United States
Environmental Protection
Agency
Atmospheric Sciences
Research Laboratory
Research Triangle Park NC 27711
Research and Development
EPA/600/S3-84/115 Mar. 1985
<>EFy\ Project Summary
Outdoor Chamber Study to
Test Multi-Day Effects:
Volumes I, II, and
William P. L Carter, Margaret C. Dodd, William D. Long, Roger Atkinson, and
Marcia C. Dodge
A series of single- and multi-day
indoor and outdoor environmental
chamber experiments have been carried
out at the Statewide Air Pollution
Research Center (SAPRC) at the Uni-
versity of California in Riverside to
derive data suitable for testing chemical
models for multi-day photochemical air
pollution episodes. Two environmental
chambers were used during this program,
-6400-/ indoor Teflon chamber with
blacklight irradiation and a ~50,000-/
dual-mode outdoor Teflon chamber
that employed natural sunlight as the
light source. A total of 32 indoor and 55
outdoor chamber experiments were
completed during this program. These
chamber experiments consisted pri-
marily of multi-day NO»-air irradiations
of an eight-component hydrocarbon
surrogate designed to represent emis-
sions of reactive organics into urban
atmospheres from all sources, and
associated control and characterization
runs. Most of the multi-day surrogate-
NO«-air experiments in the outdoor
chamber were run with the chamber in
the dual mode, where each experiment
consisted of simultaneous irradiation of
two different mixtures under the same
temperature and lighting conditions. In
addition, an isobutene-NOx-air irradia-
tion was carried out in the indoor
chamber to test chemical computer
models for the NO,-air reactions of this
surrogate component.
Problems with side inequivalency in
the dual-mode outdoor chamber exper-
iments were encountered in preliminary
experiments carried out in 1982. These
were subsequently resolved, and good
side-to-side equivalency was obtained
in the experiments carried out in 1983.
During that time, irradiations of a
"standard" surrogate-NOx-air mixture
were carried out under a variety of
temperature and light intensity condi-
tions, allowing the combined effects of
these to be determined. Good repro-
ducibility of experiments carried out
under similar weather conditions was
observed.
The multi-day surrogate-NOx-air
irradiations were carried out in both
chambers at a variety of initial hydro-
carbon and NOX concentrations. In
general, it was found that runs that
were less reactive on the first day with
respect to O3 formation tended to be
more reactive on subsequent days, and
vice versa. As a result the maximum Oa
levels in multi-day runs were much less
dependent on initial reactant concen-
trations than expected on the basis of
results of single-day irradiations.
This Project Summary was developed
by EPA's Atmospheric Sciences Re-
search Laboratory. Research Triangle
Park. NC, to announce key findings of
the research project that is fully
documented in a separate report of the
same title (see Project Report ordering
information at back).
Introduction
Multi-day air pollution episodes are
characteristic of most urban airsheds,
especially those in the eastern and
northeastern United States and the
California South Coast Air Basin. Airshed
computer models must thus incorporate
chemical packages designed specifically
for such multi-day conditions. During the
-------
last decade there have been significant
advances in our understanding of the
chemistry of photochemical air pollution,
resulting in the development of chemical
mechanisms that have been tested
against environmental chamber data and
that are currently in use in airshed
computer models. However, these chem-
ical models have been validated only for,
and thus are applicable only to, single-
day conditions. Since the development of
chemical models requires the availability
of environmental chamber data against
which these models can be tested, the
Atmospheric Sciences Research Labora-
tory of the U.S. Environmental Protection
Agency contracted the Statewide Air
Pollution Research Center (SAPRC) at the
University of California, Riverside, to
conduct a series of indoor and outdoor
chamber experiments simulating multi-
day conditions using a hydrocarbon
mixture designed to simulate that present
in urban airsheds.
Procedure
Two different environmental chambers
were employed in this program: a
~50,000-/ outdoor all-Teflon chamber
and a ~6400-/indoor all-Teflon chamber.
These two chambers were employed
because each has its own unique set of
advantages. The outdoor chamber has a
lower surface/volume ratio and has more
realistic lighting conditions, while indoor
chamber experiments can be carried out
under more controlled conditions. Thus
the indoor and outdoor chamber experi-
ments are complementary, allowing a
more thorough and comprehensive
testing of the chemical models.
Both environmental chambers em-
ployed in this program were constructed
from replaceable FEP Teflon film that
could collapse as samples were withdrawn
for analyses, thus avoiding dilution of
their contents during the experiments.
The ~50,000-/outdoor chamber could be
operated in either single or dual mode. The
majority of the experiments conducted in
the outdoor chamber in this program
were carried out in the dual mode, and
hence two separate experiments were
simultaneously run with the same
lighting and temperature conditions. The
indoor chamber was operated solely in
the single mode.
In all experiments, the reactants were
injected prior to the beginning of the
irradiation. In the outdoor runs, the
irradiations generally were initiated by
uncovering the chamber around 1000
PST, and the chamber was covered at
around 1500 PST in the evening (to
avoid differential irradiation of the
chamber sides caused by shadows). On
subsequent days of multi-day irradiations,
this chamber was uncovered at 0900
PST. For the indoor chamber runs, the
lights were turned on for 12 h at a
constant intensity (corresponding to an
NOa photolysis rate of 0.30 min~1, as
determined by separate actinometry
experiments), alternating with 12 h of
darkness to simulate nighttime. Ozone,
NO-NO2-NO* temperature and (for the
outdoor runs) UV intensity and the NO2
photolysis rate were monitored contin-
uously; and organic reactants and
selected products were monitored per-
iodically by gas chromatography during
the daytime.
Experiments Carried Out
The principal experimental runs in this
program consisted of irradiations of
"urban surrogate"-NO«-air mixtures,
which were carried out for 2 to 4 days
each, with NOX being injected on sub-
sequent days for some of these experi-
ments to simulate NO, emissions into an
aged air mass, and to allow renewed
photochemical 03 production to occur.
The majority of these experiments were
carried out to determine the effects of
varying the initial reactant concentra-
tions, with (in the case of the dual-mode
outdoor experiments) a "standard"
surrogate-NOx mixture being irradiated
on one side of thechamber, and a mixture
with differing initial surrogate and/or
NOx concentrations on the other. The
standard reactant mixture consisted of
~4 parts per million of carbon (ppmC) of
this surrogate mix together with ~0.4
ppm of NO, (at an initial [NO]/[N02] ratio
of ~2), although many experiments were
carried out with differing surrogate and
NOx concentrations to investigate the
effects of varying the initial reactant
concentrations. The "urban surrogate"
was a mixture of eight hydrocarbons
chosen to simulate emissions into urban
atmospheres (Whitten and Killus, private
communicaton), and its composition is
given in Table 1. Isobutene was included
in this mixture primarily to represent
formaldehyde, which is an important
constituent of urban emissions, but
which is difficult to handle experimental-
ly. Isobutene reacts rapidly to form
formaldehyde (together with acetone) in
NOx-air photooxidations. This was con-
firmed in a separate isobutene-NOx-air
indoor chamber irradiation carried out
under this program.
In addition, a number of conditioning,
control, and characterization runs were
carried out in order to make the data of
Table 1. Composition of the "Urban
Surrogate" Hydrocarbon Mixture
P*rr.*nt PP Component
Hydrocarbon
Ethene
Propene
Isobutene
n-Butane
n-Pentane
Isooctane
Toluene
m-Xylene
Carbon ppmC Surrogate
5.0
5.6
14.3
15.8
19.6
14.8
12.6
12.3
0.025
0.019
0.036
0.040
0.039
0.018
0.018
0.015
maximum utility for model testing pur-
poses. These consisted of (a) propene-
NOx control and conditioning runs, (b)
ozone dark decay experiments, (c) acet-
aldehyde-air irradiations to measure NO,
offgassing (from the rate of formation of
PAN in the absence of added NOX), (d)
radical tracer-NOx-air and CO-tracer-
NOx-air irradiations to measure the
chamber radical source, and (e) runs for
the outdoor chamber side equivalency
tests.
Results
Outdoor Chamber Experiments
These experiments were carried out in
two separate phases, the first in the fall
and the early winter of 1982, and the
second during the summer and fall of
1983. During the first phase, serious
problems with side equivalency were
encountered, and most runs carried out
during that time were side equivalency
tests aimed at investigating this problem.
Although the results of these tests are
somewhat ambiguous, it was concluded
that the problem was probably due to the
method of injecting the surrogate com-
pounds into the reaction bag (the liquid
components were injected into one side
before the bag was divided, and the
gaseous components were injected into
the other), possibly combined with in-
adequate conditioning of the reaction
bags. This problem was corrected prior to
carrying out the second phase of outdoor
experiments.
A total of 12 multi-day surrogate-NOx
runs were successfully completed during
the second phase of the outdoor chamber
experiments. Four of these runs were
side equivalency tests in which the same
surrogate-NOx mixture was irradiated on
both sides of the chamber, and except for
the first such run, which was apparently
carried out using an insufficiently
conditioned chamber, good side equiva-
lency was observed. As an example.
Figure 1 shows the results of one run
where good side equivalency occurred up
to the time NO was injected into side 2 on
the second day of the irradiation. Except
-------
0.8 r
0.6"
Ozone
Side 1
Side 2
0.06 T
0.04. .
§ 0.02
O
0.00
PAN n
O
0 S/rfe 1
O S/e
0°0
D O
ODD
i ,
0.4 T
0.3-
0.2. .
0.1
0.0
Formaldehyde
a
o
8
900
2/00
Clock Time (PST)
9O0
Figure 1.
Concentration-time plots for Oa,PAN and formaldehyde for run OTC-199. Arrow
indicates the time that «%» 0.5 ppm NO was added to side 2.
for one run. Which was carried out in such
a manner that the NOX offgassing rate on
day 2 was measured while the irradiation
was being carried out, the multi-day runs
consisted of a series of divided chamber
runs where initial levels of surrogate
and/or NOx were varied. For several of
these runs, NO was injected on subse-
quent days of the run to regenerate
photochemical reactivity.
Many of these runs involved irradiation,
either on one side of the divided chamber
or in the undivided chamber, of a
"standard" ~3.5- to 4- ppmC surrogate
plus ~0.4-ppm NOx mixture, with the
average temperature (for the first day)
ranging from 10°C to 42°C, and the
average N02 photolysis rate (ki) ranging
from 0.12 min"1 to 0.32 min"1. A
comparison of the concentration-time
profiles for 03 and N02 (the latter
uncorrected for interferences due to
organic nitrates and HN03) for these
irradiations is shown in Figure 2. It can be
seen that although some irradiations give
remarkably similar results, other runs,
particularly those carried out during the
winter months, exhibited significantly
lower reactivities.
An obvious explanation for these
different reactivities shown in Figure 2 is
the differences in light intensity for these
irradiations. Indeed, the first-day maximum
Os yield in the standard runs was highly
correlated with the average temperature
and the average light intensity data, with
a 0.93 correlation coefficient in both
cases. Reasonably good reproducibility
was observed in separate runs carried out
under similar lighting and temperature
conditions.
In addition to these irradiations of the
"standard" ~3.5- 4-ppmC surrogate plus
0.4-ppm NOx mixture, irradiations were
carried out in which the initial hydrocarbon
or the initial NO* concentration was
varied, usually with the standard mixture
on one side of the divided chamber and a
modified mixture on the other. For
example, Figure 3 shows selected con-
centration-time profiles for two variable
hydrocarbon surrogate runs, and Figure 4
shows selected profiles for two variable
NOx runs that were carried out under
similar conditions of temperature and
light intensity. As for other such irradia-
tions carried out in this program, and as
expected from existing environmental
chamber and computer modeling data,
the reactivity on the first day tended to
increase with increasing initial surrogate
concentation and, at least at the reactant
levels and conditions employed here, to
decrease with increasing initial NOX
-------
concentration. However, as seen from
these figures, this was not necessarily
the case for the second day of the
irradiation, where in many cases the
amount of Oa formed on the second or
subsequent days tended to be negatively
correlated with the reactivity on the first
day.
In many experiments, NO was added on
the second and/or subsequent days of
the irradiation to obtain data concerning
its effect on 03formation. Atypical result
of NO addition is shown in Figure 1,
which shows the effect of adding -O.15
ppm of NO to side 2 of the chamber at the
beginning of the second day of a matched
surrogate-NOx (side equivalency test) run.
It can be seen that while the addition of
NO caused an initial large drop in the Oa
level, because of the rapid reaction
between Oa and NO, subsequently rapid
Oa formation occurred resulting in much
higher Os levels at the end of the day than
for the side without added NO, where no
photochemical Oa formation occurred on
that day.
Indoor Chamber Experiments
The results of the six indoor chamber
multi-day surrogate runs were qualita-
tively similar to those for the outdoor
chamber runs discussed above. For
example. Figure 5 compares the results of
experiments with'differing initial NO,
levels. It can be seen that increasing the
initial NOX resulted in relatively little Oa
formation on the first day, but resulted in
greater Oa formation on the second day,
unless, as was the case for run ITC-635,
the initial NO was so high that it
suppressed Os formation on the second
day as well. The effects of varying the
initial hydrocarbon concentrations were
also similar to those observed in the
outdoor chamber experiments.
As for the outdoor chamber irradiations,
the addition of NO on subsequent days of
the experiments resulted in additional Oa
formation, provided that the amount of
NO added was not so high that Oa
formation was suppressed by the presence
of large levels of excess NO. For example.
Figure 6 compares the results of two
surrogate-NOx runs (ITC-626 and 637)
when NO was added on subsequent days.
These two runs had approximately the
same initial reactant concentrations, and
good reproducibility was observed on the
first day of the irradiation. In run ITC-626,
the amount of NO added was so high
(0.8 ppm) that it completely suppressed
Oa formation, and it is interesting to note
that 2 h after the first NO addition, the
oxidation of NO essentially stopped.
High T. UV
LowT.UV
u
a
900 1400 1900 2400
Clock Time (PST)
500
1000
1500
Figure 2. Concentration-time plots for ozone end uncorrected NOi for the standard surrogate-
NOi irradiations, where the runs are classified as either "high T.k\" runs (average
T >32 C. average *i XX2 min "V or "low T. k-\" runs (average T, *i both below those
values).
indicating a very unreactive mixture.
However, when relatively low levels of
NO (0.2 ppm) were added each day for
the subsequent 3 days of the run, then
significant Oa formation occurred, al-
though the increase of the Oa concentra-
tion following each NO injection decreased
monotonically each day. This indicates a
gradual decrease in the reactivity of the
reacting surrogate mixture with time.
Conclusions
The smog chamber data collected in
this study were supplied to Systems
Applications, Inc. (SAI), for analysis under
EPA Contract No. 68-02-3738. The
analysis of these data by SAI resulted in
the development of a multi-day chemical
mechanism suitable for use in regional
oxidant models. The purpose of this
project report is to make the experimental
data available to the scientific community
at large to enable other researchers and
modelers to further the development of
chemical transformation models.
The project report is composed of three
volumes. Volume I contains a description
of the experimental facility, methods of
procedure, and analytical techniques.
Volume II contains printouts of the
detailed data that were collected in these
experiments. Volume III includes docu-
mentation on the computer-readable
magnetic tape that contains the data
collected in this study. All three volumes
and the computer data tape are available
through the National Technical Informa-
tion Service, Springfield, Virginia, (see
ordering information at back).
-------
0.8 T
0.6-
0.4"
0.2.
Ozone
0.12-
^ 0.09
|
1 0.06
|
0.03-
0.00.
' PAN 0 High
0 STD
0 A Z.OW
a
QD
0°
o
0 A ^A ^
B^** 1 1 1 4d£-i
0.4 T
0.3
0.2 ..
0.1
0.0
Formaldehyde
4-
00DO
-+-
HI
900
2100
500
Clock Time (PST)
2/00
500
Figure 3. Concentration-time plots of O» PAN and formaldehyde for the variable surrogate
runs OTC-194 and 195. "High," "low" and "STD" refers to data for mixtures with
initial surrogate levels of ~7. ~2 or ~4 ppmC, respectively.
-------
0.0
0.6 r
!
NOi + Nitrates
0.0
0.08
Iso-Octane
0.06
o
A*
A Low
O STD
a High
O.04
0.02
0.00.
-4-
1000
1000
1000
1000
Clock Time (PST)
Figure 4. Concentration-time plots of Oa, uncorrected NOi and isooctane for the variable
initial /VO, runs OTC-204 and 205. "High." "low" and "STD" refer to curves for
mixtures with initial /VO« levels of 0.8, 0.2 and ~0.4 pom, respectively. Data
obtained after subsequent /VO, injections not shown.
-------
0.8 T
0.6
Ozone
Low
Light
Figure 5.
i i i
12 Dark 24 Light 36
Elapsed Time (hours)
36 Dark 48 Light
60
Concentration-time plots ofO* uncorrected NOa and NO observed in indoor Teflon
chamber runs 633, 635 and 637. "Low," "STD" and "high" refer to runs with
0.3. ~0.6 and 1.2 ppm initial NO* respectively. Data obtained'for subsequent NO,
injections not shown.
-------
"£ °-3 '
t
1
« 0.2-
I
c
o
<> 0.1.
0.0 .
0.15
0.12
\
1
\
L
PAN
O
a
B °
V.
D A
0 B
0.09
e
u.uo Q
a
0.03 O
o
a
0
So 0
&
<*n
D
12 24
36
O
<*& */*
48 60 72 84
Elapsed Time (hours)
Dark
Dark
Dark
Light
Light
Figure 6. Concentration-time plots for Oi NO and PAN for indoor chamber runs 626 and 63 J
Arrows indicate times NO was injected. "A" = run 626, "B" = run 637.
-------
W. P. L Carter, M. C. Dodd, W. D. Long, andR. Atkinson are with the Statewide Air
Pollution Center, University of California, Riverside, CA 92521; the EPA author
Marcia C. Dodge (also the EPA Project Officer, see below) is with the
Atmospheric Sciences Research Laboratory, Research Triangle Park, NC
27711.
The complete report consists of four volumes, entitled "Outdoor Chamber Study to
Test Multi-Day Effects:"
"Volume!. Results and Discussion," (Order No. PB35-161 628; Cost: $14.50)
"Volume II. Environmental Chamber Data Tabulations," (Order No. PB 85-161
610; Cost: $47.00)
"Volumelll. Computer-Readable Environmental Chamber Data," (Order No. PB
85-161 602; Cost: $7.00)
"VolumelV. Magnetic Tape,"(Order No. PB85-161 636; Cost: $140.00)
The above reports and magnetic tape will be available only from: fcosts subject to
change)
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Atmospheric Sciences Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
*USGPO: 1985-559-111-10795
-------
g 01
CD
C
q
8
5
I
O
i
OOOC329 PS
U S EKVIR PROTECTION AGENCY
STREET
It
REGION 5
S OEAR8CRN
------- |