United States
Environmental Protection
Agency
Office of Mobile Sources
Emission Control Technology Division
2565 Plymouth Road
Ann Arbor, Michigan 48105
EPA 460/3-85-009a
September 1985
&EPA
Air
Outdoor Smog Chamber Experiments:
Reactivity of Methanol Exhaust
-------
Outdoor Smog Chamber Experiments:
Reactivity of Methanol Exhaust
H. E. Jeffries, K. G. Sexton
M. S. Holleman
Department of Environmental Sciences
and Engineering
School of Public Health
University of North Carolina
Chapel Hill, N.C. 27514
Prepared under Subcontract with
Southwest Research Institute
Contract No. 68-03-3162
Work Assignment 30
EPA Project Officer: Craig A. Harvey
Technical Representative: Penny M. Carey
Prepared for
U. S. ENVIRONMENTAL PROTECTION AGENCY
Office of Mobile Sources
Emission Control Technology Division
2565 Plymouth Road
Ann Arbor, MI 48105
September 1985
-------
DISCLAIMER
This report has been reviewed by the Emission Control Technology Division, U. S. Environ-
mental Protection Agency, and approved for publication. Approval does not signify that the contents
necessarily reflect the views and policies of the U. S. Environmental Protection Agency, nor does
mention of trade names or commercial products constitute endorsement or recommendation for use.
-------
Abstract
The purpose of this project was to provide an experimental smog chamber database especially
designed to test photochemical kinetics mechanisms that would be used to assess the effects of
methanol fuel use in automobiles. The mechanisms would be used in urban air quality control
models to investigate the advantages of large scale use of methanol fuel in automobiles. The smog
chamber experiments were performed during three summer months. They have been added to
the existing UNC database for photochemical mechanism validation and testing, bringing the total
number of dual-experiments in the database to over 400.
Three different hydrocarbon mixtures were used: a 13-component mixture representing syn-
thetic automobile exhaust; an 18-component mixture representing synthetic urban ambient hydro-
carbons; and a 14-component mixture derived from the synthetic automobile exhaust by the addition
of n-butane. Three different synthetic methanol exhaust mixtures were used: 80% methanol/20%
formaldehyde; 90% methanol/10% formaldehyde; and 100% methanol. All experiments used a tar-
get initial concentration of 0.35 ppm oxides of nitrogen, which was 80% nitric oxide. Two basic
levels of the hydrocarbon mixture were used: 3 ppmC and 1 ppmC which gave approximately a 9
to 1 and a 3 to 1 hydrocarbon to nitrogen oxide ratio.
In the experiments, a reference mixture of 100% of the hydrocarbon mixture was reacted on
one side of the dual chamber while a "substituted" mixture of 67% carbon hydrocarbon mixture,
33% carbon synthetic methanol exhaust (with one of the three levels of formaldehyde) was reacted
on the other side of the dual.chamber. In this manner, the relative reactivities of the two systems
can be directly compared and models must reproduce both sides of the chamber with one set of
simulation assumptions.
Twenty-nine dual smog chamber runs were conducted. Eighteen of these experiments are satis-
factory for model testing and fourteen are excellent. The other 11 experiments, while having poorer
sunlight, which complicates model testing, are still quite useful to support the trends or directional
effects of the substitution process.
Synthetic methanol exhaust substitution in these experiments never resulted in an increase in
ozone maximum or a shorting of time to events over that of the reference side, even for a synthetic
methanol mixture with 20% formaldehyde.
For the highly reactive synthetic automobile exhaust at the 3 ppmC level, there was essential
no reduction in ozone when synthetic methanol was substituted. This was primarily because these
systems were limited by available oxides of nitrogen and not, by the organic reactant. There was a
delay in time to events that was reduced to almost no delay in the 20% formaldehyde experiments.
At the 1 ppmC level, there was 30-40% reduction in maximum ozone when the synthetic methanol
exhaust was substituted, depending upon the level of formaldehyde in the methanol exhaust.
For the less reactive synthetic urban mixture at the 3 ppmC level, there was ss!5% reduction in
maximum ozone when the synthetic methanol exhaust was substituted. At the 1 ppmC level, which
produced less than 0.15 ppm ozone, there was an 0-80% reduction in ozone maximum depending
upon the formaldehyde content of the synthetic methanol exhaust.
A small demonstration modeling exercise suggested that even the newest version of the Carbon
Bond mechanism has difficulties correctly simulating the range of conditions in these experiments
and further model testing with these data are strongly recommended.
111
-------
IV
-------
Contents
1 Introduction i
Need for Data . 1
Photochemical Reaction Mechanisms For Urban Air shed Models 1
Previous Methanol Test Data 3
Previous UNC Model Testing Data 6
Simple Systems Project 7
Reactivity Project .7
Automobile Exhaust Project 7
Complex Systems Project 9
Purpose ''. 9
Approach 10
Report Audience '. 10
2 Design 12
Design of Synthetic HC Mixtures 12
SynAuto 12
SynUrban 14
Synthetic Methanol Exhaust 19
Matched, Reduction, and Substitution Experiments '. . 19
Time Available for Experiments 20
Experimental Design 21
Initial Experiments 21
New Experiments Added In Second Summer "...'. 21
3 Methods 26
Outdoor Chamber and Analytical Facilities ' 26
Production of Synthetic HC Mixtures 26
SynAuto Mixture . . . '. 26
SynUrban Mixture 29
Synthetic Methanol Exhaust 31
-------
4 Results 34
Overview of Data Set 34
Detailed Experimental Data and Plots 36
Example Description of Information Presented 36
5 Discussion 101
Meaning of Reactivity 101
Experimental Findings 103
SynAuto Experiments 103
SynUrban Experiments 106
SynAutUrb Experiments at 3 ppmC 112
Modeling of Selected Experiments . 115
The Mechanisms Selected 115
Modifications and Assumptions . . . ; 132
The Simulations 133
Discussion of Model Results 134
6 Conclusions . 157
Appendix 163
A Facility Description 163
Chamber Description 163
Location . . 163
Materials . 163
Physical Design 163
Orientation 164
Air Handling System 164
Laboratory 168
Injection System ' . 169
Data Analysis, Validation and Reporting 169
Data Acquisition System . 169
Standard Operating Procedure 171
Data Treatment Procedures 171
General : . . 171
During A Run .172
Data Processing Steps 173
GC Calibration Processing 179
vi
-------
B Analytical Methods 181
Introduction 181
Hydrocarbons : . 181
Mass Spectrometry 202
Formaldehyde by Automated Colorimetry 202
Carbonyl Analysis 208
PAN Analysis 210
Alkynitrates 210
vn
-------
Tables
1. USC Synthetic HC Mixtures 4
2. HC mixtures used at UNC 8
3. Composition of Synthetic HC Mixtures 13
4. Atmospheric Compounds Accounting for 75% Each HC Class 16
5. Default Carbon Bond Class Fractions 17
6. Average Carbon Number at NECRM Sites 17
7. Average Composition at NECRM Sites 18
8. Carbon Bond 3 Class Fraction for SynUrban Mixture 19
9. Methanol Fuel Reactivity Experiments 23
10. Characterization Experiments 25
11. Analysis of Synthetic Auto Exhaust Mixture 28
12. Final SynAuto Mixture Composition 28
13. Initial Conditions in Four SynAuto Experiments 30
14. Summary of Experimental Conditions 39
18. Summary of Model Simulations 135
19. Model Simulation Wall Assumptions 135
20. Summary of Results for Ozone and Time to Events 161
Al. Processing System For Instrument Data . 176
A2. Processing System For DVM Data 177
A3. Processing System For Documentation Steps 178
A4. Processing System For Calibration Data 180
Bl. Analytical Methods, Characteristics, and Operation Methods 182
B2. Calibration Sources for Gases 183
B3. Species Measured by Major Site Instruments 184
B4. Calibration Species for PE Sigma 2 GC 199
vui
-------
Figures
1. University of Santa Clara Indoor Chamber Experimental Results 5
2. Comparison of SynAuto and Dodge Charger Exhaust 15
3. Methyl Nitrite Production Apparatus 32
4. July 17, 1984 Experimental and Meteorological Data 44
5. July 19, 1984 Experimental and Meterological Data 46
6. July 20, 1984 Experimental and Meterological Data 48
7. July 21, 1984 Experimental and Meterological Data 50
8. July 22, 1984 Experimental and Meterological Data 52
9. July 25, 1984 Experimental and Meterological Data 54
10. July 26, 1984 Experimental and Meterological Data 56
11. July 28, 1984 Experimental and Meterological Data 58
12. August 2, 1984 Experimental and Meterological Data 60
13. August 3, 1984 Experimental and Meterological Data 62
14. August 4, 1984 Experimental and Meterological Data 64
15. August 5, 1984 Experimental and Meterological Data 66
16. August 6, 1984 Experimental and Meterological Data 68
17. August 7, 1984 Experimental and Meterological Data 70
18. August 8, 1984 Experimental and Meterological Data 72
19. August 9, 1984 Experimental and Meterological Data 74
20. August 22, 1984 Experimental and Meterological Data ' 76
21. August 25, 1984 Experimental and Meterological Data 78
22. August 28, 1984 Experimental and Meterological Data 80
23. September 1, 1984 Experimental and Meterological Data 82
24. September 2, 1984 Experimental and Meterological Data 84
25. September 2, 1984 Experimental and Meterological Data 86
26. September 8, 1984 Experimental and Meterological Data 88
27. September 9, 1984 Experimental and Meterological Data 90
28. September 17, 1984 Experimental and Meterological Data 92
29. September 19. 1984 Experimental and Meterological Data 94
30. September 21, 1984 Experimental and Meterological Data 96
31. June 26, 1985 Experimental and Meterological Data 98
32. June 28, 1985 Experimental and Meterological Data 100
33. SynAuto Experiments at 3 ppmC, at Three Levels of Formaldehyde 104
34. August 5 and 7, 1984 low-ratio SynAuto Experiments Plot '. . 107
35. Combined August 5 and 7, 1984 low-ratio SynAuto Experiments Plot 108
36. 3 ppmC SynUrban Experiments Plot 110
37. Combined Sept. 1 and August 22 high ratio SynUrban Experiments Plot Ill
38. 3 ppmC SynAutUrb Experimental Plots 113
39. CB3, August 5, High Wall Assumptions 139
40. CB3, August 5, No Wall HCHO 140
41. CBX, August 5, High Wall Assumptions 141
42. CBX, August 5, Extra Wall Assumptions 142
43. ALW, August 5, High Wall Assumptions 143
44. CBS, August 6, High Wall Assumptions 144
45. CBS, August 6, No Wall HCHO 145
46. ALW, August 6, High Wall Assumptions 146
47. CBX, August 6, High Wall Assumptions 147
48. CBX, August 6, Extra Wall Assumptions 148
IX
-------
49. ALW, August 7, High Wall Assumptions 149
50. CBX, August 7, High Wall Assumptions 150
51. CBX, August 7, Extra Wall Assumptions 151
52. CBX, August 8, High Wall Assumptions 152
53. CBX, August 8, Extra Wall Assumptions 153
54. ALW, August 8, High Wall Assumptions 154
55. CBX, August 22, High Wall Assumptions 155
56. CBX, September 1, High Wall Assumptions 156
Al. The University of North Carolina Smog Chambers 165
A2. Schematic of UNC Outdoor Smog Chambers 166
A3. Orientation of UNC Outdoor Smog Chamber 167
A4. Solar Altitude and Zenith Angle at Noon 167
A5. AARF Site Data Acquisition System 170
A6. Example DATRAN Command File 173
Bl. Beckman Total HC example 186
B2. Carle 1 Total HC example 188
B3. Carle Column and Valve Configuration 189
B4. Maximum sensitivity of Carle chromatograph on attenuation of XI and gain
of 2 190
B5. Analysis of Ci-Ca Hydrocarbons on Carle II 192
B6. Plumbing Diagram to Carle HI Gas Chromatograph 193
B7. Analysis of Aromatics and C4- €5 HC from Dilute automobile Exhaust 194
B8. Plumbing Diagram for Perkin Elmer Capillary GC 195
B9. Calibration Chromatogram on PE 900 FID Gas Chromatograph. 196
BIO. Analysis of CC-C12 Aromatics on PE 900 FID Gas Chromatograph 197
Bll. Chromatograph of Calibration Mixture on Auto Sampling PE Sigma 2 199
B12. Reconstructed Ion Chromatograph of 1972 Dodge Charger Exhaust using EPA
Summer Gasoline 203
B13. Reconstructed Ion Chromatograph of EPA Summer Gasoline Using DB-1 Col-
umn 204
B14. MS of napthalene 205
B15. MS of Peak 43 206
B16. Hydrocarbon Species Identified in Cryocondenser Auto Exhaust 207
BIT. Response of CEA automobile Formaldehyde Instrument to Injected HC OH 209
B18. Response of HPLC with DNPH Method 211
B19. Response of Two UNC Electron Capture Detectors to 0.12 ppmV PAN. . 212
B20. Detection of Alkylnitrates with Automated PAN GC, Varian 940 ECD 213
-------
Acknowledgements
This project was assisted by several university staff and many students. Richard Kamens
managed the construction of the methyl nitrate synthesis apparatus and David Benham performed
the synthesis work. The site operation? were managed by John Suedbeck.
The following students performed data processing tasks: Lynn Clark, Jeffrey Hoffner, Charles
McDowell, Jennifer Jeffreys. Greg Yates, and Cindy Stock. The following students performed ex-
periments at the site: Jeff Arnold, David Benham, Lisa McQuay, and Joe Simmeonssen.
XI
-------
Introduction
Need for Data
Methanol is an alternative fuel for automobiles. The Environmental Protection
Agency,J various groups in southern California,2 The Department of Energy, E.I.
DuPont de Nemurs Co., and ARCO Petroleum Co.3 have been very active in inves-
tigating the use of methanol fuel (95% methanol, 5% isopentane or 85% methanol,
15% gasoline) in modified passenger cars.
To assess the effects that large scale use of methanol fuel might have on an urban
air shed, complex urban air shed models (UAM) are needed. These models combine
the effects of reactive emissions, local transport, vertical mixing, and chemistry to
predict the distribution of pollutants (e.g. ozone) in the air shed. With these models,
the effects of changes in emissions can be predicted and thus the probable effects of
a particular control strategy can be assessed before complex and costly policies are
promulgated.
A natural question that arises in the application of such models is "How accurate
are the predictions?" This can only be answered by testing the model components
individually and the model as a whole. Testing usually means comparing the model
predictions against measurements made in real situations. One of the most com-
plex components of the urban air shed model is the chemical mechanism that is
supposed to describe the urban atmospheric chemistry. This component is often
developed independently of the UAM and the same mechanism can be incorpo-
rated into models with varying degrees of meterological complexity (e.g. trajectory
models vs. multi-layer grid models).
-------
Introduction Photochemical Reaction Mechanisms For Urban Air shed Models
Photochemical Reaction Mechanisms For Urban Air shed Models
In an urban air shed model simulation, chemical transformation is only one process
affecting pollutant concentrations. Emissions, transport, dilution, dry deposition,
and reaction all change the concentrations. In this complex situation, it is not possi-
ble to unambigiously determine if the representation of the chemical transformation
process is adequate. In addition, the atmosphere contains many hundred chemical
species and not all of these can be represented in the chemical model. To keep
the UAM model solvable, some generalization processes (e.g. treat all aldehydes
as if they were acetaldehyde), and some deletion processes (e.g. no need to repre-
sent acetone), and some distortion processes (e.g. apply the same rate constant to
all paraffins larger than butane) must be used in constructing mechanisms for use
in UAMs. Different choices in implementing these simplifying processes leads to
different overall representations of the chemical transformations by different model
developers; it is not clear that there is one best mechanism.
To improve the belief that the chemical transformation process is adequately
represented, especially in light of the simplifications needed to treat the large num-
ber of species, photochemical mechanisms are tested for their ability to represent
events in situations where chemistry is the dominant process affecting concentra-
tions. Thus smog chambers are used to create various degrees of chemical complex-
ity and the models are tested for their ability to represent the chemical transforma-
tions in these chamber systems.
It is important to recognize, however, that the most dominant factors affecting
concentrations in the urban atmosphere are dilution (which can easily be five-fold
over the course of a day) and emissions of precursors into small morning mixing
heights. Except for a few experiments performed at UNC, smog chamber experi-
ments do not normally include these important factors. Therefore, caution must be
used when attempting to extrapolate chamber results to the ambient atmospheric
conditions. This is the primary function of the urban air shed model: to combine
the effects of all important processes.
The chamber test situation has to be complex enough to include important as-
pects of the urban situation and yet simple enough to explicitly test the chemical
mechanism. Experiments are usually designed to proceed from a simple situation
to an approximation of an urban situation in a successive series of increasingly
complex experiments. Thus test conditions usually begin with simple one-HC sys-
tems in static operating conditions and proceed to simple-HC mixtures and then to
complex-HC mixtures. When these systems can be adequately represented by the
-------
Previous Methanol Test Data Introduction
chemical model, more realistic urban-like operating conditions are added. For ex-
ample, in addition to using complex HC mixtures, large dilution is used throughout
the experiment and reactive emissions are injected continuously. A demonstration
that models work for these conditions certainly enhances the belief that these mod-
els would give reasonably good predictions in the urban atmosphere simulation.
Certainly, one would have to be suspicious of a model that could not adequately
simulate the chamber conditions which are clearly simpler than the urban situation.
Previous Methanol Test Data
The University of Santa Clara (USC) conducted a series of indoor smog chamber
experiments to investigate the impact that exhaust from cars using methanol fuel
might have on urban photochemistry. In these experiments, the effect of a reduction
of 33% of the volatile organic compounds (VOC) in a synthetic hydrocarbon mix-
ture was compared with substitution of 33% of the VOC with synthetic methanol
exhaust.3 The experiments were performed at three hydrocarbon (HC) to nitrogen
oxides (NOX) ratios: 3:1, 9:1, and 27:1. Table.1 lists the composition of the surrogate
mixtures used in the study. Figure 1 shows an example of the ^9:1 HC-to-NOx data
that were produced in this study.
System Applications Inc. (SAI) used the chamber results to test the ability of
the Carbon Bond III photochemical mechanism to represent the effects of methanol-
fuel substitution.3 Some problems with the mechanism's representation of aromatics
were encountered in these tests and the mechanism was expanded to include more
detailed aromatics representation. The new mechanism was then used in a series of
air shed model simulations to estimate the benefits of large scale use of methanol-
fuel in the South Coast Air Basin of Los Angeles.4'' Whereas the original air shed
simulations with CBS mechanism suggested ==22% reduction, the new, expanded
CBM suggested «18% reduction in the maximum 1-hour ozone (03) level when
100% of the mobile source VOC (approximately one-half the total urban VOC)~was
replaced with 90% methanol and 10% formaldehyde on a per carbon basis. The
study also found that the results were very sensitive to the assumption of how
much formaldehyde was in the methanol exhaust.
Considering the discussion above, there are some aspects of the USC/SAI study
that are troublesome:
1) The experimental conditions of the indoor chamber resulted in very fast exper-
iments and sometimes the initial conditions were not very typical of urban-like
conditions (e.g. initial NOX=1.2 ppm). This means that models that were tested
with the chamber data must be "extrapolated" significantly when they are ap-
plied to urban ambient atmospheric conditions.
-------
Introduction
Previous Methanol Test Data
Table 1. University of Santa Clara Synthetic HC Mixtures
(Units are fraction of total carbon)
HC Mixture
Compound Fraction Adj. Fraction0
n-butane
n-pentane
2,3,4-trimethyl pentane
ethylene
propylene
toluene
m-xylene
iso-butene
(formaldehyde)
(acetone)
paraffin
olefin
aromatic
aldehydes
and ketones
Total
0.150
0.200
0.150
0.050
0.050
0.125
0.125
0.150
(0.038)
(0.112)
0.500
0.100
0.250
0.150
1.000
0.169
0.225
0.169
0.056
0.056
0.141
0.141
(0.042)
0.563
0.113
0.282
0.042
1.000
Methanol Mixture
Compound
Fraction Adj. Fraction
methanol
isobutene
(formaldehyde)
(acetone)
Total
0.692
0.308
(0.077)
(0.231)
1.000
0.900
0.100
1.000
USC used iso-butene as a surrogate for formaldehyde. It was assumed that, in a reactive system, iso-butene would
be rapidly converted to formaldehyde and acetone by hydroxyl radical attack. It was further assumed that acetone
was unreactive and could be omitted in calculating the composition of the mixture. The carbon fractions omitting
the acetone are given in the column headed "Adj. Fraction."
-------
"j
I—••
c
cl
a
3'
o
•"*»
CAi
p
a
rt-
t»
O
ef
P>
a"
a,
o
O
tr
I
cr
o>
1-
a
(T*
P>
n
en
1200-r
NO
N02 1000-
OZONE
DATE iJUL 28
DATE lAUG 01
DATE iAUG 24
TYPE i
TYPE •
TYPE i
BS BASE
2/3 BASE
as + MS
PPBC/NOx i
PPBC/NOx i
PPBC/NOx •
7.69
5.30
7.54
2/3 => 2/3 8-SURROGATE BASELINE
8S => 8-SURROGATE BASELINE
OZONE. 85 BASELINE
OZONE. HEOH SUBSTITUTION
OZONE. 2/3 8S BASELINE
240 360
MINUTES OF IRRADIATION
480
600
-------
Introduction Previous UNC Model Testing Data
2) The surrogate mixture did not contain formaldehyde, but instead used iso-
butene as a substitute for formaldehyde in both the urban mixture and in the
methanol exhaust. This was based upon the assumption that iso-butene would
rapidly react to produce formaldehyde. The HO reaction with iso-butene, how-
ever, produces an RO, radical that converts an NO to an NO2 for each iso-butene,
before making formaldehyde and acetone. This "extra" reactivity would not oc-
cur with formaldehyde. Also the delay in the production of HCHO may not have
correctly represented the importance of the role of the formaldehyde from the
methanol exhaust, particularly at the low HC-to-NOx ratio. In addition, acetone
is a photolytic species that appears explicitly in the newer CBX mechanism
at a non-negligible photolysis rate, contributing additional new radicals. Thus
substituting for formaldehyde leads to the need for another "extrapolation" of
models tested with these data when used for urban simulations.
3) The CBS mechanism had difficulty simulating the surrogate HC mixture used;
this was attributed to the presence of 12.5% m-xylene in the synthetic HC mix-
ture. The CBS mechanism was revised- to include a new representation for
xylenes. The new CB mechanism was not explicitly tested for its ability to
represent either m-xylene or other higher aromatics. It predicted a significantly
smaller reduction in the air shed simulations. It is not possible with just the
USC data to tell if the adjusted model predictions are correct or not.
4) When a model has difficulty simulating aspects of the USC chamber, there is
no simpler set of data for this chamber that could be used to test the com-
ponent parts of mechanisms, and thus, in this chamber, it would be difficult
to determine if a problem existed with the mixture chosen, with the chamber
characterization, with the operating conditions, or with the chemical mechanism
used.
These observations suggested that additional test data were needed to assure
that urban air shed simulation models were adequate to assess the effects of large
scale methanol fuel use.
Previous UNC Model Testing Data
The University of North Carolina, through an extensive program with EPA, has
been producing smog chamber data from a unique large dual outdoor smog cham-
ber for more than 12 years. The primary thrust of this program has been to pro-
duce data to test the adequacy of developing photochemical kinetics mechanisms to
represent various critical phenomena in the chemical transformation processes. A
logical extension of this work was to include methanol-fuel exhaust in the reactivity
-------
Automobile Exhaust Project Introduction
tests already being conducted. Because of the extensive data set available, this pro-
gram would be tightly focused on methanol-related phenomena. Yet, because these
methanol experiments would not stand alone, they would still have a generality and
wide range of applicability in testing mechanisms because they would fit logically
into an extensive test series.
Four projects using the UNC Outdoor Smog Chamber have contributed the
majority of data for model testing. Although these projects had their own goals
and purposes, they were not designed or performed totally independently of each
other nor of this project. Instead each project depended upon data from other
projects to address specific issues. Thus data from these projects are needed to
complete the partial description provided by the specific work performed in this
project. These other projects are briefly described below.
Simple Systems Project
Between 1977 and 1981 this project produced 114 dual-experiments designed to
test explicit mechanisms for aldehydes, olefins, paraffins, and simple two-component
mixtures. These data were described in a final report5 and a magnetic tape con-
taining the data is available from UNC. Data were sent to several EPA-sponsored
model-development groups including SAL These data were part of the set used to
test and develop the CBS mechanism.6
The final report described guidance for modeling the data including recommen-
dations on treatment of light data, dilution rates, wall losses, and use of water vapor
and temperature data. These guidelines are also applicable to the data described
in this report.
Reactivity Project
Another 70 dual-experiments were produced as part of a 1981-83 EPA grant' that
investigated how well models could represent reactivity changes in simple and com-
plex HC mixtures. Table 2 lists the composition of some of the mixtures used in
these tests. Blends of the mixtures in Table 2, at various ratios, were also used. The
best of these experiments have been distributed to the model development groups
at SAI and the University of California at Riverside (UCR). UCR is up-dating and
revising the Atkinson, Lloyd, and Winges(ALW) mechanism8 for use in EPA air
shed and EKMA models.
Automobile Exhaust Project
Another 31 dual-experiments using automobile exhaust from two vehicles were pro-
7
-------
Table 2. Composition of Hydrocarbon Mixtures Used in
UNO Smog Chamber Experiments.
(unils are percent of total carbon)
Compound
SIMMIX UNCMIX SIMARO COMARO SYNAUTO P/B
P/B/T BASM1X
butane
pentnne
isopentane
2-inelliylpentane
2,4-climethylpent ane
2,2,4-lriinethylpenlane
el'hylene
propylenc
1-biilene
lrans-2-but.ene
cis-2-bulene
2-nielhyl-l-butene
2-inelliyl-2-bnt,enc
benzene
toluene
ni-xylene
o-xylene
1,2,4-triinethylbenzeiie
n- pro pyl benzene
s(!C-hulylbenzene
form aldehyde
total paraffin
total olefin
total aromatic
0.4002
0.3183 0.2531
0.1484
0.0996
0.0804
0.1202
0.1631 0.1167
0.1184 0.0524
0.0254
0.0313
0.0347
0.0317
0.4886 0.2482
0.3880 0.1882
0.1.234 0.2803
0.1371
0.1522
0.7185 0.7077 0.0000 0.0000
0.2815 0.2922 0.0000 0.0000
0.0000 0.0000 1.0000 1.0000
0.0391
0.0519
0.1121
0.2391
0.0416
0.0196
0.0196
0.0538
0.2115
0.1026
0.0481
0.0564
0.0200
0.2031
0.3199
0.4724
0.7663 0.5352 0.3140
0.1268
0.0650
0.2337 0.1648 0.0943
0.3000 0.2000
0.2000
0.7663 0.5352 0.4408
0.2337 0.1648 0.1593
0.0000 0.3000 0.4000
-------
Purpose Introduction
duced as part of a 1982-84 EPA cooperative agreement. These experiments were
described in detail in a final report,9 and the processed data have been sent to SAI
and UCR for model testing.
In the automobile exhaust work, an HC-mixture, which imitated the reactivity of
automobile exhaust, was produced by direct side-to-side comparison experiments.
This mixture was used in the methanol work and will be described in Chapter 2.
Complex Systems Project
This 1981-83 project produced 128 experiments focused on aromatics chemistry,
complex mixtures, and dynamic chamber conditions such as large dilution (five-fold)
and continuous injection of reactants. The final reportjo for this project presents
an organizational scheme for all 346 experiments from all of the projects through
the end of 1983. It also discusses three examples of selecting runs to test an explicit
toluene mechanism, to test dynamic operating conditions, and to test an EKMA-
type mechanism. The best of these experiments have been distributed to the model
development groups at SAI and UCR.
Purpose
The purpose of this research was to produce data, using an outdoor smog chamber,
that could be used by model developers to test their chemical mechanisms. The
mechanisms would then be used in urban air shed simulation models to predict the
effects of large scale methanol fuel use in urban areas.
The reaction mechanism's ability to correctly predict the effects of large changes
in the composition of the HC mixture caused by substituting methanol and formalde-
hyde for mixture carbon is the most important factor that was tested in the- ex-
periments performed in this project. In addition, a large number of well matched
experiments using well characterized and complex HC mixtures that had a range of
reactivity and were based on extensive urban field analysis and automobile exhaust
analsysis were to be produced.
It was NOT the purpose of this project to:
o reproduce the USC work;
o simulate urban air shed conditions;
o simulate the effect of evaporative emissions; and
o validate mechanisms.
-------
Introduction Approach
Some comparisons of model predictions and chamber data were performed as
part of this project. The focus of this exercise was to demonstrate the utility of
the data produced and not to validate or test mechanisms. As described in the
introduction, such testing would require a much more extensive set of chamber
data and thus would have to use information from other projects.
Approach
The basic tests consisted of side-by-side experiments in which the chemistry of a
typical synthetic auto-exhaust or synthetic urban hydrocarbon mixture, at typical
HC-to-NOx ratios, was compared with the chemistry of a mixture in which one-third
of the original mixture was substituted by a synthetic methanol-exhaust mixture.
In these so called "substitution" experiments, the overall reactivity of the original
auto-exhaust or urban mixture was compared with the reactivity of the methanol-
exhaust substituted mixture.
The first priority was to compare the reactivity of 100% synthetic automobile
exhaust (no evaporative emissions) with a mixture of 33/tC synthetic methanol
exhaust and 67%C synthetic automobile exhaust. This would test the model's
ability to represent a very reactive mixture and to correctly represent the replaced
fraction of the urban mixture in the air shed simulations.
The second priority was to compare the reactivity of a less reactive, more typical,
urban mixture (which included contributions from evaporative emissions, station-
ary source emissions, and automobile exhaust) with the same methanol exhaust
substituted mixture. This would test the model's ability to represent the "typical"
mixture and to test that the model had not been "over-tuned" to a single mixture
composition (e.g. the auto-exhaust mixture).
When new mixtures are used in the UNC chamber, matched side-to-side com-
parison experiments are needed to demonstrate that both sides do yield the same
reactivity when the same material is injected on both sides. In addition, standard
chamber characterization experiments and background test experiments must be
performed routinely. Thus the first experiments to be performed were various test
experiments needed for Quality Assurance (QA) purposes.
Report Audience
This report is primarily written for a technical reader, especially one interested
in using the chamber data produced in this project to test mechanisms. We have
assumed that this reader has also read:
10
-------
Report Audience Introduction
o the 1982 UNC report5 "Outdoor Smog Chamber Experiments To Test Photo-
chemical Models" which contains guidance on representing light intensity and
other chamber characteristics;
o the 1985 UNC reportJ0 "Outdoor Smog Chamber Experiments to Test Photo-
chemical Models: Phase IF which contains a complete index to all pre-1984
chamber experiments and suggestions on run selection for model testing;
o the 1985 UNC report9 "Outdoor Smog Chamber Experiments Using Automobile
Exhaust" which contains information on the characteristics of the vehicles used
and the results of 31 dual experiments including a series comparing the synthetic
mixture used in this project with automobile exhaust.
Much of the layout and information in this report is designed to assist readers in
deciding which experiments they should use for their tests. For this reason, the
plots and lists of conditions in Chapter 4 are very important.
A second reader is expected to be a technically oriented policy advisor. To aid
this reader in understanding a complex subject, we have included expanded expla-
nations of various processes, more definitions, and some interpretation of findings.
In spite of the fact that the chamber data can be interpreted in a relative sense,
independently of validating models, this reader is cautioned not to simply extrap-
olate chamber findings to the urban atmosphere. As described above, there are
many factors that influence the urban atmosphere other than chemistry and urban
air shed simulation models are needed to attempt to deal with all the factors at
once.
11
-------
Design
Design of Synthetic HC Mixtures
SynAuto
One of the tasks in the UNC automobile exhaust project was to test the hypothesis
that a synthetic mixture of 10-15 compounds could adequately represent the reac-
tivity of dilute automobile exhaust. In this project, four gas chromatographs (GC)
covering the range from C2 to Cn were used to determine the composition of both a
controlled and an uncontrolled vehicle used in the study. GC/mass spectra analysis
was used on cryocondensed samples to identify the species. From the concentration
and identity of the compounds in the exhaust, we prepared a 13-component mix-
ture that would mimic the overall reactivity of the exhaust. The mixture, listed in
Table 3, contained 3 paraffins. 4 olefins, 5 aromatics, and formaldehyde. Benzene,
because of its non-negligible rate constant and its relatively large concentration in
exhaust, was included among the aromatics in the synthetic mixture. Acetylene,
a "non-reactive" and major HC compound in exhaust, was not included because it
would have created analytical problems in our routine GC analysis (the effect of
omitting acetylene is minor, mainly affecting how much synthetic exhaust would
be "equivalent" to real exhaust in NMHC units). Details on the types of vehicles,
comparisons with a large test fleet, and other experimental results are provided in
the final report.9
In the project, four dual experiments were performed in which exhaust from one
of the vehicles was injected into one side of the chamber and an "equivalent" amount
of synthetic exhaust was injected into the other chamber. In all experiments, the
12
-------
SynAuto Design
Table 3. Composition of Hydrocarbon Mixtures.
(Units are fraction of total carbon)
Compound UNCMIX SynAuto SynUrban
butane
pentane
isopentane
2-methylpentane
2,4-dimethylpentane
2,2,4-trimethylpentane
ethylene
propylene
1-butene
trans- 2-butene
cis-2-butene
2-methyl- 1-butene
2-methyl-2-butene
benzene
toluene
m-xylene
o-xylene
1 ,2,4-trimethylbenzene
formaldehyde
total paraffin
total olefin
total aromatic
0.2531
0.1484
0.0996
0.0864
0.1202
0.1167
0.0524
0.0254
0.0313
0.0347
0.0317
0.7077
0.2922
0.0000
0.0391
0.0519
0.1121
0.2391
.0.0416
0.0196
0.0196
0.0538
0.2115
0.1026
0.0481
0.0564
0.0200
0.2031
0.3199
0.4724
0.1000
0.1367
0.0801
0.0538
0.0467
0.1340
0.0630
0.0238
0.0137
0.0169
0.0187
0.0171
0.0331
0.1304
0.0633
0.0296
0.0347
0.0200
0.5404
0.1546
0.2854
13
-------
Design _ SynUrban
two sides were relatively similar in reactivity, but only when the sides were carefully
matched in CO, as well as NMHC. were the 03 concentrations within 10% on the
two sides. One example of a well-matched system is shown in Figure 2. In this
experiment, 2.6 ppmC of exhaust from a 1972 V8 Dodge Charger and 2.2 ppmC of
the synthetic mix (less because of the acetylene which was «15% of the exhaust)
showed a remarkably similar reactivity (time to NO-to-NOj crossover agreed to within
8 minutes and O3 maxima agreed to within
Thus we believe that the synthetic mixture called SynAuto in Table 3 is a rea-
sonable approximation of automobile exhaust reactivity.
SynUrban
A second mixture, more typical of urban-like conditions, was needed for use in this
study. Both emissions inventory and atmospheric hydrocarbon data were the basis
for the compositional design of this synthetic urban mixture (SynUrban).
In 1972 Kopczynski investigated ambient HC compositions in both Los Angeles
and Cincinnati." Part of these data are shown in Table 4. Kopczynski used these
data to design a mix for smog chamber use. His mix was used by UNC in 1973 as
a guide to preparing a synthetic urban mixture, called UNCMIX which is also listed
in Table 3. UNCMIX has two parts: paraffins and olefins, which are in a pressurized
tank, and aromatics, which because of their significantly lower vapor pressures, are
stored as liquid mixture. UNCMIX has been used in more than 100 experiments in
our chambers; most of these experiments, however, have used only the paraffin and
olefin portion of the mixture.
In conducting this study, we wished to use newer ambient data for the mixture
design. Killus and Whitten12 conducted a review of ambient data to provide the
basis for their recommended default composition for air quality modeling in cases of
insufficient support data. They found that NMHC composition did not vary greatly
between cities. GipsonJ3 recently completed an analysis of atmospheric data col-
lected for the Northeast Corridor Regional Modeling Project. He also found little
variation among the 435 sites for both composition and average carbon number of
each HC class fraction.
It was decided to design a synthetic urban mixture which would come close to
matching the default composition of Killus and Whitten, once converted to Carbon
Bond Units. These values are listed in Table 5.
14
-------
I.U
0.9
E 0.8
a.
a 0.7
1 0.6
X
0 0.5
c
<5
O) 0.4
1 0.3
0.2
0.1
r\f\
1 | 1 | 1 | I I 1 | 1 | 1 | 1 | 1 | ! | 1 | 1 | 1 j i
- October 4, 1983 -
— —
-
- 03 -
~ /. ^~^^ -
- / / -
I /
'/
~ / ~
~ NO?/
f~~"5<7~ ^~— ~~ :
J\" \" i 1 i Til i^>*C 1 i 1 i 1 i I i 1 i 1 i 1 i 1 i
I.U
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
nn
/^
N
O
p
v«
U
1
8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
2.0
w
_
g>1.5
O
11.0
a
cc.
'0.5 -
0.0
5 6 7 8 9 10 11 12 13 14 15 16 17 18
HOURS, EOT
Figure 2.
Top: (Solid) 2.6 ppmC DODGE CHARGER EXHAUST;
(Dashed) 2.2 ppmC SYNAUTO;
Bottom: RED chamber air temperature (top solid line, °F);
RED (solid line) and BLUE (dashed line) chamber dewpoint (°F);
ambient total solar radaiation (solid line, cal-cm" -see' )•
ambient ultraviolet radiation (dashed line,mcal-cm -sec ).
15
-------
Design
Synllrban
Table 4. Analysis of Gas Chromatographic Data Of Los Angeles (Kopczynski) 75% Of Each
Major HC Class.
ID NUM.
COMPOUND
CAR. FRACTION
17
12
3
19
6
38
33
25
ISOPEMTANE 0.0917
ti-BUTANE 0.0840
ETHAJIE 0.0561
N-PENTAME 0.0502
PROPANE 0.0411
2.4-DIMETHYLPEKTANE 0.0262
N-HEXAHE 0.0241
CYCLOPENTANE 0.0203
2
5
83
10
11
21
ETHYLEI.'E 0.0444
PROPYLENE 0.0176
2-METHYL-2-BUTEHE 0.0085
1-BUTENE 0.0069
ISOBUTYLEKE 0.0069
TRAi;S-2-PE!iTENE 0.0065
55
68
67
71
78
79
65
TOLUENE 0.0796
M-XYLENE 0.0632
P-XYLEME 0.0294
0-XYLEHE 0.0256
1.2. 4-TRIKETHYLBE'JZENE 0.0201
SEC-BUTYLBENZENE 0.0201
ETHYLBENZEIJE 0.0197
ACETYLENE 0.0470
45 COMPOUNDS TOTAL
PAR FRAC 0.5040 PAR CARBON NUM. 4.1363 15 COMPOUNDS
OLE FRAC 0.1131 OLE CARBON MUM. 2.8861 13 COMPOUNDS
ARO FRAC 0.3341 ARO CARBOll NUM. 8.2082 15 COMPOUNDS
16
-------
SynUrban . Design
Table 5. Default Carbon Bond Class Fractions
as Recommended by Killus and Whitten for CB3J*
Bond Group Meaning
PAR all single bonded carbon
ETH ethylene carbon
OLE all double bonded carbon
ARO all aromatic ring carbon
CAR carbonyl group carbon
NR non-reactive carbon
Range
0.50 - 0.70
0.02-0.11
0.02 - 0.07
0.10-0.40
0.03 - 0.10
0.05 - 0.20
Default
0.58
0.04
0.03
0.19
0.05
0.15
Although these indicate the total class fractions, they do not give the class average
carbon numbers and therefore these do not suggest the types of species that should
be included in the mixture.
Average carbon numbers, as determined i>y the Gipson analysis, were used to
estimate the species needed in each class (see Table 6).
Table 6. Average Carbon Number at
Northeast Corridor Regional Modeling Sites
Paraffin Olefin Aromatic
mean 5.4 2.9 7.7
sd 0.4 0.7 0.02
The average composition of all 435 sampling sites was" also used as a guide to
designing the synthetic hydrocarbon mixture (see Table 7).
17
-------
Design Synllrban
Table 7. Average Composition at
Northeast Corridor Regional Modeling Sites
(analysis of ambient HC data)
Class mean %C s.d.
Paraffin
Olefin
Aromatic
Acetylene
Ethane
56.4
10.5
27.1
2.7
3.3
9.5
4.9
8.2
1.6
2.5
We observed that the paraffin/olefin portion of the UNCMIX was very close to
both the class fractions and average carbon number in the Gipson and in the Killus
and Whitten studies. The paraffins in the UNCMIX needed a slightly lower average
carbon number which could be achieved by the addition of n-butane. The original
aromatic portion of UNCMIX. however, had too high an average carbon number (8.3)
compared to the new Gipson values. The aromatic portion of the SynAuto mixture,
on the other hand, had an avera.ge value much closer to the value in Table 6 and
was therefore used as part of the SynUrban mixture as well. This aromatics portion
in the SynAuto mixture was made up as a liquid in a separate container and thus
could be easily used for both mixtures.
A computer program written to analyze hydrocarbon mixtures was utilized
to determine the best, ratio of the tank portion of UNCMIX, the liquid portion of
the SynAuto mixture, and n-butane to achieve the closest agreement with the regu-
lar composition and average carbon numbers of Gipson. and the resulting Carbon
Bond composition unit default values of Killus and Whitten. The best mixture
of UNCMIX, the liquid portion of the synthetic auto exhaust, and n-butane was a
54:36:10 (carbon ratio). " -\
The composition of the designed SynUrban mixture is given in Table 3. It con-
sisted of 18 compounds: 6 paraffins, 6 olefins, 5 aromatics, and formaldehyde. Its
composition in terms of the Carbon Bond Mechanism input species is given in Ta-
ble 8; these can be compared with the default values listed in Table 5 and with the
synthetic mixture used in the University of Santa Clara listed in Table 1.
18
-------
Matched, Reduction, and Substitution Experiments Design
Table 8. Carbon Bond 3 Class Fraction for SynUrban Mixture.
(units are fraction of total carbon)
Class
PAR
ETH
OLE
ARO
CARB
NR
Carbon %
0.5498
0.0533
0.0281
0.1710
0.0299
0.1680
Synthetic Methanol Exhaust
The Emission Control Technology Division of EPA has tested a methanol fueled
Ford Escort and a VW Rabbit for emissions using the Federal Test Procedure
which employs a dynamometer.1 The results show that, while the total organic
emissions were higher than the gasoline counterparts, the exhaust mostly consisted
of unburned methanol (MeOH) and formaldehyde (HCHO), with a small amount of
other HCs. In addition, very low levels of methyl nitrite (MeN02) were detected in
the methanol samples, but only under certain testing conditions. MeNO2 formation
may be an artifact of the sampling techniques used. The fraction of HCHO in the
methanol exhaust can vary from a few percent to nearly 25%.
To simulate the methanol exhaust in our experiments, we used three mixtures
of MeOH/HCHO: 80%/20%; 90%/10%; and 100% MeOH. These were called high, mid,
and low HCHO exhaust. In a few tests, 1% of the total MeOH was replaced with
MeN02.
Matched, Reduction, and Substitution Experiments
Because the UNC chamber is divided into two side-by-side halves, two experiments
are conducted at one time and the same set of instruments are used to measure
the reactant and product concentrations on both sides in a timeshared manner.
Thus experiments are designed to take advantage of the side-to-side nature of the
chamber. In the methanol work, three basic types of side-to-side experiments were
used:
• Matched Conditions—Both sides of the chamber had exactly the same chem-
ical systems. The primary purpose for this type experiment was to show that
19
-------
Design Time Available for Experiments
the two chamber halves perform identically for the various chemical systems be-
ing tested. This allows stronger conclusions about cause-and-effect to be drawn
in other experiments with different conditions on the two sides.
• Reduction Conditions—The IMOX concentration and the organic reactant mix-
ture being studied were the same on both sides of the chamber, but the HC
concentration was reduced by one third on one side of the chamber. These ex-
periments test the model's ability to predict AO3/AHC correctly with all other
conditions being the same.
• Substitution Conditions—Conditions on the two sides of the chamber were
the same except for the carbon species identity. In the organic mixture on one
side, approximately one-third of the carbon was replaced with the same amount
of carbon, but with a different species identity. In the methanol substitution
experiments, one-third of the SynAuto or SynUrban mixture carbon was replaced
with a SynMethanol mixture. The formaldehyde content of the SynMethanol mixture
was 0, 10, or 20% of the total SynMethanol mixture. These experiments test the
model's ability to correctly predict the effects of composition change in a given
mixture at a constant total HC concentration.
In addition to experiments that address the basic questions being asked, experiments
were needed for quality assurance and to assess the potentially varying chamber ar-
tifact processes such as the magnitude of wall emissions of old reactant material.
These experiments are called Characterization Experiments and were run rou-
tinely.
Time Available for Experiments
Because of the funding starting date (July 1. 1984) and the termination of the
funding source on the fiscal year end (September 30. 1984) there was only time for
twenty-one scheduled methanol reactivity runs, and nine scheduled chamber test
runs (special characterization runs). Based on previous experience and assuming
no mechanical or analytical difficulties and 50% good weather, at most about 11
methanol reactivity experiments would be reasonably useful, and maybe 8 of these
experiments would be very good. The planning for the experiments also took into
account other projects that were scheduled for the chamber.
Not every day could be devoted to making runs; the instruments needed cali-
brating and chamber characterization runs had to be performed for QA purposes.
A series of estimated schedules were used to program, the runs for the chamber, but
the schedule was change rapidly depending upon meteorological events or instru-
ment status. That is, we may have actually had a calibration day on what was to
20
-------
New Experiments Added In Second Summer Design
be a run day, but because it rained, the run was aborted and calibrations were done
instead. The planned characterization runs, however, generally were scheduled to
occur on the weekends. These runs normally required much less labor because there
was no complex sampling and data processing. They were also the slack runs for
the week. In other words, if it had been a bad week in terms of successful runs,
then the weekend characterization runs were omitted in favor of one of the other
runs for that week. This was not done every weekend, however, because of labor
costs and the need to have a minimal set of Characterization runs.
Experimental Design
The experimental program had to accomplish a number of objectives:
• demonstrate the reactivity of surrogate exhaust from methanol fueled cars rel-
ative to the reactivity of surrogate exhaust from conventional gasoline fueled
cars for outdoor conditions at several HC concentrations (NOX constant);
• provide validation data for models intended to simulate the effects of methanol
fuel substitution in control situations;
• investigate the sensitivity of the methanol-VOC mixture reactivity to changes in
composition, such as the formaldehyde content of the exhaust, and the detailed
composition of the urban mixture.
Initial Experiments
We choose NMHC concentrations based upon previous chamber results and our initial
estimates of the relative reactivity of the synthetic mixtures. The HC levels chosen
were: 3.0, 2.0, 1.0, 0.66 ppmC. All experiments were to be performed at 0.35 ppm
NOX, which is a reasonable approximation of urban maximum NOX conditions. Both
SynAuto and SynUrban mixtures would be used. The degree of substitution was de-
signed to be 33%, i.e., 33% of the SynAuto or SynUrban mixture carbon was replaced
with an equal amount of carbon in the form of MeOH and HCHO.
The target conditions for the methanol test experiments are given in Table 9.
Experiments for characterization are given in Table 10. These tables describe the
purposes of each experiment and predict the outcome. (These predictions were
made before any experiments were run and have not been changed for this report.)
New Experiments Added In Second Summer
Results from the modeling (reported in Chapter 5) and from the SynUrban experi-
ments, showed that the 1.0 ppmC and 0.66 ppmC SynUrban conditions were too low,
and new experiments were required at a higher concentration, but less than the
21
-------
Design New Experiments Added In Second Summer
3.0 ppmC of the August 22, 1984 experiment. We estimated that approximately
2 ppmC SynUrban and 1.4 ppmC SynUrban should give 0.3 and 0.15 ppm O3. The
new target conditions were used in a baseline (e.g. no MeOH) experiment, and in a
normal MeOH exhaust substitution experiment.
22
-------
Table 9. Methanol Fuel Reactivity Experiments
(All experiments have 0.35 ppm NOX)
Num
1
2
3
4
5
6
TYPE
Reduction, 33%
Substitution
normal HCHO
Substitution
low HCHO
Substitution
high HCHO
Reduction, 33%
Substitution
normal HCHO
REACTANTS
First Side
3.00 ppmC SynAuto
3.00 ppmC SynAuto
3.00 ppmC SynAuto
3.00 ppmC SynAuto
1.000 ppmC SynAuto
1.000 ppmC SynAuto
REACTANTS
Second Side
2.00 ppmC SynAuto
2.00 ppmC SynAuto
0.89 ppmC MeOH
0.01 ppmC MeNOo
O.lOppmC HCHO
2.00 ppmC SynAuto
0.99 ppmC MeOH
0.01 ppmC MeNOo
2.00 ppmO SynAuto
0.70 ]>pmC MeOH
0.01 ppmC MeNO2
0.20 ppmC HCHO
0.666 ppmC SynAuto
0.666 ppmC SynAuto
0.300 ppmC MeOH
0.003 ppmO MeNO2
0.030 ppmC HCHO
PURPOSE
To determine the effect of 33% reduction in HC in an auto-
ez/iaust-like environment al a typical HC-to-NOx ratio. Expect
30% reduction in ozone maximum.
To determine the reactivity of the most likely mcthanol fuel
exhaust in an uu
-------
Table 9, cont'd. Methanol Fuel Reactivity Experiments
(All experiments have 0.35 ppm NOX)
Num
7
8
9
10
11
12
TYPE
Reduction, 33%
Substitution
normal HCHO
Substitution
high HCHO
Substitution
normal HCHO
Substitution
normal HCHO
Chemistry
REACTANTS
First. Side
3.00 ppmO SynUrban
3.00 ppmC SynUrban
3.00 ppmC SynUrban
1.000 ppmC SynUrban
6.00 ppmC SynAuto
l.OOppmC HCHO
REACTANTS
Second Side
2.00 ppmC SynUrban
2.00 ppmC SynUrban
0.80 ppmC MeOH
0.01 ppmC MeNO2
O.lOppmC HCHO
2.00 ppmC SynUrban
0.79 ppmC MeOH
0.01 ppmC MeNO2
0.20 ppmC HCHO
0.6fi6 ppmC SynUrban
0.300 ppmC MeOH
0.003 ppmC MeNO2
0.030 ppmC HCHO
4.00 ppmC SynAuto
1.78 ppmC MeOH
0.02 ppmC MeNO2
0.20 ppmC HCHO
1.00 ppmC HCHO
1.00 ppmC MeOH
PURPOSE
To determine the effect of 33% reduction in HC in an urban'
like environment at a typical HC-to-NOx ratio. Expect, 30%
reduction in ozone maximum
To determine the reactivity of the most likely methanol fuel
exhaust in an urfcan-like environment at a typical HC-to-NOx
ratio. Expect 20% reduction in ozone maximum.
To determine the reactivity of a highly reactive methanol fuel
exhaust in an ur6nn.-like environment at a typical HC-to-NOx
ratio. Expect 10%, reduction in ozone maximum.
To determine the reactivity of a typically reactive methniiol
fuel exhaust in an uriara-like environment at a low HC-to-NOx
ratio. Expect large reduction in ozone maximum.
To determine the reactivity of the most likely methanol fuel
exhaust in an aitto-exhaust-Yike environment at a high HC-to-
NOx ratio. Expect 20% reduction in ozone maximum.
To illustrate the chemistry of methanol in a highly reactive
environment.
-------
Table 10. Methanol Fuel Characterization Experiments
Num
1C
20
3C
40
50
fiO
TYPE
Cliaracter
Radical Src.
Character
Nitrogen Src.
Character
Match Test
Oliaracter
Match Test
Oliarncter
Match Test
Oliaracter
Photolysis Test
REACTANTS
First Side
0.35 ppm NOX
background air
0.0 ppni NOX
1.0 ppniC RCHO
0.50 ppm NOX
background air
0.35 ppni NOX
3.00 ppniC SynAuto
0.35 ppm NOX
3.00 ppmC UNCMIX
0.50 ppm NOX
1.00 pp.nC HCHO
REACTANTS
Second Side
0.35 ppm NOX
50.0 ppm CO
0.0 ppm NOX
1.0 ppmC RCHO
50.0 ppm CO
0.50 ppm NOX
background air
0.35 ppm NOX
3.00 ppmC SynAuto
0.35 ppm NOX
3.00 ppmC UNCMIX
0.50 ppm NOX
1.00 ppmO HCHO
50.0 ppm OO
PURPOSE
To test, for chamber sources of radicals capable of oxidizing
NO to NO2-
To test for chamber sources of Nitrogen Oxides.
To test for matched chamber sources of radicals capable of
oxidizing NO to NOo.
To test for matched performance.
To test for matched performance for comparison with past
studies.
To test photolysis rates in chamber.
-------
Methods
Outdoor Chamber and Analytical Facilities
The UNC Outdoor Smog Chamber is located in a rural area away from major pol-
lution sources." The chamber is a 300,000-liter, rigid external A-frame, Teflon
film chamber divided into two equal halves by a Teflon film wall. The dual cham-
ber design is used to perform side-by-side comparison experiments. It uses natural
sunlight and ambient temperatures and humidity. It is purged overnight with ru-
ral ambient air. Instruments are timeshared on the two sides. The chamber and
instrument operation and the data acquisition system are controlled by a PDP-11
computer. Complete facilities descriptions are given in Appendix A. Complete an-
alytical system descriptions are given in Appendix B. A complete description of the
data processing and quality assurance system is given in a separate volume of this
report that also includes a description of the magnetic tape data file format.15
x
Production of Synthetic HC Mixtures
SynAuto Mixture
The synthetic auto-exhaust mixture (SynAuto) used in this study was listed in
Table 3. Because of the need to have repeatable injections of this mixture over
the course of this program, an injection tank containing «10,000 ppm carbon was
needed. Ordering such a tank from gas suppliers requires up to three months for
delivery and the tank would have cost $1500. In addition, the tank contents would
have had to be analyzed by us to check the manufacturer's certification. Because
of the time constraint and the need to do the analysis anyway, we elected to make
26
-------
SynAuto Mixture Methods
our own injection tank. The time constraint, however, meant that the tank had to
be put to use before a complete analysis was finished.
The five aromatic compounds in the mixture would not have remained vaporized
in the tank and these compounds, as well as the tri-methyl-pentane, were made up
as a liquid mixture. The liquid mixture was injected by microliter syringe. The
remaining six compounds were mixed in a high pressure gas tank with N2- After
evacuating the tank, pressure was used to estimate the amount of each compound
added to the tank.
A SynAuto injection consists of three parts:
• a timed injection from the tank at a constant flowrate, accounting for 41% of
the carbon;
• a microliter syringe injection from the liquid mixture accounting for 57% of the
carbon;
• a subliming of a weighted amount of paraformaldehyde into the chamber air,
accounting for 2% of the carbon.
The analysis of initial conditions in several experiments and of high concentra-
tion injections into the chamber on calibration days were used to determine the
tank and liquid composition. The concentrations are based on calibration factors
obtained from standard, certified hydrocarbon tanks used to calibrate the gas chro-
matographs. The "Data Processing and Quality Assurance System Description"
document75 provides details on these calibration factors.
There are two items of concern for these mixtures, especially for the tank: the
internal ratios of compounds and the total carbon concentration in each. Table 11
shows both of these items for the desired ideal mixture and for the August 5,
supposedly 1.00 ppmC, injection of both the tank and the liquid.
In the injections, ethylene was 1.8% higher and each of the other compounds
in the tank were within ±0.6% of their target composition. In the liquid injection,
the analysis shows that tri-methyl-pentane was 1% too high, and that the aromatic
compounds were all within 0.5% of their target compositions. This is excellent
precision and accuracy.
The column marked "design-to-analyzed ratio", however, reveals a problem.
The tank had more total carbon than was thought. Likewise, more liquid was in-
jected than was thought. The target tank total concentration was 10,000 ppmC, but
27
-------
Methods
SynAuto Mixture
Table 11. Composition of Synthetic Auto Exhaust Mixtures.
Design Analyzed Design-to-Analyzed
Compound ppmC Percent ppmC Percent. Ratio
Tank Mixture
ethylene
propylene
1-butene
t-2-butene
butane
i-pentane
sub-total
0.239 58.2
0.042 10.1
0.020 4.8
0.020 4.8
0.039 9.5
0.052 12.6
0.411 100
Liquid Mixture
tri-me-pentane 0.112 19.2
benzene 0.054 9.2
toluene
m-xylene
o-xylene
tri-me-benzene
sub-total
Total Tank
Total Liquid
Total Inj
Table
0.212 36.2
0.103 17.6
0.048 8.2
O.OSfi 9.7
0.585 100
0.411 41.3
0.585 58.7
1.00
0.389
0.066
0.034
0.031
0.056
0.079
0.649
0.134
0.060
0.242
0.123
0.045
" 0.063
0.667
0.649
0.667
1.32
60.0
9.5
5.0
4.8
8.6
12.0
100
20.0
9.0
36.2
18.4
6.7
9.5
100
49.3
50.7
0.614
0.636
0.588
0.645
0.696
0.658
0.640
0.836
0.900
0.876
0.837
1.067
0.889
0.901
0.640
0.901
12. Final SynAuto Mixture Composition
Compound
butane
i-pentane
tri-me-pentane
ethylene
propylene
1-butene
t-2-butene
benzene
toluene
m-xylene
o-xylene
tri-me-benzene
total
PAR
OLE
ARO
Design
0.0391
0.0519
0.1121
0.2391
0.0416
0.0196
0.0196
0.0539
0.2115
0.1026
0.0481
0.0564
1.0000
0.2031
0.3199
0.4724
Composition
Actual
0.0405
0.0552
0.1063
0.2850
0.0433
0.0258
0.0238
0.0455
0.1942
0.0928
0.0365
0.0510
1.0000
0.2020
0.3780
0.4200
28
-------
SynUrban Mixture Methods
a quick analysis after it was made suggested that only 8.000 ppmC actually made it
into the tank. The 8,000 value was used as the basis for calculating injections for sev-
eral very good experiments before the preliminary analysis was completed. Table 11
shows that about 20% more tank carbon and about 10% more liquid carbon was
injected compared to the targets giving a total concentration of 1.32 ppmC instead
of 1.00 ppmC and a 49/51 tank/liquid mixture instead of the 42/58 tank/liquid
mixture desired. Because these experiments were so good, we decided to keep the
49/51 mixture for the rest of the work with the SynAuto mixture. This gives a new
final composition of the SynAuto Mixture as shown in Table 12.
The total effect of all the errors for the SynAuto mixture was to shift 5% of
the carbon from the aromatics class to the olefin class (essentially all in the form
of ethylene). The effect of this change on the overall reactivity of the mixture
is quite minor. Furthermore, this small variation is well within the variability of
exhaust composition as reported in both the UNC automobile study9 and the EPA
46-Vehicle exhaust study.16 The exhaust composition, for example, can be greatly
affected by the type of gasoline used. Finally, for testing photochemical mechanisms,
it is more important that the composition be well known than that the composition
exactly fits some target composition: the model must work for a whole range of
compositions and a 5% variation is very minor for model performance testing.
The analysis of the initial HC composition and concentrations for both sides of
four experiments, two near 3 ppmC and two near 1 ppmC, are shown in Table 13.
The HC initial concentrations on one side of each of these days were supposed to be
67% of the HC initial concentration on the other side (i.e. a 33% reduction). The
table shows that, even though the initial HC concentrations were about 30% too
high, the actual side-to-side ratio achieved for these four days was between 64%
and 69% compared to the target of 67% side-to-side ratio.
SynUrban Mixture
The SynUrban mixture consists of four parts:
• an injection from the UNCMIX tank, accounting for 54% of the carbon;
• an injection from an n-butane tank, accounting for 10%i of the carbon;
• a liquid injection from the aromatic mix portion of SynAuto mix, accounting for
36% of the carbon;
• a formaldehyde injection by subliming a weighted amount of paraformaldehyde
directly into the chamber.
29
-------
Table 13.
Initial Conditions in Four Methanol Experiments
CO
o
Compound
sthy lene
propy lene
1 -b u\.ene
t-2-butene
butane
1 -pentane
sub-tot
tr i -me-pentane
benzene
tol uene
ni-xyl . 15%
31.. 14%
li.53%
fi.70%
H.50%
S? . 17%
"'.>!. 06%
E . 42%
!-. . 1 0%
c.68%
1 . .48%
K", .42%
?.:: .11%
I .59%
:-.. .99%
: . 66'i
'• . 3 4 K
: . 72%
5::. .58%
ratio
0.658
0.688
0.616
0.689
0.641
0.701
0.664
0.691
0.657
0.622
0.649
0.667
0.622
0.647
0.655
ratio
0.671
0.684
0.680
0.650
0.653
0.692
0.672
0.681
0.683
0.672
0.674
0.675
0.672
0.675
0.674
-------
Synthetic Methanol Exhaust Methods
The UNCMIX was a 2% certified tank from the manufacturer; the n-butane tank
was a commercial blend of 10.000 ppmC of n-butane in N-_>; the liquid analysis was
described above. The paraformaldehyde purity was based upon manufacturer assay
and was 92% pure.
Synthetic Methanol Exhaust
A synthetic methanol exhaust injection was in three parts:
• a microliter syringe injection of pure MeOH. accounting for 79 to 100% of the
substituted carbon;
• an additional formaldehyde injection made by subliming a weighed amount of
paraformaldehyde directly into the chamber, accounting for 0 to 20% of the
substituted carbon;
• an optional injection of Mel\IC>2 made by flushing a known pressure of pure MeNOz
in a known volume into the chamber, accounting for 0 to 1% of the substituted
carbon.
Production of Methyl Nitrite
MeNO2 boils at -12°C and is unstable; it can not be placed in a tank or kept in a bag.
It photolyzes readily to produce radicals. MeNO2 must therefore be synthesized and
purified in a laboratory and stored in liquid nitrogen. The UNC glass shop produced
the needed apparatus. Figure 3 shows the setup for synthesis, storage, and injection.
We produced the needed quantities of MeNO2 each week and stored and transported
it in liquid HZ-
-------
methyl nitrite synthesis
saturated
in methanol
dry ice/
acetone bat
Dewar
Figure 3. Methyl Nitrite Production Apparatus
-------
methyl nitrite storage and sampling
liq. N2
bath
^j>
volumetric JL
gas sample
tube
vacuum
gauge
vacuum
pump
Figure 3. Methyl Nitrite Production Apparatus
-------
Results
Overview of Data Set
The experimental work began in mid-July, 1984 and had to end by September 22.
1984. July had 6.22" of rain and the whole year was 12" above the normal rainfall
at the beginning of August. August, however, was 2" below normal rainfall for the
month. The first part of September had fairly good sun, but Hurricane Diana came
to NC for five days in the middle of the month. In addition, the UNCMIX tank was
totally exhausted in early September preventing completion of the SynUrban series.
In 1985, three experiments were conducted in June with a new UNCMIX tank; two of
these were successful, sufficiently completing the SynUrban series for adequate model
testing.
The sequence of experimental testing was determined by a set of priorities and
by estimates of having successfully completed higher priority runs based upon ex-
amination of run results immediately after each run. The priorities were: "
• conduct experiments with SynAuto first;
> demonstrate matched reactivity in both chambers with SynAuto mix;
> demonstrate effects of 33% reduction at 3.0 ppmC;
t> demonstrate effects of 33% substitution with normal HCHO methanol ex-
haust at 3.0 ppmC;
t> demonstrate effects of 33% reduction at 1.0 ppmC;
t> demonstrate effects of 33% substitution with normal HCHO methanol ex-
haust at 1.0 ppmC;
t> demonstrate effects of 33% substitution with low HCHO methanol exhaust;
34
-------
Overview of Data Set Results
i> demonstrate effects of 33% substitution with high HCHO methanol ex-
haust;
• conduct experiments with SynUrban second;
> demonstrate effects of 33% reduction at 3.0 ppmC;
> demonstrate effects of 33% substitution with normal HCHO methanol ex-
haust at 3.0 ppmC;
> demonstrate effects of 33% reduction at 1.0 ppmC;
> demonstrate effects of 33% substitution with normal HCHO methanol ex-
haust at 1.0 ppmC;
> demonstrate effects of 33% substitution with low HCHO methanol exhaust;
When a particular experiment was not successful (usually because of bad weather),
the run was repeated as soon as possible before proceeding to lower priority runs.
Superimposed upon the experimental goals were the operational needs of pro-
viding actual injections for analysis of the blended SynAuto tank composition, of
determining abilities to inject a three part mixture in the proper ratios and for
testing general chamber performance.
Table 14 is a summary table of the conditions for 29 dual experiments conducted
in this program.
In the table, the column headed SegFile gives the name of the experimental
data file on the magnetic tape; a blank in this column means that the run was
not considered satisfactory for model testing, usually because of poor sunlight,
and it was not processed other than to produce the data plots—these runs have
approximate initial conditions. There were 11 such runs excluded from the tape.
There are 18 runs with completely processed data that are included on the magnetic
data tape for distribution to model testers.
The HC Mix column indicates the type of mix used; in addition to the SynAuto,
SynUrban, and UNCMIX mixes described in Chapter 2, three other mixtures were used
for a few experiments:
• a first version of the SynAuto mixture that had incorrect internal ratios of tank
species, designated as SynAutol in Table 14; this was discarded after July 22,
1984.
• a SynAutUrb mixture that was used in late September after the UNCMIX tank had
been exhausted. This mixture was a blend of 90% SynAuto mixture and 10%
n-butane used to establish a more urban-like mixture than SynAuto. This was
35
-------
Results Detailed Experimental Data and Plots
used in four experiments; three of these were satisfactory for model testing and
are on the magnetic tape. These provide yet a third and related compositional
mix to test mechanism responses to compositional changes.
• a mixture called HMWMIX—High Molecular Weight Mixture—was just the liquid
portion of the SynAuto and SynUrban mixtures. In this mixture, five aromatics
comprised 80% of the carbon and tri-methyl-pentane was the other 20% of the
carbon.
Detailed Experimental Data and Plots
This section lists details of the initial conditions and other information for each of
the 29 methanol experiments performed in this program. Plots of the NO, NC>2, and
03 profiles and for the chamber air temperature, dew point and ambient total solar
and ultraviolet radiation (TSR and UV) are shown.
Example Description of Information Presented
We will use August 4, 1985 as an example to explain the information presented. The
reader should turn to page 64 to examine the information presented for August 4
and to look at the plot on the next page before reading the following description.
The first page of information is a summary of the experimental initial conditions.
At the top of this page is the run date and a general description of the run conditions,
e.g.
SynAuto 1.2 ppmC vs 0.83 ppmC/0.3 ppm MeOH/0.028 ppmC HCHO
This means that the experiment was a substitution experiment with the SynAuto
mixture and that the SynMethanol mixture did contain formaldehyde at approximately
10%.
Below the title section is listed information from the run documentation file
on the magnetic tape. This information gives the basic results of, the run, in this
case, the Oz maximum concentrations produced and the time of the maximum in
(). Then the initial conditions are listed for both sides of the chamber. Some of the
entries require explanation.
NMHC This value is the sum of all the injected HC. It therefore includes the
tank and liquid injections for the SynAuto, the MeOH injection and the
HCHO injection. It is the total amount of injected organic carbon available
for reaction (it does not include any chamber background organics; see
modeling discussion in Chapter 6).
36
-------
Example Description of Information Presented Results
SYH-AUTO This is the tank and liquid portion of the SynAuto mixture injection. It
does not include the formaldehyde portion of the SynAuto mixture.
MEOH This is the amount of MeOH injected. It also does not include any
formaldehyde.
HCHO This is the total HCHO injection. It includes the HCHO component of the
SynAuto as well as any HCHO that was part of the surrogate methanol
exhaust.
HC species
These are the component HC species of the SynAuto mixture.
An examination of the above values for the August 4, 1984 experiment reveals
the following:
• The substituted side had 6% less carbon than the baseline side (1.179 vs.
1.249 ppmC). This was because the SynAuto mixture was reduced on the BLUE
side to 0.664 of the RED side concentration by omitting 0.41 ppmC of the
SynAuto carbon, but only 0.35 ppmC of SynMethanol was added back, an error of
0.06 ppmC.
• Too little HCHO was injected on both sides. On the RED, baseline side, the actual
composition of the SynAuto mixture for this day was 1.52% HCHO rather than the
target 2%. Likewise, the amount of the total HCHO injected on the BLUE side
was too low; of the 0.041 ppm HCHO injected, 0.013 ppm was for the SynAuto
mixture, leaving 0.028 ppm for the SynMethanol. This resulted in the SynMethanol
mixture being 8% formaldehyde instead of the target 10% formaldehyde.
This precision for target conditions is typical—it is extremely difficult to achieve
much better than 2% overall reproducibility.
In the concentration plot for August 4, the RED chamber data are shown as
solid lines and the BLUE chamber data are shown as dashed lines (this is true for
all concentration data plots). The data in these plots are taken every four minutes,
but alternate sides so that the data for one side is every eight minutes. In the
meterological data plots, TSR is shown as a solid line and UV is shown as a fine
dashed line. Dewpoint is shown on both sides of the chamber. The data in these
plots are taken every four minutes.
On the August 4, 1984 data plots (Figure 14) there are three "holes," one about
0850-0910 EOT, one about 1150-1330 EOT and one about 1615 EDT. These were
periods when the computer data collection was stopped, either because of computer
problems (e.g. power failure) or because the operator needed to transfer data or
37
-------
Results Example Description of Information Presented
perform other required maintenance. On this day. the large hole caused by power
problems at the rural site.
In the meteorological data plot of Figure 14, the chamber was initially at sat-
uration (dewpoint — air temperature), but as the sun warmed the chamber, the
air temperature rose significantly above the dewpoint, which remained relatively
constant at about 80°F. The air temperature rose to 102°F. Small cumulus clouds
in the afternoon frequently passed in front of the sun, blocking direct sunlight from
the chamber and sensors, causing the "dips" in the TSR and UV data. The sunlight
on this day is considered to be "good," approximately an 8 on a scale of 0-10. The
sunlight on August 5, 1984 is better and the sunlight on July 17 is "bad," i.e. a 0
or 1 on a scale of 0-10. Experiments with such poor light are essentially useless for
model testing and no conclusions should be based upon their outcomes. They are
included here for completeness.
It should be recalled that the chemiluminscent NOX meter used in our work
responds 100% to PAN as well as to NC>2, but not to HNO3. Therefore, the data after
the N02 peak are actually the sum of PAN and NO2. Processed data sets contain
independent data for PAN based upon gas chromatography. In addition, there is a
12-15 second transport time from the chamber to the instruments and thus reaction
can occur between 0° and NO when both are not near zero. This causes O3 and NO
to be lower, and N02 to be higher, than the value in the chamber. This process is
only important between NO-and-N02 crossover and NO2 maximum and only effects
the concentrations by about 10 ppb.
38
-------
Table 14. Summary of Experimental Initial Concentrations
(units are ppm or ppmC)
(BLUE data on first line, RED data on second line)
Date
July 17, 1984
July 19, 1984
July 20, 1984
July 21, 1984
July 22, 1984
July 25, 1984
July 26, 1984
July 28, 1984
SegFile
JL2584
HC Mix Type
SynAutol mid ratio
SynAutol matched
SynAutol mid ratio
SynAutol matched
SynAutol mid ratio
SynAutol matched
UNCM1X 3.3% sub. low HCHO
UNCM1X mid ratio
SynAuto mid ratio
SynAuto 33% sub. low HCHO ,
SynAuto 33% sub. norm HCHO
SynAuto mid ratio
SynAuto mid ratio
SynAuto 33% sub. low HCHO
SynAuto 33% sub. low HCHO
SynAuto mid ratio
NOX NO N02 HC MeOH HCHO
0.33 0.18 0.15 3.54 0.00 0.00
0.33 0.20 0.12 3.54 0.00 0.00
0.30 0.21 0.09 3.54 0.00 0.00
0.31 0.22 0.09 3.54 0.00 0.00
0.25 0.20 0.05 3.54 0.00 0.00
0.26 0.20 0.05 3.54 0.00 0.00
0.26 0.21 0.05 2.36 0.98 0.00
0.27 0.22 0.05 3.54 0.00 0.00
0.21 .0.15 0.05 3.54 0.00 0.00
0.21 0.16 0.06 2.36 0.98 0.00
0.33 0.25 0.08 2.36 0.90 0.10
0.33 0.25 0.08 3.54 0.00 0.00
0.31 0.24 0.07 3.54 0.00 0.00
0.31 0.24 0.07 2.36 0.99 0.00
0.30 0.23 0.07 2.36 0.85 0.00
0.30 0.23 0.07 3.54 0.00 . 0.00
-------
Table 14. Summary of Experimental Initial Concentrations, cont.
(units are ppm or ppmC)
(BLUR dat.a on first line, RED data on second line)
Date
Aug. 2, 1984
Aug. 3, 1984
Aug. 4, 1984
Aug. 5, 1984
Aug. 6, 1984
Aug. 7, 1984
Aug. 8, 1984
Aug. 9, 1984
Aug. 22, 1984
Aug. 25, 1984
SegFile
AU0484
AU0584
AU0684
AU0784
AU0884
AU0984
AU2284
AU2584
HO Mix Type
SyuAul.o mid ratio
SynAuto 3.'$% reduction
SynAuto mid ratio
SynAuto 33% sub. norm HCHO
SynAuto low ratio
SynAuto 33% sub. norm HCHO
SynAuto low ratio
SynAuto 33% reduction
SynAuto mid rntio
SynAuto 33% reduction <
SynAuto 33% sub. norm HCHO
SynAuto low ratio
SynAuto mid ratio
SynAuto 33% sub. high HCHO
SynAuto 33% sub. high HCHO
SynAuto low ratio
Sy n Urban 33% sub. norm HCHO
SynUrban . mid ratio
Sy n Urban low ratio
SynUrban 33% sub. norm HCHO
NOX NO N02 HC McOH HCHO
0.41 0.31 0.09 3.54 0.00 0.00
0.42 0.32 0.10 2.36 0.00 0.00
0.44 0.32 0.12 3.73 0.00 O.OG
0.45 0.32 0.13 2.62 0.90 0.13
0.36 0.28 0.07 0.82 0.32 0.04
0.37 0.29 0.08 1.23 0.00 0.02
0.35 0.27 0.08 1.32 0.00 0.02
0.35 0.27 0.08 0.91 0.00 0.01
0.36 0.28 0.07 3.24 0.00 0.06
0.35 0.28 0.07 2.25 0.00 0.04
0.38 0.30 0.08 0.87 0.30 0.06
0.39 0.30 0.08 1.32 0.00 0.04
0.34 0.2C 0.08 3.68 0.00 0.06
0.34 0.26 0.08 2.48 0.79 0.23
0.39 0.30 0.09 0.86 0.26 0.08
0.39 0.30 0.09 1.28 0.00 0.02
0.32 0.25 0.07 2.04 0.87 0.13
0.32 0.25 0.07 3.04 0.00 0.06
0.34 0.27 0.07 1.07 0.00 0.02
0.35 0.27 0.08 0.72 0.29 0.04
-------
Table 14. Summary of Experimental Initial Concentrations, cont.
(units are ppm or ppmC)
(BLUE data, on first line, RED data on second line)
Date
Aug. 28, 1984
Sept. 1, 1984
Sept.. 2, 1984
Sept. 3, 1984
Sept. 8, 1984
Sept.. 9, 1984
Sept.. 17, 1984
Sept.. 19, 1984
Sept.21, 1984
SegFile
ST0184
ST0284
ST0384
ST0884
ST1784
ST1984
ST2184
HC Mix Type
SynUrban low ratio
SynUrban 33% sub. low HCHO
SynUrban mid ratio
SynUrban 33% sub. low HCHO
SynUrban 33% reduction
SynUrban low ratio
SynUrban low ratio
SynUrban 33% sub. high HCHO
SynAntUrb mid ratio
SynAutUrb 33% reduction
SynAntUrb low-mid ratio
SynAntUrb 33% reduction
SynAntUrb 30% sub norm HCHO
SynAutUrb low-mid ratio
IIMWMIX 40% reduction
I1MWMIX
SynAutUrb low-mid ratio -1 HCHO
SynAntUrb low-mid ratio
NOX NO NO2 HC MeOH HCHO
0.31 0.25 O.Ofi 1.16 0.00 0.02
0.32 0.26 0.07 0.81 0.32 0.01
0.30 0.24 0.06 3.31 0.00 0.06
0.31 0.25 0.06 2.66 0.97 0.04
0.32 0.26 0.06 0.76 0.00 0.01
0.34 0.28 0.06 1.10 0.00 0.02
0.35 0.24 0.11 0.78 0.26 0.07
0.35 0.24 0.11 1.10 0.00 0.02
0.33 0.21 0.13 2.79 0.00 0.05
0.33 0.20 0.12 1.84 0.00 0.04
0.34 0.29 0.06 2.14 0.00 0.04
0.35 0.29 0.06 1.42 0.00 0.02
0.34 0.27 0.07 1.42 0.57 0.08
0.34 0.27 0.07 2.14 0.00 0.04
0.34 0.25 0.09 2.63 0.00 0.00
0.34 0.25 0.09 4.43 0.00 0.00
0.36 0.27 0.09 2.42 0.00 0.18
0.36 0.28 0.09 2.43 0.00 0.00
-------
Table 14. Summary of Experimental Initial Concentrations, cont.
(units arc ppm or ppmC)
(BLUE data on first line, RED data on second line)
Date
June 26, 1985
June 28, 1985
SegFile
JN2685
JN2885
HC Mix Type
SynUrhan mid ratio
SynUrban 33% reduction
SynUrban low-mid ratio
SynUrban 33% sub norm HCHO
NOX NO NO2 IIC McOH HCHO
0.35 0.28 0.07 4.01 0.00 0.06
0.35 0.28 0.07 2.44 0.00 0.04
0.35 0.28 0.07 1.78 0.58 0.09
0.35 0.28 0.07 2.65 0.00 0.04
-------
July 17, 1984
SynAutol Matched 3.5 ppmC (no HCHO)
RESULTS: 03 MAX: BLUE 0.4605 PPM(1720); RED 0.4424 PPM(1724) .
INITIAL CONDITIONS: BLUE RED
MO 0.181 0.201
N02 0.147 0.124
NMHC " 3.540 3.540
SYN-AUTO(TANKtLIQUID) 3.540 3.540
RUN NOT PROCESSED
43
-------
I
CO
I
<§
0)
c?
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
i i r i i i i i i T r i i i iill i r
1 T 1 ' I
I^^T 1 * T
- NO
July 17, 1984
O3 -
NO2
6 7 8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
"
2.0
co
X
0)
I
11.0
&
e
'0.5
0.0
I "I ' T
i i i 1 r T r
100
90
80
70
60
50
40
30
20
10
o>
p
I
c
i
5
£.
o'
8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
Figure 4.
Top: (Solid) 3.54 ppmc SYNAUTO (FIRST), no HCHO, no MeOH;
(Dashed) 3.54 ppmC SYNAUTO (FIRST), no HCHO, no MeOH;
Bottom: RED chamber air temperature (top solid line, °F);
RED (solid line) and BLUE (dashed line) chamber dewpoint (°F);
ambient total solar radaiation (solid line, cal-cm" -sec" ):
ambient ultraviolet radiation (dashed linejncal-cm" -sec" ).
44
-------
July 19, 1984
SynAutol Matched 3.5 ppmC (no ECHO and MeOH)
RESULTS: 03 MAX: BLUE 0.8058 PPM(1304); RED 0.7889 PPK(1300).
INITIAL CONDITIONS: BLUE RED
NO 0.214 0.224
N02 0.090 0.085
NMHC ' 3.540 3.540
SYN-AUTO(TANKtLIQUID) 3.540 3.540
ETHYLENE
PROPYLENE
1-BUTENE 0.068 0.071
H-BUTANE 0.130 0.135
TRAHS-2-BUTENE 0.074 0.075
ISOPE!!TAHE 0.173 0.155
2,2.4-TRIKETHYLPEKTAHE
BE1!ZE!!E 0.157 0.160
TOLUENE 0.683 0.728
!-!-XYLEt!E 0.347 0.353
0-XYLE1IE 0.133 0.165
RUN HOT PROCESSED
45
-------
I.W
0.9
C U»O
Q.
a 0.7
1 0.6
X
0 0.5
c
0)
o) 0.4
I 0.3
0.2
0.1
n n
i | i | i | i | i | i | i | i | i | i | i | i | i | i ^
- Jijjy 19, 1984 -
'^ ~M~^~~^~\r^
- r"~::^
-
— —
: NO N02 :
~- ^\/ ^— _ ~-
~ i i i i . i i J
-------
July 20, 1984
SynAutol Matched 3.5 ppmC (no HCHO and MeOH)
RESULTS: 03 MAX: BLUE 0.4649 PPM(1648); RED 0.5157 PPM(1636).
INITIAL COIiDITIOHS: BLUE RED
HO 0.199 0.202
1102 0.050 0.053
NMHC " 3.540 3.540
SY!i-AUTO(TA!IKtLIQUID) 3.540 3.540
ETHYLEliE
2.2,4-TRIMETHYLPENTA1IE
BEMZEME 0.152 0.103
TOLUENE 0.665 0.459
M-XYLEUE 0.356 0.242
0-XYLEUE 0.124 0.103
RUH HOT PROCESSED
47
-------
I
a
0)
TJ
O n* _
a>
July 20, 1984
g> 0.4 -
8 9 10 11 12 13 14 15 16 17 18
HOURS, EOT
19
2.0
*
0)1.5
.o
£1.0 -
0.5 -
0.0
5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
Figure 6.
Top: (Solid lines) 3.54 ppmC SYNAUTO (FIRST), no HCHO, no MeOH;
(Dashed lines) 3.54 ppmC SYNAUTO (FIRST), no HCHO, no MeOH;
Bottom: RED chamber air temperature (top solid line, °F);
RED (solid line) and BLUE (dashed line) chamber dewpoint (T);
ambient total solar radaiation (solid line, cal-cm" -see' ):
ambient ultraviolet radiation (dashed line^ncal-cm" -sec" ).
48
-------
July 21, 1984
UNCMIX, subst., 3.0 ppmC vs 2.0 ppmC and 1 ppm MeOH (no HCHO)
RESULTS: 03 MAX: BLUE 0.4250 PPM(1712); RED 0.6060 PPM(1700).
INITIAL COHDITIOHS: BLUE RED
110 0.210 0.218
N02 0.047 0.050
UNCMIX " 2.360 3.540
MEOH 0.980 0.000
ETHYLEME
PROPYLEUE 0.086 0.127
1-BUTEHE 0.055 0.080
TRAUS-2-BUTEHE 0.058 0.071
ISOPEHTANE . 0.274 0.402
!!-PE!ITA!IE 0~.457 0.694
RU1! HOT PROCESSED
49
-------
I
a
0)
0)
I I I I I I I I I I I I I I I I I I I
July 21, 1984
9 10 11 12 13 14 15 16 17 18
HOURS, EOT
19
2.0
at
X
o
1.0
0.5
0.0
iiiiiiir
I ' I ' I ' I
5 6 7 8 9 10 11 12 13 14 15 16 17 18
HOURS, EOT
19
100
90
80
70
60
50
40
30
20
10
0
f
C
i
a
5
Figure 7.
Top: (Solid) 3.54 ppmC UNCMK, no MeOH, no HCHO;
(Dashed) 2.36 ppmC UNCMIX, 0.98 ppm MeOH, no HCHO;
Bottom: RED chamber air temperature (top solid line, °F);
RED (solid line) and BLUE (dashed line) chamber dewpoint (°F>,
ambient total solar radaiation (solid line, cal-cm~ -sec" ):
ambient ultraviolet radiation (dashed line^ncal-cm" -sec" ).
50
-------
July 22, 1984
SynAuto 3.5 ppmC vs 2.0 ppmC/1.0 ppm MeOH (no HCHO)
RESULTS: 03 MAX: BLUE 0.3843 PPM(1736); RED 0.3359 PPH(1756).
INITIAL CONDITIONS: BLUE RED
110 0.153 0.155
N02 0.054 0.057
HMHC - 3.540 3.340
SYN-URBAtKTAHK&LIQUID) 3.540 2.360
MEOH 0.000 0.980
ETHYLEHE
PROPYLENE 0.035 0.028
ISOPEMTANE 0.115 0.097
N-PEHTANE 0.195 0.146
2,2.4-TRIMETHYLPENTANE 0.319 0.235
BEI1ZEHE 0.131 0.101
TOLUENE 0.645 0.457
M-XYLENE 0.310 0.226
0-XYLEHE 0.135 0.123
1.2,4-TRIMETHYLBENZEIIE 0.236 0.193
RUN HOT PROCESSED
51
-------
a
a
OT
T3
2
a>
z
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1 h
0.0
1 I ' I ' I ' I ' I ' I ' I ' I ' I ' I ' I ' I ' I '
July 22, 1984
- NO
6 7 8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
1.0
0.9
0.8
0.7
0.6 |
0.5 •»
TJ
0.4 1
0.3
0.2
0.1
0.0
2.0
in
.
eS
.9
1.0
'0.5
0.0
I T | I I I
r i i i i
I ' I
5 6 7 8 9 10 11 12 13 14 15 16 17 18
HOURS, EOT
19
100
90
80
70
60
50
40
30
20
10
0
CD
3
p
\
I
c
Figure 8.
Top: (Solid) 2.36 ppmC SYNAUTO (FIRST), 0.98 ppm MeOH, no HCHO;
(Dashed) 3^4 ppmC SYNAUTO (FIRST), no MeOH, no HCHO;
Bottom: RED chamber air temperature (top solid line, °F);
RED (solid line) and BLUE (dashed line) chamber dewpoint (T);
ambient total solar radaiation (solid line, cal-cm" -sec" ):
ambient ultraviolet radiation (dashed linejncal-cm" -sec~ ).
52
-------
July 25, 1984
SynAuto 3.5 ppmC vs 2.0 ppmC plus 0.9 ppm MeOH and 0.1 HCHO
RESULTS: 03 MAX: BLUE 0.7711 PPM(1352) ; RED 0.7655 PPM(1332).
INITIAL CONDITIONS: BLUE RED
HO 0.249 0.254
IJ02 0.076 0.076
1JMHC . 3.540 3.540
SYH-URBAN(TAMKtLIQUID) 2.360 3.540
MEOH 0.900 0.000
HCHO 0.100 0.000
53
-------
Q.
Q.
6
0)
i i i i i i i i i
July 25^1984
8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
0.0
2.0
w
.
a
•Mw h
«
o
-------
July 26, 1984
SunAuto 3.5 ppmC vs 2.4 ppmC plus 1.0 ppm MeOH (no HCHO)
RESULTS: 03 MAX: BLUE 0.7050 PPM(1520) ; RED 0.7211 PPH(1S24) .
INITIAL CONDITIONS: BLUE RED
NO 0.238 0.241
1102 0.071 0.071
NMHC ' 3.540 3.350
SYN-URBAN(TANK4LIQUID) 3.540 2.360
MEOH 0.000 0.990
RUN MOT PROCESSED
55
-------
Q.
Q.
CO
<0
TJ
'x
O
fl>
o>
1 ' I ' I ' I ' I ' I ' I ' I ' I ' I ' I ' I
July 26, 1984
O3
8 9 10 11 12 13 14 15 16 17 18
HOURS, EOT
2.0
CO
.
«J
1.0
'0.5
0.0
5 6 7 8 9 10 11 12 13 14 15 16 17 18
HOURS, EOT
19
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
l
100
90
80
70
60
50
40
30
20
10
0
(0
p
O
CD
TJ
c
i
a
o
Figure 10.
Top: (Solid) 2.36 ppmC SYNAUTO, 0.99 ppm MeOH, no HCHO;
(Dashed) 3.54 ppmC SYNAUTO, no MeOH, no HCHO;
Bottom: RED chamber air temperature (top solid line, °F);
RED (solid line) and BLUE (dashed line) chamber dewpoint (°F);
ambient total solar radaiation (solid line, cal-cm" -sec" ):
ambient ultraviolet radiation (dashed linejncal-cm" -sec" ).
56
-------
July 28, 1984
SynAuto 3.5 ppmC vs 2.4 ppmC plus 0.8 ppm MeOH (no HCHO)
RESULTS: 03 MAX: BLUE 0.5050 PPM(1440); RED 0.4867 PPM(1436).
INITIAL C01IDITIOKS: BLUE RED
NO 0.229 0.231
N02 0.067 0.070
HMHC " 3.210 3.540
SYN-AUTO(TANK&LIQUID) 2.360 3.540
MEOH 0.850 0.000
RU1! NOT PROCESSED
57
-------
at
3
0)
g>
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
1 I ' I ' I ' I ' I ' I ' I ' I ' I ' I ' I ' I
July 28, 1984
NO
2.0
(A
"
BJ
O
11.0
13
'0.5
0.0
O3
8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
I
100
90
80
70
60
50
40
30
20
10
O
(D
I
c
I
8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
Figure 11.
Top: (Solid) 3.54 ppmC SYNAUTO, NO MeOH, no HCHO;
(Dashed) 2.36 ppmC SYNAUTO, 0.85 ppm MeOH, no HCHO;
Bottom: RED chamber air temperature (top solid line, °F);
RED (solid line) and BLUE (dashed line) chamber dewpoint (°F);
ambient total solar radaiation (solid line, cal-cnf -«ec" ):
ambient ultraviolet radiation (dashed line,mcal-cm~ -sec~ ).
58
-------
August 2, 1984
SynAuto Reduction 3.5 to 2.4 ppmC (no ECHO)
INITIAL CONDITIONS:
NO
N02
NMHC
SYN-AUTO(TANKfcLIQUID)
ETHYLENE -
PROPYLENE
1-BUTEME
N-BUTANE
TRANS-2-BUTENE
ISOPENTAHE
2,2.4-TRIMETHYLPENTANE
BENZENE
TOLUENE
M-XYLEIIE
0-XYLE!iE
BLUE
0.312
0.094
3.540
3.540
100730 1.099
00730 0.177
(00730 0.117
80730 0.173
80730 0.082
80730 0.194
80730 0.344
(00730 0.126
80730 0.695
80730 0.345
80730 0.143
RED
0.318
0.099
2.360
2 . 360
0.776
80700 0.122
80700 0.081
80700 0.121
80700 0.060
80700 0.145
80700 0.238
80700 0.091
80700 0.490
80700 0.265
80700 0.099
RESULTS: 03 MAX: BLUE 0.8712 PPM(1312); RED 0.7996 PPM(1612) .
RUN NOT PROCESSED
59
-------
1.0
0.9
E 0.8
a
a 0.7
OT
0.6
0.5
a)
g> 0.4
z 0.3
0.2
0.1
0.0
iii
NO
I ' I ' I ' I ' I ' I ' I ' I '
August. 2,. 19 84
1,1,1,1,1,1
8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
1.0
0.9
0.8
0.7
0.6 |
0.5 -»
TJ
0.4 |
0.3
0.2
0.1
0.0
2.0
w
.
«J
1.0
'0.5
0.0
5678
10 11 12 13 14 15 16 17 18 19
HOURS, EOT
100
90 H
CD
80 p
70 £
S
eo ;s
50 g
40 i?
a
30 K-.
o
20 3
10
0
N
Figure 12.
Top: (Solid) 2.36 ppmC SYNAUTO, no MeOH, no HCHO;
(Dashed) 3.54 ppmC SYNAUTO, no MeOH, no HCHO;
Bottom: RED chamber air temperature (top solid line, °F);
RFJD (solid line) and BLUE (dashed line) chamber dewpoint (°F);
ambient total solar radaiation (solid line, cal-cm" -sec" ):
ambient ultraviolet radiation (dashed lineoncal-cm' -sec" ).
60
-------
August 3, 1984
SynAuto 3.79 ppmC vs 2.67 ppmC plus 0.9 ppm MeOH and 0.09 HCHO
RESULTS: 03 MAX: BLUE 0.8927 PPH(1416); RED 0.8900 PPM(1412).
INITIAL COI.'DITIOIIS: BLUE RED
NO 0.317 0.324
H02 0.120 0.129
IIMHC • 3.785 3.656
SYN-AUTO(TANK&LIQUID) 3.728 2.625
MEOH 0.000 0.897
HCHO 0.057 0.134
ETHYLENE 1.055 0.718
PROPYLENE 0.210 0.149
1-BUTENE 0.111 0.077
TRANS-2-BUTEHE 0.098 0.070
M-BUTAHE 0.176 0.116
ISOPEKTANE 0.278 0.193
2.2.4-TRIMETHYLPE!!TA!IE 0.363 0.260
BENZEKE 0.162 0.119
TOLUEME 0.667 0.484
M-XYLE1IE 0.311 0.220
0-XYLENE 0.122 0.093
1.2.4-TRIMETHYLBEL'ZElIE 0.175 0.127
RUN NOT PROCESSED
61
-------
a
o.
(A
•§
O n* _
(D
I
I I I I I I I I I I I I I I
1 ' I ' I ' ' I
8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
0.0
2.0
g>1.5
.g
1.0
Si
0.5
0.0
I i I
o>
100
90
80
70 %
60 £
50 C
40 $
a
30 £
o
20 3
10
5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
Figure 13.
Top: (Solid) 2.62 ppmC SYNAUTO, 0.9 ppm MeOH + 0.13 ppm HCHO;
(Dashed) 3.73 ppmC SYNAUTO, no MeOH, 0.06 ppm HCHO;
Bottom: RED chamber air temperature (top solid line, °F);
RED (solid line) and BLUE (dashed line) chamber dewpoint (°F);
ambient total solar radaiation (solid line, cal-cm~ -sec" ):
ambient ultraviolet radiation (dashed line^ncal-cm" -sec ).
62
-------
August 4, 1984
SynAuto 1.25 ppmC vs 0.83 ppmC/0.3 ppm MeOH/0.028 ppmC HCHO
RESULTS: 03 MAX: BLUE 0.3284 PPH(1744); RED 0.5154 PPM(1732).
INITIAL CONDITIONS: BLUE RED
1JO 0.284 0.293
N02 0.071 0.077
HMHC - 1.178 1.250
SYN-AUTO(TAMK&LIQUID) 0.817 1.231
HEOH 0.320 0.000
HCHO 0.041 0.019
ETHYLEME 0.244 0.365
PROPYLE1IE 0.034 0.053
1-BUTEIJE 0.021 0.031
TRAHS-2-BUTENE 0.019- 0.029
1J-BUTA1IE 0.031 0.046
ISOPEMTAHE 0.047 0.067
2,2.4-TRIMETHYLPE1ITAIIE 0.085 0.127
BEHZEHE 0.036 0.055
TOLUENE 0.154 0.236
H-XYLEtlE 0.076 0.115
0-XYLEHE 0.029 0.045
1.2,4-TRIMETHYLBENZENE 0.041 0.062
63
-------
I.U
0.9
E 0.8
Q.
a 0.7
in
1 0.6
0 0.5
c
''""" ••'<'''.
\^ y "^^^ -
^7 \ . i i i i i i-ife^iiM.. i i i i i i i i i TTT~I i ~
I.U
0.9
0.8
0.7
o
0.6 g
3
0.5 •*
T3
0.4 3
0.3
0.2
0.1
nn
8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
w
X
JD
C
nl
0.0
8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
Figure 14.
Top: (Solid) 1.23 ppmC SYNAUTO, no MeOH, 0.02 ppm HCHO;
(Dashed) 0.82 ppmC SYNAUTO, 0.32 ppm MeOH + 0.04 ppmC HCHO;
Bottom: RED chamber air temperature (top solid line, °F);
RED (solid line) and BLUE (dashed line) chamber dewpoint (°F>,
ambient total solar radaiation (solid line, cal-cm -sec" )•
ambient ultraviolet radiation (dashed line,mcal-cm -sec ).
64
-------
August 5, 1984
SynAuto 1.3 ppmC vs 0.93 ppmC
RESULTS: 03 MAX: BLUE 0.5946 PPMU728); RED 0.3352 PPH(1756)
I1JITIAL CONDITIONS: BLUE RED
110 0.268 0.272
1102 0.077 0.079
1IMHC - 1.335 0.926
SY!i-AUTO(TAMK&LIQUID) 1.316 0.914
HCHO 0.019 0.012
ETHYLE1IE 0.389 0.265
PROPYLE1IE 0.062 0.042
1-BUTE11E 0.034 0.022
TRANS-2-BUTEHE 0.031 0.021
II-BUTANE 0.054 0.038
ISOPEi.'TAi.'E 0.079 0.053
2,2. 4-TRIMETHYLPEI.'TA!IE 0.134 0. 099 '
BENZENE 0.060 0.041
TOLUENE 0.242 0.172
M-XYLENE 0.123 0.085
0-XYLENE 0.045 0.031
1.2,4-TRIMETHYLBENZENE .0.064 0.045
65
-------
a
a
o>
§>
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
NO
1 I ' I ' I ' I ' I '
August 5, 1984
6 7 8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
2.0
0)1.5
(0
1.0
'0.5
0.0
i r
100
90
80
70
60
50
40
30
20
10
a>
p
i
I
c
i
P
£*.
o
5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
Figure 15.
Top: (Solid) 0.91 ppmC SYNAUTO, no MeOH, 0.01 ppm HCHO;
(Dashed) 1.32 ppmC SYNAUTO, no MeOH, 0.02 ppm HCHO;
Bottom: RED chamber air temperature (top solid line, T);
RED (solid line) and BLUE (dashed line) chamber dewpoint OF);
ambient total solar radaiation (solid line, cal-cm" -sec" )•
ambient ultraviolet radiation (dashed linejncal-cm" -sec" ).
66
-------
August 6, 1984
SynAuto 3.2 vs 2.3 ppmC (no ECHO or MeOH)
RESULTS: 03 MAX: BLUE 0.9398 PPM(1408); RED 0.8874 PPM(1420).
INITIAL CONDITIONS: BLUE RED
HO 0.283 0.279
H02 0.072 0.072
MMHC • 3.293 2.289
SYN-AUTO(TANK&LIQUID) 3.236 2.250
HCHO 0.057 0.039
ETHYLEHE 0.963 0.666
PROPYLEME 0.149 0.104
1-BUTE1IE 0.085 0.059
TRANS-2-BUTENE 0.079 0.052
I!-BUTANE 0.136- 0.095
ISOPENTANE 0.186 0.133
2.2,4-TRIMETHYLPENTANE 0.336 0.225
BENZENE 0.122 0.088
TOLUENE 0.602 0.422
M-XYLENE 0.301 0.212
0-XYLENE 0.115 0.081
1,2,4-TRIMETHYLBEiiZENE 0.162 0.115
-------
I
w
o>
TJ
O n* -
g> 0.4 -
8
2.0
I
1.0
0.5
0.0
I i I i I i" I 7 I i "1
9 10 11 12 13 14 15 16 17 18 19*"
HOURS, EOT
5 6 7 8 9 10 11 12 13 14 15 16 17 18
HOURS, EOT
19
100
90
80
70
60
50
40
30
20
10
0
5"
p
f
C
Figure 16.
Top: (Solid) 2.25 ppmC SYNAUTO, no MeOH, 0.04 ppm HCHO;
(Dashed) 3.24 ppmC SYNAUTO, no MeOH, 0.06 ppm HCHO;
Bottom: RED chamber air temperature (top solid line, °F);
RED (solid line) and BLUE (dashed line) chamber dewpoint (°F);
ambient total solar radaiation (solid line, cal-cm~ -sec" \
ambient ultraviolet radiation (dashed line^ncal-cm" -sec" ).
68
-------
August 7, 1984
SynAuto 1.36 ppmC vs 0.89 ppmC/0.3 ppm MeOH/0.029 ppm HCHO
RESULTS: 03 MAX: BLUE 0.4109 PPM(1749); RED 0.6012 PPM(1737) .
INITIAL CONDITIONS: BLUE RED
1!0 0.296 0.302
1102 0.084 0.083
UMHC • 1.220 1.361
SY!!-AUTO(TA!!KfcLiqUID) 0.865 1.320
HCHO 0.056 0.041
MEOH 0.299 0.000
ETHYLE1IE 0.251 0.381
PROPYLE11E 0.040 0.057
1-BUTEHE 0.020 0.033
TRANS-2-BUTE!.TE 0.021 0.030
M-BUTAliE 0.037 0.057
ISOPEIITA1IE 0.052 0.073
2.2.4-TRIMETHYLPEJiTAliE 0.095 0.138
BE1IZE1IE 0.041 0.063
TOLUEHE 0.155 0.249
M-XYLEIJE 0.083 0.128
0-XYLEKE 0.031 0.046
1.2,4-TRIMETHYLBEUZEllE 0.041 0.065
69
-------
I
Q.
OT
0>
T>
0)
I | I | I | I | I | I
August 7, 1984
O n*; -
g> 0.4 -
8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
2.0
w
at
o
11.0
(0
oc
'0.5
0.0
100
90
80
70
60
50
40
30
20
10
O
CD
c
8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
Figure 17.
Top: (Solid) 1.32 ppmC SYNAUTO, no MeOH, 0.04 HCHO, no MeNO2;
(Dashed) 0.87 ppmC SYNAUTO, 0.3 MeOH, 0.06 HCHO, 3 ppb MeNO2;
Bottom: RED chamber air temperature (top solid line, °F);
RED (solid line) and BLUE (dashed line) chamber dewpoint (°F);
ambient total solar radaiation (solid line, cal-cm* -sec" i
ambient ultraviolet radiation (dashed line^ncal-cm" -sec" ).
70
-------
August 8, 1984
SynAuto 3.7 ppmC vs 2.5 ppmC/0.79 ppm MeOH/0.19 ppm HCHO
RESULTS: 03 MAX: BLUE 0.8329 PPM(1432); RED 0.8364 PPM(1428) .
INITIAL CONDITIONS: BLUE RED
NO 0.264 0.262
1102 0.078 0.076
NMHC . 3.734 3.491
SYH-AUTO(TANKftLIQUID) 3.677 2.478
MEOH 0.000 0.785
HCHO 0.057 0.228
ETHYLEtJE 1.050 0.705
PROPYLENE 0.158 0.108
1-BUTENE 0.095 0.064
TRANS-2-BUTEME 0.089- 0.058
N-BUTANE 0.151 0.099
ISOPENTANE 0.200 0.139
2,2.4-TRIMETHYLPENTANE 0.389 0.265
BENZENE 0.166 0.113
TOLUENE 0.715 0.481
M-XYLENE 0.341 0.230
0-XYLENE 0.134 0.096
1,2,4-TRIMETHYLBENZENE 0.188 0.126
71
-------
I I I I I I I I I I I I I I I I I I I I I I I I I I I
Augusf38, 1984
6 7 8 9 10 11 12 13 14 15 16 17 18 19
HOURS. EOT
0.0
2.0
w
x
JD
0)1.5
a
o
1.0
'0.5
0.0
5 6 7 8 9 10 11 12 13 14 15 16 17 18
HOURS, EOT
19
100
90
80
70
60
50
40
30
20
10
0
o>
p
I
I
c
ff
a
Figure 18.
Top: (Solid) 2.48 ppmC SYNAUTO, 0.79 MeOH, 0.23 HCHO, 10 ppb MeNO2;
(Dashed) 3.68 ppmC SYNAUTO, no MeOH, 0.06 HCHO, no MeNO2;
Bottom: RED chamber air temperature (top solid line, °F);
RED (solid line) and BLUE (dashed line) chamber dewpoint (°F);
ambient total solar radaiation (solid line, cal-crrf -sec" ):
ambient ultraviolet radiation (dashed line,mcal-cm~ -sec" ).
72
-------
August 9, 1984
SynAuto 1.3 ppmC vs 0.87 ppmC/0.26 ppm MeOH/0.06 ppm HCHO
RESULTS: 03 MAX: BLUE 0.3213 PPM(1616); RED 0.5205 PPM(1612).
INITIAL CONDITIONS:
NO
N02
NMHC
SYN-AUTO(TANK&LIQUTD)
MEOH
HCHO
ETHYLEHE
PROPYLENE
1-BUTENE
TRAKS-2-BUTEHE
N-BUTANE
ISOPEHTANE
2.2.4-TRIMETHYLPENTANE
BENZENE
TOLUENE
M-XYLENE
0-XYLENE
1,2.4-TRIMETHYLBENZENE
BLUE
0.298
0.089
1.196
0.858
0.262
0.076
0.238
0.038
0.021
0.019
0.036
0.049
0.101
0.039
0.165
0.081
0.031
0.043
RED
0.302
0.087
1.303
1.284
0.000
0.019
0.345
0.056
0.032
0.029
0.052
0.070
0.146
0 . 059
0.256
0.122
0.049
0.067
73
-------
1.0
0.9
E 0.8
a
a 0.7
8
•
0)
0.6
0.5
0.4
0.3
0.2
0.1
0.0
III! I I T
I ]\
- NO
August 9, 1984
6 7 8 9 10 11 12 13 14 15 16 17 18
HOURS, EOT
19
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
2.0
o
1.0
i
0.5
0.0
5 6 7 8 9 10 11 12 13 14 15 16 17 18
HOURS, EOT
19
100
90
80
70
60
50
40
30
20
10
0
D
(D
I
C
i
I'
o'
•\
Figure 19.
Top: (Solid) 1.28 ppmC SYNAUTO, no MeOH, 0:02 ppm HCHO, no MeNO2;
(Dashed) 0.86 ppmC SYNAUTO, 0.26 MeOH, 0.08 HCHO, 3 ppb MeNO2;
Bottom: RED chamber air temperature (top solid line, °F);
RED (solid line) and BLUE (dashed line) chamber dewpoint (°F>,
ambient total solar radaiation (solid line, cal-cnf -sec" ):
ambient ultraviolet radiation (dashed linejncal-cm" -sec" ).
74
-------
August 22, 1984
SynUrban 3.1 ppmC vs
RESULTS: 03 MAX: BLUE
INITIAL CONDITIONS:
HO
N02
HMHC
SYN-URBAN (TANKtLIQUID)
MEOH
HCHO
ETHYLEKE
PROPYLEHE
1-BUTEME
TRANS-2-BUTENE
t! -BUTANE
ISOPENTANE
2-METHYL-l-BUTENE
2-METHYL-2-BUTE1JE
ti-PENTANE
2-METHYLPEHTAHE
2.4-DIMETHYLPEHTAHE
2 . 2 . 4-TRIMETHYLPEKTANE
BENZENE
TOLUENE
M-XYLENE
0-XYLENE
1 . 2 . 4-TRIMETHYLBENZEHE
2.1 ppmC/0.87 ppm MeOH/0.092 ppm HCHO
0.6345 PPMU624); RED
BLUE
0.247
0.068
3.039
2.042
0.867
0.130
0.148
0.056
0.033
0.034-
0.167
0.176
0.059
0.030
0.263
0.100
0.144
0.185
0.053
0.278
0.145
0.065
0.108
0.6572 PPM (1628).
RED
0.254
0.070
3.097
3.041
0.000
0.056
0.231
0.091
0.054
0.047
0.253
0.266
0.082
0.047
0.386
0.154
0.221
0.275
0.079
0.403
0.203
0.088
0.160
75
-------
I I I I I I I I I I I I
August 22, 1984
8
9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
2.0
_fl)
_l
I
i
'1.5
'0.5
0.0
1,1,1,1,1
5 6 7 8 9 10 11 12 13 14 15 16 17 18
HOURS, EOT
19
100
90
80
70
60
50
40
30
20
10
0
I
I
c
I
§
Figure 20.
Top: (Solid) 3.04 ppmC SYNURBAN, no MeOH, 0.06 ppm.HCHO, no MeNO2;
(Dashed) 2.04 ppmC SYNURBAN, 0.87 MeOH, 0.13 HCHO, 10 ppb MeN
Bottom: RED chamber air temperature (top solid line, T);
RED (solid line) and BLUE (dashed line) chamber dewpoint (°F);
,-K
ambient total solar radaiation (solid line, cal-cm" -sec"1);
ambient ultraviolet radiation (dashed linejucal-cm" -sec" ).
-------
August 25, 1984
SynUrban 1.09 ppmC vs 0.73 ppmC/0.29 ppm MeOH/0.03 ppm HCHO
RESULTS: 03 MAX: BLUE 0.0963 PPM(1744); RED 0.0744 PPM(1732).
INITIAL CONDITIONS: BLUE RED
NO 0.262 0.272
H02 0.068 0.076
IIMHC ' 1.089 1.058
SYN-URBAN(TANKftLIQUID) 1.070 0.720
MEOH 0.000 0.295
HCHO 0.019 0.043
ETHYLEME 0.129 0.098
PROPYLENE 0.040 0.031
1-BUTENE 0.023 0.016
N-BUTANE 0.027- 0.017
CIS-2-BUTENE 0.021 0.012
ISOPENTANE 0.095 0.066
II-PE11TA1IE 0.120 0.097
2.2,4-TRIMETHYLPENTAKE 0.117 0.086
BENZENE 0.030 0.025
TOLUENE 0.141 0.103
M-XYLENE 0.069 0.053
0-XYLENE 0.031 0.024
77
-------
I.U
0.9
E 0.8
a.
a 0.7
(/>
1 0.6
0 0.5
c
O) 0.4
£
z 0.3
/> A
0.2
0.1
r»n
_ i I i I i I i I i I i I i I i I i I i I i I i I i I i _
7 August 25, 1984 ~
— —
-
— —
-
— —
NO
t-TW^-r. »-~^_ NO9
„ ~* " ' " "-"•^Vy^^ IllVB/^ _
>v»^ ^^-^^.
_^^*^^
^ PI i i i i i i i_ i i i [^^^^^i^ir^ i i i i ~
i.u
0.9
0.8
0.7
o
0.6 g
D
0.5 .*
TJ
0.4 3
0.3
Of\
.2
0.1
nn
8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
2.0
8L
JS
o>1.5
(0
o
^1.0 -
at
CC
100
'0.5 -
0.0
8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
Figure 21.
Top: (Solid) 0.72 ppmC SYNURBAN, 0.29 ppm MeOH, 0.04 ppm HCHO;
(Dashed) 1.07 ppmC SYNURBAN, no MeOH;
Bottom: RED chamber air temperature (top solid line, °F);
RED (solid line) and BLUE (dashed line) chamber dewpoint (°F);
ambient total solar radaiation (solid line, cal-cm" -sec' ):
ambient ultraviolet radiation (dashed line^ncal-cm" -sec" ).
78
-------
August 28, 1984
SynUrban 1.18 ppmC vs 0.82 ppmC/0.33 ppm MeOH
RESULTS: 03 MAX: BLUE 0.0560 PPM(1456); RED 0.0307 PPM(1452).
INITIAL COMDITIOtJS: BLUE RED
NO 0.250 0.256
N02 0.062 0.068
IIMHC - 1.176 1.149
SYN-URBA1UTAMK&LIQUID) 1.157 0.809
MEOH 0.000 0.327
HCHO 0.019 0.013
ETHYLENE 0.142 0.123
PROPYLEME 0.031 0.020
1-BUTEHE 0.020 0.010
TRANS-2-BUTEHE 0.015- 0.013
H-BUTANE 0.118 0.088
ISOPENTAME 0.080 0.053
2-METHYL-1-BUTEHE 0.022 0.013
2-METHYL-2-BUTENE 0.016 0.009
N-PENTANE 0.137 0.086
2-METHYLPENTANE 0.053 0.031
2,4-DIMETHYLPEHTA!IE 0.076 0.045
2.2,4-TRIMETHYLPENTANE 0.109 0.074
BENZENE 0.022 0.022
TOLUENE 0.158 0.094
M-XYLEUE 0.145 0.052
0-XYLENE 0.035 0.030
1.2.4-TRIMETHYLBEHZENE 0.048 0.045
RUN NOT PROCESSED
79
-------
I.U
0.9
E 0.8
a
a 0.7
(/>
t\\
2 0.6
X
0 0.5
c
0>
O) 0.4
i 0.3
00
.41
0.1
nn
_ i I i I i I i I i I i I . I i I i I i I i I i I i I i _
7 August 28, 1984 ~
— ' —
-
— —
-
— —
—
— —
-
— —
I NO -
- - ~. >*~WVlr>_^^^ NU2
- ^^^^ 03 :
i i i i i i i i i i i i i i L^tiiiS^ i^-'i i i i i i i i ~
I.U
0.9
0.8
0.7
o
0.6 j?
3
0.5 «a
•a
0.4 1
0.3
00
.c.
0.1
nn
8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
2.0
w
X
o
n.5
o
11.0
i
o
CO
0.5
0.0
100
90
80
70
60
50
40
30
20
10
CO
p
<*
O
0)
I
c
5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
Figure 22.
Top: (Solid) 0.81 ppmC SYNURBAN, 0.32 ppm MeOH, 0.01 ppm HCHO;
(Dashed) 1.16 ppmC SYNURBAN, no MeOH, 0.02 ppm HCHO;
Bottom: RED chamber air temperature (top solid line, °F);
RED (solid line) and BLUE (dashed line) chamber dewpoint (°F);
ambient total solar radaiation (solid line, cal-cnf -$ec~ y.
ambient ultraviolet radiation (dashed line,mcal-cm~ -sec ).
80
-------
September 1, 1984
SynUrban 3.37 ppmC vs 2.70 ppmC/.97 ppm MeOH
RESULTS: 03 MAX: BLUE 0.6457 PPM(1608) ; RED 0.5459 PPMU708) .
INITIAL CONDITIONS: BLUE RED
HO 0.243 0.252
1102 0.059 0.062
MMHC • 3.365 3.663
SYM-URBANCTANK4LIQUID) 3.310 2.660
MEOH 0.000 0.966
HCHO 0.055 0.037
ETHYLENE 0.200 0.164
PROPYLEHE 0.092 0.061
1-BUTENE 0.048 0.033
H-BUTANE 0.380- 0.678
CIS-2-BUTENE 0.042 0.028
ISOPENTAHE 0.242 0.167
N-PEHTANE 0.420 0.310
2,2.4-TRIMETHYLPEHTANE 0.289 0.189
BENZENE 0.076 0.049
TOLUENE 0.408 0.275
M-XYLENE 0.204 0.152
0-XYLEHE 0.087 0.076
81
-------
a
a
(A
0)
T3
'X
o
p
D
CD
c
i
8.'
6'
Figure 23.
Top: (Solid) 2.66 ppmC SYNURBAN, 0.97 ppm MeOH;
(Dashed) 3.31 ppmC SYNURBAN, no MeOH;
Bottom: RED chamber air temperature (top solid line, °F);
RED (solid line) and BLUE (dashed line) chamber dewpoint (°F);
ambient total solar radaiation (solid line, cal-cm -sec" ):
ambient ultraviolet radiation (dashed line,mcal:cm~ -sec" ).
82
-------
September 2, 1984
SynUrban 1.12 ppmC vs 0.77 ppmC
RESULTS: 03 MAX: BLUE 0.0199 PPH(1656) ; RED 0.1193 PPM(1732)
INITIAL CONDITIONS: BLUE RED
NO 0.259 0.284
1J02 0.063 0.060
NMHC . 0.768 1.118
SYM-URBA1! (TANKfcLIQUID) 0.756 1.099
HCHO 0.012 0.019
ETHYLEME 0.056 0.089
PROPYLEHE 0.028 0.036
1-BUTENE 0.014 0.019
TRANS-2-BUTENE 0.010 0.016
N-BUTANE 0.10Q 0.130
ISOPEHTANE . 0.054 0.079
2-METHYL-1-BUTE1IE 0.015 0.019
2-METHYL-2-BUTEHE 0.011 0.014
N-PENTANE 0.090 0.090
2-METHYLPEUTAHE 0.036 0.046
2.4-DIMETHYLPENTAt!E 0.051 0.066
2.2.4-TRIMETHYLPENTAIIE 0.096 0.157
BENZENE 0.020 0.028
TOLUENE 0.083 0.130
M-XYLENE 0.044 0.063
0-XYLENE 0.028 0.033
1.2.4-TRIMETHYLBENZEHE 0.000 0.000
83
-------
a
a
CO
0)
•a
0)
O)
£
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
NO
September 2, 1984.
NO2
8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
1.0
0.9
0.8
0.7
0.6 g
0.5 -®
TJ
0.4 1
0.3
0.2
0.1
0.0
2.0
in
0)1.5
(0
1.0
'0.5
0.0
I
I I i I I
5 6 7 8 9 10 11 12 13 14 15 16 17 18
HOURS, EOT
19
100
90
80
70
60
50
40
30
20
10
0
a>
p
i
c.
i
Figure 24.
Top: (Solid) 1.10 ppmC SYNURBAN, no MeOH, 0.02 ppm HCHO;
(Dashed) 0.76 ppmC SYNURBAN, no MeOH, 0.01 ppm HCHO;
Bottom: RED chamber air temperature (top solid line, °F);
RED (solid line) and BLUE (dashed line) chambej dewpoint (°F>,
ambient total solar radaiation (solid line, cal-cm" -sec" )•
ambient ultraviolet radiation (dashed line,mcal-cm -sec ).
84
-------
September 3, 1984
SynUrban 1.03 ppmC vs 0.79 ppmC/0.26 ppm MeOH/0.06 ppm HCHO
RESULTS: 03 MAX: BLUE 0.0578 PPM(1728); RED 0.0479 PPM(1724).
IIIITIAL CONDITIONS: BLUE RED
HO 0.240 0.241
1J02 0.112 0.111
1JMHC - 1.116 1.033
SYN-URBAM(TANK&LIQUID) 0.778 1.014
HEOH 0.264 0.000
HCHO 0.074 0.019
ETHYLENE 0.067 0.089
PROPYLEME 0.021 0.024
1-BUTENE 0.015 0.016
TRAHS-2-BUTEME 0.012 0.014
1!-BUTANE 0.082 0.118
ISOPEHTANE 0.051 0.078
2-METHYL-l-BUTEHE i 0.016 0.018
2-METHYL-2-BUTENE 0.011 0.013
!!-PE!!TA!lE 0.095 0.095
2-KETHYLPENTA1IE 0.037 0.043
2.4-DIMETHYLPEl!TAi!E 0.053 0.062
2.2.4-TRIMETHYLPEIJTA1JE 0.083 0.114
BEl.'ZEHE 0.027 0.031
TOLUENE 0.098 0.136
M-XYLENE 0.047 0.065
0-XYLEHE 0.027 0.038
1,2.4-TRIMETHYLBEtIZENE 0.035 0.058
85
-------
1.0
0.9
E 0.8
a
a 0.7
w
;§ 0.6
'x
O 0.5
04
o
i 0.3
0.2
0.1
September 3, 1984
N02
8
10 11 12 13 14
HOURS, EOT
15
17 18 19
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
2.0
(0
g
1.0
sl
'0.5
0.0
i i r
100
90
80
70
60
50
40
30
20
10
O
(D
I
e
5678
9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
Figure 25.
Top: (Solid) 1.01 ppmC SYNURBAN, no MeOH, 0.02 ppm HCHO;
(Dashed) 0.78 ppmC SYNURBAN, 0.26 ppm MeOH, 0.07 ppm HCHO;
Bottom: RED chamber air temperature (top solid line, °F);
RED (solid line) and BLUE (dashed line) chamber dewpoint (°F);
ambient total solar radaiation (solid line, cal-cm" -sec" ):
ambient ultraviolet radiation (dashed line^ncal-cm" -sec" ).
86
-------
September 8, 1984
SynAuto 2.8 vs 1.8 ppmC (no MeOH)
RESULTS: 03 MAX: BLUE 0.7472 PPM(1456); RED 0.5659 PPM(1708)
INITIAL CONDITIONS:
NO
1102
HMHC
SYti-AUTO (TAllKtLiqUID)
HCHO
ETHYLEHE
PROPYLEME
1-BUTEHE
N-BUTANE
CIS-2-BUTE11E
ISOPENTAUE
K-PE1ITA1IE
2.2,4-TRIMETHYLPEHTAKE
BENZENE
TOLUENE
H-XYLEHE
0-XYLENE
1.2,4-TRIHETHYLBEI!ZE!IE
BLUE
0.205
0.126
2.840
2.790
0.050
0.979
00702 0.134
00702 0.075
00702 0.446
00702 0.053-
00702 0.215
00702 0.128
80702 0.218
00702 0.047
00702 0.088
00702 0.044
00702 0.079
00702 0.053
RED
0.204
0.121
1.880
1.840
0.040
0.641
00732 0.088
00732 0.044
00732 0.321
00732 0.053
00732 0.122
00732 0.000
00832 0.126
00832 0.036
00832 0.229
00832 0.125
00832 0.062
00832 0.096
87
-------
a.
a
w
'x
O
0)
O)
2
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
NO
1 i ' I ' I ' I ' I ' I '
September 8, 1984
O3
I i I i I
8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
2.0
_
0)1.5
at
O
1.0
o
en
0.5
0.0
5 6 7 8 9 10 11 12 13 14 15 16 17 18
HOURS, EOT
19
100
90
80
70
60
50
40
30
20
10
0
n>
3
p
o
CD
I
c
i
Figure 26.
Top: (Solid) 1.8 ppmC SYNAUTO, 0.2 ppmC Butane, no MeOH;
(Dashed) 2.7 ppmC SYNAUTO, 0.3 ppmC Butane, no MeOH;
Bottom: RED chamber air temperature (top solid line, °F);
RED (solid line) and BLUE (dashed line) chamber dewpoint (°F);
ambient total solar radaiation (solid line, cal-cm" -sec" )•
ambient ultraviolet radiation (dashed linejjical-cm" -sec" ).
88
-------
September 9, 1984
SynAuto 2.1 vs 1.4 ppmC (no MeOH)
RESULTS: 03 MAX: BLUE 0.6062 PPM(1600); RED 0.4203 PPM(1636)
I1IITIAL COt.'DITIONS: BLUE RED
NO 0.285 0.294
N02 0.055 0.060
HMHC ' 2.180 1.440
SY1J-AUTO(TAKK4LIQUID) 2.140 1.420
HCHO 0.040 0.020
ETHYLEtlE 0.554 0.392
PROPYLEHE 0.110 0.075
1-BUTEHE 0.052 0.030
M-BUTAUE 0.094 0.066
TRANS-2-BUTEHE 0.054 0.037
ISOPENTA1IE 0.135 0.103
2,2,4-TRIMETHYLPEHTAlIE 0.187 0.133
BEUZENE 0.056 0.038
TOLUENE 0.373 0.284
M-XYLEUE 0.188 0.145
0-XYLEHE 0.084 0.057
1.2.4-TRIHETHYLBENZEIIE 0.093 0.060
89
-------
Q.
Q.
OT
0)
T>
-------
September 17, 1984
SynAuto 2.18 vs 1.45 ppmC/0.57 ppm MeOH
RESULTS: 03 MAX: BLUE 0.4836 PPM(1632); RED 0.5389 PPH(1524).
INITIAL CONDITION'S: BLUE RED
NO 0.269 0.270
N02 0.067 0.070
MHHC • 2.070 2.180
SYN-AUTO(TAlIKiLIQUID) 1.420 2.140
MEOH 0.570 0.000
HCHO 0.080 0.040
ETHYLENE 0.767 1.142
PROPYLEME 80730 0.164 0.219
1-BUTENE 80730 0.076 0.112
M-BUTANE (80730 0.114 0.213
TRANS-2-BUTEHE 80730 0.073 0.122
ISOPENTANE 0.167 0.283
2.2,4-TRI!.!ETHYLPE!lTAl!E 0.302 0.429
BENZENE 0.102 0.124
TOLUENE 0.577 0.845
M-XYLENE 0.302 0.435
0-XYLEt'E 0.177 0.169
1.2,4-TRIMETHYLBE!!ZE!!E 0.000 0.000
91
-------
1.0
0.9
E 0.8
a
°- 0.7
w
3 0.6
x
° 0.5
0.4
i 0.3
0.2
0.1
0.0
1 I ' I ' I ' I ' I ' I ' I ' I ' I ' I ' I ' I ' I '
September 17, 198-
O3
I i I i I
6 7 8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
2.0
o
1.0
*
o
'0.5
0.0
i i i r
100
90
80
70
60
50
40
30
20
10
O
(D
I
s±
O
5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
Figure 28.
Top: (Solid) 2.14 ppmC SYNAUTO, 0.2 ppmC Butane, no MeOH, 0.04 HCHO;
(Dashed) 1.42 ppmC SYNAUTO, 0.13 Butane, 0.57 MeOH, 0.08 HCHO;
Bottom: RED chamber air temperature (top solid line, °F);
RED (solid line) and BLUE (dashed line) chamber dewpoint (°F);
ambient total solar radaiation (solid line, cal-cm" -sec" ):
ambient ultraviolet radiation (dashed lineoncal-cm" -sec" ).
92
-------
September 19, 1984
SynAuto HMW 2.6 vs 4A ppmC
RESULTS: 03 MAX: BLUE 0.4386 PPH(1512); RED 0
INITIAL CONDITIONS: BLUE
110 0.252
1J02 0.086
2.2,4-TRIMETHYLPENTANE 0.520
BEHZEME - 0.147
TOLUEME 1.011
M-XYLEHE 0.489
0-XYLENE 0.190
1,2,4-TRIMETHYLBENZEHE 0.269
MMHC 2.633
MORN
SUNLIGHT AND WEATHER: 9.8/1.0
TEMPERATURE F: 45
EXPERIMENT STARTS:
INITIAL CONDITIONS ESTABLISHED:
EXPERIMENTS ENDS:
.3828 PPM(1204).
RED
0.253
0.086
0.841
0.229
1.645
0.825
0.337
0.550
4.426
AFTERNOON
9.8/10
83
0700 (SUNRISE)
0600
1800
93
-------
1.0
0.9
E 0.8
a
a 0.7
w
I 0.6
° 0.5
c
0)
g> 0.4
i 0.3
0.2
0.1
0.0
I ' I ' I ' I ' I ' I ' I ' ! ' I ' I ' I ' I ' I '
September 19, 198"
- NO
8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
1.0
0.9
0.8
0.7
0.6 |
0.5 -®
TJ
0.4 ^
0.3
0.2
0.1
0.0
2.0
(A
OJ1.5
as
O
1.0
a
&
0.5
0.0
100
90 H
-------
September 21, 1984
SynAutUrb 2.43 ppmC vs 2.42 ppmC/0.18 ppm ECHO
RESULTS: 03 MAX: BLUE 0.7212 PPM(1448) ; RED 0.6706 PP1!(1556).
INITIAL COtlDITIOHS:
NO
N02
MMHC
SYN-AUTO(TANK&LIQUID)
HCHO
ETHYLE1IE
PROPYLEME
1-BUTEME
II-BUTAME
TRAl.'S-2-BUTEIIE
ISOPEHTAtJE
2,2.4-TRIMETHYLPEHTA1JE
BE!!ZE!1E
TOLUEUE
M-XYLEliE
0-XYLE1IE
1.2,4-TRIl.!ETHYLBE!!ZEHE
BLUE
0.273
0.086
2.600
2.420
0.180
0.596
0.119
0.062
0.105
0.066-
0.145
0.233
0.082
0.460
0.259
0.118
0.177
RED
0.275
0.088
2.430
2.430
0.000
00715 0.602
80715 0.119
80715 0.062
80715 0.101
80715 0.071
80715 0.151
80715 0.228
80715 0.081
80715 0.465
80715 0.266
80715 0.103
80715 0.182
95
-------
September 21, 198
6 7 8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
0.0
2.0
1.0
'0.5
0.0
100
90
80
70
60
50
40
30
20
10
a>
p
c
Q.
S'
r*
5'
5678
9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
Figure 30.
Top: (Solid) 2.43 ppmC SYNAUTO, no MeOH, no HCHO;
(Dashed) 2.42 ppmC SYNAUTO. no MeOH, 0.18 ppm HCHO;
Bottom: RED chamber air temperature (top solid line, °F);
RED (solid line) and BLUE (dashed line) chamber dewpoint (°F);
ambient total solar radaiation (solid line, cal-cm -sec" ):
ambient ultraviolet radiation (dashed linejncal-cm" -sec~ ).
-------
June 26, 1985
SynUrban
RESULTS: 03 MAX: BLUE 0.79
IHITIAL CONDITIONS:
NO
1102
HHHC
SYN-URBAH (TANKfcLIQUID)
MEOH
HCHO
ETHYLEl.'E
PROPYLEIIE
1-BUTEIIE
TRANS-2-BUTENE
tJ -BUTANE
ISOPENTANE
2-HETHYL-l-BUTEUE
2-METHYL-2-BUTEIIE
I.'-PENTANE
2-METHYLPENTANE
2,4-DIMETHYLPENTANE
2.2. 4-TRIMETHYLPEHTANE
BENZENE
TOLUENE
M-XYLEHE
0-XYLENE
1 . 2 . 4-TRIMETHYLBENZEHE
4.07 ppmC vs
PPM (1505); RED 0.
BLUE
0.266
0.034
4.072
4.012
0.000
0.060
0.220
0.093
0.050
0.047.
0.278
0.285
0.178
0.127
0.445
0.423
0.605
0.385
0.107
0.384
0.179
0.077
0.130
2.48 ppmC
63 PPM (1700).
RED
0.264
0.035
2.483
2.443
0.000
0.040
0.157
0.062
0.036
0.036
0.201
0.196
0.063
0.045
0.324
0.149
0.213
0.269
0.079
0.293
0.140
0.061
0.119
97
-------
I I I I I I I I I I I I I I I I I
June 26, 1985
O3
8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
0.0
2.0
CO
«J
1.0
"
0.5
0.0
8 9 10 11 12 13 14
HOURS, EOT
15 16 17 18 19
100
90
80
70
60
50
40
30
20
10
0
o>
p
O
CD
I
c
i
fa
f+
O
Figure 31.
Top: (Solid) 4.01 ppmC SYNURBAN;
(Dashed) 2.44 ppmC SYNURBAN;
Bottom: RED chamber air temperature (top solid line, °F);
RED (solid line) and BLUE (dashed line) chamber dewpoint (°F);
ambient total solar radaiation (solid line, cal-cm -sec ):
ambient ultraviolet radiation (dashed line,mcal-cm -sec" ).
98
-------
June 28, 1985
SynUrban 2.69 ppmC
RESULTS: 03 MAX: BLUE
INITIAL CONDITIONS:
NO
1102
H1-1HC
SYN-URBAN (TANK&LIQUID)
HEOH
HCHO
ETHYLEME
PROPYLENE
1-BUTENE
N-BUTAME
CIS-2-BUTENE
ISOPENTANE
2-METHYL-l-BUTEHE
2-HETHYL-2-BUTENE
N-PENTANE
2-METHYLPENTANE
2 , 4-DIMETHYLPEHTANE
2 , 2 . 4-TRIMETHYLPENTAHE
BENZENE
TOLUENE
M-XYLEHE
0-XYLEHE
1.2, 4-TRIMETHYLBENZENE
vs 1.81 ppmC/0.58
0.27 PPMU655); RED 0.24
BLUE
0.275
0.112
2.452
1.779
0.580
0.093
0.090
0.031
0.018
0.090-
0.017
0.113
0.071
0.051
0.176
0.168
0.240
0.283
0.046
0.178
0.085
0.037
0.087
ppm MeOH/0.07 HCHO
PPM (1655).
RED
0.273
0.107
2 . 690
2.650
0.000
0.040
0.144
0.056
0.030
0.409
0.036
0.184
0.058
0.042
0.230
0.138
0.197
0.425
0.073
0.270
0.127
0.054
0.108
99
-------
1.0
0.9
E 0.8
a
a 0.7
O)
3 0.6
| 0.5
0)
g>0.4
2 0.3
0.2
0.1
0.0
1 I ' I ' I ' I ' I ' I ' I ' I ' I ' I ' I ' I ' I
June 28, 1985
6 7 8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
2.0
w
0)1.5
a
.§
1
'0.5
0.0
I
5 6 7 8 9 10 11 12 13 14 15 16 17 18
HOURS, EOT
19
100
90
80
70
60
50
40
30
20
10
0
a>
I
C
ff
a
p»'
Figure 32.
Top: (Solid) 2.65 ppmC SYNURBAN 0.04 ppm HCHO;
(Dashed) 1.78 ppmC SYNURBAN, 0.58 ppm MeOH, 0.09 ppm HCHO;
Bottom: RED chamber air temperature (top solid line, °F);
RED (solid line) and BLUE (dashed line) chamber dewpoint (°F);
ambient total solar radaiation (solid line, cal-cm~ -sec" ):
ambient ultraviolet radiation (dashed line,mcal-cm -sec ).
100
-------
Discussion
This chapter will discuss the results of the experiments from two different view-
points. In the first part, plots of the primary .experimental data will be compared
for different conditions to illustrate the effects of MeOH substitution. In the second
part, three photochemical mechanism models will be used to simulate selected ex-
periments. This will illustrate the usefulness of the data in testing some current
mechanisms that might be used to predict the effects of MeOH substitution.
Meaning of Reactivity
The term reactivity has been used to mean many things relative to smog cham-
ber work. Reactivity scales were created by researchers attempting to quantify
smog chamber results. These scales have essentially fallen from use because they
over-simplify the understanding of complex smog chamber results and are therefore
misleading. The simple reactivity concepts have been replaced by the photochemi-
cal kinetics mechanism models that have the potential to represent all the aspects
of the situation that impact on "reactivity." Kinetics models, however, are complex
representations that have to be manipulated by computers and many people prefer
the simpler description provided by "this system is more reactive than that system."
The application of the term "reactivity" to experimental results, however, im-
plies that there is a "scale" that can be used to judge experimental outcomes and
that particular experiments can be assigned a location on this scale and thus vari-
ous experiments can be compared. The units that have been used to calibrate this
scale, however, are varied: e.g.
101
-------
Discussion Meaning of Reactivity
o maximum 03 produced
o rate of NO oxidation
o time to NO-to-N02 crossover
o rate of HC consumption
c HO rate constant for the HC species
o rate of O3 production
o AOs/AHC is an important measure for control calculations
A big problem with these scales is that they are not linear, not absolute, and not
monotonic (i.e. system A may be simultaneously rated lower than system B on one
scale and rated higher than system B on another scale), that is, they lack in all the
things people readily understand. It is easy to forget these limitations, however,
and in this mapping of results on convenient scales, the reader is cautioned not to
confuse the simple representations with the complex realities of the chamber results.
In the present situation, relative comparisons of the plots of the NO, NO2, and
Os are certainly preferred to a table giving maximum 03. Even this comparison of
plots is dangerous because the weather conditions were often significantly different
for different runs, and so conclusions drawn by comparing runs on different days
must be checked by examining the TSR and temperature profiles for the different
days to see if they are similar. Visual comparison of TSR data can also be misleading.
The day may appear clear, but a careful overlay comparison of the plots can, for
example, show that a thin overcast might have attenuated the light intensity for the
whole day. Other atmospheric factors can change the UV-to-TSR ratio. Likewise,
there can be day-to-day variations in the injections and in the compositions of the
mixtures and these too must be examined. Models of these days would, of course,
use individual detail data for each day and would, therefore, take these important
factors into account.
We have selected days for comparison and modeling. We have examined the
conditions and data for these days in detail and are comfortable making day-to-day
direct comparisons of the experimental data using these selected days. In describing
the outcomes in these experiments, we sometimes say that a certain system is "more
reactive" or "less reactive" than another system and in doing so we have made a
subjective judgment. In this process we intuitively applied weighting factors to the
various aspects of the experiments to produce a single scale. An example of such
a judgment is saying that the substitution side (RED) of July 26, 1984 is "less
reactive" than the baseline side (BLUE) of July 26 because, although both sides
eventually made the same amount of Os, the baseline side made it significantly
102
-------
SynAuto Experiments Discussion
faster, and the whole baseline experiment was faster as shown by a shorter time
to reach NO-to-NO* crossover and NOj maximum. In many cases, relative timing to
events is as important as the magnitudes of the secondary products produced in the
chamber. In the outdoor chamber, something that delays the progress of a system
can result in later events occurring under decreasing light intensity and thus can
affect the magnitude of the later process. That is, systems can be "light-limited"
as well as reactant limited.
Experimental Findings
SynAuto Experiments
The basic experimental plan called for 3 ppmC, mid-ratio (e.g. 9:1 HC-to-NOx ratio)
experiments to be performed first, and to investigate the effect of the HCHO fraction
in the methanol exhaust. These HCHO fractions were called low for 0% HCHO, normal
for 10% HCHO, and high for 20% HCHO. Note that these are percentages of the
methanol exhaust component, not of the total NMHC. Also recall that the SynAuto
mixture itself contains 2% HCHO, so that even in the HC reduction experiments,
HCHO was present initially on both sides of the chamber.
After satisfactory experimental results were obtained at the 3 ppmC level, the
experimental plan called for the 1 ppmC, low-ratio, e.g. 3:1 HC-to-NOx ratio exper-
iments to be performed with the SynAuto mixture. The completion of the full set of
combinations of HCHO fractions at this level, however, was not possible because of
time restrictions.
SynAuto Experiments at 3 ppmC
Figure 33 shows the NOX and 03 profiles for six 3-ppmC experiments. The top row
of plots are for solar radiation conditions that would be somewhat difficult for most
models, while the bottom row of plots are for days that had excellent solar radiation
conditions.
On the left side of Figure 33 are the «33% reduction experiments, i.e. no
methanol-exhaust was added. This shows the effect of direct HC reduction at this
HC level and for this particular mixture. This system was NOx-limited. That is, the
system consumed all the NOX, converting it to HN03 and PAN. The downward sloping
NO2 line after 1400 LDT actually is 100% PAN. Because of the high thermal decom-
position rate for PAN, the PAN was in equilibrium with a very low concentration of
NO2. The high 03 concentration was a good source of HO radicals (from photolysis)
and thus even the small amount of NOj produced from the PAN decomposition was
103
-------
u>
OS
i"
B 07
I OB
So*
I 0.4
2 03
02
0.1
OO
i ' I • I • i ' i ' i • I • I ' i ' i ' i i
Ol
Reduction August. 2,..1.9 84
NO
678
9 10 11 12 13 14 15 16 17 18 19
HOURS. EOT
_ Substitution
0% HCHO
0.0
July 26, 1984
O)
8 9 10 11 12 13 14 IS 16 17 IB 19
HOURS. EOT
8 9 10 11 12 13 14 15 16 17 18 19
0X1
HOURS. EOT
10 11 12 13 14
HOURS. EOT
I ' I • I ' I ' I I I I I I I '
Substitution July 25 1984
7 10% HCHO ftl
- NO
8 9
10 11 12 13 14 15
HOURS. EOT
16 17 18 19
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
1.0
O9
0.8
0.7
Of,
0.5
O4
03
0.2
0.1
0.0
~ Substitution
- 20% HCHO
_NO
. 1984
9 10
H O
11 12 13
U R S. ED
14 15 16 17 18
T
1X1
O9
as
0.7
OS
05
O4
03
02
Ol
00
19
Figure 33.
Comparison of high ralio SynAuio experiments.
Top row: 1'oorei Miiiii»i)i condii'iona.
Botiorn row: Good sunli»lii comTn ions;.
-------
SynAuto Experiments Discussion
converted to HNO3 by
HO -t- NO2 = HN03
Since we do not measure HMOs, it appeared that nitrogen was being lost.
In an NOx-limited system, the maximum O3 is essentially determined by the
availability of NOX and not by the HC in the system. Picture the initial location as
being on the right and below the 03 maximum ridge line on an O3 isopleth diagram.
In this area. 03 maximum concentrations are not very dependent upon HC. The time
to reach O3 maximum, however, is dependent upon the reactivity and concentration
of the HC.
The side with less HC was about one hour slower in O3 production, but produced
almost the same 03 eventually («6% less). This is in contrast to the University of
Santa Clara results at this ratio (see Figure 1), which showed about the same O3
production as our experiments at the "basecase." but showed a greater effect for
HC-reduction, giving about 25% reduction in 03 for a 33% reduction in HC. This
may be in part due to differences in the basic compositions of the two mixtures.
Also note that there was a more rapid loss of NOX in the slightly slower runs in the
USC chamber that lead to less NOX at the end of the run (see the difference in NO;
maxima in Figure 1 compared to those in Figure 33). A system that removes NOX
more rapidly will generally make less 03.
Without modeling both the USC and the UNC chamber with the same kinetics
mechanism model and then comparing the two mixtures under the same set of
photolytic conditions, it is not possible to determine how consistent the two datasets
might be.
The top middle plot in Figure 33 (July 26) shows the effect of adding ssl.O ppm
MeOH. with no additional HCHO—the low formaldehyde condition. Although the
sun was not totally clear on this day, the effect adding only MeOH was almost not
detectable when compared to the effect of merely reducing the HC, as shown in the
two reduction runs on the left of the figure. That is, MeOH by itself is not very
reactive under these conditions (less than a 6% effect on the O3 maximum and no
effect on the timing of events).
The bottom middle plot in Figure 33 (July 25) shows the effect of adding syn-
thetic methanol exhaust with the expected normal amount of HCHO, 10%. Com-
pared to the "baseline" case, there was no difference in the initial timing of events
in the two halves, the delay in the formation of O3 was decreased, and the max-
imum amount of 03 produced was identical on the two sides. Although the sun
105
-------
Discussion SynUrban Experiments
was not as good on the August 3 experiment, it confirms the July 25 experimental
results. Even 10% HCHO had a significant effect on the relative reactivity of the
system. That is. the two sides on the July 25 experiment with HCHO included in the
substitution were much more similar than the two sides on the July 26 experiments
without HCHO included in the substitution.
The August 8 experiment in Figure 33 shows the effect of adding SynMethanoi
with 20% HCHO. There was only a small delay in O3 formation in the side with
MeOH/HCHO. Relative to their baseline sides, the 20% HCHO case was only slightly
more reactive than the 10% HCHO case and both of these cases were much more
reactive than was the MeOH-only case. That is there was less difference between the
two chamber sides for the experiment with 20% HCHO substituted (Aug. 8) than
for the experiment with 10% HCHO substituted (July 25) but not nearly as much
difference as between the experiments with 10% HCHO substituted (July 25) and the
experiment with 0% HCHO substituted (Aug. 6).
Although the SynAuto mix is quite reactive at the 3 ppmC (9:1 HC-to-NOx ratio)
level, the effect of 10% or 20% HCHO in the SynMethanoi mixture produces a system
as equally reactive as the original HC it replaced. This is different from the results
obtained by USC and shown in Figure 1. It should be recalled that the USC study
did not use HCHO but substituted iso-butylene.
SynAuto Experiments at 1 ppmC
The August 5, 1984 experiment was the baseline reduction experiment for this series.
In this experiment, reducing the initial HC from 1.31 ppmC to 0.91 ppmC, a 31%
reduction, reduced the maximum O3 from 0.60 ppm to 0.34 ppm, a 43% reduction.
The August 7. 1984 experiment was the SynMethanoi (normal HCHO) substitution
experiment. Figure 34 shows both the August 5 and the August 7 experimental
plots "overlaid" on two plots. The bottom plot of Figure 34 compares the two
"reference" sides of the experiments, which were well matched. The top plot shows
the effect of the added SynMethanoi4-10% HCHO. Instead of a 43% reduction in 03,
there was only a 32% reduction in 03. The total effect is better illustrated in
Figure 35 in which both sides of the August 5 and the substituted side of August 7
are shown in one plot.
SynUrban Experiments
Fewer SynUrban experiments were performed, because they were assigned a lower
priority than the direct comparison of gasoline exhaust and methanol exhaust (e.g.
106
-------
1.0
0.9
E 0.8
a
a 0.7
S 0.6
<•> 0.5
a>
§> 0.4
i 0.3
0.2
0.1
0.0
1 I ' I ' I ' I ' I ' l ' I ' I ' I ' I ' I ' I ' I '
August 5, 1984
Test Sides August 7, 1984 .
J 1 I
MeOH/10% HCHO -
8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
1.0
0.9
0.8
0.7
0.6 I
0.5 «ffl
T3
0.4 3
0.3
0.2
0.1
0.0
O
N
O
-------
1 I ' I ' I ' I ' I ' i '
August 5, 1984
August 7, 1984
:.;eoH/io% HCHO
8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
0.0
Figure 35.
Comparison of low high ratio SynAuto experiments.
Effect of MeOH 1017,HCHO substitution vs. reduction.
108
-------
Synllrban Experiments Discussion
the SynAuto/SynMethanol experiments) and the poor weather conditions in July forced
the SynUrban experiments into the last month of the project. In addition, the high
concentration experiments consumed the UNCMIX tank in early September, termi-
nating the SynUrban experiments that depended upon the UNCMIX mixture. In 1985,
three SynUrban experiments were performed with a new UNCMIX tank. Two of these
experiments are included in this report.
SynUrban Experiments at 3 ppmC
Figure 36 shows the three basic conditions for the 3 ppmC (9:1 HC-to-NOx ratio)
SynUrban experiments. The June 26, 1985 HC reduction experiment occurred under
higher solar radiation than the August and September experiments and therefore
produced more ozone. The relative outcome in the June 26 experiment is in good
agreement with that in the September 1 experiment, which only had MeOH substi-
tuted and had no additional HCHO other than the 2% HCHO included in the SynUrban
mix. The September 1 and August 22 experiments were in excellent agreement as
shown by the "overlay" plot in Figure 37.
As would be expected, at the same total HC, the SynUrban mix was less reac-
tive than the SynAuto mix because the SynUrban mix design took into account all
urban sources, many of which are much less reactive than automobile exhaust. Fur-
thermore, as shown in the June 26 run. there was a much larger AOs/AHC for
the SynUrban mix than for the SynAuto mix, another reflection of its lower overall
reactivity.
As in the SynAuto case, substitution of 100% MeOH for 1/3 of the carbon, resulted
in essentially the same outcome as a reduction of 33% of the SynUrban carbon.
Substitution of the most reactive version of the SynMethanol mix (20% HCHO/80%
MeOH) for 1/3 of the SynUrban mix, however, results in essentially no reduction of
ozone production. In the top plot of Figure 37. the difference between the two Os
lines was the difference between 0.14 ppm of total HCHO and 0.04 ppm of total HCHO
in a mixture of about 3 ppmC! In terms of the methanol exhaust fraction, it is the
difference between about 15% HCHO emissions and essentially no HCHO emissions
(some of the HCHO in the experiment was part of the SynUrban mixture). This too is
in agreement with the results of the SynAuto mixture.
These results suggest that HCHO is the dominant factor in the reactivity of the
MeOH-exhaust system in both a reactive background such as the SynAuto mixture, as
well as in the less reactive, and more typical, SynUrban mixture.
109
-------
I T I ' I ' I ' [ ' I '
June 26, 1985
DJ
8 9 10 11 12 13 14 15 16 17 18
HOURS. EOT
1.0
0.9
E 0.8
<* 0.7
I 0.6
2 0.5
c
I 0.4
I 0.3
0.2
0.1
0.0
Substitution
0% HCHO
September 1, 1984.
_ NO
1.0
0.9
0.8
0.7
0.6 j?
0.5 •*
0.4 |
0.3
0.2
0.1
8 9 10 11 12 13 14 15 16 17 18
HOURS, EOT
19
0.0
Figure 36.
Comparison of high ratio SynUrban experiments.
Top left: reduction;
Top right: substitution with no HCHO;
Bottom: substitution with 15% HCHO.
I
2
i.o
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
1 I I I ' I ' I • I
Substitution
15% HCHO
' i ' I '• I ' I ' i i i
August 22, 1984
_ NO
9 10 U 12 13 14 15 16 17 18
HOURS, EOT
19
1.0
0.9
0.8
0.7
0.6 I
0.5 -0
0.4 I
0.3
0.2
0.1
0.0
-------
1.0
0.9
E0.8
a
a0.7
CO
10.6
X
°0.5
0)
0)0.4
h~
io.3
0.2
0.1
0.0
I ' I ' I ' I ' I > I ' I ' \ ' I ' I ' I ' I ' I
September 1, 1984:
Test Sides A , rtrt -««r> A -
August 22, 1984 _
157, HCHO
5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
Figure 37.
Comparison of two high ratio SynUrban experiments.
Top: Effect of MeOH/15%HCHO substitution vs. MeOH/0%HCHO
substitution.
Bottom: Reference sides for both days. i.e. 100% SynUrban with
no reduction or substitution.
Ill
-------
Discussion SynAutUrb Experiments at 3 ppmC
SynUrban Experiments at 1 ppmC
There were three experiments performed at these conditions, one under poor sun.
These were sufficient to confirm that the reactivity of the SynUrban mixture at the
1.0 ppmC level was quite low. The September 2 experiment showed that the system
was very sensitive to HC concentration, in that 1.1 ppmC of SynUrban made about
0.12 ppm Oz and 0.8 ppmC of SynUrban made only 0.02 ppm 03. The September 3
experiment showed, however, that even in a system as sensitive to HC as this one,
MeOH substitution gave results similar to those at higher level of HC, i.e. SynMethanol
with 20% HCHO had a reactivity similar to the SynUrban mixture itself.
Based on the outcome of these experiments, we selected a higher HC concen-
tration for the 1985 substitution experiments with SynUrban. These were 2 and
1.5 ppmC of HC. One experiment, June 28, was performed at the 2 ppmC level and
it produced about 0.3 ppm Os under poor sunlight conditions. The substitution
of the normal level (actually 11% in this case) HCHO SynMethanol mixture for about
1/3 of the carbon resulted in only a small difference in ozone production, however,
the side with the MeOH/HCHO produced slightly more Os. The sunlight conditions
were very poor during the Os production and the significantly different levels of
HCHO may have had an unusual influence upon the outcome. Because of the poor
sunlight, this experiment must be treated as questionable.
SynAutUrb Experiments at 3 ppmC
When the UNCMIX tank was exhausted in the 1984 SynUrban experiments, we switched
to a mixture inbetween SynAuto and SynUrban that was produced by blending the
SynAuto mixture with n-butane that was used in the SynUrban mixture. Because
the SynUrban mixture already used the aromatic portion of the SynAuto mixture and
because the SynUrban mixture already had n-butane at the same level, this new
SynAutUrb mixture was equivalent to varying the paraffin and olefin fractions and
specific species. This type of experiment would ideally complement the SynUrban
and SynAuto experiments to see if the models could accurately track the changes in
basic composition of the mixtures.
Figure 38 shows three 3-ppmC (9:1 HC-to-NOx ratio) experiments with the
SynAutUrb mixture. The top two plots show the effects of 1/3 reduction, while the
bottom plot shows the effect of 1/3 substitution with a SynMethanol mixture (10%
HCHO).
The basic reactivity of this mixture is intermediate between the SynAuto and the
SynUrban mixture, being closer to the SynAuto at higher concentrations, and closer to
112
-------
1
a
I
y
J5
o
i
I.U
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0:2
0.1
0.0
_ ' l ' l ' l ' i ' l ' l ' l ' i ' l ' l ' l ' l ' l ' _
~ „ , x. September 8, 1984~
Reduction Os
.-•" '"•--.. -
-
/r -
/ /
/
TNO ,,-52/_ / ~
- ~I^><^ / >C>-^
".j 1 , ! , 1 . 1 ..X^- S-4_ 1 , 1 , 1 , 1 , 1 , 1 , 1 . -
5 6 7 8 9 10 11 12 13 14 15 16 17 18 1
I.U
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
9
HOURS. EOT
Figure 38.
Comparison of 3 ppmC SynAutUrb experiments.
Top left: reduction;
Top right: reduction (poor sun);
Bottom: substitution with 10% HCHO.
E
a
a
O ">
s i
3 X
-« 0
TJ C
1 ft
O
z
(A
*
X
O
§
L.
i
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
f\ 1
0.1
0.0
_ ' I
•^ |i
i 6
1.0
0.9
0.8
0.7
0.6
0.5
0.4.
0.3
0.2
0.1
nn
1
—
-
-
_
-
-
NC
-
_
i
i 1 i 1 . 1 i 1 i 1 i 1 i 1 i 1 i I i l i 1 > 1 i _
September 9, 1984"
Reduction
-
O3
/
*
/-- ~
r~^\ N02-/ / ~-
^^^^^^^-^^ ~
f< i , i i i i i.-i r^-^L j [ , i , i i i i i i i i
7 8 9 10 11 12 13 14 15 16 17 18 1
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
r\ •*
U. 1
0.0
9
HOURS. EOT
i I i I i I i I i I i I i I i I i , > I i I i I i _
_ . ... .. September 17, 198<
Substitution r
10% HCHO
-
_
03
— -— - J
/ "' MeOH 2
/ _. «ide
NO2 / ,-•'
^^y^^^^^i ~-
_^^\- / ••' ^^ ^^^ ^
i 1 i 1 i 1 . L^t-.i. 1 i 1 , 1 i 1 i 1 , 1 i 1 i '
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
nn
ff
o
.
T3
1
-.
N
i
1
6 7 8 9 10 11. 12 13 14 15 16 17 18 19
HOURS. EOT
-------
Discussion SynAutUrb Experiments at 3 ppmC
the SynUrban at lower concentrations. The effect of substitution is also intermediate
between the effects shown in Figure 33 and in Figure 36. The lower 10% fraction of
HCHO in the September 17 experiment resulted in an intermediate response between
that for the September 1 and the August 22 SynUrban experiments.
114
-------
The Mechanisms Selected Discussion
Modeling of Selected Experiments
A primary purpose for producing this new data was to use it to test models for
their correct representation of the effects of methanol substitution. To confirm that
the data were useful for such model testing, we selected six days for demonstration
modeling. We say demonstration modeling because actual testing of the models
would involve a significant effort and would require the use of chamber data from
the 400 run UNC database to test parts of the mechanism that must be assumed
to be correct for the purposes of this demonstration.
The days that were selected were two 1-ppmC SynAuto runs, two 3-ppmC SynAuto
runs, and two 3-ppmCSynUrban runs. The two 1-ppmC SynAuto runs were the reduc-
tion experiment, August 5. and the normal-HCHO, MeOH-substituted experiment,
August 7. The two 3-ppmC SynAuto runs were the reduction experiment, August 6,
and the high-HCHO, MeOH-substituted experiment, August 8. The two 3-ppmC
SynUrban runs were the MeOH-substituted experiment with no HCHO, September 1,
and the MeOH-substituted experiment with 20% HCHO. August 22.
The Mechanisms Selected
The mechanism recommended by EPA for EKMA control strategy simulations is
the Carbon Bond III (CBS). This mechanism was also used by SAI in the air shed
simulations to test the effects of MeOH fuel substitution described in Chapter I.2
SAI has been developing a newer, and more complex, version of the CBS called the
Carbon Bond Extended (CBX) mechanism. An early version of the CBX was used
in the air shed simulations described in Chapter 1 and it predicted a small benefit
for MeOH fuel substitution than did the CBS. The CBS mechanism used in this work
is listed in Table 15. The CBX mechanism used in this work is listed in Table 16.
The CBS mechanism uses a single species, CARB, to represent all aldehydes and
the composition of CARB has been built-in to the structure and rate constants of the
mechanism. CBS therefore is not very suitable to model chamber runs with high
HCHO concentrations without extensive modification of the mechanism. We used
the CBS only to model the "baseline" and HC reduction experiments (two days, two
sides).
115
-------
Discussion
The Mechanisms Selected
Table 15.
CARBON BOND MECHANISM III PER APPENDIX A OF DRAFT REPORT OF GUIDELINES
FOR USE OF CARBON-BOND MECHANISM IN OZIPM/EKMA JULY 1983 (ORIG NUMS)
RESTRUCTURED TO PKSS INPUT CONVENTIONS. LAST REVISED 5/22/84
(1) N02 = NO + 0
(2) 0 = 03
(3) 03 + HO = N02
# 1.0 /LI ; (1)
# 4.4E+06 ; (2)
# 3.452E+03 0-1450. ;(3)
(4) 0 + M02 = HO
# 1.30E+04;(5)
(5) 03 + N02 = N03
(6) HO + N03 = 2.0 * N02
(7) H03+ N02 = 2.0 * WHN03
# 178.6 ID -2450. ; (4)
# 2.8E+04 ; (11)
# 4.63E-19 0 10600. /V.' ; (12)
(8) 03 = 0
(9) 03 = DID
(10) DID = 0
(11) DID = 2.0 * OH
# 0.0584 /LI; (76)
# 0.004 /LI; (73)
# 4.44E10; (74)
# 3.4E5 /¥ ; (75)
(12) 03 -i- OH = H02
(13) 03 + H02 = OH
# 2.867E+03 0-1000.0 ;(6)
# 4.006E+02 0-1525. ; (7)
(14) NO + NO = 2.0 * H02
(15) OH + HO = HOMO
(16) HOHO = OH + NO
# 1.5E-04 ; (10)
# 9770.; (72)
#0.179 /LI; (71)
(17) NO + H02 = OH + H02
# 1.2E+04 ; (13)
(18) OH + N02 = HH03
# 1.6E+04 ; (8)
(19) OH + CO = H02
(20) OH (+CH4) = ME02
# 440. ; (9)
# 28.0; (NONE)
(21) H02 + H02 = H202
# 1.5E+04 ; (14)
116
-------
The Mechanisms Selected
Discussion
Table 15. cont.
( CARBOIIYL CHEMISTRY )
(22)
(23)
CARB = CO # 0.00248 /LI; (37)
CARB = 1.773 * H02 + 0.227 * ME02 +
0.227 * X + CO # 0.00220 /LI; (38A-D)
(24) OH " + CARB = CR02 + X
(25) OH + CARB = H02 + CO
(26) OH + CARB X + AC03
(27) CR02 + 110 = N02 + CARB + AC03 + X
( DICARBOIILY CHEMISTRY )
# 100.; (34)
# 9.0E+3; (35)
# 8.20E+3;(36)
# 1.2E4; (69)
(28) DCRB = H02 + AC03 + CO
(29) OH + DCRB = AC03 + CO
# 0.02 /LI; (67)
# 2.5E+04; (70)
( PAH CHEMISTRY )
(30) 110 + AC03 = 1102 + ME02 + C02
(31) 1)02 + AC03 = PAH
(32) PAN = AC03 + 1J02
(33) H02 + AC03 =
( ME02 CHEMISTRY )
# 1.04E+04; (26)
# 7000.; (39)
# 1.040E+18 8 -13500.; (40)
# 1.5E+04; (41)
(34) NO + ME02 = M02 + CARB + H02
(35) NO + ME02 = 1102 +• CARB t- ME02
(36) 110 + ME02 = NRAT
(37) H02 + ME02 =
# 7400. ; (30)
# 3700.; (29)
# 900. ; (31)
# 9000.; (42)
( PARRAFFIN CHEMISTRY )
(38) OH + PARC = ME02
(39) PARC + X =
# 5559. 0 -560. ; (16)
# l.OE+05 ; (15)
117
-------
Discussion The Mechanisms Selected
Table 15. cont.
( ETHYLE1IE CHEMISTRY )
(40) 0 + ETHC = 0.5 * ME02 + 0.5 * H02 + 0.5 * CO +
0.5 * CARB + 0.5 * PARC # 17582. « -800 ; (22.23)
(41) OH + ETHC = RB02 # 3330. « 382.; (24)
(42) 110 + RB02 = 1102 + 2.0 * CARB + H02 # 1.2E+04; (27)
(43) 03 " + RB02 = 2.0 * CARB + H02 #5.00; (32)
(44) 03 + ETHC = CARB + CRIG # 12.91
-------
The Mechanisms Selected
Discussion
Table 15. cont.
( AROMATIC CHEMISTRY )
(58) OH + AROC = RARO
(59) !!0 + RARO = 1102 + PHE1! + H02
(60) OH + PHEN = H02 + PARC + 2.5 * CARB
0.5 * DCRB + 1.5 * CO +
: 0.5 * X
(61) OH + PHEt! = PHO
(62) N03 + PHEI! = PHO + HN03
(63) N02 + PHO = !JPHN
(64) H02 + PHO = PHEN
# 26211. 9 -600.; (56)
# 4000.; (58)
# 3.E4; (66,60,61)
#1.0E4; (68)
# 5000.; (62)
# 4000.; (63)
# 5.00E4; (64)
(65) OH + AROC = H02 + OPEN
(66) OPEN + NO = 1!02 + 1.5 * DCRB + 1.5 * X
1.5 * CARB + 1.5 * CO
(67) OPEN +03 = 1.5 * DCRB + 1 . 5 * X
1.5 * CARB + 1.5 * CO
# 13397. S -400.; (57)
# 6000.; (59,60,61)
#40.; (65,60,61)
( 110 HETHANOL CHEMISTRY ADDED BECAUSE OF CARB )
(WALL PROCESSES )
NOV/ALL = !!02
FORMVALL = CARB
# 0.01/L1;
# 0.02/L1;
119
-------
Discussion The Mechanisms Selected
Table 16.
CARBON BOND MECHANISM CBM-X PER APPENDIX C OF DRAFT REPORT "USING THE
EXPANDED CARBON-BOND MECHANISM (CBM-X) Hi EKHA WITH COMPUTER CODE
OZIPM-3, SYSAPP-85/194, 23 APR 1985 (ORIGINAL NUMBERS 111 PARE1JS AT END)
RESTRUCTURED TO PKSS INPUT CONVENTIONS. [02]=210000 PPM [M]=1000000 PPM
LI: 1102 L2:03=01D L3:HCHO=H02 L4:HCHO=H2 L5:RCHO=RAD
(*********** INORGANIC CHEMISTRY **********)
( 1)
( 2)
( 3)
( 4)
( 5)
( 6)
( 7)
( 8)
( 9)
( 10)
( 11)
( 12)
( 13)
( 14)
( 15)
( 16)
( 17)
( 18)
( 19)
( 20)
( 21)
( 22)
( 23)
03
0
0
0
03
1)03
N03
1)03
1103
03
03
HO
NO
OH
OH
HONO
1102
0
+ HO
+ 1102
+ 1102
+ NO
+ 1102
= 110 +0
= 03
= N02
= NO
= 1103
= N02
= 1)03
= 0.85*1102 + 0.85*0 +
+ NO
+ 1102
+ 1102
11205
11205
03
03
* OH
+ H02
+ NO
f 1)02
+ NO
+ HONO
+ HONO
HONO
= 2.0*1102
= NO -i- 1102
= 11205
= 1103 + 1102
= 2.0*HN03
= 0
= 0.868*0 + 0.
= H02
= OH
= 2.0*N02
= 2.0*HONO
= HONO
= 1)02
= 110 + 1102
= OH + NO
# 1.0/L1
# 4 . 60247E+05 8
# 3.2278E-I-03 8-
# 1.38E+04
# 309.8 8 600
- # 785.6 8 411
# 176.3 8-2450.
0.15*110 # 30.6 /LI
; (1)
690. ;(2)
1430. ;(3)
;(4)
;(5)
;(6)
;(7)
;(30-1)
# 1.2144E+04 8 250 ; (14)
# 36.6 8-1230
# 831.9 8 226
;(15)
;(16)
# 1.96E-H6 8-10840 ; (18)
# 1.9E-6 /V;
# 0.042 /LI
2644*OH # 1.0E-3/L2
# 2343.8 8-940 ;
# 21.0 8-580 ;i
# 2.6E-5 8 530 ;l
# 1.6E-11 /W ;|
# 2331.3 8 427 ; 1
# 9770. ;!
# 1.5E-5 ;l
# 0.18 /LI ;(
;(17)
;(9)
; (8,10,11)
( 12)
( 13)
C 25)
C 19)
[ 23)
: 28)
: 20)
; 21)
( 24) H02 t- NO = OH + 1102 # 5497.2 8 240 ;( 24)
120
-------
The Mechanisms Selected Discussion
Table 16. cont.
( 25) H02 + 1)02 = P!!A # 205.6 C 617 ;( 34)
( 26) PtlA = H02 + N02 # 7.85E-H5 ffl-10420 ;( 35)
( 27) OH + H02 = HN03 # 2489.2 0 560 ;( 22)
( 28) OH + HN03 = 1103 # 14.11 0 778 ; ( 29)
( 29) H02 + H02 = H202 # 87.39
-------
Discussion
The Mechanisms Selected
Table 16. cont.
(
(
(
(
(
(
(
(
(
(
(
(
(
(
(
(
49)
50)
51)
52)
53)
54)
55)
56)
57)
58)
59)
60)
61)
62)
63)
64)
(**********
(
(
(
(
(
65)
66)
67)
68)
69)
ME02 +•
OH +
RC03 +
RC03 +
H02 +
1103 +
0 +
OH +
OH +
ME02 f
ME02 +
RC03 +
ME02 +
RC03 +
1102
MP1IA
RCHO
NO
N02
PAN
RCHO
RCHO
RCHO
MEH03
KE1I02
ME02
RC03
RC03
H02
H02
= MPIIA
= ME02 + 1102
= RC03
= M02 i
= PAN
• ME02
= RC03 + 1102
= ME02 +
HCHO
= RC03 + HH03
= RC03 +
= HCHO
= HCHO
= 0.696*
= MEO +
= 2.0 *
= (ROOH)
= (ROOH)
DICARBOI.'LY CHEMISTRY *
OH +
OH +
MGPX +
GLY
GLY
MGLY
MGLY
HO
= 1.785*
= H02 +
= RC03 f
= MGPX
= RC03 +
OH
+ N02 -
+ no
MEO + 0.652
ME02
ME02
*********)
#
#
#
#
#
#
#
#
#
#
#
503.32
8 735 ;(53)
1.32E+178-10400 ; (54)
10372.
7130.9
3889.6
8 250 ; (46)
8 250 ; (50)
8 250 ;(51)
5.62E+188-14000 ; (52)
5.0
3.7
17394.7
7430.3
7417.4
*HCHO (+ 0.
#
#
#
#
#
240.44
4400.0
3700.0
113.45
9600 . 0
;(49)
;( 47)
8-986 ; ( 45)
8-360 ; (60)
8-340 ;(61)
652*MEOH)
8220 ; (63-4)
;(65)
;(66)
8 1300 ; (67)
;(68)
CO + 0.213*HCHO + 0.186*H02
2.0*CO
H02 + CO
H02
#
#
#
#
#
0.0075/L1 ;(72-4)
15000.0
0.02/L1
26000 . 0
12000.0
;(75)
;(76)
;(77)
;(78)
(*************** BACKGROUND METHANE ***************)
( 70)
OH (•»• CH4) = ME02
# 21.0
;(79)
122
-------
The Mechanisms Selected Discussion
Table 16. cont.
(*************** PARAFIl! CHEMISTRY ***************)
( 71) OH + PAR = 0.13*PAR02 + 0.87*PAR02R # 1150.0 ;(80-1)
( 72) PAR02 + 110 = 1102 + H02 + RCHO + X # 12000.0 ; (82)
( 73) PAR02R + NO = 0.923*1102 + 0.923*PAROR ( + 0.077*PAR1I03)
# 13000.0 ; (83-4)
( 74) PAROR = RCHO + D + X # 1.43E+15 C-7000 ;( 87)
( 75) PAROR = 0.385*KETOHE + 0.385*H02 + 0.615*ACTOtlE
+ 0.615*0 + 1.23*X # 390000.0 ;(86,88)
( 76) PAROR + N02 = (PARN03) # 22000.0 ;(85)
(77) D + PAR = 0.30*PAR02 +-Q.70*A02 + 1.40*X # 10000.0;(90-1)
( 78) A02 + 110 = 1102 f H02 + ACTOtlE # 12000.0 ; (93)
(79) D + KETOIIE = RC03 + X # 10000.0 ;( 92)
(80) X + PAR = # 10000.0 ;(89)
(*************** KETONE/ACETOHE CHEMISTRY ***************)
( 81) ACTOtlE = ME02 + RC03 # 0.00004/L1 ; (69)
( 82) OH + ACTOME = ACOC02 # 580.0 ;(70)
( 83) ACOC02 + MO = N02 + HCHO + RC03 # 12000.0 ;(71)
( 84) KETOIIE = RC03 + PAR02 + 2.0*X # 0.0003/L1 ; (94)
123
-------
Discussion
The Mechanisms Selected
Table 16. cont.
(*************** ETHYLENE CHEMISTRY ***************)
( 85)
( 86)
0 + ETH HE02 + H02
OH + ETH = ME02 + HCHO
CO # 15824.0(8-800 ; (104)
# 3330.2 8 382 ;(105)
( 87)
03 + ETH = HCHO + 0.37*CRIG + 0.3654*CO +
0.1260*H02 (+ 0.0441*FACID)
#37.188-2840 ;(106-7,108-11)
(*************** QLEFIIJ CHEMISTRY ***************)
(88) 0 + OLE = 0.80*RCHO + 0.15*PAR02 + 0.15*H02 + 0.15*CO
+ 0.05*ME02 + 0.05*RC03 + 0.20*X
# 17560.0 C-324;(95-7)
( 89)
( 90)
OH + OLE = ME02 + RCHO + X # 6928.6 « 537 ; (98)
03 + OLE = X + 0.50*RCHO + 0.50*HCHO + 0.30*CRIG
1- 0.30*MCRG + 0.116*CO + 0.04*H02 (+ 0.014*FACID)
+ 0.144*HE02 + 0.096*CO + 0.096*H02 + 0.016*HCHO
+ 0.080*OH # 10.46 0-1897 ;(99-102.108-15)
( 91)
( 92)
( 93)
( 94)
( 95)
( 96)
( 97)
( 98)
( 99)
(100)
(101)
(102)
CRIG +
CRIG +
CRIG f
CRIG +
MCRG +
MCRG +
MCRG +
MCRG +
N03
N02R02
HO
l!02
CRIG
HCHO
RCHO
NO
N02
MCRG
HCHO
RCHO
+ OLE
+ NO
= HCHO + 1102
= HCHO + 1103
= (FACID)
= (OZD)
= (OZD)
= RCHO + 1102
= RCHO + 1103
= (AACID)
= (OZD)
= (OZD)
= I102R02
= (0.091*011102)
+ 0.909*H02 +
;(119)
# 10000.0
#1000.0
# 0.006 l\
#30.0
#30.0
# 10000.0
# 1000.0
# 0.006 /W
#30.0
# 30.0
# 11.4
+ (0.909*PARN02)
0.909*1102 # 11000.0 ; (126-7)
124
-------
The Mechanisms Selected
Discussion
(103)
(104)
(105)
(106)
(107)
(108)
(109)
(110)
(111)
(112)
(113)
(114)
(115)
(116)
(117)
* AROMATIC
TOL + OH
OPE1! + HO
B02 + HO
BZA
OH + BZA
BZ02 f HO
BZ02 + H02
PBZH
PH02 f HO
PHO + 1102
PHE11 + 1J03
XYL + OH
XYLO + HO
TLA + OH
TL02 + 1JO
Table 16. cont.
CHEMISTRY ***************)
= 0.564*OPE1! + 0.564*GLY +
+ 0.359*H02 t- 0.359*PAR +
= H02 ••• H02 + MGLY f GLY
= H02 + H02 + BZA
= (PROD)
= BZ02
= 1102 + PH02 + CO
= PBZ1I
= BZ02 + 1102 # 5.
= H02 + PHO
= HPHH
= PHO + H1I03 -
= 0.056*XYLO + 0.278*H02 +
+ 0.556*PAR + 0.666*OPE!J +
= H02 + H02 + TLA
= TL02
= H02 + PHO + 2.0 * PAR
0.359*PHEH
0.077*802
# 9750.0
# 10000.0
# 12000.0
# 0.004/L1
# 20000.0
# 3700.0
# 2500.0
57E+18 fi-14000 ;(137)
# 12000.0 ;(140)
# 20000.0
# 14000.0
0.278*PHE!I
0.666*MGLY
# 36000.0
# 12000.0
# 20000.0
# 4000.0
(128-30)
(131)
(136)
;(138)
(133)
(135)
(134)
;(139)
(132)
; (145-7)
;(142)
(METHAHOL CHEMISTRY )
OH + MEOH
MEOHO
= HCHO + H02
= HCHO + H02 + 110
# 1550 ;
# 0.2/L1
( WALL CHEMISTRY )
1-10%'ALL = H02
FORMV.'ALL = HCHO
# O.I/LI
# 0.2/L1
125
-------
Discussion The Mechanisms Selected
A popular alternative to the Carbon Bond approach for constructing mecha-
nisms has been the the Atkinson, Lloyd, and Winges (ALW) reaction mechanisms.
This mechanism is currently undergoing significant up-dating and testing under
EPA contract using data from our chamber as well as from the UCR indoor cham-
ber. The original ALW mechanism as used in this study is listed in Table 17.
The ALW and CBX mechanisms are much more complex than the CBS mecha-
nism, and are much more expensive to use. but they provide explicit representation
of HCHO as well as many of the other species in the SynAuto mixture. The ALW
mechanism was used to model four SynAuto days and the CBX mechanism was used
to model the four SynAuto days and two SynUrban days.
126
-------
The Mechanisms Selected Discussion
Table 17.
THE ATKINSON ET AL. REACTION MECHANISM AS PRESENTED IN
TABLE A. 2 OF THE LEOl.'E AND SEINFELD REPORT PART 2
RENUMBERED AND RE-ORGAlilZED BY JEFFRIES
( INORGANIC REACTIONS )
( 1) N02 = 110 + 03 #1.0 /L;
( 2) 110 + 03 = N02 #3355.7 8-1450;
( 3) 1102 + 03 = 1103 #177.9 8-2450;
( 4) 110 + N03 = 2.00*N02 #28188;
( 5) N03 = 0.30*110 + 0.70*H02 + 0.70*03 #15.5 /L;
( 6) 1102 + N03 = 11205 #104027 8-1100;
( 7) 1J205 = N02 + M03 #3.5E18 8-12280;
( 8) 1J205 = 2.00*HN03 #4.46E-6 A';
(9) 03 = 2.00+OH #2.3E-8 A1 /L;
(10) OH + NO = HONO #9796.9;
(11) HOI10 = OH + NO #0.17 /L;
(12) OH + M02 = HN03 #16891;
(13) H02 + NO = OH + 1102 #12416;
(14) H02 + N02 = H021J02 #1689;
(15) H02N02 = H02 + 1102 #7.8E15 8-10420;
(16) H02 + H02 = H202 #114.09 81100;
(17) H02 + H02 = H202 #6.53E-10 85800 /'//;
(18) H202 = 2.00*OH #7.1E-4 /L;
(19) OH + 03 = H02 #2349 fi-940;
(20) H02 + 03 = OH #16.11 8-580;
(21) OH + CO = H02 #436.2;
127
-------
Discussion
The Mechanisms Selected
Table 17. cont.
( FORMALDEHYDE CHEMISTRY )
(22) HCHO = 2.00*H02 + CO
(23) HCHO = CO
#3.1E-3 /L;
#3E-3 /L;
(24) OH + HCHO = H02 + CO
#14765;
( ACETALD'EHYDE CHEMISTRY )
(25)
CCHO = COO + H02 + CO
#6E-4 /L;
(26) OH + CCHO = CC03
(27) CC03 + NO = 1102 + COO
(28) CC03 + H02 = PAt!
(29) PAH » CC03 + H02
#10067 0250;
#10403;
#7047;
#1.2E18 8-13543;
(30) COO f 110 = HCHO + H02 + 1102
(31) RC2CHO = CC02 + CO + H02
#10403;
#8.4E-4 /LI;
(32) OH + RC2CHO = RC2C03
(33) RC2C03 + NO = CC02 + N02
#30872;
#10403;
(34) RC2C03 + 1102 = HIGHPANS
(35) HIGHPANS = RC2C03 + N02
#7047;
#1.2E18 S-13543;
(36) CC02 + NO = CCHO + H02 + 1102
#10403;
( DICARBONYL CHEMISTRY )
(37) GLY = HCHO + CO
(38) OH + GLY = H02 + CO
(39) MGLY = CC03 + H02 + CO
(40) OH + MGLY = CC03 + CO
#1E-10 /LI;
#29530;
#0.15 /LI;
#22148;
128
-------
The Mechanisms Selected Discussion
Table 17. cont.
( KETOIJE CHEMISTRY )
(41) ACETOKE = CC03 + COO #1.7E-3 /LI;
(42) ETHMEKET = CC03 + CC02 #1.7E-3 /LI;
(43) OH + ETHHEKET = RRRCG2 #14765 8-330;
(44) RRRC02 + HO = H02 + CCHO + CC03 #10403;
( ALKA1IE ' CHEMISTRY )
(45) OH + PROPANE = PROPAC02 #22148 0-680;
(46) PROPAC02 + HO = H02 + 1J02 + ACETOUE #10403;
(47) OH + ALKANES = RC3C02 #22148 6-400;
(R=H,CH3,ETH.PRO, . . . )
(48) RC3C02 + MO = -0.80*110 +• 1.70*N02 +-0.9*H02 + 0.15*HCHO +0.30*CCHO
f 0.10*RC2CHO + 0.30*ACET01IE + 0.45*ETHHEKET
#10403;
( OLEFIU CHEMISTRY )
(49) OH + ETHEKE = -1.00*110 + M02 + H02 + 2.00*HCHO #3255 9380;
(50) OH + PROPEHE = -1.00*NO + 1102 + H02 + HCHO + CCHO #6040 0540;
(52) 03 + ETHE11E = HCHO + 0.40*CH2DIOX + 0.40*CO + 0.12*H02
#14.09 (B-2560;
(53) 03 * PROPEKE = 0.50*HCHO + 0.50*CCHO + 0.20*CH2DIOX + 0.20*ETHDIOX
+ 0.30*CO + 0.20*HQ2 + 0.10*OH + 0.20*COO
#10.40 fi-1900;
(55) CH2DIOX + MO = HCHO + 1102 #10403;
(56) CH2DIOX •»• N02 = HCHO + N03 #1040;
(57) CH2DIOX = (PRODUCT1) #5.03E-3 /W;
(58) ETHDIOX + MO = CCHO + 1102 #10403;
(59) ETHDIOX + N02 = CCHO + M03 #1040;
(60) ETHDIOX = (PRODUCT2) #5.03E-3 A';
129
-------
Discussion The Mechanisms Selected
Table 17. cont.
( AROMATIC CHEMISTRY )
(61) OH + BEl.'ZEUE = 0.25*CRESOL + 0.25*H02 + 0.75*AROADD01 #1778.5;
(62) OH + TOLUENE = 0.20*CRESOL + 0.20*H02 + 0.65*AROADD01
+ 0.15*AROC02 #9060.4;
(63) OH + RRRC8ARO = 0.25*CRESOL + 0.25*H02 + 0.75*AROADD01 #33557;
(64) AROADD01 + NO = 0.75*N02 + 0.75*H02 + 0.75*GAMDIALS + 0.75*HGLY #10403;
(65) AROC02 + HO = 0.75*1102 + 0.75*H02 + 0.75*BE!,'ZALD #10403;
(66) OH •*• GAMDIALS = GRADICAL #43624;
(67) GRADICAL + NO = -2.00*110 + 3.00*1102 + 0.55*H02 + 0.55*GLY + 0.45*CC03
+ 0.45*MGLY + 0.55*CO #10403;
(68) GRADICAL + N02 = GRADN02 #7047;
(69) GRADN02 = GRADICAL + 1102 - #1.2E18 C-13543;
(70) OH + CRESOL = AROADD02 #63758;
(71) AROADD02 + NO = 0.75*H02 + 0.75*H02 + 0.75*GAMDIALS #10403;
(72) H03 + CRESOL = H1I03 + PHEMOXY #22148;
(73) BENZALD = (PRODUCT4) #4.5E-3 /LI;
(74) OH + BEHZALD = BZC03 #19128;
(75) BZC03 + HO = U02 + BZ02 #10403;
(76) BZC03 + 1102 = PBZN #7047;
(77) PBZH = ti02 + BZC03 #1E17 C-13025;
(78) BZ02 + NO = PHEHOXY + !!02 #10403;
(79) PHEHOXY + 1102 = (PRODUCTS) #22148;
(80) OH + (CH4) = COO # 28. ;
130
-------
The Mechanisms Selected Discussion
Table 17. cont.
( METHAI10L CHEMISTRY )
OH 1- MEOH = HCHO + H02 #1550 ;
ME01IO = HCHO + H02 + NO #0.2/L1;
( WALL PROCESSES )
HOY/ALL = N02 #0.01/L1;
FOR1WALL = HCHO #0.02/L1;
( DILUTI011 ENTRAPMENT )
( CONTIHENTIAL BACKGROUND AIR )
= 0.30*CO + 0.04*03 + 0.0005*1)02 +
0.005* HCHO + 0.0025 * CCHO +
0.001*ETHENE + 0.0005*PROPE!1E + 0.0008*TOLUEIIE #1.0/E;
KEY : :
# -- rate constant at 300 deg K or A-factor
0 -- activation energy
/L -- reaction depends upon light
/Y.' -- reaction depends upon water
; -- end of reaction
( ) -- comments
UNITS :: PPM and MIHS
131
-------
Discussion Modifications and Assumptions
Modifications and Assumptions
The chemistry of HO + MeOH was added to ALW and CBX in the form of explicit
chemistry for MeOH. See the end of each mechanism's listing for the reactions that
were added.
All chambers have some wall artifact processes that can potentially affect the
experimental results. We have investigated these processes in a number of cham-
ber characterization experiments. The major observed chamber processes are the
slow appearance of gas phase NOX in experiments in which no NOX was added to
the chamber and the more rapid oxidation of significant NO concentrations in the
absence of injected organics. The appearance of NOX in the chambers is believed to
be from the release of nitrogen material that was adsorbed on the chamber walls in
previous experiments. For example, nitric acid is a major nitrogen end product in
most chamber experiments and it has a high affinity for chamber walls. The rapid
oxidation of NO in the absence of injected organics is an indication of a radical
source associated with the chamber itself. This too is believe to be from material
adsorbed on the chamber walls in previous experiments.
Modeling studies of the UNC chamber have suggestedj7 that an adequate rep-
resentation for the NOX process was
u;a//.NOx - N02
at a rate proportional to the light intensity. The amount of u>a//.NOx needed to
account for the gas phase appearance was a function of the chamber history. Typical
values were between 5-20 ppb of u>a//.NOx; this amount was introduced into the
chamber over a 10 hour period. Likewise, an adequate representation of the NO
oxidation processes was found to be represented by introduction of HCHO, also from
the walls by
wa//.HCHO = HCHO
at a rate proportional to the light intensity. The amount of u>a//.HCHO needed to
account for the NO oxidation was also a function of the chamber history. Typical
values were between 5-35 ppb of u;a//.HCHO.
In addition to these processes, injected gas-phase NO2 may react with the cham-
ber walls to produce HONO, a powerful radical source, at a rate dependent upon
the surface conditions. Further, some HONO may be present initially as a result of
high concentrations of NOX occurring during chamber injections. We believe that
the most common chamber initial condition is 0.0 ppb of initial HONO. Sometimes
we have found that simulations require 1-2 ppb initial HONO.
132
-------
The Simulations Discussion
The additional reactions for chamber processes are clearly indicated at the bot-
tom of each of the mechanism listings.
Reactions that add continental background HC in the dilution air were also
added to the mechanism. In addition, 38 ppbC of continental background HC was
added to the chamber injections, as well as background CO (0.3 ppm) and CH4.
Several potential refinements, such as a zenith angle dependence of the ratio of
formaldehyde photolysis to NC>2 photolysis and optimally adjusting the photolysis
rates for the UVR-to-TSR ratio for specific days were not included in these simula-
tions.
The ALW mechanism was designed for conditions in which the NO concentration
would never be below 10~4 ppm. Although this may be true in urban simulations, it
certainly was not true in reactive smog chamber simulations such as those of August
6 and 8. In these simulations, the simulation program had to be stopped after about
400 minutes when the NO concentration fell below 0.1 ppb so these model plots are
short compared to the data and the lower concentration simulations.
In comparing the data and model profiles, remember that the NOo experimental
data also includes PAN and at the end of the fast runs this data is almost 100% PAN.
The plotted model data includes both PAN and HN03 profiles; these are labeled on
the plots.
A Caution
As in the reactivity issue, there is no single adequate measure for model-experiment
comparison. The best understanding comes from a direct comparison of the species
profiles from the experiment and the simulation and model-to-model comparison
when different assumptions have been made. Simple measures such as the maximum
O3 prediction error are misleading. In some of the cases described below, different
assumptions have been used to force a better fit for the purpose of illustrating the
magnitude of model-data disagreement. For these cases, reporting errors for O3
maximum predictions is meaningless.
The Simulations
We performed 40 simulations. Entries in columns and rows of Table 18 indicate
which conditions were simulated. The columns of Table 18 represent the mecha-
nisms and wall assumptions used. The entries in the table body are the difference
in minutes between the experiment and simulation NO-to-NOj-crossovers (which is a
133
-------
Discussion Discussion of Model Results
measure of the model's initial reactivity), and the magnitude of the model's 03 max-
imum relative to the experiment (which is a measure of the model's final reactivity).
A j means the model over- predicted and a j means that the model underpredicted
the ozone; an = means that the model and data Os maximum agreed.
Table 19 describes the wall assumption terms used in Table 18.
Figure 39, and following figures, show the model and data profiles for NO, NC>2,
and 03 and the model profiles for PAN and HNO3. The top plot in each figure is
RED chamber data, and the bottom plot is BLUE chamber data; the solid lines are
experimental data and the dashed lines are model predictions.
Discussion of Model Results
Overall Results
The large majority of all the simulation results were slower than the experiments,
even with extreme assumptions of wall processes. For example,
• CBX simulations without any wall assumptions (the minimal case) for the basic
SynAuto mixture (i.e. no SynMethanol added) experiments were 70 to 125 minutes
slow (plots of these simulations are not shown);
• CBX simulations with high wall assumptions for the basic SynAuto mixture ex-
periments were still 40 to 100 minutes slow;
• CBX simulations with extreme wall assumptions for the basic SynAuto mixture
were slow by 15 to 45 minutes;
• for CBX, in these SynAuto simulations, there was a direct inverse relationship
between the initial HC concentration and the lateness of the crossover time;
• CBS simulations were also slow relative to the data, but these underpredicted
the O3 because of higher NC>2 losses at the end of the simulations (i.e. the pre-
dicted PAN exceeds the NOo measurement, which includes PAN);
• CBS was less sensitive to wall radical sources than was CBX;
• ALW simulations were very similar to the CBX simulations; in fact, overlaying
the two mechanism's simulations for August 5 shows essentially a perfect match
in all species profiles predicted by the two mechanisms;
A general trend, is that all three mechanisms appear to have too little reac-
tivity in the beginning of SynAuto simulations and too much reactivity in the end.
When the initial reactivity is artificially increased by the use of extremely high wall
assumptions and radical sources, the models tend to do better, but now tend to
overpredict the latter stages of the experiments.
134
-------
Discussion of Model Result? Discussion
Table 18. Model Simulations General Performance
(numbers are data - model times to NO-to-NO2-crossover)
(t is over prediction of ozone, | is under prediction of ozone)
CBS CBS ALW CBX CBX CBX
Date HC/SynMeOH high extra high none high extra
SynAutb Experiments
AU05 0.91 -100] -50 I -100 j -125 | -105 | -45 =
1.31 -60 t -30 | -90] -80 I -801 -30 T
AU06 2.25 -30 j -10 ] -60 | -85= -501 -20 =
3.23 -201 0| -40 T -70 T -38 | -1ST
AU07 1.32 -751 -851 -40 |
0.87/0.36 -751, -651 -40 T
AU08 2.48/1.02 -35 t +25 f -32 |
3.68 -20 | -35 t -8 |
SynUrban Experiments
AU22 3.04 +25 ||
2.04/1.00 -fSOTT
ST01 3.31 -j-8 TT
2.66/1.04 +8 tl
Table 19. Model Simulation Wall Assumptions
(ppb of initial material on walls or reaction rate)
Condition u;a//.NOx u;a//.HCHO MONO N02+ walls
none
typical
high
extra
0
10
25
25
0
25
50
50
0
0
0
5
0
0
0
1.6 x 10~4
135
-------
Discussion Discussion of Model Results
It was not the purpose of this work to test the mechanisms. We modeled a
few experiments in the UNC database to gain insight into what might be causing
the problems; these simulations are not shown here. CBX was used to model
an experiment with UNCMIX that had m-xylene added to one side and tri-methyl-
benzene added to the other side. The experiment produced essentially matched
results for the two sides. With high wall sources necessary to have the initial timing
of the simulation agree with the experiment, the CBX simulation likewise produced
matched results for the two sides, but over predicted the Os production in both
sides by about 40%.
In another experiment a four component simple mix (n-butane, pentane, ethy-
lene, propylene) was compared with UNCMIX; both mixtures had the same total
carbon fraction in the paraffin and olefin classes (0.70/0.30) and neither mix had
aromatics. In the experiment, the simple mix was slightly more reactive than the
UNCMIX. In the CBX simulation with high wall sources, the UNCMIX side was simu-
lated reasonably well, but the simple mix side was underpredicted by nearly 50%!
Because all the paraffin carbon is treated the same way in the CBX, the difficul-
ties between the simple mix and the UNCMIX side (which had the same amount of
paraffin carbon) must be in the treatment of the olefins. The simple mix had 20%C
as ethylene, as did the SynAuto experiments in which CBX also performed poorly.
On the other side. UNCMIX only had 10%C as ethylene and the SynUrban. for which
the CBX performed better, had only 6%C as ethylene. This suggests that the new
explicit chemistry for ethylene in CBX needs to be tested.
It appears that the CBX mechanism has too little reactivity in the simplest
paraffin/olefm portion of the mechanism and too much reactivity in the aromatics
portion of the mechanism.
We recommend that the mechanism be further tested before it is used in air
shed simulations for methanol fuel scenarios.
Analysis of Substitution Effects
An analysis technique described by Jeffries18 in which integrated reaction rates
are used to compute a process mass balance and a pathway flowchart was used
to investigate the relative effects of the SynMethanol substitution for the August 8,
3-ppmC 9:1 HC-to-NOx ratio experiment with »33% substitution using 20% HCHO.
The ALW model simulation was used as the source of the data. Although ALW
was somewhat slow on this day, it did approximate the side-to-side differences in
the data reasonably well.
136
-------
Discussion of Model Results Discussion
Analysis of the model predictions after 600 minutes showed that although there
was less concentration of each species in the SynAuto/SM side, the high methanol
and formaldehyde resulted in 16% higher HO and HO2-production. Thus, in spite of
the concentration differences in the "reactive" HCs, there was nearly the same total
HC consumption in both simulations due to the higher HO in the MeOH-substituted
side.
The 100% SynAuto side consumed a total of 0.60 ppmV of 0.90 ppmV initial HC
(66%) and the methanol substituted side consumed 0.39 ppmV of 0.58 ppmV initial
HC (67%). In addition, the methanol side consumed 0.15 ppmV of MeOH bringing
the total consumption to 0.54 ppmV. in relatively close agreement with that on the
higher HC, 100% SynAuto side.
A major difference was that the 100% SynAuto side formed a total of 1.04 ppmV
of aldehydes and reacted 0.70 ppmV, giving a net formation of 0.34 ppmV aldehydes
(0.308 ppmV HCHO). On the MeOH-substituted-side, only 0.77 ppmV was formed and
0.80 ppmV was reacted (the difference was taken from the initial aldehyde) giving no
net aldehyde formation. Because there was an initial 0.24 ppmV of HCHO, however,
a final concentration of 0.22 ppmV aldehyde (0.148 ppmV HCHO) resulted on the
MeOH-substituted side. The amount of HCHO formed from MeOH was 0.15 ppmV.
Taken altogether, the 100% SynAuto side reacted 1.30 ppmV of HC and aldehydes
and the 33%. SynMethanol substituted side reacted 1.34 ppmV of HC and aldehydes.
There was 1.41 ppm N02 created per ppm HO reacted in the 100% SynAuto side and
1.23 ppm NO2 created per ppm HO reacted in the SynMethanol-substituted side, a
reflection of shorter organic oxidation pathways in the methanol side. As indicated
above, however, 16% more HO was produced in the methanol side, which essentially
made up for the difference in NOj-production. That is there were more short cycles
on the substituted side, resulting in approximately the same conversion of NO.
This model suggests a clear explanation as to why the two systems would pro-
duce the same O3 in this case. In the actual data for the August 8 experiment, the
100% SynAuto side did make more HCHO (peak value was 0.32 ppm at 1000 EDT)
than the MeOH-substituted side (peak value about 0.30 ppm at 1200 EDT). Note,
however, that the values discussed above were the sums and differences of total
throughputs of various reactions at the end of the 10-hour simulation and thus
should not be directly compared with maximum concentration measured in the ex-
periment. This analysis also shows the complexity of the chemical situation and
illustrates why photochemical kinetics models are needed to understand the effects
137
-------
Discussion Discussion of Model Results
of major compositional changes. Improved model fits are needed, however, before
more extensive analysis would be worthwhile.
138
-------
1.0
0.9
E 0.8
a
a 0.7
I 0.6
x
° 0.5
0)
g.0.4
i 0.3
0.2
0.1 f-
0.0
I ' !
1 ; ' ' ' I ' i ' ! '
August 5, 1984
Sol id
Dashed
Lines: Data
Lines: Model
NO
NO,
6 7 8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
1.0
0.9
0.8
i i i i r i i r \ i r r
Sol id Li nes: Data
Dashed Lines: Model
8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
Figure 39. Only High NOx Wall Conditions.
CB3 model; 0 ppb Wall HCHO. 25 ppb Wall NOx, 0 ppb initial HONO;
no formation of HONO on walls.
Top: 0.91 ppmC SYNAUTO, no MeOH, 0.01 ppm HCHO
Bottom: 1.31 ppmC SYNAUTO, no MeOH, 0.02 ppm HCHO
139
-------
1.0
0.9
I I I I I I I I i I i i i I I I i i 1 I I i i I
August 5, 1984
Sol id Lines: Doto
. Dashed Lines: Model
8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
1.0
0.9
0.8
I ' I
Sol id Lines: Data
Dashed Lines: Model
6 7 8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
Figure 40. NO Wall HCHO Wall Conditions.
CB3 model; 0 ppb Wall HCHO, 25 ppb Wall NOx, 5 ppb initial HOMO;
normal formation of HOMO on walls.
Top: 0.91 ppmC SYNAUTO, no MeOH, 0.01 ppm HCHO
Bottom: 1.31 ppmC SYNAUTO. no MeOH, 0.02 ppm HCHO
140
-------
1.0
0.9
I ' I ' I ' I ' i ' I ' I '
August 5, 1984
Sol id Lines: Dota
Dashed Lines: Model
6 7 8 9 10 11 12 13 14 15 16 17 18
HOURS, EOT
19
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
1.0
0.9
0.8
i i i r
I ' I ' I
Sol id Lines: Data
Dashed Lines: Model
6 7 8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
1.0
0.9
0.8
0.7
0.6 g
0.5 $
T3
0.4 1
0.3
0.2
0.1
0.0
Figure 41. High Wall Conditions.
CBX model; 50 ppb Wall HCHO, 25 ppb Wall NOx. 0 ppb initial HONO;
no formation of HONO on walls.
Top: 0.91 ppmC SYNAUTO, no MeOH, 0.01 ppm HCHO
Bottom: 1.31 ppmC SYNAUTO, no MeOH, 0.02 ppm HCHO
141
-------
August 5, 1984
Sol id Li nes: Data
Dashed Lines: Model
8 9 10 11 12 13 14 15 16 17 18
HOURS, EOT
1.0
0.9
i i i r
i T l I i I I f
Sol id Lines: Data
Dashed Lines: Model
03
6 7 8 9 10 11 12 13 14 15 16 17 18
HOURS, EOT
19
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
I
CD
TJ
TJ
Figure 42. Extra Wall Conditions.
CBX model; 100 ppb Wall HCHO, 25 ppb Wall NOx, 5 ppb initial HONO;
normal formation of HONO on walls.
Top: 0.91 ppmC SYNAUTO, no MeOH, 0.01 ppm HCHO
Bottom: 1.31 ppmC SYNAUTO, no MeOH, 0.02 ppm HCHO
142
-------
1.0
0.9
E 0.8
a
a 0.7
(/>
0.6
0)
0.5
n j : i I j i I l j I -
August 5, 1984
Sol id Lines: Data
Dashed Lines: Model
8 9 10 11 12 13 14 15 16 17 18
HOURS, EOT
19
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
g
1.0
0.9
E 0.8
a.
a 0.7
OT
I 0.6
x
0 0.5
0)
g>0.4
i 0.3
0.2
0.1
0.0
Sol id Lines: Data
Dashed Lines: Model
NO
8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
Figure 43. High Wall Conditions.
ALW model; 50 ppb Wall HCHO, 25 ppb Wall NOx, 0 ppb initial HONO;
no formation of HONO on walls.
Top: 0.91 ppmC SYNAUTO, no MeOH. 0.01 ppm HCHO
Bottom: 1.31 ppmC SYNAUTO, no MeOH, 0.02 ppm HCHO
143
-------
1.0
0.9
0.8
I ! I I I I I I I I :
Sol id Lines: Data
Dashed Lines: Mode!
'• I ' I ' I ' I ' I ' !
August 6, 1984
/
os
t-r-l—r-4-. I i I
8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
1.0
0.9
0.8
0.7
0.6 g
0.5 $
TJ
0.4 |
0.3
0.2
0.1
0.0
1.0
0.9
g 0.8
a
a 0.7
w
i 0.6
3 0.5
0)
g> 0.4
I 0.3
0.2
0.1
0.0
Sol id Lines: Data
Dashed Lines: Model
HN03
-- PAN
£:?:>-!' I i I T-h-r-l—h-1— t— I—T I V
5678
9 10 11 12 13 14 15 16 17 18
HOURS, EOT
19
1.0
0.9
0.8
0.7
0.6
Oc
.5
0.4
0.3
0.2
0.1
0.0
§>
3
Figure 44. Only High NOx Wall Conditions.
CB3 model; 0 ppb Wall HCHO. 25 ppb Wall NOx, 0 ppb initial HONO;
no formation of HONO on walls.
Top: 2.25 ppmC SYNAUTO, no MeOH. 0.04 ppm HCHO;
Bottom: 323 ppmC SYNAUTO, no MeOH, 0.06 ppm HCHO;
144
-------
E
Q.
1.0
0.9
0.8
0.7
0.6
0.5
i i i i i
I ' i
Sol id Lines: Data
Dashed Lines: Model
1 I ' I ' ! ' I ' I ' I ' \
August 6, 1984
8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
Lines: Data
Lines: Model
•j-H—,—1—,---1—i—I—i. lili
8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
0.0
Figure 45. NO Wall HCHO Wall Conditions.
CBS model; 0 ppb Wall HCHO, 25 ppb Wall NOx, 5 ppb initial HONO;
normal formation of HONO on walls.
145
-------
E
a
1.0
0.9
0.8
I i
' i ' I '
August 6, 1984 J
Sol id Li nes: Data
Dashed Lines: Model
6 7 8 9 10 11 12 13 14 15 16 17 18
HOURS, EOT
19
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
1.0
0.9
E 0.8
a
a 0.7
| 0.6
x
° 0.5
0)
§>0.4
z 0.3
0.2
0.1
0.0
Sol id Lines: Data
Dashed Lines: Model
Li i_L I
6 7 8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
1.0
0.9
0.8
0.7
0.6 |
0.5 ?
TJ
0.4 ?
0.3
0.2
0.1
0.0
Figure 46. High Wall Conditions.
ALW model; 50 ppb Wall HCHO, 25 ppb Wall NOx, 0 ppb initial HONO;
no formation of HONO on walls.
Top: 2.25 ppmC SYNAUTO, no MeOH, 0.04 ppm HCHO
Bottom: 3.23 ppmC SYNAUTO, no MeOH, 0.06 ppm HCHO
146
-------
1.0
0.9
1 I ' I ' I ' I ' I ' \ ' \ ' i ' I ' I ' I ' I '
Augustjv 1984
Sol id Lines: Data
• Dashed Lines: Model
HN03
»j—f—i--."T i i "f-1—i—h:; i T i
8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
1.0
0.9
E 0.8
a.
^ 0.7
1 0.6
O 0.5
§
°>0.4
i 0.3
0.2
0.1 h
0.0
I ' I
Sol id Lines: Data
Dashed Lines: Model
os
HNOg
PAW
T-t—i—I—i—I—i I T
6 7 8 9 10 11 12 13 14 15 16 17 18
HOURS, EOT
19
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
O
N
§
P
T3
Figure 47. High Wall Conditions.
CBX model; 50 ppb Wall HCHO, 25 ppb Wall NOx, 0 ppb initial HONO;
no formation of HONO on walls.
Top: 2.25 ppmC SYNAUTO, no MeOH, 0.04 ppm HCHO
Bottom: 3.23 ppmC SYNAUTO, no MeOH. 0.06 ppm HCHO
147
-------
1.0
0.9
E 0.8
a
a 0.7
co
3 0.6
x
° 0.5
a>
g>0.4
L_
i 0.3
0.2
0.1
0.0
Sol id Lines: Data
Dashed Lines: Model
NO
August 6r 1984 1
os
:L.4—<--r-r i i i -|—f—i—i—F"; i . i
6 7 8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EDT
1.0
0.9
0.8
0.7
0.6 |
0.5 ?
TJ
0.4 |
0.3
0.2
0.1
0.0
1.0
0.9
E 0.8
^ 0.7
co
^ 0.6
Sol id Lines: Data
Dashed Lines: Model
6 7 8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EDT
0.0
Figure 48. Extra Wall Conditions.
CBX model; 100 ppb Wall HCHO, 25 ppb Wall NOx, 5 ppb initial HONO;
normal formation of HONO on walls.
Top: 2.25 ppmC SYNAUTO, no MeOH. 0.04 ppm HCHO
Bottom: 3.23 ppmC SYNAUTO, no MeOH. 0.06 ppm HCHO
148
-------
1.0
0.9
E 0.8
Q.
a 0.7
I 0.6
x
0 0.5
§
£0.4
i 0.3
0.2
0.1
0.0
I i I ' I ' I 'i 'I i i ' I ' I ' I ' i ' I ; I
August 7, 1984
Solid Lines: Data
Dashed Lines: Model
I
5 6 7 8 9 10 11 12 13 14 15 16 17 18
HOURS, EOT
19
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
1.0
0.9
i i i r
Sol id Lines: Data
Dashed Lines: Mode
I I i i
8 9 10 11 12 13 14 15 16 17 18 19
HOURS. EOT
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
Figure 49. High . Wall Conditions.
ALW model; 50 ppb Wall HCHO, 25 ppb Wall NOx, 0 ppb initial MONO;
no formation of HONO on walls.
Top: 1.32 ppmC SYNAUTO, no MeOH, 0.04 HCHO, no MeNO2
Bottom: 0.86 ppmC SYNAUTO, 0.3 MeOH, 0.06 HCHO, 3 ppb MeNO2
149
-------
1.0
0.9
0.8
I ' I
August 7, 1984 J
Sol id Lines: Doto
Dashed Lines: Model
8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
1.0
0.9
0.8
0.7
E
a
1.0
0.9
0.8
0.7
I 0.6
x
° 0.5
0)
g> 0.4
z 0.3
0.2
0.1
0.0
Soli d Li nes: Data
Dashed Lines: Model
03
Cll'V-k-L-J I I I I I I .
8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
Figure 50. High Wall Conditions.
CBX model; 50 ppb Wall HCHO, 25 ppb Wall NOx, 0 ppb initial HONO;
no formation of HONO on walls.
Top: 1.32 ppmC SYNAUTO, no MeOH. 0.04 HCHO, no MeNO2
Bottom: 0.86 ppmC SYNAUTO, 0.3 MeOH. 0.06 HCHO, 3 ppb MeNO2
150
-------
August 7, 1984
Lines: Doto
Lines: Model
...I—I---—— f-4--f i
8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
0.0
E
a
co
1.0
0.9
0.8
0.7
0.6
x
° 0.5
(0
0)0.4
u
S 0.3
0.2
0.1
0.0
Sol id Li nes: Data
Dashed Lines: Model
03
^JH^^H-.g.. | i | i | i | -i | i |
5 6 7 8 9 10 11 12 13 14 15 16 17 18
HOURS, EOT
19
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
Figure 51. Extra Wail Conditions.
CBX model; 100 ppb Wall HCHO, 25 ppb Wall NOx, 5 ppb initial HONO;
normal formation of HONO on walls.
Top: 1.32 ppmC SYNALTO, no MeOH, 0.04 HCHO. no MeNO2
Bottom: 0.86 ppmC SYNAUTO, 0.3 MeOH, 0.06 HCHO, 3 ppb MeNO2
151
-------
1.0
0.9
E 0.8
a
a 0.7
w"
* 0.6
Sol id Lines: Data
Dashed Lines: Model
i l--r"r~ I 'i 'i r
8, 1984
03
8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
1.0
0.9
0.8
0.7
0.6 |
0.5 *®
T>
0.4 |
0.3
0.2
0.1
0.0
1.0
0.9
E 0.8
a
a 0.7
1 0.6
0 0.5
0)
§> 0.4
i 0.3
0.2
0.1
0.0
T I 1 I I >
' .L.-l-f"1— I""1'
Solid Li nes: Data
Dashed Lines: Model
HN03
PAN
l7'-h-r-)—r--f-r-t—t- I i I
6 7 8 9 10 11 12 13 14 15 16 17 18
HOURS, EOT
19
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
O
N
O
a>
T3
Figure 52. High Wall Conditions.
CBX model; 50 ppb Wall HCHO, 25 ppb Wall NOx, 0 ppb initial HONO;
no formation of HONO on walls.
Top: 2.47 ppmC SYNAUTO, 0.79 MeOH, 0.23 HCHO, 10 ppb MeNO2
Bottom: 3.67 ppmC SYNAUTO, no MeOH, 0.06 HCHO, no MeN02
152
-------
E
1.0
0.9
0.8
I ' I ' I
Soli d Li n«a: Data
Dashed Lines: Model
1 ' I '--I'"1" ! ' ! ' I !
Atigust 8, 1984
03
HN03
i "ri—i—1--V--I-T-i--r i . i
6 7 8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
1.0
0.9
E 0.8
a
°; 0.7
I 0.6
I ' I ' I ' I
Sol id Lines: Data
Dashed Lines: Model
1—|—r IM..J-4—'—F^ I r
03
HMO.
I , T'"i—I--I-H—i—I—.—t—*r,
8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
1.0
0.9
0.8
0.7
0.6 |
0.5 $
•o
0.4 1
0.3
0.2
0.1
0.0
Figure 53. Extra Wall Conditions.
CBX model; 100 ppb Wall HCHO, 25 ppb Wall NOx, 5 ppb initial HONO;
normal formation of HONO on walls.
Top: 2.47 ppmC SYNAUTO, 0.79 MeOH, 0.23 HCHO, 10 ppb MeNO2
Bottom: 3.67 ppmC SYNAUTO, no MeOH, 0.06 HCHO, no MeNO2
153
-------
E
a
1.0
0.9
0.8
0.7
0.6
0 0.5
0)
§>0.4
i 0.3
0.2
0.1
0.0
1 i ' I ' l ' I ' i ' I ' l ' I ' I ' I ' I ' I !
August 8, 1984
Solid Lines: Doto
Dashed Lines: Model
I V--L..., I i I i I , I , I
03
6 7 8 9 10 11 12 13 14 15 16 17 18
HOURS, EOT
19
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
1.0
0.9
E 0.8
a
a 0.7
ui
I 0.6
° 0.5
-------
1.0
0.9
E0.8
a
a0.7
w
.-§0.6
x
°0.5
0)
0)0.4
L.
io.3
0.2
0.1
0.0
1.0
0.9
E0.8
a
a0.7
CO
10.6
°0.5
a>
g>0.4
|0.3
0.2
0.1
0.0
! I ' I > i ' I .'.. I ' I
August 22, 1984
Sol id Lines: Data
Dashed Lines: Model
- WO
LN02
..-r-r"." I . 1"r-+-r-+-i- I
I I I
5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EDT
1.0
0.9
0.8
0.7
0.6 |
0.5 -a
T3
0.4 ^
0.3
0.2
0.1
0.0
i i i r
Sol id Li nes: Data
Dashed Lines: Model
-NO
-NO,
"I—r-l—t I
i i i
5 6 7 8 9 10 11 12 .13 14 15 16 17 18
HOURS, EDT
19
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
Figure 55. High Wall Conditions.
CBX model; 50 ppb Wall HCHO, 25 ppb Wall NOx, 0 ppb initial HONO;
no formation of HONO on •walls; no initial MeNO3 included on BLUE.
Top: 3.04 ppmC SYNURBAN, no MeOH, 0.06 ppm HCHO, no MeNO2
Bottom: 2.04 ppmC SYNURBAN, 0.87 MeOH, 0.13 HCHO, 10 ppb MeNO2
155
-------
1.0
0.9
£0.8
Q.
a0.7
co
10.6
°0.5
a>
g>0.4
L.
io.3
0.2
0.1
0.0
Sol id Li nes: Data
Dashed Lines: Model
-NO
i j \ i i } * I r— j i i i j ; [—
September ,1; 1984:
03
8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
1.0
0.9
E0.8
a
°-0.7
in
30.6
x
°o.5
a>
§>0.4
S0.3
0.2
0.1
0.0
Sol id Li nes: Data
Dashed Lines: Model
::ii-r-r"i r*i-i—,—!--,—i—..-.i PAN
8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
Figure 56. High Wall Conditions.
CBX model; 50 ppb Wall HCHO, 25 ppb Wall NOx, 0 ppb initial HOMO;
no formation of HONO on walls.
Top: 2.0 ppmC SYNURBAN, 1.0 ppm MeOH
Bottom: 3.0 ppmC SYNURBAN, no MeOH
156
-------
Conclusions
This project clearly met its goals in terms of producing quality experiments designed
to address the issue of methanol-exhaust reactivity:
1. Twenty-nine dual smog chamber runs were conducted. Eighteen of these exper-
iments are satisfactory for model testing and fourteen are excellent, exceeding
the estimates of the yield of good runs made in the planning memorandum. The
other 11 experiments, while having poorer sunlight, which complicates model
testing, are still quite useful to support the trends or directional effects of the
substitution process.
2. Three different hydrocarbon mixtures were used:
o SynAuto—a 13-component mixture developed by a series of direct compar-
isons of the reactivity of the mixture with automobile exhaust in side-by-side
chamber experiments;
o SynUrban—an 18-component mixture that conforms with the EPA recom-
mended "default" mixture composition for use with the Carbon Bond Model
in urban ozone control calculations; and
o SynAutUrb—a 14 component mixture of intermediate reactivity between SynAuto
and SynUrban.
3. Fifteen dual experiments were conducted with the SynAuto mixture; eight dual
experiments were conducted with the SynUrban mixture; four dual experiments
were conducted with the SynAutUrb mixture. (In addition to these main exper-
iments, there was one experiment with UNCMIX and one experiment with only
the aromatic portion of the SynAuto/SynUrban mixture.)
The major conclusions that can be drawn from this study are:
157
-------
Conclusions
» Synthetic methanol exhaust substitution in these experiments never resulted in
an increase in ozone maximum or a shorting of time to events over that of the
reference side, even for a SynMethanol mixture with 20% formaldehyde.
• In experiments with 100% synthetic auto-exhaust with 3 ppmC on one side
and 2 pprnC on the other side of the chamber (i.e. a 1/3 reduction), a 3-6%
difference in the peak ozone was observed. In addition, the 2 ppmC side was
slower in producing the ozone by about 60 minutes.
> In the 3 ppmC. 9-to-l HC-to-MOx ratio experiments with 100% synthetic auto-
exhaust on one side and 2/3 synthetic auto-exhaust,1/3 synthetic methanol ex-
haust (with variable amounts of formaldehyde) on the other side of the chamber,
no difference was seen between the two chamber sides for peak O3 production.
That is, the peak ozone in these experiments was essentially independent of
the formaldehyde content of the synthetic methanol exhaust. This was because
these system were limited by the amount of nitrogen oxides available, not by
the amount or reactivity of the organic reactants. In these experiments, how-
ever, the systems were slower to produce ozone as formaldehyde in the synthetic
methanol exhaust was decreased from 20% (almost no delay) to 0% (about 60
minutes delay).
In the 1 ppmC, 3-to-l HC-to-NOx ratio experiments with synthetic auto-exhaust,
a 32% reduction in peak ozone occurred when synthetic methanol exhaust con-
taining 10% formaldehyde was substituted for 1/3 of the carbon in the synthetic
auto-exhaust mixture.
In the 3 ppmC. 9-to-l HC-to-NOx ratio experiments with 100% of the much less
reactive synthetic urban mixture on one side of the chamber and with 2/3 of
the synthetic urban, 1/3 synthetic methanol exhaust (with variable amounts of
formaldehyde) on the other side of the chamber, no difference in peak ozone
production or in time to events was seen between the two sides at the 20%
formaldehyde level. At the 0% formaldehyde level, however, there was an 15%
decrease in ozone maximum when methanol was substituted for synthetic ur-
ban mixture carbon. As in the synthetic auto-exhaust case, substitution of
100% MeOH for 1/3 of the carbon, resulted in essentially the same outcome as a
reduction of 33% of the total carbon.
In the 1 ppmC vs 0.66 ppmC experiments with 100% of the synthetic urban
mixture, an 83% difference in peak ozone was observed.
In the 1 ppmC, 3-to-l HC-to-NOx ratio experiments with 100% of the synthetic
urban mixture on one side of the chamber and with 2/3 of the synthetic urban,
1/3 synthetic methanol exhaust (with variable amounts of formaldehyde) on the
other side of the chamber, no difference in peak ozone production or in time to
158
-------
Conclusions
events was seen between the two sides at the 20% formaldehyde level. At the
0% formaldehyde level there was a 50% reduction in peak ozone and at the 10%
formaldehyde level there was a 30% reduction in peal ozone.
• The initial fraction of formaldehyde is the major factor in the reactivity differ-
ences seen in the experimental data.
Table 20 summarizes the relative ozone and time-to-events in the principal ex-
periments.
The smog chamber conditions were designed to test models, not to duplicate the
urban atmosphere. Large dilution and continuous injection are dominant factors in
the urban atmosphere and neither of these processes were included in these exper-
iments. Readers are cautioned about simple extrapolation of the results reported
here to the urban atmosphere.
A limited amount demonstration model was conducted to show the utility of the
data for model testing. General observations about the model performance are:
• Three photochemical mechanisms often used in control calculations when used
to simulate both the synthetic auto-exhaust and the synthetic urban exeriments
at different HC-to-NOx-ratios, were able to reproduce the general trends in se-
lected experiments. The absolute predictions of time to events and magnitudes
of concentrations of secondary products such as ozone, nitric acid, PAN, and
aldehydes, however, were poor for all models when using the usual assumptions
of chamber characteristics. The newest model was used to simulate selected syn-
thetic urban mixture experiments; its performance differed significantly between
simulations of the synthetic auto-exhaust and the synthetic urban mixtures. It
was generally too slow for the auto- exhaust mixture experiments and generally
too reactive for the urban mixture experiments.
• The large majority of all the model simulation results were slower than the
experiments, even with extreme assumptions of smog chamber wall processes.
Increasing the initial reactivity artificially by the use of extremely high wall
assumptions and radical sources, causes the models to agree with the data better
in the first part of the experiment, but then when the mixture composition is
changed these models greatly overpredict the ozone produced in the experiment.
• The relative responses of the models to substitution of methanol are in approx-
imate agreement with the data. For example, if the model plots are laid over
the data plots and slid to the left by 40 to 100 minutes, the model-data agree-
ments for ozone and oxides of nitrogen are very good for the 1 ppmC synthetic
auto-exhaust experiments.
159
-------
Conclusions
Until the mechanisms are tested with smog chamber data designed to test for
the correct representation of the various components in the mechanisms, it is
not possible to determine with certainty the cause of the differences between
the model simulations and the chamber experiments described here. Further
testing of these models is recommended.
160
-------
Conclusions
Table 20. Summary of Relative Reactivity For Methanol Substitution.
Change Relative to
100% Syn Mixture
Date
Run Type
% O3 red. timing
3 ppmC (9:1 HC-to-NOx)
AU06 2/3 SynAuto
2/3 SynAuto, 1/3 SynMeth
JU26 0% HCHO
JU25 10% HCHO
AU08 20% HCHO
JN26 2/3 SynUrban
2/3 SynUrban. 1/3 SynMeth
ST01 0% HCHO
10% HCHO
AU22 20% HCHO
1 ppmC (3:1 HC-to-NOx)
AU05 2/3 SynAuto
2/3 SynAuto, 1/3 SynMeth
0% HCHO
AU07 10% HCHO
AU09 20% HCHO
ST02 2/3 SynUrban
2/3 SynUrban, 1/3 SynMeth
AU28 0% HCHO
AU25 10% HCHO
ST03 20% HCHO
0
0
0
20
15
0
42
slower
slower
slow
slower
slower
slower
32 slow
38* «same
83 slower
50 slow
30 same
0 slow*
these days not clear sky conditions
161
-------
References
1 Alson J., Baines, T.M.,"Emissions and Energy Efficiency Characteristics of Methanol-Fueled
Engines and Vechicles," presented at the Institute of Gas Technology's Nonpetroleum Vehic-
ular Fuels, Arlington, Va., Oct. 14, 1982.
2 O'Toole, R. "California Methanol Assessment, Vol. II: Technical Report" JPL Publication
83-18, March 1983.
3 Pefley R.K., Pullman B.; Whitten G., "The Impact, of Alcohol Fuels on Urban Air Pollution:
Methanol Photochemistry Study," Final Report DOE/CE/50036-1, Department of Energy,
Washington D.C. November 1984.
4 Whitten G., Hugo H. "Impact of Methanol on Smog: A Preliminary Estimate," Prepared for
ARCO Petroleum Products Co. by System Applications Inc., San Rafael, CA, 1983.
5 Jeffries H.E., Kamens R.M., Sexton K.G., and Gerhardt A.A., "Outdoor Smog Chamber Ex-
periments To Test Photochemical Models", Final Report, Environmental Protection Agency
Co-operative Agreement 805843, 1982; NTIS No. PB 82-198 508.
6 Whitten G., Killus J., Hugo H., "Modeling of Simulated Photochemical Smog with Kinetic
Mechanism. Vol. 1, Final Report," EPA-600/3-80-28a, U.S. Environmental Protection Agency,
Research Triangle Park NC, 1980.
7 Jeffries, H.E. "Modeling Observed Hydrocarbon Mixture Reactivity Effects", Technical Nar-
rative, EPA Grant Request 807762, 1981
8 Atkinson R., Lloyd A., Winges, L., "An Updated Chemical Mechanism for Hydrocarbon/
NOx/SOo Photooxidations." Atmos. Envr., 16. 1341-1355, 1982.
9 Jeffries H.E., Sexton K.G.. Morris, T.P., Jackson M., Goodman R.G.. Kamens R.M., Holle-
man M.S, "Outdoor Smog Chamber Experiments Using Automobile Exhaust", Final Re-
port (EPA/600/S3-85/032), Environmental Protection Agency, Research Triangle Park, N.C.,
June 1985; NTIS No. PB 85-191 708/AS.
10 Jeffries H.E., Sexton K.G., Kamens R.M.. Holleman M.S., "Outdoor Smog Chamber Ex-
periments to Test Photochemical Models: Phase II," Final Report (EPA/600/S3-85/029),
Environmental Protection Agency. Research Triangle Park. N.C., June 1985; NTIS No. PB
85-191 542/AS.
11 Kopczynski S. et al., "Photochemistry of Atmospheric Samples in Los Angeles," Enviro.
Sci. Technol., 6, 342, 1972.
12 Killus J., Whitten G., "Technical Discussion Relating to the Use of the Carbon Bond Mech-
anism in OZIPM/EKMA," EPA-450/4-84-009, U.S. Environmental Protection Agency, Re-
search Triangle Park, North Carolina, 1984.
13 Gipson G.. Freas W., "An Analysis of Organic Species Data Collected in the Northeast Cor-
ridor Regional Modeling Project." U.S. Environmental Protection Agency, Research Triangle
Park, North Carolina, presented at 77th APCA meeting, San Francisco, CA, 1984.
14 Jeffries H.E., Fox D.L., Kamens R.M, "Outdoor Smog Chamber Studies: Effects of Hydro-
carbon Reduction on Oxides of Nitrogen," EPA-650/3-75-011, 1975.
15 Jeffries H.E., Sexton K.G., "Outdoor Smog Chamber Experiments: Reactivity of Methanol
Exhaust Vol. II, Data Processing and Quality Assurance System Description," Final Report,
Sept. 1985.
16 Sigsby J.E., Tejada S., and Ray W.,"Volatile Organic Compound Emissions from 46 In-use
Passenger Cars," Mobile Sources Emissions Research Branch, U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina, 1985.
17 Jeffries H.E., Sexton K.G."Background Chamber Reactivity"
18 Jeffries H.E.,"A Photochemical Reaction Mechanism Analysis Method Applied to Two Mech-
anism," submitted Envr.Sci. and Tech., 1985
162
-------
Facility Description
Chamber Description
Location
The site is approximately 32 kilometers from the University of North Carolina at
Chapel Hill. It is in Chatham County, North Carolina and is approximately 10
kilometers from the small town of Pittsboro. Chatham County is one of the most
rural, least industralized counties in North Carolina and is heavily wooded. The
background concentrations of NOX and nonmethane hydrocarbons are usually less
than 5 ppb and less than 80 ppbC. More importantly, the air exhibits very low
reactivity in the chamber.
Materials
The chamber surfaces are Fluorinated Ethylene Propylene (FEP) Teflon. The film's
transmission in the UV and visible regions of the solar spectrum is excellent, and
it has only a few absorption bands in the IR, a property necessary to reduce the
"greenhouse effect". It has a very low permeability for most chemical species and
can be heat-sealed to form large durable panels. For this application its worst
property is its ability to hold a static charge for long periods of time. Type A film,
0.13 mm thick is used.
Physical Design
Inlet and outlet doors, stirring fans, manifolds and other fittings come in through
a solid floor. The sides are free for light entry. The floor of the chamber is elevated
approximately 1.2 m to allow for easy access under the chamber.
163
-------
Facility Description Orientation
The design is an A-frame 9.14 m wide, 12.10 m long and 6.10 m high at the
peak on a plywood floor 1.22 m above the ground. Wooden beams, 5.08 cm by
20.32 cm, located on edge at 99.1 cm centers form an exterior framework on top of
the plywood floor. Continuous 16.46 m lengths of film are attached to the inside by
aluminum u-channels, screwed firmly to the wooden beams, thus compressing the
film against the external support. See Figure Al and Figure A2.
One piece, heat sealed Teflon film panels are used in such places as the triangular
end panels and floors. The floor-to-side seals are achieved by a 0.61 m overlap of side
panel film over floor panel film. A rubber strip under the film and an aluminum strip
over the top of the film complete the seal. All other seals are Teflon-to-Teflon under
pressure of the aluminum u channel. A single unsupported heat sealed Teflon panel
similar to the end panels is used to separate the chamber into two halves of equal
volume. It is sealed to the floor and side panels in the same manner as described
above. Aluminum foil is placed under the film on the floor to reflect the light and
heat back up through the chamber. This is_ necessary to reduce solar heating of
the air to a value that is within normal urban environments and to compensate for
transmission losses through the Teflon film.
Orientation
The chamber is oriented with its long axis approximately north to south. The actual
long axis orientation is along a 27-207 true heading. The orientation with respect
to the sunrise and sunset at different times of the year and the altitude of the sun
at noon for different times of the year are illustrated in Figure A3 and Figure A4.
Air Handling System
There are four air handling systems in each half of the chamber, one for exhausting,
one for sampling, one for recirculating through dehumidifiers, and one for mixing.
The exhaust system consists of two intake stacks, 0.61 m x 0.91 m intake doors.
0.61 m x 0.61 m exhaust doors, and an exhaust blower. The exhaust blower is
a dual blower on a single shaft driven by a 1.5 horsepower motor. Air enters the
system through the two 5.49 m high by 30.5 cm diameter stacks. This system
is designed to permit rapid exhausting of chamber contents and replacement with
ambient air. The filling rate is 7190 1/min. The chamber can be flushed with a
99.6% decay of initial contents in 2 hours.
The second air handling or manifold system is for sampling and injection of
pollutant materials into the chamber halves. To insure representative sampling, a
3.17 cm I.D. glass manifold runs from a point 1.83 m above the floor in the center
164
-------
Figure Al. The University of North Carolina Smog Chamber. (Courtesy of the Greensboro Daily News)
/
-------
END
SIDE
BLUE RED
(A)
IB)
STRUCTURE OF CHAMBERS
EXHAUST DOORS-
(0
(D)
\I
V O
/ffl \!l/
INTAKE/
DOOR
FLOOR PLAN OF CHAMBERS
so1
VINTAKE
DOOR
SAMPLING LINES
MANIFOLD
Figure A2. Schematic of UNC Outdoor Smog Chamber.
-------
W
JUNE 22
MAY2I.JW.T23
A>*H. 20.AUO 2«
MAR2',!Ef>T 23
MN2I.NOV23
DEC 22
Figure A3. Orientation of UNO Outdoor Smog Chamber with respect to seasonal sunrise and sunset
positions.
Outdoor Dual
Smog Chamber
Figure A4. Solar altitude and zenith angle at noon at the UNC Outdoor Smog Chamber for each
month.
167
-------
Facility Description Laboratory
of each chamber half down through the floor and over to a sampling laboratory.
The sampling volume required by all the instruments does not exceed 5 1pm but
to reduce losses due to long resident times it is necessary to have high flows in
the manifold. The flow rate in the manifold is 60 1pm. The manifold system is
wrapped with a controlled heating tape to maintain the sample slightly warmer
than the chamber temperature. To avoid the necessity of makeup air. the sampling
manifold is a closed loop. Squirrel cage blowers with housings and fans that are
Teflon coated are used to circulate air through the manifolds. The unused sampler
air is then returned through a 3.17 cm I.D. glass manifold to the chamber. These
return manifolds provide a convenient method for injecting the initial reactants.
There is also a special heated Teflon manifold for formaldehyde sampling.
Inside each chamber half are two mixing fans located in opposite corners. These
provide circulation and mixing of the chamber contents. The fans are 50.8 cm
diameter cast aluminum units that are FEP Teflon coated. They operate in a
horizontal position. 0.76 m above the floor on 2.54 cm diameter Teflon coated steel
shafts that extend through the chamber floor. Under the floor. 1/4 horsepower,
1750 rpm motors provide power through a belt and pulley system to each fan. Each
fan operates at approximately 31.15 m3/min.
Laboratory
The laboratory is adjacent to the chamber. It is a 3.66 m W by 15.24 m L by 3.05 m
H temperature-controlled wooden structure oriented perpendicular to the chamber
and 3.66 m away from it to avoid any shadowing.
The first 5.49 m nearest the chamber contain the instrumentation, manifolds,
and calibration systems. At the sample inlet of each gas instrument is a three
way Teflon AC solenoid valve. Since there are two intake manifolds (one for each
chamber half), air from either manifold can be drawn through the three-way valve
and into the instrument. In this manner timesharing of one instrument between the
two chamber halves is possible.
The next 5.49 m contain the data acquisition computer system .(described in
the Data Acquisition System section) and the operations area. The last 4.27 m of
the laboratory is a utility area with running water and storage facilities.
Gas tanks necessary to operate the instruments and perform calibration are
located in a 1.22 m x 1.83 m room completely closed off from, but attached to,
the laboratory. The injection system gas tanks and valves are housed in a second
168
-------
Data Acquisition System Facility Description
1.52 m x 1.52 m well-insulated heated room adjacent to the end of the laboratory
nearest the chamber.
Injection System
Pollutants are injected into the chamber sides via the return side of the sampling
manifolds. The return manifolds enter the chamber sides under one of the mixing
fans. The injection process uses gas cylinders containing pollutants at high concen-
tration (1000-10,000 ppm range), two-stage stainless steel diaphragm regulators,
on-off solenoid valves, and a precision needle valve. The flow rate of injected ma-
terial into the manifolds is established by a mass flow meter with a 5 millisecond
response time. The total injection volume is accurately controlled as a function
of the time the solenoid valve is open. The open time of each solenoid valve is
controlled by the computer system on command. Conditions can be varied suffi-
ciently to have the injection time range from a few seconds for each component to
1-2 hours for a programmed injection used to simulate the buildup of pollutants in
urban areas.
Data Analysis, Validation and Reporting
Data Acquisition System
A computer based data acquisition and control system (DAS) is used to acquire,
process and record data for the chamber instrument system. The complete system
is shown schematically in Figure A5.
Output signals from each instrument are wired to a crossbar scanner (or analog
signal multiplexer). Under control of programs in the computer, the scanner con-
nects the selected signal leads to the input of the digital voltmeter. The digital volt-
meter, which has excellent noise and spurious signal suppression and can measure
a 1 volt signal with a resolution of 10 microvolts, is triggered to acquire a reading
and to supply the 5-1/2 digit binary coded decimal number to the computer. The
computer processes the information and then commands the scanner to move to the
next channel and repeat the process. Further data processing is described below in
the Data Treatment Procedures Section. The timing of the computer processes is
under the control of a digital date and time-of-day clock which signals event times
to the computer. Another digital timer operates the Teflon solenoid valve on the
inlet of each instrument to connect the instruments to one chamber half during a
given cycle of data acquisition.
This system provides fully automatic acquisitions and processing of data during
a run and provides the operator with immediate, full data in physical units. Given
169
-------
•t' ** \ Vi <\
>l -J
Figure A5. Data Acquisition and Control System.
-------
General Facility Description
this information about what is happening, the operator can then concentrate on
what he wants to do. It also allows the massive amount of information generated
during a run to be processed in a more effective manner after the run is over.
Standard Operating Procedure
Most experiments are initiated early in the morning by the site computer accord-
ing to a detailed set of commands that have been programmed the night before.
Prior to an experimental run the computer is programmed to automatically purge
the chamber overnight with background ambient air. Several hours before dawn
the DAS begins recording measurements of the background concentration. Single
instruments are timeshared on an alternating four minute cycle between the two
sides. The chamber intake and exhaust doors are then sealed by the computer and
the drying fans in each side of the chamber activated to reduce humidity and mini-
mize chamber wall effects. Finally, the initial concentrations of reactants of interest
are injected before sunrise. To do this, the computer selects the proper tank and
calculates the correct time to achieve the desired initial concentrations. Injections
are normally short "slug" injections; most experiments have initial hydrocarbon,
NO, and N(>2. Monitoring of the chemical and physical species mentioned earlier
continues usually until 1900 EDT when the experiment is terminated. The com-
puter then performs several close-down procedures, including turning off certain
instruments, opening chamber vent doors, and turning on the exhaust fan to purge
the chamber. If instructions for the next experiment have not been programmed
during the day, the operator does so at this time.
Variations of the standard operating procedures include: 1) not changing hu-
midity before the experiment, 2) injecting pollutants slowly during the experiment,
3) diluting the chamber(s) during the experiment to simulate mixing height profiles,
4) transferring the contents of one chamber side to the other, and 5) injecting initial
conditions into the chamber containing part or all of the product of a day-old exper-
iment. Most pollutants are injected from high concentration gas cylinders (10.000
ppm) which can be controlled by the computer. Some pollutants are liquids with
vapor pressure too low to have a high vapor concentration cylinder prepared, and
are injected manually. Carefully measured samples are heated in a U-tube in which
warm zero air is passed through to the return manifold.
Data Treatment Procedures
General
The operations are carried out on four different computers and the Broomall plotter.
171
-------
Facility Description During A Run
The PDF 11/40 is at the chamber site and is responsible for acquiring the data.
The LSI/11, the VAX, the IBM 370/360. and the Broomall plotter are on campus
grounds, and complete the data treatment procedure after a run.
During A Run
DATCOL and DATRAN are written in assembly language and PASCAL. DATCOL
performs several functions, such as accepting instructions from the keyboard and
interpreting a command file which prepares the chamber, makes injections, turns
on the necessary instruments, and finally purges the chamber and turns off the
instrumentation at the end of the experiment. Command files can be written in an
easily readable English text style. An example is shown in Figure A6. DATRAN
translates this file into a format which DATCOL can interpret. DATCOL also
controls the data acquisition process, and records:
1) the instrument analog outputs;
2) the GC files produced by the Perkin Elmer Sigma 10 GC integrator and pro-
cessor;
3) the adjustment file (computer readable instructions for data adjustment); and
4) the comments files,
all on the RK05 hard disk. It also maintains status information to restart and
continue an experiment after a power failure.
The operator has several choices for printing data with DATCOL. He can have
the data printed during the whole experiment, print data for only selected times
during the experiment, have earlier data printed, or not print data at all.
The operator can also, at any time during the experiment, enter through the
keyboard information such as calibration corrections for any instrument or that an
instrument is not functional and for what time interval this information applies.
This type of information is entered by DATCOL into an Adjustment file. This
file can be read and interpreted by other computers and can be directly applied to
correcting and processing data in later steps. The operator can at any time also
enter general comments through the keyboard which DATCOL stores in a Comment
file. Such information, when transferred along with the digital data, greatly aids in
processing.
Strip charts are operated for almost every instrument as a back-up to the com-
puter.
172
-------
RUN DXI:DATRAN
SPECIFY OPTIONS DESIRED (AS optionroption,option : l-FEB-80
ERRORS ARE FLAGGED BY CNN3
CTHIS IS AN ACTUAL EXAMPLE RUN3
IMMED VENT; CLR CHARTS;
AT 3:30 SET CHARTS? OSC 35 ON 45 SEC EVERY 15 MIN? CCO:
AT 3:50 CLOSE;
AT 3:59 SET 6? CCARLE 1 AND 23
AT 4:00 UNTIL 19:20 START DVM?
AT 5:00 INJECT NO 13 TO 400 PPBC IN BOTH PURGE?
AT 5:iO INJECT UNCMIX 10 TO 4 PPMC IN BOTH PURGE?
GAS
NO
N02
PROPYL
UNCMIX
ETHYL1
N2
AUTOMX
ACETAL
BUTAN1
BUTPPL
CO
PROPAN
ETHYL2
BUTAN2
CONC(PPMC)
10700.
20500.
21396.
10471.
2028.
1000000.
10487.
4110.
92400.
11456.
100000.
22068.
20663.
20760.
C 13
C 23
WARNING- INJECTION TOO LONG,
WARNING- INJECTION TOO LONGf
BEING SPLIT
BEING SPLIT
AT 5:30 INJECT N02 12 TO 100 PPBC IN BOTH PURGE?
AT 19120 VENT? HALT?
RSTART 0:00 VENT? CLR CHARTS?
RSTART 3:30 SET CHARTS? CLOSE?
RSTART 3:30 MOD 1 OSC 35 ON 45 SEC EVERY 15 MIN?
RSTART 4*,00 START DVM?
Figure A6. Sample DATCOL Command File.
173
-------
Facility Description Data Processing Steps
Data Processing Steps
Data processing involves three types of staff
c Project Coordinator (PC)—senior staff, makes judgements about data qual-
ity, decides on calibration factors, gives directions to other data processing staff;
uses computer run tracking system to direct attention of PPs and to CTs to
work that is needed by indicating the status of processing in the run tracking
data base;
c Peak Peaker (PP)—seasoned staff, skilled at converting strip chart peaks
into computer files using a high-resolution electronic digitizer pad connected to
a computer; throughly familiar with the data output of each chromatograph;
makes plots of digitized data for QA by PC: corrects picking mistakes by exam-
ining the raw data plots; basically works independently using computer reports
from run tracking system to indicate needed work; updates run tracking system
to move work to next stage;
c Computer Technician (CT)—staff, highly skilled at use of various computer
systems (LSI-11/23, IBM PC, VAX-11/780, IBM 4341), with a through knowl-
edge of file transfer and storage among these systems: responsible for running
programs that read and convert site floppies into input for other processing pro-
grams: responsible for creating, naming, and backing up to magnetic tape all
data files on all systems: in addition, the CT is usually an apprentice processing
program developer and maintainer.
Data processing is organized by the type of data. The major processing effort oc-
curs for the chromatographic type instruments. Data processing for the continuous
instruments {DVM data) is less difficult than the chromatographic instruments, be-
cause the data come from the site in digital form. Both of these processing streams,
however, dep>end upon having calibration factors to convert the voltages and raw
displacements into physical units of concentration.
The "auto-cals" can be automatically stripped from the DVM data by the CT
who moves the data from the site floppies. These stripped cals can be processed
rather quickly by computer programs once the calibration source concentrations are
known. The chromatographic calibration data, however, must be examined by a
PC, picked by a PP, and stored by a CT to await the development of calibration
source concentrations. Furthermore, each instrument may have 10 specie calibration
factors.
Table Al illustrates, in a flowchart, the stages in processing the GC instrument
data. The flow of processing in these charts is from left to right. The three letter
174
-------
Data Processing Step? Facility Description
codes describe events in processing the data and the lines indicate which events
depend on other events having occurred. As the processing events occur, the three-
letter codes are entered into the on-line run processing tracking system by whoever
did the step. Simple commands allow the PC, PP, and CT to create appropriate
printed reports showing which runs and which instruments are at which stages of
processing.
Table A2 similarly shows the flowchart for processing DVM data. Some of the
processing steps for the DVM "auto-cal" data occurs on this chart also. The same
program that makes the "Raw Plot" also strips the "auto-cals" into a separate file
(the ACS-Auto Cals Stripped—event on the chart). These data are then processed
by a program that results in a data file of instrument span responses and instrument
zero values for the 0$ and NOX monitors. An example list of these will be shown in
the next section.
Once the actual run concentrations are known for both the HC and the NOX, the
run documentation, to be included in the combined data file, can be written. Here
all aspects of the run are drawn together to provide a description of the run. its
quality, problems, and outcome.
Table A3 illustrates the events in the documentation and in the creation of a
single segmented file containing all the data for the run in a uniform format.
175
-------
Table Al.
Processing System for Instrument Data
INTB!
/
tot inst
\
!PID!
\
cal processing.
\
SCFD!
__
Pers
\
\
!CQA! ---- (F2V) ---- (V2U)
---- — !PQA!
!sss! == stages determined by Project coordinator (PC)
== stages determined by Peak Pickers (PP)
(sss) == stages determined by Computer Techs (CT)
RUNENTBY stages
Meaning
PC
PC
PC
PP
PP
PC
PP
PP
PC
CT
CT
PC
PID
NTB
CFD
PRU
FPT
PQA
CRU
HBU
CQA
F2V
V2U
BAD
Pick inst det
Not to be proc
Cal fac det
Pick run
FPlot done
FPlot QA
CALCON run
HCANAL run
Comp CS QA
File 2 VAX
Ready 2 merge
Stop processing
Charts marked, Inst status sheet in folder
No futher processing marked on Inst Status
sheet
Cal factors for each comp entered on Inst
status sheet in folder
Data has been digitized and P-file exists
P-file has been plotted and plot is in folder
Fast plot has been marked OK (no bad data
points in P-file)
C and K files exist on floppies
HC analysis printout in run folder
Instrument QA completed on C and K files
Initial Conditions updated
C tmd K files moved to VAX
C files merged together, K files merged
together
Something is wrong with this data
176
-------
Table A2.
Processing System for DVM Data
cal processing
/ \
tot runs (CFS)
\ / \
\ (ACS) \
\ / \
(URU) (BPC) !PQA!---!CQA!--!B2M!
\ /
CUTV) /
\ /
(V2T)
!sss! == stages determined by Project coordinator (PC)
== stages determined by Peak Pickers (PP)
(sss) == stages determined by Computer Techs (CT)
Pers
CT
CT
CT
CT
PC
CT
PC
CT
PC
CT
PC
PC
CT
DVM data processing steps
Step Meaning
UBU
U2V
V2T
BPC
PQA
ACS
CFS
DBU
CQA
B2M
NTB
BAD
NDV
Unpack run
Data to VAX
to archive tape
Raw Plot done
Plot QA
AutoCals strip
Cals done
DVMFIX run
Cone QA
Beady 2 merge
Not to be proc
Stop processing
No DVM data
Site floppy data expanded to ASCII Ufiles
Ufiles moved from LSI disk to VAX
Data moved from VAX to tape
Plot of voltages to examine data
Initial quality check
Cals separated for processing
Calibration factors determined
Voltages changed to concentrations
Concentration Quality Assurance
Merge process on VAX can be run
Not to be processed
Something is wrong with this data
177
-------
Table A3.
Documentation and Final File Production Steps
Pers
PC
CT
PC
DOCUMENTATION data processing steps
Step Meaning
SUM : Run sum done
D2V : Doc to VAX
DQA : Doc QA
DVM, Instrument, and Raw Quality Combined.
Documentation transferred to VAX
Documentation Quality Assurance
FINAL FILE data processing steps
Pers
Step
Meaning
CT
CT
CT
PC
PC
PC
MRS
ALT
FPC
FQA
NTB
BAD
Merged 3 sect
Files to tape
Final plot done
Final QA
Not to be proc
Stop processing
Merge pieces to final file
Ascii labeled tape made
Correct cone plot made
Final Quality Assurance
Something is wrong with this
data
178
-------
GC Calibration Processing Facility Description
GC Calibration Processing
The GC instrument calibration processing is complex enough to require its own
tracking system. Table A4 gives the flowchart for these calibration data.
A major time-consuming task on this chart is the Official Calibration Source
(ocs) event. This event signals the existence of an DCS form that names a source
and describes the concentration of each specie in the source. These paper forms
are the results of the manual calibration comparisons described above. Filling out
the form requires extensive analysis and reconciliation to arrive at a certifiable set
of concentrations in the calibration sources. The OCS form requires the validation
date and name of the individual certifying the calibration source.
These OCS values are entered into another data base and appropriate S-files
(source concentration) files are created. These are used as input to the CALFAC
program to match the instrument responses (stored in R-files by the PP) with
the calibration sources to compute the calibration factors for each species. These
calibration factors are, in turn, entered into another data base and its output is a file
of calibration factors that are subjected to graphical and statistical analysis by the
CALANA (calibration analysis) program. Using these plots and statistics, and other
information such as the manual injections, and previous calibrations calculated for
runs nearby, the PC selects a calibration factor for each instrument-species.
The actual calibration factors used in processing the runs are entered into an-
other data base and reports of these are available for plotting and further analysis.
The actual factors are also included in the automatic program documentation p- o-
duced by the DVMFIX and CALCON programs (K-files) and are included as a
permanent part of the documentation section of the segmented file.
179
-------
Table 4
Processing System for Instrument Auto-Calibration Data
tot inst cal analysis
Step
ICPI!
\
\
IOCS!
\
\
-
!sss!
(sss)
\
-~ (Q2V) ~ (DTE) ~ (C2L) — ~ ! CQA!
stages determined by Project coordinator (PC)
stages determined by Peak Pickers (PP)
stages determined by Computer Techs (CT)
CALENTRY stages
Meaning
PC
PP
PC
PP
CT
CT
CT
PP
PC
PC
PC
CPI
DPR
DCS
CFR
Q2V
DTR
C2L
CLR
CQA
NTB
BAD
CAL pick Inst
Digpik run
DCS ok
CALFAC run
Qfile to vax
DTR file updated
Cals to LSI
CALANA run
Cal QA
Not to be proc
Stop processing
Charts marked, cal status sheet in folder
Cal data digitized and R-f ile exists
Official Cal Source exists in DTR
Cal Factor computed, Q-file created
Q-file moved to VAX
Cal Data added to DTR data base
DTR CAL file loaded on LSI-11
Cal Factors analyzed
Cal Factor Quality Assurance
Not to be processed
Something is wrong with this run
180
-------
Analytical Methods
Introduction
The analytical instrumentation in the gas laboratory adjacent to the large out-
door chamber at UNC includes: six packed column and two capillary gas chro-
matographs for hydrocarbon, organic nitrate, and inert tracer analysis; one liquid
chromatograph for DNPH-aldehyde analysis; an automated wet chemical formalde-
hyde instrument; two ozone and NOX chemiluminescent monitors; and one UV ozone
photometer. The laboratory also contains all of the associated calibration systems
used for these instruments. In addition, solar and UV radiation sensors are mounted
on top of a 7 meter tower next to the laboratory. Chamber temperature measure-
ments are made with an inline thermistor located in the inlet of each chamber
sampling manifold. The manufacture and calibration methods for all of these sys-
tems are listed in Table Bl and Table B2. A list of the species measured on each
site instrument is given in Table B3.
Hydrocarbons
Total hydrocarbons (THC), methane (CH4) and carbon monoxide (CO) are analyzed
with a Beckman 6800 gas chromatograph. Figure Bl shows a typical Beckman 6800
measurement cycle. This cycle occurs every 5 minutes and each side of the chamber
is sampled for 30 minutes. Unfortunately, methane responds differently than non-
methane HC on this type instrument requiring a separate methane calibration as
well as a NMHC calibration.
A manually operated six port Carle mini-valve (Carle Instruments, Inc., Ana-
heim, Calif. Model Mk. II, Cat. No. 5621) on a Carle 211 gas chromatograph (desig-
181
-------
Sub.tance
Method
Table Bl. Analytical Methods, Characteristics, and Operation Methods
Manufaci and Model MDC Range Mode
Time Shared
O3
NO. NO», NO2
HC'« alkenei.alkuie.,
aromatic.
THC, CO, CH4
HCHO
Aeetaldehyde,
Chemilumine.cence
Chemilamineicence
Automated multiple column GC-
FID
Automated multiple column GC-
F1D
Automated Colorimetic(rever»e
Weit-Gaeke)
Automated multiple column GC-
Bendix 8002
Bendix 8101-B
Carle 211
Beckman 6800
CEA 666
Carle 211 M
0.001
0.006
0.006-0.010
0.010
0.01
0.02-0.06
0-2
0-1
0-100
0-80
0-2
0-100
Continuou.
Cyclic, 4 min
Cyclic, 16 min
Cyclic, 16 min
Continaout
Cyclic, 80 min
Ye«
Ye.
Ye.
Ye.
Ye.
Ye.
propionaldehyde, MErFID
Aromatic. HC,
aromatic aldehyde.,
phenol.
GC-FID
GC-EC variable freq. pulled
bai. detector
PAN and Alkyl nitrateGC-EC DC ba.ed EC Model
HNOS
H202
640
Perkin-Elmer 600
Analog Tech. Corp.
Varian Aerograph
Chemilumiueicence, molybdenumBendix 8101-B
converter, nylon .crabber
.ample taken manually with
0.010
0.0004 '
0.0003-0.0008
0.003
0.0001
0-100
0-1
0-1
0-1
0-1
Manual
Cyclic, 30 min
Cyclic, 30 min
Cyclic, 16 min
Manual .
Ye.
Ye.
Ye.
No
Ye.
bubbler; aualy.ii u.ing automated
Temperature
TSR
UV
De-w Point
chemilum. method
Thermistor
Pyranometer
Pyranometer
Cooled mirror
YS1
Eppley Black and White
Eppley UV
EGandG 880
0.2°F
0.001 Langly
0.1 m Langly
0.1°F
20-110°F
0-2 Langly
0-100 mLangiy
-40 to 120°F
Continuou.
Continuou.
Continuou.
Continuou.
No
No
No
Ye.
-------
OO
to
Species
Table B2. Calibration Sources for Gases
Source Calibration Method
References
NO
NO2
CH4, CO, THC
HC's
HCHO
UV-quartz Os generator
Cylinder OsUV photometry
Oxidaton of NO
Cylinder
Cylinder
paraformaldehyde and 37%
HCHO
Acetaldehyde, propional- pure compounds
dehyde, MEK
Aromatics pure compounds
PAN
Alkyl nitrates
HN03
H202
PAN purified
pure alkyl nitrates
HNOa (gaseous)
liquid H2O2
UV photometry Dasibi EPA
Model 1008-AH
Certified commercial analysis EPA
and transfer via Gas Phase
titration
Gas Phase Titration EPA
Certified commercial analysis EPA
Certified commercial analysis EPA
Prepared static calibra-
tion standards verified with
chromotropic acid bubbler
technique
Prepared static calibra-
tion standards
Prepared static calibra-
tion standards
Reference method and trans-
fer standard
Prepared static calibra-
tion standards
Dynamic source (gas) calibrated Spicer, C., Adv. Envr.
by colorimetric (chromotropic Sci. and Tech. 7 p!83 (1980)
acid) bubbler method •
Liquid standards and iodometric
analysis
Winer, A. ct d. EST 8
p!116, (1974)
-------
Table B3.
Species Measured by Major Site Instruments
Bendix Ozone Monitor
t> Ozone
Bendix Nitrogen Oxide Monitor
> nitric .oxide, NO
> nitrogen dioxide, NOj and PAN
> sum, NO*
CEA Formaldehyde Monitor
> formaldehyde
Carle I FID Gas Chromatograph
> Cg to Cg alkanes
e.g. propane, butane, isopentane, pentane, 2-methylpentane, 2,2,4-trimethylpentane,
octane
> Ce to Cg aromatic9
e.g. benzene, toluene, m-xylene
> Cj to Cg alkenes
e.g. ethylene, propylene, 1-butene, cis-2-butene, trans-2-butene
> acetylene
> Total Hydrocarbon
Carle II FID Gas Chromatograph
> methane
t> ethane
> propane
> ethylene
> propylene
> acetylene
> acetaldehyde
t> propionaldehyde
> acetone
> methylethylketone
184
-------
Table B3.
Carle III FID Gas Chromatograph
> €3 to Cg alkanes
e.g. propane, butane, isopentane, pentane
> C3 to €5 alkenes
e.g. ethylene, propylene, isobutene, 1-butene, cis-2-butene, trans-2-butene
> Cg to CIQ alkanes
e.g. octane, 2,2,4-trimethylpentane
t> Ce to C9 aroma tics
e.g. benzene, toluene, m-xylene, o-xylene, 1,2,4-trimethylbenzene
> methanol
Varian BCD Gas Chromatograph
> PAN
> alkylnitrates
> biacetyl
> tracers
e.g. CCU, C3Cl4, C2C13F3
Perkin Elmer 900 FID Gas Chromatograph
> Ce to Cia aromatics
> aromatic oxygenates and nitrogen-containing products
Perkin Elmer Sigma 2 FID Gas Chromatograph
> C3 to Ci2 alkanes
e.g. propane to dodecane
> C3 to C6 alkenes
e.g. propylene to hexene
> Ce to GU aromatics
e.g. benzene to dimethylnapthalene
> aromatic oxygenates and nitrogen-containing products
Beckman 6800 NMHC, Methane and CO FID Gas Chromatograph
> Total Hydrocarbon
> methane
> CO
185
-------
Figure Bl. Example Chromatograms from Beckman 6800 Gas Chromatograph. Peaks: (l) Total Hydrocarbons — 4.0 ppmC, Attenuation
x 2; (2) Methane — 2.5 ppmC, Attenuation X 1; (3) CO — 9.9 ppmC, Attenuation x 2; (4) CO — 37.2 ppmC, Attenuation x 2.
-------
Hydrocarbons Analytical Methods
nated Carle I) is also used to obtain total hydrocarbon data. This system introduces
a sample directly into a blank column and then to the FID. Figure B2 shows how
this instrument is used to match the THC injected in the two chamber halves. The
increase in the chamber THC is clearly shown as well as the degree of relative match-
ing of the two chamber sides. This THC method is used primarily in the morning
during injection and the regular operation of the Carle I is used for the rest of the
run. By characterizing the response to zero air and methane through this system.
and measuring methane separately on another column, it is possible to subtract the
air and methane response from the THC response and obtain NMHC concentrations.
The exact procedure that we have used is described in detail in Appendix D.
Three automated Carle Model 211 packed column FID gas chromatographs are
used for Cj-Cio analysis. The first of these chromatographs, Carle I, is used to
perform a gross analysis on C2~C10 hydrocarbons. Three chromatographic columns
perform the analytical separation. An additional 2' x 1/8" column of carbowax 400
was installed just before the flame detector to balance column bleed and prevent
excess baseline displacement after valve switches.
The columns used in Carle I are:
Column 1 6' x 1/8" carbowax 1540 -r'5% Apeizon L on Supelcoport 80/100
mesh support to separate Ce-Cg paraffins, olefins and aromatics
Column 2 6' x 1/8" N-octane/Poracil C 100/120 mesh (baked at 240°C for 2
hrs.) -j- 2' x 1/8" 20% Apeizon L on Supelcoport 80/100 mesh to
separate C4 and Cg olefins and some C(. paraffins
Column 3 2' x 1/8" FL alumina treated with 5% NaCl to separate C^Cs com-
pounds.
Bidirectional electric motors (Carle model 4201) are used to actuate 6-poft
mini-valves and these shunt the sample to the appropriate columns. The valves are
controlled by a 30 minute 110 volt Carle programmable timer (Carle model 4102).
The column configurations for each valve position are shown in Figure B3. The
indicated restrictors are adjustable S.S. 1/16" Nupro needle valves and are used to
match the pressure drop across various columns as these columns are switched in
and out of the flow circuit.
Automatic injection is made with all three columns in series. One and one half
minutes are required for the individual compounds to become distributed into the
proper columns. Since column 1 is the first to be encountered, Ce-Cg are retained,
while Ci-Cs pass on to column 2. After this first 1.5 minutes with the columns
187
-------
OO
OO
d
O
M
tn
^
u
0)
o
0)
o
4J
C
•H
o
•H
4J
O
C
•H
4J
U)
3
(0
A
X M
0) 0)
-P m
3 xl
(d o
c -0
•H Q)
0)
PQ
Q)
M
C
•H
0)
-P
0)
1
O
o
C
o
•H
4J
0)
O XJ
0)^
•n id
H
O
10
15
C
•H
-P
(n
(d
0)
T) o
C -P
,x
0
(d
XI
G V-i
O QJ
'H, -S
H -P g
0) u
-------
Hydrocarbons
Analytical Methods
INJECT
Col 1.2,3,
in Series
Col. 3
First Valve Suitch
Col. 1 in Series
with FID .
C.I-3
Third Valve Suitch
Col. 1 & 2 in Series
Figure B3. Column and Valve Configuration for CI-CJQ Automatic Carle Chromatograph.
in series, a valve switch removes column 2 and 3 from the flow path and permits
the compounds in column 1 to elute into the FID. The system remains in this
configuration for 5.5 minutes so that Ce,-C10 paraffins and toluene can be eluted.
We also extend this time to 14 minutes to measure xylenes or a-pinene.
A second valve switch then briefly (1 minute) places columns 1, 2. and 3 back in
series so that methane and ethylene can be detected. During this period, propane
and propylene move from column 2 to column 3 but do not elute before the third
valve switch. This switch removes column 3 and leaves columns 1 and 2 in series with
the detector. Butenes, pentanes and pentenes then elute from column 2. Finally,
with only 2.5 minutes remaining in the cycle, the columns are placed back in series
so that propane, propylene and acetylene can be measured.
A chromatogram of the UNC mix is shown in Figure B4. The operating sen-
sitivity can be as low as 7 to 15 ppbC (at a signal to noise ratio of 2 on the most
sensitive scale). A limitation of this system is that ethylene and ethane, acetylene
and propylene. and some C4 and Ce compounds are not completely resolved.
A second packed column GC (Carle II) is used to measure Ci-Cs hydrocarbons
and selected carbonyl species. It employs a 2' x 1/8" ss. 60/80 mesh Porapack
189
-------
CONCENTRATIONS OF INDIVIDUAL SPECIES IN
0.39 PPMC NMHC UNC MIX
5
8
7
9
9
10,
II
COMPOUND
2-METHYLPENTENE
2,4-DIMETHYLPENTENE
2.2.4-TRINETHYLPENTANE
ETHYLENE
BUTENE-1
BUTENE-2
ISOPENTANE
N-PENTANE
2-METHYLBUTENE-l
2-METHYLBUTENE-2
PROPYLENE/ACETYLENE
CONCENTRATION (PPMC)
0.035
0.039
0.040
0.042
0.010
0.013
0.053
0.092
0.008
0.007
0.047
NMHC « 0.39 PPMC
SAIN • 2
Figure B4. Maximum sensitivity of Carle chromatograph on attenuation of XI and gain of 2.
190
-------
Hydrocarbons Analytical Methods
Q column for Ci-C3 separation and a 3 m 1/8" SS 10% carbowax 20M (80:100
supelcoport) column for the carbonyls; toluene and xylenes also elute from this
column. As with Carle I, Carle II uses the same type motor-actuated valves and
a programmable valve controller. A 5-cc air sample is injected onto each column
every 15 minutes. While one column is in series with the flame detector, the other
is being back flushed. A typical 30 minute analysis on this system is shown in
Figure B5.
The third.packed column chromatograph (Carle III) is used to resolve C4 and
Cs hydrocarbons on one column and aromatics on another. It uses a GP 80/100
mesh. Carbopack 19% picric acid, 1/8" SS x 3-m column for C4 and C5 and 1/8" x
3-m, GP 10% TCEP, on 100-120 Chromosorb AW column for aromatics. A valve
diagram of this system and a chromatographic analysis of automobile exhaust are
shown in Figure B6 and Figure B7.
As can be seen, many compounds such- as propylene, butenes. toluene and
xylenes can be detected by two or even all three of these chromatographs. Most
compounds can be quantitatively determined at concentrations of 0.05-0.1 ppmC
and, in instances when peaks are sharp, at levels of 0.01 ppmC. The output from
each of these instruments is monitored with a Perkin Elmer Sigma 10 integrator
and also with stripchart recorders. Since 30 minutes is required for one complete
analysis a sample from each chamber side is taken every hour.
Aromatic compounds are detected with greater sensitivity and resolution by
two capillary FID gas chromatographs. The first system was built around a Perkin
Elmer 900 gas chromatograph and the second uses a Perkin Elmer Sigma 2.
A plumbing diagram for the first system is shown in Figure B8. Programmed
signals from the Sigma 10 integrator autoactuate two 6-port, pneumatically oper-
ated valves which allow multiple 11-cc samples to be trapped in a freeze-out loop.
The sample is then automatically desorbed onto a 30 m DB 1 (J & W) or SP2100
(H.P.) fused silica column.
The freeze-out loop has an internal volume of 50 to 150 /^liters. It is packed
with 60/80 mesh glass beads to increase the internal surface area and decrease
the internal dead volume. The effective internal volume is «25-75^1iters. A dry
ice/methanol bath is manually used to cool the trap. The loop is wrapped with
fiberglass insulated nichrome wire for the thermal desorption and will heat the loop
from liquid nitrogen temperatures to 200° C in less than one minute.
191
-------
0)
c-
(O
J=
•»->
UJ-
cu
cu.
^
4J
L^
cu
c
cu
o
S-
a.
•a
c
03
CU
c
cu
Carle I
Figure B5. Analysis of Cj-Cs hydrocarbons in 4 ppmC dilute chamber auto exhaust on Carle II (Model 211) packed column gas
chromatograph (July 1, 1982).
-------
Vent
Restrictor to
balance flow
Aromatic
analytical
column
-------
(0
.u
Figure B7. Analysis of aromatics and C.4-C6 hydrocarbons from dilute auto exhaust injected into UNO outdoor chambers, October
4, 1983.
-------
Hydrocarbons
Analytical Methods
20 psl Helium Carrier Gas Supply
Figure B8. Plumbing Diagram for Perkin Elmer Capillary GC for Aromatic Analysis.
The operating cycle is controlled by a solid state controller. After an initial hold
of 6 minutes for valve switching, the column is programmed from 40°C to 150°C at
6.5°C per minute.
This system will separate hydrocarbons from Ce through C14. A wider range
is possible using sub-ambient column temperatures. A typical calibration chro-
matogram from a mixture of hydrocarbons injected into the chambers is shown in
Figure B9 and an analysis of dilute chamber automobile exhaust in Figure BIO. Ben-
zaldehyde, tolualdehydes. cresols, nitro benzenes, nitro and toluenes, nitro cresols
and dimethyl phenols can also be monitored with this system. The analytical sen-
sitivity of this GC to most compounds is in the 5 ppbC range. A complete analysis
requires 45 minutes, including oven cooldown and injection. Thus one sample is
taken from each chamber every 1.5 hours.
The fully automated capillary system was designed at UNC around a Perkin
Elmer Sigma 2 GC and a Spectra Physics 4100 integrator. A separate controller box
was built to drive a LNj cryogenic sample concentrator and then thermally degas
the trapped HC species onto a 30m DB-1 capillary column. Prior to the start of
an analysis the oven is cryogenically cooled with LN2 to -40°C and then, during the
195
-------
Ol
c
O) O) , 01 » C
C X Ol 0) :
O) i •— c;
3 e >» *
•— . x c:
O 08 1 O
•1
4
tL
1 F*l I
'
t
{
1
t
h'
tf
I
i
)
t
1
i
j J
«
a**
*
t
W 1° M"' '2
1 • ^
X X jfiocor
JOT
j«sir
it M /"
' :
I;
t
f.
\
':
ico
*"••£>
^_
p
Ol Q
C
J5
o
o
1 -
; •
C
CD
;
j !
•4
CD
C
L
c
1
:j
.-
^
H";
a:
i
*
!
;f 1
. »p A'
T T
j
^
, .,
P
J •
r •
i*rt i ^
LLJ L r
U 1 • J^
1
Ol
1 c
i °
•a
L-
1
1
1
>
.
•^*
^ cb cb
I cr- !•
5 f %
i i ^7
ill
: ? i rii
i; >
•; - *
ii i 1 1:.
!
1
D
i
• r-<=? •
1 ^1 !N
i TiT* i
[' (fTP M
; i ;! i ! ;
.ii ! : ' '; ' " ; ;
Ol
c
fO
— tetradec
• -i
,
i -
1 • e
. $
i!
Figure B9. Calibration Chromatogram on PE 900 FID Gas Chromatograph 30 m
DB-1 fused silica column (J&W), 22 cc air sample concentrated in 1/16"
x 8" SS tube packed with 60/80 mesh glass beads (internal volume 25-75
ul) cooled with MeOH dry ice and thermally desorbed at 200°C with heat
tape. Initial oven temperature = 40°C and programmed at 6.5°C/min to
150°C.
196
-------
10 min.
15 min.
20'min.
Figure BIO. Analysis of C6Ci3 aromatics on PE 900 FID capillary gas chromatograph from a 3 ppmC dilute auto exhaust chamber run
on June 30, 1982. Compound identities and concentrations are: l) benzene, 0.337 ppmC 2) toluene, 0.366 ppmC 3) ethylbenaene,
0.051 ppmC 4) m&p-xylene, 0.187 ppmC 5) styrene, 0.059 ppmC 6) o-xylene, 0.078 ppmC 7) nonane 8) isopropylbenzene 9) n-
propylbenzene, 0.012 ppmC 10) m-ethyltoluene, 0.057 ppmC 11) 1, 3,5-trimethyIbenzene, 0.032 ppmC 12) o-ethyltoluene, 0.017
ppmC 13) 1,2,4 trimethylbenzene + tert-butylbenzene, 0.097 ppmC 14) decane + secbutylbenzene 15) 1,2,3 trimethylbenzene,
0.019 ppmC 16) dimethylethylbenzene, 0.062 ppmC 17) napthylene, 0.083 ppmC 18) 2-methylnapthylene, 0.0313 ppmC 19) 1-
methylnapthylene, 0.015 ppmC.
-------
Analytical Methods Hydrocarbons
analysis, it is programmed to 180°C; one complete cycle requires 45 minutes and
compounds in the Cs-Cio range can be analyzed..with a sensitivity of 1-3 ppbC. The
advantage of this system is its increased sensitivity and the potential to ultimately
monitor CS-GIJ hydrocarbons in one chromatographic cycle. This permits us to
analyze most hydrocarbons on one system and provide better estimates of paraffin
to olefin ratios. An example chromatogram of a calibration mixture is shown in
Figure Bll.
A list of pure compounds and standard mixtures used to calibrate this gas
chromatograph is given in Table B4.
198
-------
(U
C\J
I
10
c
to
S-
(O
QJ
•c
Q)
IM
C
O)
d>
c
O d ^*
cai Q) a> E
>
(O a. _c
C O) 3 r— C 4-»
O) CD n— >> CU 1
Q. O X J- C ._•<*•
^_
>,
c z: cu .c"
QJ £=
a.
o
10
H- 1
4
,
0
f
f
CO
I
ro
^
S "
• N
^ 2 ^
.
j\
ID N
,S ..£
-p
(U
1
OJ
fv
ID
M
-"
i — i >>a> «
E -Pr- CVJ
n U,
M
.
2
to >i •
X .—
r^
**.
ji
,-)
^,
m
M
1
O
ID
in
M
OJ
i
» in
( ,
E
Figure Bll. Chromatograph of calibration mixture on auto sampling Perkin Elmer Sigma 2 capillary gas chromatograph with 30 m
DB-1 column (J&.W). 30 cc air sample (with 0.5 ppmC of each alkane and 1 ppmC of each aromatic) to auto LNj trap, oven =
-40°C hold (5 min hold) program rate 5"C/min to 150°C, final hold 5 min. Signal attenuated 10 fold.
-------
Analytical Methods Hydrocarbons
Table B4 Species Use for Identification and Calibration on PE Sigma 2
Species Source
propylene 1
propane 1
propadiene 1
butane 1
isobutane 2
1-butene 1
1,3-budtadiene 1.6
cis-2-butene 1
trans-2-butene 1
2-me-l,3-butadiene 1
2-methylpropene 5a.6
2-methyl-l-butene 3
2-methyl-2-butene - 3
isopentane 1
n-pentane 1
1-pentene 2
cyclopentene 4
2.2-dimethylpropane 5a.6
benzene 2
2-methylpentane 2
3-methylpentane 4
2-methyl-l-pentene 5a.6
4-methyl-l-pentene 5a,6
cyclohexene 4
cyclohexane 2
n-hexane 2
n-heptane 2
2,3-dimethylpentane 2
3-methylhexane 5a
200
-------
Hydrocarbons Analytical Methods
Table B4 Species Used for Identification, continued
Species Source
toluene 2
methylcyclohexane 4
n-octane 1
m-xylene 2
o-xylene 2
styrene 4
ethylbenzene 2
o-ethyltoluene . 4
m-ethyltoluene 4
1,2.3-trimethylbenzene 4
1,2,4-trimethylbenzene 4
1,3,5-trimethylbenzene 4
n-propylbenzene 4
isopropylbenzene 4
n-butylbenzene 4
s-butylbenzene 4
t-butylbenzene 4
diethylbenzene 5b
dimethylethylbenzene 4
naphthalene 5b
1-methylnaphthalene 4
2-methylnaphthalene 4
1 Certified Cal. tank, LOMW
2 Certified Cal. tank, HIMW
3 Calibrated Injection tank. UNCMIX
4 UNC General Chemical Supply
5a Mobile Sources Emissions Research Branch. EPA
5b Mobile Sources Measurements
Research Section, EPA
6 William Lonneman, Gas Kinetics and Photochemistry
Branch, EPA
201
-------
Analytical Methods Mass Spectrometry
Mass Spectrometry
Individual identification of hydrocarbon species in mixtures as complex as automo-
bile exhaust is performed at UNC on a VG micro-mass 707F mass spectrometer
interfaced to a Hewlett Packard 5710A capillary gas chromatograph. One to two
cc of raw automobile exhaust are injected onto a 30 m, SE54. or DB-1 fused silica
column. The first 0.3 meters of the column1 are coiled into a cup of liquid nitro-
gen so that on-column trapping of the sample can be accomplished. The oven is
then programmed from 10°C to 150°C at 5cC/min. Identification is performed by
comparison to the mass spectrum from authentic samples and to literature spectra
available in the EPA-NIEHS mass spectra libraries. An analysis of raw automobile
exhaust and gasoline is shown in the reconstructed ion chromatographs in Fig-
ure B12 and Figure B13. Figure B14 is a library mass spectrogram of napthalene
and Figure B15 is the mass spectrogram of peak 43 in Figure B12. and there is little
doubt that peak 43 is napthalene.
Quantification of individual hydrocarbon species is performed on an "off line"
Carlo Erba 4130 FID gas chromatograph. The same fused silica column, cryotrap
and temperature program which were used for the GCMS analysis are used here.
Identifications are made by comparing the FID chromatogram (Figure Bl6) with
the GCMS reconstructed ion chromatograms. Response factors are determined from
a mixture of 15 authentic compounds which are analyzed on the Carlo Erba.
Formaldehyde by Automated Colorimetry
Since 1979. UNC has used an automated colorimetric formaldehyde instrument
manufactured by CEA corporation (Westwood. NJ). This system employs a reverse
West and Gaeke techniquej and hence requires a constant supply of wet chemicals
which include highly purified pararosanaline, mercuric chloride, sodium chloride,
and sodium sulfite.
Liquid standards are prepared from 37% formalin solution and diluted in the
same sodium tetra mecurate (TMC) used in the instrument. A dynamically gen-
erated HCHO airstream using a formaldehyde permeation tube (Metronics Corpora-
tion. Santa Clara, CA) is used to supply known levels of gas phase formaldehyde.
In addition, calibration samples of formaldehyde are prepared in the chamber. The
chamber is usually dried to a dew point of less than 60°F and known quantities of
solid paraformaldehyde are then injected. During the course of a six month data
season when the chambers are in operation, we have found this system's calibration
not to change by more than 10%.
202
-------
100 -
80 .
o
CO
60
% I
40 -
20
400
800
Figure B12. Reconstructed ion chromatograph of EPA summer gasoline using DB-1 column. Identities based on comparison with
literature mass spectra. Compounds listed in order of elution are: l) n-butane 2) 2-methylbutane 3) n-pentane 4) methyl-
butene or 2-pentane 5) 2-methylpentane 6) 3-methylpentane 7) n-hexane 8) hexadiene or a methylcyclopentene -f a hexene 9)
e-methyl-2-pentene 10) methylcyclopentane ll) benzene 12) 2- methylhexane 13) 2,3-dimethylpentane 14) 3-methylhexane 15) 2, 2,4-
trimethylpentane 16) n-heptane 17) methylcyclohexane 18) 2 ,5-dimethylhexane 19) 2,4-dimethylhexane 20) 2,3,4- trimethylpentane
or 4-methylheptane 21) toulene 22) 2,3- dimethylhexane 23) 2-methylheptane 24) 3-methylheptane 25) 2,2 ,4- or 2,2,5-trimethylhexane
26) n-octane 27) ethylbenzene 28) m- and/or p-xylenes 29) methylpentanes 30) 3-ethylheptane 31) xylene 32) n-propylbenzene 33)
m- and/or p-ethyltolune 34) 1,3 ,5-trimethylbenzene 35) 2- or 4-methylnonane 36) p-ethyltoluene 37) 1,2,4-brimethylbenzene 38)
1,2,3-trimethylbenzene 39) n- decane and other CiO alkylbenzene 40) indan or a methylstyrene 41) sec-butylbenzene and other C4
alkylbenzene (m-tp-n-propyltoluenes and maybe some tert-butylbenzene) 42) n- butylbenzene and other C4 alkylbenzene (maybe
some diethylbenzenes) 43) n-propyltoluene and some undecane 44) dimethylethylbenzenes 45) a dimethylethylbenzene 46) tetram-
ethylethylbenzenes 47) C6 alkylbenzene and C4 alkenebenzene 48) C4 alkenebenzene 49) naphthalene.
-------
15
100 _
24
20
19
-------
100
80.
BO.
* I .
MQ_
O
Cn
20.
20
.1. 11.,.,.. I
I —•••'
60
100
1MO
Figure B14. Library mass spectrogram of Naphthalene on the VG Micro-Mass 707F Mass Spectrometer.
-------
io6
eo.
zoj
o 1
100*
1C.
eo.
10.
10.
o
100
00.
•
to.
J 1
10.
to
c_
,,
1
50
e
•* iiillJiiiipi
TO
' *P
SO
' • "(• **" '
f , 1
7
1
T
1
90
,., i,,;u
••'« ...,i.......ii....i.-..i—"
so
.,
s
;
T i"
130
Jl
'1
iii|iniiiiii|
130
| «
"I"" '
130
NAPHTHALENE
mi iii|i inn iii| inn mil mn'mrprmi in iniiiiii iii '
170
A METHYLNAPHTHALENE
i
1 SB ' '"
C2*-NAPHTHALENE
i
A-DIMETHYLNAPHTHALENE
1 S
.170
Figure B15. Mass spectrograms of Peaks 43, 44, and 45 of Figure 12, Auto Exhaust.
-------
15
20
tc
o
a:
•x
11
19
27
22
24
Time in Minutei
Figure B16. Hydrocarbon species identified in cryocondenser auto exhaust (7-28-83) on Carlo Erba-FID gas chromatograph 30 m
'DB-1 column (J&W) on column liquid oxygen freeze out. Temperature programmed 10°C hold 7 minutes program at 5°C/min
to 150°C. Identities bases on comparison with reconstructed ion chromatographs generated under similar operating conditions, l)
2,3-dimethylbutane 2) 2-methylpentane 3) 3-methylpentane 4) n- hexane 5) methylcyclopentane 6) 2,2- and/or 2,4-dimethylpentane
7) benzene and some methylcyclopentene 8) cyclohexane 9l 2- methvlhexane 10) 3-methylhexane 11) 2,2,4-trimethylpentane 12)
n-heptane 13) methylcyclohexane 14) 2,4-dimethylhexane 15) toluene 16) 2,3-dimethylhexane 17) n-octane 18) 2,5- dimethylheptane
19) ethylbenzene 20) m-& p-xylene 2l) 3- methyloctane 22) o-xylene 23) n-propylbenzane 24) m-& p- ethyltoluene 25) 1,3,5-
trimethylbenzene 26) o-ethyltoluene 27) 1,2,4-trimethylbenzene 28) 1,2,3-trimethylbenzene 29) indan 30 ) C4-benzene 3l) methylin-
dans 32) naphthalene.
-------
Analytical Methods Carbonyl Analysis
The minimum detectable concentration for this instrument is 20 ppb and a con-
tinuous sample on each chamber is taken every 30 minutes. Using the solutions
recommended by CEA we observed a 5% positive interference from acetaldehyde in
the concentration range of 0.1-1 ppmV. When appreciable acetaldehyde is gener-
ated, the HCHO data are adjusted to subtract out the positive acetaldehyde interfer-
ence. Interferences from other carbonyls such as acetone, propionaldehyde, MEK
or methacrolein produced are not significant. Finally, the system that we have been
using exhibits a 30-50% zero drift over a 3-4 hour period. This drift does not affect
the span response but requires frequent zero samples. Examples of the system's
response to 1.2 and 0.5 ppmC formaldehyde air samples are shown in Figure B17.
Carbonyl Analysis j
The analysis of Cj-Cg carbonyl species in ambient air by reaction with 2,4 dinitro-
phenylhydrazone (DNPH) and subsequent reverse phase high pressure liquid chro-
matography of the hydrazone derivatives has been reported by Kuwata tt a/.2 This
procedure is fairly specific and can also be used to measure dicarbonyls. We have
used this method to monitor C2 and greater carbonyls and have generally used
the previously described automated colorimetry method to obtain more continuous
formaldehyde measurements.
Ambient, aldehydes are sampled with an impinger arrangement into an absorb-
ing reagent which contains acidified DNPH. Three to four recrystallizations of the
DNPH (Eastman Kodak cat. no. 1866. 10% water added) with acetonitrile (ACN)
are required to remove impurities from the DNPH. The absorbing reagent is pre-
pared by dissolving 0.25 g of purified DNPH in one liter of HPLC grade ACN.
This is followed by the addition of 0.2 ml of H2SO4. Two ml of absorbing reagent
are used in a Mae West type bubbler with a sample flow rate of one 1/min, as de-
scribed by Kuntz tt al.3. We have observed collection efficiencies of 75-85% for most
of the carbonyl compounds which have been tested. These include formaldehyde,
acetaldehyde, propionaldehyde, methylvinylketone. benzaldehyde, glyoxal. methyl-
glyoxal, biacetyl and o-tolualdehycle.
Analysis is performed on a Varian 5010 HPLC with a Varian micropak MCH-10
reverse phase, 30 cm, 9 mm column and a fixed 254 nm detector. To eliminate
interferences from toluene and xylenes a 365 nm cut. off filter can be used. The mo-
bile phase is run isocratically at a 60:40 ACN:H2O mixture and a solution flow of
1.6 ml/min. High concentration liquid standards in the gas phase equivalent range
of 2-6 ppmV were prepared by direct dilution of 5 n\ of authentic knowns (liquids
available from Aldrich or Eastman, purity 98-99%, except for 40% gloxal in water,
40% methylglyoxal, 37% formaldehyde, 85% methacrolein and 95% biacetyl. These
208
-------
Figure 17. Response of CEA Auto Formaldehyde Instrument to 0.5 and 1.2 ppm
of HCOH injected in dry UNO chambers on August 4, 1979 and August 5,
1979.
209
-------
Analytical Methods PAN Analysis
standards were used as received and not further purified.) to twenty //I of this solu-
tion were then added to 2 ml of DNPH absorbing reagent. The final DNPH samples
were heated in 2 ml Supelco reaction vials at 70°C for 30 minutes. Comparisons
with gas phase and liquid calibration samples are made. A sample chromatogram
of a mixture of aldehydes, ketones and dicarbonyls is shown in Figure B18.
PAN Analysis
PAN measurements are made on a 1/8" x 36" glass column packed with 10%
Carbowax 600 on Gaschrom G (60-80 mesh) and detected with a Varian 940 electron
capture detector. Both the column and the detector are held at room temperature
and 5% methane-argon is used as the carrier gas. A 6-port automatic sample valve
was installed so that unattended sample injection and analysis could be performed.
With a standing current of 20 x 10~& amps we "found that the response was linearly
between 0.01 ppmV and 0.5 ppmV of PAN./The response of this system to 0.12
ppmV of PAN is shown in Figure B19.
PAN calibration samples are prepared from an irradiated SppmC biacetyl/NOx
bag system. The bag is irradiated for ] hour in midday sun so that essentially all of
the NOj is converted to PAN. The PAN is purified with a 10% Carbowax 600 prep-
column (25 cm x 7 mm) and a liquid nitrogen freeze-out concentrator. PAN from
the cryotrap is then diluted into a clean Teflon bag and measured with a calibrated
chemiluminescence NOX analyzer. The response of the chemiluminescent monitor to
PAN is 100% and thus a pure PAN sample can be standardized.
When the above procedure is not possible, then 4 ppmC of propylene and 0.5
ppm NOX (40% NOz) are irradiated outdoors for 3-4 hours during the mid-morning
and the early afternoon. With a chemiluminescent NOX meter, the NC>2 concentra-
tion is followed beyond the NOj peak until NO2 stabilizes. The PAN GC responses
are then related to chemiluminescent PAN readings assuming that all N©2 that, is
reported by the meter is actually PAN.
Alkynitrates
The C]-C5 alkylnitrates can be detected on the same system that is used to measure
PAN. Calibration for ethyl, isopropyl. propyl, and butylnitrates were performed by
injecting microliter volumes of the authentic samples into the outdoor chambers
and applying the appropriate dilution and temperature factors (Figure B20). Cal-
ibration factors for compounds for which we did not have authentic samples (e.g.,
sec-butylnitrate and isopentylnitrate) were estimated from the ECD response of
n-butylnitrate. An uncertainty of ±40% is assigned to these compounds with no
authentic standards.
210
-------
4 6
TIME IN MINUTES
10
Figure B18. An example of the response of the HPLC with DNPH Method.
211
-------
ATC detector x210i !
I. .
Figure B19. Response of two UNC electron capture detectors to 0.12 ppmV PAN May 1978.
212
-------
Figure B20. Detection of alkylnitrates with automated PAN GO, Varian 940 ECD. N2 flow = 5 cc/iain, oven and detector at 24.6°C,
standing current on 32xl(T0 amps = 91% of full scale, chart = .1"/van, atten. = x32xlO-10 amps.
-------
Reference?
References
1 West. P.W., Gaeke G.C. Anal. Chem., 28, 1956, p.1916 and Lyles G.R., Dowling F.B.,
Blanchard V.T., JAPCA, 20, 1965, p.106
2 Kuwata K., Uebori M., Yamasaki Y., Journal of Chrom. Sci., 1979. 17. pp.264-268
3 Kuntz P., Loiineinaii VV., Namie G., Hull L.A. Anal. Lett., 1980, 13, pp. 1409-1415
214
------- |