United States
Environmental Protection
Agency
Environmental Sciences Research
Laboratory
Research Triangle Park NC 27711
Research and Development
EPA-600/S3-83-086 Dec. 1983
Project Summary
Evaluation of Chemical Reaction
Mechanisms for Photochemical
Smog: Part 1. Mechanism
Descriptions and Documentation
Gregory J. McRae, Joseph A. Leone, and John H. Seinfeld
Over the past ten years or so a great
deal of effort has been devoted to
developing chemical reaction mechanisms
for photochemical air pollution. Because
the actual number of atmospheric
organic species is too large for the
detailed chemistry of each to be
included in a mechanism, it has been
necessary to reduce the number of
organic species to a manageable set by
a process referred to as lumping. The
manner in which this lumping has been
carried out constitutes one of the major
differences among existing mechanisms.
It has recently been demonstrated that
different chemical mechanisms predict
different degrees of hydrocarbon and
NOX control to achieve the same level of
ozone reduction under identical condi-
tions. Because of the necessity of using
reaction mechanisms for photochemical
smog in determining air pollution
control strategies, these results point to
a serious need to analyze the funda-
mental behavior of such mechanisms
and to understand the key elements of
their behavior. Such an analysis is the
subject of this two part report. The
current volume. Part I, contains a
detailed description of six mechanisms
that have been developed to describe
photochemical smog chemistry, includ-
ing analyses of the treatments of the
basic chemistry, of photolysis reactions
and organic lumping in initial conditions
and rate constants. It is found that the
mechanisms differ in virtually all
aspects. Part II is devoted to a detailed
analysis of the behavior of each of the
mechanisms.
This Project Summary was developed
by EPA's Environmental Sciences Re-
search Laboratory, Research Triangle
Park, NC, to announce key findings of
the research project that is fully docu-
mented in a separate report of the same
title (see Project Report ordering in-
formation at back).
Introduction
Over the past ten years or so, a great
deal of effort has been devoted to
developing chemical reaction mechanisms
for photochemical air pollution, and, as a
result, several mechanisms exist at the
present time. Since the atmosphere
contains literally -scores of hydrocarbon
species, it is virtually impossible to write a
reaction mechanism that includes the
detailed chemistry of each hydrocarbon
species present. Although each mecha-
nism is based, more or less, on the same
body of experimental kinetic data, the
manner of treatment of the hydrocarbon
chemistry varies among them. Because
the actual number of hydrocarbon species
is, as we have noted, too large for the
detailed chemistry of each to be included
in a mechanism, it has been necessary to
reduce the number of hydrocarbon
species to a manageable set by a process
referred to as lumping. The manner in
which this lumping has been carried out
constitutes one of the major differences
among existing mechanisms.
The differences among the several
existing reaction mechanisms would not
be of concern if their predictions were in
essential agreement over the range of
conditions of interest for atmospheric
-------
predictions. It has recently been demon-
strated, however, that different mecha-
nisms predict rather different degrees of
hydrocarbon and NO, control to achieve
the same level of ozone reduction under
identical conditions. Because of the
necessity of using reaction mechanisms
for photochemical smog in determining
air pollution control strategies, these
results point to a serious need to analyze
the fundamental nature of such mecha-
nisms and to understand the key elements
of their behavior. Such an analysis is the
subject of this work.
Lumped Photochemical
Mechanisms
Figure 1 depicts the ways in which
hydrocarbon lumping has been approached
in photochemical smog mechanisms. In
the surrogate species approach, typified
by the mechanism of Dodge (1977), the
entire atmospheric mixture is represented
by a small number of surrogate hydrocarbon
species. In the mechanism of Dodge
(1977), propene and n-butane, together
with small quantities of formaldehyde
and acetaldehyde, represent the entire
atmospheric mix.
The other lumping approaches are
indicated in Figure 1. Lumped molecule
approaches refer to those in which the
atmospheric hydrocarbons are lumped
into identifiable molecular species. For
example, in the surrogate species lumped
molecule approach, certain hydrocarbons
serve as surrogate species for an entire
group of actual species. For example, n-
butane can be used to represent all
atmospheric alkanes, and the alkane
portion of the full mechanism then
consists of the detailed explicit chemistry
of n-butane. Thus, a mixture of many
alkanes is represented by a comparable
concentration of n-butane. A mechanism
based on this approach is that of Atkinson
et al. (1982). We should point out that the
mechanism of Dodge (1977), referred to
above, is, in essence, also a surrogate
species mechanism. However, the n-
butane and propene are not necessarily
identified with alkane and alkene species,
respectively. Rather, propene and n-
butane represent the entire atmospheric
mix, where only the total concentration of
non-methane hydrocarbons in parts per
million by carbon is matched to the initial
concentrations of propene and n-butane.
The other type of lumped molecule
approach is that of generalized species
lumping, in which an entire group of
compounds is represented by a generalized
species, the chemistry of which reflects
the common features of that of the whole
Photochemical Reaction
Mechanisms
Surrogate Species
- Dodge (1977)
Lumped Molecule
Lumped Structure
- Killus and Whitten
(1982)
Generalized Species
Used to Represent
Chemistry of Each
Lumped Class
- Demerjian (1982)
- McRae and Seinfield (1983)
- Penner and Walton (1982)
Surrogate Species
Used to Represent
Chemistrfof Each
Lumped Class
- Atkinson et al(1982)
Figure 1. Classification of photochemical reaction mechanisms used in evaluation.
group. For example, alkanes could be
represented by a species called ALKANE,
whose rate constants and reaction
mechanisms are, in some manner, an
average of those of all alkanes. Mechanisms
in this class include those of McRae and
Seinfeld (1983), Demerjian (1982), and
Penner and Walton (1982).
Finally in the lumped structure approach,
lumped species represent various classes
of structural units, such as, single-
bonded carbon atoms, double-bonded
carbon atoms, and carbonyl carbon
atoms. An initial mixture of organics is
therefore apportioned by bond type rather
than by molecule type to obtain the
lumped species. The one mechanism of
this type is that of Killus and Whitten
(1982). The Killus and Whitten mechanism,
referred to as the Carbon Bond mechanism,
treats the reactions of six different types
of carbon atoms: 1 -alkene carbon atoms
except ethene, ethene, single bonded
carbon atoms, reactive aromatic rings,
carbonyl carbon atoms including carbon
atoms from internal olefins, and cc-
dicarbonyls.
It is important to point out that any
reaction mechanism for the atmospheric
chemistry of photochemical smog must
involve some aspect of hydrocarbon
lumping. It is sometime mistakenly
assumed that lumping is not involved in a
surrogate mechanism, since a surrogate
mechanism consists of the explicit
chemistry of the surrogate species.
Although the mechanism itself contains
only explicit chemical steps, not involving
any generalized species, the representation
of an atmospheric mixture requires that
many species not explicitly included in
the mechanism be apportioned to the
surrogate species. It is, in fact, the way
in which species are apportioned that
really constitutes the differences among
mechanisms, although mechanisms do
differ in the values of rate constants used
and in the importance of mechanistic
steps where the available experimental
information is open to interpretation.
Procedure
The particular mechanisms chosen for
initial testing in this study are shown in
Figure 1 and Table 1. Several different
criteria were used to select representative
mechanisms including: the availability of
support! ng docu mentation and the degree
of testing against smog chamber experi-
ments. One additional requirement was
that each mechanism had to have been
implemented in an air quality model and
used in emissions control calculations.
With this background the following
material was requested from each
investigator responsible for the develop-
ment of the reaction mechanism: (1) All
the available open and report literature
references that describe the scientific
basis and testing of the reaction mechanism;
(2) A computer listing of the reactions,
species names and a duplication of a test
case. This latter information was requested
to ensure that the mechanism has been
-------
Table 1. Photochemical Reaction Mechanisms Considered for Evaluation — Their Characteristics and Their Validation Base
Reaction Mechanism
Atkinson et al.
(1982)c
Number of
Reactions/
Type" Species
LM 81/52
Number of
Organic
Classes"
14
Validation Basis
Experiment
Numbers"
EC -178
EC- 143
EC-146
EC-216
EC-340
EC-344
EC-161
EC-237.EC-242.EC-246
Type of
Experiment
n-butane/NOi
ethene/NOi
tr-2-butene/NO*
propene/NO,
toluene/NO*
m-xylene/NOn
4 alkenes/NO*
7 hydrocarbons/ NO*
AGC-119AGC-133AGC-134.
AGC- 135.AGC- 138AGC- ISO,
AGC-156
AP-28AP-30.AP-35AP-37
Demerjian (1982)
LM
45/30
Dodge (1977)
Kit/us and Whitten
(1982)
LS
76/39
79/41
McRae and Seinfeld (1983)
Penner and Walton
(1982)
LM
LM
52/32
59/22
EC-231,EC-232.EC-233.
EC-237,EC-238.EC-242.
EC-243,EC-245,EC-246.
EC-247
EC-237
SUR-119J,SUR-121J,
SUR- 126J.SUR- 132J.
SUR-133J.SUR-134J
EC-231 .EC-232.EC-233,
EC-237.EC-238.EC-241.
EC-242.EC-243.EC-245
surrogate (murti-
hydrocarbonJNO*
hydrocarbon/NOi/SOz
40 Bureau of Mines smog
chamber experiments with
dilute auto exhaust and
added /VO«. Several UCR
aromatic hydrocarbon/7VOX
experiments
17 Bureau of Mines smog
chamber experiments with
dilute auto exhaust and
added NO*
7 hydrocarbons/NO *
experiments at SAPRC
2 day urban hydrocarbon
mix at UNC
7 hydrocarbons/NO*
surrogate atmospheric
mix
several individual
hydrocarbon/NO,
experiments
7 hydrocarbons/NO*
BThe mechanism type refers to: LM - Lumped molecule, S - Surrogate and LS - Lumped structure.
"Number of reactive organic groupings for which either emissions or initial conditions must be specified in practical applications.
cThe mechanism presented here corresponds to that in the most recent reference cited.
"This mechanism was developed by Falls and Seinfeld (1978), with updated rate constants presented by McRae and Seinfeld (1983).
"The experiment numbers refer to the following laboratories:
EC.SUR Statewide Air Pollution Research Center (SAPRC) of
the University of California, Riverside
UNC University of North Carolina
implemented in the manner intended by
the developer. A summary of the mechanism
characteristics, documentation sources
and bases for validation is shown in Table
1. The validation bases of the Atkinson et
al. (1982), Killus and Whitten (1982) and
Penner and Walton (1982) mechanisms
are experiments conducted at the University
of California, Riverside, whereas the
Demerjian (1982) and Dodge (1977)
mechanisms were evaluated on Bureau
of Mines smog chamber data. The Killus
and Whitten (1982) mechanism has also
been tested against outdoor smog chamber
data from the University of North Carolina.
-------
Chapter 2 contains summaries of the
mechanisms.
Results
We discuss the basis of the lumping
approaches of the five lumped mechanisms:
Atkinson et al. (1982), Demerjian (1982),
Killus and Whitten (1982), McRae and
Seinfeld (1983), and Penner and Walton
(1982). Our discussion of the chemistry of
photochemical smog is based almost
exclusively on the extensive review of
Atkinson and Lloyd (1983). We then
examine the specification of photolysis
rates and initial conditions in each of the
mechanisms.
Conclusions
This report has presented a detailed
description of six lumped reaction
mechanisms for photochemical smog:
Atkinson et al. (1982)
Demerjian (1982)
Dodge (1977)
Killus and Whitten (1982)
McRae and Seinfeld (1983)
Penner and Walton (1982)
The description includes a discussion of
the basic assumptions in each mechanism's
treatment of the fundamental chemistry,
photolysis reactions, and organic lumping
via initial conditions and organic rate
constants. In general, it is found that the
mechanisms differ in virtually all aspects,
and, without detailed, quantitative,
numerical comparisons, it is difficult to
predict simply from inspection how each
mechanism will perform in a particular
application. Nevertheless, if one removes
those mechanism aspects included only
to account for smog chamber-dependent
radical sources, such as a wall source of
OH radicals or initial HONO, it is possible
to rank the mechanisms according to
overall "reactivity." We find on the basis
of our analysis that such a ranking is:
Killus and Whitten (1982)
Demerjian (1982)
McRae and Seinfeld (1983)
Atkinson et al. (1982)
Penner and Walton (1982)
most
reactive
less
reactive
The Dodge (1977) mechanism cannot be
ranked in this scale because the split
between propene and n-butane is fixed.
Our object is to be able to tell specifically
what aspects of each mechanism are key
to its performance, as measured by its
"reactivity." To do so requires the de-
tailed, quantitative, numerical comparisons
referred to above. Such comparisons are
the subject of Part II.
References
Atkinson, R. and Lloyd, A.C. (1983), "Eval-
uation of Kinetic and Mechanistic
Data for Modeling of Photochemical
Smog," J. Phys. Chem. Ref. Data (in
press).
Atkinson, R., Lloyd, A.C. and Winges, L
(1982), "An Updated Chemical Mechan-
ism for Hydrocarbons/NOx/SOa
Photooxidations Suitable for Inclusion
in Atmospheric Simulation Models,"
Atmospheric Environment, 16, 1341-
1355.
Demerjian, K.L (1982), "Personal Com-
munication."
Dodge, M.C. (1977), "Combined Use of
Modeling Techniques and Smog
Chamber Data to Derive Ozone —
Precursor Relationships," U.S. Envi-
ronmental Protection Agency Report,
EPA-600/3-77-001a, 881-889.
Falls, A.M. and Seinfeld, J.H. (1978),
"Continued Development of a Kinetic
Mechanism for Photochemical Smog,"
Environmental Science and Technolo-
gy, 12. 1398-1406.
Killus, J.P. and Whitten, G.Z. (1982), A
New Carbon-Bond Mechanism for
Air Quality Modeling, U.S. Environ-
mental Protection Agency Report No.
EPA-600/3-82-041.
McRae, G.J. and Seinfeld, J.H. (1983),
"Development of a Second-Generation
Mathematical Model for Urban Pollu-
tion II: Model Performance Evalua-
tion," Atmospheric Environment, 17,
501-523.
Penner, J.E. and Walton, J.J. (1982) Air
Quality Model Update, Lawrence
Livermore Laboratory Report UCID -
19300, Lawrence Livermore National
Laboratory, University of California,
Livermore, California, 55 pp.
G. J. McRae, J. A. Leone, and J. H. Seinfeld are with the California Institute of
Technology, Pasadena, CA 91125,
Marc/a C. Dodge is the EPA Project Officer (see below).
The complete report, entitled "Evaluation of Chemical Reaction Mechanisms for
Photochemical Smog: Part I. Mechanism Descriptions and Documentation."
(Order No. PB 83-263 251; Cost: $16.00, subject to change) will be available
only from:
National Technical Information Service
5285 Port Royal Road
Springfield. VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Environmental Sciences Research Laboratory
U. S. Environmental Protection Agency
Research Triangle Park, NC 27711
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
irUS GOVERNMENT PRINTING OFFICE 1983-659-017/7237
Official Business
Penalty for Private Use $300
------- |