0
-1.7
Fit
46.2
41.7
14.4
5.5
5.8
4.9
_— RADM-P vs RADM (b) —
Max. - Average - %
*Chg.
-30.2
-16.5
-13.2
-5.2
-5.0
-4.1
JChg.
-36.0
-34.8
-13.0
-5.2
-5.3
-4.4
*/Max
0.0
-1.8
-10.6
-0.4
-0.9
-1.6
Fit
44.1
41.8
13.9
5.4
5.4
4.5
            (a)  See Footnote  (a)  in Table 37 for a description of  the notation used  to  indicate the test
                problems.
            (b)  See Section 6.2 for a discussions of the specific  measures of differences between the
                mechanisms.

-------
           OZONE
                                         0.005 -,
                                                 HYDROGEN PEROXIDE
    0.04-
    0.00'
   l  ' I  ' I  ' I '  I '  I '  I ' I  ' I  ' 1
5  11  17 23 29 35 41  47 53  59  65
  n  ' i  * i * i »  i »  i *  i *  i
5  11  17 23 29 35 41  47 53
                                                                         i ' i
                                                                        59 65
Q.
a
O
h-
UJ
O

O
O
   0.003 -i
   0.002 -
   0.001 -
           SULFATE
   0.000 •
5  11  17 23 29 35 41  47 53  59  65


   FORMIC  ACID
                                          0.000'
   I  i I  r I  . j T y^T '  I '  I '
5  11  17  23 29 35 41  47 53
                                                 METHYL HYDROPEROX1DE
                                                                         I  ' I '
                                                                         59 65
         T ' I  ' I '  I '  I '  I '  I '  1
5  11  17  23 28 35 41 47 53 59  65
      I  ' I  ' I  ' I '  1 '  I '  I
5  11  17  23  29 35 41 47 53
                                                                         I  ' I
                                                                         59  65
                                HOURS AFTER MIDNIGHT
   Figure  89.   Concentration-Time Plots  for  Selected Species Calculated
                Using the RADM, RADM-M, and RADM-P Mechanisms for  the  Summer
                Surface Static Urban Test Problem with Initial NOX of  6?  ppb
                and a ROG/NOV Ratio of 10.
                            A
                                        307

-------
       OZONE
                                    0.0012-1
                                    0.0009-
                                    0.0006-
                                    0.0003 -J
  HYDROGEN PEROXIDE
                                    0.0000
       17 29  41  53  65 77  89 101 113


       SULFATE
5  17 29  41  53 65 77  89  101 113


  NITRIC ACID
        I ' • I ' ' I  ' 'T T M ' ' 1 '  ' I ' ' I ' ' I
        17 29 41  53 65 77  89 101 113
       FORMIC ACID
   17  29  41  53 65  77  89 101 113
   METHYL HYDROPEROX1DE
        l •  • l ' • i • • i ' ' l ' ' l ' ' l ' ' l 'T T
        17  29 41 53 65  77 89 101  113
   17  29  41  53  65  77 89 101 113
                            HOURS AFTER MIDNIGHT
Figure 90.  Concentration-Time Plots for Selected  Species Calculated
            Using  the  RADM,  RADM-M, and RADM-P  Mechanisms for the Winter
            Surface  Static Urban Test Problem with Initial NOX of 67 ppb
            and  a  ROG/NOX Ratio of 10.
                                     308

-------
where the  predictions are most different for  the other peroxides and  the
acid species.   On the  other  hand,  the modifications  have  much less of a
consistent effect on  predictions of nitric acid and  sulfate, with positive
as well as negative  changes,  and fewer cases with relatively large differ-
ences being observed.
     As expected,  the differences  between the two versions of the  recom-
mended modified  mechanisms  are much less than the differences  between  the
modified  and  the  March  1988  mechanisms.   The   differences  between  the
modified mechanisms  in  predictions  of ozone,  H^Op,  sulfate, HNO?, and  the
organic acids  are  generally minor,  with discrepancies  only being observed
in  predictions  of  peroxy  or  acid  species  in   a  few cases  where  their
calculated  concentrations are  very  low.   There  are   some cases  of  non-
negligible   differences  in  predictions  of  the   organic  hydroperoxide
species,  peroxyacetic  acid,  and   PAN.,   Since   PAN   is  not   represented
explicitly in  any of these  mechanisms, and since  there are  no  chamber data
available  to test predictions  of  organic hydroperoxide  or acid species,
there is no  way  of  determining which  of  the two modified  mechanisms  is the
more reliable  in their  predictions  of these  species.   In view  of the many
chemical uncertainties  in the mechanisms regarding reactions  forming these
species and  the  sensitivity of predictions to  relatively  small differences
in the mechanisms,  predictions for  these species  in regional  model  simula-
tions should be  viewed  as highly uncertain.
     These results  indicate that the  recommended  modifications to the RADM
mechanism  may  have  a non-negligible impact on  predictions of species of
interest in  regional modeling applications.   Because the  modified versions
perform  somewhat better  in  simulating  the  aro.raatic-NOx-air chamber runs,
predictions  obtained with  these mechanisms  probably  should be considered
to  be  somewhat  more reliable than  those obtained  with  the March  1988
mechanism.   On  the  other hand, it probably  does not  matter  which  of the
two  modified versions  is adopted,  except that the  RADM-M  version  employs
fewer  species  and thus is  more efficient.   The  magnitude of the  differ-
ences  between  the modified and the March 1988 mechanisms also gives us an
approximate  indication  of the sensitivity of regional model predictions to
the  uncertain  aspects of aromatics photooxidation chemistry,  since that is
the  major  aspect of  the mechanism that was changed.
                                    309

-------
                        7.   SUMMARY AND CONCLUSIONS

     One  of the  more  significant advances  in  regional  acid deposition
modeling  in the  1980s  has  been  the  incorporation of  gas phase  chemical
mechanisms  that  predict  the  formation  of  sulfate,  nitric  acid, ozone,
organic nitrates,  hydrogen  peroxide,  organic  peroxide,  and organic acids
from emissions  of S02, NOX,  and  reactive organic  gases  (ROG).    The  gas-
phase chemical mechanism is  a particularly  important  component of  regional
acid  deposition  models because   it  controls  the rates  of  radical   and
oxidant generation, which  in turn controls the production  rates of acidic
species.   The  chemical  mechanism is  also  the major  potential  source  of
nonlinearity in the relationships between acid deposition  and  emissions of
SOo  and  NOX.    As such,   it is  one  of the  major  reasons  why control
strategies  for  acid  deposition   cannot  rely  on analyses  based  on simple
rollback,  but  require  use  of complex  models  such as  RADM.   Thus,  it  is
particularly important  that the  gas-phase chemical mechanism  used in  such
models  be  evaluated  as comprehensively as  possible,  to  assure  that  it
represents  the  current state-of-the-science of atmospheric chemistry.   In
addition,  since the  RADM model is to be used for  policy  development,  its
major components  such as the gas-phase chemical mechanism  must be  accepted
as  representing the  best science by  not only the  technical community but
by  the  public as  a whole.
     Peer  review is  the  method  presently  used  to assure the quality  of
scientific  research.   The  essence of peer  review is  that  the research be
evaluated  by experts  in  the field  other  than   those  who developed  the
research  being  considered.   It is required  before results  are  published in
the most  reputable   scientific   journals,  and agencies  such as  the  EPA
periodically  have peer review panels to assure the  continuing  quality of
the research they support.   Therefore,  peer  review  is clearly an  appro-
priate  means  to assure the quality and credibility of important  components
of  the  RADM model such as  the gas-phase chemical  mechanism.  However,  peer
review  as  it  is normally  practiced  does not   lend  itself  to  in-depth
assessment of  highly complex model  components such as  gas-phase  atmos-
pheric  reaction mechanisms.   Journal space is much  too limited  to  permit
in-depth  discussion   of the  many  components  involved  in  such mechanisms,
and thus  they  cannot  be   evaluated  adequately  by  the  reviewers.    Peer
                                    310

-------
review  panels  of  agency  research  programs  generally   have  relatively
limited time  to  assess  very  broad  areas of research and usually cannot be
expected  to carry  out  in-depth analyses  of  individual  components.   The
study described  in  this report,  however,  is an in-depth peer review of an
important model  component — the RADM gas-phase mechanism.   The authors of
this report have  been  involved for  years in  the development  and evaluation
of  gas-phase  mechanisms  for  use  in  airshed  models,  but  had  not parti-
cipated  in  the development of  the RADM gas-phase  mechanism.   Because of
the  many  aspects involved  in  evaluating  a  chemical mechanism,  including
the  need  to test it against the results of over  500 environmental  chamber
experiments, this review took more  than a year.
     The  results of this review  indicated  that for  the most part  the  RADM
mechanism  represents the  state  of our present  knowledge  of  atmospheric
chemistry and  performs  as  well as  reasonably can  be expected in simulating
the  results of  environmental chamber  experiments.   However, the  state of
knowledge of  atmospheric chemistry  is  constantly advancing, and  over  the
course  of this study, several  sets of modifications to  the  RADM  mechanism
and   its   interface  to  the  emissions   inventory  were  identified   and
recommended.   These recommendations and areas of suggested improvement are
described  in  detail in the body of this report and  are  briefly summarized
below.   These recommendations do  not  involve major changes to  the  basic
nature  of the mechanism as conceived by its developers,  but instead should
be  considered as refinements and  enhancements to an already high quality
product.   However,  we believe  that the implementation of these  recommenda-
tions  will  result  in  improvements in  representation of several  important
aspects of  the atmospheric  chemistry  of organic compounds.
      As stated in  the  preface to  this report, the  developers  of the  RADM
model  cooperated fully with this  review and actively  participated in  some
of  its elements.   The  principal developer  of the RADM gas-phase mechanism
kept us fully informed  of  the  exact status  of the  mechanism  as it under-
went changes, evaluated  our  recommendations,  and  worked directly with us
in  those aspects of this  evaluation where  his  participation  proved  most
valuable.   He and  other members  of  the RADM team  actively solicited our
participation in the development  of  the emissions  processing and aggrega-
tion methods,  re-directing our  efforts  to  this  critical  aspect of the
 implementation of  the  mechanism in the model.  Without the high degree of
                                     311

-------
interest and  cooperation  by  the  RADM team,  this program  could  not have
been  nearly  as  successful  as  we  believe  it was.   Most of  the major
recommendations developed during  this  evaluation  were accepted for  imple-
mentation in the RADM-JI  mechanism.
     It should  be  emphasized that although  we believe that the  resulting
RADM  gas-phase  mechanism and  its associated  emissions  processing  system
represents the current state-of-the-science,  this does not  mean that there
are no significant  areas of  uncertainty.    There remain  many  important
aspects of the gas-phase atmospheric reaction  mechanisms of organics which
are unknown  and where  basic laboratory  research is  still needed  before
reliable mechanisms  representing them can  be developed.    The  available
environmental  chamber  data  are  not  suitable  for  testing  many  critical
aspects  of  the mechanism,  particularly  those  of   the  peroxy  and acid
species which are  of concern in acid deposition models.   In addition,  the
need  to  represent  the highly complex chemistry  of  the many  hundreds  of
organic compounds  emitted into the  atmosphere with mechanisms using only
limited numbers of species and  reactions  introduces uncertainties in model
predictions whose  effects are not fully  understood.   This is  in addition
to  the uncertainties  in  the emissions   inventories  themselves,  which  —
particularly  for   biogenic  emissions  —  often dwarf the  effects of  any
uncertainties in the chemical mechanism.   Thus, users of regional  models,
even those such as RADM  that contain the  most up-to-date,  thoroughly peer-
reviewed gas-phase chemical  mechanisms, must view their predictions with a
certain amount  of  skepticism and  with  an  appreciation of the magnitudes of
the various uncertainties involved.
     The scope  of  work for  the study  consisted of four major  tasks.   The
research findings  from each  task  are summarized below.  Following that, we
briefly summarize our recommendations  for future work.

7.1  Evaluation of Kinetic and Mechanistic Parameters
     At  the  time  this  study was  initiated,  the RADM mechanism was in a
state  of transition  between  the  RADM-I and RADM-II versions of  the mech-
anism.    An  initial  review  of the  preliminary  RADM-II  mechanism  was
performed, and  a number  of changes  in both the inorganic and organic reac-
tions  in the mechanism  were recommended.   A procedure  was developed  for
determining  rate  constants  and  products  for  the  reactions  of  lumped
                                    312

-------
organic species based on  the  detailed  speciation  of  the national  emissions
inventory.    Many  of  these  initial  recommendations were  adopted   and
incorporated  into  the   March   1988   version   of  the RADM-II   mechanism
(referred  to  as the RADM).   The comprehensive evaluation was carried  out
for this version of the RADM mechanism.
     On  the  basis  of  testing   the  RADM mechanism  against  environmental
chamber  data,  two alternate  versions  of the mechanism were  developed  and
fully  evaluated.    The  first   alternate  mechanism,  referred to  as  the
recommended   modified   RADM  mechanism  (RADM-M),   offers   the   following
features:

     •   Changes  in  the representation of the  unknown  aspects of  the
         aromatic  photooxidation  mechanism  which improve  its perform-
         ance  in  simulating chamber data  and  reduce  the  number of
         species  needed;

     •   A more  condensed representation of the >C3 alkanes which has
         comparable  performance with fewer species;

     •   Some  changes in  rate constant and  product  yield parameters
         for lumped species to correspond more closely with the most
         recent  analysis of national VOC emissions data;

      •   Correction of  an error  in the  methylglyoxal  quantum yields
         which resulted in overestimation of its photolysis rates;

      •   Addition of several  reactions omitted from the NC^  + alkene
         reaction system; and

      •  Revised  representation  of  the  reactions  of   nitrophenols
         which improves the chemical fidelity of the reactions.

      Because implementation  of  the RADM mechanism  was almost completed  by
 the time  the RADM-M was developed and evaluated, a second  alternate mech-
 anism was proposed that  involved  primarily changes  in rate  constants and

                                     313

-------
product yields.   This  version is referred to  as  the recommended modified
parameter version of the RADM (RADM-P).   This incorporates:

     •  An  improved aromatic oxidation  mechanism  which  performs
        approximately as well as  the  RADM-M mechanism, but uses the
        larger number of species  employed in the RADM mechanism;

     •  Some changes in rate constant and  product yield parameters
        for lumped  species  to correspond more closely with the most
        recent analysis of national VOC emissions data;

     •  Reduced methylglyoxal quantum yields; and

     •  Addition of  several  reactions omitted from the NO? + alkene
        reaction system.

The  RADM developers have  adopted  the  modified parameter  version  of  the
mechanism and  this version is now  being  implemented into the  RADM trans-
port model (Stockwell,  private communication,  1989).

7.2  Development of the Linkage Between the VOC Emissions Inventory and
     the RADM Mechanism
     In  recent  years,  modelers have  begun  to appreciate fully the  impor-
tance  of carefully  interfacing   a chemical  mechanism with  VOC emission
inventories.    At  the  start  of  this  program,  only a  very  preliminary
emission  inventory interface had  been developed  for  the RADM mechanism.
At  the request  of the  RADM developers, a large effort in this  program was
directed  towards developing emissions aggregation procedures for use with
the  RADM  model.   A new two-step  emissions aggregation  system was  developed
for  this  purpose.   The first step involved  the aggregation  of the detailed
VOX;  emissions  data into  a  32-class mechanism-independent  VOC grouping
system based  on reactivity and relative  contributions to total emissions.
The  second  step  involved  further aggregating these emissions  groups into
the  more limited number of VOC  species  used  in the RADM mechanism.   This
effort was  carried out jointly as part of this program and  as  part of the
RADM development effort.
                                    314

-------
     The assignment of species  in the emissions  groups  to the  surrogate
species used in  the RADM  mechanism involves use  of a  new  reactivity  weigh-
ing scheme.  This  scheme  allows  VOCs  reacting  at different  rates,  but  with
similar mechanisms,  to be represented by the same model species.  This  is
based on estimation of how much  of the  VOCs and  the model surrogate  under-
go chemical reaction in the model  simulation,  and  adjusting the  amounts  of
the latter  so  they are approximately equal. Although approximate, the use
of reactivity  weighing is preferable to the alternatives of either  ignor-
ing compounds  of low  but non-negligible reactivity,  or representing  them
by much more reactive  model  surrogates.
     Another  feature  of  this aggregation  system  is the  use  of  detailed
emissions data to  derive  the rate  constants and  product yield  coefficients
of lumped  species  in  the mechanism  based  on those  for  the aggregate  of
compounds they will be representing.   A software system recently developed
as part of  another  program and  the  1985 NAPAP  anthropogenic  emissions
inventory were used for  this purpose.   This  is considered a  significant
advancement that optimizes the linkage between  the chemical mechanism and
the emissions  inventory.

7.3  Mechanism Evaluation  Against  Chamber Data
     An  important  step  in  the model  design  and evaluation  process  is
evaluation  of   the gas  phase  chemistry  against  environmental  chamber
data.   The gas  phase chemistry module  is  one  of  the  few  modules  in the
atmospheric  modeling  system  that can be  independently  tested  against
experimental data.   The  present chamber  data  base, however, is  not suit-
able for  testing all  aspects of the  mechanism,  particularly those involv-
ing predictions  of acid or  peroxy species  or  simulations of very low NOV
                                                                          A
conditions  that  prevail  in  rural  regions.   The present chamber  data  base
is best  suited for  testing  the mechanism's predictions of  ozone formation
and rates  of NOX  oxidation under  urban conditions.   To a  lesser extent,
the data can be  used to test the mechanisms' ability  to simulate formation
of   peroxyacetyl   nitrate   and   formaldehyde   under  urban   conditions.
Simulation  of  urban-like  conditions  is  important  because  urban  areas
represent  the  major sources of  the  pollutants which  are transported  to
remote  areas with sensitive receptors.  Ozone,  NOX,  PAN, and  formaldehyde
play critical  roles in long-range  transport.
                                    315

-------
     It is also  important  to recognize that even  though  the present data
base does  not  allow for direct evaluation  of  mechanisms'  predictions for
sulfuric acid and nitric acid, the testing  protocol provides some  indirect
evidence for evaluation of performance for  these species.  First,  accurate
prediction of  the rate of  NOX oxidation and  ozone formation necessitate
accurate predictions of OH radical concentrations.   According to present
theory, OH is  the sole  oxidant converting $©2  to sulfuric acid and NOp  to
nitric  acid   in   the  daylight  hours.    Second,  since  NOV  is  primarily
                                                           A
oxidized to  nitric  acid or  PAN,  the  evaluation results  for  PAN provide
some information  on  whether  the mechanism  is oxidizing NOX to the correct
products.  In  general,  if the  model cannot  at  least simulate ozone and NOX
correctly  in  one-day  smog  chamber  experiments,  it  certainly  cannot  be
trusted to  simulate accurately concentrations  of  these and other species
under multi-day, long-range transport conditions.
     A comprehensive evaluation of the RADM mechanism and the recommended
modified  versions  of  RADM  mechanism   against   available  environmental
chamber data was performed  in this study.   The approach  involved simula-
ting a large number  (-550) of  experiments from four different  chambers and
statistically  evaluating  the  mechanisms'  performance.   Experimental  data
collected  in the following chambers was used for the evaluation:

     •  The  SAPRC evacuable  indoor chamber  (EC),
     •  The  SAPRC indoor Teflon chamber (ITC),
     •  The  SAPRC outdoor Teflon chamber  (OTC), and
     •  The  University  of North Carolina  outdoor chamber (UNC).

Prior  to  conducting the  evaluation,  the  procedures  for  characterizing
chamber  dependent processes  were reviewed  and  updated where new  data were
available.   A  consistent  set of procedures for representing these was used
for all experiments from  a  given chamber.  Organic compounds  in experi-
ments  employing  complex  mixtures of organics were  aggregated using  the
types  of procedures which will be employed when the mechanism  is used  in
RADM  simulations.    Thus  there  was  no  run-to-run  adjust  of  chamber-
dependent  or mechanistic parameters employed in the evaluation protocol.
     The major  findings  from the  evaluation  of the  RADM,  RADM-M,  and
RADM-P mechanisms against  environmental chamber data are as  follows.
                                    316

-------
     •  Testing  against  experiments using  complex  organic mixtures,  that
are surrogates  for  atmospheric VOC mixtures, showed that  the RADM mechan-
ism performs as  well  as  other  chemically  up-to-date  mechanisms  (Lurmann  et
al.,  1987;  Gery et  al.,  1988;  Carter,  1988).   The  mechanism  is able  to
predict the  maximum ozone and NOX  oxidation  rate  with a  small bias and a
mean  error  of  2Q%  on the  average.   Even  though  the good performance  on
complex mixtures can in  some  cases be  a result of  partially compensating
errors  in  the mechanism,  the  mechanism  does provide ozone and NOX  oxida-
tion  rate  predictions that  are  consistent with the  available  data.   The
performance  of  the  two alternative versions of the  RADM  are  comparable  to
that  for the RADM on  complex mixture simulations.
      •  Testing  of   the  RADM  mechanism  against   single  organic-NOx-air
experimental  data  produced mixed results.   With a  few notable  exceptions,
the RADM mechanism  performed as  well as can reasonably be expected against
experiments  with the species  for which  the  mechanism was designed.   The
areas where  the predictions  of  the  mechanism can  be  considered to  be
comparable to those  of other current mechanism  are as  follows:
      •  The  predictions  for  experiments with  formaldehyde and pro-
        pene were reasonably good.
      •  The  predictions  for  acetaldehyde and  methyl ethyl  ketone
        systematically  underpredicted  ozone and NOX oxidation rates,
        but  there  are too few experiments to draw definitive conclu-
        sions on the adequacy  of this  portion of the mechanism.
      •  The  mechanism  has a  bias  towards  underprediction of  reac-
        tivity   in  ethene runs.    This is  due  to  the  omission of a
        reaction which  we  found to be  unimportant  under conditions
        where the model  will  be applied.  If this reaction is added
        to   the  mechanism,   its   performance  in  simulating  these
        experiments is acceptable.
      •  The  mean error  in  the ozone predictions in  the 1-butene  runs
        are  greater  than is  the case  for propene,  but the perform-
        ance is considered acceptable.
      •  The  predictions  for  a  limited  number  of  trans-2-butene are
         poor.   However,  there  are too  few  of these experiments to
         adequately test  this portion of  the mechanism.
                                     317

-------
     •  The  mechanism  had  large  errors  in  simulating  results  of
        alkane runs.  This  is  not  necessarily due to any failure in
        the mechanism,  but to  the extreme sensitivity of simulations
        of alkane-NOx-air experiments  to  chamber effects which tend
        to  vary  unpredictably  from run  to  run.   Compared  to the
        errors, the biases in  the predictions of ozone and NO oxida-
        tion rates were relatively low.
     The  performance  of  the  RADM-M and  RADM-P  mechanisms  in simulating
these experiments  was  similar.  The  only difference was  that the RADM-M
mechanism performed slightly worse  in  simulating some of  the higher alkane
runs because of its more condensed representation of these compounds.
     •   The  performance  of  the RADM  mechanism in  tests  against single
aromatic species was not  as good as that  reported for other chemically  up-
to-date mechanisms.  The  testing showed that  the aromatic  reactions in  the
RADM mechanism have a  bias towards overpredicting ozone and NOV oxidation
                                                                A
rates  for the more slowly reacting  compounds  and  underpredicting  those
same characteristics  for  the  faster reacting compounds.   This results in
fairly  large  errors  in  ozone  predictions  on  the average  for  toluene,
xylenes, and mesitylene.
     •  The  RADM-M and  RADM-P  mechanisms  were designed to  improve  the
performance  of  the  mechanism  in simulating   the  toluene,   xylene,  and
mesitylene  runs.   This design objective was  achieved.  The performance of
these mechanisms  in simulating the results  of aromatic  runs was comparable
to other  current mechanisms.
     •   Because  of the importance of  biogenic  emissions  in regional  model
applications,  the  RADM  mechanism has  a  separate species  to  represent
isoprene.    However,  the  performance of  this mechanism  in  simulating
isoprene  experiments was  not  as  good as  those  for  some other  alkenes.
This  is  probably  because  the mechanism assumes that  isoprene reacts to
form   the  same   photooxidation products   that   are  formed  when  propene
reacts.   Improving  the  performance of the RADM mechanism in  this  regard
would  probably  require  adding additional  species  to  the  mechanism to
represent these  products.  Research in this  area is still underway  at our
laboratories,  and  thus  no modifications in  this area are recommended at
the present time.  The RADM-M and  RADM-P therefore have the  same isoprene
mechanism  as  RADM  and   thus perform  the  same  in  simulating  these
experiments.

                                    318

-------
     •  On  the  other hand,  the RADM mechanism performed surprisingly well
in simulations  of the limited  number  of a-pinene  runs,  despite the fact
that  it  does  not represent  this  compound  explicitly.   The  RADM-M and
RADM-P mechanisms perform similarly.
     •  The  RADM mechanism's  performance for species other than  those for
which  it  was   designed,   such  as propionaldehyde,   1-hexene,  benzene,
naphthalenes, and tetralin  was  generally poor.   Much  better simulations  of
experiments  employing  those  compounds  can  be  obtained  in   mechanisms
containing  species   designed  to  represent  them.   However,  the level  of
performance  of  the   RADM mechanism  in  this  regard  is certainly  comparable
to what one  would expect from other condensed mechanisms  designed  for use
in  Eulerian  atmospheric models.    Satisfactory  performance  in  simulating
these  compounds would probably require  adding new species to  the  mechan-
ism.   This is probably not worthwhile given the relative  contributions  of
these  compounds to  overall VOC reactivity in  most  regional model applica-
tions.
     •   The  RADM,  RADM-M, and  RADM-P mechanisms  do not predict  maximum
formaldehyde  concentrations as  accurately  as  they predict  maximum  ozone
concentrations  or NO oxidation rates.   They tend to  overpredict formalde-
hyde yields  in  alkene and  alkane experiments and underpredict formaldehyde
yields  in  experiments with  single aromatic  compounds  and  organic  mix-
tures.   The mean errors associated with the  formaldehyde predictions  are
fairly  large in alkane, aromatic,  and organic mixture  experiments.  While
one  would  like  to   see  better  formaldehyde  predictions,  this does  not
necessarily  present  a problem with respect  to using  the mechanisms  for
long-range transport  and  acid deposition  modeling,  especially  given  the
uncertainty  of  the formaldehyde measurements.
      •   The mechanism  consistently overpredicts  measured  PAN  concentra-
tions.   This positive bias is expected since  the species PAN  is  used  in
the mechanism  to represent PAN  plus  numerous analogous organic nitrates.
However,  the magnitude  of the bias,  especially in  the  surrogate mixture
runs,  suggests  the  mechanism  overpredicts  the  PAN  and  analogous species
concentrations.   This  is  of  some concern  for regional  acid  deposition
modeling  because the model may underpredict nitric  acid  in  the first few
 100 km downwind of  sources (i.e.,  if NOX is preferentially oxidized to PAN
 rather than HNOo) and overpredict  the distances for  which nitrogen species
                                     319

-------
are  transported  (since  PAN  is  the  long  range  transport product  of NO
emissions).
     •  Lastly,  it  is  very unlikely  that mechanisms  could achieve the high
level  of  performance  on  ozone and  NOX  oxidation rates  found  in complex
mixture runs  without  accurately  simulating the production and destruction
of the OH radical.   These results  indirectly suggest that the mechanisms
are  able  to reasonably  predict  the  OH concentrations and, therefore, the
probable yields of nitric acid and sulfuric acid in  daylight hours.

7.4  Sensitivity Testing of Alternate Mechanistic Assumption
     Given  the nonlinear couplings  in  mechanisms,  it is  often difficult  to
predict a-priori the  conditions  and  species for  which changes in  mechanis-
tic  assumptions  will  have large  effects.  For this  reason, it is particu-
larly  important  in  evaluating alternative mechanistic assumptions to make
sure the  testing is performed over  a wide  range  of conditions.  Testing  of
alternative mechanistic  assumptions  employing an insufficiently comprehen-
sive   range   of  conditions  can  often  lead  to  misleading  conclusions.
Unfortunately,  such  testing  can not  rely on  environmental  chamber data
alone  because the  chamber data do not  cover all  the species and conditions
of  interest  for  acid deposition modeling.   To  overcome this  problem,  a
large  number  (-90)  of  test  problems  were  developed in  this  program  to
represent a wide range  of chemical conditions for  purposes of evaluating
selected  mechanistic assumptions.   Considerable  thought and analysis went
into the  design  of the test problems,  and we believe they will be as use-
ful  in future mechanism  evaluation programs as  they  have  been in  this
program.
      Based  on  our  review  of  the kinetic  and  mechanistic  assumptions
employed  in  the RADM mechanism  and the results of  the  evaluation  against
chamber data, a limited number  of alternative chemical  assumptions  were
 identified  for evaluation in sensitivity studies.   The topics for investi-
gation included:

      • The effects of various levels of condensation of the peroxy-
         peroxy radical reactions in the  RADM mechanism;

      •  The  effects  of  condensation of the  alkane reactions  in  the
         RADM-M mechanism; and
                                     320

-------
     •  The  effects  of  the  modifications   incorporated  into  the
        RADM-M and  RADM-P mechanisms on  the  predictions of species
        of interest in regional acid deposition models.

     Testing was  conducted to compare  the  RADM mechanism to an alternate
version of the mechanism  which included all possible  peroxy-peroxy  radical
combinations.   The  results  showed  that even  for hydrogen  peroxide and
organic  peroxides,  which  are  very  sensitive  to  changes in  the  peroxy-
peroxy  radical   reactions,   the   RADM  mechanism's   treatment   of   peroxy
radicals  is  justified.    Predicted  concentrations   of  key  species  were
almost  identical  for all  test problems  with the RADM  mechanism and the
detailed peroxy radical version of the  RADM mechanism.
     The method used  to represent peroxy radical  combinations in the  RADM
mechanism  is  clearly satisfactory,  and no  changes  in  this  regard are
recommended.   However,  tests  were  carried  out  to  determine  if  a  more
condensed  representation  of peroxy-peroxy radical reactions could  be  used
in the  RADM  model.   The condensed representation did  not yield predictions
which  agreed  as  closely  with those  obtained with  the detailed  radical
mechanism,   though  the   differences  are  probably not  significant.   The
benefits of  changing to the  slightly more  condensed  peroxy-peroxy  radical
reactions  scheme  are  not  considered  great enough to warrant the  change.
     Tests  were  carried out  to determine if a more  condensed  representa-
tion of the  reactions of alkanes could be employed in the RADM model.  The
results indicated  there  are no  significant  differences  in model  predic-
tions  if the two higher alkane classes in RADM are combined,  provided the
combined  class has the appropriate  mechanistic parameters.   These  results
served   as   the   basis   for   the   representation   of   the  alkanes  in  the
recommended  RADM-M  mechanism.
     Test  calculations indicated  that  the  versions  of  the  RADM mechanism
incorporating  our  recommended modifications (RADM-M and RADM-P)   can  in
some  cases  be quite  different from the original  mechanism  in predictions
of  species of interest in acid deposition models.  The differences between
the  mechanisms tend  to be the  greatest in  the simulations of urban condi-
tions.   These are  due  primarily  to  the recommended changes  in the mechan-
isms   for  the aromatics.    The  modified  mechanisms  tended   to  predict
slightly  lower  levels  of  ozone and  in  some  cases significantly  lower
                                    321

-------
levels of  peroxide  or acid species  under some conditions.   On the other
hand, the  modifications had  smaller  effects on predictions of nitric acid
or sulfate.   This indicates that  implementing the recommended changes in
the mechanism  might  in some cases have  non-negligible effects on results
of regional model simulations.
     The test calculations also indicate  that  the  two  recommended modified
versions of  the  RADM  mechanism are very  close, but  are not identical, to
each other in their predictions of  species of interest  in regional acid
deposition models.   Their predictions  of  PAN and of  the higher organic
hydroperoxides and acids  differed  in some  cases by  25% or more.  Differ-
ences in  predictions  of PAN are expected because  of different degrees of
condensation  in  the  two mechanisms — PAN  in  RADM-M is used to  represent
the  PAN   analogues   formed   from   the   unknown   aromatic   photooxidation
products,  while  RADM-P,  like RADM, uses  a  separate  species for  these  PAN
analogues.   In  the  case  of the predictions  of higher organic acids  and
hydroperoxides,  the  uncertainties  concerning  the  rates and mechanisms of
their formation  are  so great — and  the  data  suitable to  test models  for
their predictions so  inadequate — that  both mechanisms should  probably be
considered to be equally unreliable  in  this regard. Discrepancies on  the
order  of  25% or  even 50%  are well within  the  uncertainty  bounds  for
predictions  of such species.   Therefore,  for all practical purposes  the
predictions  of these  two  mechanisms can be considered to  be  essentially
equivalent.

7.5  Recommendations
     As  indicated  above,  although  we  believe that   the  RADM  gas-phase
chemical  mechanism,  along with its  emissions aggregation  system,  repre-
sents  the current  state-of-the-science,  there  remain significant  uncer-
tainties   in  this   important   model  component.     Therefore,   continuing
research  is   needed  in areas  related to  its  development  and  implementa-
tion.   Our specific recommendations  in  this regard  are given below.   The
longer  term  critical  research  needs  will be discussed  first, followed by  a
discussion of specific  shorter-term research  needs which  became  evident
during  the course of this program.
                                    322

-------
     7.5.1  Longer Term Research Needs
            The  only way  the  gas-phase  chemical mechanisms  in  airshed
models  can  be  tested  in  their  entirety,  but independently  of the many
other uncertainties  in the models  such  as emissions,  meteorology, etc.,  is
by  use  of  environmental  chamber  data.    The  presently available  chamber
data are  suitable  only  for testing the ability of models  to  predict  ozone
formation  and  NOX  oxidation  under  the  relatively  high-NOx  conditions
representative  of the  most polluted  urban  atmospheres.    Even for this
purpose,  there  are problems due  to poorly characterized  chamber effects.
There is  a  critical  need for chamber data  suitable  for testing  mechanisms
in  the  low-NOx concentration  regime  important  in   regional  and   global
models, and  for testing their abilities to predict  formation of peroxide
and  acidic  species.    Chamber  wall  effects are  such that usable  data  of
this  type cannot  be obtained  using existing environmental  chamber  tech-
nology.   This need  can  only be addressed by first  carrying out research
into  new  environmental   chamber  technologies   and  associated   surface
effects.    This  would  require  a significant  commitment  of  resources.
Unless  such  a   commitment  for  long-term  research   is  made, the  present
unsatisfactory situation will not  change  significantly.
     There still  remain many critical  areas of uncertainties in  our under-
standing  of, and  thus  our  ability  to model,  the  gas-phase  atmospheric
chemistry of many classes of organics.   The  uncertainties associated with
the  chemistry  of aromatics have  been  dwelt upon   in  "Recommendations"
sections  in  reports  such  as  this for at least a decade.  The modifications
we  recommended  to the RADM  mechanism  in this area are  based on empirical
curve-fits  to  chamber  data which are  largely  outside  the  concentration
regime  even  of  urban  conditions,  and  we have  no  way  of knowing  how
accurately  this largely empirical model  will  perform when extrapolated to
ambient conditions.   The  situation  with the higher  alkanes  only seems to
be  better because the  present  chamber data are  unsuitable for  indicating
Just  how good  or bad  the present mechanisms  really are.   An  additional
area  which  is now recognized as  being important  is  the  mechanisms of  the
biogenic  hydrocarbons.   Recent  work concerning isoprene indicates that its
OH  reaction mechanism is not as well understood  as  we once thought it  was
—  roughly  ^0% of the reaction routes  are not accounted for by  identified
products.   In addition,  there  are large  and  significant uncertainties in
                                    323

-------
the mechanism  for  its reaction with ozone.   Even less is known about  the
terpenes  — the  fact  that  the  RADM  mechanism  happens to  be  able   to
simulate  the results  of four a-pinene outdoor chamber experiments  carried
out within  a one-month  period is  probably  largely coincidental.   It  is
unknown how  well  the model  can simulate  the reactions of other biogenic-
ally emitted species such as beta-pinene or other terpenes.
     In some cases,  important uncertainties  can  be addressed by applying
presently available techniques to compounds  which have not been adequately
studied   previously.     An   example of   this  concerns   the  non-isoprene
biogenics, which is discussed  in the next  section.   However,  for compounds
such  as  the  aromatics,  the  higher  alkanes,  and even  to  some   extent
isoprene,  we have  probably approached  the  point  where  the experimental
techniques traditionally used  to study reaction mechanisms are unlikely to
yield  significant  new  information.   It  is  clear  that  the  aromatics  and
many other  compounds  of interest  form  products which  cannot  be  quanti-
tatively  measured  using present  techniques.    Many  of  these  compounds
appear  to be  highly  labile  and  contribute significantly to the  overall
reactivities of the  emitted organics.   In the case of the higher alkanes,
polyfunctional  compounds are  formed which probably do not  remain   in  the
gas phase.   Until  research laboratories  interested  in  atmospheric reac-
tions  of  organics  have the  type  of equipment  necessary to measure  and
study  such  compounds,  significant  further  progress  in  this  area   is
unlikely.    For example,  tandem  MS equipment  appears  to  be  promising  in
this regard, but  few research laboratories can afford such equipment,  and
those that have them  apparently cannot afford  to  dedicate  them to the long
periods of  time required  for the  types  of  fundamental  research projects
that are  needed.    Research into  new  types  of  equipment,   and  means  of
making  them  available for  researchers for use in fundamental studies,  is
also needed.   This  is another area  where significant new  resources  are
needed if further progress  is to be made.
     The  emissions  inventories may  not be  as  interesting  scientifically as
the subjects discussed  above,  but they  probably  represent  the greatest
area of uncertainty  in the  overall  area of gas-phase  chemistry   in  the
models.   Even  if the  gas-phase chemical  mechanism were perfectly  charac-
terized,  it  would  do us no  good if we do not know the input rates  of  the
species whose  reactions are  being  modeled.   The biogenic inventories  are
                                    324

-------
particularly uncertain; order-of-magnitude errors  in  the  present  inventor-
ies may  well be  the case.   Until  they are  better  characterized, it  may
well  not  be worth  the  effort to add  new  species in  the model  to better
represent them.  The  anthropogenic  emissions  inventories  also  have signif-
icant uncertainties,  particularly for  the  VOCs.   If there are  major errors
in the anthropogenic inventory,  it  will not matter how accurate  the  mech-
anism  is  for the  aromatics or other  anthropogenic  emissions.   Improving
the quality  of these  inventories may  be a  difficult and expensive process,
but is is absolutely  essential  if airshed  models  are  to be used for policy
assessment purposes.  This  is an area which should be given top priority.
     7.5.2   Shorter Term Needs
             We believe  that this program has  resulted in  an advance in the
processing of  emissions data for regional  models.  The aggregation of the
huge  detailed  emissions data base  into a manageable data set  which can be
used  by a variety  of models will permit for much  more flexibility in  using
alternative  chemical  mechanisms and for carrying  out  model intercomparison
studies.   However,  there  is clearly  need for further  near-term  work in
this  area.   The  first priority  should be to  improve  the software presently
used  for  processing  the NAPAP  emissions data bases.   The software intro-
duces  artificial limits to the level  of  chemical accuracy  and  range of
applicability  in the emissions aggregation  system;  for  example,  the same
aggregation  process cannot  be used for both  the  RADM and the widely-used
Carbon Bond  mechanism.   In addition,  some  categories  presently used in the
VOC  speciation  source  profiles are  chemically ambiguous and need  to be
replaced  by  those which  are not.   Examples  include "Mineral  Spirits,"
which is  not a compound but  an uncharacterized mixture,  and  "C5 Olefins,"
which is a  mixture of  isomeric compounds with quite different reactivi-
ties.   In  some cases the speciation of the emissions sources may be highly
uncertain.    However,  those  responsible for  compiling emissions profiles
are  in a better position to  make estimates concerning  the nature of these
profiles  than the chemical  mechanism  developers.  These  are not, strictly
speaking,  problems in  the  area of  chemical mechanism development, but are
areas where experts  in atmospheric   chemistry  could well   provide  much
needed guidance.
      Another area  where we believe this study has advanced the representa-
tion  of  VOC emissions  in models is the use of model surrogates with param-
                                     325

-------
eters which  are adjusted based  on the  set of  compounds  they represent.
The RADM mechanism has  been able to take advantage  of some, but not all,
of the  recent advances  in  this area.   Under  funding  for the California
ARB,  we  developed  software  and  procedures  where mechanism parameters can
be readily re-adjusted as emissions inventory change.  However, at present
the  parameters  in the  RADM mechanism  are  fixed  —  they were estimated
based on  the 1985 NAPAP anthropogenic inventory  —  but cannot readily  be
changed if this inventory changes.  It probably would not  be  too difficult
to modify the RADM mechanism and the model  software  so that  the parameters
can be changed.
     The  representation  of biogenically  emitted alkenes  is a particular
concern  in  regional  models.   Although  the  RADM mechanism represents  iso-
prene explicitly,  its performance in simulating  the chamber data is  less
than  totally satisfactory.    Probably  it  will be  necessary  to  add new
species  to  the model  to  represent  isoprene's  products satisfactorily.
However,  we  were not  able  to  complete  our investigation of how best  to
represent isoprene in the mechanism within the time and funding limits  of
this  program —  we  are  continuing  work  in   this area  under   separate
funding.
     The  terpenes  are  also   important   in the  biogenic  inventory,  and
research  is  needed  on how  best to represent them in the model.  Chamber
studies  on  the  reactivities of  the  terpenes and possibly other  biogenics
under a wide range of  conditions are clearly  needed to  test  mechanism's
abilities  to simulate at least  their effects  on  ozone  formation and NOX
oxidation.   In  addition, such data are  needed  to assess  the  ozone  forma-
tion  reactivities  of   these   compounds  relative  to  anthropogenic  VOC
emissions.   Until more  data  are  available to  test mechanisms for  these
compounds, no change  in how they are represented  in the RADM mechanism  is
warranted.
     The  scope  of this study  was  such  that we were unable  to investigate
effects  of  condensation in  the mechanism as  comprehensively as we had
hoped at the time the program was conceived.   We  believe  that research  is
still needed to investigate  the appropriateness of the tradeoffs between
condensation and chemical  accuracy  which  are  employed  in RADM and  other
condensed  chemical mechanisms.    We  also  believe  that  the  test  problems
developed  in this program are well  suited  for  this purpose.   Examples  of
                                    326

-------
areas which  need  to be investigated concern the degree of condensation of
the species  used to  represent  reactive products.   Preliminary studies of
condensation  methods  we  carried  out  for  the SAPRC  mechanism show  that
condensing   the  number  of  species  used  to  represent  primary   emitted
organics  has less of an  effect on model  predictions than condensing  the
number  of species used to  represent  reactive organic products.   The  RADM
mechanism is relatively detailed  in the number of species used to repre-
sent  primary  organic emissions,  but  is  relatively  condensed  in  its
representation of reactive products.   We  did  not  investigate  this problem
as part of this  program because by the time the other necessary tasks were
completed,  it was  too late  for  the  RADM  team  to  consider  implementing
major  changes in the mechanism for RADM-II.   However,  the  possibility of
such  changes should be considered in  time for the next major update of the
RADM  gas-phase chemical mechanism.
                                     327

-------
                                REFERENCES
Atkinson, R.,  W.  P.  L.  Carter, K.  R.  Darnall,  A.  M.  Winer,  and J.  N.
       Pitts,  Jr.  (1980), "A Smog Chamber and  Modeling Study of the Gas
       Phase N0x-Air Photooxidation of Toluene and the Cresols," Int J.
       Chem. Kinet., J12,  779-836.

Atkinson, R.,  W.  P.  L.  Carter, and A. M. Winer (1983), "Evaluation of
       Hydrocarbon Reactivities for Use in Control Strategies," Final
       Report, California Air Resources Contract No. AO-105-32, May.

Atkinson, R. and  W.  P.  L. Carter (1984), "Kinetics and Mechanisms of the
       Gas-Phase  Reactions of Ozone with Organic Compounds Under
       Atmospheric Conditions," Chem. Rev., 84, 437-470.

Atkinson, R. (1986), "Kinetics and Mechanisms  of the Gas Phase Reactions
       of the Hydroxyl  Radical with Organic Compounds Under Atmospheric
       Conditions," Chem Rev., 86, 69-201.

Atkinson, R. (1987), "A Structure-Activity Relationship for the
       Estimation of Rate Constants for the Gas-Phase Reactions of OH
       Radicals with Organic Compounds," Int.  J. Chem. Kinet., ^9,
       799-828.

Atkinson, R. (1988), "Gas-Phase Atmospheric Chemistry of Organic
       Compounds," Appendix A of Final Report, California Air Resources
       Board Contract No. A5-122-32, October.

Atkinson, R. (1989), "Gas-Phase Tropospheric Chemistry of Organic
       Compounds:  A Review," Atmos. Environ., in press.

Atkinson, R., S.  M. Aschmann, J. Arey, and W.  P. L. Carter (1989),
       "Formation of Ring-Retaining Products from the OH Radical-
       Initiated Reactions of Benzene and Toluene," Int. J. Chem.
       Kinet., in press.

Burton, C. S. (1988), "Ozone  Air Quality Models - Critical Review
       Discussion Papers," J. Air Poll. Cont.  Assoc., 38, 1119-1128.

Carter, W. P. L., A. C. Lloyd, J. L. Sprung, and J. N. Pitts, Jr.  (1979),
       "Computer Modeling of  Smog Chamber Data:  Progress in Validation
       of a Detailed Mechanism for the Photooxidation of Propene and
       n-Butane in Photochemical Smog," Int. J. Chem. Kinet,  VI,
       45-101.

Carter, W. P. L., R. Atkinson, A. M. Winer, and J. N. Pitts, Jr.  (1982),
       "Experimental Investigation of Chamber-Dependent Radical Sources,"
       Int. J. Chem Kinet., _14, 1071-1103.

Carter, W. P. L. and R. Atkinson  (1985),  "Atmospheric Chemistry of
       Alkanes," J. Atmos. Chem., 3, 377-405.
                                    328

-------
Carter, W. P. L., F. W. Lurmann, R. Atkinson, and A. C. Lloyd (1986),
       "Development and Testing of a Surrogate Species Chemical Reaction
       Mechanism," EPA-600/3-86-031, U.S. Environmental Protection Agency,
       Research Triangle Park, NC.

Carter, W. P. L., A. M. Winer, R. Atkinson, S. E. Heffron, M. P. Poe, and
       M. A. Goodman (1987), "Atmospheric Photochemical Modeling of
       Turbine Engine Fuels.  Phase II.  Computer Model Development,"
       Report, USAF Contract No. F08635-83-0278, Engineering and
       Services Laboratory, Air Force Engineering and Services Center,
       Tyndall Air Force Base, Florida,  August.

Carter, W. P. L.  (1988a), "Documentation of a Gas Phase Photochemical
       Mechanism  for Use in Airshed Modeling," Appendix B to Final Report,
       California Air Resources Board Contract No.  A5-122-32, October.

Carter, W. P. L.  (1988b) "Documentation  for  the  SAPRC  Atmospheric
       Photochemical Mechanism Preparation and Emissions  Processing
       Programs  for  Implementation  in Airshed Models," Appendix C  to
       Final Report, California Air Resources Board Contract No.
       A5-122-32, October.

Carter, W. P. L.  and R.  Atkinson  (1989a),  "Alkyl Nitrate  Formation from
       the Atmospheric  Photooxidation of Alkanes:   A Revised Estimation
       Method,"  J.  Atmos. Chem.,  in press.

Carter,  W. P. L.  and R.  Atkinson  (1989b),  "A Computer Modeling  Study of
        Incremental  Hydrocarbon  Reactivity,"  Environ. Sci.  Technol.,
        in press.

Dodge, M. C. (1989),  "A Comparison of  Three  Photochemical Reaction
       Mechanisms," J.  Geophys.  Res.,  9^.  5121-5136.

Dunker,  A.  M.,  S. Kumar, and  P.  H.  Berzins (1984),  "A Comparison  of
        Chemical Mechanisms  Used in Atmospheric  Models,"  Atmos.
        Environ., 18,  311-321.

 EPA (1984),  "Guideline for  Using the  Carbon Bond Mechanism in  City
        Specific EKMA," EPA-450/4-84-005, U.S. Environmental Protection
        Agency,  Research Triangle Park,  NC.

 EPA (1987),  "Workshop on Evaluation/Documentation of Chemical
        Mechanisms," EPA-600/9-87-024.

 Fox, D.  G.  (1981), "Judging Air Quality Model Performance," Bull. Am.
        Meteorol. Soc., 62,  599-609.

 Gery, M. W., D. L. Fox, J.  E. Jeffries, L. Stockbuger, and W.  S.  Weathers
        (1985),  "A Continuous Stirred Tank Reactor  Investigation of  the
        Gas-Phase Reactions of Hydroxyl  Radicals and Toluene," Int.  J.
        Chem. Kinet., J7, 931-955.
                                     329

-------
Gery, M. W., G. Z. Whitten, and J. P. Killus (1988),  "Development and
       Testing of the CBM-IV For Urban and Regional Modeling,"
       EPA/600/3-88/012, U.S. Environmental Protection Agency, Research
       Triangle Park, NC.

Gipson, G. L. (1984), "Users Manual for OZIPM-2:  Ozone Isopleth Plotting
       With Optional Mechansims/Version 2," EPA-450/4-84-024, U.S.
       Environmental Protection Agency, Research Triangle Park, NC.

Grosjean, D. and K. Fung (1984), "Hydrocarbons and Carbonyls in Los
       Angeles Air," J. Air Poll. Control Assoc., 34, 537-543.

Hough, A. M. (1988), "An Intercomparison of Mechanism for the Production
       of Photochemical Oxidants," J. Geophys. Res.,  93, 3789-3812.

Jeffries, H. E., K. G. Sexton, and C. N. Salmi (1981), "The Effects of
       Chemistry and Meteorology on Ozone Control Calculations Using
       Simple Trajectory Models and the EKMA Procedure," EPA-450/4-81-
       034, U.S. Environmental Protection Agency, Research Triangle Park,
       NC.

Jeffries, H. E., K. G. Sexton, R. M. Kamens, and M. S. Holleman (1985),
       Outdoor Smog Chamber Experiments to Test Photochemical Models:
       Phase II.  EPA/600/3-85/029, U.S. Environmental Protection  Agency,
       Research Triangle Park, NC.

Jeffries, H. E., K. G. Sexton, J. R. Arnold, and T. L. Kale  (1989a),
       "Validation Testing of New Mechanisms with Outdoor Chamber  Data.
       Volume 3: Calculation of Photochemical Reaction Photolysis  Rates
       in the UNC Outdoor Chamber," EPA/600/3-89/010c, U.S.  Environmental
       Protection Agency, Research Triangle Park, NC.

Jeffries, H. E., K. G. Sexton, J. R. Arnold, and T. L. Kale  (1989b),
       "Validation Testing of New Mechanisms with Outdoor Chamber  Data.
       Volume  1: Comparison of CB4 and CAL Mechanisms," EPA/600/3-89/010a,
       U.S.  Environmental Protection Agency, Research Triangle Park, NC.

Jeffries, H. E., K. G. Sexton, and J.  R. Arnold  (1989c), "Validation
       Testing of New  Mechanisms with  Outdoor Chamber Data.  Volume 2:
        Analysis of VOC Data  for the  CB4 and CAL  Photochemical
       Mechanisms,"  EPA/600/3-89/010b, U.S. Environmental Protection
        Agency, Research  Triangle  Park, NC.

Jeffries, H. E.,  K.  G. Sexton, J. R. Arnold, and T. L. Kale  (1989d),
        "Validation Testing of New Mechanisms with  Outdoor Chamber  Data.
        Volume  4:   Appendices  to Photochemical Reaction Photolysis  Rates  in
        the  UNC Outdoor Chamber,"  EPA/600/3-89/010d, U.S. Environmental
        Protection  Agency,  Research Triangle Park,  NC.

Killus,  J.  P.  and  G.  Z.  Whitten  (1982),  "A New  Carbon-Bond Mechanism for
        Air  Quality  Simulation Modeling,"  EPA-600/3-82-841, U.S.
        Environmental Protection  Agency,  Research Triangle Park,  NC.
                                    330

-------
Leone, J. A. and J. H. Seinfeld (1984), "Evaluation of Chemical Reaction
       Mechanisms for Photochemical Smog Part II - Quantitative
       Evaluation of the Mechanisms," EPA-600/3-84-063, U.S.
       Environmental Protection Agency, Research Triangle Park, NC.

Lurmann, F. W., A. C. Lloyd, and R. Atkinson (1986), "A Chemical
       Mechanism for Use in Long-Range Transport/Acid Deposition
       Modeling," J. Geophys. Res. 9J., 10905-10936.

Lurmann, F. W. and P. K. Karamchandani (1987), "An Updated Gas Phase
       Chemical Mechanism with Chlorine Chemistry for the TADAP Regional
       Model," Prepared for The Umweltbundesamt, West Germany,
       Environmental Research and Technology Document P-E142-201,
       September.

Lurmann, F. W., W. P. L. Carter, and L. A. Coyner (1987), "A Surrogate
       Species Chemical Reaction Mechanism for Urban-Scale Air Quality
       Simulation Models.  Volume I - Adaptation of the Mechanism,"
       EPA/600/3-87/Ol4a, U.S. Environmental Protection Agency,
       Research Triangle Park, NC.

Milford, J. B. (1988), "Photochemical Air Pollution Control Strategy
       Development," Ph.D. Thesis, Carnegie Mellon Univ., Pittsburgh, PA.

NASA  (1985), "Chemical Kinetics and Photochemical Data for Use in
       Stratospheric Modeling.  Evaluation Number 7," JPL Publication
       85-37, Jet Propulsion Laboratory, Pasadena, CA, July.

NASA  (1987),  "Chemical Kinetics and Photochemical Data for Use in
       Stratospheric Modeling.  Evaluation Number 8," JPL Publication
       87-41, Jet Propulsion Laboratory, Pasadena, CA, September.

NCAR  (1987),  "Development and Implementation of Chemical Mechanisms for
       the  Regional Acid Deposition Model (RADM)," EPA Interagency
       Agreement DW49930144-01-4, U.S. Environmental Protection Agency,
       Research Triangle Park, NC.

Peterson, J. T. (1976),  "Calculated Actinic Fluxes (290 - 700 nm)  for
       Air  Pollution Photochemistry Applications," EPA-600/4-76-025,
       U.S. Environmental Protection Agency.

Pitts, J. N., Jr., K. Darnall, W. P. L. Carter, A. M. Winer, and R.
       Atkinson (1979), "Mechanisms of Photochemical Reactions in  Urban
       Air," EPA-600/3-79-110, U.S. Environmental Protection Agency.

Pitts, J. N., Jr., R. Atkinson, W. P. L. Carter, A. M. Winer, and  E. C.
       Tuazon  (1983), "Chemical Consequences of Air Quality Standard and
       of Control  Implementation Programs," Final Report, California
       Air  Resources Board Contract No. A1-030-32, April.

Pitts, J. N., Jr., E. Sanhue2a, R. Atkinson, W. P. L. Carter,  A. M.
       Winer, G. W. Harris, and C. N. Plum  (1984), "An Investigation of
        the  Dark Formation of Nitrous Acid in Environmental Chambers,"  Int.
       J. Chen. Kinet., ^6, 919-939.


                                    331

-------
Plum, C. N., E. Sanhueza, R. Atkinson,  W.  P. L. Carter, and J.  N. Pitts,
       Jr. (1983), "OH Radical Rate Constants and Photolysis Rates of
       Alpha-Dicarbonyls," Environ. Sci. Technol., V7, 479-484.

Shafer, T. B. and J. H. Seinfeld (1985),  "Evaluation of Chemical Reaction
       Mechanisms for Photochemical Smog Part III - Sensitivity of EKMA
       to Chemical Mechanisms and Input Parameters," EPA-600/3-85-042.
       U.S. Environmental Protection Agency, Research Triangle Park, NC.

Shareef, G. S.f W. A. Butler, L. A. Bravo, and M. B. Stockton (1988), "Air
       Emissions Species Manual Vol 1.  Volatile Organic Compound Species
       Profiles," EPA-450/2-88-003a, U.S.  Environmental Protection Agency,
       Research Triangle Park, NC.

Stockwell, W. R. (1986), "A Homogeneous Gas Phase Mechanism for Use in a
       Regional Acid Deposition Model," Atmos. Environ., 20, 1615-1632.

Stockwell, W. R. and F. W. Lurmann (1989), "Intercomparison of the ADOM
       and RADM Gas-Phase Chemical Mechanisms," State University of New
       York, Draft Report to the Electric Power Research Institute, Palo
       Alto, CA.

Tuazon, E. C., R. Atkinson, C. N. Plum, A. M. Winer, and J. N. Pitts,
       Jr. (1983), "The Reaction of Gas-Phase N205 with Water
       Vapor," Geophys. Res. Lett., H), 953-956.

Wagner, J. K., R. A. Walters, L. J. Maiocco, and D. R. Neal (1986),
       "Development of the 1980 NAPAP Emissions Inventory,"
       EPA-600/7-86-57a, U.S. Environmental Protection Agency, Research
       Triangle Park, NC.

Weir, B. R., A. S. Rosenbaum, L. A. Gardner, G. Z. Whitten and W. Carter
       (1988), "Architectural Coatings in the South Coast Air
       Basin: Survey, Reactivity, and Toxicity Evaluation," Final Report,
       South Coast Management District, SYSAPP-88/137, Systems
       Applications, Inc., San Rafael, CA, December.

Westburg, H. and L. MacGregor (1987), "Nonmethane Organic Carbon
       Concentrations in Air Masses Advected into Urban Areas  in the
       United States," EPA-600/3-87/045, U.S. Environmental Protection
       Agency, Research Triangle Park, NC.

Whitten, G. Z., H. Hogo, and J. P. Killus (1980), "The Carbon-Bond
       Mechanism:  A Condensed Kinetic Mechanism for  Photochemical Smog,"
       Environ. Sci. Technol.,  14, 690-287.

Zafonte, L., P. L. Rieger, and J. R. Holmes (1977), Environ. Sci. Technol.
       H, 483.
                                    332

-------
                        APPENDIX A






SELECTED RESULTS OF SIMULATIONS OF ALL EXPERIMENTS MODELED




                 USING THE RADM MECHANISM
                             333

-------
             Appendix A.  Selected Results of Simulations of All E»p»r1ments Modeled Using the RAOM Mechanism.
Page  1
Experiment



1 . Pure A1 r
ITC940
ITC955
ITC100B
JN06B2R
OC0684R
OC06B4B
Group Average
S. Dev.
Avg. Abs. Value
S. Dev.
Initial
Concentrat Ions

NOx
(ppm)

0.02
0.03
0.01
0.00
0.02
0.02
0.02
0.01





Ave
Temp


Ave
H20

HC HC/NOx
( ppmC )

0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0


(degK)

0.
0.
0.
0.
0.
0.
0.
0.



0
0
0
0
3
3
1
2



303
303
303

296
296
300
3



.0
.0
.0
•
.5
.5
.4
.6


(ppm)

20300.
20300.
18200.
14200.
20000. •
20000. •
18833.
2404.


Maximum Concentrat
OZONE

Expt
(ppm)

0.072
O.OB4
O.OBB
0.203
0.097
0. 1 19
0.107
0.051



Calc
(ppm)

0.073
0.073
0.073
0.203
0.094
0.09B
0. 102
0.051


Calc
-Expt
(ppm)

0.001
0.009
-0.014
0.001
-0.003
-0.021
-0.005
0.01 1
O.OOB
O.OOB
1 on
Calc
-Expt
/Avg

0.01
0. 13
-0. IB
0.00
-0.04
-0. 19
-0.04
0. 12
0.09
O.OB
CO
co
                                                                                                                     (cont1nued)

-------
            Appendix  A.
Selected Results
(cont1nued).
of Simulations of All Experiments Modeled Using the RADM Mechanism
                                                                                                                       Page  2
U)
U)
U1
Experiment
Initial
Concent rat Ions
NOx HC HC/NOx
(ppm) (ppmC)
1. N0x-A1r
EC436
EC440
EC442
EC457
EC464
EC597
EC599
ITC695
ITC826
ITCB82
OTC185
JN2782B
AU0282R
AU2082R
AU2282R
AU2382R
OC0882R
OC0882B
ST0582R
JL2483R
JL24B3B
JL2783B
AU06B3B
Group Average
S. Dev.
Avg. Abs. Value
S. Dev.
2. CO-NOx
ITC625
ITC634
OTC188
OTC201A
OTC201B
JN2782R
AU0282B
AU2082B
AU2282B
AU2382B
ST0582B
JL2783R

1 .79
0.76
0.58
0.50
0. 19
0.56
3.40
0.50
0.90
0.70
0.28
0.44
0.39
0.41
0.46
0.43
0.30
0.30
0.50
0.31
0.48
0.43
0.37
0.65
0.68



0.28
0.60
0.34
0.37
0.76
0.45
0.40
0.41
0.46
0.43
0.50
0.47

0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.8
0.0
0.0
0.0
0.0
0.0
0.0
o.o
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.1
0.2



0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0

0,0
0.0
0. 1
0. 1
0.2
0. 1
0.0
0. 1
0.9
0.0
0.1
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
o.'o
0.0
0. 1
0.2



0. 1
0. 1
0. 1
0. 1
0. 1
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Ave
1 emp
(degK)

303,0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
298. 1
301 .3
301 .6
302.0
300.0
302.8
294.3
294.3
297 .3
305.0
305 .0
300.7
304.2
301 .6
3.0



303.0
303.0
308.8
310.9
310.8
301 .3
301 .6
302. 1
300.0
302.8
297.3
302.8
Ave
H20
(ppm)

14900.
17500.
21400.
19600.
19300.
22000.
20800.
20300.
17200.
17900.
5000. *
29300.
27700.
21000.
25600.
20200.
18100.
1B100.
14300.
19600.
19600.
20000. •
20200.
19548.
4721 .



20300.
20300.
5000. »
3880.
3000.
29300.
27700.
21000.
25600.
20200.
14300.
20000. *
Final - In1t
NO
Expt
(ppm)

-0.098
-0.063
0.069
-0.066
-0.019
-0.061
2.276
-0.021
-0.349
-0.001
-0.002
-0. 1 13
0.035
-0.011
0.000
0.056
-0.024
-0.017
0.070
-0.047
-0.071
-0.090
0.001
0.063
0.490
0. 155
0.468

-0.069
-0.040
-0.090
-0. 160
-0. 123
-0.320
-0.054
-0.111
-0.086
-0.004
0.019
-0.289
Calc
(ppm)

-0.052
-0.049
0.093
-0.042
-0.002
-0.031
1 .878
-0.008
-0. 182
-0.027
-0.004
-0.118
0.039V
-0.010
0.022
0.057
-0.032
-0.032
0.085
-0.063
-0.071
-0.076
-0.004
0.060
0.401
0. 130
0.383

-0.085
-0.061
-0. 136
-0. 157
-0.115
-0.339
-0.062
-0. 146
-0.089
-0.051
0.026

Calc
-Expt
(ppm)

0.046
0.015
0.024
0.023
0.016
0.030
-0.398
0.014
0. 167
-0.027
-0.002
-0.005
0.004
0.002
0.023
0.001
-0.008
-0.014
0.015
-0.015
0.000
0.012
-0.005
-0.004
0.094
0.038
0.085

-0.016
-0.021
-0.046
0.003
0.008
-0.019
-0.008
-0.035
-0.003
-0.047
0.007

Final - In1t
N02-UNC
Expt
(ppm)

-0.071
0.024
-0. 172
0.008
-0.002
0.041
-2.685
0.011
0.263
-0.010
-0.003
0.050
-0. 129
-0.049
-0. 147
-0. 144
0.005
0.009
-0.171
0.011
0.014
0.036
-0.090
-0. 139
0.563
0. 180
0.551

0.068
0.034
0.072
0. 162
0.096
0. 202
-0.002
0.072
0.003
-0.042
-0.086
0.214
Calc
(ppm)

-0.057
0.013
-0. 160
-0.008
-0.012
0.005
-2.536
0.001
0. 135
-0.014
-0.001
0.025
-0. 120
-0.058
-0.116
-0. 14B
-0.006
-0.005
-0. 157
-0.004
-0.010
0,009
-0.088
-0. 144
0.526
0. 160
0.521

0.078
0.052
0. 1 25
0. 143
0. 100
0.239
-0.026
0.076
-0.009
-0.046
-0.098

Calc
-Expt
(ppm)

0.014
-0.01 1
0.012
-0.017
-0.010
-0.036
0. 149
-0.010
-0. 128
-0.003
0.002
-0.025
0.009
-0.009
0.031
-0.005
-0.011
-0.014
0.014
-0.015
-0.024
-0,027
0,002
-0.005
0.045
0.025
0.037

0.010
0.017
0.053
-0.019
0.004
0.038
-0.023
0.004
-0.012
-0.004
-0.012

                                                                                                                    {cont1nued)

-------
             Appendix A.  Selected Results of Simulations of All Experiments Modeled Using the RADM Mechanism
                          (continued).
                                                                                                                       Page  3
Experiment

Croup
Avg.


Average
S. Dev.
Abe. Value
S. Dev.
Initial Ave Ave
Concentrations Temp H20
NOx
(ppm)
0.45
0.12
HC HC/NOx
(ppmC) (degK) (ppm)
0.0 0.0 303.7 17760.
0.0 0.0 4.1 8754.
Final - I nit
NO
Expt
(ppm)
-0.112
0.099
0. 1 15
0.095
Calc
(ppm)
-0.110
0.092
0.115
0.065
Calc
-Expt
(ppm)
-0.016
0.020
0.019
0.017
Final - In1t
N02-UNC
Expt
(ppm)
0.063
0.089
0.083
0.069
Calc
(ppm)
0.058
0.097
0.090
0.064
Calc
-Expt
(ppm)
0.005
0.024
0.018
0.015
                                                                                                                     (contInued)
u>

-------
             Appendix A.   Solected Results of Simulations of All Experiments Modeled Using the RADM Mechanism
                          (cont1nued) .
                                                                                                                       Page  4
to
UJ
Experiment



1. HCHO-A1r
EC250
EC255
JL17B2R
JL1782B
OC0784R
OC07B4B
OC1684R
OC1684B
Group Average
S. Dev.
Avg. Abs. Value
S. Dev.
Initial
Concent rat

NOx
(ppm)

0.02
0.02
0.01
0.01
0.00
0.01
0.00
0.00
0.01
0.01



HC
(ppmC)

0.3
0.3
0.5
0.6
0.4
1 .0
0.4
0.3
0.5
0. 2


1ons

HC/NOx
Ave
Temp


Ave
H20

Maximum Concentrat
OZONE


Expt
(degK)

14.5
20.4
71.2
86.0
96. 1
125.7


69.0
43.8



303
303
303
303
292
292
299
299
299
4



.0
.0
.3
.3
.6
.6
.7
.7
.7
.6


(ppm)

17400.
15700.
14700.
14700.
20000.*
20000.*
20000.*
20000. •
17B13.
24B3.


(ppm)

0.
0.
0.
0.
0.
0.
0.
0.
0.
0.



215
203
239
264
087
138
153
133
179
061



Calc
(ppm)

0.258
0.255
0.329
0.335
0. 124
0. 181
0. 152
0. 134
0. 221
0.085


Calc
-Expt
(ppm)

0.043
0.052
0.090
0.071
0.037
0.043
-0.001
0.000
0.042
0.031
0.042
0.031
1 on
Calc
-Expt
/Avg

0. 18
0.23
0.32
0. 24
0.36
0. 27
-0.01
0.00
0.20
0.13
0. 20
0.13
                                                                                                                     (contInued)

-------
             Appendix  A.
Selected Results of Simulations of All  Experiments Modeled Using the RADM Mechanism
(cont1nued).
                                                                                                                       Page  5
CO
Experiment
Initial
Concent rat
NOx HC
(ppm) (ppmC)
1ona
HC/NOx
Ave
Temp
(OegK)
Ave
H20
(ppm)
Maximum Concentrat
OZONE
Expt
(ppm)
Calc
(ppm)
Calc
-Expt
(ppm)
1 on
Calc
-Expt
/Avg
Maximum Concentrat
PAN
Expt
(ppm)
Calc
(ppm)
Calc
-Expt
(ppm)
1 on
Calc
-Expt
/Avg
1. Acetaldehyde-A1r
EC253
ITC627
ITC636
ITCB25
ITC957
ITC974
ITC1009
OTC200A
OTC200B
OTC206A
OTC206B
OTC234A
OTC234B
JL2683R
JL2683B
AU0483R
AU0483B
OC1584R
OC1584B
Group Average
S. Oev.
Avg. Abs . Value
S. Dev.
0.02
0.03
0.03
0.01
0.04
0.03
0.04
0.01
0.01
0.01
0.01
0.04
0.04
0.02
0.01
0.03
0.02
0.02
0.02
0.02
0.01


1 . 1
0.8
0.7
O.B
1 . 1
0.9
0.9
1 . 1
o.e
1.3
1 .0
.0
.5
. 1
. 1
.0
. 1
.0
.0
1 .0
0.2


53.8
26. 1
26.6
55.3
27.6
29.8
27.0
96.9
68.8
136.8
94.0
26.2
39.3
48.3
87.6
34.3
47.1
48.5
45.2
53.6
30.5


303.0
303.0
303.0
303.0
303.0
303.0
303.0
309.5
309.4
302.2
304.5
307.5
307.7
302.8
302.8
304.1
304. 1
296.5
296.5
303.6
3.4


17100.
20300.
20300.
17300.
20300.
19000.
20100.
3950.
2870.
8740.
7B40.
5000.*
5000.*
25900.
25900.
29300.
29300.
15200.
15200.
16242.
8541.


0. 137
0.060
0.047
0.037
0.076
0.085
0.078
O.OBB
0.076
0.030
0.023
0.083
0.084
0.422
0.331
0.548
0.431
0. 140
0. 193
0. 156
0.156


0. 100
0.050
0.050
0.017
0.051
0.048
0.052
0.081
0.080
0.063
0.087
0. 138
0. 138
0.364
0.235
0.456
0.342
0. 149
0. 195
0. 142
0. 124


-0.036
-0.010
0.003
-0.019
-0.025
-0.037
-0.026
-0.007
0.004
0.033
0.064
0.055
0.054
-0.057
-0.096
-0.092
-0.090
0.009
0.002
-0.014
0.048
0.038
0.031
-0.31
-0. 19


-0.40
-0.56
-0.40
-0.08
0.05


0.50
0.49
-0. 15
-0.34
-0. 18
-0.23
0.06
0.01
-0. 12
0.30
0.26
0. IB
0.040
0.013
0.011
0.005
0.013
0.006
0.008
0.008
0.006
0.020
0.013
0.004
0.007
0.026
0.027
0.040
0.040
0.023
0.035
0.018
0.013


0.046
0.014
0.014
0.008
0.016
0.015
0.015
0.007
0.008
0.011
0.010
0.012
0.012
0.035
0.035
0.043
0.047
0.025
0.023
0.021
0.014


0.006
0.001
0.003
0.003
0.004
0.009
0.007
-0.001
0.002
-0.009
-0.003
0.008
0.005
0.010
0.008
0.003
0.007
0.002
-0.012
0.003
0.006
0.005
0.003
0. 15
0.10
0.27
0.47
0.25
0.84
0.59
-0.08
0. 23
-0.58
-0.31
0.97
0.53
0.31
0.26
0.07
0. 16
0.08
-0.40
0.20
0.39
0.35
0.26
                                                                                                                    (contInued)

-------
             Appendl x  A .
Selected Results
(cont1nued).
of Simulations of Alt Experiments Modeled Using the RAOM Mechanism
                                                                                                                       Page  6
to
CO
Experiment
Inlt 1al
Concent rat Ions
NOx HC HC/NOx
(ppm) (ppmC)
1 . HCHO-NOx
EC251
EC252
EC3B9
EC391
EC392
ITCB64
OTC235A
OTC235B
AU0179B
AU0279B
AU0479B
AU0579B
JL2381B
OC0984R
OC0984B
Group Average
S. Dev.
Avg. Aba. Value
S. Dev.
2. Acetaldehyde-NOx
EC164
EC254
AU0179R
JN1482R
AU2482B
Group Average
S. Dev.
Avg. Abs . Value
S. Dev.

0. 1 1
0,49
4.75
4.43
8.05
0.54
0.56
0.58
0.35
0.21
0.23
0.54
0.43
0.56
0.50
1.49
2.33



0.51
0.11
0.36
0.31
0.32
0.32
0. 14



0.2
0.4
9. 1
17.7
9.5
0. 1
0. 1
0. 1
1 .0
1 .0
0.5
1 .2
1 .5
1 .0
1 .0
2.9
5. 1



0.7
1 .0
2.0
3. 1
1 .9
1 .7
1 .0



1 .7
0.7
1 .9
4.0
1 .2
0. 1
0. 1
0. 1
2.8
4.7
2. 1
2.2
3.5
1 .7
1 .9
1 .9
1 .4



1 .4
8.5
5.7
9.9
6.0
6.3
3.2


Ave
Temp
(degK)

303.0
303.0
303.0
303.0
303.0
303.0
299.2
297.9
SOB. 2
307 . 2
304.8
296.5
306. 1
295. 1
295. 1
301 .9
4.2



303.0
303.0
308.3
300. 2*
304.0
303 . 7
2.9


Ave
H20
(ppm)

17600.
17100.
2080.
2310.
877.
16000.
4430. *
3940.*
19400.
20000. *
20800.
12500.
20000.*
10900.
10900.
11922.
7438.



34000.
18900.
31300.
22300.
17400.
24760.
7462.


Maximum Concentration
OZONE
Expt
(ppm)

0.271
0.032
0.002
2.371
0.000
0.001
0.273
0.308
0.618
0.606
0.378
0.508
0.637
0.666
0.301
0.465
0.5BO



0.086
0.264
0.930
0.731
0.972
0.596
0.401


Calc
(ppm)

0.300
0.025
0.031
2. 190
0.008
0.017
0. 143
0.267
0.519
0.612
0.392
0.547
0.792
1 . 187
0.403
0.496
0.573



0.048
0. 185
0.580
0.653
0.665
0.426
0.289


Calc
-Expt
(ppm)

0.029
-0.007
0.029
-0. 181
0.006
0.017
-0. 130
-0.041
-0.099
0.006
0.014
0.040
0. 155
0.521
0. 102
0.031
0.160
0.092
0. 132

-0.037
-0.079
-0.350
-0.078
-0.307
-0. 170
0. 146
0. 170
0. 146
Calc
-Expt
/Avg

0. 10


-0.08


-0.63
-0.14
-0.17
0.01
0.04
0.07
0.22
0.56
0. 29
0.02
0.30
0.21
0. 21

-0.55
-0.35
-0.46
-0.11
-0.37
-0.37
0. 17
0.37
0. 17
Average
d( (03) -
Initial
[NO] )/dt
Calc
Expt Calc -Expt
(ppb/mln) --

3.25
1 .94
50.01
102.89
13.52
0.60
2.00
2. 16
2.06
2.12
1.17
1 .95
2.70
2.11
1.51
12.67
27.92



2.03
1 .33
1 .98
1 .62
1 .68
1 . 73
0. 29



3.77
2.13
50.67
97. 24
30.58
0.71
1 . 74
2.09
1 .96
2.06
1.11
1 .94
2.80
2.73
1 .76
13.55
27.07



1 .49
0.99
1 .69
1 .50
1 .54
1 .44
0.26



0.51
0. 19
0.66
-5.64
17.06
0.11
-0.26
-0.06
-0.10
-0.07
-0.06
-0.01
0.10
0.62
0.25
0.89
4.72
1 .71
4.47

-0.54
-0.33
-0.29
-0.12
-0.14
-0.29
0. 17
0.29
0. 17
Calc
-Expt
/Avg

0. 15
0.09
0.01
-0.06
0. 77
0.17
-0.14
-0.03
-0.05
-0.03
-0.05
0.00
0.04
0. 26
0. 15
0.09
0.22
0. 13
0. 19

-0.31
-0.29
-0. 16
-0.08
-0.09
-0. 18
0.11
0. 18
0.11
3. Propr1onaldehyde-N0x
JN1482B
0.30
3. 1
10.5
300. 1*
18900.
0.733
0.621
-0.112
-0.17
1 .75
1 .34
-0.41
-0.27
(cont 1nued)

-------
Appendix A.  Selected Results of Simulations of All Experiments Modeled Using  the RADM Mechanism
             (contInued).
                                                                                                                       Page  7
Experiment

Group
Avg.


Average
S. Dev.
Abs. Value
S. Dev.
Initial Ave Ave
Concentrations Temp H20
NOx
(ppm)
0.32
0.02
HC HC/NOx
(ppmC) (degK) (ppm)
2.5 8.0 302.1 18150.
0.9 3.4 2.8 1061.
Maximum Concentrat
OZONE
Expt
(ppm)
0.837
0. 147
Calc
(ppm)
0.608
0.019
Calc
-Expt
(ppm)
-0.229
0. 166
0.229
0. 166
Ion
Calc
-Expt
/Avg
-0.31
0.20
0.31
0.20
Average Initial
d( (031 - (NO) )/dt
Expt Calc
(ppb/mln)
1.79 1.38
0.05 0.05
Calc
-Expt
-0.41
0.00
0.41
0.00
Catc
-Expt
/Avg
-0.26
0.01
0.26
0.01
to
-Cr
O
             4. Acetone-NOx


               JN0480R
                    0.17
                            4.8
                                  27.5   302.2
                                                 17700.
                                                            0.233   0.397   0.164   0.52
                                                                                              0.48   0.52    0.04    0.09
S. Methy l«thyl ketone-NOx
OC2079R
JN0480B
Group Average
S. Dev.
Avg. Aba. Value
S. Oev.
6. Ethene-NOx
EC142
EC143
EC156
EC285
EC286
EC287
ITC926
ITC936
AU0479R
AU0579R
OC0584R
OC1 184R
OC1284R
OC0584B
Group Average
S. Dev.
Avg. Abs. Value
S. Dev.
0.22
0.18
0.20
0.03



0.48
0.50
0.50
1 .01
0.94
0.53
0.51
0.50
0.23
0.64
0.36
0.35
0.72
0.37
0.55
0.22


13.6
3.8
8.7
6.9



1 .9
4. 1
4.0
3.9
7 .5
8.0
7.9
3.9
0.9
4. 1
3.2
2.9
2.7
1 .8
4. 1
2.3


60.5
21 .6
41 .0
27.5



4. 1
8. 1
8.0
3.9
8.0
15.1
15.6
7.8
3.9
6.4
8.8
B.2
3.7
5.0
7.6
3.8


295
302
298
4



303
303
303
303
303
303
303
303
298
292
297
297
295
297
300
3


.3
.2
.8
.9



.0
.0
.0
.0
.0
.0
.0
.0
.9
.4
.3
.2
.6
.3
. 2
.6


27800.
18100.
22950.
6859.



26500.
22700.
29000.
20400.
20500.
19700.
21 100.
20300.
32500.
29 100.
20000.*
12700.
13600.
20000. •
22007.
5611 .


0
0
0
0



0
1
1
0
1
0
0
0
0
1
0
0
0
0
0
0


.563
.652
.607
.063



.782
.087
. 105
.840
.081
.965
.982
.940
.729
.294
.856
.858
.495
.675
.906
.204


0.466
0.534
0.500
0.048



0.474
0.791
0.751
0.892
1.171
1 .026
0.920
0.838
0.554
1 . 142
0.971
1 .039
0. 185
0.646
O.B14
0. 275


-0.096
-0. 1 18
-0. 107
0.015
0. 107
0.015

-0.307
-0.296
-0.354
0.052
0.090
0.061
-0.062
-0. 102
-0. 175
-0. 152
0.114
0. 182
-0.310
-0.029
-0.092
0. 179
0. 163
0.111
-0. 19
-0.20
-0. 19
0.01
0. 19
0.01

-0.49
-0.32
-0.38
0.06
0.08
0.06
-0.06
-0. 1 1
-0.27
-0. 12
0. 13
0. 19
-0.91
-0.04
-0. 16
0.30
0.23
0. 24
1 .
1 .
1 .
0.



3.
8.
8.
5.
1 1 .
13.
6.
2.
1 .
3.
2.
2.
1 .
1 .
5.
4.


25
13
19
09



20
SO
89
05
76
89
96
72
60
17
16
22
5B
48
23
10


1 .71
1 .21
1 .46
0.35



2.20
4.51
4.50
4.89
10. 13
13.00
4.91
2.34
0.99
2.27
1 .98
2.01
1 . 17
1 .33
4.02
3.53


0.46
0.08
0.27
0.26
0.27
0.26

-1 .00
-3.99
-4.39
-0.16
-1 .63
-0.88
-2.05
-0.38
-0.61
-0.91
-0. IB
-0.21
-0.41
-0. 15
-1.21
1 .39
1.21
1 .39
0.31
0.07
0. 19
0. 17
0. 19
0. 17

-0.37
-0.61
-0.66
-0.03
-0. 15
-0.07
-0.35
-0.15
-0.47
-0.33
-0.09
-0. 10
-0.30
-0. 1 1
-0. 27
0.20
0.27
0.20
                                                                                                                     (contInued)

-------
             Appendix A.   Selected Results of Simulations of All Experiments Modeled Us1n8 the RADM Mechanism
                          (cont1nued).
                                                                                                                        Page   8
oo
-fc
Experiment
Concent rat 1 ons
NOx HC HC/NOx
(ppm) (ppmC)
Avc
Temp
(degK)
Ave
H2O
(ppm)
Maximum Concentration
OZONE
Expt
(ppm)
Calc
(ppm)
Calc
-Expt
(ppm)
Calc
-Expt
/Avg
Average
d( [03] -
Initial
(NO) )/dt
Calc
Expt Calc -Expt
(ppb/m1n) --
Calc
-Expt
/Avg
7. Propene-NOx
EC 121
EC177
EC216
EC21 7
EC230
EC256
EC257
EC276
EC277
EC278
EC279
EC314
EC315
EC316
EC317
ITC693
ITC810
ITCB60
ITC925
ITC938
ITC947
ITC960
OTC186
OTC191
OTC210
OTC233
OTC236
JA1078R
OC1278B
OC207BR
OC207BB
OC21 78R
OC2578B
JN1279R
JN1279B
JN1379R
AU0279R
AU27BOB
ST0482B
ST1382B
JL1783R
JL2183R
JL2983B
JL3183R
ST2383B
0.51
0.46
0.52
0.48
0.52
0.56
0.56
0.52
0. 1 1
0.49
0.97
0.93
0.94
0.98
0.54
0.49
0.52
0.52
0.54
0.52
0.53
0.50
0.55
0.54
0.57
0.46
0.53
0.46
0.48
0.46
0.46
0.50
0.44
0.50
0.49
0.45
0.22
0.48
0.23
0.33
0.27
0.22
0.21
0.21
0.38
1.5
1 .5
1 .5
0.6
1 .9
0.4
0.7
1 .6
1 . 7
3. 1
3.5
3. 2
2.9
3.2
1 .5
3.5
2.6
3.0
2.8
2.8
1 .9
2.8
3.6
3.7
2.7
0. 1
3.3
3. 1
1 .4
1 .3
3.5
3.9
1 .3
1 .0
1 .5
2.9
1 .5
1 .9
1 . 1
1 . 1
1 . 1
1 . 1
1 . 1
1 . 1
1 .6
2.9
3.2
3.0
1 .2
3.7
0.7
1 .3
3.2
15.7
6.2
3.5
3.5
3. 1
3.3
2.8
7.2
5.4
5.8
5.2
5.3
3.6
5.5
6.6
6.9
4.8
0.2
6.3
6.9
2.9
2.9
7.7
7.9
2.9
2. 1
3.0
6.5
7.0
4.0
4.9
3.3
3.9
5.0
5.3
5.1
4.3
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303 .0
303.0
303.0
303.0
288. 7
312.0
303.0
303.0
303.0
303.0
303 . 0
303.0
303.0
303.0
299.8
309.3
304.7
303.6
300.0
265. 1
294.3
290.4
290.4
290.3
288.4
293.5
293.5
294.9
307.2
302.2
300.5
302.6
307.2
308.5
303. 1
305.0
292.5
26500.
26900.
18100.
21400.
17200.
20200.
20000.
14500.
15700.
16100.
14200.
24000.
10600.
44900.
24200.
20300.
16800.
17100.
16300.
19000.
20300.
20300.
5000. »
5000. *
3/20. *
5000. »
6510. •
20000.*
16300.
9590.
9560.
20000. •
12800.
20000. *
20000. *
20000.*
20000.*
28400.
1 1800.
24100.
29700.
23BOO.
23300.
19800.
23400.
0.506
0.540
0.564
0. 149
0.344
0.002
0.068
0.388
0.313
0.625
0.679
0.728
0.344
0.955
0.615
0.779
0.782
0.585
0.779
0.729
0.710
0.721
0.822
0.903
0.972
0.633
0.848
0.363
0.461
0.340
0.727
0.670
0.230
0.382
0.673
0.974
0.788
1 .044
0.658
0. 731
0.848
0.804
0.697
0.719
0.405
0.448
0.466
0.550
0. 227
0.309
0.010
0.094
0.435
0.386
0.685
0.692
0.776
0.638
1 .043
0.563
0.702
0.705
0.665
0.714
0.704
0.701
0.697
0.853
1 .082
0.989
0.916
0.990
0.583
0.387
0.378
0.924
0.893
0.243
0.285
0.508
0.861
0.622
1 . 180
0.607
0.678
0.713
0.725
0.733
0.682
0.654
-0.057
-0.074
-0.014
0.07B
-0.035
0.007
0.025
0.046
0.073
o.oeo
0.013
0. 048
0. 294
0.088
-0.052
-O.O78
-0.078
0.081
-0.065
-0.025
-0.009
-0.024
0.030
0. 179
0.017
0.285
0. 142
0.220
-0.074
0.037
0. 197
0. 223
0.013
-0.097
-0. 165
-0.113
-0. 166
0. 136
-0.052
-0.054
-0. 134
-0.079
0.037
-0.037
0.250
-0. 12
-0.15
-0.03
0.42
-0. 1 1

0.31
0. 1 1
0.21
0.09
0.02
0.06
0.60
0.09
-O.09
-0.11
-0. 10
0.13
-0.09
-0.03
-0.01
-0.03
0.04
0. 18
0.02
0.37
0. 15
0.46
-0.17
0. 10
0.24
0.29
0.05
-0 . 29
-0.28
-0.12
-0. 24
0.12
-0.08
-O.OB
-n. 17
-0.10
0.05
-0.05
0.47
7.46
3.69
4. 18
0.60
3.06
0.9B
3.35
3. 24
8.27
7 .72
6.64
7.21
4.33
10.64
4.06
5.07
4.25
3,57
3.72
3.61
3.34
4.23
5. 16
12.18
6.70
3.69
6.91
0.46
1 .40
1 . 23
2.83
2.53
1 .03
0.85
1 . 24
2. 69
2.41
3. 26
1 .46
1 .69
2.10
1 . 76
1 .66
1.81
1 . 23
3.80
3.71
5. 28
1 . 25
4.41
1 .03
4.06
4. 06
7. 60
9. 03
7. 38
9. 47
5.96
10.99
4. 28
5. 29
4. 22
4. 50
3.81
3.91
3. 82
4.05
7. 22
1 2. 79
5.95
5.12
7.83
0.48
1 .39
1 .33
3.42
3.32
1.10
0.74
1 .09
2. 75
2.01
2.79
.54
. 57
.72
.83
. 70
.62
.62
-3.66
-0. 18
1.10
0. 45
1 . 35
0 . 05
0.71
0.82
-0.67
1 . 30
0 .73
2 . 26
1 . 62
0.35
0. 22
0.23
-0.03
0.94
0.09
0 .30
0.48
-0.18
2.06
0.61
-0. 75
1 .43
0.93
0.01
-0.01
0. 09
0.59
0. 79
0.07
-0.11
-0. 15
0.06
-0. 40
-0.47
0.06
-0.12
-0. 38
0. 07
0. 04
-0.19
0.39
-0.65
-0 . 05
0 . 23
0 . 44
0. 36
0 . 05
0.19
0.22
-0. 08
0.16
0.10
0 . 27
0 . 32
0.03
0 . 05
0.04
-0.01
0 . 23
0. 02
0 . 08
0.13
-0.04
0.33
0. 05
-0.12
0. 32
0.13
0. 03
-0.01
0. 07
0.19
0. 27
0 .06
-0.14
-0.13
0 .02
-0.18
-0.16
0 .06
-0 . 07
-0 . 20
0 .04
0 .03
-0.11
0. 28
                                                                                                                    (contInued)

-------
             Appendix A.   Selected Reeulta of Simulations of All Experiments Mod*lad Utlng the RAOM Mechanism
                          (continued).
Page  9
ru
Experiment

Initial
Concentrat lone
NOx
Ave
Temp
Ave
H20
HC HC/NOx
(ppm) (ppmC)
OC04B4B
0011848
OC12B4B
Group Average
S. Dev.
Avg. Aba. Value
S. Dev.
8. 1-Butene-NOx
EC122
EC 123
EC124
ITC927
ITC928
ITC930
ITC935
ST23B3R
ST2583R
ST25B3B
ST2783R
Group Average
S. Dev.
Avg. Aba. Value
S. Dev.
0.36
0.36
0.68
0.49
0. 18



0.50
O.SI
0.99
0.31
0.67
0.32
0.66
0.40
0.46
0.42
0.45
0.52
0.20


1.0
2.2
2.0
2. 1
1.0



0.9
1.6
1 .7
3.6
3.B
7.2
7.6
1.5
1 .6
2.9
1 .6
3.1
2.3


2.9
6.3
2.9
4.6
2.5



1 .7
3.2
1 .7
12.3
5.7
22.2
1 1 .6
3.7
3,6
6.8
3.6
6.9
6.2


(degK)
296.7
297.2
295.9
299.9
7.5



303.0
303.0
303.0
303.0
303.0
303.0
303.0
292.6
292.3
292.3
295.5
299.4
5.0


(ppm)
20000.*
12700.
13500.
18134.
7349.



22000.
27200.
24500.
19000.
18400.
21 100.
19700.
23400.
18800.
18900.
19400.
21127.
2856.


Maximum Concentration
OZONE
Expt
(ppm)
0.446
0.674
0.432
0.608
0.238



0.227
0.506
0.247
0.646
0.022
0.717
0.872
0.206
0.266
0.594
0.28S
0.417
0.263


Calc
Calc -Expt
(ppm) (ppm)
0.457 0.011
0.757 0.083
0.43B 0.006
0.635 0.026
0.249 0.112
0.087
0.075

0.073 -0.154
0.267 -0.238
0.121 -0.126
0.712 0.066
0.037 0.015
0.807 0.090
0.949 0.077
0.431 0.225
0.427 0.161
0.767 0.174
0.430 0.145
0.457 0.039
0.316 0.150
0. 134
0.068
Calc
-Expt
/Avp
0.02
0.12
0.01
0.05
0.20
0.15
0. 13

-1.02
-0.62
-0.69
0. 10

0. 12
O.OB
0.71
0.46
0.26
0.40
-0.02
0.56
0.45
0.31
Average
d( [03] -
Expt
Calc
Initial
[NO] )/dt
Calc Calc
-Expt -Expt
(ppb/m1n) -- /Avg
1.25
2.38
1 .67
3.62
2.60



2.29
4.16
1 .99
3.24
1 .36
7.91
5.39
0.96
1 . 16
1.78
1 . 19
2.86
2.18


1.17
2.36
1.57
3.88
2.81



1.31
2. 14
1 .58
4. 19
1 .58
10.39
6. 56
1. 12
1.27
2. 13
1 . 27
3.05
2.95


-0.08 -0.07
-0.02 -0.01
-0.10 -0.06
0.26 0.06
0.87 0.19
0.57 0.14
0.70 0.13

-0.98 -0.54
-2.02 -0.64
-0.41 -0.23
0.95 0.26
0.23 0.15
2.49 0.27
1.20 0.20
0.16 0.16
0.11 0.09
0.35 0.18
0.08 0.06
0.20 0.00
1.16 0.32
0.81 0.25
0.81 0. 18
9. trans-2-Butene-NOx
EC146
EC147
EC157
ST27B3B
Group Average
S. Dev.
Av0. Abs . Value
S. Dev.
10. Isobutene-NOx
ITC694
0.51
0.9B
0.53
0.43
0.61
0.25



0.51
0.9
1 .7
0.9
2.0
1 .4
0.6



4,6
1.8
1 .7
1 .7
4.7
2.5
1.5



9. 1
303.0
303.0
303.0
295.5
301 . 1
3.7



303.0
26900.
22300.
27500.
19400.
24025.
3860.



20000.
0.247
0. 154
0.205
0.523
0.282
0. 165



0.900
0.109 -0. 137
0.084 -0.070
0.091 -0.114
0.544 0.021
0.207 -0.075
0.225 0.070
O.OB5
0.051

0.714 -0.186
-0.77
-0.59
-0.77
0.04
-0.52
0.3B
0.54
0.35

-0.23
5.87
9.83
5.96
3.00
6. 17
2.80



8.84
3.99
7.66
3.64
3.36
4.66
2.01



5.59
-1.88 -0.38
-2.18 -0.25
-2.32 -0.48
0.35 0.11
-1.50 -0.25
1.25 0.26
1.68 0.31
0.90 0. 16

-3.25 -0.45
(cont Inued)

-------
             Appendix A.  Selected Results of Simulations of All Experiments Modeled Using the RADM Mechanism
                          (continued).
                                                                                                                        Page  10
4r
U)
Experiment
Initial
Concent rat
Ions
NOx HC HC/NOx
(ppm) (ppmC)
11. 1-Hexene-NOx
ITC929
ITC931
ITC934
ITC937
Group Average
S. Oev.
Avg. Abs. Value
S. Oev.
12. Isoprene-NOx
EC520
EC522
EC524
EC525
EC527
ITC81 1
ITCB1 2
JL1680R
JL1680B
JL1780R
JL1780B
JL2381R
ST09B1R
Group Average
S. Oev.
Avg. Abs. Value

13. a-P1nene-NOx
JL1580R
JL1580B
JL2580R
JL2580B
Group Average
S. Dev.
Avg. Abs. Value
S. Dev.

0.51
0.49
1 .00
0.99
0.75
0.26



0.49
0.92
0.96
0.54
0.50
0.50
0.52
0. 18
0.18
0.46
0.47
0.43
0. 17
0.49
0.25


0. 18
0. 19
0.25
0.25
0.22
0.04


5. 1
10.3
9.7
0. 1
6.3
4.8



2.2
2.3
4.7
4.5
2. 2
3.4
1 .8
4.6
6.4
1 .0
2.6
1 .4
1 .0
2.9
1 .7


1 . 1
2.6
1 .0
0.3
1 .3
1 .0


10.0
21 . 1
9.7
0. 1
10.2
8.6



4.5
2.4
4.9
8.3
4.5
6.8
3.4
26/0
36.6
2. 1
5.5
3.4
6. 1
8.6
10.4


5.8
13.8
4.0
1 .4
6.3
5.3

Ave
Temp
(UegK)

303.0
303.0
303.0
303.0
303.0
0.0



303.0
303.0
303.0
303.0
303.0
303.0
303.0
305.0
301 .9
305.6
302.8
306. 1
297.5
303. 1
2. 1


302.4
299.6
302.4
300 . 1
301 . 1
1 .5

Ave
H20
(ppm)

19000.
20100.
20300.
20300.
19925.
624.



13000.
13000.
13000.
13000.
13000.
17300.
17300.
.30200.
26000.
27000.
25000.
20000.*
20000. •
19123.
6284.


30300.
29200.
31700.
2630O.
29375.
2291 .

Maximum Concentration
OZONE
Expt
(ppm)

0. 298
0.606
0.428
0.007
0.335
0.252



0.503
0.276
0. 759
0.691
0.547
0.919
0.768
0.652
0.837
0.606
1 .298
0.750
0.506
0.716
0.246


0.201
0.470
0.377
0.334
0.346
0.112

Calc
(ppm)

0.647
0.799
0.876
0.047
0.593
0.376



0.300
0. 1 18
0.429
0.630
0.317
0.577
0.253
0.712
0.600
0.420
0.835
0.691
0.457
0.488
0.208


0.299
0.420
0.326
0.211
0.314
0.086

Calc
-Expt
(ppm)

0.349
0. 193
0.448
0.040
0.258
0. 179
0.258
0. 179

-0.203
-0. 158
-0.330
-0.061
-0. 230
-0.311
-O.M5
0.060
-0. i'37
-0. 386
-0.463
-0.058
-0. 049
-0.229
0. 174
0. 238
0. 160

0.098
-0.051
-0.05 1
-0.124
-0.032
0.093
0.08 1
0.036
Calc
-Expt
/AVH

0.74
0.27
0.69
0.57
0.25
0.57
0.25

-0.50
-0.80
-0.55
-0.09
-0.53
-0.46
-1.01
0.09
-0.33
-0.63
-0.43
-0.08
-0.10
-0.42
0.31
0.43
0.29

0.39
-0. 1 1
-0.14
-0 . 45
-0 .08
0.35
0.28
0.17
Average
d( [03] -
Initial
(NO] )/dt
Calc
Expt Calc -Expt
(ppb/m1n) --

1.27
2.82
1 .84
0.33
1 .57
1 .04



5.30
6.00
15.57
10. 13
5.97
10.00
4.86
3. 18
2.88
1 .96
4. 20
2. 13
1 .46
5.66
4.06


0.61
1 .59
0.84
0.84
0.97
0.43


3.06
9.84
4.59
0.56
4.51
3 .92



7.56
6.75
15.94
22. 16
7.95
8.95
2.42
3.00
2.43
1 .36
2.87
2.01
1 .45
6.53
6.29


0.69
2.02
0.83
0.69
1 .06
0.64


1 .79
7.02
2.75
0.22
2.95
2 9 1
2.95
2.91

2.26
0. 76
0.37
12.03
1 .97
-1 .05
-2.43
-0. IB
-0.44
-0.60
-1 .33
-0.12
-0.01
0.66
3.58
1 .81
3. 18

0.08
0.43
-0.01
-0.15
0 .09
0.25
0.17
0.18
Calc
-Expt
XAvg

0.63
1.11
0.86
0.50
0.62
0 25
0.82
0.25

0.35
0.12
0.02
0.75
0.28
-0 1 1
-0.67
-0.06
-0.17
-0.36
-0.38
-0.06
0.00
-0.02
0.36
0. 26
0. 24

0.12
0. 24
-O.O 1
-O. 2O
0 04
0. 19
0. 14
0.10
                                                                                                                    (cont1nued)

-------
Appendix A.   Selected Results of Simulations of All  Experiments Modeled Using the RADM Mechanism
             (centInued).
                                                                                                          Page 11
Experiment
Initial
Concantrat Ions
NOx HC HC/NOx
(ppm) (ppmC)
14. Ethane-NOx
ITC999
t5. n-Butana-NOx
EC130
EC133
EC 134
EC137
EC 162
EC 163
EC16B
EC 178
EC304
EC305
EC306
EC307
EC30B
EC309
ITC507
ITC533
ITC770
ITC939
ITC946
OTC211
JL2178R
JL2178B
JL2278R
JL227BB
STI87SB
OC0979R
OC1879B
Group Average
S. Dev.
Avg. Abs. Value
S. Dev.
16. C4+ Branched
EC 165

0.09

0. 10
0.50
0.51
0.50
0.51
0.49
0.49
0.10
0.47
0. 10
0. 19
0. 10
0.4B
0.47
0.09
0.12
0.52
0.51
0.26
0.55
0.24
0.24
0.55
0.55
0.21
0.21
0.20
0.34
0. 18


Alkane-NOx
0. 10

45.4 534.1

17.6 179.3
8.6 17.1
8.3 16.3
8.7 17.3
8.2 16.3
9.0 18.3
8.0 16.2
7.B 79.6
17.1 36.7
15.7 159.7
25.6 138.2
25.8 252.9
16.2 33.6
17.2 36.3
15.2 165.0
11.9 99.8
37.9 72.8
14.8 28.9
10.0 38.2
42.8 77.5
7.2 29.9
15.4 63.9
7.9 14.4
17.5 31.5
21.2 103.1
14.6 71.0
14.3 71.8
15.7 69.8
8.8 61.6



11.3 114.3
Avp
Temp
(dagK)

303.0

303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
302.6
303.0
288.7
310.9
303.0
303.0
303.0
303.0
303.0
300. 1
302.9
302.9
305.2
305.2
298.3
297.2
295.0
302. 1
3.8



303.0
Ave
H20
(ppm)

20500.

26200.
27800.
20000."
20000.*
27300.
28500.
20700.
20900.
26800.
25400.
24200.
29700.
8750.
12400.
20300.
20800.
16800.
20300.
20300.
3530.
20000.
20000.
20000.
20000.
20000.
15300.
28100.
20892.
6033.



1B600.
Maximum Concentration
OZONE
Expt
(ppm)

0.243

0.459
0.249
0.034
0.042
0.112
0.454
0.655
0.384
0.362
0.398
0.535
0.420
0.047
0.545
0. 149
0. 165
0.042
0.017
0.054
0.008
0.763
0.986
0. 166
0.788
0. 185
0. 191
0. 208
0.312
0.268



0.48B
Calc
(ppm)

0.202

0.45B
0.052
0.042
0.050
0.056
0. 158
0.340
0.432
0.494
0.547
0.731
0.606
0. 240
0.678
0.397
0.356
0.059
0.028
0. 159
0.258
0.639
0.910
0. 161
0.484
0. 129
0.320
0. 193
0.332
0.245



0.482
Calc
-Expt
(ppm)

-0.042

-0.001
-0. 197
0.008
o.ooa
-0.056
-0.296
-0.315
0.049
0. 132
0. 149
0. 196
0. 186
0.193
0. 133
0.24B
0. 191
0.017
0.01 1
0. 105
0.250
-0. 124
-0.077
-0.006
-0.304
-0.056
0. 129
-0.015
0.021
0. 163
0. 128
0.101

-0.006
Calc
-Expt
/Avg

-0. 19

0.00
-1.31


-0.67
-0.97
-0.63
0. 12
0.31
0.32
0.31
0.36

0. 22
0.91
0.73


0.99

-0. IB
-0.08
-0.03
-0.48
-0.36
0.51
-0.08
0.00
0.59
0.46
0.36

-0.01
Average Initial
d( (O3] - |NO] )/dt
Calc
Expt Calc -Expt
(ppb/mln) --

1 .32

4.41
2.42
0.94
1 .02
1 .73
3.31
2.03
1 .61
2.09
2.39
2.38
2.61
1 .04
2.00
0.69
0.61
1 .56
0.36
0.53
0.56
1.17
1 . 64
0.9O
1 .55
0.51
0.60
0.60
1 .53
0.98



1 .77

1.41 0.08

2.72 -1.70
1.13 -1.29
1.36 0.42
1.33 0.32
1.27 -0.45
1.74 -1.57
1.21 -0.82
1.80 0.20
2.49 0.40
3.74 1.35
4.59 2.21
6.34 3.74
2.84 1.80
2.69 0.70
.68 1.00
.45 0.84
2.13 0.57
0.68 0.32
.17 0.64
.51 0.94
.09 -0.08
.50 -0.13
0.77 -0.13
1.33 -0.22
0.45 -0.06
0.62 0.03
0.54 -0.06
1.86 0.33
1 .32 1.13
0.81
0.84

1.60 -0.16
Calc
-Expt
/Avg

0.06

-0.48
-0.73
0.37
0.27
-0.30
-0.62
-0.51
0. 12
0. 17
0.44
0.63
0.83
0.93
0.30
0.84
0.82
0.31
0.61
0.75
0.91
-0.07
-O.OB
-0.16
-0. 15
-0. 13
0.04
-0.11
0.19
0.50
0.43
0.29

-0. 10
(cont 1nued)

-------
            Appendt x A.
Selected Results
(cont Inued).
of Simulations of All Experiments Modeled Using  the  RADM  Mechanism
                                                                                                                        Page 12
4=
VJ1
Experiment

Initial
Concentrat Ions
NOx
HC HC/NOx
(ppm) (ppmC)
EC171
OC1B79R
OC2079B
AU1983R
AU1983B
Group Average
S. Dew.
Avg. Abs. Value
S. Dev.
17. C5+ n-Alkane-NOx
EC 135
OC0979B
EC 131
ITC559
ITC538
ITC540
ITC552
ITC761
ITC762
ITC763
ITC797
ST1879R
EC155
ITC1001
Group Average
S. Dev.
Avg. Aba. Value
S. Dev.
0. 10
0.20
0.22
0.38
0.37
0.22
0.11



0. 10
0.21
0. 10
0. 19
0. 1 1
0. 1 1
0. 13
0.52
0.27
0.28
0.52
0.21
0. 10
0. 1 1
0.21
0. 14


3.5 35.7
16.4 81.9
12.6 56.5
4.7 12.6
4.1 10.9
8.2 47.9
5.2 38.7



20.4 212.7
15.1 73.3
24.6 251.1
279.4 1441.1
60.3 529.0
274.8 2421 .3
428.8 3278.4
75.2 145.9
74.7 280.4
7.7 27.7
7.3 14.0
6.3 30.4
37.3 385.1
2.4 22.4
93.9 650.9
133.5 1015.3


Ave
Temp

(degK)
303.0
295.0
295.3
307. 1
307. 1
301 .9
5.0



303.0
297.2
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
298.3
303.0
303.0
302.3
1 .9


Ave
H20

(ppm)
19000.
28100.
27800.
21300.
21300.
24143.
5463.



25000.
15200.
25000.
20300.
18900.
20300.
20300.
17500.
17000.
17100.
12200.
20000.*
26200.
17500.
19464.
3919.


Maximum Concentrat
OZONE
Expt
(ppm)
0.403
0.236
0.217
0.088
0.057
0.283
0.181



0.435
0. 184
0.393
0.377
0. 150
0.360
0.315
0.030
0. 105
0.041
0.004
0. 122
0.264
0.036
0.201
0. 153


Calc
(ppm)
0.318
0. 120
0.056
0.213
0.087
0.231
0. 157



0.549
0.226
0.528
0.601
0.313
0.384
0.358
0.036
0. 186
0.073
0.01 1
0.332
0.524
0.078
0.300
0.204


Calc
-Expt
(ppm)
-0.085
-0. 116
-0. 161
0. 125
0.030
-0.052
0. 106
0.096
0.059

0.114
0.042
0. 135
0. 224
0. 162
0.024
0.042
0.006
0.081
0.032
0.007
0.210
0.260
0.042
0.099
0.086
0.099
0.086
1 on
Calc
-Expt
/Avg
-0.24
-0.65
-1.18
0.83
0.41
-0. 17
0.67
0.53
0.39

0. 23
0.21
0. 29
0.46
0. 70
0.06
0.13

0.55


0.92
0.66

0.42
0. 28
0.42
0. 28
Average
d( (03] -
Expt
Calc
Initial
(NO) )/dt
Calc
-Expt
-- (ppb/m1n) —
1 .30
0.64
0.66
0.61
0.53
0.93
0.46



2.92
0.59
1 .92
1 .79
0. 74
1 .85
1 . 19
1 .08
0.83
0.68
0.64
0.40
1 .33
0. 15
1.15
0.75


0.82
0.49
0.38
0.66
0.55
0.75
0.41



2.89
0.53
2.69
2.05
1 .05
1.12
1 .09
1 .27
1 .53
1 .45
0.98
0.45
1 .69
0.43
1 .37
0.76


-0.48
-0. 15
-0.28
0.05
0.02
-0. 18
0. 19
0.20
0. 16

-0.02
-0.06
0.77
0.26
0.31
-0.73
-0. 10
0. 19
0.70
0.77
0.34
0.05
0.36
0.28
0. 22
0.40
0.35
0.28
Calc
-Expt
/Avg
-0.46
-0.27
-0.54
0.08
0.04
-0.22
0.24
0.26
0.19

-0.01
-0. 1 1
0.33
0. 14
0.35
-0.49
-0.08
0. 16
0.59
0.72
0.42
0. 12
0.24
0.97
0.24
0.37
0.34
0.28
18. Methyl eye lohexane-NOx
ITC765
ITC766
ITC767
ITC800
Group Average
S. Dev.
Avg. Abs. Value
S. Dev.
0.53
0.26
0.55
0.54
0.47
0.14


0.0 0.1
0.0 0.1
0.1 0.1
0.0 0.1
0.0 0.1
0.0 0.0


303.0
303.0
303.0
303.0
303.0
0.0


17900.
17100.
17900.
17500.
17600.
383.


0.022
0.121
0.041
0.015
0.050



0.012
0. 169
0.031
0.01 1
0.056



-0.010
0.047
-0.010
-0.004
0.006

0.018


0.33


0.33

0.33

0.86
0.87
1 .09
0.86
0.92
0. 12


0.70
1 .37
1 .08
0.76
0.98
0.31


-0.16
0.50
-0.01
-0.09
0.06
0.30
0. 19
0. 22
-0.20
0.45
-0.01
-0. 1 1
0.03
0.29
0. 19
0. 19
                                                                                                                     (contInued)

-------
 Appendix A.   Selected Result* of  Simulations  of  All  Experiments  Modeled Using the RAOM Mechanism
              (cont1nued) .
                                                                                                           Page 13
Experiment

19. Benzene-NOx
ITC560
ITC561
ITC562
ITC698
ITC710
ITC831
Group Average
S. Oev.
Av0. Abs. Value
S. Oev.
20. Toluene-NOx
EC264
EC265
EC266
EC269
EC270
EC271
EC272
EC273
EC327
EC336
EC337
EC339
EC340
ITC699
ITC828
JL3080R
AU2780R
AU2782B
OC2782R
AU0183R
Group Average
S. Dev.
Avg . Aba. Value
S. Dev.
Initial
Concent rat Ions
NOx
(ppm)

0.12
0, 11
0.56
0.50
0.55
1 .01
0.47
0.33



0.4B
0.48
0.49
0.47
0.46
0.21
0.48
0. 11
0.45
0.44
0.45
0.44
0.43
0.5t
1.02
0. 18
0.48
0.43
0.39
0.39
0.44
0. 18


HC
(ppmC)

332.3
79.1
83.8
83.5
83.6
12.2
1 12.4
1 t 1 .3



8. 1
7.5
8.4
4.0
4.2
8.0
4. 1
4. 1
4.0
7.2
7.9
5.0
4. 1
10.5
3.0
3.9
2.3
3.0
4.5
4.6
5.4
2.3


HC/NOx

2874.4
694.2
149.7
167.4
151 .0
12. 1
674.6
1103.2



17.0
15.6
17.0
8.4
9.0
37.4
8.5
37.2
8.9
16.3
17.7
11.3
9.5
20.8
3.0
21.3
4.8
7.0
11.7
11 .8
14.7
9.2


Ave
Temp
(degK)

303.0
303.0
303.0
303.0
303.0
303.0
303.0
0.0



303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
304.5
302.2
301 .7
289.8
306. 1
302,5
3. 1


Ave
H20
(ppm)

17100.
17600.
18800.
21900.
21000.
18300.
19117.
1920.



18900.
21900.
19400.
19100.
20000.
20500.
19400.
21200.
22400.
26900.
27000.
26800.
25900.
20300.
17200.
35200.
32000.
24700.
19500.
23700.
23100.
4692.


Maximum Concentration
OZONE
Expt
(ppm)

0.323
0.273
0.412
0.374
0.367
0.021
0.295
0. 142



0.419
0.393
0.405
0.318
0.369
0.296
0.410
0.215
0.376
0.396
0.325
0.225
0.344
0.485
0.021
0.273
0.736
0. 1 16
0. 123
0.458
0.335
0. 154


Calc
(ppm)

0.347
0.285
0.556
0.507
0.542
0.007
0.374
0. 21 1



0.427
0.426
0.432
0.380
0.420
0.307
0.374
0.223
0.423
0,463
0.427
0.404
0.419
0.215
0.006
0.484
0.985
0.487
0.426
0.722
0.423
0. 190


Calc
-Expt
(ppm)

0.024
0.013
0. 144
0. 133
0. 176
-0.014
0.079
0.081
0.084
0.075

0.008
0.032
0.026
0.063
0.051
0.012
-0.036
0.008
0.047
0.067
0. 102
0. 179
0.074
-0. 270
-0.015
0.210
0. 249
0.371
0.303
0.264
0.087
0. 143
0.119
0.116
Calc
-Expt
/Avg

0.07
O.OS
0.30
0.30
0.39

0.22
0. 15
O.22
0.15

0.02
0.08
0.06
0. 16
0. 13
0.04
-0.09
0.04
0.12
0. 16
0. 27
0.57
0. 19
-0.77

0.56
0.29
1 .23
1 . 10
0.45
0.24
0. 43
0. 33
0.36
Average
d( [03] -
Initial
(NO) )/dt
Calc
Expt Calc -Expt
(ppb/mln) --

7.01
4.75
2.85
2. 87
2.70
0.17
3.39
2.30



4.46
3.56
4.62
2.55
3.72
6.56
3.69
5.90
2.49
6.06
2.55
1 .53
2.50
4. 72
0.49
1 .20
1 .76
0.63
0.66
1.75
3.07
1 .87



9.61
4,79
4.40
4.37
4.21
0.30
4.61
2.96



4.94
4.83
5. 10
2.78
4. 16
4.92
2.61
3.09
3.05
6.82
4. 22
3.12
2.90
2.34
0.88
1. 10
1 .65
0. 72
0.88
1 .88
3. 11
1 .68



2.59
0.04
1 .55
1 .49
1.51
0. 13
1 .22
0.97
1 .22
0.97

0.49
1 .27
0.48
0.23
0.44
-I .64
-0.88
-2.81
0.56
0. 76
1 .67
1 .59
0.40
-2.38
0.39
-0. 10
-0. 1 1
0.08
0.22
0. 13
0.04
1.17
0.83
0.80
Calc
-Expt
/Avg

0.31
0.01
0.43
0.41
0 .44
0.56
0.36
0.19
0.36
0. 19

0.10
0.30
0.10
0.09
0.11
-0.29
-0. 27
-0.63
0.20
0.12
0 .49
0.68
0. 15
-0.67
0.57
-0.08
-0.06
0.12
0. 28
0.07
0,07
0.34
0. 27
0.22
21.  Xylene-NOx
                                                                                                        (cont1nued)

-------
            Appendix  A.
Selected Results
(continued).
of Simulations of All Experiment* Mod*lad Using the RADM Mechanism
                                                                                                                       Page 14
U>
Experiment

Initial
Concent rat 1 ons
NOx
HC

HC/NOx
(ppm) (ppmC)
EC343
EC344
EC345
EC346
ITC702
ITC827
JL3080B
AU2782R
OC27B2B
AU0183B
Group Average
S. Oev.
Avg. Abs . Value
S. Dav.
22. Mes1tylene-N0x
EC900
EC901
EC903
ITC703
ITC706
ITC709
ITC742
1TC826
Group Average
S. Dev.
Avg. Abs. Value
S. Dev.
23. Tetral1n-N0x
ITC739
ITC747
ITC748
ITC750
ITC832
Group Average
S. Dev.
Avg. Abs. Value
S. Dev.
0.28
0.67
0.28
0.26
0.52
1 .07
0. 18
0.43
0.39
0.37
0.4S
0.26



0.53
0.51
1 .00
0.50
0.49
0.99
0.48
0.90
0.68
0.24



0.52
0.50
0.22
0.53
1.00
0.55
0.28


4
4
3
3
4
1
2
2
2
2
3
1



5
2
4
5
2
4
4
0
3
1



2
93
84
44
39
52
36


.2
.0
.7
.9
.0
. 2
. 2
.0
.8
.7
. 1
.0



.4
. 7
.7
.3
. 7
.7
.6
.6
.9
. 6



.4
. 2
.0
.5
.4
.7
.7


14.9
5.9
13.3
14.8
7.8
1 . 1
12.4
4.6
7.0
7.2
8.9
4.7



10.2
5.2
4.7
10.6
5.4
4.7
9.7
0.9
6.4
3.4



4.6
187 .4
385. 1
84.6
39.5
140.2
153.2


Ave
Temp

(degK)
303.0
303.0
303.0
303.0
303.0
303.0
304.5
301 .7
289.8
306.0
302.0
4 .4



303.0
303.0
303.0
303.0
303.0
303.0
303,0
303,0
303.0
0.0



303.0
303.0
303.0
303.0
303.0
303.0
0.0


Ave
H20

(ppm)
29500.
28900.
25200.
25900.
20300.
14400.
28000.
24700.
19500.
23700.
24010.
4737.



20000. •
20000.*
20000. *
21000.
21000.
21000.
18100.
17200.
19788.
1419.



17900.
17100.
17500.
17100.
16800.
17280.
427.


Maximum Concentration
OZONE
Expt
(ppm)
0.283
0.589
0.396
0.384
0.627
0.021
0.555
0.491
0.396
0.688
0.443
0. 195



0.381
0.384
0.502
0.707
0.641
0.779
0.773
0.022
0.524
0.258



0.002
0.508
0.370
0.482
0.073
0.287
0.235


Calc
(ppm)
0.354
0.515
0.362
0.349
0.338
0.009
0.519
0.601
0.422
0.777
0.425
0.202



0.417
0.402
0.591
0.373
0.318
0.214
0.406
0.007
0.341
0.171



0.202
0.659
0.484
0.599
0.025
0.394
0.271


Calc
-Expt
(ppm)
0.071
-0.073
-0.033
-0.034
-0.2B9
-0.012
-0.036
0.111
0.026
0.089
-0.018
0. 1 13
0.077
0.081

0.036
0.019
0.089
-0.334
-0.322
-0.565
-0.367
-0.015
-0. 183
0.243
0.218
0.206

0.200
0. 150
0.114
0.117
-0.048
0. 107
0.093
0. 126
0.056
Calc
-Expt
/Avg
0.22
-0. 13
-0.09
-0.09
-0.60

-0.07
0.20
0.06
0. 12
-0.04
0. 25
0. IB
0. 17

0.09
0.05
0.16
-0.62
-0.67
-1.14
-0.62

-0. 39
0.50
0.48
0.40


0.26
0. 27
0.22
-0.98
-0.06
0.62
0.43
0.37
Average
d( f03] -
Expt
Calc
Initial
[NO] )/dt
Calc
-Expt
(ppb/mln) --
8. 18
10.72
11 .35
7.65
7.79
1.12
1.97
1.31
1 .43
2.39
5.39
4. 14



3.85
8.88
14.96
14.59
7.20
1 1 .74
13.14
1 .68
9.50
4.97



0.63
2.73
2.60
2.10
1 .38
1 .89
0.88


9.39
9.23
9.13
9.14
3. 10
1 .25
1.96
1 .69
2.04
2.91
4.98
3.69




5.07
8.16
3.49
2. 13
3. 10
2.95
0.85
3.68
2.36



1 .47
31.12
20. 16
18.59
0.99
14.47
13.01


1.21
-1 .50
-2.22
1 .49
-4.69
0.13
-0.01
0.38
0.61
0.52
-0.41
1 .89
1 .28
1 .39


-3.81
-6.80
-11.10
-5.08
-8.64
-10.20
-0.83
-6.64
3.67
6.64
3.67

0.84
28.40
17.56
16.48
-0.38
12.58
12.21
12.73
12.01
Calc
-Expt
/Avg
0.14
-0.15
-0.22
0. 18
-0.86
0.11
0.00
0.26
0.35
0.20
0.00
0.35
0. 25
0. 23


-0.55
-0.59
- 1 . 23
-1 .09
-1.16
-1 .27
-0. 66
-0.93
0.32
0.93
0.32

0.60
1 .68
1 .54
1 .59
-0.33
1 .06
0.85
1 . 19
0.60
                                                                                                                    (cont\nued )

-------
            Appendix A.
Selected Reaults of Simulations of AM
(continued).
Experiment* Modeled Using the RAOM Mechanism
                                                                                                                       Page IS
Xr
CO
Experiment
Initial
Concent rat
NOx HC
(ppm) (ppmC)
Ions
HC/NOx
Ave
Temp
(degK)
Ave
H20
(ppm)
Maximum Concentrnt
OZONE
Expt
(ppm)
Calc
(ppm)
Calc
-Expt
(ppm)
1 on
Calc
-Expt
/Avg
Average Initial
d( |03] - [NO] )/dt
Ca Ic Ca 1 c
Expt Calc -Expt -Expt
(ppb/m1n) -- /Avg
24. Naphtha lene-NOx
ITC751
ITC755
ITC756
ITC79B
ITC802
Group Average
S. Dev.
Avg. Abe. Value
S. Dev.
0.52
0.24
0.26
0.53
0.53
0.42
0.15


7.5
14. 1
27.4
19.4
8.4
15.4
6.3


14.3
58.3
106.8
36.8
15.9
46.4
38.2


303.0
303.0
303.0
303.0
303.0
303.0
0.0


15100.
17500.
17200.
16400.
16800.
16600.
93b.


0.113
0.259
0.282
0.204
0. 124
0. 196
0.077


0.505
0.392
0.451
0.521
0.503
0.474
0.053


0.392
0.132
0. 168
0.317
0.379
0.278
0. 120
0.278
0. 120
1 .27
0.41
0.46
0.88
1 .21
0.84
0.40
0.84
0.40
1 .08
1.57
2.20
1 .66
1 .42
1.59
0.41


4.76 3.68 1.26
8.68 7.11 1.39
13.91 11.71 1 .45
10.47 8.80 1.45
5.34 3.91 1.16
8.63 7.04 1.34
3.78 3.39 0.13
7.04 1.34
3.39 0.13
25. 2. 3-D1methy1naphtha lene-NOx
ITC775
ITC771
ITC806
ITC774
Group Average
S. Dev.
Avg. Abs. Value
S. Dev.
26. Simple Mixture
EC 144
EC145
EC 160
EC 149
EC150
EC151
EC152
EC153
EC161
OC1278R
OC2578R
AU0180R
AU14BOR
EC166
EC172
ST0682R

0.29
0.26
0.33
0.56
0.36
0. 14


Runs
0.51
0.99
0.99
0.99
1 .00
2.06
0.50
0.97
0.51
0.48
0.44
0.56
0.47
0.10
0. 10
0.46

1 .7
4.8
5.9
4.0
4. 1
1 .8



4.7
3.4
3.2
2.0
3.5
5.2
3.7
6.6
3.2
1 .4
1 .4
0.5
1 .4
9.2
2.8
2.8

5.8
18.0
17.7
7. 1
12.2
6.6



9.3
3.4
3.3
2.0
3.5
2.5
7.3
6.8
6.4
3.0
3. 1
0.8
3.0
92.0
28.9
6.2

303.0
303.0
303.0
303.0
303.0
0.0



303.0
303.0
303.0
303.0
303. U
303.0
303.0
303.0
303.0
294.3
288.4
304.9
305.2
303.0
303.0
299. 1

16800.
20300.
17100.
22700.
19225.
2806.



24400.
24400.
29200.
29300.
29900.
30400.
26400.
25400.
36900.
16100.
12800.
23400.
34700.
27700.
21600.
22200.

0.274
0.293
0.360
0.341
0.317
0.040



1 .065
0.777
0.874
0. 286
0. 799
0. 147
0. 791
1 .050
U.B57
0.260
0. 147
0. 256
0.863
0.462
0.369
0.378

0. 185
0.339
0.388
0.313
0.306
0.087



0.773
0.365
0.316
0. 125
0.327
0.082
0.665
0.820
0.600
0.250
0. 125
0. 107
0.484
0.395
0.310
0.619

-0.089
0.045
0.028
-0.028
-0.011
0.061
0.048
0.029

-0.292
-0.411
-0.559
-0.161
-0.472
-0.064
-0. 126
-0.230
-0.25B
-0.010
-0.022
-0. 148
-0.379
-0.067
-0.059
0.240

-0.39
0. 14
0.08
-0.08
-0.06
0.24
0. 17
0. 15

-0.32
-0.72
-0.94
-0.78
-0.84
-0.56
-0.17
-0.25
-0.35
-0.04
-0.16
-O.B2
-0.56
-0. 16
-0.17
0.48

1 .73
2.67
2.69
2.84
2.48
0.51



10.53
5.11
5.86
10.83
6.30
8.43
10.41
19.23
9.94
1 . 14
0.94
0.98
2.13
2. 1 1
1 .00
1.11

1 .28 -0.44 -0. 29
3.20 0.53 0.18
3.35 0.66 0.22
2.61 -0.23 -0.08
2.61 0.13 0.01
0.94 0.55 0.24
0.47 0.19
0.18 0.09

6.09 -4.44 -0.53
3.71 -1.40 -0.32
3.37 -2.49 -0.54
5.12 -5.71 -0.72
4.07 -2.22 -0.43
6.84 - 1 .58 -0.21
7.93 -2.48 -0.27
12.23 -7.00 -0.44
5.42 -4.53 -0.59
1.22 0.08 0.07
1.01 0.07 0.07
0.64 -0.34 -0.42
1 .39 -0. 75 -0.42
1.50 -0.61 -0.34
0.73 -0.27 -0.31
1.40 0.29 0.23
(cent 1 nued)

-------
            Appendix A.  Selected  Reaults  of  Simulations  of  All  Experiments  Modeled Uslns the  RADM Mechanism

                         (continued).
Page 16
to
jr
vO
Exper Iment

Initial
Concent rat Ions
NOx
Ave
Temp
Ave
H20
HC HC/NOx
(ppm) (ppmC)
EC 106
EC1 13
eci 14
eci 15
ECI 16
EC335
EC329
EC330
EC334
EC338
EC328
JL1581R
JL1881R
ST24B1R
AU2781R
OC0382R
OC0382B
NV1582R
JL2281R
JL21B1B
NV15B2B
DE0782B
JN1379B
Group Average
S. Dev.
Avg. Abs. Value
S. Dev.
27. 3-Component
ITC479
ITC584
ITC579
ITC472
ITC474
ITC581
ITC5B5
ITC47B
ITC482
ITC488
ITC492
ITC494
ITC498
ITC500
ITC502
0.50
0.11
1 .00
0.51
0.49
0.44
0.45
0.29
0.45
0.45
0.45
0.27
0.26
0.23
0.23
0.25
0.25
0. 18
0.26
0.24
0.18
0.19
0 .44
0.49
0.37


and "M1n1"
0.09
0.10
0.10
0.10
0.09
0.09
0.10
0.10
0.10
0.09
0.09
0.09
0. 10
0. 10
0.09
9.2
9.5
17.3
12.7
16.6
7.7
4.2
4.3
8.1
15.0
12.1
2.3
2. 1
1 .6
2. 1
2.7
1 .9
2.6
2.4
1 .3
2.7
3.5
6. 1
5.2
4.5


18.3
85.0
17.3
25.2
37.6
17.4
9.2
14.6
18.2
33.7
27.1
6.4
8.0
6.7
9.2
11.0
7.7
14.4
9. 1
5.4
14.9
1&.7
13.7
15.5
19.2


(degK)
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
304.6
303.2
292.2
299.5
297.2
296.9
283.8
302.3*
304. 2*
283.8
286.4
294.9
300.2
5.7


(ppm)
20900.
19500.
28800.
22000.
27300.
24700.
27700.
26000.
27800.
27800.
27500.
20000.*
20000.*
22800.
30900.
21200.
21 100.
20000.*
20000. *
20000.*
20000.*
13400.
20000.*
24160.
5188.


Maximum Concent rat
OZONE
Expt
(ppm)
0.592
0.352
0. 744
0.590
0.743
0.398
0.403
0,344
0.408
0.484
0.523
0.485
0.658
0.211
0.554
0.247
0. 158
0.061
0.626
0. 237
0. 167
0.093
0.756
0.492
0.270


Calc
(ppm)
0.626
0.441
0.721
0.630
0.786
0.426
0.444
0.380
0.456
0.549
0.564
0.694
0.597
0.269
0.498
0. 145
0.039
0.046
0.635
0.437
0.246
0.304
0.80B
0.443
0.226


Calc
-Expt
(ppm)
0.033
0.089
-0.023
0.040
0.044
0.027
0.041
0.036
0.04B
0.065
0.041
0.209
-0.062
O.U58
-0.056
-0.102
-0.119
-0.015
0.009
0.200
0.079
0.212
0.052
-0.049
0. 183
0. 133
0. 134
1on
Calc
-Expt
/Avg
0.05
0 . 23
-0.03
0.06
0.06
0.07
0. 10
0.10
0.11
0. 13
0.08
0.35
-0. 10
0.24
-0.11
-0.52
-1.21
-0.29
0.01
0.59
0.38
1 .07
0.07
-0.12
0.45
0.34
0.32
Average
d( [03] -
Expt
Calc
Initial
[NO] )/dt
Calc
-Expt
(ppb/m1n) --
3.80
5. 26
6.03
3. 78
8. 28
4.64
3.27
3.91
5.59
5. 19
3.50
1 .08
1.31
0.62
0.95
0.76
0.62
0.42
1 . 18
0.63
0.71
0.55
2.44
4.10
4.06


3.95
6.93
5.89
3.41
8 .99
4.43
3.43
3.47
5. 14
4.96
3. 58
1 .21
1 .21
0.65
0.92
0.55

0.34
1.12
0.64
0.76
0.74
2.77
3.32
2.76


0.15
1 .67
-2.14
-0.36
0.70
-0. 22
0.16
-0 .44
-0.45
-0.24
0.08
0.13
-0. 10
0.03
-0.04
-0.21

-0.08
-0.06
0.01
0.05
0. 19
0.33
-0.87
1.81
1 . 09
1 . 68
Calc
-Expt
/Avg
0.04
0.27
-0.31
-0. 10
O.OB
-0.05
0.05
-0.12
-0.08
-0.05
0.02
0.12
-0.08
0.05
-0.04
-0.32

-0.22
-0.05
0.02
0.07
0.30
0.13
-0.13
0 . 26
0 . 22
0.18
Surrogates
4.0
4.0
3.4
3.6
4.2
4.3
4.8
5.0
1 .2
6.0
8. 1
8.5
3.9
4. 1
4.6
43.4
40.7
34.7
35. 1
45. 1
50.0
50.8
52.6
12.3
66.2
88. 4
92.5
40.2
42.7
49.5
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
20300.
20300.
15800.
13700.
15600.
16100.
17100.
18300.
18300.
19600.
15000.
21 100.
19300.
1B700.
18100.
0.332
0.375
0.356
0. 258
0.293
0.352
0.311
0.320
0. 279
0.293
0.326
0.321
0.320
0.286
0. 271
0.304
0.303
0. 270
0.304
0.320
0.307
0.325
0.344
0. 277
0.315
0.340
0.349
0.317
0.305
0.291
-0.028
-0.071
-0.086
0.046
0.027
-0.045
0.014
0.024
-0.002
0.021
0.014
0.027
-0.003
0.019
0.021
-0.09
-0.21
-0.28
0.16
0.09
-0.14
0.04
0.07
-0.01
0.07
0.04
0.08
-0.01
0.06
0 .07
2.95
3.05
1 .54
1 .72
3.40
3.74
5.11
4.73
2.44
3.06
3. 28
3.30
2. 16
5. 18
7.07
3.47
2.80
1 .25
1 .75
4 .84
4.41
6.10
7.30
3.00
4.00
4.30
4.36
2.21
6.84
12.43
0.52
-0. 25
-0.28
0.03
1 .44
0.67
0.99
2.58
0.55
0.94
1 . 02
1 .06
0.05
1 . 65
5.35
0.16
-0.08
-0.20
0.02
0.35
0.17
0.18
0.43
0 . 20
0 . 27
0 . 27
0. 28
0 .02
0.28
0.55
                                                                                                                     (cont1nued)

-------
             Appendix A.
Selected Results
(contInued).
of Simulations of AH Experiment* Modeled Using the RADM Mechanism
                                                                                                                       Page 17
U>
Ul
O
Experiment


ITC466
ITC468
ITC451
ITC455
ITC977
ITC9B5
ITC997
ITC979
ITC992
ITC961
ITC993
JL15B1B
ST24818
JN0982B
JN1463R
JN2783B
AU18B3B
AU2683R
JL1B81B
JN09B2R
Group Average
S. Oev.
Avg . Aba. Value
S. Dev.
Initial
Concent rat Ions
NOx
(ppm)
0. 10
0.09
0. 10
0.09
0. 13
0.12
0.12
0.12
0. 12
0. 14
0.11
0.28
0.23
0.2B
0.22
0.26
0.28
0.32
0.27
0.29
0.14
0.08


Ave
Temp
Ave
H20
HC HC/NOx
( ppmC )
6.5
4.6
S.2
4.4
3.2
2.7
2.7
2.9
2.6
6.5
7. 1
2.3
1.9
3.1
2.6
2.9
0.6
2.6
2.3
3.1
4, 1
1 .9



62.7
49.0
52.0
47.8
25. 1
22.9
23.0
23.1
20.9
45.0
66.9
8.0
8.1
11 .0
11.6
11.1
2.0
8.1
8.5
10.7
37.0
23.7


(degK)
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
304.6
292.2
299.7*
301 .2
305. 1
303.0
302.2
303.2
299.7*
302.6
2.0


(ppm)
18200.
18000.
19BOO.
20200.
20900.
20100.
20100.
20100.
20900.
19000.
20100.
20000.*
14200.
14300.
26100.
24400.
23100.
26400.
20000.*
14700.
19061 .
3028.


Maximum Concentration
OZONE
Expt
(ppm)
0. 131
0.176
0.312
0.304
0.329
0.336
0.316
0.404
0.384
0.32B
0.276
0.474
0.246
0.667
0.585
0.511
0.556
0.646
0.693
0.714
0.366
0. 148


Calc
Calc -Expt
(ppm) (ppm)
0.274 0.143
0.290 0.114
0.295 -0.016
0.292 -0.012
0.312 -0.017
0.282 -0.054
0.281 -0.035
0.401 -0.003
0.393 0.009
0.360 0.033
0.347 0.071
0.657 0.184
0.342 0.095
0.604 -0.063
0.558 -0.027
0.633 0.123
0.578 0.023
0.647 0.001
0.644 -0.049
0.647 -0.067
0.383 0.016
0.134 0.066
0.049
0.047
Calc
-Expt
/Avg
0.71
0.49
-0.05
-0.04
-0.05
-0. 17
-0. 12
-0.01
0.02
0. 10
0. 23
0.32
0.32
-0. 10
-0.05
0.21
0.04
0.00
-0.07
-0. 10
0.07
0.24
0. 15
0. 19
Average Initial
d( [03) - (NO) )/dt
Expt
Calc
Calc -Expt
(ppb/mln) --
0.77
1 . 19
3.95
3.52
3.94
3.95
3.89
4.B8
5.57
2.56
2.45
1 . 10
0. 71
.37
. 10
.01
0.70
. 24
.91
.77
2.81
1 .61


2.57 .80
2.80 .61
3.92 -0.03
3.55 0.03
2.36 - .58
2.17 - .78
2.15 - .74
3.21 - .66
3.59 - .99
2.43 -0.13
2.06 -0.39
1.17 0.07
0.73 0.02
.22 -0.15
.10 -0.01
.16 0.15
0.66 -0.04
.21 -0.03
.81 -0.10
1.60 -0.17
3.13 0.33
2.28 1.37
0.90
1.07
Calc
-Expt
/Avg
1 .08
0.81
-0.01
0.01
-0.50
-0.56
-0.5B
-0.41
-0.43
-0.05
-0. 17
0.06
0.03
-0. 12
0.00
0. 13
-0.05
-0.03
-0.05
-0. 10
0.08
0.38
0.27
0.26
2B, SAPRC 7-Component Surrogates
EC231
EC232
EC233
EC237
EC23B
EC241
EC242
EC243
EC245
EC246
EC247
Group Average
S. Dev.
Avg. Aba. Value
S. Oev.
0.49
0.49
0. 10
0.48
0.95
0.49
0.50
0.50
1.00
0.51
0.51
0.55
0.24


13.2
9.3
9.5
10.5
10. 1
5.0
12.9
9.7
12.9
8.6
6.2
9.8
2.6


26.9
18.9
92.5
21.6
10.6
10.2
25.6
19.5
12.9
17.0
12.2
24.4
23.3


303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
0.0


17400.
17100.
21400.
23200.
25400.
20700.
24600.
16400.
20400.
22100.
22700.
21036.
3020.


0.620
0.342
0.330
0.652
0.691
0.40B
0.6B2
0.716
0.894
0.574
0.657
0.597
0. 173


0.698 0.07B
0.408 0.066
0.381 0.051
0.639 -0.013
0.708 0.017
0.407 -0.001
0.65B -0.074
0.700 -0.016
0.821 -0.073
0.474 -0.100
0.606 -0.052
0.591 -0.006
0.149 0.056
0.044
0.032
0.12
0. 18
0. 14
-0.02
0.02
0.00
-0.04
-0.02
-0.08
-0. 19
-0.08
0.00
0. 1 1
0.08
0.07
6.99
3. 12
4.07
7.36
5. 10
2.87
17.53
14.50
13.41
2.65
7.49
7.74
5. 16


7.72 0.73
3.04 -0.08
4.67 0.60
6.06 -1.30
4.98 -0.12
3.03 0.16
16.87 -0.66
13.10 -1.41
12.91 -0.50
2.09 -0.57
6.49 -1.00
7.36 -0.38
4.85 0.71
0.65
0.45
0. 10
-0.03
0. 14
-0. 19
-0.02
0.06
-0.04
-0. 10
-0.04
-0.24
-0.14
-0.05
0. 12
0. 10
0.07
                                                                                                                    (cont1nued)

-------
            Appendix A.  Selected Results of  Simulations  of  All  Experiments Modeled Using the RAOM Mechanism
                         (continued).
Page 18
UJ
VJl
Experiment
Initial
Concentrat
1 ons
NOx HC HC/NOx
(ppm) (ppmC)
Ave
Temp
(degK)
Ave Maximum Concentration
H20 OZONE
Expt
(ppm) (ppm)
Calc
(ppm)
Calc
-Expt
(ppm)
Calc
-Expt
/Avg
Average
d( (03] -
In1t 1al
[NO] )/dt
Calc
Expt Calc -Expt
(ppb/mln) --
Calc
-Expt
/AvB
29. SAPRC 8-Component Surrogates
ITC626
ITC630
ITC631
ITC633
ITC635
ITC637
1TC665
ITC867
ITC868
ITC871
ITC872
ITCB73
ITC874
ITCB77
ITC880
ITCSBl
:TC885
ITC886
ITC888
ITC891
OTC189A
OTC189B
OTC190A
OTC190B
OTC192A
OTC192B
OTC194A
OTC194B
OTC195A
OTC195B
OTC196B
OTC197A
OTC197B
OTC19BA
OTC19BB
OTC199A
OTC199B
OTC202A
OTC202B
OTC203A
OTC203B
OTC204A
OTC204B
OTC205A
OTC205B
0.30
0.31
0.32
0.64
1 .21
0.31
0.28
0. 28
0.37
0.37
0.38
0.39
0.38
0.38
0.73
0.73
0.64
0.73
0.33
0.32
0.45
0.45
0.41
0.42
0.47
0.47
0.38
0.38
0.41
0.41
0.41
0.38
0.77
0.64
0.40
0.36
0.37
0.73
0.40
0.39
0. 19
0.35
0. 17
0.84
0. 14
4.0
1 .9
1 .0
4.0
4,0
4.0
8.4
4.8
2.9
1 .7
2. 1
1 .3
2. 1
2.3
2.2
2.3
1 .5
2.3
4.7
4.4
3.3
4.0
3.8
3.8
4. 2
4. 2
7.4
3.7
1 .8
4. 1
3.7
3.6
3. 1
5.4
3.7
3.5
3.5
6.0
2.6
3.5
3,5
3.6
3.5
3.8
3.6
13.4
6.3
3.2
6.2
3.3
12.8
29.9
17.2
7.8
4.6
5.7
3.4
5.7
6.2
3.0
3.1
2.4
3.1
14.5
13-. 7
7.4
8.9
9.2
9.0
8.9
8.9
19.3
9.7
4.5
9.9
9. 1
9.5
4. 1
6.4
9.4
9.6
9.6
8.2
6.7
9.0
18.5
10.2
20.3
4.5
26.5
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
295.4
298.0
302.7
299.5
297.8
298. 4
296. 1
295.7
351 .6
360.0
337.5
353.2
358.0
353 .5
347.5
299.7
298 .9
325. 1
324. 1
293.3
293.7
322.2
348.8
464.9
433.4
14300. 0.618
16300. 0.284
17500. 0.043
16600. 0.231
17500. 0.006
16900. 0.617
15100. 0.632
14600. 0.631
17100. 0.518
14600. 0.376
14600. 0.213
14600. 0.160
13200. 0.191
14800. 0.250
13700. 0.031
14000. 0.012
14600. 0.012
20000.* 0.012
18000. 0.579
20000.
5000.
5000.
5400.
5410.
5000.
5000.
5000.
5000.
4660.
3180.
5000.
5000.
5000.
5000.
5000.
5000.
5000.
5000.
4560.
4850.
4880.
4330.
4330.
5000.
5000.
0.602
0.576
0.712
0.702
0.669
0.735
0.754
0.737
0.603
0. 164
0.681
0.597
0.621
0. 152
0.704
0.674
0.591
0.606
0.714
0.522
0.231
0.394
0.302
0.381
0.039
0.392
0.531
0.265
0.069
0.267
0.022
0.529
0.491
0.532
0.409
0.259
0.234
0. 1 19
0. 164
0.243
0.051
0.037
0.029
0.024
0.505
0.494
0.789
0.869
0.867
0.818
0.874
0.798
0.890
0.713
0. 126
0.789
O.B30
0.563
0. 130
0.575
0.640
0.670
0.654
0.509
0.441
0.291
0.436
0.373
0.460
0.075
0.542
-0.087
-0.019
0.027
0.036
0.016
-0.088
-0. 141
-0.099
-0. 109
-0. 1 16
0.022
-0.041
-0.026
-0.008
0,020
0.025
0.018
0.012
-0.074
-0. 108
0.214
0. 157
0. 164
0. 149
0. 139
0.045
0. 152
0.110
-0.038
0. 108
0.233
-0.058
-0.021
-0. 129
-0.034
0.079
0.048
-0.205
-0.080
0.060
0.041
0.071
0.079
0.035
0. 150
-0.15
-0.07

0.14

-0. 15
-0.25
-0. 17
-0.24
-0.37
0.10
-0.29
-0.15
-0.03




-0. 14
-0. 20
0.31
0. 20
0.21
0. 20
0.17
0.06
0. 19
O. 17
-0. 27
0. 15
0.33
-0. 10
-0. 15
-0. 20
-0.05
0.12
0.08
-0. 34
-0.17
0. 23
0.10
0.21
0. 19

0.32
2. 18
0.86
0.47
1 .26
0.95
2.31
2.60
2.23
2. 10
1 . 18
1 .20
0.85
0.83
1 . 14
0.85
1 .03
0.44
1 .24
1 .65
2.67
3.79
3.70
4.24
3.73
4 .25
3.84
7.85
2.95
2.05
4.14
3.86
4.12
2.86
6.09
3.86
3 .86
3.40
6.95
3.82
2.34
2. 20
3.27
2.62
4.85
4.49
1 .57
0.80
0.44
1.12
0.78
1 .57
2.44
2.75
1 .20
0.97
1 .20
0.71
0.79
1 . 19
0.86
0.99
0.68
0.63
1.53
2.07
3.48
3.63
3.27
2.88
3.32
2.77
7.07
2.93
1.21
3.66
3. 26
2.59
2.07
3.54
2.50
2.95
2.84
1 1 .67
3.10
2.21
2.46
3.20
2.71
2. 16
3.45
-0.60
-0.06
-0.03
-0.15
-0. 17
-0.74
-0. 15
0.52
-0.90
-0.21
0.01
-0. 13
-0.04
0.05
0.01
-0.04
0.25
-0.61
-0.12
-0.60
-0.31
-0.07
-0.98
-0.85
-0.93
-i .07
-0.78
-O.O2
-0.84
-0.48
-0.60
-1 .52
-0.80
-2.56
-1 .36
-0.91
-0.56
4.71
-0.72
-0.13
0.26
-0.06
0.09
-2.69
-1 .04
-0.32
-0.07
-0.06
-0. 12
-0.20
-0.3B
-0.06
0.21
-0.55
-0.20
0.01
-0. 17
-0.05
0.04
0.01
-0.04
0.44
-0.65
-0.08
-0. 25
-0.09
-0.02
-0. 26
-0. 26
-0. 25
-0.32
-0.10
-0.01
-0.51
-0.12
-0.17
-0.45
-0.32
-0.53
-0.43
-0.27
-0.18
0.51
-0.21
-0.06
0.11
-0.02
0.03
-0.77
-0.26
                                                                                                                    (cont1nu«d)

-------
             Appendix  A.   Selected Result*  of  Simulation*  of  All  Experiments Modeled Using the RADM Mechanism
                          (continued).
Page 19
rv>
Experiment

Initial
Concentrat Ions
NOx
Ave
Temp
HC HC/NOx
(ppm) (ppmC)
OTC215B
OTC217A
OTC217B
OTC221A
OTC221B
OTC222A
OTC222B
OTC223A
OTC223B
OTC224A
OTC224B
OTC226A
OTC228A
OTC228B
OTC229A
OTC229B
OTC230A
OTC230B
OTC237A
OTC237B
OTC238A
OTC238B
OTC239A
OTC239B
OTC240A
OTC240B
OTC241A
OTC241B
OTC242A
OTC242B
OTC243A
OTC243B
OTC248A
OTC248B
OTC249A
OTC249B
Group Average
S. Dev.
Avg. Abs. Value
S. Dev.
30. UNC Mlsce) 1
ST2081R
DE0782R
0.44
0.50
0.51
0.41
0.42
0.44
0.43
0.36
0.40
0.34
0.34
0.45
0.41
0.41
0.46
0.46
0.41
0.41
0.52
0.52
0.50
0.50
0.49
0.50
0.50
0.50
0.32
0.31
0.45
0.46
0.47
0.38
0.48
0.46
0.48
0.46
0.45
0.16


4. 1
4.9
4.6
2.4
2.0
3.4
2.6
4.8
3.5
4.4
4.3
2.5
2.4
2.3
3.0
1 .7
3.1
1 .7
4.3
4.4
2.9
3.9
2.7
2.5
1 .8
2.0
4.7
4.0
2.6
2.1
4.7
3.7
3.7
2.9
5.6
4.9
3.5
1 .3


9.3
9.6
9.0
5.8
4.7
7.8
6.0
13.4
8.8
12.7
12.6
5.5
6.0
5.6
6.6
3.8
7.6
4.2
8.2
8.5
5.8
7.8
5.4
5.0
3.6
4. 1
14.8
12.9
5.8
4.7
10.0
9.7
7.7
6.3
1 1.8
10.7
8.7
4.9


(degK)
293.4
295.5
294.2
290.6
288.8
298.5
296. 1
314.9
314.7
303.5
303.8
312.5
293.2
294.3
295.7
296.0
297.9
298.2
292.3
292.8
292.5
292.0
290.2
290.9
93.4
104.2
290.7
291 . 1
302.0
302. 1
276.5
277.4
271 .6
268.9
286.9
286.2
302.3
43.6


Ave Maximum Concentrat
H20 OZONE
Expt
(ppm) (ppm)
4420.
7460.
7460.
4460.
4460.
5000.
5000.
4270.
4270.
6410.
6410.
6840.
5000.
5000.
5000.
5000.
5000.
5000.
3810.
3810.
4810.
4810.
5150.
5150.
6210.
6210.
4760.
4760.
4670.
4670.
4720.
4720.
5000.
5000.
3540.
0.866
0.483
0.831
0.235
0.333
0.909
0.940
0.953
0.771
0.776
0.813
0.751
0.246
0. 296
0.253
0. 168
0.489
0.271
O.B07
0.757
0.406
0.702
0.343
0.234
0.034
0.217
0.671
0.674
0. 182
0.639
0.142
0. 152
0.056
0.081
0.348
3540. 0.325
7630. 0.454
4865. 0.272


Calc
(ppm)
0.937
0.633
0.937
0.312
0.372
1.030
1.169
0.893
0.756
0.804
0.847
0.598
0.483
0.507
0.517
0.460
0.873
0.624
0.865
0.739
0.441
0.738
0.449
0.212
0.049
0.153
0.694
0.696
0.21 1
0.694
0. 202
0. 180
0.069
0. 103
0.360
0.392
0.498
0.291


Calc
-Expt
(ppm)
0.069
0. 151
0.105
0.077
0.039
0. 121
0.229
-0.060
-0.015
0.029
0.035
-0. 153
0.237
0.211
0.264
0.292
0.384
0.353
0.058
-0.018
0.035
0.036
0. 106
-0.022
0.015
-0.064
0.023
0.022
0.029
0.055
0.060
0.028
0.013
0.022
0.012
0.067
0.044
0.111
o.oea
0.081
1on
Calc
-Expt
/Avg
0.08
0.27
0.12
0.28
0.11
0.12
0.22
-0.06
-0.02
0.04
0.04
-0.23
0.65
0.53
0.69
0.93
0.56
0.79
0.07
-0.02
0.08
0.05
0.27
-0. 10

-0.35
0.03
0.03
0. 15
0.08
0.35
0.17
0. 21
0. 24
0.03
0.19
0.09
0. 26
0. 20
0.18
Average Initial
d( [03] - [NO] )/dt
Expt
Calc
Calc -Expt
(ppb/m1n) —
5.33
2.55
4.64
1 .63
1 .63
3.83
5.55
4.55
3. 15
3.67
4.94
3.12
1 .49
1 .79
1.56
1 .38
2.43
1 .66
4.10
3.48
2.58
4. 06
2.b6
1.85
1 .08
1 .79
3.26
3.30
1 .47
3.04
1.41
1 .40
1 .40
1 .61
2.55
2.83
2.81
1.51


5.33 0.00
2.51 -0.05
4.10 -0.54
1.48 -0.15
1.57 -0.06
3.80 -0.03
6.22 0.67
3.94 -0.61
2.83 -0.32
2.92 -0.76
4.19 -0.75
2.40 -0.72
1.92 0.43
2.57 0.78
1.98 0.42
1.82 0.44
3.05 0.62
2.16 0.50
3.41 -0.69
2. 98 -0.49
2.21 -0.37
3.58 -0.47
2.53 -0.12
1.51 -0.34
0.94 -0.14
1.42 -0.37
2.68 -0.38
3.23 -0.07
1.38 -0.09
2.72 -0.32
1.73 0.32
2.11 0.71
1.21 -0.19
1.70 0.09
2.31 -0.24
2.99 0.16
2.54 -0.27
1.59 0.82
0.54
0.67
Calc
-Expt
/Avg
0.00
-0.02
-0. 12
-0.09
-0.04
-0.01
0.11
-0. 14
-0.11
-0.23
-0. 16
-0.26
0.25
0.36
0.24
0.28
0.23
0. 26
-0. 18
-0.15
-0. 15
-0.12
-0.05
-0.20
-0. 14
-0.23
-0.12
-0.02
-0.07
-0. 1 1
0.21
0.40
-0. 14
0.05
-0. 10
0.06
-0. 10
0. 23
0. 19
0. 16
aneous Surrogates
0.23
0. 19
2.3
3.4
10.0
18.3
294. 1
286.4
21900. 0.403
13400. 0.076
0.479
0.369
0.076
0. 293
0. 17
1 .31
0.95
0.52
1.11 0.15
0.78 0.26
0. 15
0.40
                                                                                                                     (cont1nu«d)

-------
            Appendix A.
Selected Results
(continued).
                                           of  Simulations  of  All  Experiments Modeled Using the RADM Mechanism
                                                                                                                       Page 20
oo
ui
CO
Experiment
Initial
Concent rat Ions
Ave
Temp
Ave
H20
NOx HC HC/NOx
(ppm) (ppmC)
AU2681B
AU2781B
ST03B1R
ST1081R
ST2081B
JL208IB
ST1682R
JL2081R
JL2281B
OC1481R
ST1682B
ST2981R
ST2981B
OC1481B
ST0381B
ST1081B
JL0882R
JL0882B
Group Average
S. Dev.
Avg. Abs. Value
S. Dev.
31. UNC "Synurban"
AU2284R
AU25B4R
AU2584B
ST0184R
ST0184B
ST0284R
ST02B4B
JN2685R
JN2685B
JN2885R
JN2885B
Group Average
S. Oev.
Avg. Abs. Value
S. Dev.
0.24
0.23
0.24
0.25
0.23
0.42
0.43
0.41
0.26
0.28
0.43
0.24
0.24
0.29
0.23
0.24
0.29
0.28
0.28
0.08


Surrogate
0.32
0.34
0.33
0.31
0.30
0.34
0.33
0.30
0.30
0.38
0.39
0.33
0.03


2.0
2.0
1 .6
2.8
2. 1
1 .8
3.2
2.7
2.9
3.3
3. 1
2.5
2.5
2.9
2.0
1 .0
2. 1
2. 1
2.4
0.6



0.4
0. 1
0. 1
0.2
0.2
0. 1
0. 1
0.2
0.3
0. 1
0.2
0.2
0. 1


8.5
B.8
7.6
11.3
9.2
4.3
7.5
6.6
11.2
1 1 .9
7.2
10.3
10.4
9.9
8.6
4. 1
7.3
7.4
9.0
3.0



1 .2
0.3
0.4
0.6
0.8
0.4
0.3
0.6
0.9
0.4
0.5
0.6
0.3


(degK)
295.7
296.4
301 .0
297.9
292.6
305.0*
304 .4
305.0*
302.3*
291 .3
304.4
293.7
291 .6
288 .9
297.9
295.3
304 .2
304.2
297.7
5.8



302.7
302.0
302.0
302.5
302.5
304.6
304.5
303.6
303.6
299.8
299.8
302.5
1 .6


(ppm)
27500.
25000.
23200.
22300.
15200.
20000. *
23000.
20000.*
20000. •
9470.
23000.
18700.
14400.
10300.
16800.
17700.
16000.
16000.
1901B.
4902.



8470.
22600.
22600.
9380.
9360.
19900.
19900.
16800.
16800.
1 1800.
1 1800.
15403.
5427.


Maximum Concentrat
OZONE
Expt
(ppm)
0.544
0.623
0.541
0.610
0.414
0. 165
0.410
0.635
0.722
0.462
0.840
0.294
0.485
0.458
0.61 1
0.626
0.598
0.541
0.503
0. 175



0.657
0.075
0.096
0.546
0.646
0.119
0.020
0.629
0.788
0.23B
0.275
0.372
0.284


Calc
(ppm)
0.468
0.493
0.474
0.596
0.449
0.581
0.584
0.765
0.665
0.542
0.817
0.460
0.479
0.550
0.513
0.463
0.754
0.743
0.561
0. 122



0.802
0. 145
0. 181
0.725
0.743
0.343
0. 192
0.721
0.791
0.48B
0.353
0.499
0.265


Calc
-Expt
(ppm)
-0.077
-0. 130
-0.067
-0.015
0.035
0.417
0. 174
0. 130
-0.057
0.080
-0.023
0. 166
-0.006
0.091
-0.098
-0. 163
0. 156
0.201
0.058
0. 145
0.118
0.099

0. 145
0.070
0.085
0. 179
0.097
0.224
0. 172
0.093
0.003
0.250
0.078
0. 127
0.074
0. 127
0.074
1 on
Calc
-Expt
/Avg
-0. 15
-0.23
-0.13
-0.02
0.08
1.12
0.35
0. 19
-0.08
0.16
-0.03
0.44
-0.01
0.18
-0. 17
-0.30
0.23
0.31
0. 17
0.40
0.27
0.33

0.20
0.64
0.61
0.28
0. 14
0.97

0. 14
0.00
0.69
0.25
0.39
0.31
0.39
0.31
Average
d( (03) -
Expt
Calc
Initial
[NO] )/dt
Calc
-Expt
(ppb/m1n) --
1 .22
1 .27
1 .48
1 .59
1 .06
0.82
1 .28
1 .55
1 .80
1 .53
2.39
0.76
1.35
1 .48
1 .70
1 .36
1 .25
1.12
1 .32
0.40



1.61
0.70
0.69
1 .32
1 . 72
0.73
0.40
1 .34
1 .74
0.85
0.87
1 .09
0.47


1 .07
1 . 13
1 . 27
1 .60
1 .05
0.97
1 .21
1 .53
1 . 76
1 .64
2. 17
0.91
1 .36
1.58
1 .57
1 .06
1 .41
1 .36
1 .32
0.33



1 .70
0.66
0.68
1 .39
1 .66
0.78
0.51
1 .54
1 .93
0.94
0.92
1 . 16
0.50


-0. 15
-0. 14
-0.21
0.00
-0.01
0.16
-0.07
-0.02
-0.04
0.11
-0. 22
0.15
0.01
0. 10
-0.13
-0.30
0.16
0.24
0.00
0.16
0. 13
0.09

0.09
-0.04
-0.01
0. 07
-0.06
0.05
0. 1 1
0. 20
0. 19
0.09
0.06
0.07
0.08
0.09
0.06
Calc
-Expt
/Avg
-0.13
-0. 12
-0. 15
0.00
-0.01
0. 17
-0.06
-0.01
-0.02
0.07
-0. 10
0. 18
0.01
0.07
-0.08
-0.25
0.12
0.19
0.02
0. 15
0. 1 1
0.10

0.06
-0.05
-0.01
0.05
-0.03
0.07
0.24
0. 14
0.10
0. 10
0.06
0.07
0.08
0.08
0.06
             32.  UNC  "Synauto"  Surrogate
                                                                                                                     (cont\nued)

-------
             Appendix A.   Selected Results of Simulations of All Experiments Modeled Using the RADM Mechanism
                          (continued).
Page 21
LO
Experiment

Initial
Concent rat Ions
NOx
Ave
Temp
Ave
H20
HC HC/NOx
(ppm) (ppmC)
AU0484R
AU04B4B
AU0584R
AU05B4B
AU0684R
AU06B4B
AU0784R
AU0884B
AU0984R
ST0884R
ST06B4B
ST1784R
ST17B4B
ST2184R
ST21B4B
Group Average
S. Oev.
Avg. Abs. Value
S. Dev.
0.37
0.36
0.36
0.35
0.35
0.35
0.38
0.34
0.39
0.34
0.33
0.34
0.34
0.36
0.36
0.35
0.02


0.4
0.3
0.3
0.4
0.7
1 . 1
0.4
1 . 1
0.4
0.6
1 .0
1 . 1
0.8
0.6
0.7
0.7
0.3


1.0
O.B
0.7
1.0
2. 1
3.2
1.0
3.1
0.9
1.9
2.9
3.3
2.3
1 .7
2.0
1.9
0.9


(degK)
307.5
307.5
307. 1
307.2
307.6
307.6
308.0
307.3
307.8
29B.5
298 . 5
294.4
294.4
302.2
302.2
303.9
5. 1


(ppm)
34900.
34700.
44900.
44900.
36800.
36BOO.
34700.
29000.
31500.
6920.
6920.
24BOO.
24800.
7350.
7350.
27089.
13697.


Maximum Concentrat
OZONE
Expt
(ppm)
0.515
0.328
0.335
0.595
0.887
0.940
0.602
0.834
0.521
0.566
0.750
0.539
0.484
0.671
0.721
0.619
0. 182


Calc
(ppm)
0.701
0.622
0.535
0.700
0.891
0.935
0.727
0.846
0.648
0.769
0.877
0.687
0.685
0.833
0.847
0.754
0. 1 14


Calc
-Expt
(ppm)
0. 186
0.294
0.200
0. 105
0.004
-0.005
0. 125
0.012
0. 127
0.203
0.127
0. 146
0.201
0. 162
0. 126
0. 134
0.082
0. 135
0.081
1on
Calc
-Expt
/Avg
0.31
0.62
0.46
0. 16
0.00
0.00
0. 19
0.01
0.22
0.30
0. 16
0.24
0.34
0.22
0. 16
0.23
0. 17
0.23
0. 17
Average
d( 103] -
Expt
Calc
Initial
INO] )/dt
Calc
-Expt
(ppb/m1n) --
1.45
1 . 16
1 .08
1.48
2.53
3. 16
1 .61
2.92
1 .51
1 .27
1 .90
1 .59
1 .38
1.78
2.28
1 .81
0.63


1.46
1.17
1 .03
1 .44
2.17
2.58
1 .55
2.44
1 .33
1 .44
1 .73
2.35
2.03
2.05
2.64
1 .83
0.53


0.01
0.01
-0.06
-0.04
-0.36
-0.58
-0.06
-0.48
-0. IB
0.17
-0. 18
0.76
0.65
0.27
0.35
0.02
0.38
0.28
0.25
Calc
-Expt
/Avg
0.01
0.01
-0.05
-0.03
-0. 15
-0. 20
-0.04
-0. 18
-0. 13
0.12
-0. 10
0.39
0.38
0. 14
0. 14
0.02
0. 18
0. 14
0. 12
33. UNC Auto Exhaust
OC0483B
OC0783R
OC0783B
JN2582R
JN2582B
JN2982R
JN2982B
JN3082R
JN3082B
JL0283B
JL0883B
ST2982B
OC0682R
AU1 183R
AU1 183B
JL0182R
JL01B2B
AU03B2R
AU0382B
ST1782R
ST1782B
ST29B2R
OC0682B
0.25
0.33
0.34
0.65
0.65
0.24
0.25
0.32
0.32
0. 19
0.37
0.39
0.46
0.22
0.23
0.37
0.35
0.44
0. 16
0.26
0.25
0.39
0.46
0.4
2.7
2.7
0.6
0.6
2.5
2.5
2.8
0.6
1 .7
1 .7
1 .7
2.0
2.2
0.7
3.5
3.6
2.5
1 .2
2.2
2.4
1 .7
2.0
1 .7
8.1
7.9
0.9
0.9
10.3
9.8
8.7
1 .7
9. 1
4.6
4.3
4.3
9.9
3.1
9.6
10.2
5.7
7.3
8.6
9.5
4.4
4.3
299.7
295.5
295.4
301.1*
301 . 2*
303. 1
303. 1
302.5*
302.3*
304.0
299.5
295.7
300.2
306.8
306.9
299.5
299.3
302.6
302.6
300. 1
SOU. 1
295.7
300.2
18800.
21600.
21500.
22700.
22900.
30500.
30500.
20000.*
20000.*
24500.
21200.
24800.
16900.
21400.
21400.
27700.
27400.
27400,
27400.
27400.
27400.
24800.
16900.
0.642
0. 178
0.451
0.003
0.003
0.704
0.766
0.81 1
0.840
0.697
0.879
0.205
0.355
0.850
0.601
0.740
0.759
0.344
0.576
0.562
0.637
0.073
0.450
0.706
0.469
0.637
0.016
0.018
0.814
0.817
0.965
0.973
0.850
0.898
0.079
0.348
0.887
0.696
0.998
0.946
0.527
0.513
0.598
0.621
0.063
0.406
0.064
0.291
0.186
0.014
0.014
0. 109
0.052
0. 154
0. 133
0. 153
0.019
-0. 126
-0.007
0.036
0.096
0.258
0.187
0. 183
-0.063
0.036
-0.017
-0.010
-0,043
0. 10
0.90
0.34


0.14
0.07
0.17
0. 15
0.20
0.02
-0.89
-0.02
0.04
0. 15
0.30
0.22
0.42
-0.11
0.06
-0.03
-0.14
-0.10
1 .86
0.94
1.41
0.64
0.68
2.32
2.67
2.55
2.68
1 .65
1 .76
1.12
1 .37
2.55
1 . 26
2.31
2.41
1 .26
1 .06
1 .63
1.72
0.86
1 .49
2.02
1 . 18
2.00
0.59
0.59
2. 14
2.21
2.46
2.46
1 .75
1 .52
0.82
1.23
2. 19
1 . 29
2.47
2.59
1 .27
0.95
1 . 19
1.30
0.70
1 .34
0.17
0.24
0.59
-0.05
-0.09
-0. 19
-0.46
-0.09
-0.22
0.09
-0.24
-0.30
-0.14
-0.35
0.03
0.16
0.18
0.01
-0.12
-0.43
-0.41
-0. 16
-0. 15
0.09
0. 23
0.34
-0.08
-0. 14
-0.08
-0. 19
-0.04
-0.08
0.06
-0. 15
-0.31
-0. 11
-0. 15
0.03
0.07
0.07
0.01
-0. 12
-0.31
-0.27
-0.20
-0. 11
                                                                                                                     (cont tnued)

-------
             Appendix A.   Selected Results of Simulations of AM Experiments Modeled  Using  the RADM Mechanism
                          (concluded).
Page 22
uo
ui
VJ1
Experiment

Initial
Concentrat
NOx
(ppm)
JL0683R
JL1583R
JL1583B
OC0483R
Group Average
S. Dev.
Avg. Abs. Value
S. Dev.
34. SAPRC Synthetic
ITC781
ITC784
ITC785
ITC805
ITC795
ITC796
ITC799
ITC801
Group Average
S. Dev.
Avg. Abs. Value
S. Dev.
35. SAPRC Synthetic
ITC963
ITC965
ITC967
ITC968
Group Average
S. Dev.
Avg. Abs. Value
S. Dev.
0.37
0.35
0.35
0.25
0.34
0. 12


Jet
0.51
0.50
0.26
0.52
0.50
0.54
0.51
0.55
0.49
0.09


Jet
0.49
0.46
0.26
0.49
0.42
0.11


HC
(ppmC)
1 .7
2.2
2.3
2.6
2.0
0.8


Fuel
43.0
88.0
45.0
98.0
45.0
97.0
94.0
41 .0
68. 9
27.3


Exhaust
4.4
5.2
4.4
8.7
5.7
2. 1


Ions
HC/NOx
Ave
Temp


(degK)
4.6
6.2
6.6
10.3
6.5
3. 1



83.5
177.6
170.7
189.5
89. B
178.9
184. 1
75. 1
143.7
50.8



9.1
11.3
17.2
17.8
13.8
4.3


299
306
306
299
301
3



303
303
303
303
303
303
303
303
303
0



303
303
303
303
303
0


.5
. 2
.2
. 7
.2
.3



.0
.0
.0
.0
.0
.0
.0
.0
.0
.0



.0
.0
.0
.0
.0
.0


Ave
H20

(ppm)
21200.
25700.
25700.
18800.
23607.
3802.



16600.
16600.
17BOO.
17200.
17900.
17200.
17200.
17000.
17200.
499.



19400.
20000.
18300.
19400.
19275.
709.


Maximum Concentration
OZONE
Expt
(ppm)
0.756
0.866
0.922
0.603
0.567
0. 272



0.751
0.746
0.598
0.791
0.761
0.597
0.840
0.881
0.746
0. 102



O.B22
0.863
0.586
0.852
0.781
0. 131


Calc
(ppm)
0.778
0.800
0.896
0.668
0.632
0.299



O.B47
0.905
0.625
0.934
0.804
0.876
0.953
0.885
0.854
0. 104



0.705
0.714
0.580
0.777
0.694
0.082


Calc
-Expt
(ppm)
0.022
-0.066
-0.026
0.066
0.065
0. 101
0.090
0.078

0.096
0.160
0.027
0. 144
o.oin
0.279
0.113
0.004
0. 108
O.OU8
0.108
0.008

-0.117
-0. 149
-0.006
-0.075
-0.087
0.062
0.087
0.062
Calc
-Expt
/Avg
0.03
-0.08
-0.03
0.10
0.08
0.29
0. 19
0.23

0.12
0.19
0.04
0. 17
0.06
0.3B
0. 13
0.00
0.14
0. 12
0.14
0. 12

-0. 15
-0. 19
-0.01
-0.09
-0. 1 1
0.08
0.11
0.08
Average
d( [03] -
Expt
Calc
Initial
[NO] )/dt
Calc
-Expt
(ppb/m1n) --
1 .52
2. 10
2.20
1 .76
1 .68
0.61



2.77
3.93
2.74
3.27
3.05
4.43
4.59
2.98
3.47
0.74



4.00
5. 10
5.79
1 1 .76
6.66
3.48


1 .43
1 .87
2.08
1.81
1 .60
0.60



3.33
4. 15
3.05
4.02
4.30
5.31
3.75
2.94
3.86
0.78



3.72
4.73
5.59
10.28
6.08
2.90


-0.09
-0. 23
-0. 12
0.05
-0.08
0.23
0. 19
0. 14

0.56
0.22
0.31
0.75
1 .25
0.88
-0.84
-0.05
0.38
0.64
0.61
0.40

-0.29
-0.37
-0.20
-1 .48
-0.58
0.60
0 .58
0.60
Calc
-Expt
/Avg
-0.06
-0. 1 1
-0.06
0.03
-0.06
0. 15
0.13
0.09

0. 18
0.05
0. 11
0.21
0.34
0. 18
-0.20
-0.02
0. 1 1
0.16
0.16
0. 10

-0.07
-0.08
-0.04
-0. 13
-0.08
0.04
0.08
0.04

-------
                          APPENDIX B
SELECTED RESULTS OF SIMULATIONS OF ALL ORGANIC-NOX EXPERIMENTS




              MODELED USING THE RADM-M MECHANISM
                              356

-------
            Appendix B.  Selected Results of Simulations of  Individual  Organlc-NOx-AIr Experiments Using the
                         RADM-M Mechanism.
Page  1
uo
VJ1
Experiment
Initial
Concent rat Ions
NOx HC HC/NOx
(ppm) (ppmC)
1. Ethene-NOx
EC142
EC143
EC 156
EC265
EC2B6
EC2B7
ITC926
ITC936
AU0479R
AU0579R
OC0584R
OC1 184R
OC1284R
OC05B40
Group Average
S. Dev.
Avg. Abs. Value
S. Dev.
2. Propene-NOx
EC121
EC177
EC216
EC217
EC230
EC256
EC257
EC276
EC277
EC278
EC279
EC314
EC315
EC316
EC3I7
ITC693
ITC810
ITC860
ITC925
ITC938
ITC947
ITC960

0.48
0.50
0.50
1 .01
0.94
0.53
0.51
0.50
0.23
0.64
0.36
0.35
0.72
0.37
0.55
0.22



0.51
0.46
0.52
0.48
0.52
0.56
0.56
0.52
0.11
0.49
0.97
0.93
0.94
0.98
0.54
0.49
0.52
0.52
0.54
0.52
0.53
0.50

1
4
4
3
7
B
7
3
0
4
3
2
2
1
4
2



1
1
1
0
1
0
0
1
1
3
3
3
2
3
1
3
2
3
2
2
1
2

.9
. 1
.0
.9
.5
.0
.9
.9
.9
. 1
.2
.9
.7
.8
. 1
.3



.5
.5
.5
.6
.9
.4
.7
.6
.7
. 1
.5
.2
.9
.2
.5
.5
.8
.0
.8
.8
.9
.8

4. 1
B. 1
8.0
3.9
8.0
15. 1
15.6
7.8
3.9
6.4
8.8
8.2
3.7
5.0
7.6
3.8



2.9
3.2
3.0
1 . 2
3.7
0.7
1 .3
3.2
15.7
6.2
3.5
3.5
3. 1
3.3
2.8
7.2
5.4
5.8
5.2
5.3
3.6
5.5
Ave
Temp
(degK)

303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
298.9
292.4
297.3
297.2
295.6
297.3
300. 2
3.6



303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
288.7
312.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303. 0
Ave
H20
(ppm)

26500.
22700.
29000.
20400.
20500.
19700.
21 100.
20300.
32500.
29100.
20000.*
12700.
13600.
20000.*
22007 .
561 1 .



26500.
26900.
18100.
21400.
17200.
20200.
20000.
14500.
15700.
16100.
14200.
24000.
10600.
44900.
24200.
20300.
16800.
17100.
16300.
19000.
20300.
20300.
Maximum Concentration
OZONE
Expt
(ppm)

0.782
1 .087
1 . 105
0.840
1 .081
0.965
0.982
0.940
0.729
1 . 294
0.856
0.858
0.495
0.675
0.906
0. 204



0.506
0.540
0.564
0. 149
0.344
0.002
0.068
0.3B8
0.313
0.625
0.679
0.728
0.344
0.955
0.615
0.779
0.782
0.585
0.779
0.729
0.710
0. 721
Calc
(ppm)

0.475
0.791
0.751
0.892
1 . 172
1 .026
0.920
0.83B
0.554
1 . 142
0.970
1 .039
0. 185
0.646
0.814
0.275



0.460
0.471
0.555
0.230
0.329
0.01 1
0.098
0.447
0.395
0.702
0.706
0.786
0.661
1 .051
0.570
0.718
0.721
0.680
0.728
0.719
0.715
0.712
Calc
-Expt
(ppm)

-0.307
-0.296
-0.354
0.052
0.091
0.061
-0.062
-0. 102
-0. 175
-0. 152
0.114
0. 182
-0.310
-0.029
-0.092
0. 179
0. 163
0.111

-0.046
-0.069
-0.009
0.081
-0.015
O.OOB
0.029
0.059
0.082
0.076
0.027
0.05B
0.316
0.096
-0.046
-0.062
-0.061
0.096
-0.051
-0.010
O.OOb
-0.009
Calc
-Expt
/Avg

-0.49
-0.32
-0.3B
0.06
0.08
0.06
-0.06
-0. 1 1
-0.27
-0. 12
0.12
0. 19
-0.91
-0.04
-0.16
0.30
0.23
0. 24

-0.09
-0.14
-0.02
0.43
-0.05

0.35
0.14
0.23
0.12
0.04
0.08
0.63
0.10
-0.08
-0.08
-0.08
0. 15
-0.07
-0.01
0.01
-0.01
Average
d( 103] -
Initial
[NO] )/dt
Calc
Expt Calc -Expt
(ppb/m1n) --

3.20
8.50
8.89
5.05
1 1 . 76
13.89
6.96
2.72
1 .60
3.17
2. 16
2. 22
1.58
1.48
5.23
4.10



7.46
3. 89
4. 18
0.80
3.06
0.98
3.35
3.24
8. 27
7.72
6.64
7.21
4.33
10.64
4.06
5.07
4.25
3.57
3.72
3.61
3.34
4.23

2.20
4.51
4.50
4.89
10. 13
13.00
4.91
2.34
0.99
2.27
1 .98
2.01
1.17
1 .33
4.02
3.53



4. 10
4.00
5.69
1 .30
4.78
1 .08
4.20
4.41
8.06
9.83
8.08
10.36
6.42
1 1 .93
4.71
5.72
4.57
4. 88
4. 14
4. 24
4.14
4.40

-1 .00
-3.99
-4.39
-0. 15
-1 .63
-0.88
-2.05
-0.3B
-0.61
-0.91
-0. 18
-0.21
-0.41
-0. 15
-1.21
1 .39
1.21
1 . 39

-3.36
0.11
1.51
0.50
1 .72
0. 10
0.84
1 . 16
-0. 21
2.11
1 .43
3. 15
2.09
1 .29
0.65
0.65
0.31
1.31
0.42
0.63
0.80
0. 17
Calc
-Expt
/Avg

-0.37
-0.61
-0.66
-0.03
-0. 15
-0.07
-0.35
-0. 15
-0.47
-0.33
-0.09
-0.10
-0.30
-0.11
-0.27
0.20
0.27
0.20

-0.58
0.03
0.31
0.48
0.44
0.10
0.22
0.30
-0.03
0.24
0. 19
0.36
0.39
0.11
0.15
0. 12
0.07
0.31
0.11
0.16
0.21
0.04
                                                                                                                    (cont1nued)

-------
            Appendix  B.
Selected Results of Simulations of

RAOM-M Mechanism (continued).
Individual  Organ1c-N0x-A1r Experiments Using the
                                                                                                                       Page  2
ui
CD
Experiment

Initial
Concent rat
NOx
Ions
Ave
Temp
Ave
H20
HC HC/NOx
(ppm) (ppmC)
OTC186
OTC191
OTC210
OTC233
OTC236
JA107BR
OC1278B
OC2078R
OC2078B
OC2178R
OC2578B
JN1279R
JN1279B
JN1379R
AU0279R
AU2780B
ST0482B
ST1382B
JL1783R
JL2183R
JL29B3B
JL3183R
ST2383B
OC0484R
OC0484B
OC1184B
OC1284B
Group Average
S. D»v.
Avg. Abs . Value
S. Oev.
3. 1-Butene-NOx
EC122
EC 123
EC124
ITC927
ITC928
ITC930
ITC935
ST2383R
ST2583R
ST2583B
ST2783R
0.55
0.54
0.57
0.46
0.53
0.46
0.48
0.46
0.46
0.50
0.44
0.50
0.49
0.45
0.22
0.46
0.23
0.33
0.27
0.22
0.21
0.21
0.38
0.36
0.36
0.36
0.68
0.49
0. 18



0.50
0.51
0.99
0.31
0.67
0.32
0.66
0.40
0.46
0.42
0.45
3.6
3.7
2.7
0. 1
3.3
3. 1
1 .4
1 .3
3.5
3.9
1 .3
1 .0
1 .5
2.9
.5
.9
,
t
f
t
,

.6
2. 1
1 .0
2.2
2.0
2.1
1 .0



0.9
1 .6
1 .7
3.8
3.8
7.2
7.6
1 .5
1 .6
2.9
1 .6
6.6
6.9
4.8
0.2
6.3
6.9
2.9
2.9
7.7
7.9
2.9
2. 1
3.0
6.5
7.0
4.0
4.9
3.3
3.9
5.0
5.3
5. 1
4.3
5.9
2.9
6.3
2.9
4.6
2.5



1 .7
3.2
1 .7
12.3
5.7
22.2
1 1 .6
3.7
3.6
t'.e
3.6
(degK)
299.8
309.3
304.7
303.6
300.0
265. 1
294.3
290.4
290.4
290.3
286.4
293.5
293.5
294.9
307.2
302.2
300.5
302.6
307.2
306.5
303. 1
305.0
292.5
296.8
296.7
297.2
295.9
299.9
7.5



303.0
303.0
303.0
303.0
303.0
303.0
303.0
292.6
292.3
292.3
295.5
(ppm)
5000.
5000.
3720.
5000.
6510.
20000.
16300.
9590.
9560.
20000.*
12800.
20000.*
20000.*
^0000.*
20000.*
28400.
1 1800.
24100.
29700.
23800.
23300.
19800.
23400.
20000. »
20000.*
12700.
13500.
18134.
7349.



22000.
Z7200.
24500.
19000.
18400.
21 10U.
19700.
23400.
18600.
16900.
19400.
Maximum Concentration
OZONE
Expt
(ppm)
0.822
0.903
0.972
0.633
0.848
0.363
0.461
0.340
0.727
0.670
0.230
0.382
0.673
0.974
0.788
1 .044
0.658
0. 731
0.848
0.804
0.697
0.719
0.405
0.645
0.446
0.674
0.432
0.608
0. 238



0.227
0.506
0. 247
0.646
0.022
0.717
0.872
0.206
0.266
0.594
0.285
Calc
(ppm)
0.939
1 . 155
1 .006
0.934
1 .015
0.604
0.413
0.403
0.948
0.913
0.269
0.306
0.515
0.874
0.632
1 . 183
0.614
0. 607
0.720
0.731
0.739
0.6B7
0.666
0.759
0.474
0.772
0.471
0.651
0.253



0.081
0.280
0.132
0.727
0.045
O.B27
0.972
0.451
0.449
0.783
0.452
Calc
-Expt
(ppm)
0.117
0.252
0.034
0. 301
0. 167
0.240
-0.048
0.063
0.221
0.243
0.039
-0.076
-0. 158
-0.099
-0. 156
0. 139
-0.044
-0.045
-0. 127
-0.072
0.042
-0.032
0.261
0. 1 14
0.029
0.098
0.039
0.043
0.117
0.093
0.081

-0. 147
-0.226
-0.115
0.08 1
0.023
0.110
0. 100
0.244
0. 182
0. 190
0. 167
Calc
-Expt
/Avg
0. 13
0.24
0.03
0.38
0. 18
0.50
-0.11
0. 17
0.26
0.31
0. 16
-0.22
-0.27
-0.11
-0.22
0. 12
-0.07
-0.06
-0. 16
-0.09
0.06
-0.05
0.49
0. 16
0.06
0. 14
0.09
0.08
0.20
0. 16
0. 14

-0.95
-0.58
-0.61
0.12

0.14
0. 1 1
0. 74
0.51
0.28
0.45
Average
d( [03] -
Expt
Calc
Initial
{NO] )/dt
Calc
-Expt
(ppb/mln) --
5.16
12. 18
6.70
3.69
6.91
0.46
1 .40
1 .23
2.83
2.53
1 .03
0.85
1 .24
2.69
2.41
3.26
1 .46
1 .69
2. 10
1 .76
1 .66
1 .81
1 .23
2.07
1 .25
2.38
1 .67
3.62
2.60



2.29
4. 16
1 .99
3. 24
1.36
7.91
5.39
0.96
1 . 16
1.7B
1 . 19
7.81
13.94
6.36
5.50
8.41
0.48
1 .45
1 .38
3.53
3.44
1 .15
0.77
1 . 10
2.83
2.06
2.84
1 .59
1 .61
1 .75
1 .87
1 . 74
1 .67
1 .67
2. 15
1 .21
2.45
1 .64
4. 15
3.08



1 .40
2.30
1 . 71
4. 54
1 .69
11.13
7. 15
1 . 15
1.31
2. 19
1.31
2.65
1 .77
-0.34
1 .81
1.50
0.02
0.05
0. 15
0.70
0.91
0. 12
-0.09
-0. 14
0.14
-0.35
-0.42
0. 13
-O.OB
-0.35
0. 10
0.08
-0.14
0.44
0.08
-0.04
0.06
-0.03
0.53
1 .01
0.76
0.85

-0.89
-1 .86
-0.27
1.31
0.34
3.22
1 .76
0.20
0. 15
0.41
0.12
Calc
-Expt
/Avg
0.41
0. 14
-0.05
0.39
0.20
0.04
0.03
0.12
0.22
0.31
0. II
-0.11
-0.12
0.05
-0. 16
-0. 14
0.08
-0.05
-0. 18
0.06
0.05
-0.08
0.31
0.04
-0.04
0.03
-0.02
0. 11
0. 19
0. 17
0.14

-0.48
-0.58
-0. 15
0. 34
0. 22
0. 34
0. 26
0. 19
0.12
0.21
0. 10
                                                                                                                    (contInued)

-------
            Appendix B.  Selected Results of  Simulations  of  Individual
                         RADM-M Mechanism (continued).
Orgun1c-N0x-A1r Experiments Using the
                                               Page  3
to
VJI
Experiment
Initial
Concent rat 1 ons
NOx HC HC/NOx
(ppm) (ppmC)
Group Average
S. Dev.
Avg. Abs . Value
S. Dev.
0.52
0.20
3. 1
2.3
6.9
6.2
Ave
Temp
(degK)
299.4
5.0
Ave
H20
(ppm)
21 127.
2856.
Maximum Concentration
OZONE
Expt
(ppm)
0.417
0.263
Calc
(ppm)
0.472
0.321
Calc
-Expt
(ppm)
0.055
0. 154
0. 144
0.066
Calc
-Expt
/Avg
0.02
0.55
0.45
0.29
Average
d( [03] -
Initial
[NO] )/dt
Calc
Expt Calc -Expt
— (ppb/mln) —
2.86
2. 18
3.26
3.18
0.41
1.34
0.96
0.99
Calc
-Expt
/Avg
0.05
0.32
0.27
0.15
4. t rans-2-Butene-NOx
EC146
EC147
EC157
ST2783B
Group Average
S. Dev.
Avg. Abs. Value
S. Dev.
5. Isobutene-NOx
ITC694
6. 1-Hexene-NOx
ITC929
ITC931
ITC934
ITC937
Group Average
S. Dev.
Avg. Abs. Value
S. Dev.
7. Isoprene-NOx
EC520
EC522
EC524
EC525
0.51
0.9B
0.53
0.43
0.61
0.25

0.51

0.51
0.49
1 .00
0.99
0.75
0.28

0.49
0.92
0.96
0.54
0.9
1 . 7
0.9
2.0
1 .4
0.6

4.6

5. 1
10.3
9.7
0. 1
6.3
4.8

2.2
2.3
4.7
4.5
1 .8
1 .7
1 .7
4.7
2.5
1.5

9. 1

10.0
21 . 1
9.7
0. 1
10.2
8.6

4.5
2.4
4.9
8.3
303.0
303.0
303.0
295.5
301 . 1
3.7

303.0

303.0
303.0
303.0
303.0
303.0
0.0

303.0
303.0
303.0
303,0
26900.
22300.
27500.
19400.
24025.
3860.

20000.

19000.
20100.
20300.
20300.
19925.
624.

13000.
13000.
13000.
13000.
0.247
0. 154
0.205
0.523
0.282
0. 165

0.900

0.296
0.606
0.428
0.007
0.335
0.252

0.503
0.276
0.759
0.691
0. 109
0.083
0.091
0.543
0. 207
0.224

0.731

0.670
0.819
0.894
0.060
0.61 1
0.379

0.300
0. 1 IB
0.430
0.630
-0. 138
-0.071
-0. 1 14
0.019
-0.076
0.069
O.OB5
0.052

-0. 169

0.372
0.213
0.466
0.053
0.276
0. 182
0.276
0. 182

-0.203
-0. 158
-0.330
-0.061
-0.77
-0.60
-0.77
0.04
-0.53
0.38
0.54
0.35

-0.21

0.77
0.30
0.71
0.59
0.26
0.59
0. 26

-0.50
-0.80
-0.55
-0.09
5.87
9.83
5.96
3.00
6.17
2.80

8 .84

1 .27
2.82
1 .84
0.33
1 .57
1 .04

5.30
6.00
15.57
10. 13
3.84
7.39
3.54
3. 16
4.48
1.96

6.03

3.32
10.67
5.00
0.57
4.89
4.26

7.57
6.75
15.94
22. 16
-2.02
-2.45
-2.42
0. 15
-1 .68
1 .24
1 .76
1 .09

-2. 81

2.05
7.84
3. 16
0.24
3.32
3.25
3.32
3.25

2.26
0.76
0.37
12.03
-0.42
-0.28
-0.51
0.05
-0.29
0.24
0.32
0.20

-0.38

0.89
1.16
0.92
0.52
0.88
0.26
0.88
0.26

0.35
0.12
0.02
0.75
                                                                                                                    (cont1nued)

-------
             Appendix  B.   Selected Results of Simulations of Individual Organ1c-N0x-A1r Experiments Using the
                          RADM-M Mechanism (continued).
Page  4
U)
CT>
O
Experiment


EC527
ITC811
1TC812
JL16BOR
JL1680B
JL1780R
JL1780B
JL23B1R
ST0981R
Croup Average
S. Oev.
Avg. Abs. Value
S. Dev.
8. a-P1nene-NOx
JL1580R
JL1580B
JL2580R
JL2580B
Group Average
S. Dev.
Avg. Abs. Value
S. Dev.
9. Ethane-NOx
ITC999
10. n-Butane-NOx
EC130
EC133
EC 134
EC137
EC162
EC163
EC 168
EC178
EC304
Initial
Concentrat Ions
NOx
(ppm)
0.50
O.SO
0.52
0. IB
0. IB
0.46
0.47
0.43
0. 17
0.49
0.25



0.1B
0. 19
0.25
0.25
0.22
0.04



0.09

0. 10
0.50
0.51
0.50
0.51
0.49
0.49
0. 10
0.47
HC
( ppmC )
2.2
3.4
1 .8
4.6
6.4
1 .0
2.6
1 .4
1 .0
2.9
1 .7



1 . 1
2.6
1.0
0.3
1 .3
1 .0



45.4

17.6
6.6
8.3
8.7
8.2
9.0
B.O
7.8
17.1
HC/NOx

4.5
6.8
3.4
26.0
36.6
2. 1
5.5
3.4
6.1
8.6
10.4



5.8
13.8
4.0
1 .4
6.3
5.3



534. 1

179.3
17.1
16.3
17.3
16.3
18.3
16. 2
79.6
36.7
Ava
Temp

(degK)
303.0
303.0
303.0
305. 0
301 .9
305.6
302.8
306. 1
297.5
303. 1
2. 1



302.4
299.6
302.4
300. 1
301 . 1
1 .5



303.0

303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
Ave
H20

(ppm)
13000.
17300.
17300.
30200.
26BOO.
27000.
25000.
20000. •
20000.*
19123.
6284.



30300.
29200.
31700.
26300.
29375.
2291 .



20500.

26200.
27800.
20000."
20000.*
27300.
28500.
20700.
20900.
26800.
Maximum Concentration
OZONE
Expt
(ppm)
0.547
0.919
0.768
0.652
0.837
0.806
1 .298
0.750
0.506
0.716
0.246



0.201
0.470
0.377
0.334
0.346
0.112



0.243

0.459
0.249
0.034
0.042
0.112
0.454
0.655
0.384
0.362
Calc
(ppm)
0.317
0.577
0.253
0.712
0.600
0.420
0.835
0.691
0.457
0.488
0.208



0.297
0.421
0.327
0.211
0.314
0.067



0.203

0.442
0.043
0.036
0.043
0.048
0. 151
0.308
0.418
0.422
Calc
-Expt
(ppm)
-0.230
-0.342
-0.515
0.060
-0.237
-0.3B6
-0.463
-0.058
-0.049
-0.229
0. 174
0.238
0.160

0.095
-0.049
-0.050
-0. 123
-0.032
0.092
0.079
0.036

-0.041

-0.017
-0.206
0.002
0.001
-0.064
-0.304
-0. 347
0.034
0.060
Calc
-Expt
/Avg
-0.53
-0.46
-1 .01
0.09
-0.33
-0.63
-0.43
-0.08
-0. 10
-0.42
0.31
0.43
0.29

0.3B
-0. 11
-0. 14
-0.45
-O.OB
0.34
0.27
0. 17

-0. 18

-0.04
-1.41


-0.80
-1 .00
-0.72
0.09
0. 15
Average
dC [03J -
Expt
Calc
In1t 1al
[NOJ )/dt
Calc
-Expt
(ppb/m1n) --
5.97
10.00
4.86
3. 18
2.88
1 .96
4.20
2. 13
1 .46
5.66
4.06



0.61
1.59
0.84
0.84
0.97
0.43



1 .32

4.41
2.42
0.94
1.02
1 .73
3.31
2.03
1 .61
2.09
7.95
8.95
2.42
3.00
2.43
1 .36
2.87
2.01
1 .45
6.53
6.29



0.67
1 .96
0.83
0.6B
1 .03
0.62



1 .42

2.44
1 .05
1 .27
1 . 24
1 . 19
1 .69
1.12
1 .65
2.26
1 .97
- 1 .05
-2.43
-0. IB
-0.44
-0.60
-1 .33
-0.12
-0.01
0.86
3.58
1.81
3. 18

0.06
0.37
-0.02
-0. 16
0.06
0. 22
0. 15
0. 16

0.09

-1.97
-1 .37
0.33
0.23
-0.54
-1.62
-0.91
0.04
0. 17
Calc
-Expt
/Avg
0.28
-0.11
-0.67
-0.06
-0.17
-0.36
-0.38
-0.06
0.00
-0.02
0.36
0.26
0.24

0. 10
0.21
-0.02
-0.21
0.02
0. IB
0. 13
0.09

0.07

-0.58
-0.79
0.30
0.20
-0.37
-0.65
-0.58
0.03
0.08
                                                                                                                     (cont1nued)

-------
           Appendix  B.   Selected  Results  of  Simulations  of  Individual  Organ1c-N0x-A1r  Experiments  Using the
                         RADM-M  Mechanism  (continued).
Page  5
ON
Experiment
Initial
Concentrations
Ave
Temp
Ave
H20
Maximum Concentrat
OZONE
Calc
NOx HC HC/NOx
(ppm) (ppmC)
EC305
EC306
EC307
EC30B
EC309
ITC507
ITC533
ITC770
ITC939
ITC948
OTC21 1
JL2178R
JL2178B
JL2278R
JL2278B
ST1879B
OC0979R
OC1879B
Group Average
S. Dev.

Avg. Abs. Value
S. Dev.
1 1 . C4+ Branched
EC 165
EC 169
EC171
OC1879R
OC2079B
AU1983R
AU1983B
Group Average
S. Dev.

Avg. Abs. Value
S. Dev.
0. 10
0. 19
0. 10
0.48
0.47
0.09
0.12
0.52
0.51
0.26
0.55
0. 24
0.24
0.55
0.55
0.21
0.21
0.20
0.34
0. IB


15.7
25.8
25. B
16.2
17.2
15.2
11.9
37.9
14. B
10.0
42.8
7. 2
15.4
7.9
17.5
21 -2
14.6
14.3
15.7
8.8


159.7
138.2
252.9
33.6
36.3
165.0
99.6
72.8
28.9
38.2
77.5
29.9
63.9
14.4
31 .5
103. 1
71.0
71 .8
69. B
61 .6


(degK)
303.0
302.6
303.0
288. 7
310.9
303.0
303.0
303.0
303.0
303.0
300. 1
302.9
302.9
305. 2
305.2
298.3
297.2
295.0
302. 1
3.8


(ppm)
25400.
24200.
29700.
8750.
1 2400.
20300.
20BOO.
16800.
20300.
20300.
3530.*
20000.*
20000.*
20000. *
20000.*
20000. *
15300.
28100.
20892.
6033.


Expt
(ppm)
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.


398
535
420
047
545
149
165
042
017
054
008
763
986
166
788
185
191
208
312
268


Calc
(ppm)
0.528
0.705
0.585
0.204
0.601
0.350
0.321
0.048
0.025
0. 137
0.181
0.586
0.861
0. 132
0.406
0.111
0.266
0. 153
0.300
0.233


-Expt
(ppm)
0
0
0
0
0
0
0
0
0
0
0
-0
-0
-0
-0
-0
0
-0
-0
0

0
. 130
. 170
. 165
. 157
.056
.201
. 156
.006
.008
.083
. 173
. 177
. 125
.034
.382
.075
.075
.055
.011
. 161
121
. 104
1 on
Calc
-Expt
/Avg
0.28
0.27
0.33
0. 10
0.81
0.64

0.87
-0.26
-0.14
-0.23
-0.64
-0.50
0.33
-0.30
-0.10
0.59
0.47
0.36
Average
d( [03] -


In1 t lal
[NO] )/dt
Calc
Expt Calc -Expt
(ppb/mln) --
2.39
2.38
2.61
1 .04
2.00
0.69
0.61
1 .56
0.36
0.53
0.56
1.17
1 .64
0.90
1 .55
0.51
0.60
0.60
1 .53
0.98


3.38
4.12
5.81
2.54
2.42
1 .52
1 .33
1 .94
0.63
1 .08
1 .37
1 .05
1 .45
0. 75
1 .27
0.43
0.59
0.51
1.71
1.19


0.99
1 . 74
3.20
1.51
0.42
0.83
0.72
0.38
0.27
0.55
0.81
-0.12
-0. 19
-0. 15
-0.28
-0.08
-0.01
-0.09
0. 18
1 .04
0.72
0.76
Calc
-Expt
/Avg
0.34
0.54
0.76
0.84
0.19
0.75
0.75
0.22
0.55
0.68
0.83
-0.11
-0. 12
-0.18
-0.20
-0.17
-0.01
-0.16
0.12
0.49
OA 1
. *1 I
0.28
Alkane-NOx
0. 10
0. 19
0. 10
0.20
0.22
0.38
0.37
0.22
0.11


11.3
4.5
3.5
16.4
12.6
4.7
4. 1
8.2
5.2


114.3
23.5
35.7
81 .9
56.5
12.6
10.9
47.9
38.7


303.0
303.0
303.0
295.0
295.3
307. 1
307. 1
301 .9
5.0


18600.
32900.
19000.
28100.
27BOO.
21300.
21300.
24143.
5463.


0.
0.
0.
0.
0.
0.
0.
0.
0.


488
493
403
236
217
088
057
283
181


0.4B1
0.340
0.31 1
0. 106
0.051
0. 231
0.098
0.231
0. 156


-0
-0
-0
-0
-0
0
0
-0
0

0
.008
.153
.092
. 130
. 165
. 143
.041
.052
. 1 15
1 05
!oeo
-0.02
-0.37
-0.26
-0.76
-1 . 23
0.90
0.53
-0. 17
0.73
0. 58
0.41
1 .77
1 .00
1 .30
0.64
0.66
0.61
0.53
0.93
0.46


1 .46
0. 74
O.B1
0.45
0. 36
0.66
0.56
0.72
0.36


-0.30
-0.26
-0.50
-0.18
-0.30
0.05
0.03
-0.21
0.20
0 . 23
0. 16
-0. 19
-0. 30
-0.47
-0.34
-0.59
0.09
0.06
-0.25
0.25
0 . 29
0.20
12. C5 + n-Alkane-NOx
EC135
OC0979B
EC 131
ITC559
0. 10
0.21
0. 10
0. 19
20.4
15. 1
24.6
279.4
212.7
73.3
251 . 1
1441 . 1
303.0
297.2
303.0
303.0
25000.
15200.
25000.
20300.
0.
0.
0.
0.
435
184
393
377
0.543
0.205
0.524
0.516
0
0
0
0
. 108
.021
.131
. 139
0. 22
0. 11
0.29
0.31
2.92
0.59
1 .92
1 .79
2. 44
0.51
2. 26
1 .58
-0.47
-0.08
0.33
-0.21
-0. 18
-0. 15
0. 16
-0.12
                                                                                                                    (contInued)

-------
            Appendix  B.
Selected Result* of Simulation* of Individual Organlc-NOx-AIr  Experiment*  Us1no the

RADM-M Mechanism (continued).
                                                                                                                       Page   6
oo
cr>
ro
Experiment

Initial
Concentrat 1on»
ND»
HC HC/NOx
(ppm) (ppmC)
ITC53B
ITC540
ITC552
ITC761
ITC762
ITC763
ITC797
ST1879R
EC 155
ITC1001
Group Average
S Oev.
Avg. Ab«. Value
S. Dev.
0.11
0. 1 1
0.13
0.52
0.27
0.28
0.52
0.21
0. 10
0. 11
0.21
0. 14

60.3 529.0
274.8 2421 .3
428.8 3278.4
75.2 145.9
74.7 280.4
7.7 27.7
7.3 14.0
6.3 30.4
37.3 385.1
2.4 22.4
93.9 650.9
133.5 1015.3

Ave
Temp

(degK)
303.0
303.0
303.0
303.0
303.0
303.0
SOo.O
298.3
303.0
303.0
302.3
1 .9

Ave
H20

(ppm)
18900.
20300.
20300.
17500.
17000.
17100.
12200.
20000.*
26200.
17500.
19464.
3919.

Maximum Concentration
OZONE
Expt
(ppm)
0. 150
0.360
0.315
0.030
0. 105
0.041
0.004
0.122
0.264
0.036
0.201
0. 153

Calc
(ppm)
0.457
0.546
0.532
0.079
0.327
0.088
0.012
0.532
0.558
0.084
0.357
0.215

Calc
-Expt
(ppm)
0.306
0. 186
0.217
0.049
0.221
0.047
0.007
0.410
0.294
0.049
0. 156
0. 122
0. 156
0. 122
Calc
-Expt
/Avg
1 .01
0.41
0.51
1 .02
1 .25
0.72

0.58
0.39
0.58
0.39
Average
d( [03] -
Expt
Calc
Initial
fNO] )/dt
Calc
-Expt
-- (ppb/m1n) —
0.74
1.85
1. 19
1.08
0.83
0.68
0.64
0.40
1 .33
0. 15
1 . 15
0.75

1.50
1 .62
1.59
1 .85
2.30
1 .78
1 . 10
0.51
2.42
0.49
1.57
0.69

0.76
-0. 24
0.39
0.77
1 .47
1 . 10
0.46
0. 1 1
1 .09
0. 34
0.42
0.57
0.56
0.42
Calc
-Expt
/Avg
0.68
-0.14
0.28
0.53
0.94
0.89
0.53
0.23
0.58
1 .06
0.38
0.43
0.46
0.33
13. Methy 1 cycl ohexane-NOx
ITC765
ITC766
ITC767
ITC800
Group Average
S Dev.
Avg. Aba. Value
S. Oev.
14. Benzene-NOx
ITC560
ITC561
ITC562
ITC698
ITC710
ITC831
Group Average
S. Dev.
Avg. Abs. Value
S. Dev.
0.53
0.26
0.55
0.54
0.47
0. 14


0.12
0. 1 1
0.56
0.50
0.55
1 .01
0.47
0.33

0.0 0.1
0.0 0.1
0.1 0.1
0.0 0.1
0.0 0.1
0.0 0.0


332.3 2874. 4
79.1 694.2
83.8 149.7
83.5 167.4
83.6 151.0
12.2 12.1
112.4 674.8
111.3 1 103.2

303.0
303.0
303.0
303.0
303.0
0.0


303.0
303.0
303.0
303.0
303.0
303.0
303.0
0.0

17900.
17100.
17900.
17500.
17600.
363.


17100.
17600.
18800.
21900.
21000.
18300.
19117.
1920.

0.022
0.121
0.041
0.015
0.050


0.323
0.273
0.412
0.374
0.367
0.021
0.295
0. 142

0.012
0.301
0.064
0.012
0.097


0.335
0.299
0.528
0.497
0.512
0.009
0.363
0. 199

-0.010
0.180
0.022
-0.003
0.047
0.054

0.012
0.026
0.116
0. 123
0. 145
-0.012
0.068
0.067
0.072
0.062
0.85
0.85
0.85

0.04
0.09
0.25
0.28
0.33

0.20
0. 13
0.20
0.13
0.86
0.87
1 .09
0.86
0.92
0.12


7.01
4.75
2.85
2.87
2.70
0. 17
3.39
2.30

0.77
2.03
1.48
0.85
1 .28
0.59


15.99
10.63
10.21
10. 16
9.68
0.31
9.50
5.08

-0.09
1. 16
0.39
0.00
0.36
0.57
0.41
0.53

8.98
5.88
7.36
7.29
6.98
0.14
6.11
3.08
6.11
3.08
-0. 12
0.80
0.30
0.00
0.25
0.41
0.31
0.35

0.78
0.76
1.13
1.12
1.13
0.60
0.92
0.23
0 .92
0.23
                                                                                                                      (contInued)

-------
           Appendix  B.   Selected  Results  of  Simulations  of  Individual  Organ1c-N0x-A1r  Experiments  Using the
                         RADM-M Mechanism  (continued).
Page
U)
Exper 1ment
Initial
Concentrat 1 ons
NOx HC HC/NOx
(ppm) (ppmC)
15. Toluene-NOx
EC264
EC265
EC266
EC269
EC270
EC271
EC272
EC273
EC327
EC336
EC337
EC339
EC340
ITC699
ITC828
JL3080R
AU2780R
AU2782B
OC2782R
AU0183R
Group Average
S. Dev.
Avg . Abs. Value
S. Dev.
16. Xy 1 ene-NOx
EC343
EC344
EC345
EC346
ITC702
ITC827
JL3080B
AU2782R
OC2782B
AU01B3B
Group Average
S. Dev.
Avg. Abs. Value
S. Dev.

0.48
0.48
0.49
0.47
0.46
0,21
0.48
0.11
0.45
0.44
0.45
0.44
0.43
0.51
1 .02
0. 18
0.46
0.43
0.39
0.39
0.44
0. 18


0.28
0.67
0.28
0. 26
0.52
1 .07
0. 18
0.43
0. 39
0.37
0.45
0.26



8. 1
7.5
B.4
4.0
4. 2
8.0
4. 1
4. 1
4.0
7. 2
7.9
5.0
4. 1
10.5
3.0
3.9
2.3
3.0
4.5
4.6
5.4
2.3


4.2
4.0
3.7
3.9
4.0
1 .2
2.2
2.0
2.8
2.7
3. 1
1 .0



17.0
15.6
17.0
8.4
9.0
37.4
8.5
37.2
8.9
16.3
17.7
11.3
9.5
20.8
3.0
21 .3
4.8
7.0
11.7
11,8
14.7
9.2


14.9
5.9
13.3
14.8
7.8
1 . 1
12.4
4.6
7.0
7.2
8.9
4.7


Ave
Temp
(degK)

303
303
303
303
303
303
303
303
303
303
303
303
303
303
303
304
302
301
2B9
306
302
3


303
303
303
303
303
303
304
301
289
306
302
4



.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.5
. 2
.7
.8
. 1
.5
. 1


.0
.0
.0
.0
.0
.0
.5
. 7
.8
.0
.0
.4


Ave
H20
(ppm)

18900.
21900.
19400.
19100.
20000.
20500.
19400.
21200.
22400.
26900.
27000.
26800.
25900.
20300.
17200.
35200.
32000.
24700.
19500.
23700.
23100.
4692.


29500.
28900.
25200.
25900.
20300.
14400.
28000.
24700.
19500.
23700.
24010.
4737.


Maximum Concentrat
OZONE
Expt
(ppm)

0.419
0.393
0.405
0.318
0.369
0.296
0.410
0.215
0.376
0.396
0.325
0.225
0.344
0.485
0.021
0.273
0.736
0.116
0. 123
0.458
0.335
0. 154


0.283
0.589
0.396
0.384
0.627
0.021
0.555
0.491
0.396
0.688
0.443
0. 195


Calc
(ppm)

0.387
0.383
0.391
0.293
0.362
0.316
0. 288
0.236
0.354
0.426
0.375
0.325
0.341
0.430
0.007
0.414
0.666
0.261
0.334
0.537
0.366
0. 156


0.380
0.524
0.390
0.375
0.480
0.012
0.503
0.552
0.438
0.706
0.436
0. 179


Calc
-Expt
(ppm)

-0.032
-0.010
-0.014
-0.025
-0.006
0.021
-0. 122
0.021
-0.022
0.030
0.050
0. 101
-0.004
-0.055
-0,014
0. 140
0. 130
0. 146
0.212
0.079
0.031
0.081
0.062
0.060

0.097
-0.065
-0.006
-0.009
-0. 147
-0.009
-0.052
0.061
0.042
0.01B
-0.007
0.069
0.050
0.045
1on
Cole
-Expt
/Avg

-0.08
-0.03
-0.04
-0.08
-0.02
0.07
-0.35
0.09
-0.06
0.07
0.14
0.37
-0.01
-0.12

0.41
0. 16
0.77
0.93
0. 16
0.13
0.31
0.21
0.26

0. 29
-0. 12
-0.01
-0.02
-0.26
-0.10
0. 12
0. 10
0.03
0.00
0. 16
0. 12
0.10
Average
d( [03] -
Initial
[NO] )/dt
Calc
Expt Calc -Expt
(ppb/m1n) --

4.46
3.56
4.62
2.55
3.72
6.56
3.69
5.90
2.49
6.06
2.55
1 .53
2.50
4.72
0.49
1 .20
1 .76
0.63
0.66
1 .75
3.07
1.87


8. 18
10.72
1 1 .35
7.65
7.79
1.12
1 .97
1.31
1 .43
2.39
5.39
4.14



5.13
4.95
5.30
2.54
3.94
5.59
2.52
3.47
2.63
6.61
3.99
2.88
2.47
4.86
1 . 15
1 .03
1.51
0.65
0.79
1 .70
3. 19
1 .79


10.74
9.84
10.79
10.71
8. 20
1.97
1 .95
1 .63
1 .97
2. 76
6.06
4.29



0.67
1.39
0.68
-0.01
0.21
-0.97
-1.17
-2.43
0.14
0.55
1 .44
1 .35
-0.03
0.14
0.66
-0.16
-0.25
0.02
0.13
-0.06
0.12
0.91
0 .62
0.65

2.56
-0.88
-0.55
3.06
0.42
0.84
-0.02
0.32
0 .54
0.37
0.67
1 .25
0.96
1.01
Calc
-Expt
/Avg

0.14
0.33
0.14
0.00
0.06
-0.16
-0.38
-0.52
0.05
0.09
0.44
0.61
-0.01
0.03
0.80
-0. 15
-0.15
0.02
0.18
-0.03
0.07
0.31
0.21
0. 23

0.27
-0.09
-0.05
0.33
0.05
0.55
-0.01
0.22
0. 32
0.14
0. 17
0.20
0. 20
0.17
                                                                                                                         (cont1nued)

-------
Appendix B.  Selected Results of Simulations of Individual Organ1c-NOx-A1r Experiments Using the
             RAOM-M Mschanlsm (continued).
Page  B
Experiment
Initial
Concantrat
1ons
NOx HC HC/NOx
(ppm) (ppmC)
17. Mesltylene-NOx
EC900
EC901
EC903
ITC703
ITC706
ITC709
ITC742
ITC826
Group Avsrsgs
S. D»v.
Avg. Abs. Value
S. Dev.
18. TetraHn-NOx
ITC739
ITC747
ITC746
ITC750
ITC832
Group Avarage
S. Dev.
Avg. Abs. Value
S. Dav.
19. Naphthalana-NOx
ITC751
ITC755
ITC756
ITC798
ITC802
Group Avaraga
S. Dev.
Avg. Abs. Valua
S. Oev.

0.53
0.51
1.00
0.50
0.49
0.99
0.48
0.90
0.68
0. 24


0.52
0.50
0.22
0.53
1 .00
0.55
0. 28


0.52
0.24
0.26
0.53
0.53
0.42
0. 15


54
2.7
4. 7
5.3
2.7
4.7
4.6
0.8
3.9
1 .6


2.4
93.2
84.0
44.5
39.4
52.7
36.7


7.5
14. 1
27.4
19.4
8.4
15.4
8.3


10.2
5.2
4.7
10.6
5.4
4.7
9.7
0.9
6.4
3.4


4.6
187.4
385.1
84.6
39.5
140.2
153.2


14.3
58.3
106.8
36.8
15.9
46.4
38.2

Ave
Temp
(deuK)

303.0
303.0
303 .0
303.0
303.0
303.0
303.0
303.0
303.0
0.0


303.0
303.0
303.0
303.0
303.0
303.0
0.0


303.0
303.0
303.0
303.0
303.0
303.0
0.0

Ave
H20
(ppm)

20000.*
20000.*
20000.*
21000.
21000.
21000.
18100.
17200.
18788.
1419.


17900.
17100.
17500.
17100.
16800.
17280.
427.


15100.
17500.
17200.
16400.
16800.
16600.
935.

Maximum Concantrat
OZONE
Expt
(ppm)

0.381
0.384
0.502
0.707
0.641
0.779
0.773
0.022
0.524
0.258


0.002
0.508
0.370
0.482
0.073
0.287
0.235


0.113
0. 259
0.282
0.204
0. 124
0. 196
0.077

Calc
(ppm)

0.419
0.323
0.403
0.478
0.475
0.528
0.481
0.008
0.389
0. 166


0.325
0.726
0.478
0.706
0.352
0.517
0. 190


0.534
0.453
0.491
0.634
0.542
0.531
0.068

Calc
-Expt
(ppm)

0.038
-0.061
-0.099
-0.229
-0. 165
-0.251
-0.293
-0.013
-0. 134
0. 1 19
0. 144
0. 106

0.322
0.217
0. 108
0.224
0.279
0.230
0.081
0.230
O.OB1

0.421
0. 193
0. 209
0.430
0.418
0.334
0. 122
0.334
0. 122
1 on
Avarage
d( [03] -
Calc
-Expt
/Avg

0
-0
-0
-0
-0
-0
-0
-0
0
0
0

0
0
0
1
0
0
0
0

1
0
0
1
1
0
0
0
0

.09
. 17
.22
.39
.30
.38
.47
.26
. 19
.29
. 13

.35
.26
.38
.31
.57
.50
.57
.50

.30
.54
.54
.03
.26
.93
.37
.93
.37
In1t
[NO]
1al
)/dt
Calc
Expt Calc -Expt
(ppb/m1n) --

3.85
8.88
14.96
14.59
7.20
1 1 .74
13.14
1 .68
9.50
4.97


0.63
2.73
2.60
2. 10
1 .38
1 .89
0.88


1 .08
1 .57
2.20
1 .66
1 .42
1 .59
0.41


3.57
5.63
9.62
4.45
7.09
8.01
1. IB
5.65
2.86


2.57
54.85
31 . 70
46.35
1 .00
27.30
24.72


1 1 .24
22.32
29.67
29.98
12.52
21.15
9.01


-5.
-9.
-4.
-2.
-4.
-5.
-0.
-4.
2.
4 .
2.

1 .
52.
29.
44.
-0.
25.
23.
25 .
23.

10.
20.
27.
28.
1 1 .
19.
8.
19.
8.

32
33
97
75
65
13
49
66
69
bb
69

95
12
1 1
25
37
41
96
bb
77

15
75
47
32
1U
56
67
bb
67
Calc
-Expt
/Avg

-0.85
-0.91
-0.41
-0.47
-0.49
-0.48
-0.34
-0.57
0.22
0.57
0.22

1 .22
1 .81
1 .70
1 .83
-0.31
1 .25
0.91
1 .37
0.64

1 .65
1 .74
1 .72
1 .79
1 .59
1 .70
0.08
J .70
o.oe
                                                                                                        (cont1nued)

-------
Appendix B.  Selected Results of Simulations of Individual Organlc-NOx-A1r Experiments Using the
             RADM-M Mechanism (continued).
   Page   9
Experiment
Initial
Concent rat 1 ons
NOx HC HC/NOx
(ppm) (ppmC)
20.




Ave
Temp
(degK)
Ave
H20
(ppm)
Maximum Concentrat
OZONE
Expt
(ppm)
Calc
(ppm)
Calc
-Expt
(ppm)
1 on
Calc
-Expt
/Avg
Average
d( [03] -
In1 t lal
[NO] )/dt
Calc
Expt Calc -Expt
(ppb/m1n) --
Calc
-Expt
/Avg
2,3-D1methylnapnthBlene-NOx
ITC775
ITC771
ITCB06
ITC774
Grouo Average

Avg

21 .
































S. Dev.
. Abs. Value
S. Dev.
Simple Mixture
EC 144
EC145
EC160
EC149
EC 150
EC151
EC152
EC153
EC161
OC1278R
OC2578R
AU0180R
A 1)1 48 OR
EC 166
EC172
ST0682R
ST0682B
EC 106
EC1 13
EC1 14
EC1 15
EC 11 6
EC335
EC329
EC330
EC334
EC33B
EC328
JL1581R
JL1881R
ST2481R
AU27B1R
0.29
0.26
0.33
0.56
0. 36
0. 14


Runs
0.51
0.99
0.99
0.99
1 .00
2.06
0.50
0.97
0.51
0.46
0.44
0.56
0.47
0. 10
0. 10
0.46
0.45
0.50
0. 1 1
1.00
0.51
0.49
0.44
0.45
0. 29
0.45
0.45
0.45
0. 27
0.26
0. 23
0.23
1 .7
4.B
5.9
4.0
4. 1
1 .8



4.7
3.4
3.2
2.0
3.5
5.2
3.7
6.6
3.2
1 .4
1 .4
0.5
1 .4
9.2
2.8
2.8
2.9
9. 2
9.5
17.3
12.7
18.6
7. 7
4.2
4.3
8. 1
15.0
12. 1
2.3
2. 1
1 .6
2. 1
5.8
18.0
17.7
7. 1
12.2
6.6



9.3
3.4
3.3
2.0
3.5
2.5
7.3
6. '8
6.4
3.0
3. 1
0.8
3.0
92.0
28.9
6.2
6.5
18.3
85.0
17.3
25.2
37.6
17.4
9.2
14.6
18.2
33.7
27. 1
8.4
8.0
6.7
9.2
303.0
303.0
303.0
303.0
303 .0
0.0



303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
294.3
288.4
304.9
305.2
303.0
303.0
299. 1
299. 1
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303 .0
303.0
303.0
303.0
304. G
303.2
292.2
299.5
16BOO.
20300.
17100.
22700.
19225.
2806.



24400.
24400.
29200.
29300,
29900.
30400.
26400.
25400.
36900.
16100.
12800.
23400.
34700.
27700.
21600.
22200.
22200.
20900.
19500.
28800.
22000.
27300.
24700.
27700.
26000.
27800.
27800.
27500.
20000. *
20000.*
22800.
30900.
0.274
0.293
0.360
0.341
0.317
0.040



1 .065
0.777
0.874
0.286
0.799
0. 147
0.791
1 .050
0.857
0.260
0. 147
0.256
0.863
0.462
0.369
0.378
0.47B
0.592
0.352
0.744
0.590
0.743
0.398
0.403
0.344
0.408
0.484
0.523
0.485
0.658
0.211
0.554
0.275
0.383
0.432
0.514
0.401
0. 100



0.772
0.376
0.326
0.131
0.329
0.086
0.665
0.817
0.599
0.250
0. 125
0. 107
0.474
0.377
0.297
0.487
0.522
0.614
0.441
0.702
0.606
0.782
0.370
0.384
0.348
0.409
0.4B7
0.507
0.691
0.594
0.276
0.498
0.001
0.090
0.072
0. 174
0.084
0.071
0.084
0.071

-0.293
-0.401
-0.548
-0. 155
-0.470
-0.060
-0. 127
-0.233
-0.259
-0.010
-0.022
-0. 149
-0.389
-0.085
-0.072
0. 108
0.044
0.021
0.089
-0.042
0.016
0.039
-0.028
-0.019
0.004
0.001
0.003
-0.016
0.206
-0.065
0.065
-0.056
0.00
0.27
0. 18
0.41
0.21
0.17
0.21
0.17

-0.32
-0.70
-0.91
-0. 74
-0.83
-0.52
-0. 17
-0.25
-0.36
-0.04
-0. 16
-0.82
-0.58
-0.20
-0.22
0.25
0.09
0.04
0. 22
-0.06
0.03
0.05
-0.07
-0.05
0.01
0.00
0.01
-0.03
0.35
-0.10
0.27
-0. 1 1
1 .73
2.67
2.69
2. 84
2.48
0.51



10.53
5.11
5.86
10.83
6.30
8.43
10.41
19.23
9.94
1.14
0.94
0.98
2.13
2.11
1 .00
1.11
1 .35
3.80
5.26
8.03
3.78
8.28
4.64
3.27
3.91
5.59
5. 19
3.50
1 .08
1.31
0.62
0.95
2.47
8.32
8.91
5.82
6.38
2.93



6. 15
3.84
3.48
5.20
4.12
7.04
7.84
11.91
5.43
1 .22
1 .01
0.64
1 .37
1 .37
0.70
1 .27
1 .58
4.05
7.06
6.07
3.40
9.25
4.21
3. 16
3.44
4.99
4.71
3.23
1 .23
1 .23
0.67
0.92
0.74
5.65
6.22
2.98
3.90
2.54
3.90
2.54

-4.39
-1 .27
-2.39
-5.63
-2.17
-1 .38
-2.57
-7.32
-4.51
0.08
0.07
-0.35
-0.76
-0.74
-0.29
0.16
0.23
0.25
1.81
-1 .95
-0.38
0.97
-0.43
-0. 10
-0.47
-0.59
-0.49
-0.27
0. 15
-0.08
0.05
-0.03
0.35
1 .03
1 .07
0.69
0.79
0.34
0. 79
0. 34

-0.53
-0.28
-0.51
-0. 70
-0.42
-0. 18
-0.28
-0.47
-0.59
0.07
0.07
-0.43
-0.44
-0.42
-0. 35
0.13
0.15
0.06
0. 29
-0.28
-0. 10
0. 1 1
-0.10
-0.03
-0.13
-0.11
-0. 10
-0.08
0.13
-0.06
0.08
-0.03
(cont1nued)

-------
            Appendix B.  Selected Results of Simulations of Individual Organ1c-N0*-A1r Experiments Using the
                         RADM-M Mechanism (continued).
Page 10
•j
T\
Experiment

Initial
Concentrations
NOx
Ave
Temp
Ave
H20
HC HC/NOx
(ppm) (ppmC)
OC03B2R
OC03BZB
NV15B2R
JL2281R
JL2181B
NV1582B
DE0782B
JN1379B
Group Average
S. Dev.
Avg. Abs. Value
S. Dev.
22. 3-Component
ITC479
ITC5B4
ITC579
ITC472
ITC474
ITC581
ITC5B5
ITC47B
ITC4B2
ITC4BB
ITC492
ITC494
ITC498
ITC500
ITC502
ITC462
ITC466
ITC468
ITC451
ITC455
ITC977
ITC985
ITC997
JTC979
ITC992
ITC981
ITC993
JL1581B
ST2481B

0.25
0. 25
0. IB
0. 26
0.24
0. IB
0. 19
0.44
0.49
0.37


and "Mini"
0.09
0. 10
0. 10
0. 10
0.09
0.09
0. 10
0. 10
0. 10
0.09
0.09
0.09
0. 10
0. 10
0.09
0. 11
0.10
0.09
0. 10
0.09
0. 13
0. 12
0. 12
0. 12
0. 12
0. 14
0. 1 1
0.28
0.23

2.7
1 .9
2.6
2.4
1 .3
2.7
3.5
6. 1
5.2
4.5


11.0
7.7
14.4
9. 1
5.4
14.9
18.7
13.7
15.5
19.2


(degK)
297.2
296.9
283.8
302.3*
304.2*
283.8
286.4
294.9
300.2
5.7


(ppm)
21200.
21100.
20000.*
20000.*
20000.*
20000.*
13400.
20000.*
24160.
5188.


Maximum Concent rat
OZONE
Expt
(ppm)
0.247
0. 158
0.061
0.826
0.237
0.167
0.093
0.756
0.492
0.270


Calc
(ppm)
0. 137
0.038
0.042
0.638
0.441
0.245
0.303
0.809
0.428
0.220


Cslc
-Expt
(ppm)
-0.110
-0. 120
-0.019
0.012
0.204
0.079
0.210
0.053
-0.065
0. 172
0. 123
0. 136
Ion
Calc
-Expt
/Avg
-0.57
-1 .22
-0.36
0.02
0.60
0.38
1 .06
0.07
-0.15
0.43
0.32
0.32
Average Initial
d( (03) - [NO] )/dt
Expt
Calc Calc
Calc -Expt -Expt
(ppb/m1n) -- /Avg
0.76
0.62
0.42
1. 18
0.63
0.71
0.55
2.44
4. 10
4.06


0.54 -0.22 -0.33

0.33 -0.09 -0.24
1.12 -0.06 -0.05
0.65 0.01 0.02
0.74 0.03 0.05
0.71 0.17 0.27
2.78 0.33 0.13
3.30 -0.89 -0.15
2.76 1.82 0.25
1.11 0.23
1 . 69 0.18
Surrogate*
4.0
4.0
3.4
3.6
4. 2
4.3
4.6
5.0
1 . 2
6.0
B. 1
B.5
3.9
4. 1
4.6
7.5
6.5
4.6
S. 2
4.4
3.2
2.7
2.7
2.9
2.6
6.5
7. 1
2.3
1.9

43.4
40.7
34.7
35.1
45. 1
50.0
50. B
52.6
12.3
66.2
BB.4
92.5
40.2
42.7
49.5
71.3
62.7
49.0
52.0
47.8
25. 1
22.9
23.0
23. 1
20.9
45.0
66.9
8.0
B.I

303.0
303.0
303.0
303.0
303.0
303.0
30J.O
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303. 0
303.0
303.0
303.0
303.0
303.0
303 . 0
303.0
303.0
304.6
292. 2

20300.
20300.
15800.
13700.
15600.
16100.
17100.
18300.
18300.
19600.
15000.
21 100.
19300.
18700.
18100.
18300.
18200.
18000.
19800.
20200.
20900.
20100.
20100.
20100.
20900.
19000.
20100.
20000. *
14200.

0.332
0.375
0.356
0.25B
0.293
0.352
0.311
0.320
0.279
0.293
0.326
0.321
0.320
0.286
0.271
0.099
0.131
0. 176
0.312
0.304
0.329
0.336
0.316
0.404
0.384
0.328
0.276
0.474
0.246

0.317
0.314
0.292
0.319
0.327
0.313
0.332
0.351
0.292
0.327
0.350
0.358
0.330
0.318
0.302
0.267
0.280
0.299
0.294
0.295
0.324
0.294
0.292
0.414
0.404
0.377
0.365
0.619
0.274

-0.015
-0.060
-0.064
0.061
0.034
-0.039
0.021
0.031
0.013
0.034
0.024
0.036
0.009
0.031
0.031
0. 168
0. 149
0. 123
-0.018
-0.008
-0.006
-0.042
-0.024
0.010
0.020
0.049
0.089
0. 145
0.028

-0.05
-0.18
-0.20
0.21
0.11
-0.12
0.07
0.09
0.05
0.11
0.07
0.11
0.03
0. 10
0. 1 1
0.92
0.72
0.52
-0.06
-0.03
-0.02
-0. 13
-0.08
0.02
0.05
0. 14
0.28
0.27
0. 1 1

2.95
3.05
1.54
1 .72
3.40
3.74
5. 11
4.73
2.44
3.06
3.28
3.30
2. 16
5. 18
7.07
0.70
0.77
1 . 19
3.95
3.52
3.94
3.95
3.89
4.88
5.57
2.56
2.45
1 . 10
0.71

4.81 1.B7 0.4B
3.95 0.90 0.26
2.14 0.61 0.33
3.01 1.30 0.55
6.10 2.70 0.57
5.30 1.56 0.35
7.00 1.89 0.31
8.20 3.48 0.54
4.41 1.97 0.57
5.18 2.12 0.51
5.40 2.12 0.49
5. 34 2 .04 0.47
3. 63 1 . 47 0.51
7.42 2.23 0.35
12.32 5.25 0.54
2.49 1 . 79 1.12
2.93 2.16 1.17
3.46 2.27 0.98
5.68 1.73 0.36
5.08 1 .56 0.36
3.51 -0.43 -0.12
3.32 -0.63 -0.17
3.27 -0.61 -0.17
4.11 -0.77 -0.17
4.49 -1 .08 -0.22
3.73 1.17 0.37
3.21 0.76 0.27
1.15 0.05 0.05
0.70 -0.01 -0.01
(cont 1 nued)

-------
Appendix B.  Selected Results of Simulations of Individual Organlc-NOx-A1r Experiments Using the
             RAOM-M Mechanism (continued).
Page 11
Experiment
•
Initial
Concentrat Ions
NOx
Ave
Temp
Ave
H20
HC HC/NOx
(ppm) (ppmC)
JN0982B
JN1483R
JN2783B
AU1883B
AU2683R
JL18B1B
JN09B2R
Group Average
S. Dev.
Avg. Abs. Value
S. Dev.
0.28
0.22
0.26
0.28
0.32
0.27
0.29
0. 14
0.08


3. 1
2.6
2.9
0.6
2.6
2.3
3. 1
4. 1
1 .9


11 .0
11.6
11.1
2.0
a. i
a. 5
10.7
37.0
23.7


(deBK)
299.7*
301 .2
305. 1
303.0
302.2
303.2
299.7*
302.6
2.0


(ppm)
14300.
26100.
24400.
23100.
26400.
20000.*
14700.
19061 .
3028.


Maximum Concentration
OZONE
Expt
(ppm)
0.667
0.585
0.51 1
0.556
0.646
0.693
0.714
0.366
0. 148


Calc
(ppm)
0.578
0.516
0.579
0.530
0.596
0.620
0.634
0.380
0.117


Calc
-Expt
(ppm)
-0.089
-0.069
0.069
-0.025
-0.050
-0.073
-0.080
0.014
0.065
0.051
0.042
Calc
-Expt
Average Initial
d( [03] - [NO] )/dt
Expt
Calc Calc
Calc -Expt -Expt
/Av9 -- (ppb/m1n) — /Avg
-0. 14
-0. 13
0.13
-0.05
-0.08
-0. 1 1
-0.12
0.08
0.24
0.16
0. 19
1 .37
1 . 10
1 .01
0.70
1 .24
1 .91
1 .77
2.81
1 .61


1.19 -0.18 -0. 14
1.06 -0.04 -0.04
1.12 0.11 0.10
0.65 -0.05 -0.07
1.17 -0.07 -0.06
1.76 -0.15 -0.08
1.55 -0.22 -0.13
3.88 1.08 0.28
2.44 1.36 0.36
1.32 0.36
1.13 0.28
23. SAPRC 7-Component Surrogates
EC231
EC232
EC233
EC237
EC238
EC241
EC242
EC243
EC245
EC246
EC247
Group Average
S. Dev.
Avg. Abs. Value
S. Dev.
0.49
0.49
0.10
0.48
0.95
0.49
0.50
0.50
1 .00
0.51
0.51
0.55
0.24


13.2
9.3
9.5
10.5
10. 1
5.0
12.9
9.7
12.9
8.6
6.2
9.8
2.6


26.9
IB. 9
92.5
21 .6
10.6
10. .2
25.6
19.5
12.9
17.0
12.2
24.4
23.3


303.0
303.0
303.0
303.0
303.0
303.0
303 .0
303.0
303.0
303.0
303.0
303.0
0.0


17400.
17100.
21400.
23200.
25400.
20700.
24600.
16400.
20400.
22100.
22700.
21036.
3020.


0.620
0.342
0.330
0.652
0.691
0.408
0.682
0.716
0.894
0.574
0.657
0.597
0. 173


0.715
0.374
0.3B6
0.648
0.701
0.37B
0.686
0.722
0.856
0.434
0.627
0.593
0. 170


0.094
0.032
0.056
-0.004
0.010
-0.030
0.004
0.006
-0.038
-0. 140
-0.030
-0.004
0.060
0.041
0.043
0.14
0.09
0.16
-0.01
0.01
-0.08
0.01
0.01
-0.04
-0.28
-0.05
0.00
0. 12
0.08
0.08
6.99
3. 12
4.07
7.36
5. 10
2.87
17.53
14.50
13.41
2.65
7.49
7.74
5.16


7.67 0.68 0.09
2.91 -0.21 -0.07
4.34 0.27 0.06
5.89 -1 .47 -0. 22
4.71 -0.39 -0.08
2.90 0.03 0.01
16.65 -0.88 -0.05
12. 77 -1 .74 -0.13
12.82 -0.59 -0.05
1.95 -0.70 -0.30
6.49 -1.00 -0.14
7.19 -0.55 -0.08
4.83 0.73 0.12
0.72 0.11
0.53 0.09
24. SAPRC 8-Component Surrogates
ITC626
ITC630
ITC631
ITC633
ITC635
ITC637
ITC865
0.30
0.31
0.32
0.64
1.21
0.31
0.2B
4.0
1 .9
1 .0
4.0
4.0
4.0
8.4
13.4
6.3
3.2
6.2
3.3
12.8
29.9
303.0
303.0
303.0
303.0
303.0
303.0
303.0
14300.
16300.
17500.
16600.
17500.
16900.
15100.
0.618
0.284
0.043
0. 231
0.006
0.617
0.632
0.540
0.270
0.059
0. 279
0.018
0.537
0.513
-0.078
-0.014
0.016
0.046
0.013
-0.080
-0. 120
-0.13
-0.05

0.19

-0.14
-0.21
2. 18
0.86
0 .47
1 .26
0.95
2.31
2.60
2.26 0.09 0.04
1 .00 0.14 0.15
0.52 0.05 0.10
1 . 50 0 . 24 0.17
1.00 0.05 0.05
2.24 -0.07 -0.03
3.43 0.84 0.28
                                                                                                        (cent1nued)

-------
Appendix B.
         Results of Simulations of Individual Organ1c-N0x-A1r Experiment* lifting the
RADM-M MechanUm (continued).
                                                                                                          Page  12
Experiment
Initial
Concentrat lone
NO* HC HC/NOn
(ppm) (ppmC)
ITC867
ITCB6B
ITC871
ITC872
ITC873
ITC874
ITC877
ITC880
ITCB81
ITCB85
ITC8S6
ITCB8B
ITCB91
OTC189A
OTC189B
OTC190A
OTC190B
OTC192A
OTC192B
OTC194A
OTCJ94B
OTC195A
OTCI95B
OTC196B
OTC197A
OTC197B
OTC198A
OTC198B
OTC199A
OTC199B
OTC202A
OTC202B
OTC203A
OTC203B
OTC204A
OTC204B
OTC205A
OTC205B
OTC215A
OTC215B
OTC217A
OTC217B
OTC221A
OTC221B
OTC222A
0.28
0.37
0.37
0.38
0.39
0.38
0.38
0.73
0.73
0.64
0.73
0.33
0.32
0.45
0.45
0.41
0.42
0.47
0.47
0.38
0.38
0.41
0.41
0.41
0.38
0.77
0.84
0.40
0.36
0.37
0.73
0.40
0.39
0.19
0.35
0. 17
0.84
0. 14
0.45
0.44
0.50
0.51
0.41
0.42
0.44
4.8
2.9
1 .7
2. 1
1 .3
2. 1
2.3
2.2
2.3
1 .5
2.3
4.7
4.4
3.3
4.0
3.8
3.8
4.2
4.2
7.4
3.7
1.8
4. 1
3.7
3.6
3.1
5.4
3. 7
3.5
3.5
6.0
2.6
3.5
3.5
3.6
3.5
3.6
3.6
3.5
4. 1
4.9
4.6
2.4
2.0
3.4
17.2
7.8
4.6
5.7
3.4
5.7
6.2
3.0
3. 1
2.4
3. 1
14.5
13.7
7.4
8.9
9.2
9.0
8.9
8.9
19.3
9.7
4.5
9.9
9. 1
9.5
4. 1
6.4
9.4
9.6
9.6
8.2
6.7
9.0
18.5
10.2
20.3
4.5
26.5
7.8
9.3
9.6
9.0
5.8
4.7
7.8
Ave
Temp
(degK)
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
295.4
298.0
302.7
299.5
297. 6
298.4
296. 1
295.7
351 .6
360.0
337.5
353.2
358.0
353.5
347.5
299.7
298.9
325. 1
324. 1
293.3
293.7
322.2
34B.8
464.9
433.4
293. 1
293.4
295.5
294.2
290.6
288. 8
298.5
Ave Maximum Concentration
H20 OZONE
Expt
(ppm) (ppm)
14600. 0.631
17100. 0.518
14600. 0.376
14600. 0.213
14600. 0.160
13200. 0.191
14600. 0.250
13700. 0.031
14000. 0.012
14600. 0.012
20000.* 0.012
18000. 0.579
20000.
5000.
5000.
5400.
5410.
5000.
5000.
5000.
5000.
4660.
3180.
5000.
5000.
5000.
5000.
500O.
5000.
5000.
5000.
4560.
4850.
4880.
4330.
4330.
5000.
5000.
4420.
4420.
7460.
7460.
4460.
4460.
5000.
0.602
0.576
0.712
0.702
0.669
0.735
0.754
0.737
0.603
0. 164
0.6B1
0.597
0.621
0. 152
0.704
0.674
0.591
0.606
0.714
0.522
0. 231
0.394
0.302
0. 381
0.039
0.392
0.830
0.868
0.483
0.831
0.235
0.333
0.909
Calc
(ppm)
0.539
0.423
0.255
0.220
0. 109
0. 152
0.229
0.042
0.030
0.025
0.021
0.505
0.511
0. 757
0.833
0.028
0.782
O.B41
0. '47
0 . B97
0.695
0. 105
0.764
0.7B9
O.B49
0.093
0.568
0.627
0.645
0.630
0.475
0.413
0.279
0.447
0.374
0.465
0.054
0.546
0.850
0.931
0.570
0.922
0.279
0.349
0.945
Calc
-Expt
(ppm)
-0.092
-0.095
-0.121
0.007
-0.051
-0.038
-0.022
0.011
0.018
0.013
0.009
-0.074
-0.091
0.181
0. 121
0. 126
0. 1 13
0. 106
-0.007
0. 160
0.091
-0.059
0.083
0. 192
-0.072
-0.059
-0. 136
-0.047
0.054
0.024
-0.239
-0. 109
0.048
0.053
0.071
0.084
0.015
0. 154
0.020
0.063
0.087
0.091
0.044
0.016
0.036
Calc
-Expt
/Avg
-0. 16
-0.20
-0.38
0.03
-0.38
-0. 22
-0.09




-0. '4
-0 . 16
0 27
0. 16
0.16
0 16
0 13
-0.01
0 20
0. 14
-0. 44
0.11
0. 28
-0. 12
-0.48
-0. 21
-0. 07
0.09
0. 04
-0. 40
-0.23
0. 19
0. 12
0.21
0. 20

0.33
0.02
0.07
0.17
0. 10
0.17
0.05
0.04
Average
d( (03] -
Initial
INO] )/dt
Calc
Expt Calc -Expt
(ppb/m1n) --
2.23
2. 10
1. 18
1 . 20
0.85
0.63
1 . 14
0.85
1 .03
0.44
1 .24
1 .65
2.67
3.79
3.70
4.24
3.73
4.25
3.84
7.85
2.95
2.05
4.14
3.66
4. 12
2.86
6.09
3.66
3.B6
3.40
6.95
3.62
2.34
2.20
3.27
2.62
4.85
4.49
4.33
5.33
2.55
4 .64
1 .63
1.63
3.B3
3.37
1 .41
1 .09
1 .33
0.78
0.85
1 .31
0.96
1 .08
0.76
0.69
1 .84
2.84
3.55
3.67
3.30
2.90
3.35
2.75
7.55
3.00
1 . 16
3.73
3.30
2.63
2.02
3.63
2.55
3.00
2.88
12.47
3.10
2.29
2.61
3.36
2.86
2.09
3.64
3.61
5.48
2.47
4. 24
1.48
1.59
3.67
1. 15
-0.6B
-0. 10
0. 13
-0.07
0.02
0. 17
0. 1 1
0.04
0.33
-0.55
0. 19
0. 17
-0.24
-0.03
-0.95
-0.83
-0.91
-1 .09
-0.31
0.05
-0.89
-0.40
-0.56
-1.48
-0.84
-2.46
-1.31
-0.86
-0.52
5.52
-0.72
-0.05
0.41
0.09
0.24
-2.76
-0.85
-0.72
0. 15
-0.08
-0.40
-0. 14
-0.04
-0. J6
Calc
-Expt
/Avg
0.41
-0.39
-0.09
0. 10
-0.09
0.03
0.14
0.12
0.04
0.54
-0.57
0.11
0.06
-0.07
-0.01
-0.25
-0.25
-0.24
-0.33
-0.04
0.02
-0.55
-0. 10
-0.16
-0.44
-0.34
-0.51
-0.41
-0.25
-0.17
0.57
-0. 21
-0.02
0.17
0.03
0.09
-0.80
-0.2J
-0. IB
0.03
-0.03
-0.09
-0.09
-0.03
-0.04
                                                                                                        (contInued)

-------
Appendix B.  Selected Results of Simulations of  Individual  Organ1c-NO»-A1r Experiments Using the
             RAOM-M Mechanism (continued).
Page 13
Experiment

Initial
Concentrat Ions
NOx
Ave
Temp
HC HC/NOx
(ppm) (ppmC)
OTC222B
OTC223A
OTC223B
OTC224A
OTC224B
OTC226A
OTC22BA
OTC228B
OTC229A
OTC229B
OTC230A
OTC230B
OTC237A
OTC237B
OTC23BA
OTC23BB
OTC239A
OTC239B
OTC240A
OTC240B
OTC241A
OTC241B
OTC242A
OTC242B
OTC243A
OTC243B
OTC24BA
OTC248B
OTC249A
OTC249B
Group Average
S. Oev.
0.43
0.36
0.40
0.34
0.34
0.45
0.41
0.41
0.46
0.46
0.41
0.41
0.52
0.52
0.50
0.50
0.49
0.50
0.50
0.50
0.32
0.31
0.45
0.46
0.47
0.38
0.48
0.46
0.48
0.46
0.45
0. 16
2.6
4.8
3.5
4.4
4.3
2.5
2.4
2.3
3.0
1 .7
3. 1
1 .7
4.3
4.4
2.9
3.9
2.7
2.5
1 .8
2.0
4.7
4.0
2.6
2. 1
4.7
3.7
3.7
2.9
5.6
4.9
3.5
1 .3
6.0
13.4
8.6
12.7
12.6
5.5
6.0
5.6
6.6
3.8
7.6
4.2
8.2
8.5
5.8
7.8
5.4
5.0
3.6
4. 1
14.8
12.9
5. .8
4.7
10.0
9.7
7.7
6.3
1 1 .8
10.7
8.7
4.9
(degK)
296. 1
314.9
314.7
303.5
303.8
312.5
293 . 2
294.3
295.7
296.0
297.9
298.2
292.3
292.8
292.5
292.0
290.2
290.9
93.4
104 . 2
290. 7
291 . 1
302.0
302. 1
276.5
277.4
271 .6
268 .9
286.9
286.2
302.3
43.6
Ave Maximum Concentration
H2O OZONE
Expt
(ppm) (ppm)
5000.
4270.
4270.
6410.
6410.
6840.
5000.
5000.
5000.
5000.
5000.
5000.
3810.
3810.
4810.
4610.
5150.
5150.
6210.
6210.
4760.
4760.
4670.
4670.
4720.
4720.
5000.
5000.
3540.
0.940
0.953
0. 771
0.776
0.813
0. 751
0. 246
0. 296
0.253
0. 168
0.489
0.271
0.807
0.757
0.406
0.702
0.343
0.234
0.034
0.217
0.671
0.674
0. 1H2
0.639
0.142
0. 152
0.056
0.081
0.34B
3540. 0.325
7630. 0.454
4865. 0.272
Calc
(ppm)
1 . 107
0.866
0.714
0.774
0.836
0.530
0.421
0.456
0.446
0.391
0.628
0.550
0.625
0.676
0.393
0.696
0.390
0. 159
0.034
0. 1 10
0.693
0.679
0. 148
0.623
0.201
0.171
0.055
O.OB5
0.330
0.360
0.474
0.284
Avg. Abs. Value
S. Oev.
25. UNC Mlscel
ST20B1R
OE0782R
AU2681R
AU2681B
AU2781B
ST0381R
ST1081R
ST2081B






Calc
-Expt
(ppm)
0. 167
-0.086
-0.057
-0.002
0.023
-0.221
0. 175
0. 160
0. 193
0.223
0.339
0.279
0.018
-0.081
-0.013
-0.006
0.047
-0.075
0.000
-0. 107
0.022
0.005
-0.034
-0.016
0.059
0.019
-0.001
0 . 004
-0.018
0.035
0.020
0. 102
0.076
0.070
Calc
-Expt
/Avg
0. 16
0.10
0.08
0.00
0.03
-0.35
0.52
0.43
0.55
0.80
0.52
0.68
0.02
-0. 1 1
-0.03
-0.01
0.13
-0.38

-0.65
0.03
0.01
-0.20
-0.03
0.34
0. 1 1
-0.03
0.05
-0.05
0.10
0.03
0.26
0. 19
0. 17
Average
d( [03] -
Expt
Calc
Initial
[NO] )/dt
Calc Calc
-Expt -Expt
(ppb/mln) -- /Avg
5.55
4.55
3. 15
3.67
4.94
3.12
1 .49
1 .79
1 .56
1.38
2.43
1 .66
4. 10
3.48
2.58
4.06
2.66
1 .85
1 .08
1 .79
3.26
3.30
1 .47
3.04
1 .41
1 .40
1 .40
1 .61
2.55
2.83
2.81
1 .51


6. 17
4.06
2.83
2.91
4.41
2.32
1 .92
2.56
1 .96
1.83
3.07
2. 16
3.39
2.92
2. 18
3.56
2.49
1 .46
0.90
1 .36
2.94
3.26
1.31
2.64
1.78
2. 14
1 .20
1 .69
2.29
2.97
2.65
1 .65


0.62 0.1)
-0.47 -0.11
-0.33 -0.11
-0.76 -0.23
-0.53 -0.11
-0.80 -0.29
0.43 0.25
0.77 0.36
0.40 0.23
0.45 0.28
0.64 0.23
0.50 0.26
-0.71 -0.19
-0.56 -0.17
-0.40 -0.17
-0.49 -0.13
-0.16 -0.06
-0.39 -0.24
-0. IB -0. 18
-0.43 -0.28
-0.32 -0, 10
-0.05 -O.01
-0.17 -0.12
-0.40 -0. 14
0.37 0.23
0.74 0.42
-0.20 -0.16
O.OB 0.05
-0.26 -0.11
0.14 0.05
-0.16 -0.05
0.89 0.24
0.54 0.19
0.73 0.16
laneous Surrogates
0.23
0. 19
0.24
0.24
0. 23
0. 24
0. 25
0.23
2.3
3.4
2.0
2.0
2.0
1 .8
2.8
2. 1
10.0
18.3
8.4
8.5
8.8
7.6
11.3
9.2
294. 1
286.4
29B.B
295.7
296.4
301 .0
297.9
292.6
21900. 0.403
13400. .0.076
25500. 0.506
27500. 0.544
25000. 0.623
23200. 0.541
22300. 0.610
15200. 0.414
0.463
0.361
0.514
0.442
0.469
0.453
0.598
0.415
0.059
0. 285
0.007
-0. 102
-0. 153
-O.OBB
-0.012
0.000
0.14
1 .30
0.01
-0.21
-0.28
-0. 18
-0.02
0.00
0.95
0.52
. 18
.22
.27
.48
.59
.06
1 .05
0.74
.08
.00
.07
.22
.54
0.98
0.10 0.10
0.22 0.35
-0.11 -0.09
-0.22 -0.20
-0.21 -O.18
-0.26 -0. 19
-0.05 -0.03
-0.07 -0.07
                                                                                                        ( cont 1 niied )

-------
             Appendix  B.
Selected Results of Simulations of Individual Organ1c-N0x-A1r  Experiments  Using the
RADM-M Mechanism (continued).
                                                                                                                        Page  14
•J
«1
Experiment
Initial
Concentrat Ions
Ave
Temp
Ave
H20
NOx HC HC/NOx
(ppm) (ppmC)
JL2081B
ST1682R
JL2081R
JL2281B
OC14S1R
ST16B2B
ST2981R
ST2981B
OC1481B
ST038IB
ST1081B
JLOB82R
JL0882B
Group Average
S. Oev.
Avg. Aba. Value
S. Dev.
26. UNC "Synurban"
AU2284R
AU25B4R
AU2584B
ST0184R
ST0184B
ST0284R
ST0284B
JN2685R
JN2685B
JN2685R
JN2BB5B
Group Average
S. Dev.
Avg. Aba. Value
S. Dev.
0.42
0.43
0.41
0.26
0.2B
0.43
0.24
0. 24
0.29
0.23
0.24
0.29
0.28
0.28
0.08


Surrogate
0.32
0.34
0.33
0.31
0.30
0.34
0.33
0.30
0.30
0.38
0.39
0.33
0.03


1.8
3.2
2.7
2.9
3.3
3. 1
2.5
2.5
2.9
2.0
1 .0
2. 1
2. 1
2.4
0.6



0.4
0. 1
0. 1
0.2
0.2
0. 1
0. 1
0.2
0.3
0. 1
0.2
0.2
0. 1


4.3
7.5
6.6
11.2
11.9
7.2
10.3
10.4
9.9
8.6
4. 1
7.3
7.4
9.0
3.0



1.2
0.3
0.4
0.6
0.8
0.4
0.3
0.6
0.9
0.4
0.5
0.6
0.3


(degK)
305.0*
304.4
305.0*
302.3*
291 .3
304.4
293.7
291 .6
288.9
297.9
295.3
304.2
304.2
297.7
5.8



302.7
302.0
302.0
302.5
302.5
304.6
304.5
303.6
303.6
299.8
299.8
302.5
1 .6


(ppm)
20000.*
23000.
20000.*
20000.*
9470.
23000.
18700.
14400.
10300.
16800.
17700.
16000.
16000.
19018.
4902.



8470.
22600.
22600.
9380.
9380.
19900.
19900.
16800.
16800.
1 18OO.
11800.
15403.
5427.


Maximum Concentrat
OZONE
Expt
(ppm)
0.165
0.410
0.635
0.722
0.462
0.840
0.294
0.485
0.458
0.61 t
0.626
0.598
0.541
0.503
0.175



0.657
0.075
0.096
0.546
0.646
0. 1 19
0.020
0.629
0.788
0. 238
0.275
0.372
0.284


CalC
(ppm)
0.473
0.508
0.666
0.652
0.570
0.784
0.427
0.492
0.564
0.517
0.376
0.668
0.651
0.527
0.110



0.756
0.105
0. 130
0.653
0.687
0.261
0. 140
0.662
0.742
0.387
0.271
0.436
0.266


Calc
-Expt
(ppm)
0.308
0.098
0.031
-0.070
0. 108
-0.055
0. 134
0.008
0. 106
-0.093
-0.250
0.070
0. 1 10
0.024
0. 134
0. 102
0.086

0.099
0.030
0.034
0. 107
0.041
0. 142
0.120
0.034
-0.046
0. 148
-0.004
0.064
0.063
0.073
0.051
1 on
Calc
-Expt
/Avg
0.97
0.21
0.05
-0.10
0.21
-0.07
0.37
0.02
0.21
-0. 17
-0.50
0.11
0. 18
0. 11
0.40
0.25
0.32

0. 14
0.33
0.30
0. IB
0.06
0.74

0.05
-0.06
0.47
-0.01
0.22
0.25
0.24
0.23
Average
d( [03] -
Expt
Calc
Initial
[NO] )/dt
Calc
-Expt
-- (ppb/m1n) --
0.82
.28
.55
.80
.53
2.39
0.76
.35
.48
.70
.36
.25
. 12
1.32
0.40



1 .61
0.70
0.69
1 .32
1 .72
0.73
0.40
1 .34
1 .74
0.85
0.87
1 .09
0.47


0.94
. 17
.43
.70
.56
2.08
0.88
.29
.50
.55
0.95
.33
.29
1 .25
0.32



1 .63
0.64
0.65
1 .30
1 .57
0.75
0.49
1 .45
1 .84
0.8B
0.87
1 . 10
0.47


0.12
-0.12
-0. 12
-0. 10
0.04
-0.32
0. 12
-0.06
0.02
-0.15
-0.41
0.09
0. 17
-0.06
0.17
0. 15
0. 10

0.02
-0.06
-0.04
-0.02
-0. 15
0.02
0.09
0.11
0. 10
0.03
0.00
0.01
0.08
0.06
0.05
Calc
-Expt
/Avg
0. 14
-0.09
-0.08
-0.06
0.02
-0. 14
0. 15
-0.04
0.02
-0.09
-0.35
0.07
0. 14
-0.03
0.15
0. 12
0.09

0.01
-0. 10
-0.06
-0.01
-0.09
0.02
0.21
0.08
0.06
0.04
0.01
0.01
0.09
0.06
0.06
                                                                                                                     (cortt 1 nued)

-------
Appendix B.  Selected Results of Simulations of Individual Organ1c-N0x-A1r Experiments Using the
             RADM-M Mechanism (continued).
Page 15
Experiment
Initial
Concent rat Ions
NOx HC HC/NOx
(ppm) (ppmC)
Ave
Temp
(degK)
Ave
H20
(ppm)
Maximum Concentration
OZONE
Expt
(ppm)
Calc
(ppm)
Calc
-Expt
(ppm)
Calc
-Expt
/Avg
Average
d( [03] -
Initial
(NO] )/dt
Calc
Expt Calc -Expt
— (ppb/m1n) --
Calc
-Expt
/Avg
27. UNC "Synauto" Surrogate
AU0484R
AU0484B
AU0584R
AU0584B
AU0684R
AU06B4B
AU0784R
AU08B4B
AU0984R
ST08B4R
STOB84B
ST17B4R
ST1784B
ST2184R
ST21B4B
Group Average
S. Dev.
Avg. Abs. Value
S. Dev.
2B. UNC Auto Exhaust
OC0483B
OC07B3R
OC07B3B
JN2582R
JN25B2B
JN2982R
JN29B2B
JN3082R
JN3082B
JL0283B
JL0883B
ST2982B
OC0682R
AU1 183R
AU1 183B
JL0182R
JL01B2B
0.37
0.36
0.36
0.35
0.35
0.35
0.38
0.34
0.39
0.34
0.33
0.34
0.34
0.36
0.36
0.35
0.02



0.25
0.33
0.34
0.65
0.65
0.24
0.25
0.32
0.32
0. 19
0.37
0.39
0.46
0.22
0.23
0.37
0.35
0.4
0.3
0.3
0.4
0.7
1 . 1
0.4
1 . 1
0.4
0.6
1 .0
1 . 1
0.8
0.6
0.7
0.7
0.3



0.4
2.7
2.7
0.6
0.6
2.5
2.5
2.8
0.6
1 .7
1 .7
1 .7
2.0
2. 2
0.7
3.5
3.6
1.0
0.8
0.7
1 .0
2. 1
3.2
1 .0
3. 1
0.9
1 .9
2.9
3.3
2.3
1 .7
2.0
1 .9
0.9



1 . 7
B. 1
7.9
0.9
0.9
10.3
9.8
B.7
1 .7
9. 1
4.6
4.3
4.3
9.9
3. 1
9.6
10.2
307.5
307.5
307. 1
307. 2
307. 6
307.6
308.0
307.3
307.8
298.5
298.5
294.4
294.4
302. 2
302. 2
303.9
5. 1



299. 7
295.5
295.4
301 . 1»
301 . 2»
303. 1
303. 1
302.5*
302.3*
304.0
299.5
295. 7
300. 2
306.8
306.9
299.5
299.3
34900.
34700.
44900.
44900.
36800.
36BOO.
34700.
29000.
31500.
6920.
6920.
24800.
24800.
7350.
7350.
27089.
13697.



18800.
21600.
21500.
22700.
22900.
30500.
30500.
20000. *
20000. *
24500.
21200.
24800.
16900.
21400.
21400.
27700.
27400.
0.515
0.328
0.335
0.595
0.887
0.940
0.602
0.834
0.521
0.566
0.750
0.539
0.484
0.671
0.721
0.619
0. 182



U.642
0. 178
0.451
0.003
0.003
0.704
0.766
0.81 1
O.B40
0.697
0.879
0.205
0.355
0.850
0.601
0.740
0.759
0.627
0.548
0.450
0.636
0.823
O.B70
0.651
0.798
0.541
0.740
0.869
0.710
0.696
O.B11
O.B29
0.707
0. 130



0.708
0.365
0.640
0.012
0.013
0.823
O.B29
0.985
0.992
0.839
0.638
0.054
0.267
0.878
0.665
1 .032
0.989
0.112
0.220
0.115
0.042
-0.064
-0.070
0.049
-0.035
0.021
0. 174
0.119
0.171
0.212
0.141
0. 108
0.086
0.094
0.110
0.064

0.066
0.207
0. 190
0.010
0.010
0. 1 19
0.063
0. 174
0. 151
0.141
-0.041
-0. 151
-0.088
0.028
0.065
0.292
0.230
0.20
0.50
0.29
0.07
-0.07
-0.08
0.08
-0.04
0.04
0.27
0.15
0.27
0.36
0. 19
0. 14
0. 16
0.17
0. 18
0.13

0. 10
0.73
0.35


0.16
0.08
0. 19
0. 17
0. 18
-0.05
-1.17
-0. 28
0.03
0.10
0. 33
0. 26
1 .45
1 .16
1 .08
1 .48
2.53
3. 16
1 .61
2.92
1 .51
1 .27
1 .90
1 .59
1 .38
1 .78
2.28
1 .81
0.63



1 .86
0.94
1.41
0.64
0.68
2.32
2.67
2.55
2.68
1 .65
1.76
1 . 12
1 .37
2.55
1 . 26
2.31
2.41
1.37
1.11
0.97
1 .37
2. 10
2.52
1 .46
2.39
1 .25
1 .38
1 .69
2.25
1.95
2.00
2.61
1 .76
0.54



1 .97
1.11
1 .92
0.57
0.57
2. 10
2. IB
2.41
2.41
1 .72
1 .43
0.75
1.17
2.16
1 .25
2.41
2.53
-0.08
-0.06
-0.12
-0.11
-0.43
-0.64
-0. 15
-0.52
-0.26
0. 1 1
-0.21
0.67
0.57
0.22
0.33
-0.05
0.37
0.30
0.21

0.12
0.18
0.51
-0.07
-0. 1 1
-0.22
-0.49
-0.14
-0.27
0.07
-0.33
-0.38
-0. 20
-0.39
-0.01
0.09
0.12
-0.06
-0.05
-0.11
-0.08
-0. 19
-0.22
-0. 10
-0.20
-0. 19
0.08
-0. 12
0.35
0.34
0.1 1
0. 13
-0.02
0. 18
0. 16
0.09

0.06
0.17
0.30
-0.11
-0. 18
-0.10
-0.20
-0.05
-0.11
0.04
-0.21
-0.40
-0. 16
-0.17
-0.01
0.04
0.05
(cont 1nued)

-------
Appendix B.
Selected Results of Simulations of Individual  Organlc-NOx-AIr Experiments Using the
RAOM-M Mechanism (continued).
                                                                                                          Page 16
Experiment


AU03B2R
AU0382B
ST1782R
ST17B2B
ST2982R
OC0682B
JL0283R
JLOBB3R
JL15B3R
JL15B3B
OC0483R
Group Average
S. Dev.
Avg. Abs. Value
S. Dev.
29. SAPRC Synthetic
ITC781
ITC784
ITC785
ITC805
ITC795
ITC796
ITC799
ITC801
Group Average
S. Dev.
Avg. Abs. Value
S. Dev.
30. SAPRC Synthetic
ITC963
ITC965
ITC967
ITC96B
Group Average
S. Dev.
Avg. Abs. Value
S. Dev.
Initial
Concentrat Ions
NOx
(ppm)
0.44
0.16
0.26
0.25
0.39
0.46
0. 18
0.37
0.35
0.35
0.25
0.34
0. 12


Jet
0.51
0.50
0.26
0.52
0.50
0.54
0.51
0.55
0.49
0.09


HC
( ppmC )
2.5
1.2
2.2
2.4
1 .7
2.0
1 .6
1 .7
2.2
2.3
2.6
2.0
0.8


Fuel
43.0
68. 0
45.0
98.0
45.0
97.0
94.0
41.0
68.9
27.3


HC/NOx

5.7
7.3
6.6
9.5
4.4
4.3
9.0
4.6
6.2
6.6
10.3
6.5
3. 1



83.5
177.6
170.7
189.5
89.6
178.9
184. 1
75. 1
143.7
50. B


Ave
Temp

(degK)
302.6
302.6
300. 1
300. 1
295.7
300.2
304.0
299.5
306.2
306.2
299.7
301 .2
3.3



303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
0.0


Ave
H20

(ppm)
27400.
27400,
27400.
27400.
24800.
16900.
24500.
21200.
25700.
25700.
18800.
23607.
3602.



16600.
16600.
17900.
17200.
17900.
17200.
17200.
17000.
17200.
499.


Maximum Concentration
OZONE
Expt
(ppm)
0.344
0.576
0.562
0.637
0.073
0.450
0.595
0.756
0.866
0.922
0.603
0.567
0.272



0.751
0.746
0.598
0.791
0.761
0.597
0.840
0.881
0.746
0. 102


Calc
(ppm)
0.444
0.49B
0.586
0.612
0.044
0.319
0.658
0.683
0.705
O.B36
0.657
0.607
0.310



0.927
0.921
0.611
0.959
0. 799
0.851
0.993
0.974
0.880
0. 126


Calc
-Expt
(ppm)
0.100
-0.078
0.024
-0.025
-0.029
-0. 131
0.063
-0.073
-0. 161
-0.086
0.055
0.040
0. I 19
0. 102
0.072

0. 176
0. 176
0.013
0. 169
0.03B
0.254
0. 152
0.093
0. 134
0.080
0. 134
0.080
Calc
-Expt
/Avg
0.25
-0. 15
0.04
-0.04
-0.49
-0.34
0. 10
-0.10
-0.20
-0. 10
0,09
0.01
0.34
0. 23
0. 25

0. 21
0.21
0.02
0. 19
0.05
0.35
0. 17
0. 10
0.16
0.11
0. 16
0.11
Average
d( [03] -
Expt
Calc
Initial
(NO] )/dt
Calc
-Expt
(ppb/mln) --
1.26
.06
.63
.72
0.86
.49
.37
.52
2. 10
2.20
1 . 76
1 .68
0.61



2.77
3.93
2.74
3.27
3.05
4.43
4.59
2.98
3.47
0.74


1 .22
0.92
1 . 15
1 .27
0.68
1 .26
.37
.31
.73
.98
.74
1 .55
O.S9



9.61
13.15
8.64
12.47
13.25
17.30
11.31
7.82
1 1 .69
3.06


-0.03
-0.14
-0.47
-0.45
-0. 19
-0.23
0.00
-0.21
-0.36
-0.22
-0.02
-0. 14
0.23
0.21
0. 15

6.84
9.22
5.90
9.21
10. 19
12.86
6.72
4.84
8.22
2.63
8.22
2.63
Calc
-Expt
/Avg
-0.03
-0. 14
-0.34
-0.30
-0.24
-0. 17
0.00
-0. 15
-0. 19
-0.10
-0.01
-0. 10
0. 15
0. 14
0.11

. 10
.08
.04
. 17
.25
1 . 18
0.65
0.90
1 .07
0.14
1 .07
0.14
Jet Exhaust
0.49
0.46
0.26
0.49
0.42
0.11


4.4
5.2
4.4
8.7
5.7
2. 1


9. 1
1 1.3
17.2
17. B
13. B
4.3


303.0
303.0
303.0
303.0
303.0
0.0


19400.
20000.
18300.
19400.
19275.
709.


0.822
O.B63
0.586
0.652
0. 781
0. 131


0.722
0. 728
0.591
0. 793
0. 708
O.OB4


-0. 100
-0. 135
0,005
-0.059
-0.072
0.060
0.075
O.OS6
-0.13
-0. 17
0.01
-0.07
-0.09
0.08
0. 10
0.07
4.00
5. 10
5.79
11 .76
6.66
3.48


4. 16
5.31
6.21
11.21
6.72
3.11


0.16
0.21
0.42
-0.55
0.06
0.42
0.33
0. 18
0.04
0.04
0.07
-0.05
0.03
0.05
0.05
0.01

-------
                           APPENDIX C
SELECTED RESULTS  OF  SIMULATIONS OF  ALL ORGANIC-NOX EXPERIMENTS




              MODELED  USING THE RADM-P MECHANISM
                              373

-------
             Appendix C.
Selected Reaulta of Simulation* of Individual  Organlc-NOx-Atr Experiment* Using the
RADM-M Mechanlem.
                                                                                                                        Paga   1
U)
-J
-Cr
Experiment
Initial
Concantrat 1on»
NO* HC HC/NOx
(ppm) (ppmC)
1. Ethene-NOx
EC142
EC 143
EC156
EC285
EC2B6
EC287
ITC926
ITC936
AU0479R
AU0579R
OC0584R
OC11B4R
OC1284R
OC05B4B
Group Average
S. Oev.
Avg. Aba. Value
S. Dav.
2. Propene-NOx
EC121
EC177
EC216
EC217
EC230
EC256
EC257
EC276
EC277
EC278
EC279
ECS 14
EC315
EC316
EC317
ITC693
ITC810
ITC860
ITC925
ITC938
ITC947
ITC960

0.48
0.50
0.50
1.01
0.94
0.53
0.51
0.50
0.23
0.64
0.36
0.35
0.72
0.37
0.55
0.22



0.51
0.46
0.52
0.48
0.52
0.56
0.56
O.S2
0. 1 1
0.49
0.97
0.93
0.94
0.98
0.54
0.49
0.52
0.52
0.54
0.52
0.53
0.50

1 .9
4. 1
4.0
3.9
7.5
8.0
7.9
3.9
0.9
4. 1
3.2
2.9
2.7
1 .8
4. 1
2.3



1 .5
1 .5
1 .5
0.6
1 .9
0.4
0.7
1.6
1 .7
3. 1
3.5
3.2
2.9
3.2
1 .5
3.5
2.6
3.0
2.8
2.8
1 .9
2.8

4. 1
8.1
8.0
3.9
8.0
15. 1
15.6
7.8
3.9
6.4
8.8
8.2
3.7
5.0
7.6
3.8



2.9
3.2
3.0
1 .2
3.7
0.7
1.3
3.2
15.7
6.2
3.5
3.5
3.1
3.3
2.8
7.2
5.4
5.8
5.2
5.3
3.6
5.5
Ave
Tamp
(dagK)

303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
298.9
292.4
297.3
297.2
295.6
297.3
300.2
3.6



303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
2BB.7
312.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
Ave
H20
(ppm)

26500.
22700.
29000.
20400.
20500.
19700.
21 100.
20300.
32500.
29100.
20000.*
12700.
13600.
20000.*
22007.
561 1 .



26500.
26900.
18100.
21400.
17200.
20200.
20000.
14500.
15700.
16100.
14200.
24000.
10600.
44900.
24200.
20300.
16800.
17100.
16300.
19000.
20300.
20300.
Maximum Concantrat
OZONE
Expt
(ppm)

0.782
1.087
1.105
0.840
1 .081
0.965
0.982
0.940
0.729
1 .294
0.856
0.858
0.495
0.675
0.906
0.204



0.506
0.540
0.564
0. 149
0.344
0.002
0.068
0.388
0.313
0.625
0.679
0.728
0.344
0.955
0.615
0.779
0.782
0.585
0.779
0.729
0.710
0.721
Calc
(ppm)

0.474
0.791
0.751
0.892
1.171
1 .026
0.921
0.839
0.554
1 142
0.970
1.039
0. 185
0.646
O.B14
0.275



0.460
0.471
0.555
0.230
0.329
0.011
0.098
0.447
0.395
0.702
0.706
0.786
0.661
1 .051
0.570
0.718
0.721
0.681
0.728
0.719
0.715
0.712
Calc
-Expt
(ppm)

-0.307
-0.296
-0.354
0.052
0.090
0.061
-0.061
-0. 101
-0. 175
-0.152
0.114
0. 181
-0.310
-0.029
-0.092
0. 179
0. 163
0.111

-0.046
-0.069
-0.009
0.081
-0.015
0.008
0.029
0.059
0.082
0.076
0.027
0.058
0.316
0.096
-0.045
-0.061
-0.061
0.096
-0.051
-0.010
0.005
-0.009
1 on
Calc
-Expt
/Avg

-0.49
-0.32
-0.38
0.06
0.08
0.06
-0.06
-0.11
-0.27
-0.12
0.12
0. 19
-0.91
-0.04
-0. 16
0.30
0.23
0.24

-0.09
-0.14
-0.02
0.43
-0.05

0.35
0. 14
0.23
0.12
0.04
0.08
0.63
0. 10
-0.08
-0.08
-0.08
0. 15
-0.07
-0.01
0.01
-0.01
Avaraga
d( [03] -
Initial
[NO] )/dt
Calc
Expt Calc -Expt
(ppb/m1n) --

3.20
8.50
8.89
5.05
1 1 .76
13.89
6.96
2.72
1 .60
3. 17
2. 16
2.22
1 .58
1 .48
5.23
4.10



7.46
3.89
4. 18
0.80
3.06
0.98
3.35
3.24
8.27
7.72
6.64
7.21
4.33
10.64
4.06
5.07
4.25
3.57
3.72
3.61
3.34
4. 23

2.20
4.51
4.50
4.89
10. 14
13.01
4.92
2.35
0.99
2.27
1.98
2.01
1 . 17
1.33
4.02
3.53



4. 10
4.00
5.69
1 .30
4.78
1 .08
4.20
4.41
8.06
9.84
8.08
10.36
6.43
1 1 .93
4.71
5.72
4.57
4.88
4.14
4.24
4. 15
4.40

-1 .00
-4.00
-4.39
-0. 15
-1 .63
-0.88
-2.05
-0.37
-0.61
-0.91
-0.18
-0.21
-0.41
-0. 15
-1 .21
1.39
1.21
1 .39

-3.36
0. 1 1
1 .51
0.50
1 . 72
0. 10
0.84
1 . 16
-0.21
2. 1 1
1 .43
3. 15
2.09
1 . 29
0.65
0.66
0.32
1.31
0.42
0.63
0.81
0. 17
Calc
-Expt
/Avg

-0.37
-0.61
-0.66
-0.03
-0. 15
-0.07
-0.34
-0. 15
-0.47
-0.33
-0.09
-0. 10
-0.30
-0.11
-0.27
0.20
0.27
0.20

-0.58
0.03
0.31
0.48
0.44
0. 10
0.22
0.30
-0.03
0.24
0.19
0.36
0.39
0.11
0. 15
0.12
0.07
0.31
0.11
0. 16
0.22
0.04
                                                                                                                      ( con t1nu*d)

-------
            Appendix C.
Selected Results of Simulations of Individual Organ1c-N0x-A1r  Experiments  Using  the
RADM-M Mechanism (continued).
                                                                                                                       Page  2
VJ1
Experiment

Initial
Concent rat Ions
NOx
Ave
Temp
Ave
H20
HC HC/NOx
(ppm) (ppmC)
OTC186
OTC191
OTC210
OTC233
OTC236
JA107BR
OC1278B
OC207BR
OC207BB
OC217BR
OC2578B
JN1279R
JN1279B
JN1379R
AU0279R
AU27BOB
ST04B2B
ST1382B
JL17B3R
JL2183R
JL2983B
JL3183R
ST23B3B
OC0484R
OC0484B
OC1 184B
OC1284B
Group Average
S. Dev.
Avg. Abs . Value
S. Dev.
3. 1-Butene-NOx
EC122
EC123
EC 124
ITC927
ITC92B
ITC930
ITC935
ST2383R
ST25B3R
ST25B3B
ST2783R
0.55
0.54
0.57
0.46
0.53
0.46
0.48
0.46
0,46
0.50
0.44
0.50
0.49
0.45
0.22
0.4B
0.23
0.33
0.27
0.22
0.21
0.21
0.36
0.36
0.36
0.36
0.68
0.49
0.18



0.50
0.51
0.99
0.31
0.67
0.32
0.66
0.40
0.46
0.42
0.45
3.6
3.7
2.7
0. 1
3.3
3. 1
1 .4
1 .3
3.5
3.9
1 .3
1 .0
1 .5
2.9
1 .5
1 .9
. 1
. 1
. 1
. 1
. 1
. 1
1 .6
2. 1
1 .0
2. 2
2.0
2. 1
1 .0



0.9
1 .6
1 .7
3.8
3.8
7.2
7.6
1 .5
1 .6
2.9
1 .6
6.6
6.9
4.8
0.2
6.3
6.9
2.9
2.9
7.7
7.9
2.9
2.1
3.0
6.5
7.0
4.0
4.9
3.3
3.9
5.0
5.3
5. 1
4'. 3
5.9
2.9
6.3
2.9
4.6
2.5



1 .7
3.2
1 . 7
12.3
5.7
22.2
11.6
3.7
3.6
6.8
3.6
(degK)
299.8
309.3
304.7
303.6
300.0
265. 1
294.3
290.4
290.4
290.3
288.4
293.5
293.5
294.9
307.2
302.2
300.5
302.6
307.2
308.5
303. 1
305.0
292.5
296.8
296.7
297.2
295.9
299.9
7.5



303.0
303.0
303.0
303.0
303.0
303.0
303.0
292.6
292. 3
292.3
295.5
(ppm)
5000.
5000.
3720.
5000.
6510.
20000.
16300.
9590.
9560.
20000.*
12800.
20000. •
20000.*
20000.*
20000.*
2B400.
1 1600.
24100.
29700.
23800.
23300.
19800.
23400.
20000.*
20000.*
12700.
13500.
18134.
7349.



22000.
27200.
24500.
19000.
1B400.
21 100.
19700.
23400.
18800.
18900.
19400.
Maximum Concentration
OZONE
Expt
(ppm)
0.822
0.903
0.972
0.633
O.B4B
0.363
0.461
0.340
0.727
0.670
0.230
0.382
0.673
0.974
0.788
1 .044
0.65B
0.731
O.B4B
0.804
0.697
0.719
0.405
0.645
0.446
0.674
0.432
0.608
0. 238



0. 227
0.506
0. 247
0.646
0.022
0.717
0.872
0.206
0.266
0.594
0.285
Calc
(ppm)
0.939
1 . 155
1.006
0.934
1.015
0.604
0.413
0.403
0.948
0.913
0. 269
0.306
0.515
0.874
0.632
1 . 183
0.614
0.687
0. 720
0.731
0.739
0.687
0.666
0.759
0.474
0.772
0.471
0.651
0. 254



0.081
0.280
0. 132
0. 728
0.045
0.827
0.972
0.451
0.449
0.783
0.452
Calc
-Expt
(ppm)
0.117
0.252
0.034
0.301
0. 167
0.240
-0.049
0.063
0.221
0.243
0.039
-0.076
-0. 158
-0.099
-0. 156
0. 139
-0.044
-0.045
-0.127
-0.073
0.042
-0.032
0.261
0.114
0.029
0.098
0.039
0.043
0.117
0.093
O.OB1

-0. 147
-0.226
-0.115
0.082
0.023
0.110
0. 100
0.245
0. 182
0. 190
0. 167
Calc
-Expt
/Avg
0. 13
0. 24
0.03
0.38
0.18
0.50
-0. 11
0. 17
0.26
0.31
0. 15
-0.22
-0.27
-0. 11
-0.22
0. 12
-0.07
-0.06
-0.16
-0.09
0.06
-0.05
0.49
0. 16
0.06
0. 14
0.09
O.OB
0.20
0. 16
0. 14

-0.95
-0.58
-0.61
0. 12

0.14
0.11
0.74
0.51
0.28
0.45
Average
d( 103] -
Expt
Calc
Initial
(NO) )/dt
Calc
-Expt
(ppb/mln) --
5. 16
12. 18
6.70
3.69
6.91
0.46
1 .40
1.23
2.63
2.53
1 .03
0.85
1 .24
2.69
2.41
3.26
.46
.69
2.10
.76
.66
.81
1 .23
2.07
1 .25
2.38
1 .67
3.62
2.60



2. 29
4. 16
1 .99
3.24
1 .36
7.91
5.39
0.96
1.16
1 .78
1 . 19
7.81
13.94
6.36
5.50
8.41
0.46
1 .45
1.38
3.53
3.44
1 . 15
0.77
1 . 10
2.83
2.06
2.84
1.59
1.61
1 . 75
1 .87
1 . 74
1.67
1 .67
2. 15
1 .21
2.45
1 .64
4. 15
3.08



1 .40
2.30
1.71
4.54
1 .69
11.12
7.16
1 . 15
1 .31
2. 19
1 .31
2.65
1 .77
-0.34
1.81
1 .50
0.02
0.05
0. 15
0.70
0.91
0. 12
-0.09
-0. 14
0.14
-0.35
-0.42
0. 13
-0.08
-0.35
0.10
0.08
-0. 14
0.44
O.OB
-0.04
0.06
-0.03
0.53
1 .01
0.76
0.85

-0.89
-1 .86
-0.27
1.31
0.34
3.21
1 .77
0.20
0.15
0.41
0. 12
Calc
-Expt
/Avg
0.41
0.14
-0.05
0.39
0.20
0.04
0.03
0.12
0.22
0.31
0.11
-0.11
-0.12
0.05
-0.16
-0. 14
0.08
-0.05
-0.18
0.06
0.05
-0.08
0.31
0.04
-0.04
0.03
-0.02
0.11
0. 19
0.17
0.14

-0.48
-0.58
-0. 15
0.34
0.22
0.34
0. 28
0. 19
0. 12
0.21
0. 10
                                                                                                                    (cont1nued)

-------
u>
-4
en
            Appendix  C.   Selected  Reault*  of  Simulation* of  Individual  Organlc-NOx-AIr Experiment* Using the

                          RADM-M Mechanism  (continued).
                                                                                                                       Page  3
Experiment
Initial
Concent ret Ions
NOx HC HC/NOx
(ppm) (ppmC)
Group Average
S. Oev.
Avg. Abe. Value
S. Dev.
0.52
0.20
3. 1
2.3
6.9
6.2
Ave
Temp
(degK)
299.4
f. 0
Ave
H20
(ppm)
21 127.
2656.
Maximum Concent rat
OZONE
Expt
(ppm)
0
0
.417
.263
Calc
(ppm)
0
0
.473
.321
Celc
-Expt
(ppm)
0.055
0. 154
0. 144
0.066
1 on
Average
d( [03] -
Calc
-Expt
/Avg
0
0
0
0
.02
.55
.45
.29
Initial
[NO] )/dt
Calc
Expt Calc -Expt
— (ppb/mln) --
2.
2.
86
IB
3.26
3. IB
0.41
1 . 34
0.96
0.99
Calc
-Expt
/Avg
0.05
0.32
0.27
0.15
4. trans-2-Butane-NOx
EC146
EC147
EC157
ST2783B
Croup Average
S. Dev.
Avg. Abe. Value
S. Dev.
0.51
0.98
0.53
0.43
0.61
0.25
0.9
1 .7
0.9
2.0
1 .4
0.6
KB
1 .7
1 .7
4. 7
2.5
1.5
303.0
303.0
303.0
295.5
301 . 1
3.7
26900.
22300.
27500.
19400.
24025.
3B60.
0
0
0
0
0
0
.247
. 154
.205
.523
.282
.165
0
0
0
0
0
0
. 109
.083
.091
.543
.207
.224
-0. 13B
-0.071
-0.114
0.019
-0.076
0.069
0.085
0.052
-0
-0
-0
0
-0
0
0
0
.77
.60
.77
.04
.53
.38
.54
.35
5.
9.
5.
3.
6.
2.
87
63
96
00
17
80
3.85
7.39
3.54
3.16
4.48
1.96
-2.02
-2.45
-2.42
0. 15
-1.68
1 .24
1 .76
1 .09
-0.42
-0.28
-0.51
0.05
-0.29
0.24
0.31
0.20
             5. leobutene-NOx



               ITC694           0.51
4.6
9. 1
303.0
20000.
0.900   0.731  -0.169  -0.21
6.84
6.04  -2.80  -0.3B
6. 1-Hexene-NOx
ITC929
ITC931
ITC934
ITC937
Group Average
S. Dev.
Avg. Abs . Value
S. Dev.
7. Isoprene-NOx
EC520
EC522
EC524
EC525

0.51
0.49
1 .00
0.99
0.75
0.28



0.49
0.92
0.96
0.54

5. 1
10.3
9.7
0. 1
6.3
4.8



2.2
2.3
4.7
4.5

10.0
21.1
9.7
0. 1
10.2
8.6



4.5
2.4
4.9
6.3

303
303
303
303
303
0



303
303
303
303

.0
.0
.0
.0
.0
.0



.0
.0
.0
.0

19000.
20100.
20300.
20300.
19925.
624.



13000.
13000.
13000.
13000.

0.
0.
0.
0.
0.
0.



0.
0.
0.
0.

298
606
428
007
335
252



503
276
759
691

0.670
O.B19
O.B95
0.060
0.61 1
0.379



0.300
0. 1 18
0.430
0.630

0
0
0
0
0
0
0
0

-0
-0
-0
-0

.372
.213
.467
.053
. 276
. 182
.276
. 1B2

.203
. 158
.330
.061

0.
0.
0.

0.
0.
0.
0.

-0.
-0.
-0.
-0.

77
30
71

59
26
59
26

50
80
55
09

1
2
1
0
1
1



5
6
15
10

.27
.82
.84
.33
.57
.04



.30
.00
.57
. 13

3.33
10.67
5.00
0.57
4.89
4.26



7.56
6.75
15.93
22. 16

2.06
7.85
3. 16
0. 24
3.33
3. 25
3.33
3.25

2.26
0.76
0.37
12.04

0.89
1 . 16
0.93
0.52
0.88
0.26
0.88
0.26

0.35
0. 12
0.02
0.75
                                                                                                                     (cont Intied)

-------
             Appendix  C.   Selected Results of  Simulations of  Individual  Organ1c-N0x-A1r Experiments Using
                          RADM-M Mechanism (continued).
the
                                                                                                                       Page  4
to
Experiment


EC527
ITC81 1
ITC812
JL1680R
JL1680B
JL17BOR
JL1780B
JL2381R
ST0981R
Group Average
S. Dev.
Avg. Abs. Value
S. Oev.
8. a-P1nene-NOx
JL15BOR
JL15BOB
JL2580R
JL2580B
Group Average
S. Oev.
Avg. Abs. Value
S. Oev.
9. Ethane-NOx
ITC999
10. n-Butane-NOx
EC130
EC133
EC 134
EC137
EC162
EC163
EC 168
EC1 78
EC304
Initial
Concent rat

NOx
(ppm)
0.50
0.50
0.52
0. 18
0. 18
0.46
0.47
0.43
0.17
0.49
0.25


0. 18
0. 19
0.25
0.25
0.22
0.04



0.09

0. 10
0.50
0.51
0.50
0.5t
0.49
0.49
0.10
0.47

HC
(ppmC)
2.2
3.4
1 .8
4.6
6.4
1 .0
2.6
1 .4
1 .0
2.9
1 .7


1 . 1
2.6
1 .0
0.3
1.3
1 .0



45.4

17.6
8.6
8. 3
8. 7
8.2
9.0
8.0
7.8
17.1
1ons

HC/NOx
4.5
6.B
3.4
26.0
36.6
2.1
5.5
3.4
6. 1
8.8
10.4


5.8
13.8
4.0
1 .4
6.3
5.3



534. 1

179.3
17.1
16.3
17.3
16.3
18.3
16.2
79.6
36.7
Ave
Temp

(degK)
303.0
303.0
303.0
305.0
301 .9
305.6
302.8
306. 1
297.5
303. 1
2. 1


302.4
299.6
302.4
300. 1
301 . 1
1 .5



303.0

303.0
303.0
303.0
303 .0
303 .0
303.0
303.0
303.0
303.0
Ave
H20

(ppm)
13000.
17300.
17300.
30200.
26800.
27000.
25000.
20000.*
20000. •
19123.
6284.


30300.
29200.
31700.
26300.
29375.
2291 .



20500.

26200.
27800.
20000. •
20000. •
27300.
28500.
20700.
20900.
26800.
Maximum Concentration
OZONE

Expt
(ppm)
0.547
0.919
0.76B
0.652
0.837
0.806
1 .298
0.750
0.506
0.716
0. 246


0.201
0.470
0.377
0.334
0.346
0.112



0.243

0.459
0.249
0.034
0.042
0.112
0.454
0.655
0.384
0.362

Calc
(ppm)
0.317
0.578
0.253
0.712
0.600
0.420
0.835
0.691
0.457
0.488
0.20B


0. 297
0.421
0.327
0.211
0.314
0.087



0.203

0.444
0.045
0.038
0.044
0.050
0. 152
0.315
0.421
0.434
Calc
-Expt
(ppm)
-0. 230
-0.341
-0.515
0.060
-0. 237
-0.386
-0.463
-0.058
-0.049
-0. 228
0. 174
0.238
0. 160

0.095
-0.049
-0.050
-0. 123
-0.032
0.091
0.079
0.036

-0.040

-0.015
-0. 204
0.004
0.002
-0.062
-0.302
-0.340
0.037
0.073
Calc
-Expt
/Avg
-0.53
-0.46
-1.01
0.09
-0.33
-0.63
-0.43
-0.08
-0.10
-0.42
0.31
0.43
0.29

0.38
-0.11
-0. 14
-0.45
-O.OB
0.34
0.27
0. 17

-0.18

-0.03
-1 .38


-0.76
-0.99
-0. 70
0.09
0. 18
Average
d( [03] -


In1t1a
[NO] )
Ca Ic
Expt Calc -Expt
(ppb/m1n) --
5.97
10.00
4.86
3.18
2.88
1 .96
4.20
2. 13
1 .46
5.66
4.06


0.61
1 .59
0.84
0.84
0.97
0.43



1 .32

4.41
2.42
0.94
1 .02
1 .73
3.31
2.03
1 .61
2.09
7.95
B.96
2.43
3.00
2.43
1 .36
2.87
2.01
1 .45
6.53
6.29


0.67
1 .96
0.83
0.68
1 .03
0.62



1 .42

2.49
1 .07
1 .30
1 .27
1.21
1 .71
1.14
1 .68
2.31
1 .97
-1 .04
-2.43
-0. IB
-0.44
-0.60
-1 .33
-0. 12
-0.01
0.86
3.58
1 .81
3.18

0.06
0.37
-0.02
-0. 16
0.06
0. 22
0. 15
0. 16

0. 10

-1 .93
-1 .35
0.36
0.25
-0.51
-1 .60
-0.89
0.07
0.22
1
/dt
Cal c
-Expt
/Avg
0.28
-0.11
-0.67
-0.06
-0.17
-0.36
-0.38
-0.06
0.00
-0.02
0.36
0.26
0.24

0.10
0.21
-0.02
-0.21
0. 02
0. 18
0. 13
0.09

0.07

-0.56
-0.77
0.32
0.22
-0.35
-0.64
-0.56
0.04
0. 10
                                                                                                                    (cont1nued)

-------
             Appendix C.   Selected Results of Simulations of Individual Organ1c-N0x-A1r Experiments Using the
                          RADM-M Mechanism (continued).
                                                                                                                       Page
UJ
-J
CD
Experiment


EC305
EC306
EC307
EC30B
EC309
ITC507
ITC533
ITC770
ITC939
ITC94S
OTC211
JL2178R
JL2176B
JL227BR
JL227BB
ST1B7BB
OC0979R
OC1B79B
Group Average
S. Dev.
Avg. Abs. Value
S. Dev.
1 1 . C4+ Branched
EC 165
EC 169
EC171
OC1B79R
OC2079B
AU1983R
AU1983B
Group Average
S. Oev.
Avg. Abs. Value
S. Dev.
Initial
Concentrat Ions
NOx
(ppm)
0. 10
0.19
0. 10
0.48
0.47
0.09
0. 12
0.52
0.51
0.26
0.55
0.24
0.24
0.55
0.55
0.21
0.21
0.20
0.34
0.18


HC
(ppmC)
15.7
25.8
25.8
16.2
17.2
15.2
11.9
37.9
14.8
10.0
42.8
7.2
15.4
7.9
17.5
21.2
14.6
14.3
15.7
8.8


HC/NOx

159.7
138.2
252.9
33.6
36.3
165.0
99.8
72.8
28.9
38.2
77.5
29.9
63.9
14.4
31 .5
103.1
71 .0
71.8
69.8
61 .6


Ave
Temp

(degK)
303.0
302.6
303.0
288.7
310.9
303.0
303.0
303.0
303.0
303.0
300. 1
302.9
302.9
305.2
305.2
298.3
297.2
295.0
302. 1
3.6


Ave
H20

(ppm)
25400.
24200.
29700.
8750.
12400.
20300.
20800.
16800.
20300.
20300.
3530.
20000.
20000.
20000.
20000.
20000.
15300.
28100.
20892.
6033.


Maximum Concentrat
OZONE
Expt
(ppm)
0.398
0.535
0.420
0.047
0.545
0. 149
0. 165
0.042
0.017
0.054
0.008
0.763
0.986
0. 166
0.788
0.1B5
0. 191
0.208
0.312
0.268


Calc
(ppm)
0.531
0.709
0.588
0.210
0.616
0.356
0.327
0.050
0.025
0. 141
0. 190
0.594
0.868
0. 138
0.41B
0. 1 14
0.272
0. 158
0.305
0.235


Calc
-Expt
(ppm)
0.133
0. 174
0. 166
0. 164
0.071
0.207
0.162
O.OOB
0.008
O.OB7
0. 182
-0. 168
-0. 1 19
-0.028
-0.369
-0.072
0.081
-0.050
-0.006
0. 160
0. 122
0. 102
Ion
Calc
-Expt
/Avg
0.29
0.28
0.33

0. 12
0.62
0.66


0.89

-0.25
-0.13
-0. IB
-0.61
-0.48
0.35
-0.27
-0.08
0.59
0.47
0.36
Average
d( (031 -
Expt
Calc
Initial
INO] )/dt
Calc
-Expt
(ppb/m1n) --
2.39
2.38
2.61
1 .04
2.00
0.69
0.61
1 .56
0.36
0.53
0.56
1.17
1 .64
0.90
1.55
0.51
0.60
0.60
1 .53
0.98


3.43
4.20
5.89
2.62
2.48
1.54
1 .36
1 .98
0.65
1. 10
1 .40
1 .06
1 .46
0.75
1.28
0.43
0.59
0.52
1 .74
1 .21


1 .04
1 .82
3.28
1 .59
0.4B
0.85
0.75
0.42
0.28
0.57
O.B3
-0. 1 1
-0. 18
-0. 15
-0.27
-0.08
0.00
-0.09
0.21
1 .05
0.74
0.77
Calc
-Expt
/Avg
0.36
0.55
0.77
0.87
0.22
0.77
0.76
0.24
0.56
0.70
O.B5
-0. 10
-0. 12
-0. IB
-0. 19
-0.16
0.00
-0.16
0.13
0,49
0.41
0.28
Alksne-NOx
0. 10
0. 19
0. 10
0.20
0.22
0.38
0.37
0.22
0.11


11.3
4.5
3.5
16.4
12.6
4.7
4. 1
8.2
5.2


114.3
23.5
35.7
81 .9
56.5
12.6
10.9
47.9
38.7


303.0
303.0
303.0
295.0
295.3
307. 1
307. 1
301.9
5.0


18600.
32900.
19000.
28100.
27800.
21300.
21300.
24143.
5463.


0.488
0.493
0.403
0.236
0.217
0.088
0.057
0.283
0.181


0.479
0.326
0.306
0. 130
0.058
0. 197
0.078
0.225
0. 153


-0.009
-0. 168
-0.097
-0. 106
-0. 158
0. 109
0.020
-0.05B
0. 102
0.095
0.061
-0.02
-0.41
-0.27
-0.58
-1.15
0.77
0.30
-0. 19
0.62
0.50
0.37
1 .77
1 .00
1 .30
0.64
0.66
0.61
0.53
0.93
0.46


1 .65
0.70
0.79
0.50
0.39
0.65
0.55
0.75
0.42


-0.12
-0.30
-0.51
-0. 13
-0.27
0.05
0.02
-0. 18
0.20
0.20
0. 17
-0.07
-0.35
-0.49
-0.23
-0.52
0.07
0.03
-0.22
0.24
0.25
0.20
12. C5 + n-Alkane-NOx
EC 135
OC0979B
EC131
ITC559
0. 10
0.21
0. 10
0. 19
20.4
15. 1
24.6
279.4
212.7
73.3
251 . 1
1441 . 1
303.0
297. 2
303.0
303.0
25000.
15200.
25000.
20300.
0.435
0. 184
0.393
0.377
0.552
0.240
0.530
0.654
0.117
0.056
0. 137
0.277
0. 24
0.26
0.30
0.54
2.92
0.59
1.92
1.79
3. 16
0.55
2.95
2.46
0.25
-0.05
1 .03
0.67
o.oe
-0.08
0.42
0.32
                                                                                                                     (contInued)

-------
Appendix C.  Selected Results of Simulations of Individual Organ1c-N0x-A1r Experiments Using the
             RADM-M Mechanism (continued).
                                                                                                          Page  6
Experiment


ITC538
ITC540
ITC552
ITC761
ITC762
ITC763
ITC797
ST1879R
EC 155
ITC1001
Group Average
S. Dev.
Avg. Abs. Value
S. Dev.
Initial
Concentrat Ions

NOx
(ppm)
0.11
0.11
0. 13
0.52
0.27
0.26
0.52
0.21
0. 10
0.11
0.21
0. 14



HC HC/NOx
(ppmC)
60.3 529.0
274.8 2421.3
428.8 3276,4
75.2 145.9
74.7 280.4
7.7 27.7
7.3 14.0
6.3 30.4
37.3 385.1
2.4 22.4
93.9 650.9
133.5 1015.3


Ave
Temp

(degK)
303.0
303.0
303.0
303.0
303.0
303.0
303.0
296.3
303.0
303.0
302.3
1 .9


Ave
H20

(ppm)
16900.
20300.
20300.
17500.
17000.
17100.
12200.
20000.*
26200.
17500.
19464.
3919.


Maximum Concentrat
OZONE

Expt
(ppm)
0. 150
0.360
0.315
0.030
0. 105
0.041
0.004
0. 122
0.264
0.036
0.201
0. 153



Calc
(ppm)
0.300
0.370
0.344
0.033
0. 176
0.066
0.010
0.300
0.521
0.073
0.298
0.211


Calc
-Expt
(ppm)
0. 150
0.010
0.029
0.003
0.071
0.025
0.006
0. 179
0. 257
0.037
0.097
0.092
0.097
0.092
Ion
Calc
-Expt
/Avg
0.67
0.03
0.09

0.50


0.85
0.65

0.41
0.27
0.41
0.27
Average
d( [03] -


Inlt lal
[NO] )/dt
r« i r
Expt Calc -Expt
(ppb/m1n) --
0.74
1 .85
1 . 19
1 .08
0.83
0.68
0.64
0.40
1 .33
0. 15
1.15
0.75


1 .03
1 . 10
1 .07
1 . 23
1.51
1 .40
0.93
0.44
1 .65
0.42
1 .42
0.87


0. 29
-0.76
-0.12
0. 16
0. 68
0. 72
0. 29
0.04
0.32
0.27
0. 27
0.44
0.40
0.31
Ca Ic
-Expt
/Avg
0. 33
-0.51
-0. 1 1
0. 14
0. 58
0.69
0. 37
0.10
0. 22
0.95
0. 25
0. 36
0. 35
0.26
13. Methy Icycl ohexane-NOx
ITC765
ITC766
ITC767
ITC800
Group Average
S. Dev.
Avg. Abs. Value
S. Dev.
14. Benzene-NOx
ITC560
ITC561
ITC562
ITC698
ITC710
ITC831
Group Average
S. Dev.
Avg. Abs. Value
S. Dev.
0.53
0.26
0.55
0.54
0.47
0. 14

0. 12
0. 1 1
0.56
0.50
0.55
1 .01
0.47
0.33

0.0 0.1
0.0 0.1
0.1 0.1
0.0 0.1
0.0 0.1
0.0 0.0

332.3 2874.4
79. 1 694.2
83.8 149.7
83.5 167.4
83.6 151.0
12.2 12.1
112.4 674.8
111.3 1 103.2

303.0
303.0
303.0
303.0
303.0
0.0

303 .0
303.0
303.0
303. 0
303.0
303.0
303.0
0.0

17900.
17100.
17900.
17500.
17600.
383.

17100.
17600.
18800.
21900.
21000.
18300.
19117.
1920.

0.022
0. 121
0.041
0.015
0.050


0.323
0.273
0.412
0.374
0. 367
0.021
0. 295
0. 142

O.OM
0. 159
0.029
0.010
0.052


0.329
0.294
0.528
0.495
0.512
0.010
0.361
0. 199

-0.01 1
0.036
-0.012
-0.004
0.002
0.016

0.006
0.021
0. 1 16
0.121
0. 145
-0.011
0.066
0.068
0.070
0.064
0.27

0.27
0.27

0.02
0.08
0.25
0.28
0.33

0.19
0.14
0.19
0. 14
0.86
0.87
1 .09
0.86
0.92
0.12

7.01
4.75
2.85
2.87
2.70
0.17
3.39
2.30

0.66
1 .34
1 .05
0.72
0.95
0.32

15.95
10.53
10.45
10.39
9.92
0.32
9.59
5.08

-0.20
0.48
-0.04
-0.13
0.03
0.31
0.21
0. 19

8 .94
5.78
7 .60
7.52
7. 22
0.15
6. 20
3.13
6.20
3. 13
-0.26
0.43
-0.03
-0. 17
-0.01
0.31
0.22
0. 17

0. 78
0.76
1 . 14
1 . 13
1 . 14
0.61
0.93
0.24
0.93
0.24
                                                                                                       (cont1nued)

-------
             Appendix C.   Selected  Result*  of  Simulations,  of  Individual  Organ1c-N0x-A1r Experiment* Using the
                           RADM-M Mechanlem  (continued).
                                                                                                                        Page  7
u>
00
O
Experiment
Initial
Concentrat 1on»
NOx HC
(ppm) (ppmC)
IS. Tolutne-NOx
EC264
EC265
EC266
EC269
EC270
EC271
EC272
EC273
EC327
EC336
EC337
EC339
EC340
ITC699
ITC828
JL3080R
AU2780R
AU27B2B
OC2782R
AU0183R
Group Average
S. Dev.
Avg. Abs. Value
S. Dev.
16. Xylene-NOx
EC343
EC344
EC345
EC346
ITC702
ITC827
JL3080B
AU2782R
OC2782B
AU0183B
Group Average
S. Dev.
Avg. Abs. Value
S. Dev.

0.4B
0.48
0.49
0.47
0.46
0.21
0.48
0. 11
0.45
0.44
0.45
0.44
0.43
0.51
1 .02
0. 18
0.4S
0.43
0.39
0.39
0.44
0. 18



0.28
0.67
0.28
0.26
0.52
1 .07
0.1B
0.43
0.39
0.37
0.45
0.26



6. 1
7.5
6.4
4.0
4.2
B.O
4. 1
4. 1
4.0
7.2
7.9
5.0
4. 1
10.5
3.0
3.9
2.3
3.0
4.5
4.6
5.4
2.3



4.2
4.0
3.7
3.9
4.0
1 .2
2.2
2.0
2.8
2.7
3. 1
1 .0


HC/NOx

17.0
15.6
17.0
8.4
9.0
37.4
6.5
37.2
6.9
16.3
17.7
11.3
9.5
20.6
3.0
21.3
4.6
7.0
11.7
1 1 .8
14.7
9.2



14.9
5.9
13.3
14.8
7.8
1 . 1
12.4
4.6
7.0
7.2
8.9
4.7


Ave
Temp
(degK)

303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
304.5
302.2
301 .7
289. 8
306. 1
302.5
3. 1



303.0
303.0
303.0
303.0
303.0
303.0
304.5
301 . 7
289.8
306.0
302.0
4.4


Ave
H2O
(ppm)

18900.
21900.
19400.
19100.
20000.
20500.
19400.
21200.
22400.
26900.
27000.
26800.
25900.
20300.
17200.
35200.
32000.
24700.
19500.
23700.
23100.
4692.



29500.
28900.
25200.
ib900.
20300.
14400.
28000.
24700.
19500.
23700.
240)0.
4737.


Maximum Concentration
OZONE
Expt
(ppm)

0.419
0.393
0.405
0.318
0.369
0.296
0.410
0.215
0.376
0.396
0.325
0.225
0.344
0.465
0.021
0.273
0.736
0. 1 16
0. 123
0.458
0.335
0. 154



0.283
0.589
0.396
0.384
0.627
0.021
0.555
0.491
0.396
0.688
0.443
0.195


Calc
(ppm)

0.397
0.395
0.401
0.31 1
0.385
0.318
0.306
0.245
0.376
0.434
0.389
0.337
0.356
0.441
0.008
0.441
0.879
0.295
0.346
0.561
0.382
0.159



0.380
0.552
0.369
0.376
0.509
0.013
0.535
0.606
0.453
0.784
0.460
0.200


Calc
-Expt
(ppm)

-0.022
0.001
-0.004
-0.006
0.016
0.023
-0. 105
0.030
0.000
0.039
0.064
0.112
0.012
-0.043
-0.013
0. 167
0. 143
0. 180
0.223
0. 123
0.047
0.084
0.066
0.069

0.096
-0.037
-0.007
-0.007
-0. 1 18
-0.007
-0.019
0. 1 17
0.057
0.096
0.017
0.073
0.056
0.047
Calc
-Expt
/Avg

-0.05
0.00
-0.01
-0.02
0.04
0.07
-0.29
0. 13
0.00
0.09
0. 18
0.40
0.03
-0.09

0.47
0. 18
0.87
0.95
0.24
0. 17
0.31
0.22
0.28

0. 29
-0.06
-0.02
-0.02
-0.21

-0.04
0.21
0. 13
0. 13
0.05
0. 16
0.12
0. 10
Average
d( [03) -
Initial
(NO) )/dt
Calc
Expt Calc -Expt
(ppb/m1n) --

4.46
3.56
4.62
2.55
3.72
6.56
3.69
5.90
2.49
6.06
2.55
1 .53
2.50
4.72
0.49
1 .20
1 .76
0.63
0.66
1.75
3.07
1 .87



8. 18
10.72
1 1 .35
7.65
7.79
1 . 12
1 .97
1 .31
1 .43
2.39
5.39
4. 14



5. 24
5.07
5.43
2.67
4.12
5.58
2.65
3.46
2.76
6.76
4.11
2.99
2.59
4.88
1.21
1 .04
1 .53
0.67
0.81
1 .72
3.26
1 .61



10.60
9.93
10.59
10.49
6.09
2.08
1 .93
1 .64
1 .97
2.75
6.01
4. 22



0.79
1 .52
0.81
0.12
0.39
-0 .99
-1 .04
-2.44
0.27
0.70
1 .56
1 .46
0.09
0. 15
0.72
-0.16
-0.23
0.03
0. 15
-0.03
0. 19
0.94
0.68
0.66

2.41
-0.79
-0.76
2.84
0.30
0.96
-0.04
0.33
0.54
0.36
0.61
1 . 19
0.93
0.94
Calc
-Expt
/Avg

0. 16
0.35
0. 16
0.05
0. 10
-0,16
-0.33
-0.52
0.10
0.11
0.47
0.64
0.03
0.03
0.85
-0.14
-0.14
0.05
0.20
-0.02
0.10
0.31
0.23
0.23

0. 26
-0.08
-0.07
0.31
0.04
0.60
-0.02
0.22
0 .32
0.14
0.17
0.21
0.21
0. 16
                                                                                                                     (cont1nued)

-------
            Appendix  C.   Selected  Results  of  Simulations  of  Individual  Organlc-NOx-A1r Experiments Using the
                          RADM-M  Mechanism  (continued).
                                                                                                                       Page  B
Lo
OO
Experiment

17. Mes1tylene-N0x
EC900
EC901
EC903
ITC703
ITC706
ITC709
ITC742
ITCB26
Group Average
S. Dev.
Avg. Abs. Value
S. Dev.
18. Tetralln-NOx
ITC739
ITC747
ITC748
ITC750
ITC832
Group Average
S. Dev.
Avg. Abs. Value
S. Dev.
19. Naphthalane-NOx
ITC751
ITC755
ITC756
ITC798
ITC802
Group Average
S. Dev.
Av0. Abs. Value
S. Oev.
In1 1 lal
Concentrat
NOx
(ppm)

0.53
0.51
1 .00
0.50
0.49
0.99
0.48
0.90
0.68
0.24



0.52
0.50
0.22
0.53
1 .00
0.55
0.28


0.52
0.24
0.26
0.53
0.53
0.42
0. 15


HC
(ppmC)

5.4
2.7
4.7
5.3
2.7
4.7
4.6
0.8
3.9
1 .6



2.4
93.2
84.0
44.5
39.4
52.7
36.7


7.5
14. 1
27.4
19.4
8.4
15.4
8.3


Ions
HC/NOx

10.2
5.2
4.7
10.6
5.4
4.7
9.7
0.9
6.4
3.4



4.6
187.4
385. 1
84.6
39.5
140.2
153.2


14.3
58.3
106.6
36.6
15.9
46.4
3B.2


Ave
Temp
(degK)

303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
0.0



303.0
303.0
303.0
303.0
303.0
303.0
0.0


303.0
303.0
303.0
303.0
303.0
303.0
0.0


Ave
H20
(ppm)

20000.*
20000.*
20000.*
21000.
21000.
21000.
18100.
17200.
19786.
1419.



17900.
17100.
17500.
17100.
16800.
17280.
427.


15100.
17500.
17200.
16400.
16800.
16600.
935.


Maximum Concentration
OZONE
Expt
(ppm)

0.381
0.384
0.502
0.707
0.641
0.779
0.773
0.022
0.524
0. 258



0.002
0.508
0.370
0.482
0.073
0.287
0.235


0.113
0.259
0.282
0.204
0. 124
0. 196
0.077


Calc
(ppm)

0.434
0.341
0.424
0.496
0.510
0.560
0.502
0.009
0.410
0. 175



0.358
0.720
0.478
0.697
0.336
0.516
0. 182


0.548
0.450
0.486
0.624
0.551
0.532
0.067


Calc
-Expt
(ppm)

0.053
-0.042
-0.078
-0.211
-0. 130
-0.219
-0. 271
-0.013
-0.114
0.113
0. 127
0.096

0.356
0.211
0. 108
0.215
0.263
0.231
0.090
0.231
0.090

0.436
0.191
0.203
0.421
0.427
0.335
0. 127
0.335
0. 127
Calc
-Expt
/Avg

0.13
-0. 12
-0. 17
-0.35
-0.23
-0.33
-0.42

-0.21
0. 19
0.25
0. 12

0.34
0. 26
0. 36
1 .29
0.56
0.49
0.56
0.49

1 .32
0.54
0.53
1.02
1 .27
0.93
0. 38
0.93
0.38
d(
Expt

3.85
8.88
14.96
14.59
7.20
1 1 . 74
13.14
1 .68
9.50
4.97



0.63
2.73
2.60
2. 10
1 .38
1 .89
0.88


1 .08
1 .57
2.20
1.66
1 .42
1 .59
0.41


Average
[03] -
Initial
[NO] )/dt
Calc
Calc -Expt
(ppb/m1n) --


3.72
5.88
9.40
4.48
7.21
7.77
1 .29
5.68
2. 74



2.64
54.42
31 .58
45.57
1 .00
27.04
24.42


1 1 . 27
21.78
29. 17
29.50
12.55
20.85
8. 74




-5. 17
-9.08
-5. 19
-2.72
-4.53
-5.37
-0.39
-4.63
2 . 66
4.63
2.66

2.02
51 .69
28.98
43.46
-0.37
25. 16
23.67
25.30
23.47

10.18
20.22
26.97
27.84
11.13
19.27
8 . 40
19.27
8.40
Calc
-Expt
/Avg


-0.82
-0.87
-0.43
-0 .47
-0.48
-0.51
-0.26
-0.55
0 . 22
0.55
0.22

1 .23
1.81
1 .70
1 .82
-0.31
1 . 25
0.91
1 .37
0.64

1 .65
1 . 73
1 . 72
1 . 79
1 .59
1 . 70
0 . 08
1 . 70
0.08
                                                                                                                    (cont1nued)

-------
             Appendix  C.   Selected Results of  Simulations of  Individual  Organ1c-N0x-A1r Experiments Using the
                          RADM-M Mechanism (continued).
                                                                                                                       Page  9
10
OD
Experiment

Initial
Concentrations
NOx
(ppm)
HC HC/NOx
(ppmC)
Ave
Temp
(degK)
Ave
H20
(ppm)
Maximum Concentration
OZONE
Expt
(ppm)
Catc
(ppm)
Calc
-Expt
(ppm)
Calc
-Expt
/Avg
Average
d( [03] -
Initial
[NO) )/dt
Calc
Expt Calc -Expt
(ppb/ml n) , --
Calc
-Expt
/Avg
20. 2,3-D1mathy1naphtha1ene-NOx
ITC775
ITC771
ITC806
ITC774
Group Average
S. Oev.
Avg . Abs . Value
S. Dev.
21 . Simple Mixture
EC 144
EC 145
EC160
EC149
EC 150
EC151
EC152
EC153
EC161
OC1278R
OC257BR
AU0180R
AU14BOR
EC 166
EC172
ST0682R
ST0682B
EC 106
EC 11 3
EC114
EC1 15
EC1 16
EC335
EC329
EC330
EC334
EC338
EC328
JL1581R
JL1881R
ST2481R
AU2781R
0.29
0.26
0.33
0.56
0.36
0. 14


Runs
0.51
0.99
0.99
0.99
1 .00
2.06
0.50
0.97
0.51
0.48
0.44
0.56
0.47
0. 10
0. 10
0.46
0.45
0.50
0. It
1 .00
0.51
0.49
0.44
0.45
0.29
0.45
0.45
0.45
0.27
0.26
0.23
0.23
1.7
4.8
5.9
4.0
4. 1
1 .8



4.7
3.4
3.2
2.0
3.5
5.2
3.7
6.6
3.2
1 .4
1 .4
0.5
1 .4
9.2
2.8
2.8
2.9
9.2
9.5
17.3
12.7
IB. 6
7. 7
4.2
4.3
8. 1
15.0
12. 1
2.3
2. 1
1 .6
2. 1
5.8
18.0
17.7
7. 1
12.2
6.6



9.3
3.4
3.3
2.0
3.5
2.5
7.3
6.8
6.4
3.0
3.1
0.8
3.0
92.0
28.9
6.2
6.5
18.3
85.0
17.3
25.2
37.6
17.4
9.2
14.6
18.2
33.7
27.1
8.4
8.0
6.7
9.2
303.0
303.0
303.0
303.0
303.0
0.0



303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
294.3
288.4
304.9
305. 2
303.0
303.0
299. 1
299. 1
303.0
303.0
303.0
303.0
303.0
303.0
303 . 0
303.0
303.0
303.0
303.0
304.6
303.2
292.2
299.5
16800.
20300.
17100.
22700.
19225.
2806.



24400.
24400.
29200.
29300.
29900.
30400.
26400.
25400.
36900.
16100.
12800.
23400.
34700.
27700.
21600.
22200.
22200.
20900.
19500.
28800.
22000.
27300.
24700.
27700.
26000.
-'7800.
27800.
27500.
20000.*
20000.*
22800.
30900.
0.274
0.293
0.360
0.341
0.317
0.040



1.065
0.777
0.874
0.286
0.799
0.147
0.791
1.050
0.857
0.260
0. 147
0.256
0.863
0.462
0.369
0.378
0.478
0.592
0.352
0.744
0.590
0.743
0.398
0.403
0.344
0.408
0.484
0.523
0.485
0.65B
0.211
0.554
0.296
0.388
0.438
0.551
0.418
0. 106



0.772
0.376
0.326
0.131
0.329
0.086
0.665
0.817
0.599
0.250
0. 125
0. 107
0.474
0.379
0.297
0.536
0.574
0.617
0.442
0.710
0.612
0.785
0.383
0.406
0.359
0.420
0.499
0.520
0.693
0.596
0.277
0.496
0.022
0.095
0.079
0.210
0. 101
0.079
0. 101
0.079

-0.293
-0.401
-0.548
-0. 155
-0.470
-0.060
-0. 127
-0.233
-0.259
-0.010
-0.022
-0.149
-0.389
-0.083
-0.072
0.158
0.096
0.025
0.090
-0.034
0.021
0.042
-0.015
0.002
0.015
0.012
0.016
-0.002
0.208
-0.063
0.066
-0.05B
0.08
0.28
0.20
0.47
0.26
0.17
0. 26
0.17

-0.32
-0.70
-0.91
-0.74
-0.83
-0.52
-0.17
-0.25
-0.36
-0.04
-0.16
-0.62
-0.58
-0.20
-0.22
0.34
0. 18
0.04
0.23
-0.05
0.04
0.06
-0.04
0.01
0.04
0.03
0.03
0.00
0.35
-0. 10
0.27
-0.11
1.73
2.67
2.69
2.84
2.48
0.51



10.53
5. 11
5.86
10.83
6.30
8.43
10.41
19.23
9.94
1.14
0.94
0.96
2.13
2.11
1 .00
1.11
1.35
3.80
5.26
8.03
3.78
8.28
4 .64
3.27
3.91
5.59
5. 19
3.50
1 .08
1.31
0.62
0.95
2.53
8.26
8.80
5.92
6.38
2.85



6. 15
3.84
3.48
5.21
4. 12
7.05
7.84
11.91
5.43
1 .22
1.01
0.64
1 .37
1.41
0.70
1 .29
1 .59
4.08
7.09
6. 1 1
3.44
9.30
4.31
3.27
3.50
5.09
4.78
3.31
1 .23
1 .23
0.67
0.92
0.81
5.59
6.11
3.09
3.90
2.45
3.90
2.45

-4.39
-1 .27
-2.39
-5.63
-2.18
-1 .38
-2.57
-7.32
-4.52
0.08
0.07
-0.35
-0.76
-0.70
-0.30
0. 18
0.24
0.28
1 .83
-1.91
-0.34
1.01
-0.33
0.00
-0.41
-0.50
-0.41
-0. 19
0. 15
-0.08
0.05
-0.03
0.38
1 .02
1 .06
0.70
0.79
0.32
0.79
0.32

-0.53
-0.28
-0.51
-0.70
-0.42
-0. IB
-0.28
-0.47
-0.59
0.07
0.07
-0.43
-0.44
-0.39
-0.35
0. 15
0. 16
0.07
0.30
-0.27
-0.09
0. 12
-0.07
0.00
-0.11
-0.09
-0.08
-0.06
0. 13
-0.06
0.08
-0.03
                                                                                                                    (cont1nued)

-------
            Appendix C.  Selected Results of  Simulations
                         RADM-M Mechanism (continued).
of Individual Organlc-NOx-AIr Experiments Using the
                                                             Page  10
U>
oo
wo
Experiment

Initial
Concentrat Ions
NOx
Ave
Temp
Ave
H20
HC HC/NOx
(ppm) (ppmC)
OC03B2R
OC0382B
NV15B2R
JL2281R
JL2181B
NV15B2B
DE07B2B
JN1379B
Group Average
S. Dev.
Avg. Abs. Value
S. Dev.
22. 3-Component
ITC479
ITC584
ITC579
ITC472
ITC474
ITC581
ITC5B5
ITC478
ITC482
ITC4B8
ITC492
ITC494
ITC498
1TC500
ITC502
ITC462
ITC466
ITC46B
ITC451
ITC455
ITC977
ITC985
ITC997
ITC979
ITC992
ITC981
ITC993
JL1581B
ST2481B
0.25
0.25
0. 18
0.26
0.24
0.18
0. 19
0.44
0.49
0.37


and "M1n1"
0.09
0. 10
0. 10
0. 10
0.09
0.09
0. 10
0. 10
0. 10
0.09
0.09
0.09
0. 10
0. 10
0.09
0. 1 1
0. 10
0.09
0. 10
0.09
0. 13
0. 12
0. 12
0. 12
0. 12
0. 14
0. 1 1
0.28
0.23
2.7
1 .9
2.6
2.4
1 .3
2.7
3.5
6. 1
5.2
4.5


11 .0
7.7
14.4
9. 1
5.4
14.9
16.7
13.7
15.5
19.2


(degK)
297.2
296.9
283.6
302.3*
304. 2*
263.8
286.4
294.9
300.2
5.7


(ppm)
21200.
21 100.
20000.*
20000.*
20000.*
20000. •
13400.
20000.*
24160.
51B8.


Maximum Concentration
OZONE
Expt
(ppm)
0.247
0. 158
0.061
0.626
0.237
0. 167
0.093
0.756
0.492
0.270


Calc
(ppm)
0.141
0.038
0.045
0.628
0.428
0.240
0.296
0.809
0.432
0.221


Calc
-Expt
(ppm)
-0. 106
-0. 120
-U.016
0.002
0.191
0.073
0.203
0.053
-0.060
0. 175
0. 124
0. 136
Calc
-Expt
/Ave
-0.55
-1 .23
-0.30
0.00
0.58
0.36
1 .05
0.07
-0.14
0.44
0.32
0.32
Average
d( [03] -
Expt
Calc
Initial
[NO] )/dt
Calc Calc
-Expt -Expt
(ppb/m1n) -- /Avg
0.76
0.62
0.42
1.18
0.63
0.71
0.55
2.44
4. 10
4.06


0.55

0.34
1.11
0.64
0.74
O.72
2. 78
3.32
2.77


-0.21 -0.32

-0.08 -0.22
-0.06 -0.06
0.01 0.02
0.04 0.05
0.18 0.28
0.34 0.13
-0.87 -0.14
1.83 0.25
1 . 10 0.22
1.69 0.18
Surrogates
4.0
4.0
3.4
3.6
4.2
4.3
4.B
5.0
1 .2
6.0
8. 1
B.5
3.9
4. 1
4.6
7.5
6.5
4.6
5.2
4.4
3.2
2.7
2.7
2.9
2.6
6.5
7. 1
2.3
1 .9
43.4
40.7
34.7
35. 1
45. 1
50.0
50.8
52.6'
12.3
66.2
68.4
92.5
40.2
42.7
49.5
71.3
62.7
49.0
52.0
47. B
25.1
22.9
23.0
23. 1
20.9
45.0
66.9
8.0
8. 1
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303. 0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
304.6
292. 2
20300.
20300.
15800.
13700.
15600.
16100.
'7100.
18300.
18300.
19600.
15000.
?1 100.
19300.
18700.
IB 100.
18300.
18200.
18000.
19800.
20200.
20900.
20100.
i!0100.
20100.
20900.
19000.
20100.
20000.*
14200.
0.332
0.375
0.356
0.258
0.293
0.352
0.311
0.320
0.279
0.293
0.326
0.321
0.320
0.286
0.271
0.099
0. 131
0. 176
0.312
0.304
0.329
0.336
0.316
0.404
0.384
0.328
0.276
0.474
0.246
0.323
0.320
0.303
0.331
0.329
0.315
0.332
0.352
0.298
0.332
0.355
0.363
0.336
0.322
0.304
0.269
0.283
0.304
0.296
0.300
0.330
0.300
0.298
0.418
0.407
0.375
0.359
0.632
0.283
-0.009
-0.055
-0.05"
0.073
0.036
-0.037
0.021
0.032
0.019
0.039
0.029
0.041
0.016
0.036
0.034
0. 170
0. 152
0. 128
-0.016
-0.004
0.001
-0.036
-0.018
0.014
0.023
0.047
0.083
0. 158
0.037
-0.03
-0. 16
-0.16
0.25
0.12
-0.11
0.07
0.09
0.07
0.12
0.09
0. 12
0.05
0. 12
0. 12
0.92
0.73
0.53
-0.05
-0.01
0.00
-0.11
-0.06
0.03
0.06
0.13
0. 26
0. 29
0.14
2.95
3.05
1 .54
1 .72
3.40
3.74
5.11
4.73
2.44
3.06
3.28
3.30
2. 16
5. 18
7.07
0.70
0.77
1 . 19
3.95
3.52
3.94
3.95
3.89
4.88
5.57
2.56
2.45
1. 10
0.71
4.79
3.93
2.14
3. O2
6.06
5.28
6.98
8. 19
4.39
5. 17
5.39
5.33
3.60
7.40
12.30
2.51
2.96
3. 48
5.63
5.05
3.50
3.31
3.26
4.09
4.47
3.33
2.71
1.16
0.71
1.84 0.48
0.88 0.25
0.61 0.33
1.31 0.55
2.69 0.57
1.54 0.34
1.BB 0.31
3.46 0.54
1.94 0.57
2.10 0.51
2.11 0.49
2.03 0.47
1.45 0.50
2.22 0.35
5.23 0.54
1.81 1.13
2.19 1.17
2.29 0.98
1.68 0.35
1.52 0.36
-0.44 -0.12
-0.65 -0.18
-0.63 -0.18
-0.78 -0.17
-1.10 -0.22
0.77 0.26
0.25 0.10
0.06 0.05
0.00 0.00
                                                                                                                    (cont1nued)

-------
            Appendix C.
Sal acted Results of Simulations of Individual Organlc-NOx-AIr Experiments Using  the

RADM-M M»ch«n1sm (continued).
                                                                                                                       Page  11
10
CO
XT
Experiment

Initial
Concentrat Ions
NOx
Ave
Temp
Ave
H20
HC HC/NOx
(ppm) (ppmC)
JN09B2B
JN1483R
JN2783B
AU1883B
AU2683R
JL18B1B
JN0982R
Group Average
S. Dev.
Avg. Abe. Value
S. Dev.
0.28
0.22
0.26
0.28
0.32
0.27
0.29
0. U
0.08


3. 1
2.6
2.9
0.6
2.6
2.3
3. 1
4. 1
1 .9


1 1 .0
11.6
11.1
2.0
8.1
8.5
10.7
37.0
23.7


(degK)
299.7*
301 .2
305. 1
303.0
302.2
303.2
299.7*
302.6
2.0


(ppm)
14300.
26100.
24400.
23100.
26400.
20000.*
14700.
19061 .
3028.


Maximum Concentrat
OZONE
Expt
(ppm)
0.667
0.585
0.511
0.556
0.646
0.693
0.714
0.366
0.148


Calc
(ppm)
0.586
0.529
0.596
0.545
0.612
0.643
0.647
0.387
0. 121


Calc
-Expt
(ppm)
-0.081
-0.056
0.085
-0.01 1
-0.035
-0.050
-0.068
0.021
0.063
0.050
0.043
1 on
Calc
-Expt
/Avg
-0.13
-0. 10
0. 15
-0.02
-0.05
-0.08
-0. 10
0.09
0.23
0. 16
0. 19
Average Initial
d( (03] - [NO] )/dt
Expt
Calc Calc
Calc -Expt -Expt
(ppb/mln) -- /Avg
.37
. 10
.01
0.70
.24
.91
.77
2.81
1 .61


.20 -0.17 -0.13
.07 -0.03 -0.03
.13 0.12 0.11
0.65 -0.05 -0.07
.18 -0.07 -0.05
.76 -0.15 -0.08
.56 -0.21 -0.13
3.85 1.05 0.28
2.43 1 .36 0.36
1.28 0.35
1.14 0.29
23. SAPRC 7-Component Surrogates
EC231
EC232
EC233
EC237
EC23B
EC241
EC242
EC243
EC245
EC246
EC247
Group Average
S. Dev.
Avg. Abs. Value
S. Dev.
0.49
0.49
0. 10
0.48
0.95
0.49
0.50
0.50
1 .00
0.51
0.51
0.65
0.24


13.2
9.3
9.5
10.5
10. 1
5.0
12.9
9.7
12.9
8.6
6.2
9.8
2.6


28.9
18.9
92.5
21.6
10.6
10.2
25.6
19.5
12.9
17.0
12.2
24.4
23.3


303,0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
0.0


17400.
17100.
21400.
23200.
25400.
20700.
24600.
16400.
20400.
22100.
22700.
21036.
3020.


0.620
0.342
0.330
0.652
0.691
0.408
0.682
0.716
0.894
0.574
0.657
0.597
0. 173


0.711
0.377
0.385
0.645
0.701
0.378
0.678
0.717
0.847
0.433
0.625
0.591
0. 167


0.090
0.035
0.056
-0.007
0.010
-0.030
-0.003
0.001
-0.047
-0.141
-0.032
-0.006
0.060
0.041
0.043
0.14
0. 10
0. 16
-0.01
0.01
-0.08
-0.01
0.00
-0.05
-0.28
-0.05
-0.01
0.12
0.08
0.08
6.99
3. 12
4.07
7.36
5. 10
2.87
17.53
14.50
13.41
2.65
7.49
7.74
5. 16


7.96 0.97 0.13
3.03 -0.09 -0.03
4.73 0.66 0.15
5.98 -1.38 -0.21
4.79 -0.31 -0.06
2.94 0.08 0.03
16.62 -0.91 -0.05
12.74 -1 . 76 -0.13
12.84 -0.57 -0.04
1.97 -0.68 -0.29
6.48 -1.01 -0.14
7.28 -0.46 -0.06
4.78 0.83 0.13
0.77 0.12
0.52 0.08
24. SAPRC 8-Component Surrogates
ITC626
ITC630
ITC631
ITC633
ITC635
ITC637
ITC865

0.30
0.31
0.32
0.64
1 .21
0.31
0.28

4.0
1 .9
1 .0
4.0
4.0
4.0
8.4

13.4
6.3
3.2
6.2
3.3
12.8
29.9

303.0
303.0
303 . 0
303. 0
303. 0
30:1 . o
303.0

14300.
16300.
17500.
16600.
17500.
16900.
15100.

0.618
0.284
0.043
0.231
0.006
0.617
0.632

0.548
0.287
0.066
0.308
0.021
0.545
0.514

-0.070
0.003
0.023
0.077
0.015
-0.072
-0.118

-0.12
0.01

0.28

-0.12
-0.21

2. IB
0.86
0.47
1.26
0.95
2.31
2.60

2.32 0.14 0.06
1 .04 0.18 0.19
0.54 0.07 0.15
1.58 0.31 0.22
1.05 0.09 0.09
2.30 -0.01 0.00
3.51 0.91 0.30
(cont 1nued)

-------
            Appendix C.  Selected Results of Simulations of  Individual Organlc-NOx-A1r  Experiments Using  the

                         RAOM-M Mechanism (continued).
Page 12
oo
oo
Experiment
Initial
Concent rat Ions
NOx HC HC/NOx
(ppm) (ppmC)
ITC867
ITC868
ITC871
ITCB72
ITC873
ITC874
ITC877
ITCB80
ITC881
ITC885
1TC886
ITC888
ITC891
OTC189A
OTC1B9B
OTC190A
OTC190B
OTC192A
OTC192B
OTC194A
OTC194B
OTC195A
OTC195B
OTC196B
OTC197A
OTC197B
OTC198A
OTC198B
OTC199A
OTC199B
OTC202A
OTC202B
OTC203A
OTC203B
OTC204A
OTC204B
OTC205A
OTC2058
OTC215A
OTC215B
OTC217A
OTC217B
OTC221A
OTC221B
OTC222A
0.28
0.37
0.37
0.38
0.39
0.38
0.38
0.73
0.73
0.64
0.73
0.33
0.32
0.45
0.4S
0.41
0.42
0.47
0.47
0.38
0.38
0.41
0.41
0.41
0.38
0.77
0.84
0.40
0.36
0.37
0.73
0.40
0.39
0. 19
0.35
0. 17
0.64
0. 14
0.45
0.44
0.50
0.51
0.41
0.42
0.44
4.8
2.9
1 .7
2. 1
1 .3
2. 1
2.3
2.2
2.3
1.5
2.3
4.7
4.4
3.3
4.0
3.8
3.8
4.2
4.2
7.4
3.7
1 .8
4. 1
3.7
3.6
3. 1
5.4
3.7
3.5
3.5
6.0
2.6
3.5
3.5
3.6
3.5
3.8
3.6
3.5
4. 1
4.9
4.6
2.4
2.0
3.4
17.2
7.8
4.6
5.7
3.4
5.7
6.2
3.0
3. 1
2.4
3. 1
14.5
13.7
7.4
8.9
9.2
9.0
8.9
B.9
19.3
9.7
4.5
9.9
9. 1
9.5
4. 1
6.4
9.4
9.6
9.6
8.2
6.7
9.0
18.5
10.2
20.3
4.5
26.5
7.8
9.3
9.6
9.0
5.8
4.7
7.8
Ave
Temp
(degK)
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
303.0
295.4
298.0
302.7
299.5
297.8
298.4
296. 1
295.7
351 .6
360.0
337.5
353.2
356.0
353.5
347.5
299.7
298.9
325. 1
324. 1
293.3
293.7
322.2
34B.6
464.9
433.4
293. 1
293.4
295.5
294. 2
290.6
288.8
298.5
Ave
H20
(ppm)
14600.
17100.
14600.
14600.
14600.
13200.
14600.
13700.
14000.
14600.
20000.*
18000.
20000.*
5000.*
5000.*
5400.*
5410.*
5000.*
5000.*
5000.*
5000.*
4660.*
3180.*
5000.*
5000.*
5000.*
5000.*
5000.*
5000.*
5000.*
5000.*
4560. •
4850.*
4880.*
4330. •
4330.*
5000. •
5000.*
4420.*
4420.*
7460.*
7460.*
4460.*
4460. *
5000.*
Maximum Concentration
OZONE
Expt
(ppm)
0.631
0.518
0.376
0.213
0. 160
0. 191
0.250
0.031
0.012
0.012
0.012
0.579
0.602
0.576
0.712
0.702
0.669
0.735
0.754
0.737
0.603
0. 164
0.681
0.597
0.621
0. 152
0.704
0.674
0.591
0.606
0.714
0.522
0.231
0.394
0.302
0.381
0.039
0.392
0.830
0.868
0.483
0.831
0.235
0.333
0.909
Calc
(ppm)
0.544
0.435
0.265
0.228
0.115
0. 160
0.237
0.046
0.033
0.027
0.022
0.512
0.515
0.766
0.845
0.841
0.793
0.852
0.760
0.905
0.701
0.112
0.775
0.804
0.554
0. 102
0.576
0.633
0.653
0.638
0.494
0.424
0.282
0.447
0.377
0.468
0.060
0.552
O.B57
0.937
0.584
0.935
0.286
0.357
0.964
Calc
-Expt
(ppm)
-0.088
-0.083
-0.111
0.015
-0.045
-0.031
-0.013
0.016
0.021
0.015
0.010
-0.068
-0.087
0. 190
0. 132
0.138
0. 124
0. 1 17
0.006
0. 167
0.098
-0.052
0.095
0.206
-0.068
-0.049
-0. 128
-0.041
0.062
0.032
-0.220
-0.09B
0.051
0.053
O.O75
0.087
0.021
0. 160
0.027
0.069
0.101
0. 104
0.051
0.024
0.055
Calc
-Expt
/Avg
-0. 15
-0. 17
-0.35
0.07
-0.33
-0. 18
-0.05




-0.12
-0. 16
0.28
0.17
0.18
0. 17
0. 15
0.01
0.20
0. 15
-0.38
0. 13
0.29
-0. 12
-0.39
-0.20
-0.06
0.10
0.05
-0.37
-0.21
0.20
0. 13
0.22
0.21

0.34
0.03
0.08
0.19
0.12
0.20
0.07
0.06
Average
d( [03] -
Initial
(NO] )/dt
Calc
Expt Calc -Expt
(ppb/m1n) --
2.23
2. 10
1 . 18
1 .20
0.85
0.83
1.14
0.85
1 .03
0.44
1 .24
1 .65
2.67
3.79
3.70
4.24
3.73
4.25
3.84
7.85
2.95
2.05
4. 14
3.86
4.12
2.86
6.09
3.86
3.86
3.40
6.95
3.B2
2.34
2.20
3.27
2.62
4.85
4.49
4.33
5.33
2.55
4.64
1 .63
1 .63
3.83
3.41
1 .46
1.12
1 .36
0.80
0.88
1 .34
0.99
1 . 10
0.78
0.70
1 .89
2.91
3.62
3.74
3.35
2.94
3.41
2.80
7.71
3.04
1 .20
3.83
3.37
2.66
2.07
3.69
2.58
3.05
2.92
13. 77
3.17
2.36
2.65
3.46
2.91
2. 16
3.71
3.65
5.49
2.52
4. 30
1 .50
1 .60
3.75
1 . 19
-0.64
-0.06
0. 16
-0.05
0.05
0.20
0. 14
0.07
0.35
-0.54
0.24
0.23
-0. 17
0.04
-0.89
-0.79
-0.85
-1 .05
-0. 14
0.09
-0.85
-0.31
-0.49
-1 .46
-0.79
-2.40
-1 .28
-0.81
-0.47
6.81
-0.65
0.02
0.45
0.21
0.29
-2.70
-0.78
-0.68
0.16
-0.03
-0.34
-0.12
-0.03
-0.09
Calc
-Expt
/Avg
0.42
-0.36
-0.05
0. 13
-0.06
0.06
0. 16
0. 15
0.06
0.57
-0.55
0. 14
0.08
-0.05
0.01
-0.23
-0.24
-0.22
-0.31
-0.02
0.03
-0.52
-0.08
-0.14
-0.43
-0.32
-0.49
-0.40
-0.23
-0. 15
0.66
-0. 19
0.01
0. 19
0.06
0. 1 1
-0.77
-0. 19
-0. 17
0.03
-0.01
-0.08
-0.08
-0.02
-0.02
                                                                                                                    (cont1nued)

-------
             Appendix C.   Selected Results of Simulations of Individual Organ1c-NOx-A1r Experiments Using  the
                          RADM-M Mechanism (continued).
                                                                                                                        Page  13
Oa
CTi
Experiment

Initial
Concentrat


Ions

NOx HC HC/NOx
(ppm) (ppmC)
OTC222B
OTC223A
OTC223B
OTC224A
OTC224B
OTC226A
OTC22BA
OTC228B
OTC229A
OTC229B
OTC230A
OTC230B
OTC237A
OTC237B
OTC23BA
OTC23BB
OTC239A
OTC239B
OTC240A
OTC240B
OTC241A
OTC241B
OTC242A
OTC242B
OTC243A
OTC243B
OTC24BA
OTC248B
OTC249A
OTC249B
Group Average
S. Oev.
0.43
0.36
0.40
0.34
0.34
0.45
0.41
0.41
0.46
0.46
0.41
0.41
0.52
0.52
0.50
0.50
0.49
0.50
0.50
0.50
0.32
0.31
0.45
0.46
0.47
0.3B
0.48
0.46
0.48
0.46
0.45
0. 16
2.6
4.8
3.5
4.4
4.3
2.5
2.4
2.3
3.0
1 .7
3. 1
1 .7
4.3
4.4
2.9
3.9
2.7
2.5
1 .8
2.0
4.7
4.0
2.6
2. 1
4.7
3.7
3.7
2.9
5.6
4.9
3.5
1 .3
6.0
13.4
B.B
12.7
12.6
5.5
6.0
5.6
6.6
3.8
7.6
4.2
8.2
8.5
5.8
7.8
5.4
5.0
3.6
4. 1
14.8
12.9
5.8
4.7
10.0
9.7
7.7
6.3
11.8
10.7
8.7
4.9
Ave
Temp

(detiK)
296. 1
314.9
314.7
303.5
303.8
312.5
293.2
294.3
295.7
296.0
297.9
298. 2
292.3
292.8
292.5
292.0
290. 2
290.9
93.4
104. 2
290. 7
291. 1
30?. 0
302. 1
276.5
277.4
271 .6
268.9
286. 9
286.2
302.3
43.6
Ave Maximum Concentration
H20 OZONE

Expt
(ppm) (ppm)
5000.
4270.
4270.
6410.
6410.
6840.
5000.
5000.
5000.
5000.
5000.
5000.
3810.
3810.
4810.
4810.
5150.
5150.
6210.
6210.
4760.
4760.
4670.
4670.
4720.
4720.
5000.
5000.
3540.
0.940
0.953
0.771
0.776
0.813
0.751
0.246
0.296
0.253
0. 168
0.4B9
0.271
0.807
0.757
0.406
0.702
0.343
0.234
0.034
0.217
0.671
0.674
0. 182
0.639
0. 142
0. 152
0.056
O.OB1
0.348
3540. 0.325
7630. 0.454
4865. 0.272

Calc
(ppm)
1 .127
0.883
0.729
0.786
0.850
0.545
0.433
0.464
0.460
0.405
0.842
0.564
0.836
0.691
0.403
0.700
0.401
0. 172
0.037
0.119
0.698
0.683
0. 161
0.640
0.207
0. 174
0.060
0.090
0.340
0.365
0.483
0.286
Avg. Abs. Value
S. Dev.
25. UNC Mlscel
ST20B1R
DE0782R
AU2681R
AU2681B
AU2781B
ST03B1R
ST10B1R
ST2081B






Calc
-Expt
(ppm)
0.187
-0.069
-0.043
0.010
0.038
-0.205
0. 187
0. 168
0.207
0.237
0.353
0.293
0.029
-0.066
-0.003
-0.002
0.058
-0.062
0.003
-0.098
0.027
0.009
-0.021
0.001
0.065
0.022
0.004
0.009
-0.008
0.040
0.029
0. 103
0.078
0.072
Calc
-Expt
/Ava
0.16
-0.08
-0.06
0.01
0.05
-0.32
0.55
0.44
0.58
0.83
0.53
0.70
0.03
-0.09
-0.01
0.00
0.15
-0.31

-0.58
0.04
0.01
-0.12
0.00
0.37
0. 13
0.06
0. 10
-0.02
0. 12
0.05
0.25
0. 19
0. 17
Average
d( (03] -


Expt Calc
(ppb/m1r
5.55
4.55
3.15
3.67
4.94
3.12
1 .49
1 .79
1 .56
1 .38
2.43
1 .66
4. 10
3. 46
2.58
4.06
2.66
1 .85
1 .08
1 .79
3.26
3.30
1 .47
3.04
1.41
1 . 40
1 .40
1 .61
2.55
2.83
2.81
1 .51


6.25
4. 15
2.88
2.96
4.48
2.37
1 .96
2.60
2.01
1 .87
3. 14
2.21
3.46
2.97
2.22
3.60
2.54
1 .49
0.92
1 .39
2.99
3.27
1 .35
2. 70
1 .84
2.18
1 .24
1 .73
2.36
3.02
2.71
1 .76


Initial
[NO] )/dt
Calc Calc
-Expt -Expt
i) — /Avg
0.70 0.12
-0.41 -0.09
-0.27 -0.09
-0.71 -0.21
-0.47 -0.10
-0.75 -0.27
0.47 0. 27
0.81 0.37
0.45 0.25
0.49 0.30
0.71 0. 25
0.55 0.29
-0.64 -0.17
-0.50 -0.16
-0.36 -0.15
-0.46 -0.12
-0.12 -0.05
-0.35 -0.21
-0.16 -0.16
-0.40 -0.25
-0.28 -0.09
-0.03 -0.01
-0.13 -0.09
-0. 33 -0.12
0. 43 0.27
0.77 0.43
-0.16 -0.13
0.11 0.07
-0.19 -0.08
0.19 0.07
-0. 10 -0.03
1.00 0.25
0.54 0.19
0.84 0.16
laneoua Surrogates
0.23
0. 19
0.24
0.24
0.23
0.24
0.25
0.23
2.3
3.4
2.0
2.0
2.0
1 .8
2.8
2 . 1
10.0
18.3
8.4
8.5
8.8
7.6
11 .3
9.2
294. 1
286.4
298.8
295.7
296.4
301 .0
297.9
292.6
21900. 0.403
13400. 0.076
25500. 0.506
27500. 0.544
25000. 0.623
23200. 0.541
22300. 0.610
15200. 0.414
0.466
0.357
0.520
0.448
0.476
0.452
0.603
0.414
0.062
0. 281
0.014
-0.096
-0. 147
-0.089
-0.008
0.000
0.14
1 .30
0.03
-0. 19
-0. 27
-U. 18
-0.01
0.00
0.95
0.52
1 . 18
1 . 22
1 .27
1 .48
1 .59
1 .06
1 .06
0.76
1 .08
1.01
1 .07
1 .22
1 .54
0.99
0.10 0.10
0.23 0.36
-0.10 -0.09
-0.21 -0.19
-0.20 -0.17
-0.26 -0.19
-0.05 -0.03
-0.07 -0.07
                                                                                                                     (cont1nued)

-------
            AppencH x  C.
Selected Results

RADM-M Mechanism
                                           of Simulations

                                           (cont1nued).
of Individual Organ1c-N0x-A1r Experiments Using the
to
oo
-J
                                                                                                                        Page  14
Experiment
Initial
Concentrations
NOx HC HC/NOx
(ppm) (ppmC)
JL2081B
ST1682R
JL20B1R
JL2281B
OC1481R
ST1682B
ST2981R
ST29B1B
OC1481B
ST0381B
ST10B1B
JL08B2R
JL0882B
Group Average
S. Dev.
Avg. Abs. Value
S. Oev.
26. UNC "Synurban"
AU22B4R
AU2584R
AU2584B
ST0184R
ST0184B
ST0284R
ST0284B
JN2685R
JN2685B
JN2885R
JN2885B
Group Average
S. Dev.
Avg. Abs. Value
S. Oev.
0.42
0.43
0.41
0.26
0.28
0.43
0.24
0.24
0.29
0.23
0.24
0.29
0.28
0.28
0.08

1 .8
3.2
2.7
2.9
3.3
3. 1
2.5
2.5
2.9
2.0
1 .0
2. 1
2. 1
2.4
0.6

4.3
7.5
6.6
11.2
1 1 .9
7.2
10.3
10.4
9.9
B.6
4. 1
7.3
7.4
9.0
3.0

Ave
Temp
(degK)
305.0*
304.4
305.0*
302.3*
291 .3
304.4
293.7
291 .6
288.9
297.9
295.3
304.2
304.2
297.7
5.8

Ave
H20
(ppm)
20000.*
23000.
20000.*
20000.*
9470.
23000.
18700.
14400.
10300.
16800.
17700.
16000.
16000.
19018.
4902.

Maximum Concentration
OZONE
Expt
(ppm)
0. 165
0.410
0.635
0.722
0.462
0.840
0.294
0.485
0.458
0.611
0.626
0.598
0.541
0.503
0. 175

Calc
(ppm)
0.325
0.508
0.679
0.675
0.569
0.800
0.426
0.494
0.561
0.516
0.382
0.703
0.689
0.527
0. 126

Calc
-Expt
(ppm)
0. 160
0.099
0.043
-0.047
0. 108
-0.040
0.132
0.009
0. 102
-0.095
-0. 244
0. 105
0. 147
0.024
0.121
0.097
0.073
Calc
-Expt
/Avg
0.66
0.22
0.07
-0.07
0.21
-0.05
0.37
0.02
0. 20
-0.17
-0.4B
0.16
0.24
0. 10
0.37
0.24
0.29
Average Initial
d( [03] - [NO] )/dt
Calc
Expt Calc -Expt
(ppb/m1n) --
0.62
1 .28
1 .55
1 .80
1 .53
2.39
0.76
1 .35
1.48
1 .70
1 .36
1 .25
1.12
1 .32
0.40

0.82 0.01
1.17 -0.11
1 .44 -0.11
1.70 -0.10
1.57 0.04
2.08 -0.31
0.88 0.12
1 .30 -0.05
1.51 0.03
1.55 -0.15
0.96 -0.40
1 .35 0.10
1.30 0.18
1.26 -0.06
0.33 0.16
0. 14
0. 10
Calc
-Expt
/Avg
0.01
-0.09
-0.07
-0.06
0.03
-0. 14
0.15
-0.04
0.02
-0.09
-0.34
0.08
0. 15
-0.03
0. 15
0.12
0.10
Surrogate
0.32
0.34
0.33
0.31
0.30
0.34
0.33
0.30
0.30
0.38
0.39
0.33
0.03

0.4
0. 1
0. 1
0.2
0.2
0. 1
0. 1
0.2
0.3
0. i
O.Z
0.2
0. 1

1 .2
0.3
0.4
0.6
0.8
0.4.
0.3
0.6
0.9
0.4
0.5
0.6
0.3

302.7
302.0
302.0
302.5
302.5
304.6
304.5
303.6
303.6
299.8
299.8
302.5
1 .6

8470.
22600.
22600.
9380.
9380.
19900.
19900.
16800.
16800.
1 1800.
1 1800.
15403.
5427.

0.657
0.075
0.096
0.546
0.646
0. 1 19
0.020
0.629
0.788
0.238
0.275
0.372
0.284

0.773
0. 109
0. 135
0.661
0.704
0.265
0. 144
0.677
0.761
0.391
0.278
0.445
0.272

0.116
0.034
0.039
0.115
0.058
0. 146
0. 124
0.049
-0.027
0. 153
0.003
0.074
0.060
0.07B
0.053
0. 16
0.37
0.33
0.19
0.09
0.76
0.07
-0.03
0.49
0.01
0. ?4
O.ilb
0.25
0.24
1.61
0.70
0.69
1 .32
1 .72
0.73
0.40
1 .34
1 .74
0.85
0.87
1 .09
0.47

1.63 0.03
0.64 -0.06
0.66 -0.03
1.31 -0.01
1.58 -0.14
0.76 0.02
0.50 0.10
1 .46 0.12
1 .85 Oil
0.89 0.04
0.88 0.01
1.11 0.02
0.47 0.08
0.06
0.05
0.02
-0.09
-0.05
-0.01
-0.08
0.03
0.22
0.09
0 06
0.05
0.01
0 02
0.09
0.06
0.06
                                                                                                                    (cont1nued)

-------
            Appendix  C.   Selected  Results  of  Simulations  of  Individual  Organ1c-N0x-A«r  Experiments Using the
                         RADM-M  Mechanism  (continued).
                                                                                                                      Page IS
to
00
CD
Experiment
Initial
Concentret
NOx HC
(ppm) (ppmC)
27. UNC "Synauto'
AU04B4R
AU0484B
AU0584R
AU05B4B
AU0684R
AU0684B
AU0784R
AU0884B
AU0984R
ST0884R
STOB84B
ST1784R
ST1784B
ST2184R
ST2184B
Group Average
S. Dev.
Avg. Abs. Value
S. Dev.
Ions
HC/NOx
Ave
Temp
(degK)
Ave
H20
(ppm)
Maximum Concentration
OZONE
Expt
(ppm)
Calc
(ppm)
Calc
-Expt
(ppm)
Calc
-Expt
/Avg
Average
d( [03} -
Initial
{NOJ )/dt
Calc
Expt Calc -Expt
(ppb/m1n) --
Calc
-Expt
/Avg
Surrogate
0.37
0.36
0.36
0.35
0.35
0.35
0.38
0.34
0.39
0.34
0.33
0.34
0.34
0.36
0.36
0.35
0.02


0.4
0.3
0.3
0.4
0.7
1 . 1
0.4
1 . 1
0.4
0.6
1 .0
1 . 1
O.B
0.6
0.7
0.7
0.3


1 .0
0.8
0.7
1.0
2. 1
3.2
1.0
3. 1
0.9
1.9
2.9
3.3
2.3
1 .7
2.0
1 .9
0.9


30/ .S
307.5
307. 1
307.2
307.6
307.6
308.0
307.3
307.8
296.5
298.5
294.4
294.4
302.2
302.2
303.9
5. 1


34900.
34700.
44900.
44900.
36800.
36800.
34700.
29000.
31500.
6920.
6920.
24800.
24800.
7350.
7350.
27089.
13697.


0.515
0.328
0.335
0.595
0.887
0.940
0.602
0.834
0.521
0.566
0.750
0.539
0.484
0.671
0.721
0.619
0. 182


0.646
0.562
0.465
0.656
0.853
0.902
0.672
0.625
0.559
0.751
0.874
0.716
0.705
0.827
0.648
0.724
0.131


0. 133
0.234
0. 130
0.062
-0.035
-0.037
0.070
-0.009
0.039
0. 185
0. 123
0. 177
0.222
0. 157
0. 126
0. 105
0.087
0. 1 16
0.071
0.23
0.53
0.33
0. 10
-0.04
-0.04
0. 1 1
-0.01
0.07
0. 28
0. 15
0.26
0.37
0.21
0. 16
0. 18
0. 16
0. 19
0. 14
1 .45
1 . 16
1 .08
1 .48
2.53
3. 16
1 .61
2.92
1 .51
.27
.90
.59
.36
. 78
2.28
1 .81
0.63


1 .38
1.12
0.98
1.3B
2. 10
2.52
1 .48
2.39
1 . 26
1 . 38
1 .69
2. 26
1 .95
2.01
2.61
1 .77
0.53


-0.06
-0.04
-0.10
-0. 10
-0.43
-0.64
-0.13
-0.52
-0. 25
0.11
-0.21
0.67
0.58
0. 22
0.32
-0.04
0, 37
0.29
0.22
-0.04
-0.04
-0.10
-0.07
-0.18
-0.22
-0.09
-0.20
-0. 18
0. 08
-0.12
0. 35
0. 35
0.12
0. 13
-0.01
0.18
0. 15
0. 10
28. UNC Auto Exhauat
OC0483B
OC0783R
OC0783B
JN2582R
JN2582B
JN2982R
JN2982B
JN3082R
JN3082B
JL0283B
JL0883B
ST2982B
OC0682R
AU1 183R
AU1 183E)
JL0182R
JL0182B
0.25
0.33
0.34
0.65
0.65
0.24
0.25
0.32
0.32
0. 19
0.37
0.39
0.46
0.22
0.23
0.37
0.35
0.4
2.7
2.7
0.6
0.6
2.5
2.5
2.6
0.6
1 .7
1 .7
1 .7
2.0
2.2
0.7
3.5
3.6
1 .7
8.1
7.9
0.9
0.9
10.3
•9.8
8.7
1 .7
9. 1
4.6
4.3
4.3
9.9
3. 1
9.6
10.2
299. 7
295.5
295.4
301 . 1 •
30' . 2*
303. 1
303. 1
302.5*
302.3*
304.0
299.5
295.7
300.2
306.8
306.9
299.5
299.3
18800.
21600.
21500.
22700.
22900.
30500.
30500.
20000.*
20000.*
24500.
21200.
24800.
16900.
21400.
21400.
27700.
27400.
0.642
0. 178
0.451
0.003
0.003
0.704
0.766
0.811
0.840
0.697
0.879
0. 205
0.355
0.850
0.601
0. 740
0. 759
0.719
0.391
0.650
0.013
0.014
0.833
0.839
1 .005
1.011
0.850
0.845
0.057
0. 276
0.890
0.673
1 .038
0.995
0.077
0.213
0. 199
0.010
0.01 1
0. 128
0.073
0. 194
0. 170
0. 152
-0.034
-0. 148
-0.079
0.040
0.072
0.298
0.236
0. 11
0.75
0.36

0. 17
0.09
0.21
0. 18
0. 20
-0.04
-1.13
-0.25
0.05
0. 1 1
0.34
0. 27
1 .86
0.94
1.41
0.64
0. 68
2.32
2. 67
2.55
2. 68
1 .65
1 . 76
1.12
1 .37
2.55
1 . 26
2.31
2.41
1 .97
1.13
1 .93
0.58
0 . 58
2.10
2.18
2.40
2.40
1.72
1 .44
0.76
1.16
2.15
1 . 25
2.40
2.52
0.12
0. 19
0. 52
-0.06
-0.10
-0.22
-0. 50
-0. 15
-0. 28
0.07
-0.32
-0.36
-0.19
-0.39
0.00
0.09
0. 11
0 . 06
0.18
0.31
-0.10
-0.17
-0.10
-0 . 20
-0.06
-0.11
0.04
-0. 20
-0.38
-0. 15
-0.17
0. 00
0.04
0.05
                                                                                                                   (cont1nued)

-------
            Appendix  C.   Selected Results of  Simulations of  Individual  Organlc-NOx-A1r Experiments Using the
                         RADM-M Mechanism (concluded).
Page 16
OJ
00
Exper 1ment

Initial
Concentrations
NOx
(ppm)
AU0382R
AU0382B
ST1782R
ST1782B
ST2982R
OC0682B
JL0283R
JL0883R
JL1583R
JL1583B
OC0483R
Group Average
S. Dev.
Avg. Abs . Value
S. Dev.
29. SAPRC Synthetic
ITC781
ITC784
ITC785
ITC805
ITC795
ITC796
ITC799
ITC801
Group Average
S. Dev.
Avg. Abs. Value
S. Dev.
30. SAPRC Synthetic
ITC963
ITC965
ITC967
ITC968
Group Average
S. Dev.
Avg. Abs . Va 1 ue
S. Dev.
0.44
0. 16
0.26
0.25
0.39
0.46
0. 18
0.37
0.35
0.35
0.25
0.34
0. 12


Jet
0.51
0.50
0. 26
0.52
0.50
0.54
0.51
0.55
0.49
0.09


Jet
0.49
0.46
0. 26
0.49
0.42
0. 1 1


Ave
Temp
Ave
H20
HC HC/NOx
(ppmC)
2.5
1 .2
2.2
2.4
1 .7
2.0
1 .6
1 .7
2.2
2.3
2.6
2.0
0.8


Fuel
43.0
88.0
45.0
98.0
45.0
97.0
94.0
41.0
68.9
27.3


Ex haus »
4.4
5.2
4.4
8.7
5.7
2. 1


(degK)
5.7
7.3
8.6
9.5
4.4
4.3
9.0
4.6
6.2
6.6
10.3
6.5
3. 1



83.5
177.6
170.7
189.5
89.8
178 .9
184. 1
75.1
143.7
50.8



9. 1
11.3
17.2
17.8
13.8
4.3


302.
302.
300.
300.
295.
300.
304.
299.
306.
306.
299.
301 .
3.



303.
303.
303.
303.
303.
303.
303 .
303.
303,
0.



303.
303.
303.
303.
303.
0.


b
6
1
1
7
2
0
5
2
2
7
2
3



0
0
0
0
0
0
0
0
0
0



0
0
0
0
0
0


(ppm)
27400.
27400.
27400.
27400.
24800.
16900.
24500.
71200.
25700.
25700.
18800.
23607.
3802.



16600.
16600.
17900.
17200.
17900.
17200.
17200.
17000.
17200.
499.



19400.
20000.
18300.
19400.
19275.
709.


Maximum Concentration
OZONE
Expt
(ppm)
0.344
0.576
0.562
0.637
0.073
0.450
0.595
0.756
0.866
0.922
0.603
0.567
0.272



0.751
0.746
0.598
0.791
0.761
0.597
0.840
0.881
0.746
0. 102



0.822
0.863
0.586
0.852
0.7B1
0.131


Calc
(ppm)
0.450
0.501
0.590
0.616
0.047
0.328
0.675
0.694
0.737
0.860
0.672
0.617
0.314



0.914
0.906
0.602
0.942
0.794
0.840
0.972
0.957
0.866
0. 122



0.722
0.728
0.591
0.791
0.708
0.084


Calc
-Expt
(ppm)
0. 106
-0.074
0.028
-0.021
-0.026
-0.121
O.OBO
-0.062
-0. 129
-0.062
0.069
0.050
0.119
0. 104
0.074

0. 16:)
0 . 1 60
0.004
0 . 1 c> 1
0.033
0.242
0.131
0.076
0. 1 20
0.078
0. 120
0.078

-0. 100
-0. 135
0.005
-0.061
-0.073
0.060
0.075
0.056
Calc
-Expt
/Avg
0.27
-0. 14
0.05
-0.03
-0.44
-0.31
0.13
-0.09
-0.16
-0.07
0.11
0.03
0.34
0.23
0.24

0.20
0.19
0.01
0.17
0.04
0.34
0.15
0.08
0.15
0.10
0. 15
0. 10

-0.13
-0. 17
0.01
-0.07
-0.09
0.08
0.10
0.07
Average
d( [03] -
Expt
Calc
Initial
[NOJ )/dt
Calc
-Expt
(ppb/m1n) --
1 .26
1 .06
1 .63
1 .72
0.86
1 .49
1 .37
1 .52
2.10
2.20
1 .76
1 .68
0.61



2.77
3.93
2.74
3.27
3.05
4.43
4.59
2.98
3.47
0.74



4.00
5. 10
5.79
1 1 . 76
6.66
3.48


1.23
0.93
1 . 16
1 . 27
0.68
1 .27
1 .37
1 .32
1 .75
1 .99
1.75
1 .55
0.59



7.54
9.86
6.65
9.39
10.78
13.31
8.44
6. 14
9.01
2.35



4. 15
5.28
6. 14
1 1 .08
6.66
3.06


-0.02
-0.14
-0.47
-0.45
-0.18
-0.22
0.00
-0.20
-0.35
-0.21
-0.02
-0.13
0.23
0.21
0. 15

4.77
5.93
3.91
6.12
7.72
8.88
3.86
3. 16
5.54
2.01
5.54
2.01

0.14
0. 18
0.35
-0.68
0.00
0.46
0. 34
0.24
Calc
-Expt
/Avg
-0.02
-0.14
-0.34
-0.30
-0.23
-0. 16
0.00
-0.14
-0. 18
-0. 10
-0.01
-0.09
0. 15
0. 14
0. 10

0.92
0.86
0.83
0.97
1 . 12
1 .00
0.59
0.69
0.87
0. 17
0.87
0.17

0.03
0.03
0.06
-0.06
0.02
0.05
0.05
0.01

-------