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 methanei.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 concentrationanother 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
-------
ParaffinsNitrate 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 speciesCARB. 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
11 ,_
> 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-)
TO
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 ppmin 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 ModelThe 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
------- |