United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park NC 27711
EPA-450/4-84-009
May 1 984
Air
Technical
Discussions
Relating To The
Use Of The Carbon
Bond Mechanism
In OZIPM/EKMA

-------
                                    EPA-450/4-84-009

                                             May 1984
 Technical  Discussions Relating To The
Use Of The Carbon  Bond Mechanism In
                 OZIPM/EKMA
                          By

                        J P Killus
                         And
                       G Z. Whitten

                    Systems Application, Inc
                    101 Lucas Valley Road
                    San Rafael, CA 94903
                       Prepared For

                Monitoring And Data Analysis Division
              Office Of Air Quality Planning And Standards
                U S. Environmental Protection Agency
                   Office Of Air And Radiation
              Office Of Air Quality Planning And Standards
                    Research Triangle Park

                        May 1984

-------
This report has been  reviewed by the Office of Air  Quality Planning and Standards, U S,
Environmental Protection Agency, and approved for publication as received from the contractor.
Approval does not signify that the contents necessarily reflect the views and policies of the Agency,
neither  does  mention  of trade names or  commercial products constitute endorsement or
recommendation for use
                                 EPA-450/4-84-009

-------
                                   CONTENTS


Section

    1      INTRODUCTION	      1

    2      USING THE CARBON-BOND MECHANISM	      4

           Speciation of Emissions and Atmospheric Concentrations
           into Bond Categories:  The Total Carbon Principle	      5

           The Volumetric Equivalence Principle	      6

           Surrogate Compounds (Carbonyls)	      7

           Examples of Hydrocarbon Speciation in the CBM-III/EKMA...       7

    3      PARAMETERS AND LUMPED RATE CONSTANTS IN THE CBM-III/EKMA.     25

    4      PHOTOLYSIS RATES	    37

           Photolysis of N02 (kj)	        37

           Carbonyl Photolysis	    40

           Photolysis of Carbonyl Mix	    45

           Photolysis of Hono and Ozone	    46

           Photolysis of DCRB	    46

    5      INITIAL CONDITIONS, EMISSIONS, AND NMHC/NOX RATIOS	      49

           Concentrations of NMHC	    49

           Emissions	    54

    6      BACKGROUND OZONE AND ITS PRECURSORS	    63

           Ozone and NOX	      63

-------
           Reactivity of the Free Troposphere	    64

           Continental Background	    68

REFERENCES	    70

Appendix A:   Recommended Rates for Carbon-Bond Mechanism-III/EKMA
             in OZIPM	    76

Appendix B:   Bond Groups per Molecule	    80
                                       iv

-------
                                   FIGURES
Number                                                                  Page

1    Plot of the rate constant ratio (k  /k   +  k.)  against  carbon
     number in the n-alkane series	       31

2    Product of flux and cross-section for  photolysis  of ozone,
     acetaldehyde, HONO, and N02	       38

3    NOo photolysis (Ki) calculated from the algorithm of
     Demerjian et al., 1980 (solid lines) compared to  photolysis
     derived from UV data (dashed lines) and TSR data
     (dotted 1 i nes )	       41

4    Los Angeles APCD data from seven monitoring stations  for
     1971 through 1975	       52

5    Diurnal variations of area source emissions in
     Philadelphia emissions inventory	       60

6    Diurnal variation in traffic flow in the  Denver
     metropol i tan area	       61

7    Co-mixing ratios in marine air masses  over the Atlantic
     and Pacific oceans	       67

-------
                                    TABLES
Number                                                                   Page

1    Carbon Numbers of Carbon-Bond Groups ...........................       4

2    Carbon-Bond Concentrations Applied to Ambient Hydrocarbon
     Measurements [[[       8

3    Los Angeles Ambient Measurements ...............................      13

4    Example of Light and Heavier Molecular-Weight Hydrocarbon
     Composition in Ambient Los Angeles Air .........................      15

5    Pollutant Concentrations Measured by a Long-Path  Infrared
     Study of Los Angeles Smog.. ....................................      20

6    Ranges of Urban Hydrocarbon Compositions .......................      23

7    OH Abstraction per Alkyl Carbon for Several Aromatic
     Mol ecul es [[[      28

8    Fractional Yields of Alkyl Nitrates in the N0x-Air
     Photooxidation of C2 through Cg n-alkanes ......................      30

9    Nitrogen Dioxide Photolysis rates (k   (k  x 10   s   ) as a
     Function of Solar Zenith Angle .................................       39

10   Formaldehyde Photolysis Ratios to N02 (x 10  ) as a
     Function of Solar Zenith Angle .................................       43
                                                3
11   Acetaldehyde Photolysis Ratios to N02 (x 10  ) as a
     Function of Solar Zenith Angle .................................       44

12   Calculations for Ratios of Carbonyl Photolysis to
     Formaldehyde Photolysis in OZIPM ...............................       47


-------
14   CDT Average NMHC and NOX and Ratio of NMHC to NOX	      56

15   Background of Reactive Hydrocarbons	      65

16   Average Light Hydrocarbon Concentrations Measured
     over the Eastern Pacific	      66

-------
                                 SECTION 1

                                INTRODUCTION


     This document is intended to fulfill several purposes:

     (1)  Recommend the specific form of the Carbon-Bond Mechanism  (CBM-
          III/EKMA) to be used with the Empirical Kinetic Modeling
          Approach (EKMA) to estimate the impact of reducing
          emissions of volatile organic compounds (VOC) and/or oxides of
          nitrogen (NOX) in reducing peak ozone values downwind or within
          a city;
     (2)  Examine the reactive organic composition of urban air in order
          to describe its photochemical  behavior and to recommend specific
          values for C3M-III/EKMA parameters needed to simulate the
          reactivity of the urban composition;
     (3)  Apply information and experience obtained in past-modeling
          applications of the CBM-II to indicate how available emissions,
          air quality and meteorological data are used in the modeling
          procedure necessary for EKMA applications; and
     (4)  Provide one means for Air Pollution Control  Agencies to check
          the completeness and suitability of analyses performed with the
          model to support the need for urban VOC emission  controls.
     The Empirical Kinetic Modeling Approach (EKMA) is a procedure for
using an ozone isopleth diagram to estimate the impact of controlling
urban volatile organic compound (VOC) and/or NOX emissions on peak hourly
ozone concentrations.  Both ambient 0600-to-0900 NMOC/NOX ratios measured
in the urban core and peak hourly ozone measured within or downwind of the
city are needed to apply an isopleth diagram in the EKMA procedure.  Other
information requirements, supplied either by measurements or prior
assumptions, depend on the nature of the atmospheric model and the kinetic
mechanism employed.

     The atmospheric model described in this guidelines document is the
city-specific EKMA.  The city-specific EKMA modeling procedure has also
been referred to as "Level III analysis."  These two terms are
synonymous.  The conventional, city-specific EKMA consists of two
components:  The OZIPP model  and the EKMA procedure (i.e., OZIPP/EKMA)
(EPA 1978 a, b; Whitten and Hogo, 1978).  OZIPP (Ozone Isopleth Plotting

-------
Package) is a computer program that allows the  user  to  plot maximum  hourly
ozone concentrations as an explicit function of  initial  (0800)  ambient
concentrations of nonmethane organic compounds  (NMOC) and  NOX within  the
urban area (EPAc).  The OZIPP computer program  uses  the  Dodge
photochemical mechanism.  Another program--OZIPM (Ozone  Isopleth  Plotting
with Optional Mechanisms)—may use any kinetic mechanism;  this  program  is
used in conjunction with the CBM-III/EKMA.

     The conventional EKMA (OZIPP/EKMA) relies  on  an empirical  analysis of
the reactivity of automobile exhaust as described  by a  surrogate
photochemical mechanism (Dodge/EKMA).  In the validation procedure
underlying OZIPP/EKMA, a series of smog chamber  experiments using
automobile exhaust were simulated using a kinetic  mechanism for a mixture
of propene and butane.  The power of this technique  lies in its use  of
empirical calibration data to adjust the simulation  model.  The
calibration procedure will tend to compensate for  many  uncertainties  or
inadequacies in the photochemical mechanism.  Key  weaknesses  in the
approach are (1) the assumption that all urban  hydrocarbon emissions  are
equivalent to auto exhaust, and (2) the inability  of the mechanism to
adjust for changes in the state of knowledge of  photochemistry  without
extensive recalibration and revalidation.

     The Carbon-Bond Mechanism (CBM) attempts to address these
deficiencies.  The CBM has been validated with  the Bureau  of Mines auto
exhaust data, which was used to validate the Dodge/EKMA  mechanism
(Jeffries, Salmi, and Sexton, 1981).   The validation set  for the
CBM-III/EKMA also includes multicomponent smog  chamber  experiments across
a broad range of hydrocarbon speciation and outdoor  smog chamber
experiments using natural sunlight.  To our knowledge,  the Carbon-Bond
Mechanism is the most extensively validated and  applied  kinetic mechanism
that currently exists.

     Because of the wide range of applications  that  have used the Carbon-
Bond Mechanism, it is possible to draw on a considerable body of  knowledge
for use in atmospheric simulations involving the CBM.   Also,  because  the
CBM-III/EKMA is a condensation of the explicit  chemistries of known
  The version of the Carbon-Bond Mechanism  used  in  the  study  of  Jeffries,
  Salmi, and Sexton (1981) was the CBM-II,  the  immediate  precursor  of  the
  CBM-III, which this document describes.   The  CBM-II  is  also used  in  the
  Systems Applications Airshed Model, which  has  been  used in  many of the
  atmospheric studies from which information  in  this  document is
  derived.  However, the CBM-III is very  similar in structure and
  performance to the CBM-II  (differing  primarily in the technical
  description of aromatic oxidation), and information  used  to support  the
  CBM-II may also be used for the CBM-III/EKMA.

-------
compounds, it is usually possible to relate  specific  parameters  and  rate
constants in the mechanism to precise photochemical phenomena.   Many of
the relevant parameters in the CBM-III/EKMA,  therefore,  are  not  either
empirical, "adjustable" or "tuning" parameters, but instead  are  calculated
directly from the hydrocarbon speciation  in  the urban mix.

     This document is divided into six sections:   Section 2,  examines
hydrocarbon speciation in urban areas; Section 3 calculates  rate constants
and parameters that depend on the hydrocarbon mix; Section 4  presents the
same calculations for photolysis rates; Section 5  describes  ambient  data
and emission inputs into OZIPM; and Section  6, recommends the background
or "clean air" values for ozone, NOX, and hydrocarbons to be  used  in
atmospheric simulations.  With the information supplied  in this  document,
it should be possible to use the OZIPM model  to prepare  a set of city-
specific isopleths, which can then be used in an EKMA analysis.

-------
                                 SECTION 2

                      USING THE CARBON-BOND MECHANISM
     In its current form, the Carbon-Bond Mechanism (CBM-III/EKMA)  treats
the reactions of six types of carbon atoms:  (1) single-bonded carbon
atoms, whose principal constituent is paraffinic carbon molecules  (PAR),
(2) relatively reactive double-bonded carbon (OLE), (3) slow double bonds,
which are almost exclusively ethylene (ETH), (4) reactive aromatic  rings
(ARO), (5) carbonyl compounds such as aldehydes and ketones (CARB), and
(6) highly photolytic a-dicarbonyl compounds such as methyl glyoxal and
biacetyl  (OCRS).  Some other types of carbon atoms can also be treated
within this set.  For instance, highly reactive internal double-bonded
carbon atoms were shown by Whitten, Kill us, and Hogo (1980) to be equiva-
lent to two carbonyls per double bond.  Hence three levels of olefin
reactivity can be treated in the CBM (slow as in ETH, relatively reactive
terminal  olefins as in OLE, and highly reactive internal olefins as in 2
CARB per bond).  Appendix B lists the CBM fractions recommended for a
variety of organics.

     The use of the molecular bond rather than the whole molecule as the
principal  unit may at first seem confusing to those whose experience is
solely with molecular reactions.  However, several major advantages
associated with the bond-group-reaction principle make the conceptual
effort involved worthwhile.  The primary advantage is that the Carbon-Bond
Mechanism does not require the sometimes uncertain calculation of "average
molecular weight."  The carbon number of each carbon-bond group is  fixed
(Table 1):

               TABLE 1.  CARBON NUMBERS OF CARBON-BOND GROUPS
                                          Carbon Number
              Carbon-Bond Group	(carbon atoms per molecule)

                    PAR
                    ETH
                    OLE
                    ARO
                    CARB
                    DCRB
1
2
2
6
1 (plus
3 (plus




1 oxygen atom)
2 oxygen atoms)

-------
     In a lumped molecular mechanism, chemical  reactions  might  be  expected
to alter the average molecular weight of each species category.  When  this
phenomenon occurs, it is impossible to perform  mass-balance  calculations
on the reactive organic compounds remaining  in  the model  simulation.   The
Carbon-Bond Mechanism allows precise hydrocarbon mass-balance calculations
to be made, thus facilitating the estimation of the  importance  of
phenomena like long-range smog precursor transport and  day-to-day  carry-
over of pollutants.  Moreover, whereas most  lumped molecular mechanisms do
not conserve carbon, the Carbon-Bond Mechanism  conserves  carbon and
follows each hydrocarbon fraction to its end products (generally CO  or
C02> but occasionally aerosol or nonreactive hydrocarbons).

     Although the Carbon-Bond Mechanism has  been designed to minimize  the
problems of rate constant averaging and specification of  lumped
parameters, any highly condensed mechanism must contain some parametric
approximations.  The problem of rate constant averaging is somewhat
reduced in the CBM-III/EKMA because the range of reactivities of carbon
bonds is generally smaller than the range of reactivities for molecules.
The parameters that do exist in the CBM-III/EKMA may be related to
specific features of the hydrocarbon mix.  Moreover, in the  design of  the
mechanism, there was an attempt to minimize  the sensitivity  to  the
parameters.

     In the following three sections we discuss three important issues
regarding use of the CBM-III/EKMA.  (1) hydrocarbon  speciation,  (2)
specification of parameters, and (3) specification of photolysis rates.
We first treat the issue of hydrocarbon speciation using  realistic
examples of urban hydrocarbon mixes, in order to use these real  cases  for
the specification of parameters.

SPECIATION OF EMISSIONS AND ATMOSPHERIC CONCENTRATIONS
INTO BOND CATEGORIES:  THE TOTAL CARBON PRINCIPLE

     Several important principles must be considered in the  application of
the Carbon-Bond Mechanism.  First, all carbon must be accounted for.
Thus, if one adds all of the carbon in each  bond category of emissions,
the sum should equal the total carbon emitted.  Although  this principle
appears simple and obvious, there are practical complications.   Emissions
of solvents, for example, are usually given  in  kilograms  of  emissions, but
methyl alcohol (^OH) is a solvent in which most of the  weight is
represented by the oxygen atom in the methanol  molecule.   Another  example
is the case of automobile exhaust emissions, which are  usually  reported  in
gm/mi. of hydrocarbon as methane—i.e., each carbon  atom  measured  is
assumed to have a molecular weight of 16 gm/mole.  Evaporative  emissions,
on the other hand, are reported as straight  mass, which means that a lower
molecular weight is called for.  Accounting  for all  carbon is further
complicated by the fact that the procedures  used to  obtain automobile

-------
exhaust hydrocarbon estimates do not  respond  efficiently to  all  reactive
species.  Aldehydes, for example,  are  not  often measured by  standard
procedures and must be added to the exhaust emissions  estimates.

THE VOLUMETRIC EQUIVALENCE PRINCIPLE

     The second important principle to  remember when  using the  Carbon-Bond
Mechanism is that the volumetric concentrations (in ppm)  of  most  species
used with the CBM are similar to both  the  volumetric measurements and the
molar concentrations used in other mechanisms.  One ppm  of aromatic  hydro-
carbon bonds in the CBM is usually equivalent to  1  ppm of aromatic
hydrocarbons in a lumped mechanism.

     The major exception to the equivalence of  speciation between the CBM
and molecular mechanisms is the PAR species,  which  includes  not  only the
carbon in paraffinic molecules, but also single-bonded carbon  in  other
molecules.  A molecule of propylene,  for example, contains one  single-
bonded carbon in addition to the olefinic  bond:
                 OLE*
PAR,
1 - "
1
1 1
1 C -
1 1
1 H
1

"1
H '
1 '
• C I
1
1
1


H
1
-C-H
1
H


In other words, the CBM total reactive hydrocarbon  (RHC)  given  in  ppmC
must equal

             (OLE  x 2) + (ETH x 2) * (ARO x 6) * CARB

                       + (DCRB x 3) * PAR = RHC (in ppmC)
  This is not true, however, for "surrogate mechanisms"  in  which  all
  hydrocarbons are assumed to be represented by some mixture of surrogate
  hydrocarbons (e.g., propylene and butane).  Comparison of speciation  in
  the CBM with that in such a surrogate mechanism  is obviously impossible.

-------
SURROGATE COMPOUNDS (CARBONYLS)

     A minor exception to the rule of equivalent  speciation  lies  in  the
relationship of CARS as a reaction product to other  species.   Some com-
pounds, especially internal olefins  (e.g., trans-2-butene),  react much
more rapidly than do terminal olefins such as propylene.   Thus,  instead  of
creating a new species with an atmospheric lifetime  of  only  a few minutes,
we chose to treat internal olefins as if they had  already  reacted (i.e.,
as if an internal olefinic bond were already transformed to  two
carbonyls).

     In a prior report (Killus and Whitten, 1982), we suggested  that a
surrogate carbonyl be added when a cycloalkane  is  treated.   This  sugges-
tion was based on the conjecture that cycloalkanes were more reactive than
normal alkanes, due to breakage in the ring structure subsequent  to
reaction with OH.  More recent analysis of smog chamber data for  cyclo-
alkanes indicates that though they may be somewhat more reactive  than n-
alkanes, the difference is insufficient to justify the  addition  of  a
surrogate carbonyl.  Thus, cycloalkanes should  be  treated  as PAR  in  the
CBM-III/EKMA.  The effect of this change on overall  speciation is slight,
since cycloalkanes account for less  than 2 percent of the  total  reactive
carbon in typical urban mixes, and the surrogate  carbonyl  from this  source
is less than 6 percent of total carbonyl (less  than  1 percent overall
reactive organics, ROG).  This change has been  noted in the  speciation
tables in Appendix B.

EXAMPLES OF HYDROCARBON SPECIATION IN THE CBM-III/EKMA

     In this section we perform three example calculations necessary to
separate reactive organics in urban  hydrocarbon samples into the  appro-
priate carbon bond species.  The three examples are  for Los  Angeles  over a
10 year period; the changes in hydrocarbon speciation over this  period  is
an indication of the effect of the increased use  of  unleaded gasoline.   As
we will see, this change resulted in a fairly substantial  increase  in the
aromatics fraction.

Example 1

     Example 1 calculates the carbon-bond concentrations that would  be
used for the ambient hydrocarbon measurements reported  by  Kopczynski et
al. (1972).  Gas chromatographic analysis (GCA) accounted  for 90  percent
of total nonmethane hydrocarbons as  identified  by flame ionization
detection (FID).  Table 2a gives the carbon fraction allocated to each
bond category for each molecular species as calculated  from  the  bond-
splitting information given in Appendix B.  The calculated molar  concen-
tration for each bond group is also  given.  Table 2b gives the sum  of each
bond category as well as the carbon  fraction for  each bond category  for

-------













































































,_
cu
— H
a
§
x
LJ









^^

o\
*~
•"
•

CD

, i
CD

CO
C
N
O
a.
o
— '
Ul

z
LJ
s;
LJ
a:

LO
«r
z:

z
g
a:

CJ
o

o
X
z
LJ

m
y
^

O
*~
S
_j
Q_
a.

tn
z
0
cr
z
Ul
,-rf"
s
o
3E
o
m
z
s
cc
CJ


*
CM
Ul
^£
H-
































-8.
a.

C

CO
c
o
• H
CO
I*

c
CD
U
c§
•o
c

m
1
C
o
J2
eg
U
CO
— •





























CO
c
o
H
+j
CO

J^>
c
CO
o
c

£3

fc-
co
i

























o
'^\
o
CO

U.

c
o
•2
CO
CJ











c
o
j3
i-
a
u
0
U
I-

"S
CO
CO
£






cc
z
CD
°£
U


cc
a.




o
a:
^


LJ
O



X

LU




CC
z



CO
cc
CJ




cc
a.


o
s:
^c



LJ
0


^
t<~
UJ


CJ
•&
Q.






CO
c
U.
cu
o










Q O CO O VO •»
"" •* CD  CM»- VO
CSI «— •?


tA VA
lA fcA
f"» r~










IA IA ^ iA iA iA \o^O\O«£vOvOVO
00 0 0 OOOOOOO







1^1 K>
^O lA O -^ -^ K> K>
C3 0 r^ CD CD CD C3


O

^~


T-O* CM Q0<— *- lACMOCAlACN O
r- • C CU 1 C CU 1
eu o> CD "eu 4- -3 cu T a. "c i f- CD CD
CO'W4CU^; 1^4 13-w.C 1 IJCCUXZ -^
>>o.3j3cj: i i cujjjc lecocc "S
rtJ Ut 1 CO fc- QJ ^ •* 1 1 b* -"H 1 I C C
UJ Q. *— "— ' *- 2. U*— T-CMJJ CJCM^»53
T3
3
C
C
O
CJ












































































8

-------


















































•o
CO
1


c
8


CM
UI
2

>~










cn
c
o
CD

c
CO
o
t—
5
C_J
i-*
ra
—1
£









o
u
CO
t-
L"
c
o
.p
CD







c
o
£
u
CD
0
o
u
TJ

1
TJ
0)
CO
CD
X








CC
z

CO
CC
2


fv*
a.




or
«t

u
o
X
UJ
CC
z
ca
cc
C-3

CC
a.


0
§



UJ
O
UJ

C_3
.a
a.




en
o

. .
CD
O
u

















f-COOCOCOOO CM CD r» IO r-
rO r- CM -A CM *O lA CM CN
CM


1- --t -A 0s OA >o >0 fO I--. CMoocMrg CM K\ f\ co i>
K> r~ CM T~ ^~ V TO
T—








-— i— (3— i*j>-i— a. o
0)— 1-<— l.-HOijZIJ.ICD'-a'.Au J
^^^^x'ct'uji.jl-.UJ.JCM'lo'o' (-0
o _ i i I i i i ca i co - • m
i— UIQ.EOCeC....-lOCO«-«-'—
•D
CO
3
c
c
o
CJ
























































-------











































































•
-o
CD
3
C
^
C
cS


CM
UJ
5
^—










CO
c
o
1-f
^jj
co
^3
c
01
o
c
o
o

^
to
—4
£


















c
o
-H
"o

U
u_

c
0
U
CO
f \

















c
o
J2
U
CO
U
o
^
•O
X

•o
Measure







ce
z
CD
CC

LJ




§
Q.




O
ce



UJ
_j
o


X

UJ



ce


CD
cc
r^\oco^r*- f*> ^ococo^-^O^'tf ^KNCMCMCOCO IA ^j
CM ^ ~ r r- |
O
U
C"
U

ia
CO
H
03

JZ
s
co
TJ
C
3
O
CO lA lA 0
• « • O
0 O 0
tj
o
U
ft
Q.
CO
CMiAOOOOO OOOOOOOOOOOOOiA C

CO
*- to
C C
CO CO
CJ O"1
C 0
o •-
O T3
U- JZ
O
CM
(0 r-
0)
JJ T3
co c
— CD
CO C
CO O
CO u
u a
CD U
* o co \o
*— O iA ^O OJ •— CO COCMCTs\OCOOOvO«— f^CO'ACDO ^O -H
»— »- CM r^ T- r- «-CMCOO
JT
C Si

CM ^ (_ •-<
CM T- CO OJ
— »- Q. £
_^ U
§C • t-
CO -H CO CO
4- C 4) ~*
4) (» C (C C E 3
CCCOBX O-HO
COCOcXCCO ~^4JCO
CO — *- -1 CD JT 4-1-^
ccc>.coxo /-> coco
co a CD cu .c c co — < eo-4 UOE
§03 ) . o o j jg j^ fl] fj ^— jvj |»^ .^ i/\ v^ Q CO | . P^
Cc^^^^COCUXi»lC>^ CO 4-' CawCO
eocoC4jcoa>>->-CECo>.coocccccc-H o coce
CC03*JO)CO4Jj-iXQO4-'Q.>-OOOC)OC)j-i ja C-uCO
4^ U CD 1 CO 1 5^ 1 1 1 •.XlTcOCCCCCC"*
UJCLH-1 C*"1 COCM^ CCMLJf^ CSZ2ZOZ2 Z3 Z3 ZO ^^* * 4-
•D
Cu
3
c
"c
Q
CJ















































































10

-------
TABLE 2.   Concluded.
                            (b)  Carbon-Bond Speciation Category
Species
ETH
OLE
ARO
PAR
CARB
Non-Methane
Nonreactive
Z Molar
Concentrations
(ppb)
75.5
46.4
139.6
2178.0
34.4

302.8
Carbon Fraction
0.042
0.026
0.233
0.606
0.010

0.084
  Gas chromatograph accounted  for  3597 ppbC  (3294 ppbC RHC + 303 ppbC nonreactive).  Flame
  lonization analysis (FIA):   TNMHC = 4.0 ppmC (4000 ppb).  Carbon fractions based on
  3597 ppbC.
                                                 11

-------
the measured hydrocarbon mix.  This  information  could  be  directly input to
OZIPM, a computer program designed to  generate EKMA-type  isopleth diagrams
using any kinetic mechanism.

     Kopczynski et al. (1972) do not report  carbonyl data for  aldehydes or
ketones.  The response of aldehydes  and ketones  to  FID and  GCA is ineffic-
ient.  The carbon fraction shown for the  CARB  species  in  Table 2b consists
exclusively of surrogate carbonyls--compounds  such  as  internal  olefins
(which form carbonyls rapidly); precise carbonyl  data  are lacking.   As
discussed later in this chapter, the additional  carbonyls must be
estimated.

Example 2

     Example 2 (Table 3) also represents  ambient sampling in the  Los
Angeles area (Calvert, 1976).  In this case, however,  the measurements  are
reported in molar units.  To calculate CBM units from  molar units,  the
appropriate bonds-per-molecule factor  (from  Appendix B) is  multiplied by
the molar concentration.

     Calvert (1976) stated that roughly 85 percent  of  total  carbon  atoms
were detected as individual species.  Thus about 0.35  ppmC  remain
unaccounted for in the analysis.  It is probable that  a substantial
fraction of these unmeasured hydrocarbons consist of aromatic  compounds.
This is evident from  the absence of  compounds  such  as  trimethyl benzene
and ethyl toluene, which were observed in the  study described  in  Example
1, and which are known to occur in gasoline  (Mayrsohn  et  al. 1975).  The
unmeasured compounds  are likely to be those  with  a  high molecular weight and
low concentration—another indication that aromatics are  the principal
constituent.  Thus, the aromatic fraction noted  in  Table  3  is  probably  a
lower bound.

Example 3

     Example 3 (Table 4) consists of data obtained  for Los  Angeles  in 1981
and reported by Grosjean et al. (1981).  The reported  concentrations in
Mg/nr have been converted into ppbC  and then split  into carbon-bond
concentrations as in Example 1.  Overall hydrocarbon recovery  in  this data
was very high, over 96 percent.  The most noteworthy feature of the  1981
hydrocarbon data is the high proportion of aromatic hydrocarbons.  As
noted previously, this is probably due to the  increased use of unleaded
gasoline in the period 1975-1980.

     The data of Grosjean et al. also contained  measurements of carbonyl
compounds, although these were measured separately, since carbonyls  are
difficult to measure with gas chromatographs,  and flame ionization
detectors (FIDs)  have a low response to aldehydes and  ketones.  Carbonyls
                                  12

-------
Example 2
TABLE 3. LOS ANGELES AMBIENT MEASUREMENTS. (Source: Calvert, 1976).



Compound
CH,
C2H6
C2H4
C2H2
C3H8
C3H6
iso-C4H10
n-C4H1Q
1-C4H8
ISO-C^Hg
iso-C5H12
n-C5H12
Cyclo-C5H10
1-C5H10
2-Methylbutene
2,3-Dimethylbutane
2-Methylpentane
3-Methylpentane
1-Hexene
n-Hexane
**
Cyclohexane
2,2,3-Trimethylbutane
C6H6
2-Methylhexane
3-Methylhexane
(a)
[RH],ppm
Molar
Basis
2.01
0.049
0.043
0.038
0.037
0.0087
0.012
0.037
0.0015
0.0030
0.0443
0.0162
0.0026
0.004
0.0008
0.0008
0.0110
0.0100
0.0017
0.0100
0.0107
0.0077
0.0082
0.0069
0.0063
Reported in Molar Units

Bonds per Molecule x Concentration
NR OLE ETH PAR ARO CARS

0.078 0.0196
0.043
0.038 0.038
0.0555 0.0555
0.0087 0.0087
0.048
0.148
0.0015 0.0030
0.009 0.003
0.2215
0.0810
0.013
0.004 0.012
0.0024 0.0016
0.0048
0.066
0.06
0.0017 0.0068
0.06
0.0642
0.0539
0.041 0.0082
0.0483
0.0441
    Incorrectly reported as 2,2 dimethylbutene by Calvert  (1976).
    Incorrectly reported as cyclohexene by Calvert (1976).
                                             13
                                                                                   Continued

-------
TABLE 3.   Concluded.
Compound
1-Heptene
n-C?H16
Methyl eye lohexane
2,2,3- and 2,3,3-
Tr imethylpentane
2,2,4-Trimethylpentane
Toluene
1-Methylcyclohexene
2, 2, 5-Trimethylhexane
n-C8H1B
p,m-Xylenes
o-Xylene
n'C9H20
n-PrC6H5
n-C1QH22
n"C11H24
CO
Total
[RH],ppm
Molar
Basis
0.0044
0.0043
0.0037
0.0019
0.0025
0.020
0.0047
0.0010
0.0021
0.0041
0.014
0.0060
0.0013
0.0010
0.0050
0.0011
0.0010
0.0003
1.91

Bonds per Molecule x Concentration
NR OLE ETH PAR ARO CARB
0.0044 0.022
0.0301
0.0259
0.0152
0.02
0.02 0.02
0.0235 0.0094
0.009
0.0168
0.0082 0.0041
0.028 0.014
0.012 0.006
0.0117
0.003 0.001
0.02 0.005
0.011
0.011
0.0036

0.2125 0.0203 0.043 1.367 0.0501 0.014
                      (b)  Carbon Bond Splits  for  Identified Compounds
Compound
NR
OLE
ETH
PAR
ARO
CARB
MHHC
(ppm)
0.2125
0.0203
0.043
1.367
0.0501
0.014
ppmC
0.2125
0.0406
0.086
1.367
0.301
0.014
Carbon Fraction
of CWHC
0.105
0.020
0.043
0.0676
0.149
.007
                      Total  Identified ROG:  2.021
                                         14

-------
TABLE 4.  EXAMPLE OF LIGHT AND HEAVIER MOLECULAR-WEIGHT  HYDROCARBON COMPOSITION
IN AMBIENT LOS ANGELES AIR.  (Source:   Grosjean et  al.,  1981).
     Hydrocarbon
Amount
(pg/m3)
                                             Molar  Concentration  in  CB  Units (ppb)
ppbC
NR
OLE
ETH
PAR
ARO
CARB
Ethane                    13.5     18.9    15.1

Ethylene                  21.8     30.52

Acetylene                 11.1     15.54    7.77

Propane                   17.3     24.22   12.11

Propene                   10.3     14.42            4.81

Propyne

Propadiene

Isobutane

Butane

1-butene                   2.2      3.08            0.77

Isobutene

trans-2-butene

cis-2-butene

Isopentane

Pentane

3-methyl-1-butene          1.8      2.52            0.504

1,3-butadiene               -

1-pentene                  0.9      1.26            0.252
Isoprene                    -

trans-2-pentene

cis-2-pentene               -

2-methyl-2-butene

2,2-dimethylbutane

Cyclopentene

Cyclopentane               3.0      4.2

2,3-dimethylbutane         3.9      5.46

2-methylpentane           14.1     19.74

cis-4-methyl-2-pentene

3-methylpentane           11.9     16.66

2-methyl-1-pentene         4.2      5.88            0.98
                                  15.26
24.2
53.2
2.2
6.6
5.9
48.4
32.1
1.8
33.88
74.48
3.08
9.24
8.26
67.76
44.94
2.52
                                 3.8


                                 7.77
                                12.11
                                 4.81
                                          33.88

                                          74.48

                                           1.54

                                           6.93

                                           4.13


                                          67.76

                                          44.94

                                           1.51


                                           0.756
                                           4.2

                                           5.46

                                          19.74


                                          16.66

                                           3.92
                                                  2.31
                                                  4.13
                                                                                Continued
                                                15

-------
TABLE 4.   Continued.
Hydrocarbon
Hexane
trans-2-hexene
2 -met hyl -2-pentene
cis-2-hexene
Methylcyclopentane
2,2,3-trimethylbutane
2,4-dimethylpentane
1-methylcyclopentene
Benzene
Cyclohexane
2-methylhexane
2,3-dimethylpentane
3-methylhexane
Dimethylcyclopentanes
2, 2,4-trunethylpentane
Heptane
Met hyl eye lohexane
Ethylcyclopentane
2,5-dimethylhexane
2 , 4-dimet hyl hexane
2,3,4-tnmethylpentane
Toluene
2,3-dimethylhexane
2-methylheptane
3-methylheptane
2,2,5-trunethylhexane
Dimethyl eye lohexane
Octane
Ethylcyclohexane
Ethylbenzene
p- and m-xylene
St yr ene
Amount
(ug/m?)
9.3
-
-
-
10.6
-
2.7
-
10.4
4.6
4.7
7.3
7.3
5.0
5.6
5.0
5.4
-
-
3.1
2.0
138.7
0.9
4.8
2.7
-
-
2.7
1.0
24.9
109.1*
1.7

ppbC
13.02
-
-
-
14.84
-
3.78
-
14.56
6.44
6.58
10.22
10.22
7.0
7.84
7.0
7.56


4.34
2.8
194.18
1.26
6.72
3.78
-
-
3.78
1.4
34.86
152.74
3.19
Molar Concentration in CB Units (ppb)
NR OLE ETH PAR ARO CARB
13.02



14.84

3.78

12.1 2.43
6.44
6.58
10.22
10.22
7.0
7.84
7.0
7.56


4.34
2.8
27.74 27.74
1.26
6.72
3.78


3.78
1.4
8.72 4.36
38.18 19.09
0.40 0.40 0.40
                                                                                 Continued
                                              16

-------
 TABLE  4.  Concluded.
Hydrocarbon
o-xylene
Nonane
Amount
(ug/m3)
28.3
1.7

ppbC
39.62
2.38
Molar Concentration in
NR OLE ETH


CB Units (ppb)
PAR ARO
9.9 4.95
2.38

CARB


 Isopr op yl benzene

 Propylbenzene

 p-ethyltoluene

 m-ethyltoluene

 1,3,5-trimethylbenzene

 o-ethyltoluene

 tert-butylbenzene

 1,2,4-trimethylbenzene

 secbutylbenzene

 1,2,3-trunethylbenzene

 Decane

 Methylstyrene ***

 1,3-diethylbenzene

 1,4-diethylbenzene

 1,2-diethylbenzene

 Undecane

 Dodecane

 a-pinene

 B-pinene

 Litnonene

 Myrcene

                     Total

           Carbon Fraction
3.7
5.8
2.4
2.7
3.1
9.9
2.2
1.8
1.6
0.8
2.7
2.6
5.18
8.12
3.36
3.78
4.34
13.86
3.08
2.52
2.24
1.49
3.78
3.64
                 0.083
1.73
2.7
1.12
1.26
1.45
4.62
1.23
0.84
2.24
0.165
1.51
0.58
0.90
0.37
0.42
0.48
1.54
0.31
0.28

0.16
0.37!
                                 1.456   0.364
1016.46 47.08   7.40   15.26
           0.05   0.015   0.03
543.0  62.33   7.0

 0.53  0.368   0.01
                (+0.034
               measured
              carbonyls;
              see text)
    Assumed  split:  p-xylene  -  36 ppb; m-xylene  73 ppb
    27.1  ugm/m  unidentified compounds
    identified compounds > 0.96 total IUHC.
*** Assumed split:   50% A-methylstyrene, 50%  B-methylstyrene
                                                 17

-------
reported by Grosjean et  al. were  C^  to  C^  aldehydes  plus acetone and
benzaldehyde.  Data indicate a 22.4  ppb  formaldehyde concentration  and
35.83 ppb of total carbonyls.  This  amounts  to  a 0.034  fraction  of
measured carbonyls which  should be added to  the  0.01 measured  surrogate
carbonyl fraction in Table 4.  This  addition will  cause a summnation of
carbon fractions slightly greater than  1.0 because carbonyls  are not
measured by GC or FID  instruments.   It  should also be noted  that Grosjean
(1982) reported a variety of unidentified  carbonyl compounds  in  Los
Angeles measurements.  The 0.044  carbon  fraction noted  in Table  3 is
therefore a lower limit.

Carbonyl Compounds

     In the first two  examples presented in  this section, only surrogate
carbonyls could be included in the speciations because  carbonyl  compounds
per se (aldehydes and  ketones) could  not be  detected by the  instrumenta-
tion that was used.  Yet carbonyl compounds  were undeniably  present and
they are significant contributors to  smog  chemistry. Therefore, some
estimates of carbonyl  emissions must  be made.

     In addition to their role in the radical  initiation of  the  smog
process, a significant fraction of hydrocarbon reactivity results from the
oxidation of carbonyl  compounds by the  hydroxyl  radical.  Kopczynski,
Kuntz, and Bufalini (1975) compared  the reactivity of an urban air  sample
with that of a surrogate  laboratory  smog mixture containing  pure hydro-
carbons, and found the reactivity for the  urban  air  sample to  be about 40
percent greater than that of the  surrogate.   Those authors suggested that
various aldehyde species were contributing to the  reactivity of  the urban
sample.  Killus and Whitten (1982) calculated that a carbonyl  fraction of
0.09 was necessary to  explain the additional  reactivity.  However,  the
emissions fraction of .carbonyl groups might  be less  than this, since some
of the excess reactivity would be due to carbonyls produced by previous
photochemical reaction.

     Dimitriades and Wesson (1972) reviewed  available information concern-
ing the relative levels of aldehydes  and hydrocarbons found  in automobile
exhaust.  In tests performed by the  U.S. Bureau  of Mines (Sawicki,
Stanley, and Elbert, 1961; Dimitriades  and Wesson, 1972),  a mole fraction
of total aldehyde per  mole of hydrocarbon was calculated to range from
0.06 to 0.09.  Other studies indicated greater variation, with mole
fractions of aldehyde  ranging from 0.07 to 0.35  (Oberdorfer, 1967)  and
  The term "reactivity" has acquired a variety of meanings  in  smog
  chemistry.  In this context we define it as the oxidative  production  of
  peroxyl radicals, a process which then effects a conversion  of  NO  to
  N02.
                                    18

-------
from 0.12 to 0.20 (Wadowski and Weaver, 1970).  Dimitriades and Wesson
(1972) concluded that total aldehyde levels in pre-1970 auto exhaust
represented about 10 percent of total hydrocarbon on a molar basis and
5 percent on a carbon basis (the aldehydes being about 60 percent formald-
ehyde on a molar basis).

     Altshuller and McPherson (1963) measured ambient concentrations of
formaldehyde and acrolein.  Acrolein was found to make up 10 percent to 15
percent of the concentration of formaldehyde, thus indicating the probable
importance of carbonyl  species other than simple aldehydes.  Seizinger and
Dimitriades (1972) identified numerous carbonyl  compounds in automobile
exhaust — notably acrolein, acetone, and the aromatic aldehydes.  Although
Altshuller and McPherson (1963) did not report aldehydes as a fraction of
reactive hydrocarbon, the formaldehyde concentrations observed (0.01 to
0.115 ppm) were consistent with the 5 percent carbon fraction suggested by
Dimitriades and Wesson (1972).

      In an analysis of monitoring data for the Los Angeles area (Scott
Research Laboratories,  1970), Killus et al . (1980) concluded that aldehyde
emissions were similar to olefin emissions when calculated on a molar
basis.  If this assumption is made, the data given in this section
indicate aldehyde emissions that range from 0.021 to 0.031 as a fraction
of emitted reactive carbon.

     Data are very sparse for emissions from vehicles having pollution-
control devices.  However, in a review of recent data, Bui on, Malko, and
Taback (1978) found no major differences in the relative formaldehyde
emissions from controlled and uncontrolled vehicles.  Reported emission
levels were fairly low in these studies:  approximately 2 to 3 percent of
total RHC.  Note that total aldehyde emissions would be expected to be
higher and total carbonyl s higher still.

     Hanst, Wong, and Bragin (1982) conducted a long-path infrared
measurement study of Los Angeles smog.  The study is noteworthy in several
respects.  One of the infrared absorb ti on spectra used in the study was
the  so-called C-H band, which is a good measure of alkyl carbon, i.e., the
measurement corresponds almost precisely to PAR in the Carbon-Bond
Mechanism.  Total NMHC can be estimated from PAR  (assuming an additional
proportion of olefinic and aromatic bonds) and other measured compounds
can  be expressed as a fraction of total NMHC.  This  is done in Table 5 for
formaldehyde and ethene
     The relative fraction of formaldehyde to NMHC would be expected  to
increase during the daylight hours as photochemistry produces formaldehyde
from other compounds.  As can be seen in Table 5, this is what is
observed.  Similarly, ethene is destroyed by photochemical reaction and
its relative fraction decreased during the day.
                                    19

-------
TABLE 5.  POLLUTANT CONCENTRATIONS MEASURED BY A LONG-PATH INFRARED
                        STUDY  OF  LOS  ANGELES  SMOG
Time
0635
0710
0720
0810
0840
0900
0930
1010
1030
1100
1155
1220
1315
1345
1420
1445
1510
1535
1620
1645
1705
1735
1840
1905
1945
2010
2045
2115
2145
2300
0045
0110
0245
0305
0445
0505
0605
NMPC*
(ppbC)
870

1090
780


720

720


720
700

900

740


750

620
500

530

760

980


980
1210


720
660
H2CO
(ppb)
29

33
33


40

37


42
42

56

42


37

32
34

25

26

24


26
33


34
36
H2CO/NMHC**
0.024

0.022
0.030


0.040

0.037


0.042
0.043

0.044

0.040


0.035

0.037
0.049

0.034

0.024

0.018


0.019
0.020


0.034
0.039
^2^4
(ppb)
31
31


17
20

20


19


24

15

12
10

9
7

13

16

23

25
21


23



C2H4/NMHC**

0.043


0.032
0.04

0.04


0.038


0.043

0.041

0.023
0.019

0.019
0.016

0.036

0.036

0.040

0.036
0.031


0.034



                                                            Continued
                               20

-------
TABLE 5.  Concluded.
Time
0630
0710
0815
0840
0903
0930
0950
1020
1045
1215 p.m.
1240
NMPC
(ppbC)

1640
1610

1440

1320

1210

1020
H2CO
(ppb)

47
49

59

58

48

48
H2CO/NMHC**

0.02
0.022

0.029

0.031

0.028

0.034
C2H4
(ppb)
26
63

38

33

27

20

C2H4/NMHC**
0.032
0.055

0.036

0.034

0.030

0.027

    Nonmethane paraffinic carbon.  The CPIR system used by Hanst, Wong,
    and Bragin measures -CH2 and -CHg groups which correspond almost
    exactly to PAR in the Carbon-Bond Mechanism (some unreactive
    hydrocarbons such as ethane and propane would be included in NMPC).

    Nonmethane hydrocarbon is estimated as NMPC x 1.4, assuming an
    additional loading of aromatic and olefinic carbon.  Ratios are
    expressed as relative carbon fraction.
                                 21

-------
     The relative fraction of emissions for HCHO and ethene, may best be
estimated from the nighttime values.  Thus, formaldehyde seems to be about
2 percent of emissions (as carbon) and ethene is 3 to 4 percent of NMHC
(as carbon).  The ethene fraction is completely consistent with the three
examples noted in this chapter; the formaldehyde fraction corroborates the
measurements of Grosjean et al. (Example 3)

     Given the information presented in this chapter we may prepare an
outline of plausible carbonyl emissions in an urban emissions inventory.
Formaldehyde emissions alone account for perhaps 2 to 6 percent of the
carbon emitted in automobile exhaust; however, formaldehyde would account
for only 1 to 4 percent of total emitted reactive carbon, since other
emission processes (e.g., evaporation) seldom emit aldehyde per se.

     Adding other aldehydes to formaldehyde increases our estimate of
carbonyl emissions by approximately 50 percent (since formaldehyde
represents 60 percent of aldehyde emissions on a molar basis).  The
addition of other carbonyl compounds (e.g., acetone, acrolein, and
benzaldehyde) increases the carbonyl emissions rate still further, to
perhaps twice that of emitted formaldehyde.  Finally, surrogate carbonyl
in the CBM accounts for 1 to 2 percent of emitted carbon.

     In the CBM, carbonyl emissions as a fraction of total  reactive carbon
emissions would be approximately 5 percent, which is in agreement with
the assumptions used in previous EKMA studies.  With the onset of photo-
chemical smog formation,  the carbonyl  fraction increases because of
oxidation of reactive hydrocarbons to aldehydes, ketones, glyoxals, and
so forth.  This process eventually reaches a photochemical  equilibrium in
which carbonyl  carbon can represent as much as 15-25 percent of reactive
carbon.

Representativeness of the Los Angeles Area

     Because the three examples given in this section are all  taken from
Los Angeles, the question arises as to whether the hydrocarbon speciation
found in Los Angeles is representative of urban hydrocarbons in general.
The available evidence suggests that urban hydrocarbons do not vary widely
in reactivity or composition.  In a survey of hydrocarbon compositions
measured in different urban areas, the range of hydrocarbons was found to
be fairly small  (see Table 6).*  Since the observed variations in composi-
tion also reflect measurement variance, the actual  city-to-city
variability is  likely to  be even less than that reflected in Table 6.

     When compared to the hydrocarbon  composition found in  other urban
areas,  the Los  Angeles 1981 speciation does appear to be somewhat depleted
* Killus and Whitten, 1982.
                                      22

-------
TABLE 6.  RANGES OF URBAN HYDROCARBON COMPOSITIONS
(Source:  Killus and Whitten, 1982).
Bond Type
ETH
OLE
ARO
PAR
CARB
Un reactive
Nonmethane
Hydrocarbons
Range
0.02
0.02
0.10
0.5
0.03
0.05
- 0.11
- 0.07
- 0.4
- 0.70
- 0.10
- 0.22
Recommended
Value
0.04
0.03
0.19
0.58
0.05
0.15
  Killus and Whitten  (1982) reported the speciation
  as normalized to reactive hydrocarbons (RHC).
  The data in this table have been modified to reflect
  the inclusion of the unreactive hydrocarbon components
                     23

-------
in olefins and somewhat enriched  in  aromatics.  This  feature  may be traced
to the effects of rules instigated to regulate  the  reactivity and composi-
tion of gasoline in southern  California.   For example,  the  bromine number
of gasoline sold in the South Coast  Air Basin may not exceed  30..   The
result of this and other measures to reduce  the olefin  content of emitted
hydrocarbons in Los Angeles has been to increase the  aromatic fraction as
well.  Apart from this difference, it is  probable that,  nationwide, the
composition of urban hydrocarbons is similar to compositions  for Los
Angeles noted in this chapter.  In any case, in the absence of an exten-
sive study to determine the composition of hydrocarbons  in  a  particular
area, the default values noted  in Table 5  should be used.   The default
composition is suggested as a good first estimate for any U.S.  city.
Kill us and Whitten (1982) show  how hydrocarbon  splits from  several  cities
fall within the range recommended in Table 5.
£
  The bromine number corresponds to the number of grams of bromine  (B^)
  that will react with 100 grams of gasoline.  The test is exactly
  specific to the proportion of olefinic bonds in the gasoline  (although
  it cannot distinguish between internal and terminal olefins).  A  bromine
  number of 30 corresponds to an olefin bond fraction of 0.054.  In
  practice, the bromine number of gasoline in Los Angeles has been  reduced
  to 6-14, corresponding to an olefin fraction of 0.01 to 0.025.  Note
  that ethene is not a constituent of gasoline, but  is instead  a
  combustion product and is emitted from certain industrial processes.
  Ethene is also a plant hormone, i.e., it is emitted biogenically.
                                    24

-------
                                   SECTION 3

           PARAMETERS AND LUMPED RATE CONSTANTS  IN THE  CBM-III/EKMA


     As noted previously, any generalized mechanism  compact  enough  to  be  used
in atmospheric simulation models will require  some averaging  of  reaction  rate
constants and will probably  include  some  parametric  assumptions.  The  CBM-
III/EKMA contains the following reactions that depend on  hydrocarbon
composition and assumptions  about the physical state  (e.g.,  surface to volume
ratio and surface type, spectrum of  incident light)  of  the smog  system:

     >  Rate constant for the heterogeneous reaction  of ^5  with water
        (parameterized in Reaction 12)

     >  Lumped rate constants for radical-hydrocarbon reactions  (Reactions  16-
        25, 34-36, 56, 57, 66, 70)

     >  Nitrate formation rate for alkylperoxy radical  with  NO (Reaction  31)

     >  Alkoxy decomposition parameter (A-factor, Reactions  29,  30)

     >  Rates of formation of phenolic compounds relative to  ring opening
        (Reactions 56, 57)

     >  Yield of alkylated a-dicarbonyls  from  aromatics (Reactions  60-61)

     >  Photolysis rates, especially for carbonyls (Reactions 36, 37)

     In this section, we describe how these rate constants and parameters can
be estimated and recommend values on the basis of hydrocarbon composition data
given in the previous section.


Decay of NO^and Ozone

     Reaction 12 in the CBM-III/EKMA is a parameterization of the more compli-
cated reaction scheme:
                                                                      (3-1)
                                       25

-------
                     N2°5 * N°3 + N02                                 (3"2)

                     N205 * H20 surface> 2 HN03                       (3-3)


     The species ^Og is assumed to be in steady  state  and the  rate  constant
for Reaction 12 is computed on that basis.  The daytime dynamics of  ^05
formation and destruction depart from the steady-state  approximation at high
concentrations of NC^ and 03 (NC^ and 03 > 0.3 ppm).  However,  this  circum-
stance is rare.  Also, ^0$ may not be correctly  treated  in  steady state  at
night because its steady-state concentration might  exceed the NOX concentra-
tion.

     Because Reaction 3-3 is surface-dependent, its estimation  for atmospheric
circumstances is difficult.  Ideally, the destruction of  ^Oc in this manner
should be treated as a surface deposition.  This  is beyond the  current
capability of OZIPM/EKMA.  However, a simple first-order  loss,  parameterized
in Reaction 12, is acceptable because its effects on daytime  simulation
results is fairly minor in the atmosphere.

     Platt et al. (1980) measured the lifetime of the N03 radical in the  Los
Angeles atmosphere at night.  From these measurements,  it is  possible to  place
upper limit values on Reaction 3-3 and Reaction 12  in the CBM-III/EKMA.   The
daytime value for Reaction 3-3 is likely to be less than  the  value at night
due to a reduction in the surface-to-volume ratio with  the daytime increase  in
mixing depth.  The maximum lifetime of the NQ3 radical  reported by Platt  et
al. was approximately one minute, from which a value of 2.25  x  10"°  ppm"^
min~l was derived for Reaction 3-3, or a value of 1.3 x 10"^  ppm~^ min'l  for
the pseudo-third-order Reaction 12.

     The water can also be incorporated into the  reaction and rate constant,
making a pseudo-second-order reaction of N03 with fK^.  To do this,  we
recommend a water level of 20,000 ppm, which corresponds  to  a. dew point of
about 65 F, or 50 percent relative humidity, at 85  F.   The pseudo-second-order
rate constant for Reaction 12 would then be 26.0  ppm~l
Paraffins--OH Rate Constant
     In a typical urban mix of hydrocarbons,  approximately  80  percent  of  the
total alkyl groups are contained  in  alkane  (paraffin) molecules  (see Section
2).  The work of Greiner  (1970) suggests  that  the  reactivity of  most alkanes
(C > 3 apart from cyclopropane, cyclobutane,  and molecules  containing  a
                                        26

-------
tertiary-butyl group) can  be  described  in  terms of the number of primary,
secondary, and tertiary C-H bonds.   Darnall  et  al. (1978)  and Atkinson et al.
(1979) suggest a modification of  Greiner's formula:


     KQH = 1.01 x 1(T12 N]_ e823/T +  2.41 x 10'12 N2  e"428/T


         + 2.1 x 10~12N3 cm^  molecule'lsec'l


where Ni, No, and Ng are the  number  of  primary, secondary,  and tertiary C-H
bonds.  At 298 K, this equation becomes


                    283 Cp *  1696 Cs +  3110  Ct  ppm^min"1,


where CD, Cs, and C^. are the  number  of  primary, secondary,  and tertiary
carbons.

     Because the creation  of  a tertiary carbon  involves a branching that must
generally terminate in a primary  carbon group,  the effect  of branching on
hydroxyl reactivity is negligible in the modified  Greiner  formulation:

                               2CS a  Cp  t Ct     .

Thus, the overall OH-reactivity per  carbon atom is similar  for most alkanes,
increasing only gradually with increasing  carbon number.

     The average carbon number for paraffins in the  three  example mixes given
in Section 2 may be calculated as 5.3,  5.1,  and 5.5,  for  examples 1,  2, and 3,
respectively.  From the modified  Greiner formula,  we  may  calculate reaction
rate constants with hydroxyl  to be 1162 ppnT1 min"*,  1142  ppnT1 min'1, and
1182 ppm'l min"! per carbon atom  for the paraffin  molecules in these  three
mixes.

     Roughly 20 percent of the alkyl  carbon  in  the three  example mixes is to
be found in nonparaffin molecules, primarily aromatics.  The shorter  alkyl
carbon number in these molecules  reduces the average  alkyl  carbon number for
the three mixes to 4.4, 4.0,  and  4.8, respectively.   If aromatic alkyl carbon
followed the modified Greiner  formula,  this  would  reduce  the average  OH-PAR
rate constant.  However, the  aromatic ring appears to activate the alkyl group
to H atom abstraction (see Table  7).  Overall,  the aromatic linked alkyl
carbon seems to be slightly more  reactive  than  the average  paraffin mole-
cule.  The calculated average  OH-PAR rate  constant for all  three mixes is very
close to 1200 ppm"-'- min   and we  recommend that value.
                                       27

-------
TABLE 7.  OH ABSTRACTION PER ALKYL CARBON FOR
SEVERAL AROMATIC MOLECULES.  (Calculated from
data contained in Atkinson et al., 1979).
                           OH Abstraction Rate
                           per Alkyl Carbon Atom
       Compound               (ppm   min'1)
Toluene                             1536
Ethyl benzene                        1593
o-xylene                            2145
m-xylene                             720
p-xylene                             803
1,2,3-trimethylbenzene               583
1,2,4-trimethylbenzene               600
1,3,5-trimethylbenzene               624
                     28

-------
Paraffins—Nitrate Formation
     Long-chain paraffin molecules undergo a nitrate formation reaction
(Reaction 31) that depends on the length of the alky! chain (see Table 3,
Figure 1).  The distinction must be made here between alky! chain length and
carbon number.  A branched hydrocarbon such as 2,3-dimethyl butane seems to
have a nitrate formation yield more similar to n-butane than to n-hexane
(private communication, Carter, 1983).

     For this reason we base our estimate of the relevant  rate constant on the
alky! chain length of our typical urban mixes.  For our examples the calcu-
lated alkyl chain is between 4.0 and 4.4.  The ratio of Reaction 31 to the sum
of Reactions 29 to 31 would, according, to Atkinson et al.  (1982), be 0.077 to
0.1, for R0« radicals (modeled as Me02 in the CBM-III/EKMA) derived from
paraffins.  Since some Me02 radicals are derived from aldehyde oxidation,
which results in short alkyl chains, we recommend the lower value of 900
ppm"1 min"1 for the rate constant of Reaction 31.  This number also provides
simulated nitrate levels close to measured values in smog  chamber experiments
using urban mixes.
Alkoxy Radical Decomposition and Isomerization (A-factor)

     When an alkyl peroxy radical (R0«) reacts with NO it generally forms N02
plus an alkoxy radical (RO-):

                             RO. + NO > NO  + RO.


The alkoxy radical may then react in a number of ways, either with oxygen to
form aldehydes or ketones (CARB) only, or to reform an RO- radical by a number
of pathways.  For a discussion of alkoxyl chemistry, the reader is referred to
Whitten, Killus, and Hogo (1980).  In the CBM-III/EKMA, the RO- radical is
treated implicitly.  Alkoxy radical  reaction with oxygen is treated in
Reaction 29.  Alkylperoxy reformation (through decomposition or
isomerization), is treated in Reaction 30.

     The total number of alkyl peroxy radicals formed per reaction with OH is
termed parameter or factor "A" in the CBM-III/EKMA:
               K(29) = [12000 - K(31)]
                                       29

-------
TABLE 8.  Fractional yields  of alkyl  nitrates in the NO  -air
photooxidation of  t? through Cn n-alkanes.             x
n-Alkane
Ethane
Propane
n-Butane
n-Pentane
n-Hexane
n-Heptane
n-Octane
A [alkyl nitrate]/-A [n-alkane]a)
0.010
0.036
0.077
0.128
0.22
0.31
0.33
  Corrected  for loss processes of alkyl nitrates  due to reaction with OH
  radicals


   Source:   Atkinson et al.  (1982)
                               30

-------
0.5 r-
O.I  ~
0.0
                           n  in  CnH2n+2
    Source: Atkinson et al. (1982)
     FIGURE 1.  Plot of the rate constant ratio k /(k
     against carbon number in the n-alkane series.
     (k  is the rate constant
     ROo + NO reactions and L
     alRoxy radicals and NO,).
                           kh)
for nitrate formation from
 is the rate constant to
                           31

-------
               K(29) * K(30) * K(31) = 1200°


     For a typical mix of hydrocarbons, A was determined  to  be  1.5,  based
on experimental evidence and known detailed chemistry  (Killus and  Whitten,
1980).  Since a value of 900 is recommended for  K(3j\,  K/og) has  a
recommended value of 3700 ppm~* min"^ and !<3g =  7400 ppm  * min   .
Olefins

     The radical reaction rate constants for monoalkylated  (terminal)  olefin
bonds are nearly invariant for most of the compounds  in which  they  occur  (see,
for example, Ohta, 1983).  Thus, the rate constants for olefins  in  the CBM-
III/EKMA, which are derived from propene, are expected to apply  to  all  mono-
alkylated olefins (1-olefins).
Aromatics

OH Reaction Rate--

     Hydroxyl reaction with aromatic rings has two  product  channels  in  the
CBM-III/EKMA.  Reaction 60 leads to phenolic compounds  such as  cresols;  it
also mimics to some degree the formation of phenols from  subsequent  reactions
of aromatic aldehydes.  Otherwise, formation of  the aromatic  aldehydes  are
treated in alkyl group (PAR) chemistry.  The other  reaction channel  (Reaction
61) leads to opening of the aromatic ring and to  prompt formation  of alkylated
a-dicarbonyl compounds such as methyl glyoxal.

     The rate constant averaging problem in aromatics chemistry is more  severe
than for olefins and alkyl groups.  The rate constant varies  by a  factor of 10
between monoalkylated benzenes such as toluene and  the  most reactive trimethyl
benzenes (e.g., the 1, 3, 5 isomer).  Fortunately,  most of  the  aromatics have
OH rate constants that are within a factor of 2  of  the  average  value.

     As we note in Section 2, atmospheric measurements  tend to  neglect
compounds that exist at low concentrations, but  that can  nevertheless con-
stitute an appreciable fraction of total hydrocarbon reactivity.   This  is
especially true for higher molecular weight compounds such  as aromatics,
particularly the di- and  tri-alkylated compounds  for which  many isomers
exist.  Since these compounds are of greater than average reactivity, compound
reactivity averages calculated from ambient data  will tend  to underpredict the
true average reactivity.
                                        32

-------
     This underprediction problem  is  especially  evident  in Example 2 (Section
2), which does not contain data for trimethylbenzenes  or  ethyltoluenes,
compounds known to occur in  urban  emissions.   Accordingly, we have used only
Examples 1 and 3 to calculate average OH  rate  constants  for aromatics.   Using
values derived from Atkinson et al. (1979),  for  the  individual  compounds, and
subtracting the contribution of H  atom  abstraction from  the alky!  sub-
stituents, we obtain values  of 22,500 ppm~*  min"^  and  18,150 ppm~* min~* for
Examples 1 and 3, respectively.  These  numbers correspond to the sum of
Reactions 56 and 57 in the CBM-III/EKMA.

     The formation pathway ratios  of  phenolics to  ring opening  in  Kill us and
Whitten (1982) are based on  toluene and m-xylene.  The formation of phenolics
from toluene is much greater than  for the  xylenes  (possibly excepting o-
xylene) and is probably greater than  for  most  other  aromatics.   Thus, we
reduce the phenolic pathway  fraction  somewhat  for  the  aromatic  mixture.  The
recommended value for Reaction 56  is  6,000 ppm"-*- min   ,  and for Reaction 57 is
14,500 ppm'l rain'*, which sum to the  approximate average  aromatic  reactivity
for Examples 1 and 2.
Yield of Alkyl-Substituted a-dicarbonyls--

     Ring opening in alky! benzene photooxidation  systems  generally leads to
at least one alkyl-substituted a-dicarbonyl  (e.g.,  methyl  glyoxal).  For
monoalkylated benzenes such as toluene, only  a  single  alkyl  dicarbonyl  may
result.  For di- and trialkylated aromatics  a greater  yield  is possible.

     There is currently some question  as to whether  the  formation  of these
additional <*-dicarbonyl products is  prompt or is delayed by  the formation of
an intermediate unsaturated product, e.g., an olefinic dialdehyde  (see
Atkinson et al., 1980; Killus and Whitten, 1982).   Recent  smog chamber  studies
(Bessemer 1982; C. Spicer, private communication)  indicate that the yield of
the dial products is low.  Whitten,  Kill us,  and Johnson  (1983) suggest  that
the formation of the additional alkyl  a-dicarbonyl  products  is prompt—
consistent with the observations of  a  low dial  yield.

     Additional products from aromatic oxidation in  the  CBM-III/EKMA are
represented by the species APRC (Aromatic Product  Carbon).   For monoalkylated
benzenes, the product of APRC is two glyoxal  molecules,  which  are  represented
in the CBM-III/EKMA as CARB + CO.  Dialkylated  benzenes  yield  glyoxal plus an
alkyl a-dicarbonyl (DCRB), while trialkylated benzenes give  two alkyl
dicarbonyls or one glyoxal plus a dialkyl dicarbonyl (e.g.,  biacetyl).   The
formation of dialkyl dicarbonyls is  a  relatively minor pathway, however.

     The ratio of APRC products (Reactions 60 and  61)  is determined from the
average number of alkylation sites in  the aromatic mix.  In  our three
                                       33

-------
examples, the average alkylations are 1.69,  1.4,  and  1.5,  respectively.   For
an average number of alkylations equal to 1.5, Reaction  60  equals  Reaction
61.  We recommend 10^ min~* for both reactions.   If the  secondary  formation of
alky! dicarbonyls is delayed by the existence of  an intermediary dial,
Reactions 60 and 61 would be slower and would depend  on  OH  concentration.  For
a typical level of OH, a pseudo-first-order  rate  appropriate  to simulate  slow
APRC decay would be 0.004 min~^ for the sum  of Reactions 60 and 61.   The
faster reactions just noted are more likely, however, given our current       •
knowledge of the ring opening process.
Carbonyls

     Oxygenated compounds containing a carbonyl  group  (aldehydes  and  ketones)
are important intermediate species in smog chemistry.   In  the  CBM-III/EKMA,
most of these compounds are lumped into  a  single species—CARB.   The  relative
amounts of the carbonyl compounds that make up the  CARB group  determine  the
specific reaction rate constants for Reactions 34-38  in the  CBM-III/EKMA.
Reactions 37 and 38 are photolysis reactions  and are  discussed in Section  4.
We now discuss the appropriate composition of CARB  and the determination of
Reactions 34-36.
Mix of Carbonyls--

     The initial (morning) composition of the  carbonyl  category in  the  CBM-
III/EKMA is strongly influenced by two factors--fresh emissions and the
composition of the most reactive hydrocarbons,  which we have  taken  as
surrogate carbonyls in the CBM-III/EKMA.

     The data of Grosjean et al. (1981), (Example  3, Section  2) for carbonyls,
indicates an initial ratio of formaldehyde  to  total aldehydes  of about  0.6.
On the other hand, surrogate carbonyls are  primarily internal  olefins,  which
produce higher aldehydes rather than formaldehyde.  Taking  immediate produc-
tion of higher aldehydes from reactive olefins  into account,  the initial
formaldehyde fraction would be  very close to 0.5.  This is  supported by a more
extensive data set (Grosjean, 1982), which  also indicates  a morning formal-
dehyde fraction of 0.5 or less.  The data of Grosjean (1982)  also contains
information on ketone concentrations,  indicating them to be of negligible (< 5
percent) importance in the morning hours.

     Subsequent photochemical reaction tends to raise the  amount of formal-
dehyde relative to other aldehydes.  This is because the higher aldehydes
react more rapidly with OH than does formaldehyde,  and  when reaction occurs,
formaldehyde is formed as the end product.  This phenomenon is also apparent
                                       34

-------
in ambient data for aldehydes.  The overall  average  formaldehyde fraction of
total aldehydes in the data of Grosjean  (1982)  is  about  0.6.

     Photochemical reaction also  produces  glyoxal ,  primarily  from aromatic
oxidation.  Glyoxal is similar to formaldehyde  in  its  reactivity to  OH (Plum
et a!., 1983) and  in  its products.  (It  photolyzes  primarily  to stable
products, however.)

     The average ratios of formaldehyde, higher  aldehydes,  and glyoxal over
the entire lifetime of a hydrocarbon can be  estimated.   The estimation
procedure is too lengthy to describe in  detail,  but  it involves an  analysis of
the structure of the  hydrocarbon  and determination  of  which structures in the
hydrocarbon may yield which products.  For example,  formaldehyde is  primarily
produced by -CHq groups in most molecules, while glyoxal  results from non-
alkylated carbon-carbon pairs in  aromatic molecules.

     When the carbonyl estimation procedure  is  applied to the hydrocarbon
mxitures in Section 2, we obtain  the following  fractions:

          Formaldehyde          0.6
          Higher Aldehydes      0.25
          Glyoxal               0.1
          Ketones*              0.05

     The effects of the changing  mix of  carbonyl s  is seen primarily  in the
reactions of CARB with OH (Reactions 34-36).  The  ketone fraction appears to
remain quite low; the rate constant for  Reaction 34  remains around  100
ppnT^min" .  Reactions 35 and 36  change  substantially  with  photochemical
reaction.  With the 50/50 mixture of formaldehyde  and  higher  aldehydes that is
seen in the early morning, reaction 35 would have  a  value of  7000 pprrT^min"1
and Reaction 36 would be 12,000 ppm'^min"^.  As  chemical  reaction preceded
toward the average composition, these reaction  values  would become  10,500
        ~l and 6000 ppm~l min, respectively.
     While the overall rate of reaction does not change  significantly  as
reaction proceeds, the relative production of  peroxyacyl  and  hydroperoxyl
radicals changes substantially.  The effect of  this  change  is  seen  primarily
in the production of PAN; ozone formation  is little  affected.   Empirically,
PAN formation is best simulated by a ratio of Reactions  35  and 36 that  is
midway between the initial and full oxidation  average  values:   9000 ppm'^min
for Reaction 35 and 8200 ppm"^min~^ for Reaction 36.   This  corresponds  to  the
following mix of carbonyls:
  The 0.05 fraction is based upon the measurements  of  Grosjean,  1982
                                       35

-------
          Formaldehyde          0.55
          Higher Aldehydes      0.35
          Glyoxal               0.05
          Ketones               0.05

     The effects of the carbonyl mix on photolysis  will  be  discussed  in  the
next section.
                                       36

-------
                                   SECTION 4
                               PHOTOLYSIS RATES
     As the name implies, photochemical smog is activated  by  sunlight,
specifically ultraviolet light.  As may be seen in Figure  2,  different
chemical species are sensitive to different portions of the solar  spectrum,
making relative photolysis sensitive to factors that alter the  solar  spectrum,
e.g., solar zenith angle.  The CBM-III/EKMA considers the  photolysis  of  five
species:  N02, carbonyls (CARB), a-dicarbonyls  (DCRB), nitrous  acid  (HONO) and
ozone 03.  Since the photolysis of N02, often  referred to  as  kj, is the  most
studied and best understood photolysis reaction, we discuss this reaction
first.
PHOTOLYSIS OF N02

     The computer program OZIPM with which EKMA/CBM-III calculations  are
performed, calculates the rate of N02 photolysis based on the  algorithm of
Schere and Demerjian (1977).  The calculations of Schere and Demerjian, were
based on actinic irradiance values calculated by Peterson  (1975) using a
radiative transfer model developed by Dave (1972).  From these  actinic fluxes,
N02 photolysis rates were calculated from available absorbtion  and quantum
yield data.

     The calculations of Schere and Demerjian were updated  in  1980 (Demerjian
et al., 1980) and rates were reported as a function of height  (Table  9).  Note
that there is good agreement between the values calculated  in  OZIPM using the
Schere and Demerjian algorithm, and the column averaged values  for the first
two levels (0 - 0.36 km) of the light scattering model (Demerjian et  al.,
1980).

     In addition, Table 9 contains N02 photodissociation rates  as calculated
by Duewer et al., (1974), which are also based on radiative transfer  model
results, and Killus et al., (1977/1980), which were derived from measurements
of volumetric UV flux and calibrated to direct measurements of  N02 photo-
lysis.  Note that the discrepancy between these various methods for the
                                      37

-------
                                       400     420
                  Wavelength (nm)
FIGURE 2.  Product of flux and cross-section for
photolysis of ozone, acetaldehyde, HONO, and N0?.
(Source: Whitten et al., 1979).
                       38

-------






UJ
h-
1— (
•^
UJ
IVl
CC.
<£.
O
oo

o
z
o
1—


u_


oo
"*
•— ~
1
01

ro
o
.— 1
X

_ij~l
—
^^
1— <
.*

oo
UJ
I—

a:
oo
t«H
00

1
o
1—
o
:E
o.

UJ
O
»— l
X
o
1— 1
o

z
UJ
0
o
a:
L«
n~"

.
CT,
UJ
t
03
^^
I—
(O
CU-I-
t_ ^
CU CTi
CU
3
O
O
•!-> 00
CU CTi
r— 1
(/) •>•».
•— r».
r— CTi
•f~ ^™H


co 1=
,_
CU *T
^^ ^o
^ *
_1 O
0 *~"
CO ^~
CT, CM E
1—1 ,_
•> CU ^O
• > CO
i— CU •
"3 _l O
^j
CU — ^
—i E

rO f*~~
•r- CU Lf)
(_ CU •
CU _J O
E ^-^
CU
o
cu
u
13
c_
3
oo




c
TO*
•i— X
"'""3 P^
t- r>-
cu cr>
E ^
CU
Q



*
C
o
4_>
JC
0)
•^
cu
	 1

cu
CD *—~
C t/>
^ cu
CD
J= J-
4_> C7)
•r~ CU
C ^O
CU *^^
r>j

O Lf> o r^
CM r- 1 O CMO
r—( t— 1 <— t



co ^" CT> co r**- Lf> ^t~ LO r*^
CTiOO«3CMr-^OOLf>Lf)CO«d-
CTiCTiCTiCTiOOr~~VO^l'CMO





CMLftCOCO CO.— tlJO'3-CM
•a-cOf-Hi^-cMcoooo^a-
CTi CTi CTi CO 00 r^. VO ^ CM O




00.— !CTir-H^-iCMCOLOLf>00
OOOOLf)CM^OI^<*Lf)r^CO
oocooooor«-vOLOcor-iO








^ CO Lf) ^~
«5 CO CM CTi
• • • •
CTI en oo Lf>








1 — «3 CO
O <£) *± CTi
• • • •
o en oo LD
I— 1






OOOOOOOOCOVO
r"H CM CO *^ Lf) tQ 1*^- ^^ CO


















a
in
ai
X
t[

0
c
•>-
o
TO

^ZT
vo
en
—
«/)
_
c
o
4_j
JC
Q)
•^
O)
I
•II J
CD
.JE
(/)
3
<_^
Lf)
1 —
CTi
I— t
— '
cu
t_
cu
^
o


•o
c
TO

c
TO
•r™
•r^
^
cu
E
eg
o
•
s s
t— H
QJ O
• %
TO C
3
o -a
•— CU
TO t/>
o rs

* *
*












































•
V)

-------
determination of k^ is small, generally less than 20 percent.  This is also
demonstrated in Figure 3 where the surface value of N02 photolysis calculated
from Demerjian et al. (1980) is compared to the empirically calibrated methods
involving UV and TSR data (Whitten, Killus, and Johnson, 1983).

     Natural factors may cause some expected deviation from the calculated
solar function.  Cloud cover obviously has some effect (evident in Figure 3)
but correcting for clouds is a difficult and cumbersome procedure.  Since
overcast days do not usually coincide with smog episodes, we do not recommend
any attempt at treating cloud effects.

     Atmospheric turbidity  (including the effects of smog itself) has been
shown to cause as much as a 30 percent decrease in surface irradiance (Peter-
son et al., 1980).  However, Killus et al. (1980) note that turbidity redis-
tributes light in the mixed (polluted) layer, and that the effect on collumn-
averaged irradiance is less than 10 percent.

     Finally, irradiance increases with elevation.  Anderson et al. (1977)
noted a +15 percent correction for NOg photolysis in photochemical calcula-
tions for Denver, Colorado.  However, unless the elevation of the urban area
is extreme, the correction  becomes negligible.
CARBONYL PHOTOLYSIS

     Photolysis of carbonyl compounds (CARB) in the CBM-III/EKMA is a major
source of free radicals that drive the hydrocarbon oxidation process.
Photolysis of CARB can be treated in two ways:  either as a ratio to N0£
photolysis, or as a ratio to some species whose photolysis is treated
explicitly in OZIPM, e.g., HCHO.  If this latter technique is used, the
species name "CARB" must be replaced with the photolytic species identified  in
OZIPM.

     Because the  ratio of carbonyl photolysis to N02 photolysis depends  fairly
strongly on solar zenith angle, the daily average  ratio depends on latitude
and time of year.  In order that OZIPM may make these corrections auto-
matically, we recommend the second method for treating the photolysis of  CARB,
which involves replacing the species name with  "HCHO."  The ratio used for
"HCHO"  (nee CARB) photolysis in OZIPM, will depend on the species mix of
carbonyls  (discussed in Section 3) and the photolysis of the various carbonyl
species produced  (formaldehyde, higher aldehydes,  glyoxal, and ketones).  We
discuss the photolysis of these compounds next.
                                      40

-------
                                                                                                                                            O
                                                                                                                                            co
                                                                                                                                            en
                                                                                                                                             O)  C
                                                                                                                                            •r-  [/)
                                                                                                                                            •i-)   O
                                                                                                                                                 s_
                                                                                                                                             E <+-
                                                                                                                                             O
                                                                                                                                             S- T3
                                                                                                                                             <*-  O)
03
r-
J	I
                      i	I    J	L
                                                         S3
                                                         IS
                                                         CD
                                                         =2
                                                         (53
                                                         10
                                                         S3
                                                         S3
                                                         S3
                                                         S3
                                                         CM
                                                                         .    '         '
                                                                        CD
                                                                     _  OD
                                                                                      1    1    '     I    '
                                                                                                                    S3
                                                                                                                    S3
                                                                                                                    CD
                                                                                                                    S3
                                                                                                                    IS
                                                                                                                    ID
                                                                                                                    S3
                                                                                                                    S3
                                                                                                                    S3
                                                                                                                    S3
                                                                                                                    f\l
                                                                 (U  S_
                                                                 -t-J  ,
                                                                 O i—
                                                                     O
                                                                x - . 4-)
                                                                  T—O
                                                                 2*Z J=
                                                                ---  Q.

                                                                 i/>  O
                                                                 •i— -t->
                                                                 en
                                                                r— Ol
                                                                 O S-
                                                                •*-> iT3
                                                                 O Q.
                                                                J= E
                                                                 Q. O
                                                                    u
                                                                  CSJ   — »
                                                                o - — • co
                                                                z: iji  01
                                                                    O)  C
                                                                                                                                             uj -a  01
                                                                                                                                             Oi -r-  -t->
                                                                o  o  o
                                                                i— i  co T3
    S3
    CD
IS
LO
                             S3
                     S3
S3      S3
CJ      ~
  •       •
S3      SJ
                                          S3
                                            •
                                          S3
S3
CO
S3
un
S3
m
S3
CM
S3      S351
—      S3
                                                                                             S3
                                                                   41

-------
Formaldehyde

     Formaldehyde (HCHO) photolysis in OZIPM is based on the  rates determined
by Schere and Demerjian (1977).  Photolysis of HCHO to both stable and radical
products is considered; photolysis to stable products has a negligible effect
on smog chemistry, whereas photolysis to radicals is of overwhelming impor-
tance.

     In the update to their recommended photolysis rates, Demerjian et al.
(1980) applied newer quantum yield information that substantially reduced the
suggested photolysis of HCHO to stable products and somewhat  increased the
radical yield (see Table 10).

     The photolysis rates of Schere and Demerjian (1977) and  Demerjian et al.
(1980) were based on HCHO absorbtion cross sections taken at  elevated tempera-
tures  (Calvert et al., 1972).  Recently, Bass et al., (1980)  reported room
temperature absorbtion cross sections for HCHO that were considerably lower
than the high temperature data.  Applying these new cross sections to the
spectral data of Demerjian et al.  (1980), we calculated a decrease in the
rates  of photolysis to radicals of 32 percent, and a decrease of 35 percent  in
the photolysis of stable products.  This lower photolysis rate has been tested
in simulations of formaldehyde under atmospheric conditions (including natural
sunlight) by Whitten (1983), and was found to best reproduce  the experimental
HCHO/NOX data.  The radical photolysis rate of pure formaldehyde in the OZIPM
program should be multiplied by 0.81 to account for recent quantum yield  and
absorbtion cross-section data and  for the increase in photolysis with
height.  The correction factor for photolysis to stable products is 0.37.
Glyoxal

     Glyoxal photolyzes at a rate that is similar to the overall  rate  of
formaldehyde photolysis (Plum et al., 1983), but the products  of  glyoxal
photolysis seem to be mainly stable compounds.  Carbon monoxide  (CO) and  H2
are the main products, but some formaldehyde is also formed.   The conversion
of glyoxal to formaldehyde and its  subsequent photolysis to  radicals does
represent a minor source of radicals.  We estimate this source to be equiva-
lent to 10 percent of the rate of formaldehyde photolysis  to  radicals.
Conversely, glyoxal photolysis to H£ and CO occurs at nearly  twice the rate of
formaldehyde photolysis to these compounds.
Higher Aldehydes/Acetaldehyde

     The photolysis of the higher  aldehydes  under  atmospheric  conditions  has
not been studied as extensively as that of formaldehyde  and  is  highly  uncer-
tain.  Typically,  acetaldehyde is  taken to be  representative of the  higher
aldehydes; estimates of the photolysis  rate  of  acetaldehyde  vary by  more  than
a  factor of  10  (see Table 11).
                                      42

-------
TABLE 10.  FORMALDEHYDE PHOTOLYSIS RATIONS TO N02 (x 10  ) AS A FUNCTION OF
                              SOLAR  ZENITH  ANGLE
Solar Zenith
Angle (degree)
(A) HCHO + hv-*-
0
10
20
30
40
50
60
70
78
86
(B) HCHO + hv*
0
10
20
30
40
50
60
70
78
86
Schere and
Demerjian Level 1
(1977) Surface (0.16 km)
H2 + CO Ratio to
9.68

9.52

9.01

7.98



H. + HCO. Ratio
3.7

3.59

3.25

2.58



N02 + hv
5.55
5.53
5.45
5.31
5.09
4.78
4.29
3.74
3.28
2.96
to N02 + hv
4.20
4.17
4.05
3.85
3.56
3.15
2.61
2.03
1.53
1.16

5.70
5.68
5.60
5.46
5.23
4.90
4.44
3.79
2.85
2.34

4.39
4.36
4.24
4.03
3.73
3.30
2.76
2.12
1.62
1.19
Level 2
(0.36 km)












4.49
4.45
4.34
4.13
3.81
3.38
2.81
2.10
1.57
1.28
Whitten 1983

3.76
3.75
3.70
3.60
3.46
3.26
2.99
2.59
2.29
2.18

2.71
2.70
2.61
2.48
2.28
2.01
1.64
1.25
0.91
0.52
                                  43

-------
UJ
1
C3
Z
<£

i—
I-H
Z
UJ
IV!
«
O
oo

u_
0

JP*
o
t—
0
z
=3
LI_
^

OO
•t
*->
0
o
I— 1

X

CM
O


O
*~"
OO
0
1— 1
I—
^£
r^

00

00
_1
o
1—
0
3:
o_

UJ
Q

3C
UJ
O
_J
UJ
o



•
<— 1
1— 1

UJ
—1
CQ

1—
"
«/h
3

r—
•^
^
„
C

[ *
4->
*f™
2


O
CO

i-H

*
're

4->
O)
C
re
•r-
c^
(U

r—
re
c_>







"Q
c
re

a;

QJ

o
oo



^
+->

c
cu
fsj

L.
TO
^«.
o
oo
K
K
O
CO
en
fH

cT
en
o
~i™

c







i-
'oT
o
re
L.
3
10
>*— •






c
CM
00

r- 1

"
N
4->
•r—
2
O
t~.
O




r^
C>
rH

A
C
re
«f—
•"-5
t_
ai
jp
01
_
cu
O)
L.
Ol
cu
•^^
^w'

(U

a>
c:




VO CO ^f" VO ^ VO CO CO ^" OO
i— I O ^* »•"! CO CM CO ^- CO VO
O*i CTi C7^ CO f^ VO ^" CO CM I-H
oooooooooo














cviaocr>cT>.- o
r*^ r1^ UD *5f"
• • • •
0000









oooooooocovo
f-H cvj oo <^ to vo ?**- r^* oo








































•
•o
cu

re
.^
1 *
to
cu

L_
o

•a
0)
+•>
o
a.
^
(U
4_>
c
•1—

(/)
O)
3
r—
(U

fO
E
•r"
X
o
L»
a.
o.


*
















•
o
CO
en
.—1
*
r— •
re
^J
cu

c
re
•r>
L.
ai
E
cu
o
(^
o

CO
-a
• —
•2?
>.

E
3
3
[ *
C
re
3
o-
£
E

T—

U
OJ
jt
o
r~"

i
*-
T3
cu
40
re
r-w
3
U
r^-
re
o

*
*












































»
CO
•^^
^_
cu
•r-
E
3
3
j ^
C
re
3
o-

^«>
•r>
E
•r—
C_
CU
ex
Q.
3

O>
C
•r»
OO
•-^

^—

44

-------
     Using the lower limit quantum yields of Demerjian et al.  (1980), Whitten,
Killus, and Hogo (1980) calculated acetaldehyde photolysis  rates that were
then empirically tested against outdoor smog chamber experiments (Whitten,  and
Johnson, 1983).  We recommend these estimates, which indicate  acetaldehyde
photolysis at about 35 percent of formaldehyde photolysis.   This estimate can
be applied to the higher aldehydes with the caveat that the  uncertainties are
quite large.
Ketones
     There is even less information concerning ketone photolysis under
atmospheric conditions than is available for the higher aldehydes.  Carter
et al. (1979) estimated MEK photolysis to be similar to formaldehyde, whereas
Whitten,  Killus, and Johnson (1983) estimated the photolysis of acetone and
MEK to be 6 and 12 percent, respectively, of formaldehyde photolysis, based  on
simulations of outdoor smog chamber experiments.  The value estimated for MEK
photolysis, however, is probably low, because the reaction rate constant used
for MEK plus OH was faster than that indicated by recent measurements.

     Calculations for MEK photolysis based on the spectrum of Demerjian and
Schere (1977) and absorbtion and quantum yield information of Calvert and
Pitts (1966) give an estimate of roughly 50 percent of formaldehyde photo-
lysis.  Using 15 percent as a lower limit and 50 percent as an upper limit,  we
derive 30 percent as our "best guess" for the ratio of ketone to formaldehyde
photolysis.
PHOTOLYSIS OF CARBONYL MIX

     Based on the mix of carbonyls described in Section 3, we can  now estimate
the correction factor to be used in OZIPM for "HCHO,"  (to  replace  CARB).
These calculations are shown in Table 12.  Because the number of radical
products for carbonyl photolysis exceeds the number of products in a reaction
for OZIPM, we split the photolysis into four reactions:
      HCHO + QQ     0.564                                             (4-1)

      QQ -»• H02 + H02 + CO    8867 min"1                               (4-2)

      QQ -»• Me02 + Me02 + X + XCO    1133 min"1                        (4-3)

      XCO * X + CO fast (104 min"1)                                 .  (4-4)
                                      45

-------
     The factor 0.564 in Reaction 4-1 is the correction factor for photolysis
calculated in Table 12.  The ratio of Reactions 4-2 and 4-3 are determined by
the relative amounts of H02 and R0» formed in the photolysis of the different
carbonyl species.  The use of the fictitious intermediate species "XCO" and QQ
is due to the four-product species limit in OZIPM.
PHOTOLYSIS OF HONO AND OZONE

     Photolysis of HONO under atmospheric conditions is rapid, and since it
occurs at spectral wavelengths similar to N02 photolysis, the ratio is nearly
constant, (0.18).  However, this is 3.1 times the rate calculated internally
by OZIPM and this correction factor should be applied to the rate constant.

     Ozone photolyzes by two pathways; both must be included in OZIPM:


      03 + uv light 0*0                                               (4-5)

      0, + light 03P                                                  (4-6)
       O                 *


     Ozone photolysis to OP is part of a "do nothing" cycle:


                 03P                                                  (4-7)
      03P + 02 + M ->• M + 03                                           (4-8)


     Photolysis to 0 D can be important under some circumstances,  notably
those approaching clean air.  There is a correction factor of 0.71 for
updating OZIPM ozone photolysis (based on Schere and Demerjian,  1977) to be
consistent with Demerjian et al .  (1980).  However, direct measurements  of this
photolytic rate (Bahe et al., 1979) indicate a further  correction  of  0.75 for
a combined correction factor of 0.53.
PHOTOLYSIS OF DCRB

     The photolytic rate for dicarbonyls  (DCRB)  in  Killus  and  Whitten  (1982)
was set equal to ~ 0.04 x Kj, based on the known  rate  of photolysis  for
biacetyl.  Recently, Plum et al.  (1983) measured  the photolysis  rate of
methylglyoxal and found it to be  one-half that of biacetyl.  Since methyl-
glyoxal is the largest component  of DCRB, we  recommend 0.02  x  K^  as  the
                                      46

-------
TABLE 12.  CALCULATIONS FOR RATIOS OF CARBONYL PHOTOLYSIS
TO FORMALDEHYDE PHOTOLYSIS IN OZIPM.
    Species
     Fraction of
     Carbony! Mix
  Ratio to
Formaldehyde
 Photolysis
Formaldehyde
Higher aldehydes
Glyoxal
Ketones
Formaldehyde
Higher aldehydes
Glyoxal
Ketones
For Radical Products

     0.55 x 1.0
     0.35 x 0.35
     0.05 x 0.1
     0.05 x 0.35
     Weighted average

     Correction for
     quantuum yield and
     absorbtion spectrum
     updates

     CARB ("HCHO")
     photolysis to
     radicals

  For Stable Products

     0.55 x 1.0
     0.35 x ~ 0
     0.05 x 2.0
     0.05 x ~ 0
                        Correction for
                        quantuum yield and
                        absorbtion spectrum
                        updates
   0.55
   0.1225
   0.005
   0.0175
   0.696
   0.78



   0.564



   0.55

   0.1

   0.65
                           0.37

                           0.240
                            47

-------
photolysis rate for dicarbonyls.  Also, on the basis of empirical data for
yields of PAN in methylglyoxal NOX systems, we recommend a 100 percent yield
of AC03 from DCRB photolysis:


      DCRB + 1igh£ ACO^ + H09 + CO (0.02 x K, )
                      J     Cm               1
                                      48

-------
            5   INITIAL CONDITIONS, EMISSIONS, AND NMHC/NOX RATIOS
     Oxidant precursors enter the OZIPM/EKMA model in three ways:

        Initial  conditions
        Emission
        Transported air (including entrainment from aloft)

     Transport can also introduce ozone directly into the model.  Ozone and
precursor concentrations due to transport can be derived from anthropogenic
emissions upwind, which may be reduced if controls are applied.  Alterna-
tively, some ozone and precursor concentrations are due to the natural
background of those species (e.g., stratospheric intrusion for ozone, biogenic
and geogenic emissions for hydrocarbons).  The magnitude and sources of
background ozone and precursors are important topics in themselves and their
treatment is discussed in Section 6.

     In the EKMA methodology, initial conditions and emissions of ozone
precursors are assumed to be linearly related (i.e., a reduction in emissions
of NMHC or NOX is assumed to result in a proportional reduction in the initial
condition for those species).  Also, whereas the emission inputs are derived
from emission inventory information, initial conditions are derived from
ambient monitoring data, which is also used to obtain ratios of NMHC to NOX
for use in the EKMA-isopleth-derived control strategy calculations.

     To obtain a valid simulation of atmospheric photochemical processes, the
proper concentration of ozone precursors must be supplied to OZIPM/EKMA.
Furthermore, the initial conditions must be quantitatively related to post
0800 emissions;  otherwise, the emissions/concentration assumption breaks
down.  In this section we discuss these two important issues:  the use of
ambient data and the relationship between measured concentrations and input
emissions.
CONCENTRATIONS OF NMHC

     If detailed, gas chromatographic (GC) determinations of nonmethane
organic (NMOC) concentrations are available, (e.g., the three cases given in
Section 2), then that information should be used to specify initial NMOC
                                     49

-------
concentrations and NMHC/NOX ratios.  However, gas chromatographs currently
require intensive technical expertise; GC measurements are not usually
available on a routine basis.  Note also that even full GC speciation measure-
ments usually have some residual "unknown" compounds whose speciation and
reactivity must be assumed.

     In the absence of chromatographic hydrocarbon data, some other method  of
obtaining NMOC concentrations must be used.  To this end, a number of "NMHC
measurement" devices have been marketed.  These devices provide only an
indirect measurement of NMHC in that they provide separate measurements of
total organic compounds (TOC) and of methane (CH4).  The difference between
the TOC and CH4 measurements is (defined as) the NMOC measurement.  Most NMOC
analyzers are designed to perform the subtraction automatically and provide  a
direct NMOC output.

     The TOC and CH4 measurements are made with a flame ionization detector
(FID), but several methods for separating the CH4 from the TOC are in general
use.  Chromatographic analyzers use adsorbent columns to separate the CH4 from
all other organic compounds.  These analyzers tend to be rather complex, to
require special operating procedures, and to provide up to 12 analyses  per
hour rather than a truly continuous measurement.  Other analyzers use a
catalytic process to separate CH4 by oxidizing all hydrocarbons other than  CH4
in a special controlled-temperature oxidizer.  They may be either dual
channel, in which the TOC and CH4 are measured simultaneously, or cyclic, in
which TOC and CH4 are measured alternately.  In some analyzers, activated
carbon is used to remove NMOC and the CH4 determination is then made.

     Continuous NMOC analyzers suffer from a number of inherent technical
problems that limit the reproducibility of the data they provide.  A primary
problem is the necessity of subtracting two comparably-sized numbers to obtain
a measure of the NMOC.  Because the difference between the TOC and CH4
measurements — i.e., NMOC — is usually considerably smaller than either of the
individual TOC or CH4 concentrations, small errors in the TOC or CH4 measure-
ments may become large percent errors in the NMOC difference.  Furthermore,
ambient TOC and CH4 concentrations must be measured on broad, relatively
insensitive ranges of the instrument to accommodate the frequent wide excur-
sions of the TOC and CH4 ambient concentrations.   Also, FIDs are sensitive to
changes in operating conditions such as flow rates, temperature, burner
  For example, NMOC  instruments are  usually  set  to  measure  full-scale
  concentrations of  TOC of 10 ppm.   If the TOC concentration  were  5  ppm
  and the CH4 concentration were 4 ppm,  NMOC would  be  1  ppm.   Assuming a
  10% error in the TOC measurement--0.5  ppm--would  result  in  a 50% error--
  1+0.5 ppm—in the NMOC computations.
                                      50

-------
cleanliness, etc., which may result in zero and span drift.  These charac-
teristics make careful calibration and accurate balance of the TOC and CH4
channels difficult.

     Other NMOC analyzer problems over which the operator may have little or
no control  include measurement of TOC and CH4 in different samples of air due
to sequential, cyclic operation (usually not a problem for hourly averages),
nonuniform sensitivity to various organic compounds and from one analyzer
design to another, operational complexity, and potential safety hazards  from
hydrogen gas, which all FIDs require.

     Experience to date with NMOC analyzers has not been encouraging.  In the
EPA document, Guidance for Collection of Ambient Non-Methane Organic Compound
(NMOC) Data for Use in 1982 Ozone SIP Development, and Network Design and
Siting Criteria for the NMQC and NO-? Monitors,  it is noted that:
     Recent studies with commercial non-methane hydrocarbon  (NMHC) analyzers
     have established the fact that these instruments yield  unreliable data.
     Not only do instruments from different manufacturers produce different
     results, even instruments from the same manufacturer, with supposedly the
     same characteristics, yield data sometimes differing by a factor of
     two.  (For comparison, measurements made with different S02 instruments
     have a correlation coefficient of better than 0.85).  These studies were
     carried out by EPA laboratories as well as contractors, and the results
     are thus thought to be conclusive.
     The overall sense of EPA-450/4-80-011 supports the use of FIDs in
monitoring NMOC; however, this passage has been included here to emphasize the
need for properly maintaining and calibrating these instruments.  Experience
at Systems Applications in the utilization of NMOC data for atmospheric
simulations (including urban airshed modeling) underscores the expressed
concerns in the quoted passage.  Although some NMOC analyzers, particularly
those based on chromatograhic separation, seem to give reasonable estimates of
NMOC concentrations in some available data sets, the complexity of operation
tends to make data from these instruments unreliable.  Other methodologies,
particularly those that involve activated charcoal adsorbtion, appear to give
strong biases, and are not recommended.

     A good example of the errors that can arise when NMOC analyzer data is
used is found in Figure 4, which gives the NMOC and NOX data for seven
monitoring stations in the South Coast Air Basin (Los Angeles) area.  The
basin-wide average NMOC/NOX value for the five years was about 3.5:1.  The
* EPA 450/4-80-011.
                                     51

-------
   0.25r-
   0.20
 CM
O
i/i
ITS
 Q.
~ 0.15
 c
 o
4-1
 •o

-------
emission inventory for the region, however, indicates a NMOC/NOX ratio  of
greater than 5.0.  All known atmospheric processes tend to  reduce NOX faster
than NMOC.  Therefore, the NMOC/NOX ratio for the SOCAB should be greater
than 5.

     A similar discrepancy exists for data reported for the San Francisco  Bay
Area, for which the charcoal absorbtion methodology was also used to determine
CH4 and NMOC.  De Mandel et al. (1979) reported 0600 to 0900 NMHC/NOX ratios
of 1.2 and 1.7 for San Francisco and San Jose, respectively, despite an
emission ratio of 4.5.

     Fortunately, a methodology exists for obtaining NMOC estimates directly
from total hydrocarbon (THC) data.  The THC-derived NMOC methodology is more
robust than NMOC subtraction obtained from the instruments  just described.
Raskin and Kinosian (1974) first used the THC-derived NMOC  technique in a
study of hydrocarbon/NOx/oxidant relationships in the Los Angeles basin.
Unfortunately, this technique should be applied on a site-by-site basis.

     Nonmethane hydrocarbon concentrations were calculated  from measured THC
concentrations using correlations derived from chromatographic data.  Air
samples were collected at stations in Los Angeles, Azusa, and Mira Loma
between 0800 and 1000 during July, August, and September of 1971 (Bonamasa and
Mayrsohn, 1971).  Methane concentrations were subtracted from the THC concen-
trations to give the nonmethane hydrocarbon (NMHC « NMOC) concentrations.

     The reported THC-NMOC correlations were:


             Station                     THC-NMOC Correction

        Los Angeles                    NMOC * 0.64 (THC - 1.35)
        Azusa                          NMOC = 0.47 (THC - 1.4)
        Mira Loma                      NMOC = 0.73 (THC - 1.88)

     A similar correlation was used by Killus et al. (1981) for the specifi-
cation of initial conditions for the Urban Airshed Model:

                            NMHC =0.8 (THC - 1.4)

The data upon which the correlation was based were also for Los Angeles
(Mayrsohn et al. 1975; 1976).

     The regression intercept (the subtracter within the parenthesis) corre-
sponds to the background of methane for the locality.  The  proportional term
outside the parenthesis corresponds to the nonmethane fraction in urban
                                     53

-------
emissions.  This proportionality can be determined from special  gas chromato-
graphic studies similar to those just noted.  In the absence of  such data, we
recommend 0.7 as the NMOC emission fraction.

     The local methane background can be obtained from THC data  alone.   Using
a month of THC data (all hours), an average of the ten lowest THC concentra-
tions provides a good estimate of the methane background.  A small additional
correction to this estimate can be performed by subtracting 0.03 ppmC from the
average THC value as a correction for the natural background of  NMOC (see
Section 6).

     When the THC-calculated NMOC methodology is used for Los Angeles,  the
result is much more realistic than for NMOC measured by subtraction devices.
The difference between annual average NMOC concentrations is more than  a
factor of two; the ratio of calculated NMOC/NOX by the THC methodology  (Table
13) ranges from 7.19:1 to 8.58:1, values that are much more reasonable  than
those displayed in Figure 4.  When the THC-derived NMOC data are used for San
Francisco and San Jose , calculated NMOC/NOX ratios are 5:1 and  7.4:1 respec-
tively, or four times the ratios reported by De Mandel et al.  (1979).

     Further confirmation of the THC-NMOC relationship can be  seen in Table
14, which shows data obtained as part of an extensive monitoring program in
Tulsa, Oklahoma.  Two methods of NMOC measurement were used:   a  summation of
individual hydrocarbon species obtained by chromatographic separation,  and a
Beckman 6800 NMOC monitor, a subtraction device that employs a chromatograhic
adsorbent column for CH^ separation.  Data collected from these  instruments
are compared to NMOC calculated from the THC methodology just  described.

     For the instruments located at the Tulsa Post Office (Table 14a),  all
three methods for obtaining average NMOC concentrations gave identical
results.  For the health department site, however, the subtraction device
yielded results significantly different from those obtained by chromatographic
summation, while the THC-NMOC methodology was only 20 percent  greater than the
GC summation average.
EMISSIONS

     As was previously described, the OZIPM/EKMA  model  is  similar in  concept
to a Lagrangian photochemical dispersion model  in that  ozone  and  precursor
concentrations within a well-mixed  column  of  air  are  modeled  as the column
traverses the city.  The column  of  well-mixed  air is  assumed  to originate  in
  Based on maximum  hourly  averaged  concentration  for the month  of
  September  1975, using  the Los Angeles  NMOC-THC  relationship (Table 13).
                                      54

-------
TABLE 13. ANNUAL LOS ANGELES BASIN SEVEN-STATION MEAN NMOC AND NOX
FOR 0600-0900

Date
1971
1972
1973
1974
1975
NOX
0.178
0.167
0.152
0.139
0.147
Reported
NMOC
0.67
0.65
0.52
0.47
0.44
Reported
NMOC/NOX
3.76
3.89
3.42
3.38
2.99
THC
3.51
3.51
3.27
3.11
2.99
Calculated
NMOC*
1.41
1.41
1.24
1.19
1.07
Calculated
NMOC/NOy
_ A
7.9
8.45
8.15
8.58
7.29
 NMOC =0.7 (THC - 1.5)
                                55-

-------
TABLE 14.  CDT AVERAGE NMHC AND NOX AND RATIO OF NMHC TO NOX

                           (a) Tulsa Post Office


Date
8/01
02
03
07
08
09
11
12
13
14
15
18
19
20
21
22
23
24
26
Mean

ZNMHC*
ppmC
0.794
0.857
0.852
0.531
0.719
0.487
0.448
0.435
0.332
0.774
1.065
0.535
0.507
0.536
0.767
3.684
0.478
0.592
0.565
0.787

Average
6-9 CDT
NOX, ppm
0.088
0.075
0.071
0.209
0.192
0.137
0.030
0.025
0.042
0.012
0.074
0.052
0.047
0.037
0.029
0.128
0.033
0.036
0.056
0.0723

ZNMHC
NOX
9.0
11.4
12.0
2.5
3.7
3.6
14.9
17.4
7.9
64.5
14.4
10.3
10.8
14.5
26.4
28.7
14.5
16.4
10.1
10.9
_.
NMHC**
ppmC
0.980
1.110
0.900
0.380
0.600
0.690
0.470
0.320
0.350
0.330
0.940
0.740
0.710
0.530
0.830
3.110
0.720
0.470
0.430
0.77
	 "
NMHC
NOX
11.1
14.8
12.7
1.8
3.1
5.0
15.7
12.8
8.3
27. 5
12.7
14.2
15.1
14.3
28.6
24.3
21.8
13.1
7.7
10.7
"*
THC
ppmC
3.29
3.39
2.98
2.20
2.43
2.52
2.84
2.36
2.06
2.79
2.16
3.04
2.70
2.58
3.17
5.47
2.68
2.6
2.54
2.80
~*
Calculated
NMOC1"
ppmC
1.10
1.18
0.89
0.34
0.50
0.57
0.79
0.46
0.24
0.75
0.32
0.93
0.69
0.60
1.02
2.63
0.68
0.62
0.58
0.778

NMOC
NOX
12.6
15.8
12.5
1.7
2.62
4.1
26.3
18.2
5.9
62.8
4.3
17.0
14.8
16.4
35.3
20.6
20.7
17.34
10.9
10.8
 Measured by summation of  Individual  compounds  as  determined by
 GC separation.

  Measured  by Beckman 6800 NMHC  monitor.
                                  56

-------
   TABLE  14  (Concluded)

                  (b) Tulsa City/County Health Department
=============
Date
8/02
03
04
05
07
09
10
11
13
14
15
16
17
18
20
21
24
26
27
29
30
31
Mean
=
— 	 	 	
NMHC
ppmC
0.559
0.343
0.396
0.223
0.270
0.193
0.477
0.330
0.158
0.141
0.421
0.281
0.309
0.444
0.368
0.476
0.425
0.389
0.163
0.332
0.222
0.136
0.32
-
Average
6-9 COT
NOX, pprc
••••••^^ i,^ ^—
0.055
0.028
0.047
0.026
0.005
0.014
0.018
0.027
0.011
0.009
0.037
0.011
0.024
0.061
0.025
0.018
0.038
0.014
0.005
0.026
0.025
0.011
0.0243
.
	 . 	 	
NMHC
NOx ._
10.2
12.2
8.4
8.6
54.0
13.8
26.5
12.2
14.4
15.7
11.4
25.5
12.9
7.3
14.7
26.4
11.2
27.8
32.6
12.8
8.9
12.4
13.2
	
	 	 — . — . 	 — — - — —
NMHC
ppmC
0.800
0.080
0.620
0.240
0.050
0.130
0.220
0.270
0.010
0.010
0.190
0.040
0.150
0.360
0.100
0.060
0.070
0.030
0.0
0.070
0.190
0.130
0.17
_
	 	 	 — - — — — - — — —
ZNMHC
_ "°x
14.5
2.9
13.2
9.2
10.0
9.3
12.2
10.0
0.9
1.1
5.1
3.6
6.3
5.9
4.0
3.3
1.8
2.1
0.0
2.7
7.6
11.8
7.15
a— 	 	 	
Cal
THC
ppmC
••^^i— ^™-™i^— ""^"^ ""
3.08
1.98
2.52
2.29
1.79
2.03
2.12
2.49
2.08
2.02
2.31
2.02
2.08
2.82
2.24
3.01
2.56
2.05
1.88
2.14
2.06
1.9
'.26
	 ' ... -
culated
NMOCt
ppmc
«••« ^ -—••»
0.96
0.24
0.57
0.41
0.063
0.23
0.37
0.55
0.27
0.22
0.42
0.22
0.26
0.80
0.37
0.91
0.60
0.24
0.12
0.31
0.24
0.13
0.39
NMOC
NOX
17.6
8.8
12.2
15.8
12.6
16.4
20.4
20.4
24.7
24.3
11.4
29.9
10.9
13.2
14.7
50.5
15.7
16.9
24.5
11.8
9.8
11.9
15.9
* NMOC « 0.7 (THC - 1.7); The 1.7 ppmC methane  background  was  obtained
  from the average of ten lowest hourly THC measurements:  the  methane
  background for the Health Department site was identical  to that
  obtained for the Post Office site.
                                    57

-------
the urban core, and begin moving at 0800 LCT toward the site of the peak ozone
concentration.  Thus, the emissions occurring subsequent to 0800 LCT are
determined by the track of the column.  Within OZIPP, emissions data are input
to the model  for each hour after 0800 LCT, and are expressed relative to the
initial concentrations.  Thus, there are two basic and separate problems in
deriving the necessary emission information:

     (1)  Determining emissions along the track of the theoretical column;

     (2)  Expressing the emissions derived in (1) relative to the initial
          concentrations.  These initial concentrations are assumed to be the
          result of local urban emissions occurring prior to 0800.

Emission inventory data preparation is discussed in "Guidelines for Use of
City-Specific EKMA in Preparing Ozone SIPs" (EPA-450/4-80-027).  We note one
recommended modification to those guidelines:  we recommend that the NOX
emissions from large, elevated, point sources of NOX be excluded from the
emission inventory when performing EKMA simulations.  To some extent, the
questions of the degree of size and elevation need to be considered.  The
obvious reason for the exclusion of such emissions is that they may not impact
on the air parcel that contains the maximum observed ozone value.  Typically,
some local knowledge of mixing height, wind direction aloft, and stack height
can be applied to make sensible exclusions.  The need for such exclusions
arises from the major and erroneous effect large NOX values might have on the
EKMA model .

     Numerous calculations using large-scale grid models seem to indicate that
episodic ozone formation is insensitive to large point source emissions of
NOX.  This seems to be because the relatively narrow horizontal spread and
extreme vertical dispersion of concentrated NOX plumes render them unlikely to
impact a single ground monitoring site.  Moreover, the well-known suppression
of ozone by NOX suggests that the highest measured ozone in an area is
unlikely to be at a site of plume impingement.  In any case, specific point
source impacts should be assessed with a suitable plume model rather than with
OZIPM/EKMA.

     The approach recommended for computing emission fractions is one in which
the initial emission density is related to initial conditions.  The initial
emission density is calculated as the emission density necessary to generate
the initial concentrations observed within the column in one hour.  If an
initially empty column were "exposed" to the calculated initial emission
density for one hour, the concentrations of precursors in the column at the
end of the hour would equal the observed  initial  level.  The emission density
for any hour after 0800  LCT can then be related to the initial concentrations
by means of the initial  emission density.  The emission fraction can therefore
be calculated using the  following equations:
                                     58

-------
Q0 - a C0HQ    ,                                              (5-1)
                0Q

and

        E1 -S£    ,                                                   (5-2)

where
                                                       o
        Q0 = calculated initial emission density  (kg/knr)

        CQ = initial precursor concentration (ppm)

        HQ = initial mixing height (kilometers)

        Q.J = emission density for hour i from the sequence of emissions
             (kg/km2)

        E.J = emission fraction for hour i

         a = conversion factor for converting from volumetric to mass units
             (1.4 for NMOC; 0.49 for NOX)

     The factors Q^ and CQ are determined from external data, the countywide
emissions density and initial concentrations of pollutants,  respectively.  The
remaining three variables, QQ, HQ, and E^ , are interdependent; specification
of any one of these variables determines the others.

     Since initial conditions in OZIPM may be related to the phenomenon of
carryover, i.e., the percentage of the previous 24 hour's emissions  remaining
in the air column at 0800 hours, we may place some bounds on the factor E^, at
least for hydrocarbons.  Because hydrocarbons decay very slowly at night, it
is possible to estimate the mass of pollutants remaining in  the air  column
from the previous night and day using mass balance calculations.  Killus
(1981) performed such a mass balance calculation  for a high  stagnation day in
St. Louis, and estimated that carryover represented 35 percent of a  day's
emission of hydrocarbon.  More extensive calculations involving the  Urban
Airshed Model (Killus et al., 1981) also yielded  a 35 percent estimate for
carryover.

     The plausibility of the carryover of 35 percent of a day's hydrocarbon
inventory may be seen in Figures 5 and 6.  From Figure 5 one may see that, for
Philadelphia, roughly 35 percent of a day's emissions of hydrocarbon and NOX
is emitted during the period of 1800-0800, i.e.,  during the  evening, night-
time, and morning hours when winds are calm and the mixing layer is  thin and
                                     59

-------
to
c
o
.£   .08
<3   .06

M-
o



I   .04
4-J
u
     02
                                                                               RHC
                                                            I	j	I	I	I
                     0500
1000  1200


     Time of  Day
1700
2400
         FIGURE  5.   Diurnal  variations  of  area  source  emissions in Philadelphia
         emissions  inventory.
                                      60

-------
  0.10 -
  0.08 -
£ 0.06 -
         01  02  03  04   05   06   07   08  09   10  11   12  13   14   15   16   17   18   19  20  21  22  23   24
                                            Hour of the Day (MST)
  0.04 -
  0.02 -
       FIGURE 6.   Diurnal  variation  in traffic flow in the Denver metrooolitan area.
                                             61

-------
likely to trap pollutants.  A similar fraction may be derived for Denver
(Figure 6) based on diurnal traffic flow.

     Since, at most, one third of a day's NMOC emissions may be represented in
the pre-0800 column of air, and two-thirds of the day's emissions must occur
during the simulation,

                               10

                               1=1   1

or the average E^ should be greater than 0.2.  This limits the allowable
height of the 0800 mixed layer (HQ).  Thus, the morning mixing depth should
not exceed 250 meters  in any case.  However, it should also be low enough to
allow an average E^ of greater than 0.2.

     Note also that diurnal emission profiles such as those found in Figures 5
and 6 may be used to correct the daily averaged source emissions to hourly
values, provided such information exists.
* EPA-450/4-80-027.
                                     62

-------
                                   SECTION  6
                      BACKGROUND  OZONE  AND  ITS  PRECURSORS
     The photochemical modeling of polluted atmospheres requires some esti-
mation of what constitutes a "clean atmosphere."  Boundary conditions for an
urban Airshed Model, for example, should generally be specified as "clean air"
concentrations to avoid driving the model with unknown source material.  When
clean air boundaries cannot be used, as for example when long-range transport
of pollutants is a factor, control strategy calculations become more diffi-
cult, since projections of changes in long-range transport involve greater
uncertainties than projections for a single urban area.

     Conceptually, pollutant background must be divided into three com-
ponents:  tropospheric background, (biogenic and geogenic in origin) and
background due to transport of anthropogenic precursors.  Operationally, for
control  strategy calculations, the background must be divided into controll-
able and uncontrollable components.  Clearly, the tropospheric and natural
continental background is uncontrollable; in OZIPM/EKMA this background must
be specified as transported (TRANS) or ALOFT pollutants or in the kinetic
mechanism itself.  Some fraction of anthropogenic background is controllable
and appears in the initial conditions for the model.  This portion will be
rolled back with the emission control scenario.
OZONE AND NOX

     The estimation of background concentrations of ozone and nitrogen oxides
can be made fairly easily.  Ozone in remote areas has been measured in the
range of 20-60 ppb, with occasional exceptions due to photochemical destruc-
tion or generation (Routhier et al., 1980), or from stratospheric  intrusion
(Singh et al., 1980).  Nitrogen oxide data for clean atmospheres ranges from a
few ppb to less than a tenth of a ppb (Kelly et al., 1980; Noxon,  Norton, and
Marovich, 1980).  Clearly, background concentrations of NOX are very small
when compared with concentrations found in polluted atmospheres.   Therefore,
background NOX does not constitute a major perturbation in the urban photo-
chemical  system.  For example, a concentration of 3 ppb NOX will generate only
about 10 ppb of ozone before conversion to nitric acid.
                                      63

-------
     We recommend the use of a natural background of 40 ppb  (0.04 ppm) ozone
and 1 ppb (0.001 ppm) NOX.  These concentrations should be specified on TRANS
card inputs to OZIPM.
REACTIVITY OF THE FREE TROPOSPHERE

     The reactivity of unpolluted air is difficult to estimate, because it  is
the result of a great many compounds, each having a low concentration, whose
sum effect is nevertheless of significance to the smog process.   For the
purposes of this discussion, the term "reactivity" is best defined in a
specific chemical sense as the rate at which hydroxyl radicals react with
atmospheric gases to form peroxyl radicals:
                              OH + HC * ROj


     Whitten, Killus, and Johnson (1983) performed an extensive review of
background reactivity phenomena and recommended three estimates of ROG
concentrations, speciation, and OH-to-RO»conversion rates (see Table 15).
Presumably, the low, mid, and high background estimates correspond respec-
tively to tropospheric, continental, and anthropogenic transport  background.

     Singh and Sales (1982) reported measurements of methane and  light alkanes
over the eastern Pacific (Table 16 and Seiler (1974) reported background
concentrations of CO (Figure 7).  Using these data, we can calculate the
hydrocarbon and CO/OH-to-R02 conversion reactivity directly, and  can use
photochemical mechanisms to estimate the equilibrium concentration of car-
bonyls.  We arrive at an OH-to-RO» conversion rate of 94 pprrf1 min   plus 2
   These  peroxyl  radicals (HO-, CH 0-, CH CO., etc.) are responsible for
  conversion of NO to N02 in the  photochemical process  of  smog  formation
  that eventually yields ozone.   This definition of  reactivity  is  not
  always suitable for the background  effect  because  it  makes  no special
  distinction for variations in concentration  levels.   A high concen-
  tration of slowly reacting species  can  have  the  same  "reactivity"  as  a
  low concentration of quickly  reacting  species.   However, the  latter
  can only be associated with  fresh emissions  because any  true  background
  mixture must be relatively stable in both  overall  concentration  and
  speciation character.
                                       64

-------
TABLE 15.  BACKGROUND OF REACTIVE HYDROCARBONS.  (Source:
Whitten, Killus, and Johnson, 1983).
Carbon
Fraction
Low


Mid




High





0.91
0.09

0.61
0.083
0.034
0.014
0.26
0.0005
0.57
0.07
0.12
0.07
0.17
0.0005
Concentration
0.03 ppmC
0.003 ppm
0.1 ppm
0.035 ppmC
0.0008 ppm
0.001 ppm
0.0004 ppm
0.015 ppm
0.00003 ppm
0.2 ppm
0.1 ppmC
0.002 ppm
0.01 ppm
0.006 ppm
0.03 ppm
0.00008 ppm
0.5 ppm
Species
PAR
CARB
CO
PAR
ARO
ETH
OLE
CARB
DCRB
CO
PAR
ARO
ETH
OLE
CARB
DCRB
CO

0.03 ppmC

OH to R02 reactivity
= 125 min'1
0.043 ppmC
(+ 0.015 ppm CARB)


OH to R02 reactivity
= 387 min"1
0.144 ppmC
(+ 0.03 ppm CARB)



OH to R02 reactivity
1153 min'1
Note:  All estimates include 0.015 ppmC PAR as surrogate for
       background methane.
                                  65

-------
TABLE 16.    Average light hydrocarbon  concentrations
measured  over  the  eastern Pacific.
*
ConLf nt rat ion (fp^)
L«tlcud*
40-30°»
JO-20°H
20-lOe»
10-001,
0-10°S
10-20°S
20-32°S
Long tt "da
124-117
117-108
108-97
97-89
89-79
79-75
75-72
CH4
1664
1639
1601
1557
1520
1531
1526
C2H6
2.37
1.84
0.94
0.34
0.29
0.27
0.23
'A
0.12
0.10
0.05
0.11
0.07
0.07
0.08
C2H2
0.46
0.42
0.37
0.09
0.16
0.33
0.13
C3H8
0.80
0.72
0.39
0.27
0. il
0.20
0.11
C3H6
O.U5
n. is
0.30
0.16
0.15
0.28
0.07
"V.o
0.21
0.20
0.28
0.10
0.13
0.11
o.o*
"Vie
0.51
0.6n
0.65
0.30
0.19
o.n
O.V4
"V.2
0.24
0.30
0.24
0.15
0.18
0.20
0.05
n-C5Hu
0.42
0.43
0.37
0.26
0.29
0.36
0.17
   Typically baifd on an average ot 2 to 3 «an«pl«» collrnfd over thi-
   10° latitude belt.  Data collected during 30 Nov. to 21 Hoc.  1981.
                                    66

-------
          a
          a
            O.20
         ...
                    ..  =—..•'.   ••..•••.
                                       oW
                                       ..«,  -..*-•:
               60
20              0              20
 S •*—  L  ATITUDE —*•  N
60
FIGURE  7. CO-niising ratios in mrtriiie mr ma-'soH over the Atlantic and Parifir OO-HIH.  Full friiii^li-d ropre-ont
        the £llantii-fru\ti- Xo.  31  (10(>7)  from S.m  Frnnrisro I" K<-w Zf>nlnnd (Rnbinson* K'>lil.m«.  IH(i!»). The
        open  trinnf?lca rrpn-wnt the GAHP-i-xpcdition in lor.'t from  Id  S to rtn' X nl'-.ne  3d: W (>cil.-r i Juiik'f.
        I97H) over the Atliintic oconn; fhr inlirl linos  rcprosont  fhc avr-rnirr CO inixitiu rntio  fuunii b\  Kipporton
        et nl.  (1972) in the Bonicx-nrcd over  (lie Ailumir •ii-oun: D.« d|"'ii nr. \< •> rojin'«pnt  the "Sli u Uit-ton" rxjx'iii-
        tiiin in Novcinbor 11)71  over the Atlantic or»nn;  the rro~>i». tvpn'-ent the "Shm-UU-ton" <>\|)oiliti'>n  in April
        1972 over the Atlantic o''i»nn iind rhp jiomM rcpn-r-nt tho rnn-c from rdf I'SA tn rli<» Anrnn-fic (S'wmnTton
        rt •!., 1973) over the  Panfir occun.
                                                     67

-------
ppb of carbonyls, mostly formaldehyde.   The overall OH to R02 reactivity
including carbonyls is 129 min"!, ^n gOQ^ agreement with the estimates of
Whitten, Killus, and Johnson (1983).

     Because the tropospheric background concentrations of methane and light
hydrocarbons are nearly invariant, it is possible to account for their effect
as part of the kinetic mechanism itself.  The conversion of hydroxyl  to
peroxyl radicals may be modeled as two pseudo first-order reactions:

                       (+CH. + ROG)                 ,
                    OH	»• Me02    28 min'1                (6-1)


                    OH i-££ H02              66 min'1                (6-2)


The tropospheric background of CARB, because CARB is highly reactive, must be
included in the TRANS inputs to OZIPM.  As the initial  concentration  of CARB
(or "HCHO"—see Section 4) is destroyed photochemically, it is replenished by
the CARB generated by reaction 6-1 (followed by reaction 30 in the CBM-III/
EK.MA).  Thus, the equilibrium concentration of carbonyls is maintained.
CONTINENTAL BACKGROUND

     The boundary layer over continents receives a variety of biogenic,
geogenic, and anthropogenic emissions that constitute a reactive background
over and above the tropospheric background.  Whitten, Killus, and Johnson
(1983) noted a very high percentage of unidentified compounds in various
measurements of rural air.  This is not surprising.  Biogenic emissions are
known to include isoprene, terpenes, heavy paraffins (plant waxes), ethene (a
fruit hormone), esters, and alcohols, among others.  Geogenic emissions from
natural  gas and petroleum seepage, can mimic anthropogenic emissions (many of
which are, after all, derived from petroleum).

     The mid and background estimates in Table 15 should be taken as repre-
sentative of the minimum continental background; the high estimates may be
taken as associated with a substantial amount of urban transport.  The high
should not be used for OZIPM boundary conditions (TRANS) unless the trans-
ported precursors are deemed "uncontrollable", i.e., if future controls are
unlikely to affect the upwind anthropogenic emission areas.
  Kinetic theory predicts an additional 2 ppb of acetone from light alkane
  oxidation.  However, acetone is much less reactive than other carbonyl
  compounds and we have not included it in the background CARB.
                                      68

-------
     Atmospheric measurements indicate that the reactive background decreases
with height.  The data of Grosjean et al. (1981), for example, indicate a
gradual decrease in concentrations until the tropospheric background is
approached at 300 to 500 m.  The total mass loading of continental background
hydrocarbons can be represented by the mid-range background estimate taken as
a constant concentration up to 250 m.  The concentration should be inversely
proportional to mixing height; e.g., if the initial mixing depth  is taken to
be 100 m, all background concentrations should be multiplied by 2.5.

     To make the mid-range background estimate consistent with our previous
methodology for treating tropospheric background (Equations 6-1 and 6-2), we
made the following changes to the mid-range estimate.  The 0.015  PAR surrogate
is subtracted from the PAR category.  The rate constant for Reaction 6 is
increased to 88 min   to account for 0.2 ppm CO.  Also, the 0.015 ppm CARB in
the mid-range estimate of Whitten, Kill us, and Johnson includes 0.005 ppm
acetone.  Since we are neglecting acetone in our background reactivity
estimates (acetone having been shown to be relatively unreactive), we reduce
the CARB concentration to 0.01.  Thus, the following concentrations (ppm)
should be input to OZIPM as TRANS

              Fraction     Concentration (ppm)       	Species	

              0.5305             0.02                PAR
              0.1273             0.0008              ARO (0.0048  ppmC)
              0.05310            0.001               ETH (0.002 ppmC)
            .  0.0212             0.0004              OLE (0.0008  ppmC)
              0.2652             0.01                CARB
              0.0027             0.00003             DCRB (0.0001 ppmc)
      Total   0.0377

Reactions to account for tropospheric background are

                         OH * Me02    28 min"1    ,

                         OH > H02    88 min"1
                                       69

-------
                                  REFERENCES
Altshuller, A. P., and S. P. McPherson  (1963),  "Spectrophotometric Analysis of
     Aldehydes in the Los Angeles  Atmosphere,"  J.  Air  Pollut.  Control  Assoc.,
     Vol. 13, No. 3.

Anderson, G. E., et  al.  (1977),  "Air  Quality  in the  Denver Metropolitan Region
     1974-2000," EF77-222, EPA-908/1-77-002,  U.S.  Environmental  Protection
     Agency, Region  VIII, Denver,  Colorado.

Atkinson, R., K. R.  Darnall, A.  C.  Lloyd,  A.  M.  Winer,  and J.  N.  Pitts, Jr.
     (1979), "Kinetics and Mechanisms of  the  Reaction  of  the Hydroxyl  Radical
     with Organic Compounds  in the  Gas  Phase,"  Adv.  Photochem.,  Vol.  11,  pp.
     375-488.

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.,  Vol.
     12, pp. 779-836.

Bahe, F. C., W. N. Marx, and U.  Schurath  (1979),  "Determination  of the
     Absolute Photolysis Rate of Ozone  by  Sunlight,  0.,  +  ho •*•  0(0)  +
     02( A ), at Ground Level," Atmos.  Environ.,  Vol.  13,  pp.  1515-1522.

Bass, A. M., L. C. Glasgow,  C. Miller,  J.  P.  Jenson, and  D.  L.  Filkin  (1980),
     "Temperature Dependent  Absorption  Cross  Sections  for  Formaldehyde
     (Ch^O):  The Effect of  Formaldehyde  on Stratospheric  Chlorine Chemistry,"
     Atmos. Environ.. Vol. 28, pp.  675-679.

Besemer, A. C. (1982), "Formation of Chemical Compounds from Irradiated
     Mixtures of Aromatic Hydrocarbons  and Nitrogen  Oxides," j\tmos.  Environ.,
     Vol. 16, pp. 1599-1602.

Bonamassa, F., and H. Mayrsohn (1971),  "Distribution of Hydrocarbons  in the
     Los Angeles Atmosphere, Aug.-Oct.  1971," California Air Resources  Board,
     Sacramento, CA.
                                      70

-------
Bulon, H. W., J. F. Malko, and H. J. Taback (1978), Volatile Organic Compound
     (VOC) Species Data Manual, EPA-450/3-78-119, U.S. Environmental
     Protection Agency, Research Triangle Park, North Carolina.

Calvert, J. G. (1976), "Hydrocarbon Involvement in Photochemical Smog
     Formation in Los Angeles Atmosphere," Environ. Sci. Techno!., Vol.  10,
     No. 3, p. 257.

Calvert, J. 6., J. A. Kerr, K. L. Demerjian, and R. D. McQuigy  (1972),
     "Photolysis for Formaldehyde as a Hydrogen Atom Source in  the Lower
     Atmosphere," Science, Vol. 175, pp. 751-752.

Calvert, J. 6., and J. N. Pitts, Jr. (1966), Photochemistry (John Wiley &
     Sons, New York, New York).

Carter, W.P.L., et al. (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., Vol.
     11, pp. 45-103.

Darnall, K. R., R. Atkinson, and J. N. Pitts Jr. (1978). J. Phys. Chem., Vol.
     82, p. 1581.

Dave, J. V. (1971), "Development of Programs for Computing Characteristics of
     Ultraviolet Radiation," Final  Report, NAS5-21680, NASA Report CR-139134,
     National Aeronautics and Space Administration, Goddard Space Fit. Crt.,
     Greenbelt, MD.

De Mandel, R. E., L. H. Robinson, J.S.L. Fong, and R. Y. Wada (1979),
     "Comparisons of EPA Rollback,  Empirical/Kinetic, and Physicochemical
     Oxidant Prediction Relationships in the San Francisco Bay  Area," J. Air
     Pollut. Control Assoc., Vol. 29, No. 4, pp. 352.

Demerjian, K. L., K. L. Schere, and J. T. Peterson (1980), "Theoretical
     Estimates of Active (Spherically Integrated) Flux and Photolytic Rate
     Constants of Atmospheric Species in the Lower Troposphere," Adv. Environ.
     Sci. Techno!., Vo!. 9.

Dimitriades, B., and M. Dodge, Eds. (1981), Proc. of the Empirical Kinetic
     Modeling Approach (EKMA) Validation Workshop, 15-16 December, Research
     Triangle Park, NC.

Dimitriades, B., and T. C. Wesson (1972), "Reactivities of Exhaust Aldehydes,"
     Environ. Sci. Techno!., Vol. 22, No. 1, p. 33.
                                      71

-------
EPA (1977), "Uses, Limitations and Technial  Basis of Procedures  for
     Quantifying Relationships Between Photochemical  Oxidants  and  Precursors,"
     EPA-450/2-77-021a,  U.S. Environmental  Protection Agency,  Research
     Triangle Park, NC.

EPA (1978a), "Procedures for Quantifying Relationships Between Photochemical
     Oxidants and Precursors:   Supporting Documentation,"  EPA-450/2-77-021b,
     U.S. Environmental  Protection Agency,  Research Triangle Park,  NC.

EPA (1978b), "Ozone Isopleth Plotting Package (OZIPP)," EPA-600/8-78-014b,
     U.S. Environmental  Protection Agency,  Research Triangle Park,  NC.

Greiner, N. R. (1970),  "Hydroxyl  Radical Kinetics by Kinetic Spectroscopy:
     VI--Reactions with  Alkanes in the Range 300 to 500K," J.  Chem. Phys..
     Vol. 53, pp. 1970-1076.

Grosjean, D. (1982), "Formaldehyde and Other Carbonyls in  Los  Angeles Ambient
     Air," Environ. Sci. Techno!., Vol. 16,  No. 5,  p. 254.

Grosjean, D., R. Countess, K. Fung, K. Ganesan, A. Lloyd,  and  F. Lurmann
     (1981), "Deriving  Empirical  Kinetic Modeling Approach Isopleths from
     Experimental Data:   The Los  Angeles Captive-Air Study," Environmental
     Research and Technology,  Inc., Westlake Village, CA.

Hanst, P. L., N. W. Wong, and J.  Bragin (1981), "A Long-Path Infra-Red  Study
     of Los Angeles Smog." Atmos. Environ.,  Vol. 16, No. 5, pp.  969-981.

Jeffries, H. E., K. G.  Sexton, and C. H. Salmi  (1981), "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.

Kelly, T. J., D. H. Stedman, J. A. Ritter,  and  R. B. Harvey (1980),
     "Measurements of Oxides of Nitrogen and Nitric Acid in Clean  Air,"
     J. Geophys. Res.,  Vol. 85, No. C12, 20 December 1980.

Killus, J. P. (1981), "Trouble Shooting A Photochemical Grid Model—The St.
     Louis Experience,"  1981 Winter Simulation  Conference  Proceedings,  IEEE 81
     CH 1709-5 (Library  of Congress Catalog No. 80-641977).

Killus, J. P., W. R. Oliver, P. D. Gutfreund, J. E. Langstaff, T.  W. Tesche,
     and H. J. Su  (1981), "Simulation of Impacts of Nitrogen Oxide Control
     Strategies under Oxidant Episode Conditions:  Vol. 1.  Final  Report,"  SAI
     No. 81165, Systems  Applications, Inc.,  San Rafael, CA.
                                       72

-------
Killus, J. P., and G. Z. Whitten,  (1982b), "A New  Carbon-Bond Mechanism  for
     Air Quality Simulation Modeling," EPA 600/3-82-041, U.S. Environmental
     Protection Agency, Research Triangle Park, North  Carolina.

Kill us, J. P., et al. (1980), "Continued Research  in Mesoscale Air  Pollution
     Simulation Modeling--Vol. V," EF77-142R, Systems  Applications,  Inc.,  San
     Rafael, California.

Kopczynski, S. L., R. L. Kuntz, and J. J. Bufalini  (1975),  "Reactivities of
     Complex Hydrocarbon Mixtures," Environ. Sci.  Techno!.,  Vol. 9,  No.  7, p.
     649.

Kopczynski, S. L., et al. (1972),  "Photochemistry  of Atmospheric Samples in
     Los Angeles," Environ. Sci. Technol., Vol. 6,  No. 4,  p. 342.

MacCracken, M. C., and G. 0. Sauter, eds. (1975),  "Development of  an Air
     Pollution Model  for the San Francisco Bay Area,"  UCRL-51920, Vol. 1,
     Lawrence Livermore Laboratory, Livermore, CA.

Mayrsohn, H., M.  Kuramoto, J. H. Crabtree, D. Sothern, and  S. H. Mano  (1975),
     "Hydrocarbon Composition of Los Angeles Gasolines, 1974," California  Air
     Resources Board, Division of Technical Services,  Sacramento, CA.

Noxon, J. F., R.  B. Norton, and E. Marovich (1980), "N03 in  the Troposphere,"
     Geophys. Res. Lett., February 1980, Vol. 7, No. 2, pp.  125-128.

Oberdorfer, P. E. (1967), "The Determination of Aldehydes  in Automobile
     Exhaust Gas," SAE Paper 670123, Society of Automotive  Engineers,  New
     York, New York.

Paskind, J. J., and J. R7 Kinosian (1974), "Hydrocarbons,  Oxides of  Nitrogen
     and Oxidant  Pollutant Relationships in the Atmosphere  over California
     Cities," 67th Annual Meeting, Air Pollution Control Association,  9-13
     June 1974, Denver, CO.

Peterson, J. T. (1976), "Calculated Actinic Fluxes  (290-700nm) for  Air
     Pollution Photochemistry Applications," EPA-600/4-76-002, Environmental
     Science Research Laboratory,  Environmental Protection  Agency,  Research
     Triangle Park, NC.

Platt et al. (1980),  "Observations of HONO in an Urban Atmosphere by
     Differential Optical Absorption," Nature, Vol. 285, p.  312.

Plumb, I. C,, K.  R. Ryan, J. R. Steven, and M.F.R.  Mulcahy  (1981),  "Kinetics
     of the Reaction  of C2H502 with NO at 295K," Int.  J. Chem. Kinetics,
     Vol. 14, pp. 183-194.
                                      73

-------
Routhier, F., K. Dennett, D. D. Davis, H. Wartburg,  P.  Haagenson,  and  H.  C.
     Del any (1980), "Free Troposphere and Boundary Layer  Airborne  Measurements
     of Ozone Over the Latitude Range of 58'S  to  70°N,"  J.  Geophys.  Res.,
     Vol. 85, No. C12, 20 December 1980.

Sawicki, E., T. W. Stanley, and W. Elbert (1961), "The  3-Methyl-2-
     benzothiazolone Hydrazone Test," Anal. Chem., Vol.  38,  No.  1,       pp.
     93-96.

Schere, K. L., and K. L. Demerjian (1977),  "Calculation  of  Selected  Photolytic
     Rate Constants over a Diurnal Range,"  EPA-600/4-77-015,  U.S.
     Environmental Protection Agency, Research Triangle  Park,  North  Carolina.

Scott Research Laboratories (1970), "1969 Atmospheric Reaction  Studies  in  the
     Los Angeles Basin, Vol. III.  El Monte Ground Data," National Air
     Pollution Control Administration, No.  70-6,  Plumsteadville, Pennsylvania.

Seiler, W. (1974), The Cycle of Atmospheric CO,"  Tellus,  XXVI  1974 (1-2),  pp.
     116-135.

Seizinger, D. E., and B. Dimitriades (1972), "Oxygenates  in  Exhaust  from
     Simple Hydrocarbon Fuels," J. Air Pollut. Control Assoc.,  Vol.  22, No. 1,
     pp. 47-51.

Singh, H. B., W. Viezee, W. B. Johnson, and F. L. Ludwig  (1980), "The  Impact
     of Stratospheric Ozone on Tropospheric Air Quality," JAPCA, Vol.  30,  No.
     9.

Singh, H. B., and L. J. Sal as (1982), "Measurement of Selected  Light
     Hydrocarbons over the Pacific Ocean:  Latitudinal and  Seasonal
     Variations," Geophys. Res. Lett., Vol. 9, No. 8, pp. 842-845.

Wadowski, C. S., and E. E. Weaver (1970), "The Effects of Engine Parameters,
     Fuel Composition, and Control Devices on Aldehyde Exhaust  Emissions,"
     presented at the APCA West Coast Meeting, 8  October  1970,  San Francisco,
     California.

Whitten, 6. Z., J. P. Killus, R. G. Johnson (1982),  "Modeling of Simulated
     Photochemical Smog with Kinetic Mechanisms," Final Report, EPA  Contract
     No. 68-02-3281, U.S. Environmental Protection Agency, Research Triangle
     Park, NC.
                                      74

-------
Whitten, G. Z., J. P. Killus, and H. Hogo (1980), "Modeling of Simulated
     Photochemical Smog with Kinetic Mechanisms.  Vol. 1.  Final Report," EPA-
     600/3-80-028a, U.S. Environmental Protection Agency, Office of Research
     and Development, Research Triangle Park, NC.

Whitten, G. Z., and H. Hogo (1978), "User's Manual for Kinetics Models  and
     Ozone Isopleth Plotting Package," EPA-600/8-78-014a, U.S. Environmental
     Protection Agency, Research Triangle Park, NC.

Whitten, G. Z., and R. G. Johnson (1982), "Computer Modeling of Simulated
     Photochemical Smog," Monthly Technical Narrative No. 2, Systems
     Applications, Inc., San Rafael, CA.

Whitten, G. Z., and R. G. Johnson (1983), "Computer Modeling of Simulated
     Photochemical Smog," Monthly Technical Narrative No. 3, Systems
     Applications, Inc., San Rafael, CA.
                                       75

-------
                          Appendix  A




Recommended Rates for Carbon-Bond Mechanism-III/EKMA in OZIPM


1
2
3
4
5
6
7
8
9
10
11
12
13
14
15

16
17
18
19
Reaction

N02 * NO + 0
0 - (02) + (M) + 03
NO + 03 * N02 + 02
N02 + 03 + N03 + 02
N02 + 0 -»• NO + 02
OH + 03 * H02 + 02
H02 + 03 * OH + 202
OH + N02 * HN03
°2
OH + CO -4 H02 + C02
NO + NO + (02) * N02 + N02
NO + N03 * N02 + N02
N02 + N03 + H20 * 2HN03
NO + H02 * N02 + OH
H02 + H02 * H202 + 02
X + PAR *
0
OH + PAR -4 ME02 -»• H20
°2
0 + OLE -4 ME02 + AC03 + X
0 + OLE + HCHO + PAR
°2
OH + OLE -4 RA02
Rate Constant
at 298 K
if
1.0
4.40 x 106**
26.6
0.048
1.3 x 104
100
2.40
1.60 x 104
440
1.50 x 10'4
2.80 x 104
1.3 x 10-3t
1.20 x 104
1.50 x 104
105

1200
2700
2700
3.70 x 104
Activation
Energy

0
0
1450
2450
0
1000
1525
0
0
0
0
-1.06 x 104
0
0
0

560
325
325
-540
                                                           Continued
                             76

-------

20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38a

38b
38c
38d
Reaction
03 + OLE * HCHO + CRIG
Oo + OLE + HCHO + MCRG + X
°2
0 + ETH -4 ME02 + H02 + CO
0 + ETH > HCHO + PAR
°2
OH + ETH -4 RB02
Oo + ETH -»• HCHO + CRIG
°2
NO + AC03 -4 N02 + ME02
°2
NO + RB02 -4 N02 + HCHO + H02 + HCHO
°2
NO + RA02 -4 N02 + HCHO + H02 + HCHO
°2
NO + ME02 -4 N02 + HCHO + ME02 + X
°2
NO + ME02 -4 N02 + HCHO + H02
NO + ME02 * NRAT
03 + RB02 •»• HCHO + HCHO + H02 + 02
03 + RA02 * HCHO + HCHO + H02 + 02
OH + HCHO * CROo + X
°2
OH + HCHO -4 HOo + CO
°2
OH + HCHO -4 AC03 + X
HCHO -»• CO + Ho
°2
HCHO + -4 QQ
2 00
QQ 	 * H02 + H02 + CO
QQ + ME02 + ME02 + X + XCO
XCO -»• X + CO
Rate Constant
at 298 K
0.008
0.008
600
600
1.20 x 104
0.0024
1.04 x 104
1.20 x 104
1.20 x 104
3700
7400
900
5.0
20
100
9000
8200
(0.24)*
(0.564)*

8867
1133
104
Activation
Energy
1900
1900
800
800
-382
2560
0
0
0
0
0
0
0
0
0
0
0
0
0




                              Continued
77

-------

39
40
41
42
43
44
45
46
47
48
49
50
51
52

53
54

55
56
57

58
Reaction
N02 + AC03 + PAN
PAN -»• AC03 + N02
H02 + AC03 ->• Stable products
H02 + ME02 * Stable products
NO + CRIG * N02 + HCHO
N02 + CRIG ->- N03 + HCHO
HCHO + CRIG * Ozonide
NO + MCRG -»• N02 + HCHO + PAR
NO? + MCRG + NO-} + HCHO + PAR
£. - Ozonide
CRIG * CO + H20
CRIG -»• Stable products
°2
CRIG -4 H02 + H02 + CO
MCRG -»• Stable products
0
MCRG -4 ME02 + OH + CO
°2
MCRG -4 ME02 + H02
0
MCRG -4 HCHO + H02 + CO + H02
°2
OH + ARO -4 RARO + H20
°2
OH + ARO -4 HOo + OPEN
o, 2
NO + RARO -4 N02 + PHEN + H02
Rate Constant
at 298 K
(pprrT1 min'1)
7000
0.022
1.50 x 104
9000
1.20 x 104
8000
2000
1.20 x 104
8000
2000
670**
240**
90**
150**

340**
425**

85**
6000
1.45 x 104

4000
Activation
Energy
(•1C)
0
1.35 x 104
0
0
0
0
0
0
0
0
0
0
0
0

0
0

0
600
400

0
                               Continued
78

-------

59
60
61
62
63
64
65
66

67
68
69
70
71
72
73
74
75
76
77

OPEN
APRC
APRC
PHEN
Reaction
°2
+ NO -4 N02 + DCRB + X + APRC
0
-4 DCRB + HCHO + CO + X
°2
-4 HCHO + HCHO + CO + CO
+ N03 -»• PHO + HN03
PHO + N02 * NPHN
PHO
OPEN
OH +

DCRB
PHEN
CR02
DCRB
HONO
OH +
03 +
oV
(>lD
°3 +
NR +
+ H02 -»• PHEN
+ 03 •* DCRB + X + APRC
PHEN -£ H02 + APRC + PAR + HCHO
°2
-4 H02 + AC03 + CO
+ OH > PHO
°2
+ NO -4 N02 + HCHO + AC03 + X
+ OH •»• AC03 + CO
* OH + NO
NO > HONO
0lD
+W 0
+ H20 + OH + OH
o

Rate Constant Activation
at 298 K Energy
(ppm'1 min'1) (°K)
6000
104**
104**
5000
4000
5.00 x 104
40
3.00 x 104
**
(0.02 x Kj)
104
1.20 x 104
25000.
(3.1* or 0.18** x Kx)
9770
(0.53)*
4.44 x 1010
3.4 x 105
(1.0)*
0.0
0
0
0
0
0
0
0
0

0
0
0
0
0
0
0
0
0


   Sunlight-dependent; rate constant is correction factor for OZ1PM input,
   units of min
               -1
**
"*" Units of ppnT'TTun"1.




  Units of min"1.
                                     79

-------
        Appendix B



BOND GROUPS PER MOLECULE
            80

-------
TABLE B-l. Molecular Weights of Molecules
           (in Alphabetical Order)
Species
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
SAP CAD
Code
43814
43820
43813
45225
45208
45207
43218
46201
43245
43224
43312
43296
43276
43299
43291
43280
43279
43234
50001
43274
43277
43271
43278
43308
43311
43452
50002
43310
43229
43225
43228
50004
43275
43230
43223
43211
43270
43298
43295
43293
Chemical Name
1,1, l-TfilCHLOROETHANE
1,1, 2-TRICHLOROETflANE
1,1-DICHLOROETflANE
1,2, 3-TRIMETHYLBENZENE
1 , 2, 4-TRIMETHYLBENZENE
1,3, 5-TFIMETflYLBENZENE
1,3-BOTADIENE
1,4-DIOXANE
1-HEXENE
1-PENTENE
1-T-2-C-4-TM-CYCLOPENTANE
2 , 2 , 3-TRIMETflYLPENTANE
2,2, 4-TRIMETHYLPENTANE
2,2, 5-TRIMEOHYLPENTANE
2 , 2-DIMETHYLBUTANE
2, 3 , 3-TBIMETHYLPENTANE
2,3, 4-TRIMETHYLPENTANE
2, 3-DIMETflYL-l-BUTENE
2 , 3-DIMETflYLBUTANE
2 , 3-DIMETHYLPENTANE
2 , 4-DIMETHYLHEXANE
2 , 4-DIMETflYLPENTANE
2,5-DIMEOHYLHEXANE
2-BOTYLETHANOL
2-ETHGKYEOHANOL
2-E'XHGXYETHYL ACETATE
2-ETflYL-l-BOTENE
2-METflCKYETflANOL
2-METHYL PENTANE
2-METHYL-l-BDTENE
2-METflYIr-2-BDTENE
2-METHYL-2-PENTENE
2-METflYLHEXANE
3-METflYL PENTANE
3-METHYL-l-BOTENE
3-*lETflyL-l-PENTENE
3-METflYI^Tl-2-PENTENE
3-METflYLHEPTANE
3-METflYLBEXANE
4-METHYL-T-2— PENTENE
Molecular
Weight
133.42
131.66
98.97
120.19
120.19
120.19
54.09
88.12
84.16
70.13
112.23
114.23
114.22
114.23
86.17
114.22
114.22
84.16
86.17
100.20
114.22
100.20
114.22
118.17
90.12
132.00
84.16
76.09
86.17
70.13
70.13
84.16
100.20
86.17
70.13
84.16
84.16
114.23
100.20
84.16
                         81

-------
TABLE B-l. (Continued)
Species
No.
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
67
68
69
70
71
72
73
74
75
76
77
78
79
80
SAROAD
Code
43297
45221
50025
43503
43404
43551
43702
43206
43704
50015
50020
50026
45201
50024
43213
43510
50003
43115
43116
43117
43511
43512
43289
43294
43513
43290
43807
43804
43443
43803
43217
43227
50019
43248
43264
43273
43242
43292
43207
50027
Chemical Name
4-METHYLHEFTANE
A-METHYLSTYRENE
A-PINENE
ACETALDEHYDE
ACETIC ACID
ACETONE
ACETONITRILE
ACETYLENE
ACRYLONITRILE
ANTHRACENE
B-METHYLSTYRENE
B-PINENE
BENZENE
BENZYLCHIORIDE
BDTENE
BOTYRALDEHYDE
C-3-HEXENE
C-7 CYCLOPARAFFINS
C-8 CYCLOPARAFFINS
C-9 CYCLOPARAFFINS
C3 ALDEHYDE
C5 ALDEHYDE
C6 OLEFINS
C7-OLEFINS
C8 ALDEHYDE
C8 OLEFINS
CARBON TETRABROMIDE
CARBON TETRACHLORIDE
CELLOSOLVE ACETATE
CHLOROFORM
CIS-2-BOTENE
CIS-2-PENTENE
CRYOFLCORANE (FREON 114)
CYCLOHEXANE
CYCLOHEXANONE
CYCLOHEXENE
CYCLOPENTANE
CYCLOPENTENE
CYCLOPROPANE
D-LIMONENE
Molecular
Weight
114.23
118.15
136.24
44.05
60.05
58.08
41.05
26.04
53.06
178.22
118.15
136.24
78.11
112.56
56.10
72.12
84.16
98.19
112.23
126.26
58.08
86.14
84.16
98.18
128.21
112.23
331.67
153.84
132.00
119.39
56.10
70.13
170.93
84.16
98.15
82.14
70.14
68.11
42.08
136.24
                82

-------
TABLE B-l. (Continued)
Species
No.
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
SAROAD
Code
43320
43823
43802
50018
43450
50017
45103
50012
43287
43285
43202
43433
43438
43302
43812
43351
43219
43721
45203
43288
43203
43815
43370
43601
50011
43502
43368
43367
43286
43282
43232
50005
43281
43231
43214
43306
43304
43446
43451
43215
Chemical Name
DIACETDNE ALCOHOL
DICfiLQRODIFLOOROMETBANE
DICHLOROMETHANE
DIMETHYL ETHER
DIMETHYL FORMAMIDE
DIMETHYL-2 , 3 , DIHYERO-1H-INDENE
DIMETHYLETHYLBENZENE
DIMETHYLNAPHTHALENE
DOCOSANE
EICOSANE
ETHANE
ETHYL ACETATE
ETHYL ACRYLATE
ETHYL ALCOHOL
ETHYL CHLORIDE
ETHYL ETHER
ETHYLACETYLENE
ETHYLAMINE
ETHYLBENZENE
ETHYLCYCLOHEXANE
ETHYLENE
ETHYLENE DICHLQRIDE
ETHYLENE GLYCOL
ETHYLENE OXIDE
ETHYLNAPHTHALENE
FORMALDEHYDE
GLYCOL
GLYCOL ETHER
HENEICOSANE
HEPTADECANE
HEPTANE
HEPTENE
HEXADECANE
HEXANE
ISO-BOTANE
ISO-BDTYL ALCOHOL
ISO-PROPYL ALCOHOL
ISOBUTYL ACETATE
ISOBUTYL ISOBUTXRATE
ISOBUTYLENE
Molecular
Weight
116.16
120.91
84.94
46.07
73.09
146.23
134.21
156.22
310.59
282.54
30.07
88.10
100.11
46.07
64.52
74.12
54.09
45.09
106.16
112.23
28.05
99.00
62.07
44.05
156.22
30.03
62.07
62.07
296.57
240.46
100.20
98.18
226.44
86.17
58.12
74.12
60.09
116.16
144.21
56.10
          83

-------
TABLE B-l. (Continued)
Species
No.
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
SAROAD
Code
43120
45105
43109
45106
43112
45104
43106
43105
45234
43108
43107
43114
43122
43121
45108
43113
43111
45107
43110
45102
43243
43444
43119
50022
45212
45205
43201
43432
43301
43445
43801
43552
43560
43559
43209
50016
43261
43262
43272
43819
Chemical Name
ISCMERS OF BUTENE
ISCMERS OF BUTYLBENZENE
ISCMERS OF DECANE
ISCMERS OF DIETBYLBENZENE
ISCMERS OF DODECANE
ISCMERS OF ETHYLTOLUENE
ISCMERS OF HEPTANE
ISCMERS OF HEXANE
ISCMERS OF METHYLPROP. BENZENE
ISCMERS OF NCNANE
ISCMERS OF OCTANE
ISCMERS OF PENTADECANE
ISCMERS OF PENTANE
ISCMERS OF PENTENE
ISCMERS OF PROPYLBENZENE
ISCMERS OF TETRADECANE
ISCMERS OF TRIDECANE
ISCMERS OF TRIMETHYLBENZENE -
ISCMERS OF ONDECANE
ISCMERS OF XYLENE
ISOPRENE
ISOPROPYL ACETATE
IACTOL SPIRITS
M-CRESOL (3-METHYLBENZENCL)
M-ETHYLTOLCENE
M-XYLENE
METHANE
METHYL ACETATE
METHYL ALCOHOL
METHYL AMYL ACETATE
METHYL CHLORIDE
METHYL ETHYL KETDNE
METHYL ISOBUTYL KETONE
METHYL N-BUTYL KETONE
METHYLACETYLENE
METflYLANTHRACENE
METHYLCYCLOHEXANE
METHYLCYCLOPENTANE
METHYLCYCLOPENTENE
METHYLENE BROMIDE
Molecular
Weiaht
56.10
134.21
142.28
134.21
170.33
120.19
100.20
86.17
134.21
128.25
114.23
212.41
72.15
70.13
120.19
198.38
184.36
120.19
156.30
106.16
68.13
104.00
114.23
110.16
120.19
106.16
16.04
74.08
32.04
140.00
50.49
72.10
100.16
100.16
40.06
192.25
85.16
84.16
82.14
173.85
            84

-------
TABLE B-l. (Continued)
Species
No.
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
SAROAD
Code
50010
43118
45801
43212
43435
43305
43238
43255
43260
43220
43303
45209
43259
43258
45101
43284
43235
50021
45211
45204
43283
43233
43265
50023
45206
43817
45300
50006
43204
43504
43434
43205
43369
43602
50013
43208
45216
45220
43123
43309
Chemical Name
METHYLNAPHTflALENE
MINERAL SPIRITS
MONOCHLQRCBENZENE
N-BDTANE
N-BOTYL ACETATE
N-BDTYL ALCOHOL
N-DECANE
N-DODECANE
N-PENTADECANE
N-PENTANE
N-PROPYL ALCOHOL
N-PROPYLBENZENE
N-TETRADECANE
N-TRIDECANE
NAPHTHA
NONADECANE
NONANE
0-CRESOL (2-METflYLBENZENOL)
0-ETHYLTOLOENE
0-XYLENE
OCTADECANE
OCTANE
OCTENE
P-CRESOL (4-METHYLBENZENOL)
P-XYLENE
PERCHLOROETHYLENE
PHENOLS
PROPADIENE
PROPANE
PROFRIONALDEHYDE
PROPYL ACETATE
PROPYLENE
PROPYLENE GLYCOL
PROPYLENE OXIDE
PROPYLNAPHTHALENE
PROPYNE
SEC-BDTYLBENZENE
STYRENE
TERPENES
TERT-BOTYL ALCOHOL
Molecular
Weight
142.19
114.23
112.56
58.12
116.16
74.12
142.28
170.33
212.41
72.15
60.09
120.19
198.38
184.36
114.23
268.51
128.25
110.16
120.19
106.16
254.49
114.23
112.21
110.16
106.16
165.85
94.11
40.06
44.09
58.08
102.13
42.08
76.00
58.08
170.25
40.06
134.21
104.14
136.24
74.12
           85

-------
                        TfiBLE B-l.  (Concluded)
Species  SAROAD
  No.     Code
                Chemical Name
                                                        Molecular
                                                        Weight
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
45215
43390
45232
45202
43216
43226
45233
43824
43811
43821
43740
43822
50014
43241
43000
43860
45401
                  TER^BOTYLBENZENE
                  TETRAHYDROFURAN
                  TETOAMETHYLBENZENE
                  TOLDENE
                  TOANS-2-BOTENE
                  ORANS-2-PENTENE
                  ORI/TE1RAALKYL BENZENE
                  ORICHLOROE'IHYLENE
                  aFICHLOROFLOORGMETBANE
                  TRIMEOHYL AMINE
                  1RIMETBYLFLDOROSIIANE
                  TOIMETHYIflAEHTHALENE
                  DNDECANE
                  UNKNOWN SPECIES
                  VINYL CHLORIDE
                  XYLENE BASE ACIDS
134.21
 72.10
134.21
 92.13
 56.10
 70.13
148.23
131.40
137.38
187.38
 59.11
 92.00
170.25
156.30
 86.00
 62.50
230.00
                                    86

-------
TABLE B-2. BOND GROUPS PER MOLECULE
           (IN ALPHABETICAL ORDER)

Species Profiles by Bond
Saroad
NO.
43814
43820
43813
45225
45208
45207
43218
46201
43245
43224
43312
43296
43276
43299
43291
43280
43279
43234
50001
43274
43277
43271
43278
43308
43311
43452
50002
43310
43229
43225
43228
50004
43275
43230
43223
43211
43270
43298
43295
43293
43297
45221
50025
43503
43404
Chemical Name
1,1, 1-TRICHLOROETHANE
1,1, 2-TRICHLOROETHANE
1 , 1-DICHLOROETHANE
1 , 2 , 3-TRIMETflYLBENZENE
1 , 2 , 4-TRIMETflYLBENZENE
1,3,5-TRIMETHYLBENZENE
1,3-BUTADIENE
1,4-DIOXANE
1-HEXENE
1-PENTENE
1-T-2-C-4-1H-CYCLOPENTANE
2 , 2 , 3-TRIMETflYLPENTANE
2,2, 4-TRIMETHYLPENTANE
2 , 2 , 5-TR IMETflYLPENTANE
2 , 2-DIMETflYLBUTANE
2,3, 3-TR IMETflYLPENTANE
2 , 3 , 4-TR IMETflYLPENTANE
2 , 3-DIMETHYL-l-BUTENE
2 , 3-DIMETflYLBUTANE
2 , 3-DIMETHYLPENTANE
2 , 4-DIMETHYLHEXANE
2 , 4-DIMETflYLPENTANE
2 , 5-DIMETflYLHEXANE
2-BUTYLETflANOL
2-ETflOXYETHANOL
2-ETflOXYETHYL ACETATE
2-ETHYL-l-BOTENE
2-METflOXYETflANOL
2-METHYL PENTANE
2-METHYL-l-BOTENE
2-METflYL-2-BOTENE
2-METflYL-2-PENTENE
2-METflYLHEXANE
3-HETHYL PENTANE
3-METHYL-l-BOTENE
3-^ETHYL-l-PENTENE
3-METHYL-T-2-PENTENE
3-HETHYLBEPTANE
3-METflYLHEXANE
4-METflYL-T-2-PENTENE
4-METHYLHEPTANE
A-METHYLSTYKENE
A-PINENE
ACETALDEHYDE
ACETIC ACID
OLE
^^
-
-
-
-
-
2
1
1
1
—
-
-
—
-
—
-
1
—
-
-
—
—
-
—
—
1
-
—
1
—
—
-
-
1
1
—
-
—
-
—
-
1
-
-
PAR
^ .
—
-
3
3
3
-
2
4
3
8
8
8
8
6
8
8
4
6
7
8
7
8
5
3
4
4
2
6
3
3
4
7
6
3
4
4
8
7
4
8
2
8
1
1
Group
ARO
^m
-
-
I
1
1
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
—
-
-
-
-
-
-
-
-
—
-
-
—
-
-
-
-
—
-
1
-
-
-
GARB
—
-
-
-
-
-
-
2
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
1
1
2
-
1
-
-
2
2
-
-
—
-
2
-
-
2
-
1
—
1
-
ETfl
•V
-
-
-
-
—
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
DNREACT
2
2
2
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
—
-
—
-
1
            87

-------
TABLE B-2. (Continued)

Species Profiles by Bond Group
Saroad
NO.
43551
43702
43206
43704
50015
50020
50026
45201
50024
43213
43510
50003
43115
43116
43117
43511
43512
43289
43294
43513
43290
43807
43804
43443
43803
43217
43227
50019
43248
43264
43273
43242
43292
43207
50027
43320
43823
43802
50018
43450
50017
45103
50012
43287
43285
Chemical Name
ACETONE
ACETONITRILE
ACETYLENE
ACRYLONITRILE
ANTHRACENE
B-METHYLSTYRENE
B-PINENE
BENZENE
BENZYLCHLORIDE
BUTENE
BOTYRALDEHYDE
C-3-HEXENE
C-7 CYCLOPARAFFINS
C-8 CYCLOPARAFFINS
C-9 CYCLOPARAFFINS
C3 ALDEHYDE
C5 ALDEHYDE
C6 OLEFINS
C7-OLEFINS
C8 ALDEHYDE
C8 OLEFINS
CARBON TETRABROMIDE
CARBON TETRACHLORIDE
CELLOSOLVE ACETATE
CHLOROFORM
CIS-2-BOTENE
CIS-2-PENTENE
CRYOFLOORANE (FREON 114)
CYCLOHEXANE
CYCLOHEXANONE
CYCLOHEXENE
CYCLOPENTANE
CYCLOPENTENE
CYCLOPROPANE
D-LIMONENE
DIACETONE ALCOHOL
DICHLORODIFLOOROMETflANE
DICHLOROMETflANE
DIMETHYL ETHER
DIMETHYL FORMAMIDE
DIMETHYL-2 , 3 , DIHYDRO-1H-INDENE
DIMETHYLETflYLBENZENE
DIMETHYLNAFHTHALENE
DOCOSANE
EICOSANE
OLE
••>
—
—
1
—
1
1
—
—
1
—
1
—
-
-
-
-
1
1
-
1
—
-
-
—
—
—
-
-
-
1
-
1
-
1
—
—
-
-
-
-
-
-
-
—
PAR
3
1
1
1
8
-
8
1
1
2
3
4
7
8
9
2
4
4
5
7
6
-
-
4
-
2
3
-
6
5
4
5
3
3
6
5
-
-
2
-
5
4
6
22
20
ARO
•H
—
—
—
1
1
—
—
1
—
—
—
-
-
-
-
-
-
-
—
-
-
-
—
—
-
-
-
-
-
-
-
-
-
-
—
—
-
-
-
1
1
1
-
-
CARE
•V
—
-
-
-
1
-
-
-
—
1
—
-
-
-
1
1
-
-
1
-
-
-
2
-
2
2
-
-
1
-
-
-
-
2
1
-
-
-
-
-
-
-
-
-
ETH
^^
—
-
-
-
-
-
—
—
-
—
—
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
—
-
-
-
-
-
-
-
-
-
-
- .
-
-
-
-
-
DNREACT
^
1
1
-
-
-
-
5
-
-
-
—
-
-
-
-
-
-
-
-
-
1
1
-
1
-
-
2
—
-
-
-
-
-
-
-
1
1
-
3
-
-
-
-
-

-------
    TABLE B-2. (Continued)
Species Profiles by Bond Group
Saroad
43202
43433
43438
43302
43812
43351
43219
43721
45203
43288
43203
43815
43370
43601
50011
43502
43368
43367
43286
43282
43232
50005
43281
43231
43214
43306
43304
43446
43451
43215
43120
45105
43109
45106
43112
45104
43106
43105
45234
43108
43107
43114
43122
43121
45108
Chemical Name
ETHANE
ETHYL ACETATE
ETHYL ACRYLATE
ETHYL ALCOHOL
ETHYL CHLORIDE
ETHYL ETHER
ETHYLACETYLENE
ETHYLAMINE
ETHYLBENZENE
ETHYLCYCLOHEXANE
ETHYLENE
ETHYLENE DICHLQRIDE
ETHYLENE GLYCOL
ETHYLENE OXIDE
ETHYLNAPHTHALENE
FORMALDEHYDE
GLYCOL
GLYCOL ETHER
HENEIOOSANE
HEPTADECANE
HEPTANE
HEPTENE
HEXADECANE
HEXANE
ISO-BCTANE
ISO-BUTYL ALCOHOL
ISO-PROPYL ALCOHOL
ISOBUTYL ACETATE
ISOBUTYL ISOBUTYRATE
ISOBUTYLENE
ISOMERS OF BOTENE
ISOMERS OF BUTYLBENZENE
ISOMERS OF DECANE
ISOMERS OF DIETHYLBENZENE
ISOMERS OF DODECANE
ISOMERS OF ETHYLTOLOENE
ISOMERS OF HEPTANE
ISOMERS OF HEXANE
ISOMERS OF METHYLPROP. BENZENE
ISOMERS OF NONANE
ISOMERS OF OCTANE
ISOMERS OF PENTADECANE
ISOMERS OF PENTANE
ISOMERS OF PENTENE
ISOMERS OF PROPYLBENZENE
OLE
_
-
—
-
-
-
—
—
-
-
-
—
-
-
-
-
-
-
-
-
-
1
-
-
-
-
-
—
-
—
-
—
-
—
—
—
-
-
-
—
-
—
-
—
—
PAR
.4
3
3
2
-
3
4
1
2
8
-
-
2
1
6
-
1
1
21
17
7
5
16
6
4
4
3
6
7
3
2
4
10
4
12
3
7
6
4
9
8
15
5
3
3
ARO
mi
-
-
-
-
-
-
-
1
-
-
-
-
-
1
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
1
-
1
-
1
-
-
1
-
-
-
-
—
1
CARE
|...
-
2
-
-
1
-
-
-
-
-
-
-
-
-
1
1
1
-
-
-
-
-
-
-
-
-
—
1
1
2
-
-
-
-
-
-
-
-
-
-
-
-
2
-
ETH
^^
-
-
-
-
-
-
-
-
-
1
1
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
—
-
—
-
-
DNREACT
1.6
1
-
-
2
-
-
1
-
-
-
-
-
1
-
-
-
-
-
-
-
~
-
-
~
-
-
~
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
—
-
          89

-------
TABLE B-2. (Continued)

Species Profiles by Bond Group
Saroad
NO.
43113
43111
45107
43110
45102
43243
43444
43119
50022
45212
45205
43201
43432
43301
43445
43801
43552
43560
43559
43209
50016
43261
43262
43272
43819
50010
43118
45801
43212
43435
43305
43238
43255
43260
43220
43303
45209
43259
43258
45101
43284
43235
50021
45211
45204
Chemical Name
ISOMERS OF TETRADECANE
ISOMERS OF TRIDECANE
ISOMERS OF TRIMETHYLBENZENE
ISOMERS OF DNDECANE
ISOMERS OF XYLENE
ISOPRENE
ISOPROPYL ACETATE
LACTOL SPIRITS
M-CRESOL (3-METHYLBENZENOL)
M-ETHYLTOLOENE
M-XYLENE
METHANE
METHYL ACETATE
METHYL ALCOHOL
METHYL AMYL ACETATE
METHYL CHLORIDE
METHYL ETHYL KETONE
METHYL ISOBUTYL KETONE
METHYL N-BDTYL KETONE
METHYLACETYLENE
METHYLANTHRACENE
METHYLCYCLOHEXANE
METHYLCYCLOPENTANE
METHYLCYCLOPENTENE
METHYLENE BROMIDE
METHYLNAPHTHALENE
MINERAL SPIRITS
MONOCHLOROBENZENE
N-BUTANE
N-BUTYL ACETATE
N-BDTYL ALCOHOL
N-DECANE
N-DODECANE
N-PENTADECANE
N-PENTANE
N-PROPYL ALCOHOL
N-PROPYLBENZENE
N-TETRADECANE
N-TRIDECANE
NAPHTHA
NONADECANE
NONANE
0-CRESOL (2-METHYLBENZENOL)
0-ETHYLTOLUENE
0-XYLENE
OLE
^m
-
-
-
-
1
-
-
-
—
-
-
-
-
—
-
-
-
-
-
-
-
-
1
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
FAR
14
13
3
11
2
1
5
8
-
3
2
-
-
1
8
-
3
5
5
1.5
9
7
6
4
-
5
7
5
4
5
4
10
12
15
5
3
3
14
13
8
19
9
-
3
2
ARO
^
-
1
-
1
-
-
-
1
1
1
-
-
-
-
-
-
-
-
-
1
-
-
-
-
1
-
-
—
-
—
—
—
—
-
-
1
-
-
—
-
-
1
1
1
GARB
M
-
-
-
-
2
-
-
1
—
—
-
-
-
-
-
1
1
1
-
-
-
-
-
-
-
1
-
-
-
-
—
—
—
-
—
—
-
—
-
-
-
1
-
-
ETH DNREACT
•» •—
- -
- -
- -
- -
- -
- -
- -
- -
— —
- -
1
3
- -
- -
1
— -
- -
- -
1.5
- -
- -
— —
— —
1
— —
— —
1
— —
1
— —
— —
— —
— —
— —
— —
— —
— —
— —
— —
— —
- —
- -
- -
— —
      90

-------
TABLE B-2. (Concluded)

Species Profiles by Bond Group
Saroad
NO.
43283
43233
43265
50023
45206
43817
45300
50006
43204
43504
43434
43205
43369
43602
50013
43208
45216
45220
43123
43309
45215
43390
45232
45202
43216
43226
45233
43824
43811
43821
43740
43822
50014
43241
43000
43860
45401
Chemical Name
OCTADECANE
OCTANE
CCTENE
P-CRESOL (4-METHYLBENZENOL)
P-XYLENE
PERCHLQROETflYLENE
PHENOLS
PROPADIENE
PROPANE
PROPR IONALDEHYDE
PROFYL ACETATE
PROPYLENE
PROPYLENE GLYCOL
PROFYLENE OXIDE
PROPYLNAPHTHALENE
PROPYNE
SEC-BOTYLBENZENE
STYRENE
TERPENES
TERT-BUTYL ALCOHOL
TERT-BOTYLBENZENE
TETFAHYEROFORAN
TETRAMETHYLBENZ ENE
TOLUENE
TRANS-2-BDTENE
•mANS-2-PENTENE
TRI/TETRAALKYL BENZENE
TRICHLOROETHYLENE
TR ICHLOROFLOUROMETHANE
TRICHLOROTR IFLOOROETHANE
TRIMETHYL AMINE
TRIMETHYLFLDOROSILANE
TRIMETHYLNAPHTflALENE
UNDECANE
UNKNOWN SPECIES
VINYL CHLORIDE
XYLENE BASE ACIDS
OLE PAR
18
8
1 6
- -
2
_
- -
1
1.5
2
4
1 1
2
2
7
2
4
1
1 8
- -
4
3
4
1
2
3
5
— -
— -
— -
3
- -
7
11
.1 4
— -
2
ARO
^_
-
-
1
1
-
-
-
—
—
-
-
-
-
1
-
1
1
-
-
1
-
1
1
-
-
1
-
-
—
—
—
1
-
.25
-
1
CARE
*_
-
-
1
-
-
-
2
-
1
-
-
1
-
-
-
-
1
-
-
-
1
-
-
2
2
-
-
-
-
-
-
—
-
.32
-
-
ETfl
—
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
1
-
-
-
-
-
-
.16
1
-
UNREACT
— _
-
-
-
-
2
6
-
1.5
-
1
-
-
1
-
1
-
-
-
4
-
-
-
-
-
-
-
-
1
2
-
3
-
-
-
-
-
     91

-------
TECHNICAL REPORT DATA
(Please read Instructions on the rcierse before completing)
1. REPORT NO. ' 2.
EPA-450/4-84-009
4. TITLE AND SUBTITLE
Technical Discussions Relating to the Use of
Carbon Bond Mechanism in OZ1PM/EKMA
7. AUTHOR(S)
J. P. Kill us and G. Z. Whitten
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Systems Applications, Inc.
101 Lucas Valley Road
San Rafael, California 94903
12. SPONSORING AGENCY NAME AND ADDRESS
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
MDAD, AMTB, Mail Drop 14
Research Triangle Park, North Carolina 27711
15 SUPPLEMENTARY NOTES
U.S. EPA Contact: Gerald L. Gipson
3. RECIPIENT'S ACCESSION NO.
5. REPORT DATE
May 1984
6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NO.
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
Contract 68-02-3570
13, TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE

16 ABSTRACT
The document discusses the use of the Carbon Bond 3 (CB-3) mechanism with the
city-specific Empirical Kinetics Modeling Approach (EKMA). Topics addressed include:
(1) a description of the CB-3 mechanism, (2) background information of the formulation
of key mechanism parameters, and (3) discussions on the treatment of initial conditions,
emissions, background ozone, and background precursors with EKMA/CB-3.
17. KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS b.lDENTIFI
ozone EKMA
control strategies OZIPP
photochemical pollutants
models
SIPs
carbon bond mechanism
18 DISTRIBUTION STATEMENT 19.SECURI
unlimited 20 SECURI
ERS/OPEN ENDED TERMS C. COSATI Held/Group

TY CLASS (This Report j 21. NO. OF PAGES
92
TY CuASS /Tins page/ 22. PRICE
EPA Form 2220-1 (R«r. 4-77)     PREVIOUS EDITION IS OBSOLETE

-------