United States
Environmental Protection
Agency
Office of Mobile Sources
Emission Control Technology Division
2565 Plymouth Road
Ann Arbor, Michigan 48105
EPA 460/3-86-001
March 1986
c/EPA
Air
Photochemical Modeling of
Methanol-Use Scenarios in Philadelphia
-------
EPA 460/3-86-001
Photochemical Modeling of Methanol-Use
Scenarios in Philadelphia
by
G.Z. Whitten
N. Yonkow
T.C. Myers
Systems Applications, Inc.
101 Lucas Valley Road
San Rafael, California 94903
Contract No. 68-02-3870
Work Assignment 6
EPA Project Officer: Thomas N. Braverman
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, Michigan 48105
March 1986
-------
This report was furnished to the Environmental
Protection Agency by Systems Applications, Inc., 101
Lucas Valley Road, San Rafael, California, in
fulfillment of Work Assignment 6 and 2- of Contract
No. 68-02-3870. The contents of this report are
reproduced herein as received from Systems
Applications, Inc. The opinions, findings, and
conclusions expressed are those of the authors and
not necessarily those of the Environmental Protection
Agency. Mention of company product names is not to
be considered as an endorsement by the Environmental
Protection Agency.
Publication No. 460/3-86-001
-------
CONTENTS
ILLUSTRATIONS iv
TABLES vi
ABSTRACT vi i i
1 INTRODUCTION 1
2 DESCRIPTION OF MODELS USED 2
Chemical Submodel 2
The Urban Airshed Model 15
Description of the Systems Applications Trajectory Model
and Box Model Used in the Philadelphia Study 15
3 DESCRIPTION OF MODEL INPUTS 18
UAM General Input Preparation Procedures 18
Urban Airshed Model Inputs 19
Inputs Used in the 1979 Urban Airshed Model Simulations 23
Modifications to 1979 UAM Inputs for
Philadelphia Application 33
4 DISCUSSION OF MODEL RESULTS 70
Methanol Fuel Substitution 72
Sensitivity of the Model Simulations to Formaldehyde 83
Formaldehyde Concentration Levels 84
5 CONCLUSIONS 88
REFERENCES 124
m
85117ri*
-------
ILLUSTRATIONS
1 Schematic illustration of the grid used and treatment of
atmospheric processes in the Systems Applications
Airshed Model 16
2 Trajectory path for July 13 regional ozone maximum 27
3 Trajectory path for July 19 regional ozone maximum 31
4 Mixing height profiles for urban and rural cells for the
13 July 1979 simulation 35
5 Mixing height profiles for urban and rural cells for the
19 July 1979 simulation 36
6 Airshed model surface winds for 13 July 1979 38
7 Airshed model surface winds for 19 July 1979 42
8 Schematic of preparation of mobile emission and
evaporation inputs for 20 percent methanol/80 percent
gasoline simulation 56
9 Trajectory paths for 13 and 19 July 1979 64
10 Overall emissions sensitivity for 13 July 2000 base case
with low initial conditions 71
11 Methanol and NOX sensitivity for mobile source emissions
relative to 13 July 2000 (low initial) base case 73
12 Formaldehyde concentration (July 13) in ppb--year 2000
base case (lower initial conditions) 90
13 Formaldehyde change (July 13) in ppb--year 2000, 2B minus
base case 102
IV
8511 7ri* 1
-------
14 Formaldehyde concentration (July 13) in ppbyear 2000
base case for all hours 114
15 Formaldehyde change (July 13) in ppbyear 2000, 2B minus
base for all hours 115
16 Ozone concentration (July 13) in pphmyear 2000
base case (lower initial conditions) 116
8511
-------
TABLES
1 Carbon-Bond Mechanism-Ill 4
2 Carbon-Bond Mechani sm-3M 8
3 Hourly ozone predictions for CBM-3 and CBM-3B
EKMA simulations 14
4 Inputs to the Systems Applications Airshed Model 20
5 Monitored surface conditions for 13 July 1979 25
6 Monitored surface conditions for 19 July 1979 26
7 Background concentration values for 13 July at the top
of the modeling regionas initial concentrations above
the mixing height, and for all levels of all boundaries
except the levels below the mixing height on the southeast
boundary 28
8 Southeast boundary conditions for cells below the mixing
height for the simulation of 13 July 1979 29
9 Boundary conditions used for the northeast and east
boundaries below the mixing height estimated from data
collected at the Van Hiseville, New Jersey monitor 32
10 Urban and rural mixing height values used in the
diffbreak file for 13 July 1979 34
11 Urban and rural mixing height values used in the diffbreak
file for 19 July 1979 36
12 Background concentration values used at the top of the
modeling region (TOPCONC), as initial concentrations above
the mixing height and for all boundaries except the levels
below the mixing height on the northeast and east boundaries... 45
8511
vi
-------
13 Total daily emissions by source type (g-mole) in the
1979 Philadelphia inventory 46
14 County adjustment factors 48
15a Box model scenarios for 13 July 2000 49
15b Methanol, formaldehyde, methyl nitrate, and hydrocarbon
for box model scenarios as percent carbon of base case
IB mobile sources 51
16 Mobile source inventory splits for 1979 and 2000 55
17 1979 and year 2000 Philadelphia emission inventories 58
18 Background of reactive hydrocarbons 60
19 Lower background concentration values for the year 2000
for all levels of all boundaries except the levels below
the mixing height on the northeast and east boundaries 61
20 Lower initial conditions for the year 2000 62
21 Boundary conditions for the year 2000 used for the
northeast and east boundaries below the mixing height
estimated from data collected at the Van Hiseville,
New Jersey monitor 63
22 Initial conditions and emission rates used for OZIPM
calculations for the base (1A)July 13 66
23 Initial and boundary conditions for OZIPM calculations for
the base case (lA)--July 19 68
24a Methanol impact modeling for 13 July 2000 74
24b Box model sensitivity tests for 13 July 2000 75
25 Methanol impact modeling for 19 July 2000 76
26 Box model methanol impact ozone results for 13 July 2000 with
30 percent and 50 percent mobile source emissions 78
27 Product concentrations predicted for Scenario 11 81
28 Product concentrations predicted for Scenario 12 82
8511?f\ 1
vii
-------
29 Maximum hourly formaldehyde levels comparing year 2000
base case (1A) to 100 percent methanol substitution
of mobile sources (28) 85
vm
85 1 1 7"
-------
ABSTRACT
A photochemical modeling study was conducted to estimate the impact
on smog production resulting from the substitution of methanol fuel for
gasoline and diesel fuel in Philadelphia in the year 2000. Three photo-
chemical models were used: a comprehensive grid model adapted from the
Urban Airshed Model (UAM), a four-cell trajectory model (a Lagrangian
version of the UAM), and a single-cell box model. All three models used
identical chemical mechanisms and input data based on observations for two
days in July of 1979. Emission rates, initial conditions, and boundary
values were forecast to the year 2000. Mobile sources and stationary
sources related to mobile source emissions were estimated to make up 20
percent of the total volatile organic (VOC) inventory for 2000. At the 20
percent methanol substitution level, i.e., complete substitution of mobile
source VOC with 90 percent methanol/10 percent formaldehyde, methanol-
related emissions show little or no impact on smog formation unless metha-
nol is also substituted for 20 percent of the initial and boundary concen-
trations. Model sensitivity tests indicate that methanol substitution
above 20 percent of the overall VOC inventory may significantly inhibit
smog formation, but the reduction also depends strongly on other factors
such as formaldehyde emissions, methanol carryover from upwind sources or
previous days, and NOX levels. Whereas formaldehyde emissions tend to
accelerate ozone formation, methanol carryover and NOX emissions tend to
decrease ozone formation.
85117T 1
ix
-------
SECTION 1
INTRODUCTION
This report discusses the investigation of the potential impact of
methanol fuel substitution in Philadelphia based on scenarios for the year
2000. The work was performed by Systems Applications, Inc. under the
sponsorship of the Office of Mobile Sources (OMS) of the U.S. Environ-
mental Protection Agency. Previous research has shown that methanol can
be used as an alternative to gasoline and diesel fuels. The results of
studies by Whitten and Hogo (1983) and Pefley, Pullman, and Whitten (1984)
in Los Angeles indicate that conversion to methanol fuel on a large scale
could reduce ozone levels and that methanol-fueled vehicles, which emit
volatile organics (VOC) with lower reactivity than that of gasoline-fueled
vehicles thus produce less photochemical smog.
In the previous study by Whitten and Hogo (1983), the Systems Appli-
cations trajectory model, which uses inputs common to both the Urban Air-
shed Model (UAM) and EKMA, was used to predict ozone reduction resulting
from methanol fuel use in Los Angeles. The EKMA model can utilize
alternative chemical reaction schemes as input, whereas in the UAM, the
chemical mechanism is fixed to the computer code. The purpose of the
study reported here was to investigate the effects of methanol substitu-
tion in a major urban area other than Los Angeles using the Urban Airshed
Model (UAM), which is a preferred EPA model, in addition to the Systems
Applications trajectory model and a single-cell box model. Because the
substitution of methanol for gasoline involves a kinetic mechanism for
methanol chemistry and also emphasizes the chemistry of formaldehyde (its
main oxidation product), for the study reported here, a new UAM computer
code was prepared based on an updated and expanded chemical mechanism
featuring explicit formaldehyde and methanol chemistry. The methanol
component of this new mechanism has been recently validated in a smog
chamber study by Pefley, Pullman, and Whitten (1984).
Sections 2 and 3, respectively, describe the computer models and
the input data used in this study. Section 4 contains a discussion of
the results; Section 5 presents the summary and conclusions.
85117T 2
-------
SECTION 2
DESCRIPTION OF MODELS USED
Three atmospheric models (a four-cell trajectory, a single-cell box
model, and the Urban Airshed Model (DAM)) were used in this study. All
three models contain the same chemical submodel, or mechanism, which is
based on the Carbon-Bond III Mechanism (CBM-3), but which contains an
explicit treatment of formaldehyde and methanol. The trajectory model and
the box model are linked to the UAM grid model in the following manner:
the trajectory model shares input files and computer algorithms with the
UAM and is, in essence, a pilot cell for tracing a key trajectory path
through the gridded framework of the UAM. The box model uses averaged
inputs developed from the trajectory model. Because the box model is
capable of generating isopleth diagrams, sensitivity tests can easily be
implemented. The basic chemistry submodel, or mechanism, common to all
three models used in this study, is described, followed by a description
of each model.
CHEMICAL SUBMODEL
The substitution of methanol for gasoline as a transportation fuel
changes the atmospheric chemistry of an urban airshed. Therefore, the
chemical mechanism used to simulate the atmospheric chemistry of an urban
area is of primary importance in assessing the effect of this substitution
on oxidant formation. For this project, the primary urban atmospheric
chemical reactions were simulated with a modified version (CBM-3M) of the
Carbon-Bond III Mechanism (CBM-3). The unmodified version, CBM-3, is
currently approved for oxidant assessment and State Implementation Plan
(SIP) preparation by the U.S. Environmental Protection Agency (EPA,
1984). Technical descriptions of this mechanism and its use in atmo-
spheric oxidant modeling can be found in Killus and Whitten (1984) and
EPA, (1984).
The CBM-3 was designed to simulate typical gasoline-influenced urban
atmospheres and not for the specific task of simulating methanol-related
impacts. Because the reactivity of methanol to hydroxyl radical reaction
is similar to that of the species PAR in the CBM-3, the unmodified CBM-3
85U7T2 3
-------
may actually be suitable for methanol assessment. However, in this pro-
ject the actual concentrations of methanol and, perhaps even more impor-
tant, the concentrations of formaldehyde, as well as their chemistries,
are essential in understanding the overall impact of methanol fuel
substitution. Formaldehyde is the main oxidation product of methanol in
both combustion and atmospheric photooxidation:
OH + CH3OH » CH2OH + H20 , (1)
CH2OH + 02 H02 + HCHO . (2)
Because water is not a significant direct product of methanol photo-
chemistry, and because the second reaction is extremely fast and the sole
pathway of CH2OH loss, the entire chemistry of methanol can be condensed
to one reaction, which produces formaldehyde and the hydroperoxyl radical:
CH3OH + OH » HCHO + H02
The hydroxyl and hydroperoxyl radical species are central to the
atmospheric chemistry of ozone formation and are treated explicitly in all
current atmospheric chemical mechanisms. However, in the CBM-3 mechanism,
formaldehyde is treated as part of a generalized lumped species for all
carbonyl compounds (CARB). Hence, the main modification to the CBM-3 for
this project was the addition of an explicit treatment of formaldehyde
chemistry. Tables 1 and 2, respectively, present the CBM-3 described in
Killus and Whitten (1984) and the modified version (CBM-3M) used for this
project, which involved the addition of the reaction for methanol and an
explicit treatment of formaldehyde.
As can be seen in Table 2, there are several points in the CBM-3M
mechanism where formaldehdye is produced. This "delumping" of formalde-
hyde represents an improvement over the CBM-3 since the species CARB is
parameterized to account for formaldehyde plus all other carbonyl species
in CBM-3, whereas in the CBM-3M, the CARB parameters relating to formalde-
hyde have been eliminated or reduced.
Modifications that improve the chemical accuracy of the CBM-3M have
been made to reactions 21, 25, 27, 28, 43, 44, 60, and 61. In the CBM-3,
as shown in Table 1, all these reactions lead to the species CARB when
formaldehyde is known to be a product. Thus, the subsequent chemistry of
85117T 3 3
-------
TABLE 1. CARBON-BOND MECHANISM III
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Reaction
N02 * NO + 0
0 + (02) + (M) > 03
NO + 03 -» N02 + 02
N02 + 03 + N03 + 02
N02 + 0 + NO + 02
OH + 03 * H02 + 02
H02 + 03 + OH + 202
OH + N02 * HN03
°2
OH + CO -4 H02 + C02
NO + NO + (02) + N02 + N02
NO + N03 + N02 + N02
N02 + N03 + H20 * 2HN03
NO + H02 * N02 + OH
H02 + H02 * H202 + 02
X + PAR *
°2
OH + PAR -4 ME02 + H20
°?
0 + OLE -^ ME02 + AC03 + X
Q + OLE * CARB + PAR
°2
OH + OLE -£ RAO?
Rate Constant
at 298K
(ppm~ mi n~ )
*
4.40 x 106t
26.6
0.048
1.3 x 104
100
2.40
1.60 x 104
440
1.50 x 10'4t
2.80 x 104
§
1.20 x 104
1.50 x 104
105
1300
2700
2700
3.70 x 104
Activation
Energy
(K)
0
0
1450
2450
0
1000
1525
0
0
0
0
-1.06 x 104
0
0
0
560
325
325
-540
(continued)
-------
TABLE
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
03 *
°3 +
0 +
0 +
OH +
03 +
NO +
NO +
NO +
NO +
NO +
NO +
03 +
°3 +
OH +
OH +
OH +
CARB
CARB
Reaction
OLE + CARB + CRIG
OLE * CARB + MCRG + X
°2
ETH -4 ME02 + H02 + CO
ETH + CARB + PAR
°2
ETH -4 RB02
ETH + CARB + CRIG
°2
AC03 -4 N02 + ME02
Oy
RB02 -4 N02 + CARB + H02 + CARB
RA02 -1 N02 + CARB + H02 + CARB
°2
ME02 -4 N02 + CARB + ME02 + X
°2
ME02 -4 N02 + CARB + H02
ME02 * NRAT
RB02 * CARB + CARB + H02 + 02
RA02 * CARB + CARB + H02 + 02
CARB * CRO? + X
0
CARB -4 HOo + CO
0,
CARB -4 AC03 + X
> CO + H2
Qy
+ (02) -» 2/3 (2H02 + CO)
Rate Constant
at 298K
0.008
0.008
600
600
1.20 x 104
0.0024
1.04 x 104
1.20 x 104
1.20 x 104
3800
7700
500
5.0
200
500
7000
6000
(-0.001 Kj)*
(-0.002 KI)*
Activation
Energy
(K)
1900
1900
800
800
-382
2560
0
0
0
0
0
0
0
0
0
0
0
0
0
1/3 (2ME02 + CO + 2X)
(continued)
-------
TABLE 1
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
Reaction
N02 + AC03 + PAN
PAN + AC03 + N02
H02 * AC03 -» Stable products
H02 + ME02 * Stable products
NO + CRIG + N02 + CARB
N02 + CRIG * N03 + CARB
CARB + CRIG + Ozonide
NO + MCRG -> N02 + CARB + PAR
N02 + MCRG + N03 + CARB + PAR
CARB + MCRG -» Ozonide
CRIG + CO + H20
CRIG -» Stable products
0-
CRIG -4 H02 + H02 + CO
MCRG -» Stable products
°2
MCRG -4 MEOo + OH + CO
°2
MCRG -4 ME02 + H02
°2
MCRG -4 CARB + H02 + CO + H02
0-
OH + ARO -4 RARO + H20
Op
OH + ARO -4 H02 + OPEN
°2
NO + RARO -4 NO, + PHEN + HO,
Rate Constant
at 298K
7000
0.022
1.50 x 104
9000
1.20 x 104
8000
2000
1.20 x 104
8000
2000
670**
240**
90**
150**
340**
425**
85**
8000
1.45 x 104
4000
Activation
Energy
(K)
0
1.35 x 104
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
600
400
0
(continued)
-------
TABLE 1.
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
Reaction
°?
OPEN + NO -£ N02 + DCRB + X + APRC
°2
APRC -* DCRB + CARB + CO + X
°2
APRC -* CARB + CARB + CO + CO
PHEN + N03 -» PHO + HN03
PHO + N02 * NPHN
PHO + H02 -» PHEN
OPEN + 03 * DCRB + X + APRC
°2
OH + PHEN -* H02 + APRC + PAR + CARB
°2
DCRB -* 1/2 (H02 + AC03 + CO)
1/2 (ME02 + H02 + 2CO)
PHEN + OH * PHO
°2
CR02 + NO -* N02 + H02 + DCRB
DCRB + OH * AC03 + CO
HONO -» OH + NO
OH + NO * HONO
03 * O^D
O^D ii^»- 0
0^ .+ H20 » OH + OH
Rate Constant
at 298K
6000
104**
104**
5000
4000
5.00 x 104
40
3.00 x 104
(*0.04 Kj)*
104
1.20 x 104
7000
fcO.3 Ki)*
9770
H x ID'4 Kj)1
4.44 x 1010
3.4 x 105
Activation
Energy
(K)
0
0
0
0
0
0
0
0
0
0
0
0
0
0
k o
0
0
Sunlight-dependent; units of min"1.
Units of
5 Heterogeneous; pseudo third order. Equal to 591 x N205 + H20.
** .. -^ f i
Units of mm i.
7
-------
TABLE 2. CARBON-BOND MECHANISM-3M.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Reaction
N02 * NO + 0
0 + (02) + (M) * 03
NO + 03 -» N02 + 02
N02 + 03 * N03 + 02
N02 + 0 + NO + 02
OH + 03 * H02 + 02
H02 + 03 * OH + 202
OH + N02 * HN03
Qy
OH + CO -4 H02 + C02
NO + NO + (02) * N02 + N02
NO + N03 * N02 + N02
N02 + N03 + H20 * 2HN03
NO + H02 + N02 + OH
HOo + H02 * H202 + 02
X + PAR *
°2
OH + PAR -4 R02 + X + H20
°2
0 + OLE -» R02 + AC03 + XX
0 + OLE * HCHO + PAR
°2
OH + OLE -» RA02 + X
Rate Constant
at 298 K
1.0*
4.40 x 106**
26.6
0.048
1.3 x 104
100
2.40
1.60 x 104
440
1.50 x 10'4
2.80 x 104
2.600 x 10'5
1.20 x 104
1.50 x 104
105
1200
2700
2700
3.70 x 104
Activation
Energy
0
0
1450
2450
0
1000
1525
0
0
0
0
-1.06 x 104
0
0
0
560
325
325
-540
Continued
85il/ 8
-------
TABLE 2 (continued)
Reaction
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38a
38b
°3 +
03 +
0 +
0 +
OH +
o3 +
NO +
NO +
NO +
NO +
NO +
NO +
°3 +
03 +
OH +
OH +
OH +
HCHO
HCHO
XCO
OLE + HCHO + CRIG + X
OLE * HCHO + MCRG + X
°2
ETH -4 ME02 + H02 + CO
ETH -» HCHO + PAR
Oo
ETH -4 RB02
ETH -» HCHO + CRIG
n
AC03
RB02
RA02
ME02
ME02
ME02
RB02
RA02
CARB
HCHO
CARB
+ CO
*S
f PAR
-5
f?
^
°2
h
-
*
*
+
N02
N02
N02
N02
N02
NRAT
HCHO
CARB
CRO?
°2 "
-4 H0?
Hj
+
-»
ACO-
*
H2
CO +
CO
+ ME02
+ HCHO + H02 + HCHO
+ CARB + H02 + HCHO
+ CARB + R02 + XX
+ CARB + H02
+ PAR
+ HCHO + H02 + 02
+ HCHO * H02 + 02
+ X
+ CO
, + x
H02 + H02
Rate Constant
at 298 K
(ppm~ min~ )
0.008
0.008
600
600
1.20 x 104
0
1
1
1
5
6
1
5
.0024
.04 x 104
.20 x 104
.20 x 104
.00 x 103
.00 x 103
.20 x 103
.0
20
200
1
2
0
1
.50 x 104
.00 x 104
.39
Activation
Energy
(0 ^ \
/
1900
1900
800
800
-382
2560
0
0
0
0
0
0
0
0
0
0
0
0
0
105
Continued
85ii/ 8
-------
TABLE 2 (continued)
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
Reaction
N02 + AC03 * PAN
PAN + AC03 + N02
H02 + AC03 -»
H02 + ME02 -»
NO + CRIG * N02 + HCHO
N02 + CRIG * N03 + HCHO
HCHO + CRIG *
NO + MCRG * N02 + CARB + PAR
N02 + MCRG + N03 + CARB + PAR
HCHO + MCRG *
CRIG + CO + H20
CRIG *
°2
CRIG -4 H02 + H02 + CO
MCRG *
°2
MCRG -4 R09 + OH + CO + X
°2
MCRG -» R02 + H02 + X
°2
MCRG -» CARB + H02 + XCO + H02
°2
OH + ARO -4 RARO + H90
°2
OH + ARO -4 HOo + OPEN
°2
NO + RARO -4 N02 + PHEN + H02
Rate Constant
at 298 K
(ppm"^ min"^)
7000
0.022
1.50 x 104
9000
1.20 x 104
8000
2000
1.20 x 104
8000
2000
670**
**
240
90**
**
150
340**
425**
85**
6000
1.45 x 104
4000
Activation
Energy
(K)
0
1.35 x 104
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
600
400
0
Continued
85ii/ 8
10
-------
TABLE 2 (continued)
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
Reaction
°2
OPEN + NO -* N0? + DCRB + X + APRC
°2
APRC -4 DCRB + HCHO + CO + X
°2
APRC -4 HCHO + HCHO + CO + CO
PHEN + N03 * PHO + HN03
PHO + N02 * NPHN
PHO + H02 * PHEN
OPEN + Oo * DCRB + X + APRC
°2
OH + PHEN -4 H0? + APRC + PAR + CARB
0
DCRB -» H02 + AC03 + CO
PHEN + OH * PHO
°2
CR02 + NO -4 N02 + CARB + AC03 + X
DCRB + OH * AC03 + CO
HONO * OH + NO
OH + NO » HONO
03 * 00
O^D ' * 0
0^ + H20 * OH + OH
°3 + °
NR *
OH * HO,
Rate Constant Activation
at 298 K Energy
(ppm"1 min'1) (°K)
6000
104**
104**
5000
4000
5.00 x 104
40
3.00 x 104
(0.02 x Kj)**
104
1.20 x 104
25000.
(3.1* or 0.18** x Kx)
9770
1.15
4.44 x 1010
3.4 x 105
(1.0)*
0.0
88
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Continued
85ii/ 8
11
-------
TABLE 2 (concluded)
79
80
81
82
83
84
85
CARB
NO +
H02 H
CARB
CARB
OH +
XX +
Reaction
* ME02 + H02 + CO
ME02 * N02 + HCHO + H02
H R02 *
+ CRIG *
+ MCRG »
METH + HCHO + H02
PAR -» X
Rate Constant
at 298 K
(ppm~* min"1)
1
1.
9.
2.
2.
1.
1.
20 x
00 x
00 x
00 x
60 x
00 x
10
10
10
10
10
10
4
3
3
3
3
5
Activation
Energy
(Qiy \
K /
0
0
0
0
0
0
0
Sunlight-dependent; rate constant is correction factor for OZIPM input,
units of min".
^ Units of ppnfTnin .
Units of min"*.
85il/ 8
12
-------
CARB in the CBM-3 assumed that a fixed percentage (55 percent) of the CARB
reacted as formaldehyde. When substantial amounts of methanol are added
to an urban atmosphere, the fraction of carbonyls that is formaldehyde may
change from this fixed 55 percent. The formaldehyde ratio can also change
with the time of day. Hence, the explicit treatment of formaldehyde is an
improvement in accuracy over the original CBM-3 and is more appropriate
for assessing the impacts of methanol substitution on oxidant formation.
Finally, the separate treatment of formaldehyde chemistry is useful for
assessing the changes in the formaldehyde concentrations themselves
because formaldehyde is a toxic substance.
Additional improvements in the CBM-3M chemistry are associated with
the radical species ME02, R02, and AC03. First, all organic peroxyl radi-
cals containing more than one carbon are lumped into a new species and
ME02 is used exclusively for the methylperoxy radical. In the CBM-3
(Table 1), ME02 represents all organic peroxyl radicals, whereas in CBM-3M
(Table 2), the acylperoxy radical, AC03, reacts to form ME02, which yields
100 percent formaldehyde from PAN decomposition in the presence of NO
(reaction 40 followed by reaction 26). In the original CBM-3, the ME02
formation implied only 55 percent formaldehyde (reactions 40 and 26).
Although 100 percent formaldehyde represents an overestimation of formal-
dehyde production because other PAN-like compounds can lead to carbonyls
other than formaldehyde, most atmospheric measurements show that PAN
itself is by far the main peroxyacetylnitrate seen in urban atmospheres.
Hence, the 55 percent formaldehyde value in the original CBM-3 is an
underestimate.
The main parameter adjustments in the modified CBM-3M are found in
reactions 29, 30, and 31 of Table 2. The ratios of these three pathways
control the recirculation of R02 and the formation of organic nitrates.
The former represents the unimolecular decomposition and isomerization
reactions of alkoxy radicals. The organic nitrate formation pathway was
revised somewhat from that of the CBM-3 because the methylperoxy radicals,
which are treated explicitly as ME02 in CBM-3M, form very few nitrates.
The final values chosen for reactions 29, 30, and 31 in the CBM-3M were
based more on their agreement with the original CBM-3 than on fundamental
principles since such a fundamental approach was beyond the scope of this
project. The overall reactivity of the CBM-3 and CBM-3M was tested by
comparing two EKMA runs that used the two different mechanisms and identi-
cal inputs from data taken on 24 June 1980 in Philadelphia. Table 3 lists
hourly ozone predictions for both EKMA runs and shows that results
obtained with the two mechanisms agree within 8 percent. The pathway
ratios of reactions 29, 30, and 31 were adjusted within the range of
fundamental uncertainty to match the overall reactivity of the CBM-3.
85117T 3 13
-------
TABLE 3. HOURLY OZONE PREDICTIONS FOR
CBM-3 and CBM-3M EKMA SIMULATIONS
Ozone (ppm)
Hour CBM-3 CBM-3M
0800 9 x 10'3 9 x 10"3
0900 1.63 x 10"2 1.70 x 10'2
1000 3.21 x 10'2 3.40 x 10'2
1100 4.84 x 10'2 5.15 x 10'2
1200 6.88 x 10'2 7.32 x 10"2
1300 9.53 x 10'2 1.00 x 10'1
1400 1.24 x 10"1 1.28 x 10"1
1500 1.45 x 10'1 1.47 x ID'1
1600 1.55 x 10'1 1.577 x 10'1
1700 1.65 x 10"1 1.68 x 10-1
1800 1.71 x 10"1 1.74 x 10"1
85ii?r 5
-------
DESCRIPTION OF THE URBAN AIRSHED MODEL
The Urban Airshed Model (UAM) simulates the major physical and chemi-
cal processes associated with ozone formation in the polluted tropo-
sphere. These processes include gas-phase chemistry, advective transport,
and turbulent diffusion. The UAM model domain is divided into a large
array of grid cells (Figure 1). The horizontal cells are uniform 5 x 5 km
squares. Typically, four or five layers of cells represent the vertical
domain. The depth of the cells is determined by the height of the mixed
layer and the height of the top of the modeling domain. The mixed layer
typically ranges from as low as 50 m in the morning hours to 1000 m or
more in the afternoon. Emissions are injected into individual cells
depending on the location of the sources, their height release, and the
buoyant rise of individual stack gas plumes.
The UAM is first operated for some base-case situation to identify a
point in the gridded array at which an important feature occurs in both
space and time. For this project, the important feature was the time and
place of the maximum ozone concentration occurring within the entire UAM
modeling region. Other important features might have been the monitoring
station site showing the highest one-hour maximum ozone concentration or
the grid point in space and time at which the highest formaldehyde concen-
tration was simulated to occur.
DESCRIPTION OF THE SYSTEMS APPLICATIONS TRAJECTORY MODEL
AND THE BOX MODEL USED IN THE PHILADELPHIA STUDY
Two models were used in conjunction with the UAM to assess the
effects of methanol use in Philadelphia; the Systems Applications tra-
jectory model (Myers et al., 1979) was used to trace and process UAM
inputs. These inputs were then used in a single cell box model, which
served as the primary trajectory model for this study. Using the grid
point in space and time determined by the UAM, the Systems Applications'
cell trajectory model is first run backwards through the UAM wind field in
the mixed layer. Because in this study wind shear was present within the
mixed layer for the days simulated by the UAM, the winds in the two layers
that make up the mixed layer were vector-averaged to define the trajectory
path. This path was then traced backwards in time until the boundary of
the modeling region was reached or until the UAM midnight starting time
if the boundary was not reached by midnight. Two days in 1979, 13 and
19 July, formed the meteorological base cases. The trajectory for the
13 July case started within the gridded region at the UAM midnight
starting time; the trajectory for the 19 July case started from the
eastern boundary at 0700 hours.
85U?r2 3
15
-------
(a) The Area lo be Modeled
1r«fi»»ort
£
I
I
^^^^
- Infers,.,
>
>
^'-^ "^ '^ '^^^/^^-^^l
-^-^^^^' ^ ^^v ;//.,$
^ ^^ ^ ^ s sf sf -7*;,;
///y
//'
//
(b) Specification of the Grid
'1 *
1
ICMortttry 1
Clwttrd bmtiontj ^ _
I
ICMnUtr/
tlt*
-------
As the intermediate trajectory model runs forward in time, it traces
the gridded emissions inventory following a path one grid cell in width.
The same vertical cell structure and chemical algorithms are used for both
the trajectory model and the UAM. Hence, there are four cells that mix
vertically, react chemically, and entrain fresh emissions using the same
input files and computer code as the UAM itself. The only differences
between the UAM and this intermediate trajectory model, besides the vast
difference in the number of total cells simulated, are the lack of verti-
cal winds and horizontal dispersion in the intermediate trajectory
model. Hence, the lack of agreement between results of the UAM and those
of the intermediate trajectory model for the final important point defin-
ing the termination of the trajectory path becomes a measure of (1) the
sensitivity or importance of the vertical winds and (2) horizontal dis-
persion. Unfortunately, the presence of wind shear adds a confusing fac-
tor to this measure of sensitivity because the "proper" weighting of the
vector-averaging used to define the trajectory path is uncertain and wind
shear adds a large amount of horizontal dispersion. A simple weighting
scheme of equal contribution of each layer was used for t.nis study.
The primary trajectory model used in this study is a box (i.e.,
single-cell) model developed under previous EPA contracts designed to test
various components of the Empirical Kinetic Modeling Approach (EKMA). The
first of these projects was reported in Whitten et al. (1981). A more
recent model application that assessed methanol substitution in the Los
Angeles area is described by Whitten and Hogo (1983). Although much of
the computer code used to run this moving box model is identical to the
code used in EKMA, the model is considerably different from the standard
EKMA model; therefore, the name "EKMA" should not be used to describe the
box model used in this study. The standard EKMA trajectory model is a
stand-alone model that uses inputs generated by prescribed guidelines and
that is dominated by initial conditions at the 0800 starting time. The
inputs, chemical mechanism, and starting time of the box model used in
this study are all identical to those of the UAM.
Although inputs used by the box model are the same as those used
by the UAM, the box model uses only a subset of the UAM inputs. These
simplified input requirements make it easier to run sensitivity tests with
the box model than with UAM and trajectory models. To discover the effect
of changing certain parameters on UAM or trajectory model results, inputs
to the entire grid system must be changed, which can be very expensive.
However, such inputs as emissions, initial conditions, and temperature
need only be changed along the trajectory path if the box model is used.
Before the box model can be run, however, the UAM and trajectory models
must be used at least once to establish appropriate wind fields and a
trajectory path. Thereafter, the box model can be run numerous times at
a minimal cost. Finally, the UAM can be run to confirm interesting cases
discovered from exploratory tests using the box model.
85U7T2 3
17
-------
SECTION 3
DESCRIPTION OF MODEL INPUTS
URBAN AIRSHED MODEL GENERAL INPUT PREPARATION PROCEDURES
The Urban Airshed Model (UAM) is capable of providing the most
accurate air quality predictions obtainable, given the current state of
knowledge of atmospheric processes and available air pollution data.
Almost any aerometric measurements pertaining to a physical or chemical
atmospheric process can be input to the UAM. However, in many instances
it is not possible to establish a model input directly from measured
data. For example, it is not practical to measure the wind speed and
direction in every grid cell; therefore, available wind data can be used
in conjunction with a mathematical interpolation routine to generate
suitable inputs. In some instances, if sufficient data are not available
for a specific parameter, it may be necessary to assume that the value of
the parameter in one area is equal to that measured in another area.
Interpolation between data points can also be used.
In general, data input preparation procedures should be adapted to
the region of interest through algorithms that make the best possible use
of available data. Such algorithms are especially important when the data
are insufficient or otherwise inappropriate to characterize a specific
input. For example, in a Los Angeles study by Reynolds et al. (1973), the
available data were not adequate to develop complete temperature profiles
of the area. Therefore, instead of employing an algorithm that used only
the limited data available, a specific algorithm was developed that used
both available data and the results of previous studies, which indicated
that mixing depth contours should be parallel to the coastline. Straight-
forward mathematical interpolation of the data would have yielded differ-
ent, and less accurate, mixing depths.
Emission, meteorological, and air quality inputs provide further
illustrations of UAM input preparation procedures. Emission inputs are
prepared from available information pertaining to mobile, stationary, and
natural sources of organic compounds, NOX, CO, S02, and particulates.
85117T i»
18
-------
Gridded mobile emission inputs are most conveniently obtained from a
transportation model that calculates traffic volumes and then relates them
to motor vehicle emission factors. If a transportation model is not
available for use in the area of interest, then the Vehicle-Miles-Traveled
(VMT) in each grid cell can be estimated from existing traffic-count data
and used in place of the traffic volume information. Distributed area
source emission inputs (both natural and anthropogenic) can be estimated
from land-use data. Emission rates and plume rise data for large point
sources are usually available from local or state air pollution control
agencies.
The organic compounds emitted from each type of source are appor-
tioned among reactive groups in the UAM through the use of splitting fac-
tors. These factors can be derived from local measurements, or they can
be estimated from data for other similar urban areas.
Preparation of UAM meteorological inputs entails estimation of three-
dimensional wind fields from available surface and upper-level wind
observations. Theoretical wind shear relationships can also be used in
some situations to calculate wind flows aloft. Mixing depth inputs are
generated from vertical temperature soundings and ground-level temperature
observations. Atmospheric stability can be characterized through the use
of insolation or vertical temperature sounding data.
URBAN AIRSHED MODEL INPUTS
Primary inputs to the UAM include point and area source emissions for
seven species (NO, NOg, and five carbon-bond types), initial and boundary
concentrations both at the surface and aloft for eight species (seven
emitted plus Oj), and a variety of meteorological data. These include a
three-dimensional wind field, mixing depths, solar radiation (expressed as
the equivalent N02 photolysis rate), surface temperature, and exposure
class (an indicator of thermal instability). Inputs to the UAM are given
in Table 4.
Air quality inputs are needed to establish UAM initial and boundary
conditions. At the start of the simulation, the pollutant concentrations
of all species must be specified for each grid call. These inputs are
established through the use of existing air quality data and suitable
interpolation techniques. During the simulation, the upwind "background"
concentration of each pollutant must also be specified along the boundary.
These values are obtained either from actual air quality observations or
from estimates of natural background pollutant levels.
85117T i» 19
-------
TABLE 4. INPUTS TO THE URBAN AIRSHED MODEL
(a) Meteorological Inputs
Category
Required Input Data
Wind speed and direction
Temperature
Water concentration
Atmospheric pressure
Insolation
One hour-average surface obser-
vations
Pibal, radiosonde, or tower
observations (as available)
Hourly surface measurements
Radiosonde, acoustic sounder, or
tower observations for hourly
estimates of mixing depths,
vertical temperature gradient,
and atmospheric stability
Relative humidity observations
Atmospheric pressure
observations
Hourly surface observations
(b) -Air Duality Inputs
Category
Required Input Data
Initial conditions
Boundary conditions
One-hour-average ground-level
pollutant concentrations
Vertical pollutant distribution
sounding
Ground-level pollutant concen-
trations at points on the upwind
boundary of the modeling region
Pollutant concentrations aloft
at the top and upwind sides of
the modeling region.
Continued
20
-------
TABLE 4 (Continued)
(c) Emissions and Surface Uptake Inputs
Category
Required Input Data
Motor vehicle emissions
Distributed area source emissions
(including small point sources)
Large point source emissions
Surface uptake process
Gridded hourly emissions of each
pollutant
Gridded hourly emissions of
each pollutant
Source locations
Hourly emissions rate for each
pollutant
Stack height
Heat flux calculated from flow
rate and exit temperature data
Gridded estimates of surface
roughness
Gridded surface uptake vegeta-
tion factors
(d) Miscellaneous Inputs
Category
Required Input Data
Grid specification
Number of grid cells in the x,
y, and z directions
Horizontal grid spacing
Location (UTM coordinates) of
grid origin
Continued
85U7r2 5
21
-------
TABLE 4 (Concluded)
Category
Required Input Data
Program control parameters
Chemical mechanism
parameters
Start and end times for the
simulation
Initial integration time step
size
Interval over which predictions
will he time-averaged (e.g., 60
minutes for one-hour-average
preditions)
Reaction rate constants
Activation energies
85U7P2 5
22
-------
In general, the quantity of data needed by the UAM depends on the
application. The number of measurements needed is related to
(1) The sensitivity of the model to variations in the parameters of
interest. If the model is relatively insensitive to variations
in a parameter, the data used to construct inputs for that
parameter are not as important as when the model is sensitive to
the particular parameter.
(2) The degree of uncertainty in the predicted results that is
acceptable for the application. Model verification generally
requires much more data than does rough estimation of the impact
of a control strategy.
(3) The characteristics of the region, e.g., for studies that
require resolution of the primary features of the wind field,
fewer measurements would be required in flat terrain than in
hilly or mountainous areas.
The UAM inputs used for the Philadelphia methanol study utilized data for
two days--13 and 19 July 1979--previously modeled to evaluate the model as
an air quality planning tool in large metropolitan areas (Haney, 1985).
Although the Philadelphia-methanol UAM inputs are based on the 1979 UAM
application, projections were made to the year 2000. We first present a
summary of the 1979 inputs, followed by a description of the changes made
to these inputs for the year 2000 Philadelphia UAM application.
INPUTS USED IN THE 1979 URBAN AIRSHED MODEL SIMULATIONS
In this section, we summarize the inputs used for each of the two
modeling days.
Emissions
The emission inventory used in the 1979 UAM simulations was developed
by Engineering Science (EPA, 1982). The emission region has an area of
12,500 km and comprises five counties in New Jersey, five counties in
Pennsylvania, and one county in Delaware. The grid system used for the
emission inventory consists of 502 cells, five km on a side. Surrounding
each side of the modeling region is a row of cells representing the
boundary cells and containing the boundary conditions (concentrations)
used in the UAM simulations. Initial conditions for all species were
specified using all available monitoring data in the Philadelphia
region. Initial conditions in surface grid cells without monitors were
85U7P2 it 23
-------
obtained by employing a Poisson interpolation. Following the computation
of the surface field, a vertical interpolation method was employed in
which the background concentration at the top of the modeling region
(TOPCONC) was used in levels 3 and 4, and the level 2 value was obtained
by a linear interpolation between the surface (level 1) value and the
level-3 value. Using this method, all grid cells in all levels were
initialized with appropriate concentrations for all species. For further
discussion of interpolation techniques, refer to The User's Manual for the
SAI Airshed Model, (Ames et al., 1978). The monitored surface conditions
for 13 and 19 July are shown in Tables 5 and 6. The Carbon-Bond fractions
in Tables 5 and 6 were calculated by the EPA using the Philadelphia emis-
sions inventory (EPA, 1982) and the Carbon-Bond system described in Killus
and Whitten (1981).
Boundary Conditions
The physical boundaries and the trajectory path used in the 13 July
simulation are presented in Figure 2. Because of the stagnation charac-
teristics of this episode, much of the large airshed region was "blocked
off" and not included in the simulation; however, no cells containing
major emission sources were excluded by this procedure. Background values
for all species for 13 July were designated for all boundaries except the
southeast boundary (Table 7). Because an urban plume from Philadelphia
was transported to the east late on 12 July and then recirculated back by
a southeasterly flow on 13 July, there is a different set of concentra-
tions for the southeast boundary grid cells below the mixing height (Table
8); these concentrations were made to duplicate the inflow of aged air
parcels from the previous day's emissions.
Because of wind flow through the airshed on 19 July and the need to
limit simulation costs, unnecessary grid cells were eliminated for this
simulation. The physical boundaries and the trajectory path used for the
July 19 simulation are shown in Figure 3. Background values were designa-
ted for all boundaries except the east and northeast boundaries. Esti-
mates of boundary conditions for the east and northeast boundaries below
the mixing height are presented in Table 9. Observed NO and N02 for 0500-
0600 EST were used as input boundary values for 0000-0600. Since hydro-
carbons were not measured at the Van Hiseville monitor, estimates of the
influx of total reactive hydrocarbons across the northeast and east boun-
daries below the mixing height were specified by multiplying the hourly
NOX concentrations at the Van Hiseville monitor by the Philadelphia sur-
face layer emission inventory hydrocarbon/NOx ratio of 6. The total reac-
tive hydrocarbon value was then speciated into carbon-bond components
using the carbon-bond fractions of the emission inventory.
85117T2 it
24
-------
TABLE 5. MONITORED SURFACE CONDITIONS FOR 13 JULY 1979 (CONCENTRATIONS
IN PPM).
Station
AMS Lab.
Ancora
Brlgantine
Bristol
Camden
Chester
Cl aymont
Conshohocken
Defense Support
Downington
Franklin Inst.
Island Rd. Airp. Cir.
Lumberton
Norn's town Armory
Northeast Airp.
Roxy Water Pump
SE Sewage Plant
South Broad
Summit Bridge
SW Corner Broad/Butler
Trenton
Van Hi Seville
Vineland
Easting
(m)
491600
511800
546000
511000
491700
469000
461500
474500
483800
436000
485200
480300
518000
473500
499000
479500
487200
486100
441000
487000
520000
559000
498200
Northing
(m)
4428500
4392400
4377506
4440000
4419000
4410000
4406400
4435600
4418300
4426000
4422800
4414800
4423000
4440000
4436000
4433100
4417300
4421600
4376000
4428000
4452000
4439500
4371200
NO
0
-
-
-
0
-
-
-
-
0
0
-
0
-
0
-
-
0
0
-
-
0
-
.005
-
-
-
.022
-
-
-
-
.002
.010
-
.006
-
.035
-
-
.020
.000
-
-
.001
-
N02 CO
0.
--
0.
0.
0.
--
0.
0.
0.
0.
0.
--
0.
0.
--
--
0.
0.
075 1.5
--
105 2.4
030 00
045
012 0.2
065
031 1.2
075 --
060
080 3.5
003 0.2
006 0.2
024
0
0
0
0
0
0
-
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
°3
.035
.051
.037
.000
.004
.058
-
.000
.020
.073
.035
.010
.010
.000
.000
.045
.010
.035
.046
.035
.009
.005
.054
RHC
1.05
--
--
2.10
--
0.00
0.20
0.39
--
0.30
0.05
--
Carbon-Bond Fraction
RHC Component (%
PAR
OLE
ETH
ARO
CARB
as
74
2
4
13
5
Carbon)
.0
.8
.1
.2
.9
85117T 5
25
-------
TABLE 6. MONITORED SURFACE CONDITIONS FOR 19 JULY 1979 (CONCENTRATIONS
IN PPM).
Station
AMS Lab.
Ancora
Brigantlne
Bristol
Camden
Chester
Cl aymont
Conshohocken
Defense Support
Downington
Franklin Inst.
Island Rd. Airp. Cir.
Lumberton
Norristown Armory
Northeast Airp.
Roxy Water Pump
SE Sewage Plant
South Broad
Summit Bridge
SW Corner Broad/Butler
Trenton
Van Hi Seville
Vineland
Easting
(««0
491600
511800
546000
511000
491700
469000
461500
474500
483800
436000
485200
480300
518000
473500
499000
479500
487200
486100
441000
487000
520000
559000
498200
RHC
Component
PAR
OLE
ETH
ARO
CARB
Northing
(m)
4428500
4392400
4337506
4440000
4419000
4410000
4406400
4435600
4418300
4426000
4422800
4414800
4423000
4440000
4436000
4433100
4417300
4421600
4376000
4418000
4452000
4439500
4371200
NO
0.
--
0.
--
--
0.
0.
«
0.
0.
~
0.
0.
0.
000
Oil
006
030
000
000
020
000
000
0
-
-
0
0
0
0
-
-
0
0
-
0
-
0
-
-
0
0
-
-
0
0
N02 CO
.025 0.5
-
_
.039
.051 2.5
.024
.045
.
-
.015 0.2
.050 0.5
_
.028 0.0
-
.015
_
-
.045 1.5
.009 0.2
_
_
.006 0.2
.025
0
0
0
0
0
0
0
0
0
0
0
0
0
0
-
0
0
0
0
-
0
0
0
°3
.015
.015
.017
.012
.004
.026
.020
.000
.000
.024
.010
.015
.008
.030
-
.010
.005
.010
.009
-
.002
.021
.010
RHC
0.35
--
--
0.95
0.05
0.40
0.29
--
0.25
0.10
--
Carbon-Bond Fraction
(% as Carbon)
74.
2.
4.
13.
5.
0
8
1
2
9
85117T 5
26
-------
0
30
20
(D
UJ
10
n
0
NORTH
10
20
30
1600 hr
NORR
C°NRS0XY
DOHN
0000 hr
CHES
CLHY
SUMM
.t . \ t.. i .1 . i...i..i.. t.. .i. .i...t..i.. j.. A. .1.. .i. .t.: J^^=^
10
20
30
SOUTH
TREM ROBB
BR1S
20
***!"*
"*"?"*
LUHB iij CD
CL
:::: ::S
flNco ....... "C""'"''!'!-!' ?'?'"'"
10
:::H
30
FIGURE 2. Trajectory path for July 13 regional ozone maximum.
-------
TABLE 7. BACKGROUND CONCENTRATION VALUES
FOR 13 JULY AT THE TOP OF THE MODELING REGION
(TOPCONC), AS INITIAL CONCENTRATIONS ABOVE THE
MIXING HEIGHT, AND FOR ALL LEVELS OF ALL
BOUNDARIES EXCEPT THE LEVELS BELOW THE MIXING
HEIGHT ON THE SOUTHEAST BOUNDARY.
Concentration
Species (ppm)
NO 0.001
N02 0.002
03 0.08
CO 0.2
ETH 0.001
OLE 0.0004
PAR 0.040
CARB 0.010
ARO 0.0008
PAN 0.000025
BZA 0.00001
85117T 5 28
-------
TABLE 8. SOUTHEAST BOUNDARY CONDITIONS FOR CELLS BELOW THE MIXING
HEIGHT FOR THE SIMULATION OF 13 JULY 1979 (CONCENTRATIONS IN PPM)
Time
interval
0000-0100
0100-0200
0200-0300
0300-0400
0400-0500
0500-0600
0600-0700
0700-0800
0800-0900
0900-1000
1000-1100
1100-1200
1200-1300
1300-1400
1400-1500
1500-1600
1600-1700
1700-1800
NO
0.001
0.001
0.001
0.001
0.001
0.001
0.001
0.001
0.001
0.001
0.001
0.001
0.001
0.001
0.001
0.001
0.001
0.001
N02
0.009
0.009
0.009
0.009
0.009
0.009
0.009
0.007
0.004
0.002
0.002
0.002
0.002
0.002
0.002
0.002
0.002
0.002
°3
0.053
0.041
0.037
0.036
0.028
0.010
0.021
0.035
0.072
0.102
0.111
0.121
0.139
0.113
0.105
0.093
0.069
0.068
PAR
0.1041
0.1041
0.1041
0.1041
0.1041
0.1041
0.1041
0.0983
0.0904
0.0825
0.0758
0.0698
0.0644
0.0596
0.0546
0.0525
0.0506
0.0506
OLE
0.00105
0.00105
0.00105
0.00105
0.00105
0.00105
0.00105
0.00065
0.00040
0.00040
0.00040
0.00040
0.00040
0.00040
0.00040
0.00040
0.00040
0.00040
ETH
0.0026
0.0026
0.0026
0.0026
0.0026
0.0026
0.0026
0.0022
0.0018
0.0015
0.0012
0.0010
0.0010
0.0010
0.0010
0.0010
0.0010
0.0010
ARO
0.0021
0.0021
0.0021
0.0021
0.0021
0.0021
0.0021
0.0017
0.0012
0.0008
0.0008
0.0008
0.0008
0.0008
0.0008
0.0008
0.0008
0.0008
CARB
0.0261
0.0261
0.0261
0.0261
0.0261
0.0261
0.0261
0.0259
0.0244
0.0218
0.0188
0.0157
0.0129
0.0107
0.0100
0.0100
0.0100
0.0100
PAN
0.0040
0.0040
0.0040
0.0040
0.0040
0.0040
0.0040
0.0044
0.0054
0.0062
0.0062
0.0058
0.0052
0.0047
0.0040
0.0037
0.0032
0.0032
85117T 5
-------
OJ
o
TABLE 8 (Concluded)
Time
interval
1800-1900
1900-2000
2000-2100
2100-2200
2200-2300
2300-2400
NO
0.001
0.001
0.001
0.001
0.001
0.001
N02
0.002
0.002
0.002
0.002
0.002
0.002
°3
0.058
0.065
0.059
0.056
0.051
0.050
PAR
0.0506
0.0506
0.0506
0.0506
0.0506
0.0506
OLE
0.0004
0.0004
0.0004
0.0004
0.0004
0.0004
ETH
0.001
0.001
0.001
0.001
0.001
0.001
ARO
0.0008
0.0008
0.0008
0.0008
0.0008
0.0008
CARB
0.010
0.010
0.010
0.010
0.010
0.010
PAN
0.0032
0.0032
0.0032
0.0032
0.0032
0.0032
85117T 5
-------
0
30
20
in
UJ
10
10
DOHN
NORTH
20
30
SUHH
1700 hr
flMS
CHE3
LUHB
I3LH
CLBY
RNCO
. i.. t.. .1.. .t.. j... i... t. liii^^
30
20
en
cr
UJ
10
10
20
30
i!0
SOUTH
FIGURE 3. Trajectory path for July 19 regional ozone maximum.
-------
HEIGHT ESTIMATED FROM DATA COLLECTED AT THE VAN HISEVILLE, NEW JERSEY MONITOR
(CONCENTRATIONS IN PPM)
GO
r\>
Hour (EST)
0000-0100
0100-0200
0200-0300
0300-0400
0400-0500
0500-0600
0600-0700
0700-0800
0800-0900
0900-1000
1000-1100
1100-1200
1200-1300
1300-1400
1400-1500
1500-1600
1600-1700
1700-1800
1800-1900
1900-2000
2000-2100
2100-2200
2200-2300
2300-2400
NO
0.044
0.044
0.044
0.044
0.044
0.044
0.047
0.043
0.026
0.009
0.001
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.002
0.005
0.008
N02
0.036
0.036
0.036
0.036
0.036
0.036
0.041
0.0494
0.0478
0.0273
0.0099
0.0058
0.0054
0.0072
0.0047
0.0035
0.0043
0.0053
0.0072
0.0071
0.0041
0.0030
0.0060
0.0030
°3
0.018
0.009
0.000
0.000
0.000
0.000
0.000
0.004
0.015
0.051
0.083
0.078
0.075
0.072
0.072
0.062
0.059
0.052
0.044
0.036
0.020
0.010
0.000
0.001
PAR
0.3552
0.3552
0.3552
0.3552
0.3552
0.3552
0.39072
0.40996
0.32782
0.16132
0.04839
0.02575
0.02397
0.03196
0.02087
0.01554
0.01909
0.0235
0.03197
0.03152
0.0182
0.0222
0.04884
0.04884
Philadel
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
phi
OLE
00672
00672
00672
00672
00672
00672
00739
007755
0062
00305
000915
000487
0004535
000605
0003945
000294
000361
000445
000605
0005965
0003445
00042
000925
000925
ETH
0.00985
0.00985
0.00985
0.00985
0.00985
0.00985
0.010825
0.01135
0.00908
0.00447
0.00134
0.000716
0.000665
0.000885
0.000578
0.0004305
0.000529
0.000652
0.000885
0.000875
0.000505
0.000615
0.001355
0.001355
ARO
0.01056
0.01056
0.01056
0.01056
0.01056
0.01056
0.01162
0.01218
0.00975
0.0048
0.00143
0.000767
0.000717
0.00095
0.00062
0.000462
0.000568
0.0007
0.00095
0.000937
0.000542
0.00066
0.00145
0.00145
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
CARB
.02832
.02832
.02832
.02832
.02832
.02832
.03115
.03269
.02614
.01286
.00386
.00205
.00191
.00255
.00166
.00124
.00152
.001876
.00255
.00251
.00145
.00177
.00389
.00389
PAN
0.0002
0.0001
0.0000
0.0000
0.0000
0.0000
0.0000
0.0004
0.0008
0.0013
0.0019
0.0018
0.0014
0.0012
0.0007
0.0005
0.0003
0.0003
0.0002
0.0001
0.0001
0.0000
0.0000
0.0000
a Inventory Split
Carbon-Bond Fraction
Species
PAR
OLE
ETH
ARO
CARS
(%
as Carbon)
74.
2.8
4.1
13.2
5.9
-------
Mixing Height Profiles
Mixing height profiles for 13 July were estimated using temperature
soundings from JFK and Dulles airports since no soundings were available
for Philadelphia on this day. Mixing heights for 13 July are shown in
Table 10 and Figure 4. Mixing height profiles for 19 July were developed
using available radiosonde observations and sodar data for the Phila-
delphia area. Mixing heights for 19 July are shown in Table 11 and Figure
5.
Wind Fields
Three-dimensional wind fields for 13 July and 19 July were generated
from station measurements and aloft winds obtained from radiosonde data.
Wind direction on the morning of 13 July was very light and predominantly
from the north. Around 1000 EST, however, a shift occurred and winds with
higher speeds came from the southeast for the rest of the day. Winds on
the morning of 19 July were predominantly from the north. Throughout the
day a 180° shift occurred: at noon the wind direction was predominantly
from the east; during the evening it came from the south-southeast. The
gridded, smoothed surface vectors for selected hours are presented in
Figure 6 for 13 July and in Figure 7 for 19 July.
Background Concentrations
Except for the ozone concentration, the background concentrations
above the mixing height are the same for both days. On the basis of
examination of upwind monitoring data at the time of mixing, the value
specified for ozone for 19 July is 0.06 ppm, compared to 0.08 ppm for 13
July. Background concentrations for July 19 are listed in Table 12. The
gridded area source and elevated point source emission inventory was pre-
pared for EPA in 1981 by Engineering Science, Inc. (EPA, 1982). Total
daily emission values for total NOX and total hydrocarbon are presented in
Table 13.
MODIFICATIONS TO 1979 UAM INPUTS FOR PHILADELPHIA APPLICATION
For the year 2000 base case projections, the changes made to the 1979
UAM inputs involved mobile source emissions, major point, minor point and
area source emissions, and initial and boundary conditions. When modify-
ing the year 2000 base case for methanol substitution, changes made to
mobile source emissions and fueling operation emissions were limited to
85117r i* 33
-------
TABLE 10. URBAN AND RURAL
MIXING HEIGHT VALUES USED
IN THE DIFFBREAK FILE FOR
13 JULY 1979.
Time
(EST)
0000
0030
0100
0130
0200
0230
0300
0330
0400
0430
0500
0530
0600
0630
0700
0730
0800
0830
0900
0930
1000
1030
1100
1130
1200
1230
1300
1330
1400
1430
1500
1530
1600
1630
1700
1730
1800
1830
1900
1930
2000
2030
2100
2130
2200
2230
2300
2330
2400
Urban
(m)
250
250
250
250
250
250
250
250
250
250
250
250
250
270
295
375
450
680
925
1160
1200
1330
1480
1480
1480
1500
1530
1530
1530
1475
1420
1365
1310
1220
1130
1035
950
855
770
675
590
500
410
370
330
290
250
250
250
Rural
(m)
100
100
100
100
100
100
100
100
100
100
100
100
100
135
150
250
350
620
925
1160
1200
1330
1480
1480
1480
1500
1530
1530
1530
1475
1420
1365
1310
1220
1130
1035
950
730
525
320
100
100
100
100
100
100
100
100
100
34
85117T 5
-------
1700
1600
1500
1100
1300
-1200
o>1100
1000 -
.*>
-
0)
X
Legend:
_ Urban
... Rure I
0
12
Time thours)
FIGURE 4. Mixing heiqht profiles for urban and
rural cells for the 13 July 1979 simulation.
B3033r
35
-------
TABLE 11. URBAN AND RURAL
MIXING HEIGHT VALUES USED
IN THE DIFFBREAK FILE FOR
19 JULY 1979.
Time
(EST)
0000
0030
0100
0130
0200
0230
0300
0330
0400
0430
0500
0530
0600
0630
0700
0730
0800
0830
0900
0930
1000
1030
1100
1130
1200
1230
1300
1330
1400
1430
1500
1530
1600
1630
1700
1730
1800
1830
1900
1930
2000
2030
2100
2130
2200
2230
2300
2330
2400
Urban
(m)
190
190
190
190
190
190
190
190
190
190
190
190
190
190
190
240
350
460
600
740
890
1060
1240
1420
1530
1530
1480
1410
1340
1270
1200
1120
1020
940
850
750
660
570
480
410
340
300
250
250
250
250
250
250
250
Rural
(m)
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
260
440
490
540
560
580
630
880
1320
1430
1430
1380
1310
1240
1170
1100
1020
920
840
750
480
160
100
100
100
100
100
100
100
100
100
100
100
100
85117P 5 36
-------
I I I | I I I I I | I I I I I | I I I
Legend:
^_ Urban
... Ruro I
12
Time Ihoursl
FIGURE 5. Mixing height profiles for urban and
rural cells for the 19 July 1979 simulation.
63033.
37
-------
I I I I I 1
f S
HIND SfEED (M/S)
e
15
f\
fffffttfttftlt
25
15
10
0
sssss / r iff ///
35 L'
0
10 15 20
(a) 0 - 100 EST
25
30
FIGURE 6. Airshed model surface winds for 13 July 1979.
B3033r
38
-------
0
10
15
5
MIND SPEED CH/SJ
25
.' i «, ij ', i, «, \4 i, 1,
l4U444**
4 4 444U**
4-4444444H*"'"***
4-4 4 44 4 4 4 m <" « * 4
25
15
B
£//////////"""/////v//;
4/////////////«"'*////?#}
44444444444
41444444444
444444441*1
/* t ****
J**'**
vl vl vl v
*-^***+^~*~f»*~it
t«"| » |-» ^^^^»t^4^s^f>4^.
10 15 20
(b) 600 - 700 EST
25
35
FIGURE 6 (continued)
63033r
39
-------
<\\\\\\\
35
- 1300 EST
FIGURE 6 (continued)
40
-------
I I I I I I
IB
0,
IB
15 20
(d) 1800 - 1900 EST
25
30
35
FIGURE 6 (concluded)
63D33r
41
-------
0
15 20
5
N1ND SPEED IK/S)
25
35
(a) 400 - 500 EST
FIGURE 7. Airshed model surface winds for 19 July 1979.
-32
83033r
42
-------
0
I I I I I I
s
MIND SPEED (H/S)
20 25
30
"0
10
15 20 25
(b) 1200 - 1300 EST
30 35
FIGURE 7 (continued)
63C33r
43
-------
B.
0
FIGURE 7 (concluded).
15 20 25
(c) 1800 - 1900 EST
30 35
B3033r
44
-------
TABLE 12. BACKGROUND CONCENTRATIONS VALUES
USED AT THE TOP OF THE MODELING REGION
(TOPCONC), AS INITIAL CONCENTRATIONS
ABOVE THE MIXING HEIGHT, AND FOR ALL
LEVELS OF ALL BOUNDARIES EXCEPT THE LEVELS
BELOW THE MIXING HEIGHT ON THE NORTHEAST
AND EAST BOUNDARIES
Concentration
Species (ppm)
NO
N02
°3
CO
ETH
OLE
PAR
CARB
ARO
PAN
BZA
0.001
0.002
0.06*
0.2
0.001
0.0004
0.040
0.010
0.0008
0.000025
0.00001
A value of 0.05 ppm was used below
the mixing heignt for all boundaries
except the Northeast and East
boundaries.
85117P 5
-------
TABLE 13. TOTAL DAILY EMISSIONS BY SOURCE TYPE (g-mole) IN THE 1979 PHILADELPHIA INVENTORY
cr>
Source
Elevated Point
Minor Point
Area Source
Mobile Source
TOTAL
NO N02
5,738,955 257,086
2,041,679 114,837
2,705,097 207,808
6,646,688 738,496
17,132,419 1,318,227
Total HC = 63,832,
PMladel
Species
PAR
OLE
ETH
ARO
CARB
ETH OLE PAR CARB
12,144 7,792 901,823 570,765
385,284 183,632 15,846,729 818,880
295,258 194,471 26,841,639 1,669,555
827,355 664,553 11,635,952 1,337,697
1,520,041 1,050,448 55,226,197 4,396,897
101 Total NOX = 18,450,646
phla Hydrocarbon Splits
Carbon-Bond Fraction
(% as Carbon)
74.0
2.8
4.1
13.2
5.9
ARO
139,960
510,799
491,833
495,926
1,638,518
85117T 8
-------
area sources. For selected scenarios, methanol was also substituted for a
portion of initial and aloft concentrations.
This section discusses in further detail the modifications made to
the 1979 DAM inputs. The simulations chosen for this study are also
discussed. The 1979 inventory contains hourly mobile source emissions for
each of the 502 grids in the modeling region as well as daily mobile
source emissions for each county. It would have been a massive undertak-
ing in terms of resources and time to revise the hourly emissions on a
grid-to-grid basis. Instead, rough percentage reductions in the 1979
mobile source inventory were made for the year 2000 projections. This was
done for each of the 11 counties in the Philadelphia AQCR.
Exhaust HC, evaporative HC, and NOX emission factors (g/mile) for the
year 2000 for each of the 11 counties were determined using MOBILES and
were then multiplied by the appropriate year-2000 daily VMT to obtain
kg/day emissions for each county for the year-2000 base case. The MOBILES
input parameters specific to each county (e.g., average speed) as well as
the year-2000 projected daily VMT for each county are contained in the
Delaware Region's Year-2000 transportation plan (DVRPC, 1982). The
exhaust HC emissions predicted by MOBILES were multiplied by 1.016 to
account for aldehydes and ketones. The value of 1.016 is based on VOC
data for 1981 and 1982 vehicles (Sigsby et al., 1984).
County adjustment factors, which are ratios of the year 2000 to the
year 1979 kg/day emissions, were then estimated. These adjustment factors
are listed in Table 14. The adjustment factors given for each county were
then used for each grid within that county. The adjustment factors were
multiplied by the appropriate mass-per-grid-cell values contained in the
1979 mobile source inventory to obtain revised year 2000 mass-per-grid-
cell values for exhaust HC, evaporative HC and NOX.
County adjustment factors for the methanol replacement scenarios
differ from those calculated for the base case because different assump-
tions were made regarding the mass of exhaust and evaporative emissions
from methanol-fueled vehicles. NOX emissions from methanol-fueled
vehicles were assumed to be the same as those of the base case.
The specific box model scenarios are listed in Table 15. The UAM
runs are more limited in scope, i.e., not all scenarios were simulated:
85117T
47
-------
TABLE 14. COUNTY ADJUSTMENT FACTORS
County
Burlington
Exn. HC
Evap. HC
NOX
Camden
Exn. HC
Evap. HC
NOX
Gloucester
Exn. HC
Evap. HC
NOX
Mercer
Exn. HC
Evap. HC
NOX
Bucks
Exn. HC
Evap. HC
NOX
Chester
Exn. HC
Evap. HC
NOX
Delaware
Exn. HC
Evap. HC
NOX
Montgomery
Exn. HC
Evap. HC
NOX
Philadelphia
Exh. HC
Evap. HC
NOX
Salem
Exh. HC
Evap. HC
NOX
New Castle
Exh. HC
Evap. HC
NOX
1979
Base Case
(kg/day)
14127
9108
33068
16039
9188
31855
6899
4829
18247
10453
6007
21208
20646
9569
33553
14372
7057
25443
21265
8421
27610
33108
14059
48105
51702
17096
55725
4078
3218
11482
23149
9636
34250
2000
Base Case
1A
0.31
0.45
0.39
0.36
0.48
0.43
0.31
0.46
0.38
0.34
0.50
0.44
0.25
0.47
0.41
0.24
0.46
0.40
0.29
0.52
0.50
0.28
0.51
0.46
0.28
0.48
0.49
0.33
0.47
0.42
0.25
0.47
0.41
Complete
2A
0.31
0.14
0.39
0.36
0.14
0.43
0.31
0.14
0.38
0.34
0.15
0.44
0.25
0.14
0.41
0.24
0.15
0.40
0.29
0.16
0.50
0.28
0.16
0.46
0.28
0.14
0.49
0.33
0.14
0.42
0.25
0.14
0.41
Methanol
2B
0.78
0.35
0.39
0.90
0.35
0.43
0.78
0.35
0.38
0.85
0.37
0.44
0.62
0.35
0.41
0.60
0.37
0.40
0.72
0.40
0.50
0.70
0.40
0.46
0.70
0.35
0.49
0.82
0.35
0.42
0.62
0.35
0.41
Replacement
2C
1.55
0.70
0.39
1.80
0.70
0.43
1.55
0.70
0.38
1.70
0.75
0.44
1.25
0.70
0.41
1.20
0.75
0.40
1.45
0.80
0.50
1.40
0.80
0.46
1.40
0.70
0.49
1.65
0.70
0.42
1.25
0.70
0.41
Revised 2000
Base Case
IB
0.34
0.14
0.39
0.32
0.14
0.43
0.38
0.14
0.38
0.33
0.15
0.44
0.25
0.14
0.41
0.26
0.15
0.40
0.24
0.16
0.50
0.25
0.16
0.46
0.18
0.14
0.49
0.44
0.14
0.42
0.23
0.14
0.41
85117T 5
48
-------
TABLE 15a. BOX MODEL SCENARIOS FOR 13 JULY 2000
1. Base Case
1A. Year 2000 base case
IB. Revised mobile source year 2000 base case (see text for explanation)
2. Complete methanol replacement scenario for mobile sources
2A. Same exhaust HC as base case (1A), evap at standards
2B. 2.5 times base case exhaust HC (1A), evap 2.5 times standards
2C. 5.0 times base case exhaust HC (1A), evap 5 times standards
2D. 2B with surface deposition
2E. 2B with 20 % methanol for aloft and initial conditions
3. Partial methanol replacement scenario for mobile sources
3A. 80% 1A, 20% 2A
3B. 80% 1A, 20% 2B
3C. 80% 1A, 20% 2C
4. Formaldehyde sensitivity analysis for 28
4A. 5% of methanol exhaust emissions
4B. 20% of methanol exhaust emissions
5. Formaldehyde sensitivity analysis for 2C
5A. 5% of methanol exhaust emissions
5B. 20% of methanol exhaust emissions
6. Ozone isopleth diagrams based on 2B
7. Removal of all mobile source emissions
7A. From base case 1A
7B. Base NOX, zero VOC from mobile sources
8. Stationary source sensitivity analysis for 2B
8A. Elimination of mobile-related stationary source methanol contribution
8B. Upper limit assumption for mobile-related stationary source methanol
contribution; substitution of methanol (no formaldehyde) for 50% of all
VOC emissions
8C. 8B with 50% methanol in initial conditions
80. 88 with 50% methanol aloft and in initial conditions
9. Carbon percent replacement scenarios (all mobile source HC replaced with
methanol or formaldehyde on a C% basis)
9A. 100% of base case (IB) carbon replaced with methanol (no formaldehyde)
9B. 150% of base case (IB) carbon replaced with methanol (no formaldehyde)
9C. 200% of base case (IB) carbon replaced with methanol (no formaldehyde)
90. 10% of base case (IB) carbon replaced with formaldehyde (no methanol)
9E. 20% of base case (IB) carbon replaced with formaldehyde (no methanol)
49
-------
TABLE 15a (Concluded)
10. 50% reduction of total base case (IB) HC
11. 30% mobile source emissions of the total base case 1A emissions
11A. Revised year 2000 base case
11B. Revised complete methanol substitution scenario based on 2B with initial
methanol conditions
12. 50% mobile source emissions of the total base case 1A emissions
12A. Revised year-2000 base case emissions
12B. Revised complete methanol substitution scenario based on 2B with initial
methanol conditions
50
8511
-------
TABLE 15b. METHANOL, FORMALDEHYDE, METHYL NITRATE, AND HYDROCARBON
FOR BOX MODEL SCENARIOS AS PERCENT CARBON OF BASE-CASE
IB MOBILE SOURCES.**
1.
Scenario
Base Case
IB
1A
Methanol*
0
0
Formaldehyde*
0
0
Methyl Nitrate*
0
0
Hydrocarbons*
100
151
8.
Complete methanol
replacement scenario
2A
2B
2C
2D
2E
Partial methanol
replacement scenario
3A
3B
3C
Formaldehyde sensitivity
analysis (2B)
4A
4B
Formaldehye sensitivity
analysis (2C)
5A
5B
Mobile-related
stationary-source
sensitivity analysis
(2B)
63
158
315
158
158
7.33
18.2
36.6
18.2
18.2
0.36
0.90
1.80
0.90
0.90
12.6
31.5
63
1.46
3.65
7.30
0.07
0.18
0.36
167
140
333
280
9.10
36.6
183
73.3
0.90
0.90
1.80
1.80
0
0
0
0
0
121
121
121
0
0
0
0
8A
8B
8C
8D
151
104
104
104
17.6
12.1
12.1
12.1
0.04
0.59
0.59
0.59
0
0
51
Continued
-------
TABLE 15b (Concluded)
Scenario Methanol* Formaldehyde* Methyl Nitrate* Hydrocarbons*
9. Carbon percent
replacement scenarios
9A 100 0 0 0
9B 150 0 0 0
9C 200 0 0 0
9D 0 10 0 0
9E 0 20 0 0
10. 50% reduction of
base case IB mobile
source HC
50
11. 30% of total IB
emissions***
11A 000 151
11B 158 18.2 0.9 0
12. 50% of total
emissions***
12A 000 151
12B 158 18.2 0.9 0
* For IB base case mobile sources only.
** See Table 15a for complete scenario descriptions.
*** Mobile source contribution was increased to 30% by lowering the total emissions.
8511 7>* 5
-------
Scenarios
Simulated by the UAM
13 July 2000 19 July 2000
1A 1A
2B
3B
5A
The year 2000 base case 1A is based on the MOBILES projections. Base case
IB assumed that gasoline- and diesel-fueled vehicles in the year 2000 will
meet their projected exhaust and evaporative emission standards.
The second scenario assumes complete replacement of gasoline- and
diesel-fueled vehicles with methanol-fueled vehicles. Scenarios 2A, 2B,
and 2C differ with respect to the quantity of exhaust and evaporative HC
emissions assumed to be emitted by methanol-fueled vehicles. Scenario 2A
assumes that methanol-fueled vehicles will meet current exhaust and
evaporative emission standards. For example, the mass of exhaust HC emis-
sions (composed of methanol, formaldehyde, and methyl nitrite) from a
light-duty methanol-fueled vehicle total 0.42 g/mile (0.41 x 1.016).
Scenarios 28 and 2C assume that both exhaust and evaporative emissions, on
a mass basis, will equal 2.5 and 5 times the standards, respectively. It
was anticipated that these scenarios would bracket the range of emissions
possible from in-use methanol-fueled vehicles.
Scenario 3 is a partial methanol replacement scenario in which 20
percent of the gasoline- and diesel-fueled vehicles are replaced with
methanol-fueled vehicles. For this scenario, the mobile source inventory
is composed of 80 percent of the base case inventory (Scenario 1A), and 20
percent of the complete replacement inventory (Scenario 2) on a mass
basis. Three emissions inventories were examined for Scenario 3 since
three emissions inventories were available for the complete replacement
scenario on the basis of the three in-use methanol emission rates assumed.
The remaining scenarios were chosen to examine the sensitivity of the
model predictions to formaldehyde, stationary source, and mobile source
emissions in general. For Scenario 10, methanol or formaldehyde replace-
ment of mobile source VOC was made on a per carbon basis. This procedure
differs from that of the other scenarios, in which methanol substitution
was done on a mass basis. Scenario 10 was included to provide a more
direct comparison of the results of this study with results obtained in
the previous modeling study of methanol substitution in the Los Angeles
basin.
85117T2 i+
-------
The hydrocarbon splits for motor vehicles contained in the 1979
inventory are based on data from 1979 and earlier model-year light-duty
vehicles. For the year 2000 inventory, revised HC speciation data for
motor vehicles were incorporated, based on emissions from in-use 1982
light-duty vehicles (Sigsby et al., 1984). It was believed that the com-
position of emissions from in-use 1982 vehicles would more closely
approximate the composition of vehicle emissions in the year 2000 than
that contained in the 1979 Inventory. The hydrocarbon splits for motor
vehicles in the 1979 inventory can be compared to the hydrocarbon splits
for motor vehicles in the year 2000 inventory in Table 16. The hydrocar-
bon reactivity was somewhat reduced in the year 2000 as evidenced by the
increase in paraffins and the decrease in olefins, carbonyls, and ethy-
lenes, in spite of an increase in the fairly reactive aromatics.
For the complete methanol replacement scenarios, the exhaust HC emis-
sions were replaced by methanol, formaldehyde, and methyl nitrite. The
composition of the exhaust emissions on a mass basis is 89 percent
methanol, 10 percent formaldehyde and one percent methyl nitrite. The
composition of the evaporative HC emissions was assumed to be 100 percent
methanol. Figure 8 illustrates the steps taken to prepare the mobile-
source emission inputs for a 20 percent methanol, 80 percent base case UAM
run.
At every grid in the modeling region, the total reactive hydrocarbon
is first split into two parts: one representing emissions from gasoline-
fueled vehicles, the other representing emissions from methanol-fueled
vehicles. In Figure 8 (Scenario 3), 80 percent of the mass of RHC is
treated as emissions from gasoline-fueled vehicles and 20 percent is
treated as emissions from methanol-fueled vehicles. The next step is the
same for both gasoline and methanol type emissions. The emissions are
divided into exhaust and evaporative fractions on the basis of the county
in which the grid cell is located. The proportions determined from the
total emissions by county for the year 1979 are listed in Table 14.
Multiplying by the projection factor in Table 14 for the appropriate
scenario yields the mass of RHC in each of the categories. The final step
involves splitting the mass into species for modeling. The factors used
for gasoline exhaust and evaporative emissions are shown in Table 16. The
evaporative emissions for the methanol case were treated as 100 percent
methanol. Methanol exhaust emissions were split into 89 percent methanol,
10 percent formaldehyde and 1 percent methyl nitrite. The fractions of
methanol and formaldehyde for Scenario 5 were modified to test the effects
of 94/5 percent and 79/20 percent splits.
To project HC and NOX emissions for stationary and off-road mobile
sources, estimates of future HC emissions contained in a recent State
Implementation Plan revision for the region were reviewed (DVRPC, 1983).
85117T2 »t
54
-------
TABLE 16. MOBILE SOURCE INVENTORY
SPLITS FOR 1979 AND 2000
Species
PAR
OLE
ETH
ARO
CARB
Exhaust
Carbon-Bond
Fraction (%
1979
61.4
7.0
7.1
15.7
8.8
as Carbon)
2000
67.0
3.8
2.6
22.1
4.5
Evaporative
Species
PAR
OLE
ETH
ARO
CARB
Carbon-Bond
Fraction (%
1979
92.8
1.7
0
0.6
4.8
as Carbon)
2000
68.7
3.4
0.49
22.1
5.2
85117T2 5
55
-------
Total mobile source VOC
1979
80% gasoline fraction
Exhaust
fraction
Evaporated
fraction
20% methanol fraction
Exhaust
fraction
Evaporated
fraction
tn
cr>
VOC with
exhaust
reactivity
Multiply by projection
factor for mobile-source
emissions from 100%
gasoline combustion
Total exhaust VOC
projected for
2000
ETH
OLE
split
"** ARO
VOC with
evaporation
reactivity
Multiply by projection
factor for mobile-source
emissions from 100%
gasoline combustion
Total
evaporation VOC
projected for
2000
ETH'
'CARB
OLE
split
ARO
CARB
PAR
PAR
Methanol and
formaldehyde
from methanol
combustion
Methanol only
Multiply by projection
factor for mobile-source
emissions from combustion
of 100% methanol
Total methanol
related VOC
projected for
2000
Multiply by projection
factor for mobile-source
emissions from combustion
of 100% methanol
Evaporated
methanol
projected for
2000
Methanol
Methanol
Formaldehyde
FIGURE 8. Schematic of preparation of mobile emission and evaporation inputs for 20 percent
methanol/80 percent gasoline simulation.
85117
-------
Stationary sources include major and minor point sources, and area
sources. Area sources are defined as small local sources such as dry
cleaners, automotive repair shops, and residential emissions. Mobil
sources include both on-road motor vehicles and mobile-related stationary
sources. Area sources are the major contributors to reactive hydrocarbon
emissions for both the 1979 and 2000 Philadelphia emissions inventory.
Attainment of the ozone standard was projected for 1987, with a 36 percent
reduction in HC emissions from all sources other than motor vehicles.
Accordingly, it was assumed that a 36 percent reduction from 1979 HC emis-
sions would be maintained through the year 2000 by point and area
sources. Furthermore, the implementation plan revision estimated a 4 per-
cent increase in NOX emissions by 1987 from all sources other than motor
vehicles. This figure is consistent with a 4.1 percent increase in 1987
population estimated for the Philadelphia region (ESI, 1982). Since popu-
lation in the year 2000 is expected to have increased 10.7 percent over
the 1979 population (ESI, 1982), the point and area source NOX emissions
in the year 2000 were increased by 10.7 percent.
Evaporation from fueling operations (i.e., gas stations) for the year
2000 was estimated by calculating the ratio of hydrocarbon evaporation to
total hydrocarbon emissions for the 1979 inventory and then multiplying
that same ratio by the total hydrocarbon emissions from area sources esti-
mated for the year 2000; it was determined that 8.5 percent of the total
HC from area sources would be due to fueling operations. For the complete
methanol replacement scenarios, this percentage (8.5 percent) was assumed
to be methanol. This value accounts for roughly 4 percent of the total HC
inventory. For the partial replacement scenarios, 1.7 percent (0.20 x 8.5
percent) of evaporative emissions due to fueling operations were assumed
to be methanol.
Total daily emissions (tons per day) by source type for the year 2000
base case (1A) are compared to those of the 1979 inventory in Table 17.
For mobile sources only, the percentage of the HC inventory is reduced
from 31 percent in 1979 to 16 percent in the year 2000. The hydrocarbon
splits for 1979 and 2000 are also shown for comparison. Percentages of
total inventory due to mobile sources with complete methanol replacement
are 4.8, 12.3, and 24.4, respectively, for Scenarios 2A, 2B, and 2C.
Initial Conditions
The maximum initial ozone concentration for the year 2000 was set at
0.08 ppm and is an estimate of that value in the year 2000. In most cases
the 1979 initial ozone concentrations were equal to 0.08 ppm.
85117^2 4 57
-------
TABLE 17. 1979 AND YEAR 2000 PHILADELPHIA EMISSION INVENTORIES
Emissions (tons
Source File
Major Point
Minor Point
Area
Mobile-Related
Stationary
On-Road Motor
Vehicles**
1979
per day, summer weekday)
2000
RHC*
41.
386.
526.
45.
442.
5
1
4
8
9
(3%)$
(26)
(37)
(3)
(31)
304.
109.
147.
0
343.
NO
0
3
7
9
xt RHC*
(34)
(12)
(16)
(0)
(38)
26
235
336
29
117
.6
.5
.8
.3
.4
(4%)
(31)
(45)
(4)
(16)
NO
336.6
121.0
163.5
0
243.0
xt
(39)
(14)
(19)
(0)
(28)
TOTAL 1424.7 (100) 904.9 (100) 745.6 (100) 864.1 (100)
* Emissions expressed as methane.
t Emissions expressed as nitrogen dioxide.
§ Percentages of the total inventory
** Emissions based on MOBILES.
Philadelphia Inventory
Splits
Carbon-Bond
Fraction (% as Carbon)
Species 1979 2000
PAR
OLE
ETH
ARO
CARS
74.0
2.8
4.1
13.2
5.9
79.0
1.1
9.7
8.5
1.7
85M7P2 5
58
-------
Initial and boundary hydrocarbon conditions for the year 2000 were
calculated by subtracting an estimated middle hydrocarbon background value
(Table 18) from the total 1979 initial conditions. The resulting differ-
ence represents the portion of hydrocarbons that can be controlled by
changing the volume or composition of emissions in Philadelphia. In com-
parison, the hydrocarbon background is considered uncontrollable because
it remains unaffected by changes in the emissions. An example of an
uncontrollable background source is the influx of hydrocarbons emissions
from natural vegetation. The controlled portion of hydrocarbon concentra-
tion is multiplied by 0.52, which is a projection factor of the reduction
in Philadelphia's contribution to initial hydrocarbon concentrations in
the year 2000. This calculation estimates the change in initial and boun-
dary hydrocarbon concentrations in the year 2000 due only to controllable
sources. It is assumed that areas outside of Philadelphia might also con-
trol VOC emissions to the same degree as those within the Philadelphia
airshed (i.e., by 48 percent). The initial and boundary conditions in the
year 2000 are given in Tables 19, 20, 21.
Inputs Used in the Systems Applications'
Trajectory and Box Model Simulations
The meteorological and emissions files used in the Systems Applica-
tions' trajectory model are the same as those used by the DAM. By running
the Systems Applications trajectory model in a backward mode, a path to
the maximum ozone observed can be determined for the day of interest. The
trajectory paths for 13 and 19 July 1979 are shown in Figures 9a and b,
respectively. For a more detailed explanation of the use and application
of the Systems Applications trajectory model (see Myers et al., 1979).
Using the results of the Systems Applications trajectory model, emis-
sions and meteorological conditions are constructed for a Level II type
box model with Carbon-Bond chemistry (Ozone Isopleth Plotting Package)
(OZIPM/CBM). OZIPM is a computer routine used in box modeling that can
handle different chemical mechanisms. The trajectory model prints out the
emissions and mixing heights along the trajectory path at hourly inter-
vals. These mixing heights can be used as direct inputs to the OZIPM com-
puter code. The emission rates from the trajectory model are in units of
moles/hr for each of the Carbon-Bond species and are converted to ppmC/hr
for volatile organic compounds (VOC) by the following equation:
Emissions Rate (ppmC/hr) = [24450/(ZQ* 40002)][Emissions Rate (moles/hr)]
where Zg is the initial mixing height.
85117^2 1+
-------
TABLE 18. BACKGROUND OF REACTIVE HYDROCARBONS. (Source:
Killus and Whitten, 1984).
Carbon
Fraction
Low
Mid
High
0.91
0.09
0.61
0.083
0.034
0.014
0.26
0.0005
»
0.57
0.07
0.12
0.07
0.17
0.0005
Concentration
0.03 ppmC
0.003 ppm
0.1 ppm
0.035 ppmC
0.0008 ppm
0.001 ppm
0.0004 ppm
0.015 ppm
0.00003 ppm
0.2 ppm
0.1 ppmC
0.002 ppm
0.01 ppm
0.006 ppm
0.03 ppm
0.00008 ppm
0.5 ppm
Species
PAR
CARB
CO
PAR
ARO
ETH
OLE
CARB
DCRB
CO
PAR
ARO
ETH
OLE
CARB
DCRB
CO
0.03 ppmC
OH to R02 reactivity
= 125 min'1
0.043 ppmC
(+ 0.015 ppm CARB)
OH to R02 reactivity
= 387 min'1
0.144 ppmC
(+ 0.03 ppm CARB)
OH to R02 reactivity
1153 min"1
Note: All estimates include 0.015 ppmC PAR as surrogate for
background methane.
85117T2 5
60
-------
TABLE 19. LOWER BACKGROUND CONCENTRATION VALUES FOR
THE YEAR 2000 FOR ALL LEVELS OF ALL BOUNDARIES
EXCEPT THE LEVELS BELOW THE MIXING HEIGHT
ON THE NORTHEAST AND EAST BOUNDARIES
Species Concentration (ppm)
NO 0.001
N02 0.002
CO 0.2
ETH 0.001
OLE 0.0004
PAR 0.050
CARB 0.015
ARO 0.0008
PAN 0.000025
BZA 0.00001
85117T2 5
61
-------
TABLE 20. LOWER INITIAL CONDITIONS FOR THE YEAR 2000 (ppm)
AMS Lab.
Ancora
Brigantine
Bristol
Camden
Chester
Cl aymont
Conshohocken
Defense Support
Downington
Franklin Inst.
Island Rd. Airport Cir.
Lumberton
Norristown Armory
Northeast Airport
Roxy Water Pump
SE Sewage Plant
South Broad
Summit Bridge
SW Corner Broad/Butler
Trenton
Van Hiseville
Vineland
PAR
0.1602
--
--
--
0.4071
--
--
0.037
0.1811
0.1359
0.1195
0.0579
--
--
--
OLE
0.0012
--
--
0.0029
--
0.0004
0.0013
0.0010
--
--
0.0009
0.0005
ETH
0.0093
--
--
--
0.0244
--
--
0.0017
--
0.0106
0.0078
0.0068
0.0030
--
--
ARO
0.0030
--
--
__
0.0074
--
--
--
0.0008
--
0.0033
0.0025
--
--
0.0022
0.0011
--
--
--
CARB
0.0150
--
--
--
--
0.0150
--
--
0.0150
--
0.0150
0.0150
--
--
0.0150
0.0150
--
--
--
85117P2 5
62
-------
TABLE 21. BOUNDARY CONDITIONS FOR THE YEAR 2000 USED FOR THE NORTHEAST AND EAST
BOUNDARIES BELOW THE MIXING HEIGHT ESTIMATED FROM DATA COLLECTED AT THE VAN HISEVILLE,
NEW JERSEY MONITOR (ppm)
Hour (EST)
0000-0100
0100-0200
0200-0300
0300-0400
0400-0500
0500-0600
0600-0700
0700-0800
0800-0900
0900-1000
1000-1100
1100-1200
1200-1300
1300-1400
1400-1500
1500-1600
1600-1700
1700-1800
1800-1900
1900-2000
2000-2100
2100-2200
2200-2300
2300-2400
NO
0.044
0.044
0.044
0.044
0.044
0.044
0.047
0.043
0.026
0.009
0.001
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.002
0.005
0.008
N02
0.036
0.036
0.036
0.036
0.036
0.036
0.041
0.0494
0.0478
0.0273
0.0099
0.0058
0.0054
0.0072
0.0047
0.0035
0.0043
0.0053
0.0072
0.0071
0.0041
0.0030
0.0060
0.0030
PAR
0.1627
0.1627
0.1627
0.1627
0.1627
0.1627
0.1773
0.1852
0.1515
0.0831
0.0367
0.0350
0.0350
0.0350
0.0350
0.0350
0.0350
0.0350
0.0350
0.0350
0.0350
0.0350
0.0350
0.0350
OLE
0.0004
0.0004
0.0004
0.0004
0.0004
0.0004
0.0004
0.0004
0.0004
0.0004
0.0004
0.0004
0.0004
0.0004
0.0004
0.0004
0.0004
0.0004
0.0004
0.0004
0.0004
0.0004
0.0004
0.0004
ETH
0.001
0.001
0.001
0.001
0.001
0.001
0.001
0.001
0.001
0.001
0.001
0.001
0.001
0.001
0.001
0.001
0.001
0.001
0.001
0.001
0.001
0.001
0.001
0.001
ARO
0.0008
0.0008
0.0008
0.0008
0.0008
0.0008
0.0008
0.0008
0.0008
0.0008
0.0008
0.0008
0.0008
0.0008
0.0008
0.0008
0.0008
0.0008
0.0008
0.0008
0.0008
0.0008
0.0008
0.0008
CARB
0.015
0.015
0.015
0.015
0.015
0.015
0.015
0.015
0.015
0.015
0.015
0.015
0.015
0.015
0.015
0.015
0.015
0.015
0.015
0.015
0.015
0.015
0.015
0.015
PAN
0.0002
0.0001
0.0000
0.0000
0.0000
0.0000
0.0000
0.0004
0.0008
0.0013
0.0019
0.0018
0.0014
0.0012
0.0007
0.0005
0.0003
0.0003
0.0002
0.0001
0.0001
0.0000
0.0000
0.0000
85117T2 5
63
-------
ItOfTH
SOUTH
FIGURE 9a. Averaged trajectory path for 13 July 1979,
NORTH
10
20
30
30 r
10
1600
LVHB
AffCO
SUHM
BKIS
₯tN£
30
25
10
10
20
30
SOUTH
FIGURE 9b. Trajectory path for 19 July 1979.
85117r
64
-------
Initial NOX and VOC precursors for OZIPM are determined from the
initial conditions using the average of the layers below the mixing height
in the trajectory model. NOX and VOC precursors aloft are determined from
the amount of precursors that entered the mixed layer from aloft in the
trajectory model simulations. Tables 22 and 23 describe the initial con-
ditions and emission rates used for the 13 and 19 July base case (1A).
OZIPM was run for 13 July using both the lower initial conditions and
those from 1979. Emission rates were changed according to the list of
scenarios. July 19 was run with 1979 initial conditions for the base
case, which was also the only scenario simulated for that day with
OZIPM. For a more detailed explanation of the use and application of
OZIPM see EPA (1978).
85117T2 (* 65
-------
TABLE 22. INITIAL CONDITIONS AND EMISSION RATES USED FOR
OZIPM CALCULATIONS FOR THE BASE CASE (lA)--JULY 13.
(a) Initial Conditions*
Species
03
voc
NOX
Surface
0.0363 ppm
0.0891 ppmC
0.0332 ppm
Aloft
0.0782 ppm
0.0539 ppmC
0.0004 ppm
Hydrocarbon Reactivity: Carbon Fraction
OLE =
PAR =
ARO =
ETH =
METH =
HCHO =
DCRB =
CARB =
Surface
0.0263
0.6931
0.1259
0.0400
0.0000
0.0381
0.0003
0.0763
Al oft
0.0062
0.5650
0.0711
0.0285
0.0000
0.1002
0.0023
0.2266
85117P2 5
66
-------
TABLE 22. CONCLUDED.
(b)
Time (COT)
0000-1000
0100-0200
0200-0300
0300-0400
0400-0500
0500-0600
0600-0700
0700-0800
0800-0900
0900-1000
1000-1100
1100-1200
1200-1300
1300-1400
1400-1500
1500-1600
1600-1700
1700-1800
1800-1900
1900-2000
Emission
Species
VOC
(mole/hr)
1994
1577
1217
2864
7505
11671
42260
77379
72265
51171
43113
34344
26227
12866
12199
3876
1445
1380
167
Rates and Mixing Heights
NO
(mole/hr)
2056
4316
3005
2763
4972
6387
10699
15940
22218
14038
10257
6974
5399
3187
2494
856
370
271
41
Mixing Heights
at Beginning of
Each Hour (m)
250.0
250.0
250.0
250.0
250.0
250.0
250.0
295.0
450.0
925.0
1200.0
1480.0
1480.0
1530.0
1530.0
1420.0
1310.0
1130.0
950.0
525.0
* Lower initial conditions for 2000.
85 1 1 7' 4 5
67
-------
TABLE 23. INITIAL AND BOUNDARY CONDITIONS FOR OZIPM
CALCULATIONS FOR THE BASE CASE (1A)JULY 19.
(a) Initial Conditions*
Species
°3
voc
NOX
Surface
0.0034
0.5414
0.0871
Aloft
0.0659
0.0518
0.0015
Hydrocarbon Reactivity: Carbon Fraction
OLE =
PAR =
ARO =
ETH =
METH =
HCHO =
DCRB =
CARB =
Surface
0.0270
0.7139
0.1308
0.0398
0.0000
0.0292
0.0009
0.0584
Aloft
0.0085
0.6331
0.0695
0.0314
0.0000
0.0839
0.0039
0.1696
85117T2 5
68
-------
TABLE 23. CONCLUDED.
(b) Emission Rates and Mixing Heights
Species Mixing Heights
VOC NOX at Beginning of
Time (CDT) (mole/day) (mole/day) Each Hour (m)
0000-1000 4313 892 80.0
0100-0200 12323 1887 431.6
0200-0300 22721 3441 600.0
0300-0400 17367 1445 856.7
0400-0500 8115 1762 1240.0
0500-0600 9304 2217 1530.0
0600-0700 9843 2614 1480.0
0700-0800 6266 1763 1332.6
0800-0900 3640 1108 1117.8
85117T2 5
69
-------
SECTION 4
DISCUSSION OF MODEL RESULTS
Study results indicate that atmospheric sensitivity to VOC and NOX
precursors in Philadelphia in the year 2000 appears to be strongly
governed by the ratio of VOC/NOX, and only to a small degree by total pre-
cursor concentrations. This finding is most clearly illustrated by the
isopleth diagram produced specifically by OZIPM shown in Figure 10. An
isopleth diagram is a series of lines connecting the points at which a
given variable has a specified constant value. This isopleth diagram
shows the maximum one-hour ozone concentrations resulting from NO and VOC
emissions relative to the base-case emissions (the 1,1 point on the dia-
gram). VOC and NOX are measured here as a percentage of the base-case
emissions. For example, at 0.75 along the relative VOC scale and 0.60
along the relative NO scale, VOC emissions are 75 percent and NO emis-
sions are 60 percent of the base-case emissions. Therefore, the base-case
point on the isopleth diagram has the coordinates (1,1), since VOC and NOX
are equal to the equivalent emission inventory value at that point.
The isopleth diagram shown in Figure 10 also explains the unexpected
ozone insensitivity when mobile source emissions are eliminated. Since
the percent reduction in overall .emissions is slightly greater for NOX
than for VOC when mobile source emissions are eliminated, the overall
VOC/NO ratio increases somewhat. As the total emissions from the (1,1)
base case point of Figure 10 decrease and the VOC/NOX ratio is increased
to the zero mobile equivalent point, the predicted ozone closely follows
the isolines, indicating little change in ozone. Removing mobile sources
and mobile-source-related stationary emissions reduces the area-wide VOC
in the year 2000 by about 20 percent, as shown in Table 17. Hence, the
point marked "No Mobile Source Point" in Figure 10 shows about the same
ozone maximum as the (1,1) base case point.
The lack of ozone reduction from zero mobile-related emissions is
related to the slanting parallel lines of Figure 10. Essentially, the
slanted lines result from the initial "loss" of emitted NO to form N0~ via
the NO-to-NOp conversions related to VOC photooxidation. Once the initial
NO has reacted to N02, further NO-to-N02 conversions start the production
85il7r2 6
70
-------
CD
tn
1.50
1.35 -
NO Mobile Sources
00 0.15 0.30 0.45 0.60 0.75 0.90 1.05 1.20 1.35 1.50
RELflTIVE VOC
FIGURE 10. Overall emission sensitivity for the 13 July 2000 base case with low initial conditions,
-------
of ozone. In the initial conversion of NO to N02, the amount of VOC that
is photooxidized is directly related to the amount of NO; however, once
the NO has reacted, the amount of ozone formed is essentially independent
of the amount of NOX present, but closely related to the amount of VOC,
which continues to photooxidize and convert NO to N02. Hence, there is a
threshold at which the chemistry changes from NO to N02 reaction and to
ozone formation; this threshold appears on the isopleths as a slanted
straight line with a slope related to the amount of VOC expended to con-
vert the emitted NO. The ozone formed thereafter essentially depends only
on the VOC, which continues to photooxidize. Hence, a series of isolines,
all parallel to the threshold line, are formed.
The validity of this explanation was also presented by Whitten
(1983), who demonstrated that converting NOX emissions from NO to N02
virtually removes the "slant," causing the threshold to be nearly verti-
cal. Hence, an important sidelight of the present study is the reemphasis
of the importance of the high ratio of NO in NOX emissions. If NOX emis-
sions were composed mostly of N02 instead of NO, the atmospheric models
would predict much higher urban ozone levels, independent of NOX emis-
sions, until very low NOX levels were reached. Virtually all ozone iso-
pleth diagrams have isolines nearly parallel to the VOC axis for low
levels of NOX. However, the 0.12 ppm ozone isoline is typically far below
both present and near-future NO emissions levels.
METHANOL FUEL SUBSTITUTION
The effect of emission control devices on mobile sources has been
predicted to account for a factor of 3 reduction in mobile source VOC
emissions for Philadelphia between 1979 and 2000. The overall contribu-
tion to VOC emissions from mobile sources drops from 34 percent in 1979 to
only 20 percent in 2000, in addition to which total VOC emissions are
expected to be nearly half of the 1979 value by the year 2000 (Table 17).
The substitution of methanol and some formaldehyde in exchange for
gasoline-fueled auto exhaust has been shown by Pefley, Pullman, and Whit-
ten (1984) to reduce the reactivity of the atmosphere towards ozone forma-
tion. However, the ozone isopleth diagram 13 July 2000 (Figure 11), shows
virtually no change in ozone levels for Philadelphia as a result of
methanol substitution. Tables 24 and 25 also show that in many of the
scenarios studied there was no significant change in ozone levels from the
base case simulations. For the 19 July 2000 cases shown in Table 25, the
base case was already too low in ozone to justify any further considera-
tion. Hence, the present study focused on the 13 July 2000 cases, which
were clearly above the air quality standard of 0.12 ppm ozone.
85117T2 6
72
-------
CD
cn
X
o
o>
u
3
O
CO
0>
r
-O
O
00 O)
^
M
>O
"tt)
a:
<:. MB
1.80
1.60
1.40
1.20
1.00
0.80
0.60
0.40
0.20
1 1 1 1 1 1 1 U- B
'i!> " ~ 0 15-
0 JR n ir
lu 0. lo-
0.17 . 0 1?
' 0. J7
0.18
-B.18 0 jg
U'1D 0.10 0.19.
" 0-20 -0 -0 n -
""-D 0.20 _0 20
-0.21 0 o,
0.21- 0-i,j
1 1 1 1 1 1 1 1 1
00 0.10 0.20 0.30 0.40 0.50 0.60 0,70 0.80 0,9? I.
Relative Methanol Substitution
of Mobile Sources
FIGURE 11. Methanol and NOX sensitivity for mobile source emissions relative to 13 July 2000
(low initial) base case.
-------
TABLE 24a. METHANOL IMPACT MODELING FOR 13 JULY 2000
Scenario
1A
IB
2A
28
2C
2D
2E
3A
3B
3C
4A
4B
5A
5B
Description
Base case
Full methanol
(x 2.5)
(x 5.0)
2B with surface deposition
2B with 20 percent 1C,*
aloft
Partial methanol (20 percent)
Based on 26;
formaldehyde, 5 percent
20 percent
Based on 2C;
formaldehyde, 5 percent
20 percent
UAM UAM* TRAJ TRAJ* BOX
192 185 211 194 223
202 219
192 186 210 192 219
189
203
192 185 211 194 223
185 188 214
228
BOX*
204
192
183
198
216
177
198
201
205
190
206
206
233
* Lower initial and boundary conditions; all results in ppb instantaneous
ozone at 1600 hours at regional maximum site. Lower initial and boundary
conditions were calculated as previously discussed.
* 1C = initial conditions.
85 1 17ri*
74
-------
TABLE 24b. BOX MODEL SENSITIVITY TESTS FOR 13 JULY 2000
Scenario
7A
7B
8A
8B
8C
8D
9A
9B
9C
9D
9E
10A
Description
Zero mobile sources,
Base NOX, zero VOC from mobile
Stationary methanol at zero
(based on 2B)
Overall substitution 50 percent methanol
Same with 50 percent methanol in 1C*
Same with 50 percent methanol aloft and 1C
Mobile carbon replacement by methanol
(no formaldehyde)
carbon (x 1.5) methanol
(x 2.0) methanol
carbon, 0.1 formaldehyde (no methanol)
0.2 formaldehyde (no methanol)
50 percent VOC overall, no methanol;
base NOX, 1C, aloft
Ozone
(ppb)
201*
156
192
156
134
113
179
183
187
184
196
124
* Lower initial and boundary conditions; all results in ppb ozone at
1600 hours at regional maximum site.
* 1C = initial conditions.
15 1 1 7'
75
-------
TABLE 25. METHANOL IMPACT MODELING FOR 19 JULY 2000*
Scenario UAM TRAJ
Base case 1A 137 130
Full methanol for mobile 2B
source
No mobile source 7A 114
BOX
124
126
107
* All results in ppb ozone at 1600 hours at regional
maximum site; high carryover conditions.
85117T 8
76
-------
The general lack of effect from methanol substitution can be partly
explained by the relatively low (16 percent) contribution from mobile
sources and the high sensitivity of the models to the small amount of
formaldehyde emissions that are found in methanol exhaust. The lack
of sensitivity to mobile source substitution by methanol fuel may be
explained in part by the extreme slant shown in the isopleth curves of
Figure 10. Although most of the slant probably results from NO titration,
as discussed, the remainder is due to ozone and ozone precursor/carryover
either from Philadelphia or from neighboring cities. Figure 10 shows that
zero VOC emissions in conjunction with a small amount of NOX emissions can
produce simulated ozone levels in excess of 0.12 ppm. The necessary VOC
precursors in such cases come from assumed initial and aloft concentra-
tions.
The sensitivity results for cases 8C and 8D shown in Table 24b indi-
cate that initial VOC is only slightly more important than aloft VOC even
though the assumed concentrations aloft are much lower than the initial
concentrations. This nearly equal sensitivity can be explained by two
factors. First, the aloft precursors enter the simulation only as the
mixing height rises and have not dissipated when they enter the simula-
tion, whereas the initial precursors are partially dissipated through
chemical reactions and wind. Second, the ratio between the initial and
highest mixing height values accounts for a reduction due to dilution by a
factor of 6.1 for the initial precursors, which nearly equals the contri-
butions of the two carryover sources.
If the total concentration of the carryover sources of VOC in initial
and aloft conditions is lowered by 20 percent, and if methanol replaces
that missing 20 percent, reactivity is greatly reduced. The sensitivity
results in Table 24 for Scenario 2E show greater decreases in ozone than
when methanol is substituted in the primary mobile-related emissions for
Philadelphia (Scenario 2A). Assessment of the contribution of Philadel-
phia emissions to carryover and the contribution from sources upwind of
Philadelphia is beyond the scope of the present study; such an assessment
would require multiple-day regional modeling. However, the box model is
an effective tool for assessing the atmospheric sensitivity to such
changes. Scenarios 8B, 8C, and 8D were designed to show the gross sensi-
tivity to methanol substitution. The results shown in Table 24b indicate
that a 50 percent substitution of methanol for half of all VOC emissions
reduces peak ozone from the base case (1A) value of 201 ppb to 156 ppb.
Further, a 50 percent substitution for half of all the initial VOC reduces
peak ozone to 134 ppb; when aloft VOC is then added to the 50 percent
methanol substitution, the final peak predicted ozone is only 113 ppb.
85117T2 6
77
-------
These simulations assumed no reduction in NOX emissions; however, a reduc-
tion in formaldehyde levels equivalent to 50 percent resulted from substi-
tution of methanol since all VOC species were reduced by 50 percent and
methanol was added to keep the total VOC constant.
Since the lack of effect on ozone production of methanol substitution
can be partly explained by the low contribution of mobile sources (16
percent from vehicles; 4 percent from mobile-related sources) to total
emissions, two additional scenarios were set up. Scenarios 11 and 12
increased the contribution of mobile sources to 30 percent and 50 percent,
respectively. In both cases, mobile source emissions were kept the same
but stationary source emissions were decreased until the mobile source
contributions equaled 30 and 50 percent, resulting in a decrease in total
emissions of about 33 percent and 60 percent, respectively.
Decreasing the overall emissions by reducing stationary source
emissions to the point at which mobile sources contribute 30 percent of
the total emissions (Scenario 11A) significantly decreases the amount of
ozone produced by the base case 1A from 201 ppb to 136 ppb at 1600 hr.
Substituting methanol for the 30 percent mobile source contribution
(Scenario 11B), further decreases the amount of ozone produced, bringing
it close to the EPA standard of 120 ppb (Table 26).
Simply decreasing the overall emissions by reducing stationary source
emissions to the point at which the mobile sources contribute 50 percent
of the total emissions (Scenario 12A) decreases the amount of ozone
predicted by the base case 1A from 201 ppb to 90 ppb, which is already far
below the EPA standard. Full substitution of methanol for the 50 percent
mobile source contribution (Scenario 12B) reduces ozone even further
(Table 26).
Methanol was incorporated into the initial and aloft conditions for
Scenarios 11 and 12 since it is reasonable to assume that complete
methanol substitution for the mobile source contribution to emissions
would lead to the presence of methanol in VOC carryover. Ozone production
for Scenario 11B (with methanol substitution for initial and aloft
conditions) reaches the EPA standard ozone level. With either initial or
aloft methanol removed, more ozone is produced, bringing the level above
the EPA standard (Table 26). Scenario 2E also showed the sensitivity of
ozone production to the substitution of methanol for initial and aloft
conditions.
For Scenarios 11 and 12, the box model was run an additional three
hours because, though 1600 hours is the regional maximum simulated by the
UAM, the box model shows a positive slope for ozone at that time. Ozone
concentration does not begin to level off until about 1800 hour and
85117T2 6
78
-------
TABLE 26. BOX MODEL METHANOL-IMPACT OZONE RESULTS (ppb) FOR 13 JULY 2000 WITH
30 PERCENT AND 50 PERCENT MOBILE SOURCE EMISSIONS
Scenario*
Base Case:
Instantaneous ozone
at 1600 hrf
Average hourly
maximum ozone
No Initial or
Aloft Methanol
136
151
30% Mobile Emissions
Initial and Initial
Aloft Methanol Methanol Only
-_
50% Mobile
Aloft No Initial and
Methanol Only Aloft Metnanol
90
94
Emissions
Initial and
Aloft Methanol
__
Full Methanol:
Instantaneous ozone
at 1600 hr* 129
Average houly
maximum ozone 142
120
131
123
134
126
138
75
76
* All scenarios with lower Initial and boundary conditions calculated as previously discussed.
* Ozone at 1600 hr, the regional maximum site simulated by the UAM.
-------
results in a higher average maximum ozone value. This difference in the
UAM and box model results can be attributed to wind shear between the
layers below the mixing height in the UAM, starting in the late morning of
the modeling day and ending in the aftrnoon of the same day; wind shear
can not be accounted for in the single-layer box model. Since wind shear
introduces a horizontal dilution factor, limiting precursor concentrations
and diluting the ozone, the box model, which is not subject to this
dilution factor, predicts a higher ozone peak occurring later than that of
the UAM. For further discussion of wind shear see Volume III of Whitten
et al. (1985).
The concentrations of ozone and selected products predicted in
Scenarios 11 and 12 are given in Tables 27 and 28 for two different
VOC/NOX ratios--l:l and l:0.2--which are relative ratios based on the base
case values. For Scenario 11 (30 percent mobile), methanol substitution
is beneficial in decreasing ozone under low VOC/NOX conditions, but at
higher values a nominal increase in ozone is predicted (Tables 26 and
27). A similar trend is predicted for Scenario 12 (50 percent mobile).
As reported previously, a decrease in ozone production should result from
methanol substitution at a VOC/NOX ratio of 1:1. Although Table 28 shows
somewhat lower ozone for methanol substitution at 1600 hours for the high
VOC/NOX 50 percent case (Scenario 12), the overall maximum ozone compari-
son for all hours shows that the methanol substitution simulation predicts
178 ppb, which is slightly more than the 173 ppb ozone predicted without
methanol substitution. It appears that for higher VOC/NOX conditions, the
base-case Scenario (12A) was slightly more reactive earlier in the
simulation, but the methanol-substituted Scenario (12B) produced ozone for
a slightly longer period. The relative differences in these maximum
concentrations are small, however, compared to the reductions predicted in
the 1:1 ratio simulations (EPA VOC/NOX ratios projected for the year 2000
give greater weight to the 1:1 results). The effect of the VOC/NOX ratio
on ozone reduction from methanol substitution shown in these simulations
is consistant with the smog chamber results of Pefley, Pullman, and
Whitten (1984).
The predicted product concentrations for the high and low VOC/NOX
conditions used in Scenario 11 (30 percent mobile emissions) indicate that
full methanol substitution should reduce the production of PAN, nitrophen-
ols, and organic nitrates, all of which are potentially harmful to
humans. In addition, the results given in Table 28 show that these
reductions are far less dependent on VOC/NOX ratio than is ozone produc-
tion. That is, although the beneficial effect of methanol substitution on
ozone production is minimized by raising the VOC/NOX ratio through NOX
control, the predicted decreases in the concentrations of these nitrogen-
containing species occur for all cases; this is because the VOCs replaced
by methanol would have produced products that react with N02 to form PAN,
85117T2 6
80
-------
TABLE 27. PRODUCT CONCENTRATIONS PREDICTED FOR SCENARIO 11*
Species^
Ozone
PAN
Nitrophenols
Formaldehyde
Nitric Acid
Organic Nitrates
No Methanol
VOC/NOX =1:1
136
6.57
0.43
10.9
37.0
1.66
Substitution
VOC/NOX = 1:0.2
169
9.10
0.61
8.64
9.3
2.03
Full Methanol
VOC/NOX = 1:1
120
4.20
0.21
10.8
36.6
1.16
Substitution
VOC/NOX = 1:0.2
176
7.86
0.34
9.56
10.7
1.70
* VOC/NOX ratios are relative to absolute base-case values.
"*" Methanol substitution for initial conditions and aloft.
§ ppb at 1600.
85117T2 5
81
-------
TABLE 28. PRODUCT CONCENTRATIONS PREDICTED FOR SCENARIO 12*
Species*
Ozone
PAN
Nitrophenols
Formaldehyde
Nitric Acid
Organic Nitrates
No Methanol
VOC/NOX =1:1
90
2.56
0.19
7.9
29.1
0.81
Substitution
VOC/NOX = 1:0.2
167
7.71
0.43
6.7
11.0
1.58
Full Methanol
VOC/NOX =1:1
75
1.30
0.06
6.8
27.1
0.49
Substitution*
VOC/NOX = 1:0.2
173
5.79
0.16
7.37
13.2
1.92
* VOC/NOy ratios ar relative to absolute base-case values.
4. *
+ Methanol substitution for initial conditions and aloft.
§ ppb at 1600.
85 1 1 7' 2 5
82
-------
nitrophenols, and organic nitrates. Since methanol forms no such long-
chain organic products upon reaction, the production rates are lowered,
not by lack of NC^, but by the decrease in organic products with which NO?
reacts. This is emphasized in the 1:0.2 simulation; these results
indicate that some of the N02 that cannot react with the missing VOC
products was oxidized through inorganic pathways to form additional nitric
acid and ozone (Table 27).
Also of particular importance to an understanding of the chemistry is
the fact that a formaldehyde molecule is formed for every methanol
molecule reacted. This is not quite the case for other VOCs. However,
Table 27 indicates only a slight increase in hourly formaldehyde values
between the base case and the substitution simulations of Scenario 11.
That is, formaldehyde concentrations in the full methanol substitution
simulation do not exceed those of the base case by more than about 10
percent, on the average, with maximum differences in hourly values of no
more than 20 percent.
For Scenario 12, in which mobile sources represent 50 percent of
emissions, the effects of methanol substitution are more definite. Table
28 shows the predicted product yields for these simulations. Since the
percentage of methanol is larger in these cases, the predicted decreases
in PAN, organic nitrates, and nitrophenols are larger than those from the
30 percent mobile exhaust simulations. Again, however, the relative
reductions in these species show far less relationship to the VOC/NOX
conditions than do the predicted ozone maximum concentrations. Rather, as
noted, these changes relate more to the amount of methanol present.
SENSITIVITY OF THE MODEL SIMULATIONS TO FORMALDEHYDE
The sensitivity of the model simulations to formaldehyde is shown by
comparing Scenarios 4A and 4B with Scenario 2B in Table 24a. In these
three scenarios, 100 percent methanol was used; however, in Scenario 4A
formaldehyde exhaust emissions were reduced from 10 percent (Scenario 2B) to
5 percent. In Scenario 4B, the formaldehyde exhaust was assumed to be as
high as 20 percent.
The maximum ozone values produced from the 5, 10, and 20 percent
formaldehyde exhaust levels were 190, 196, and 206 ppb, respectively, for
the box model simulations using low initial conditions, and 214, 219, and
228 ppb ozone for the cases assuming higher initial conditions. The
amount of overall formaldehyde emissions for these three cases was only
3.6, 4.2, and 5.4 percent, respectively, of the entire VOC emissions. The
base case (1A) formaldehyde emissions were 3.3 percent.
85117P2 6
-------
Comparison of Scenarios 5A and 5B with 2C provides another indication
of formaldehyde sensitivity. These scenarios are similar to 4A, 2B, and
4B, except that the 5A, 2C, 5B series substitutes 2.5 times as much
methanol and formaldehyde for gasoline-related emissions. Thus the 5, 10,
and 20 percent formaldehyde exhaust assumptions for the 5A, 2C, 5B series
lead to overall percentages of formaldehyde emissions of 4.2, 5.4, and 7.8
percent, respectively. This series of formaldehyde levels produces ozone
maxima of 206, 216, and 233 ppb, respectively, in the box model simula-
tions shown in Table 24a.
A more dramatic demonstration of the model's sensitivity to formalde-
hyde is shown in the results of Scenarios 9D and 9E in Table 24b. For
these tests, 10 and 20 percent of all mobile-related emissions were
replaced by formaldehyde as carbon, i.e., no methanol or gasoline-related
emissions were present. Two or more base case points can be compared with
the 9D and 9E Scenarios. With no mobile VOC emissions (case 7B), the
maximum ozone is only 156 ppb. Adding only 10 percent of the mobile VOC
as formaldehyde (case 9D) raises the simulated ozone peak to 190 ppb.
Doubling the formaldehyde addition (case 9E) further increases peak ozone
only to 207 ppb. Hence the model shows a very high sensitivity to the
first addition of formaldehyde, raising the ozone peak to 190 ppb (from
156 ppb), which is nearly the peak predicted for 150 percent replacement
by pure methanol (case 9B). This implies a formaldehyde-to-methanol reac-
tivity ratio of 15 to 1. The overall percentages for formaldehyde in
cases 9D and 9E are 6 and 8.3 percent of the total VOC emissions. For
case 7B (no mobile source VOC emissions), formaldehyde was only 3.7 per-
cent of the emissions. Hence the model shows a very high sensitivity when
formaldehyde is increased from 3.7 to 6 percent.
FORMALDEHYDE CONCENTRATION LEVELS
The UAM is an appropriate model for assessing the spatial and tem-
poral impact of changes in formaldehyde emissions resulting from substitu-
tion of methanol fuel. Table 29 shows the contrast in maximum hourly
formaldehyde levels between the 2000 base case and 100 percent methanol
mobile source substitution. Figure 12* shows the year 2000 base case
formaldehyde levels from the beginning of the simulation at midnight until
noon, when the mixing heights have increased to such an extent that the
formaldehyde levels are no longer important because of dilution. Figure
13 presents contour lines of the change in formaldehyde due to 100 percent
substitution of mobile sources with methanol (case 2B minus base case).
Figures 14 and 15 present the overall maximum values seen in Figures 12
and 13. The area of maximum formaldehyde impact is in a region of heavy
* For the convenience of the reader, Figures 12 through 16 are placed at
the end of the report.
84
-------
TABLE 29. MAXIMUM HOURLY FORMALDEHYDE (ppb)
LEVELS COMPARING YEAR 2000 BASE CASE (1A) TO
100 PERCENT METHANOL SUBSTITUTION OF MOBILE
SOURCES (2B)
Hour
0000-0100
0100-0200
0200-0300
0300-0400
0400-0500
0500-0600
0600-0700
0700-0800
0800-0900
0900-1000
1000-1100
1100-1200
1A*
27.3
25.7
24.0
24.1
24.2
25.2
24.6
22.5
18.2
14.1
13.1
13.2
2B*
27.3
25.8
24.1
24.2
25.2
27.1
27.1
25.9
20.6
15.3
13.7
13.3
* Lower initial and boundary
conditions.
85117T 8
85
-------
traffic emissions. The area of maximum ozone in the 2000 base case simu-
lations is shown in Figure 16. The maximum ozone occurs in grid cell
22,24 between 1600 and 1700 hours for all UAM simulations on 13 July 2000,
85117T2 6
86
-------
SECTION 5
CONCLUSIONS
The principal findings regarding methanol fuel substitution for Philadel-
phia in the year 2000 were derived from the results of three models: the
DAM, the Systems Applications trajectory model, and a box model. Model
results indicate that when methanol fuel is used for mobile sources,
(1) There are minor increases in formaldehyde levels in the vicinity
of major traffic emissions;
(2) There is virtually no change in maximum ozone levels;
(3) Methanol displacement of carryover pollutants or initial
boundary condition pollutants can combine with local methanol
emissions to produce synergistic reductions in ozone maxima;
(4) Formaldehyde emissions can be very important as the reactivity
(via methanol substitution) or the total amount of overall VOC
levels is reduced; and
(5) The Philadelphia area is strongly sensitive to NOX emissions.
Regarding the last finding, this sensitivity is exhibited both in the
reactivity of NOX emissions and in the total amount of NO emissions. The
reactivity of NOX increases as the ratio of N0« to NOX Increases. For
this study, the total NOX level is such that if it is decreased, the
amount of ozone will increase. Thus, minor control of NOX (a small
decrease), or a change in the NOX emissions that increases the N02~to-N0x
ratio, will result in increased ozone levels.
The results of the present study can be compared with results of a
previous study by Whitten and Hogo (1983), which simulated methanol
substitution in the Los Angeles area. Minor increases in formaldehyde
levels were seen in both studies and both showed enhanced ozone reduction
when methanol replaced VOC in carryover and boundary pollutant concentra-
tions. However, the previous study shows substantial ozone reductions
when methanol is substituted as a transportation fuel. The present study
85117r 6
-------
shows no ozone benefit from such a substitution unless carryover pollu-
tants are also assumed to be displaced by methanol. The difference in the
results of this study and those of the previous study in the Los Angeles
area can be explained by the small percentage of mobile source VOC (20
percent) predicted for the year 2000 in Philadelphia. Sensitivity studies
performed with the box model suggest that an overall 50 percent
substitution of methanol for VOC may produce large reductions in ozone if
(1) formaldehyde emissions are minimized, (2) NOX is left at predicted
2000 levels, and (3) methanol is present to displace VOC in carryover
pollutants. However, the use of a regional multiday model would be
required to study the intercity and carryover effects of such widespread
methanol use.
85117r 6
88
-------
0
30
20
10
s
.fe
NORTH
10
20
30
I I I I I I \ I I T I 1 I ) I 1 j I 1 iII \ II I j
...3>
;. -- :i1'-
i i
1' i_J_ j 1
: >'< K > "H ;'»:
' -1J-^hMn-
^ " I -.h ^ySP***&$%>*»S
j -i i.. i i ...r..-J...^i .11..;l
10
20
30
i i l
\
_j
-j
-j
10
SOUTH
FORMALDEHYDE CONC (JULY J3) IN PPB - YEAR 2000 BASE CASE IL OVER 1C)
(a) BEJHEEN THE HOURS OF 0 AND 1
FIGURE 12
-------
NORTH
10
20
1 I i i
T j j T T» i1 1 ! i I {1 i 1 !I
30
T 1| T T 1 T
30-
l*-..l .1 't. i I. i ! ..I
It ) > i! -I I I :i 1
0
SOUTH
FORMALDEHYDE CONC (JULY 13) IN PPB - YEAR 2000 BASE CASE (LOHER 1C)
BE THE EN THE HOURS OF 1 AND 2
FIGURE 12 (continued)
90
-------
NORTH
0
10
20
II i I I t I 1 I i I I ITT I 1\ 1 i i T 1 ] !
0
SOUTH
FORMALDEHYDE CONC (JULY 13) IN PPB - YEAR 2000 BASE CASE (LOKER 1C)
(c) BE THE EN THE HOURS OF 2 AND 3
FIGURE 12 (continued)
91
-------
NORTH
30
SOUTH
0
FORMALDEHYDE CONC (JULY 13) IN PPB - YEAR 2000 BASE CASE (LONER 1C)
BE'THE'EN THE HOURS OF 3 AND 4
FIGURE 12 (continued)
92
-------
0
30
20
°
10
NORTH
10 20 30
j 5 I~~T I ! 1 I ! 1 | 1 I T~T : I ! i I I
t lit
JO
20
30
SOUTH
-I
-j
-j
K
0
FORMALDEHYDE CONC (JULY 13) IN PPB - YEAR 2000 BASE CASE (LOHER I-C1
(e) BETHEEN THE HOURS OF 4 AND 5
FIGURE 12 (continued)
93
-------
NORTH
10
20
30
30
10
j i i i i ? i i i j i i i i i i i i i i ; T i T i i i i ii i r
i! ,
^ '-
i I
10
20
30
SOUTH
-\30
4
\
j
'\: I
10
0
FORMALDEHYDE CONC (JULY 13) IN PPB - YEAR 2000 BASE CASE (LOUER 1C)
(f) BETWEEN THE HOURS OF 5 AND 6
FIGURE 12 (continued)
94
-------
NORTH
0
10
20
30
30
20
10
1 i I I i I I I i j I I 1 I 1 1 I 1~1 \ \ I! i i I i 1 T } I 1 i I i
I 11 ;
i t i £ t. ...i .a '.» ..>.'. t...r
0
10
30
SOUTH
FORMALDEHYDE CONC (JULY 13) IN PPB - YEAR 2000 BASE CASE ILOUER 1C)
(q) BE THE EN THE HOURS OF 6 AND 7
FIGURE 12 (continued)
95
-------
NORTH
0
10
20
30
30
20
fc
1 J A .J J./.l. I .1 .3.
SOUTH
.t .1
30
- 30
H
-I
\-\20
f.-i
10
o
FORMALDEHYDE CONC (JULY 13) IN PPB - YEAR 2000 BASE CASE ILOKER 1C)
(h) BE THE EN THE HOURS OF 7 AND 8
FIGURE 12 (continued)
96
-------
NORTH
0
10
20
i i i T "i rf 1 r -[ i
30
» .! t 1 . j i i I :'l l ..... j... I I > t
30
j r i
J i
-i
N
N
i i
I I i
SOUTH
FORMALDEHYDE CONC (JULY 13) IN PPB - YEAR 2000 BASE CASE (LOUER 1C)
(l) BE THE EN THE HOURS OF 8 AND 9
FIGURE 12 (continued)
97
-------
NORTH
JO
20
30
30
20
10
1 ! ! 1 i 1 ' 1 1 j \ \ \ 1 1 1 ' i I 1 i ; i 1 1 \ 1 ) j 1 i
§.«»:-:, ..-t>l,: :!
... £iiV::V-::i':?!V:
f i
JO
-:30
-
> 1 t * J 1 -t.
1 1 1 il 1 1 1 i 1 i
20
ki
10
20
30
^?
SOUTH
FORMALDEHYDE CONC (JULY 13) IN PPB - YEAR 2000 BASE CASE (LOHER 1C)
(j) BETWEEN THE HOURS OF 9 AND JO
FIGURE 12 (continued)
98
-------
NORTH
0
10
20
30
30-
20
10
Tj I
a.., it ,.t Ji 1 I i
1
-]
J
N
iLj a t..:.!.-.1 . i:.« < i i:.LiJi1oll!i. i.. i ...I t i i i
10
20
30
0
SOUTH
FORMALDEHYDE CQNC (JULY 13) IN PPB - YEAR 2000 BASE CASE (LOUER f-CJ
(k) BE THEEN THE HOURS OF 10 AND 11
FIGURE 12 (continued)
99
-------
0
30
20
°
10
i .:.t
NORTH
10
O
20
30
O
j i :i i .j .1 ...I ..^. i. i ...i .
SOUTH
-30
1
-\
J
1 t i I i
20
10
30
0
FORMALDEHYDE CONC (JULY 13) IN PPB - YEAR 2000 BASE CASE (LOHER 1C)
(1) BE THE EN THE HOURS OF 11 AND 12
FIGURE 12 (concluded)
100
-------
0
30
20
JO
0
NO? TH
10
N
20
30
50
_!
J
1
1 I t 1 1 i i i 1 } 1 1 < 1 * I ! > i 1 t ! < 1 I 1 ! 1 1 i ' 1
0
20
SOUTH
FORMALDEHYDE CHANGE (JULY 133 IN PPB - YR 2000, 28 MINUS BASE
(a) BE THE EN THE HOURS OF 0 AND 1
FIGURE 13
im
-------
20
10
10
NORTH
20
30
o
r
n\ :1 :> i i 1 i i ».- l
0
! 1 \.\-\ ' j i I i i i : j I ) 3 i i . ! 1 i
30
/i
SOUTH
FORMALDEHYDE CHANGE (JULY 13J IN PPB - YR 2000, 2B MINUS BASE
(b) BETWEEN THE HOURS OF 1 AND 2
FIGURE 13 (continued)
102
-------
30
20
10
NORTH
10
20
30
I i ! I
J'J
10
n\ j t i i t i i .t .i i i i ..) i i i J < 1 111 i i * .1 i i i I > i ; . i i i Q
SOUTH
FORMALDEHYDE CHANGE (JULY 13) IN PPB - YR 2000, 2B MINUS BASE
(c) BE THE EN THE HOURS OF 2 AND 3
FIGURE 13 (continued)
103
-------
NORTH
0
10
30
20
JO
20
30
i
H
II
1 ) 11 J ] -.:\ 1 i II 1 II I 1 II 1 I I I I i ...j i i I i 1 ) ; I
n
10
20
30
SOUTH
FORMALDEHYDE CHANGE (JULY 13) IN PPB - YR 2000, 2B MINUS BASE
(d) BE THE EN THE HOURS OF 3 AND 4
FIGURE 13 (continued)
104
-------
0
NORTH
10
20
30
30
20
10
n
J
-30
J
N
120
r
~ i .t . I Jill I J } i > 1 J i i 1 J < f J I :.!....! 1 1 I S 1 I 1 1 i I 1
10
20
30
SOUTH
FORMALDEHYDE CHANGE (JULY 13) IN PPB - YR 2000, 2B MINUS BASE .
(e) BE THE EN THE HOURS OF 4 AND 5
FIGURE 13 (continued)
105
-------
30
20
10
10
NORTH
i : ' I - I l ! I i I : ; l
20
30
-\
H
\ - 20
-to
1
j i i i ii i lit ill j t « i*. i i i i i l ) » l i i t \ l t '. l
0
30
SOUTH
FORMALDEHYDE CHANGE (JULY 13) IN PPB - YR 2000, 2B MINUS BASE
BE THE EN THE HOURS OF 5 AND 6
FIGURE 13 (continued)
106
-------
30-
20
K.
10
NORTH
10
20
30
' "1
' _ ;: 'I
\ \. ' t t i LJ i I i 1 I i t ', t i ' I i t I !' I ; r I S 'l . i
10
10
. 20
SOUTH >,
30
FORMALDEHYDE CHANGE (JULY 13) IN PPB - YR 2000, 2B MINUS BASE
(g) BETWEEN THE HOURS OF 6 AND 7.
FIGURE 13 (continued)
107
-------
0
30
20
°
10
0
NORTH
10
20
30
1
' -20
r
10
"j .1 i , .1 t...i ! t .1 I ..I i 1 I 1 ! .) ) ) > I i i I 1 1 < t t ' \ I : i I I
0
10
20
30
0
SOUTH
FORMALDEHYDE CHANGE (JULY 13) IN PPB - YR 2000, 2B MINUS BASE
(h) BETHEEN THE HOURS OF 7 AND 8
FIGURE 13 (continued)
108
-------
0
30
20
10
10
I 1 i I ! I
NORTH
20
30
J
1
-s 50
Id"
I t " } " 1 'i \ ) t I } l i I 1 1 I ? ; 'I ' I ' ''' I ' l' i t I '' I ' i I
0
30
SOUTH
FORMALDEHYDE CHANGE (JULY 13) IN PPB - YR 2000, 2B MINUS BASE
(i) BE THE EN THE HOURS OF 8 AND 9 -
FIGURE 13 (continued)
109
-------
0
30
20
10
0
NORTH
10
20
: 1 ! .' 1 ' 1 '> \ 1 ! 1 ' 1 ! 1
r
30
J
.] j
! 1 i > . 1 .1 1 i S 1 1 1 .1 '. 1 * 1. > ) 1 ; 1 i 1 1
K
10
20
30
SOUTH
FORMALDEHYDE CHANGE (JULY 13) IN PPB - YR 2000, 2B MINUS BASE
(j) BETWEEN THE HOURS OF 9 AND 10
FIGURE 13 (continued)
110
-------
30-
20
°
10
NORTH
?
-
10 20 30
\ i 1 1 I I f i .1 f : 1 : : t i . . j - . i i : ; '! "J -I
i j
~\
\
i
-50
-20
i t } -I i t 1 i 1 _J L _i 1 Jl L i 1 i I ) I I ) I i I
10
o
10
20
30
0
SOUTH
FORMALDEHYDE CHANGE (JULY 13) IN PPB - YR 2000, 2B MINUS BASE
(k) BETWEEN THE HOURS OF 10 AND 11
FIGURE 13 (continued)
111 -
-------
0
30
2C
10
10
NORTH
20
30
10
.1 i t 1 i i ) j . i ! i ) i a i i. .- s i
10
20
30
SOUTH
FORMALDEHYDE CHANGE (JULY 13) IN PPB - YR 2000. 2B MINUS-BASE
(!) BETWEEN THE HOURS OF 11 AND 12
FIGURE 13 (concluded)
112
-------
30 -
20
10
NORTH
10
20
30
'' i i r ' i t i ; t < ? i i i i t i * - i < f i i ' i i i » '-;' i ! !'").!
30
20
oo
<:
UJ
10
10
20
30
0
SOUTH
FORMALDEHYDE CONC (JULY 13) IN PPB - YEAR 2000 BASE CASE (LOWER 1C)
FOR ALL HOURS
FIGURE 14
113
-------
NORTH
10
20
30
30
20
10
I i i I ' \. i : I vl.'i »'«'..::-'. :.":'I rl...' ::!;. J .'i^r^r^J.'.^^TIi!' ;l . "T^;T..:;| ;-j, :!" ! i
0
/ ^
1 f
sf'-J ' 1 ! 1 ' , i )
/^i^'il'vjwilV^if^'lj^iiitii'S!:^^
L
10
20
30
SOUTH
H30
20
co
-------
0
30
20
10
NOR TH
10
20
30
> i \ t 1 } i r t . 1 1. 1 t i
! ,. . - 1
I i ' t i I 1 t ' » I > < t r I ) I
I s
20
10
0
0
10
20
30
SOUTH
OZONE CONC (JULY 13) IN PPHM - YEAR 2000 BASE CASE (LOHER 1C)
(a) BETWEEN THE HOURS-OF 12 AND 13
FIGURE 16
115
-------
0
30
10
10
NDRTH
20
30
J .i.. i.. .J .. i ;< i 1 I 1 1 i 1 j| i I 1 < j I I . > » 1 i j . ) i I i I i. i I i
50
20
10
n
0
10
20
30
SOUTH
OZONE CONC (JULY 13) IN PPHM - YEAR 2000 BASE CASE (LOUER 1C)
(b) BE THE EN THE HOURS OF 13 AND 14
FIGURE 16 (continued)
116
-------
NORTH
0
10
20
30
~!l ! T~~T T
30
20
10
0
i i i i : -i i t i < I i t t i ' t i r i t i' i < i i i i i -1 r I i t r' .1
0
SOUTH
OZONE CONC (JULY 13) IN PPHM - YEAR 2000 BASE CASE ILOUER 1C)
(c) BE THE EN THE HOURS OF 14. AND 15
FIGURE 16 (continued)
117
-------
0
20
JO
0
NORTH
10
20
30
'0
10
0
0
10
20
SOUTH
30
OZONE CONC (JULY 13) IN PPHM - YEAR 2000 BASE CASE fLOh'ER 1C)
(d) BETWEEN THE HOURS OF 15 AND 16
FIGURE 16 (continued)
118
-------
NORTH
0
10
30
20
10
20
30
I T I i 1.1 l.i
t > 1 1 1 ? l I } I t } I . } t t ) 'l f < 1 ! 1 t' < > i I f J ' 1 » I !
0
10
20
30
SOUTH
OZONE CONC (JULY 13) IN PPHM - YEAR 2000 BASE CASE (LONER 1C)
(e) BETHEEN THE HOURS OF 16 AND 17
FIGURE 16 (continued)
119
-------
NORTH
10
20
30
30
20
10
:? \ ..i. . j. i . ..*' j » t i J i
a ! i
lilt
10
0
0
10
20
30
SOUTH
OZONE CONC (JULY 13) IN PPHM - YEAR 2000 BASE CASE (LOUER 1C)
(f) BETWEEN THE HOURS OF 17 AND 18
FIGURE 16 (continued)
120
-------
NORTH
0
10
20
30
30
20
10
I' 1 -it' i J t I I t I t i l i
0
30
SOUTH
OZONE CONC (JULY 13) IN PPHM - YEAR 2000 BASE CASE (LONER 1C)
BETWEEN THE HOURS OF 18 AND 19
FIGURE 16 (continued)
121
-------
NORTH
t .1 i ! . i. i I i ) (: \ \ \ \ I t ! i i .i. ', i.t I \ ;
0
20
30
SOUTH
0
OZONE CONC (JULY 13) IN PPHM - YEAR 2000 BASE CASE (LOUER 1C)
(h) BE THE EN THE HOURS OF 19 AND 20
FIGURE 16 (concluded)
122
-------
REFERENCES
Ames, J., et al. 1978. "A User's Manual for the Airshed Model." U.S.
Environmental Protection Agency, Research Triangle Park, North
Carolina (SAI NO. EM78-89R2).
DVRPC. 1982. "Year 2000 Transportation Plan for the Delaware Valley
Region." Delaware Valley Regional Planning Commission, Philadelphia,
Pennsylvania.
DVRPC. 1983. "1982 Revision of the State Implementation Plan for Ozone
and Carbon Monoxide for the Pennsylvania Portion of the Philadelphia
Air Quality Control Region." Delaware Valley Regional Planning
Commission, Philadelphia, Pennsylvania.
EPA. 1978. "User's Manual for Kinetics Model and Ozone Isopleth Plotting
Package." U.S. Environmental Protection Agency, Research Triangle
Park, North Carolina (EPA-600/8-78-014a).
EPA. 1982. "Emissions Inventories for Urban Airshed Model Application in
the Philadelphia AQCR." U.S. Environmental Protection Agency,
Research Triangle Park, North Carolina (EPA-450/4-82-005).
EPA. 1984. "Guideline for Using the Carbon-Bond Mechanism in City-
Specific EKMA." U.S. Environmental Protection Agency, Research
Triangle Park, North Carolina (EPA-450/4-84-005).
ESI. 1982. "Emission Inventories for Urban Airshed Model Application in
the Philadelphia AQCR." Engineering-Science, Inc., Fairfax, Virginia
(EPA-450/4-82-005). "
Haney, J. L. 1985. "Evaluation and Application of the Urban Airshed
Model in the Philadelphia Air Quality Control Region." U.S.
Environmental Protection Agency, Research Triangle Park, North
Carolina (EPA-450/4-85-003).
85117T2 10
123
-------
Killus, J. P., and 6. Z. Whitten. 1981. "A New Carbon-Bond Mechanism for
Air Quality Simulation Modeling." U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina (EPA-600/3-82-041).
Killus, J. P., and G. Z. Whitten. 1984. "Technical Discussions Relating
to the Use of the Carbon-Bond Mechanism in OZIPM/EKMA." U.S.
Environmental Protection Agency, Research Triangle Park, North
Carolina (EPA-450/4-84-009).
Myers, T. C., D. C.' Whitney, and S. D. Reynolds. 1979. "User's Guide to
the SAI Trajectory Model, Volume IIAppendix E." U.S. Department of
Transportation, Washington, D.C. (DOT-FH-11-8529).
Pefley, R. K., B. Pullman, and G. Z. Whitten. 1984. "The Impact of
Alcohol Fuels on Urban Air Pollution: Methanol Photochemistry
Study." U.S. Department of Energy, Washington, D.C. (DOE/CE/50036-
1).
Reynolds, S. D., J. H. Seinfeld,' and P. M. Roth. ,1973. Mathematical
modeling of photochemical air pollution--!. Formulation of the
model. Atmos. Environ., 7:1033-1061.
Sigsby, J. E., S. Tejada, W.'Ray, J. M. Lang, and J. W. Duncan. 1984.
"Volatile Organic Compound Emissions from 46 In-Use Passenger
Cars." Submitted for publication.
Whitten, G. Z., H. Hogo.^'ancl R. G. Johnson. 1981. "Application of the
Empirical Kinetic Modeling Approach (EKMA) to Urban Areas." U.S.
Environmental Protection Agency, Research Triangle Park, North
Carolina (EPA-450/4-81-005A).
Whitten, G. Z. 1983. The chemistry of smog formation: A review of
current knowledge. Environ. Int., 9:447-463.
Whitten, G. Z., and H. Hogo. 1983. "Impact of Metnanol on Smog: A
Preliminary Estimate." Systems Applications, Inc., San Rafael,
California (SYSAPP-83/044).
10
124
------- |