United States
Environmental Protection
Agency
National Exposure
Research Laboratory
Research Triangle Park, NC 27711
Research and Development
EPA/600/SR-96/078 July 1996
EPA Project Summary
The University of California,
Riverside Environmental
Chamber Data Base for
Evaluating Oxidant Mechanisms:
Indoor Chamber Experiments
Through 1993
William P.L. Carter, Dongmin Luo, Irina L. Malkina, and Dennis Fitz
Data collected in the environmental
chambers operated by the University
of California, Riverside, have been
documented and are now available for
use by the scientific community for
evaluating photochemical mechanisms
for urban and regional airshed mod-
els. The compiled data include experi-
ments performed in the Statewide Air
Pollution Research Center (SAPRC)
Evacuable Chamber (EC), Indoor Teflon
Chamber #1 (ITC), Indoor Teflon Cham-
ber #2 (ETC), Dividable Teflon Cham-
ber (DTC), and Xenon arc Teflon Cham-
ber (XTC) between September of 1975
through November of 1993. The ex-
periments contained in this data base
are listed and summarized, the facility
and procedures employed are de-
scribed, and the analytical and moni-
toring methods and their calibration
data and associated uncertainties are
documented. In addition, input data
needed to conduct model simulations
of the experiments in the data base
are included, and the format of the
data sets, which are available on the
Internet by anonymous FTP, are de-
scribed. Also available on the Internet
are files that permit modeling of the
experiments in the data base using
the SAPRC-90 and the Carbon Bond IV
chemical mechanisms. Recommenda-
tions are made concerning the steps
that need to be taken before using
these data to evaluate chemical mecha-
nisms.
This Project Summary was developed
by EPA's National Exposure Research
Laboratory, Research Triangle Park, NC,
to announce key findings of the re-
search project that is fully documented
in a separate report of the same title
(see Project Report ordering informa-
tion at back).
Introduction
Urban and regional oxidant models are
important tools in the development and
assessment of regulatory strategies aimed
at reducing ground-level ozone formation.
The gas-phase photochemical mechanism
is an important component of such mod-
els because ozone is not emitted directly,
but is formed from the gas-phase photo-
chemical reactions of the emitted volatile
organic compounds (VOCs) and oxides of
nitrogen (NOX) in air. The chemistry of
ground level ozone formation is highly
complex and nonlinear and has many un-
certainties. Because of this, no chemical
model can be relied upon to give even
approximately accurate predictions unless
it has been evaluated by comparing its
predictions with experimental data.
There are essentially two ways a photo-
chemical oxidant model can be evaluated.
The first is to compare the predictions of
the complete model against data taken
during an historic ozone pollution episode.
However, this is rarely definitive because
of the many uncertainties in the input data
needed to represent any historic episode.
Furthermore, it is rarely clear which com-
ponent of the many components of com-
prehensive airshed models is the source
of any discrepancy observed, or, if the
model agrees with the data, whether there
may be compensating errors among the
different components.
-------
The other approach for evaluating mod-
els is to evaluate each of the components
separately. In the case of the gas-phase
chemical mechanism, this means evaluat-
ing the predictions of the mechanism
against results of environmental chamber
experiments. If the model can success-
fully predict the transformations under a
range of chemical conditions which en-
compass the range of variability in the
atmosphere, one has at least some basis
to conclude that the model may give rea-
sonably accurate predictions of chemical
transformations in the atmosphere, if pro-
vided with the appropriate input data.
However, modeling environmental
chamber data is not without significant
uncertainties. Analytical methods for the
reactants and products have inaccuracies
and imprecisions which might introduce
errors in the amount of initial or injected
reactants assumed in the model, or in
evaluating the extent to which the model
can predict pollutant concentrations. Simu-
lating a chamber experiment also requires
knowledge of the temperature, light inten-
sity, and spectrum of the photolyzing light,
and how they vary with time. All of these
have uncertainties that can result in errors
in reaction rates when simulating the ex-
periments. Characterizing the light inten-
sity and spectra is particularly difficult in
outdoor chamber experiments. Because
of this, it is important that chamber experi-
ments used to evaluate mechanisms have
as high quality analytical data and as well
characterized experimental conditions as
possible, and that the sources of the un-
certainties in these data are identified and
quantified. With a knowledge of the range
of uncertainty in the input or evaluation
data, one can assess the extent to which
a model agrees with the data.
Perhaps the most serious problem is
the existence of chamber wall effects (het-
erogeneous processes involving the walls)
which are known to be non-negligible in
all current-generation chamber experi-
ments and can dominate the results of
certain types of experiments. Because of
this, one should not rely on data from a
single chamber for evaluating mechanisms;
the use of data from a variety of cham-
bers is necessary to minimize the chance
for errors in the chamber model causing
errors in model simulations of experiments
where the gas-phase chemistry is being
evaluated.
With these problems in mind, the Uni-
versity of North Carolina (UNC), the State-
wide Air Pollution Research Center
(SAPRC) at the University of California at
Riverside (UCR), and several other re-
search institutions initiated an effort that
was sponsored by the U.S. Environmental
Protection Agency to develop a data base
and protocol for evaluating photochemical
mechanisms for air quality simulation mod-
els. An overview of the many factors which
must be considered when developing such
a protocol is discussed in the document
entitled "Protocols for Evaluating Oxidant
Mechanisms for Urban and Regional Mod-
els" (Jeffries et al., 1992), which was pre-
pared under this program. As discussed
there, the specifics of the evaluation pro-
tocol will depend on the nature of the
model being evaluated and the uses to
which it will be put. However, regardless
of these application-specific considerations,
the data base of chamber experiments,
and the comprehensiveness of its associ-
ated documentation, are critical to any
evaluation. Sufficient information must be
given concerning the experiments so the
evaluator can represent them in the model
appropriately and can understand and
document their uncertainties and variabili-
ties. The report of Jeffries et al. (1992)
gives standards for the necessary docu-
mentation of the data base.
The report described in this Project Sum-
mary documents the results of SAPRC
efforts to develop a data base of environ-
mental chamber experiments for mecha-
nism evaluation. The Project Report is
composed of two volumes. The first vol-
ume serves as the backing document for
the SAPRC environmental chamber data
base. That volume describes the proce-
dures employed in the experiments and
the analytical methods used to generate
the data and the calibration and other
data which can be used to estimate their
uncertainties. It also contains a compre-
hensive discussion of the light intensity
and light spectrum information, discusses
and evaluates the temperature and other
characterization data necessary to estab-
lish run conditions for modeling, discusses
corrections made to the data as a result
of reevaluations under this program, gives
recommendations for modeling these runs,
and describes the formats of the distrib-
uted data sets. The second volume con-
tains the data sets themselves. In addition
to the environmental chamber data, this
includes spreadsheets giving run condi-
tion and data summaries, data sets giving
supporting information relevant to data
quality, and programs and data which can
be employed for mechanism evaluation.
Scope
Coverage
To date, six different environmental
chambers at SAPRC have provided data
which are potentially useful for mecha-
nism evaluation. These include the
Evacuable Chamber (EC), the Indoor
Teflon Chamber #1 (ITC), the Indoor Teflon
Chamber #2 (ETC), the Dividable Teflon
Chamber (DTC), the Xenon Teflon Cham-
ber (XTC), and the Outdoor Teflon Cham-
ber (OTC). These chambers are described
in the backing documentation and are sum-
marized in Table 1. Data from the indoor
chambers are included in this release of
the data base, and data from the outdoor
chamber will be included in a later re-
lease. The DTC and XTC have since been
moved to another location and are cur-
rently generating additional data. The data
base will be updated to include these re-
Although the presently distributed data
base only includes indoor chamber data,
the backing documentation contains a de-
scription of all six chambers and their op-
erating procedures, and discusses the
analytical methods and calibration data appli-
cable for all groups of runs listed in Table 1.
Not all environmental chamber experi-
ments carried out in the chambers listed
in Table 1 are considered appropriate for
this data base. The main criteria used
when determining whether to include a
run in the data base were as follows: (1)
the data from the experiment must be in
our computer data base; (2) the experi-
ments should either have been carried
out for the purposes of mechanism evalu-
ation or characterization of chamber ef-
fects for mechanism evaluation, or they
should be judged to be potentially useful
for such purposes; and (3) the experi-
ments were carried out under conditions
which are sufficiently well characterized
for modeling. A number of experiments
which passed the initial screening and thus
were included in the data base were sub-
sequently found to have problems which
limit or prevent their use for mechanism
evaluation. Therefore, not all runs in the
data base are considered useful for
mechanism evaluation.
-------
The total number of experiments from
each chamber in the current data base
and the number of experiments consid-
ered to be potentially useful for mecha-
nism evaluation are indicated in Table 1.
Some of these experiments have higher
data quality than others, and the summa-
ries distributed with the data base indicate
runs which have problems and include
estimates of uncertainties in reactant con-
centrations and run conditions.
The run-specific data which were evalu-
ated and documented in this project are
sufficient for using them to evaluate a
mechanism's performance in predicting
ozone formation. The data that were evalu-
ated most comprehensively include the
NO and NO2 data; organic reactant data;
ozone data; the light intensity and spec-
tral distribution; the temperature and hu-
midity conditions, and information concern-
ing special run conditions. In addition, data
for PAN, formaldehyde, and other organic
products were evaluated to some extent,
but not as comprehensively as for those
listed below. Mechanism evaluators can
utilize the calibration summaries and other
information contained in the Project Re-
port and the data sets to make conclu-
sions concerning the data for individual
runs of interest.
Computer Data Sets
The computer data sets prepared for
this program are available on the Internet
for anonymous FTP at carterpc.ucr.edu in
the directory "/chdata", and also at
cert.ucr.edu in the directory "pub/carter/
chdata". The format of the data sets and
instructions on how to install them on PC-
compatible computers are discussed in
the Project Report. The distributed data
sets include the following.
Master Run Summary
Spreadsheets
For each chamber whose data are in-
cluded in this distribution, there is a series
of Excel 5 (and ASCII CSV) spreadsheet
files giving important summary informa-
tion for each run in the data base. This
includes, for each run in the data base,
the run descriptions and classification, ini-
tial reactant concentrations and their esti-
mated uncertainties, notations where ap-
plicable of special run conditions or prob-
lems, notations where applicable of spe-
cial problems with the run data, notations
where applicable indicating runs judged
not suitable for modeling, information con-
cerning light intensity and spectral distri-
bution assignments, and other data that
might be of significance to the modeler.
Individual Experiments Data
Sets
For each experiment, there is an ASCII
file containing all the run-specific data. In
addition to the experimental measurements
as a function of time, it includes (where
available) identifications of instruments and
calibration and zero corrections, uncer-
tainty estimates, recommended input data
for modeling the runs, comments from the
log book and comments documenting data
corrections and special run conditions and
input data to use when modeling. These
contain all the run-specific information
needed for conducting model simulations
and evaluating model performance for all
runs which are sufficiently well character-
ized for modeling.
Calibration Data Summary
Files.
For most of the instruments used to
monitor O3, NOx or organics, there are
calibration data files giving the calibration
data and the times the instruments were
calibrated, codes indicating how recom-
mended calibration factors are to be com-
puted, and relevant comments. Computer
programs to read these files and produce
output summary files are also included.
Table 1. Summary ofSAPRC Environmental Chambers
Chamber
Runs [a]
Tot Model
Description
EC 216 160 5800-liter evacuable, thermostatted cylindrical chamber. Teflon-coated aluminum walls. Quartz
windows on both ends. 25 KWXenon arc light "solar simulator" light source with pyrex filters to remove
UVbelow~290nm. Generally operated at-303 K and 50% RH. Experiments carried out between 9/29/75 and
11/18/83 are on the present data base.
ITC 346 329 Replaceable~6000-liter FEP Teflon bag in aluminum frame banks of blacklights on either side. Gener-
ally operated at room terperature and 50% RH. Experiments carried out in the periods 1/29/82-8/29/86
and 10/2/89-10/16/89 are on the present data base.
ETC 413 384 Replaceable~4000-literFEP Teflon bag in an aluminum frame with banks ofblacklights on the top and bottom.
All runs at room temperature re and ~5% RH. Experiments carried out between 10/25/89 and 2/25/93 are
on the present data base.
DTC 132 128 Dual Replaceable~5000-liter FEP teflon bags located next to each other, between banks ofblacklights on
either side. Allows for simultaneous irradiation of two mixtures. All runs at room temperature and all but 2 at
-5% RH. Experiments between 3/5/93 through 8/4/94 are on the present data base.
XTC 31 31 Replaceable~50,000-liter pillow-shaped FEP teflon reaction bag located in a room with reflective walls and
with 4 6.5 KW xenon arc lights on a wall 4' from the bag. Uses same enclosure as DTC, and run numbering
continues from that of DTC runs. All runs at room temperature and -5% RH. Experiments between 8/23/93
and 11/17/93 are on the present data base.
OTC (not incl.) Replaceable~50,000-liter pillow-shaped FEP teflon reaction bag located outdoors and irradiated with
sunlight. No OTC experiments are on the present data base, though the backing document discusses most
of the analytical procedures employed. Experiments from the periods 6/22/83-10/14/83, 5/13/85-11/27/85
and 6/10/92-10/26/93 may be included in future releases of the data base.
[a] "Tot" = total number of experiments whose data are in the distribution. "Model" = number of experiments with sufficiently well
characterized conditions to be potentially useful for modeling.
-------
NO2 Actinometry Results Data
Files
The results of all the NO2 actinometry
experiments, the primary method used to
measure light intensity, are included. The
modeler can use these to recompute the
light intensities if it is judged that the
method or rate constants used in this work
are inappropriate.
Recommended Spectral
Distributions
Data files are included giving spectral
distributions we recommend using for cal-
culating photolysis rates when modeling
these experiments.
Computer Programs
A number of FORTRAN computer pro-
grams that can read and process these
files are included to assist the modeler in
using these data. These include programs
which can be used to conduct model simu-
lations of the experiments and plot and
summarize the results. The source files
for these programs are available on a
separate distribution that can be found on
the Internet for anonymous FTP at
carterpc.ucr.edu in the directory "/model"
or at cert.ucr.edu in the directory "/pub/
carter/model". All executable files require
a 386 PC-compatible computer with a math
co-processor or a 486DX or better. The
programs are not fully documented, but
sample input files with comments and
batch files giving examples of these are
included.
Chemical Mechanism
Implementation Files and
Software
Computer files implementing the
SAPRC-90 and Carbon Bond IV mecha-
nisms are included so example model
simulations of the experiments can be car-
ried out. Batch files showing examples of
their use are included. The evaluator can
use these as guidelines as how these
data can be used for mechanism evalua-
tion.
Discussion and
Recommendations
This project has provided important in-
put data for the evaluation of chemical
mechanisms for use in urban and regional
oxidant models. Although many of the
chamber experiments whose data have
been compiled under this project have
been used in previous mechanism evalu-
ations, a number of potentially significant
corrections have been made to these data,
and many other potentially important ex-
periments whose results have not been
available previously have been included
in this data base.
However, it is essential that the mod-
eler recognize the limitations of this data
base. In the first place, the present distri-
bution includes data only from SAPRC
experiments carried out through the end
of 1993 and does not include the exten-
sive body of SAPRC outdoor chamber
experiments, nor the experiments we are
currently carrying out for mechanism evalu-
ation and VOC ozone reactivity assess-
ment. We expect to include these in fu-
ture distributions, to be prepared as a part
of our ongoing environmental chamber pro-
grams. In addition, UCR is not the only
laboratory where environmental chamber
experiments useful for mechanism evalu-
ation have been carried out. A large data
base of University of North Carolina (UNC)
outdoor chamber experiments is now be-
ing compiled by the UNC researchers as
part of the overall effort of preparing an
environmental chamber data base for the
EPA. Environmental chamber experiments
that are potential candidates for this data
base have also been or are being carried
out at other laboratories. As discussed by
Jeffries et al. (1992), a comprehensive
and complete mechanism evaluation re-
quires the use of data not only from differ-
ent environmental chambers, but also from
different research groups.
It is important that users of environmen-
tal chamber data recognize the many prob-
lems that can be associated with chamber
experiments which might affect the accu-
racy of the results or the modeler's ability
to accurately characterize the conditions
of the experiment. When preparing this
data base, we attempted to identify and
note all problems which might affect the
use of the experiment (or particular mea-
surements associated with it) for mecha-
nism evaluations. However, because of
the large number of runs involved, com-
bined with the number of things that can
go wrong, we cannot guarantee that all
problems have been detected and noted.
Therefore, if the model simulation is in
gross disagreement with the results of an
experiment, the possibility that the prob-
lem exists with the data and not the model
cannot be totally ruled out. In this case, it
is important that the modeler has some
understanding of the limitations associ-
ated with the data or run conditions. One
of the objectives in preparing the backing
documentation is to help the modeler ob-
tain this understanding.
References
Jeffries, H.E., M.W. Gery, and W.P.L.
Carter. Protocols for Evaluating Oxidant
Mechanisms for Urban and Regional
Models. EPA/600/R-92/112, U.S. Envi-
ronmental Protection Agency, Research
Triangle Park, NC 1992. 89pp.
-------
William P. L Carter, Dongmin Luo, Irina L Malkina, and Dennis Fitzare with the College of Engineering
Center forEnvironmentalResearch and'Technology, University of California, Riverside, CA 92521.
Marcia C. Dodge is the EPA Project Officer (see below).
The complete report consists of two volumes entitled "The University of California, Riverside
Environmental Chamber Data Base for Evaluating Oxidant Mechanisms: Indoor Chamber Experi-
ments Through 1993"
"Volume I. Backing Document and Data Base Description " (Order No. PB96-190 673; Cost: $44.00,
subject to change)
"Volume II. Appendices" (Order No. PB96-190681; Cost: $31.00, subject to change)
The above reports 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:
National Exposure Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
United States
Environmental Protection Agency
National Risk Management Research Laboratory (G-72)
Cincinnati, OH 45268
Official Business
Penalty for Private Use $300
BULK RATE
POSTAGE & FEES PAID
EPA
PERMIT NO. G-35
EPA/600/SR-96/078
------- |