1
III.
Modeling
Activities
and
Action
Plan
The
following
section
summarizes
some
of
the
modeling
activities
and
recommendations
for
both
EACs.
Modeling
continues
to
be
performed
by
a
contractor,
ENVIRON
Corporation.

Also,
on
March
30,
2004,
the
Oklahoma
Department
of
Environmental
Quality,
under
separate
cover,
submitted
the
local
Clean
Air
Plan
for
both
the
Central
Oklahoma
and
Tulsa
EACs.
Enclosed
is
an
acknowledgement
of
completion
of
this
important
milestone
from
Administrator
Greene.

Additional
EAC
technical
information
can
be
found
at
www.
deq.
state.
ok.
us.

At
the
January
22,
2004
Oklahoma
EAC
Technical
Advisory
Committee
and
Public
Meeting,
results
were
presented
for
a
2007
Base
Case
simulation
based
on
projected
2007
Base
Case
emissions
that
used
area
and
point
source
emissions
from
EPA's
Heavy
Duty
Diesel
Rulemaking.
2007
on­
road
and
non­
road
mobile
sources
were
projected
using
EPA's
MOBILE6
and
NONROAD
model
and
projected
increases
in
activity.
Projected
8­
hour
ozone
Design
Values
were
estimated
following
EPA's
guidance
and
using
five
years
of
observed
8­
hour
ozone
Design
Values
from
1999
to
2003.
Attainment
was
demonstrated
at
all
monitoring
sites
but
Skiatook
monitor
in
Tulsa.
Table
1
below
displays
the
2007
Base
Case
projected
8­
hour
ozone
Design
Values
for
the
three
sites
in
Oklahoma
that
recorded
violations
of
the
8­
hour
ozone
standard
during
1999­
2003,
two
sites
in
Tulsa
(
Skiatook
and
Tulsa)
and
one
in
Oklahoma
City
(
OSDH).
When
using
the
observed
1999,
2000,
2002
or
2003
DVs,
attainment
is
projected
at
all
sites
in
Oklahoma.
However,
when
the
2000
or
2001
DVs
are
used,
the
projected
Design
Value
at
Skiatook
is
88.0
ppb
and
85.1
ppb
both
of
which
are
85.0
ppb
or
greater
so
do
not
project
attainment.

Table
1.
Projected
2007
8­
hour
ozone
Design
Values
(
DVs)
in
Oklahoma
using
the
EPA
Heavy
Duty
Diesel
Rulemaking
2007
Base
Case
emissions
and
five
years
of
observed
DVs
from
1999
to
2003
(
attainment
demonstrated
when
project
DV
is
84.9
ppb
or
lower)
Tulsa
Skiatook
OSDH
Year
Obs
DV
2007
DV
Obs
DV
2007
DV
Obs
DV
2007
DV
1997­
1999
86
81.9
88
83.2
86
82.4
1998­
2000
89
84.7
93
88.0
84
80.5
1999­
2001
82
78.1
90
85.1
80
76.7
2000­
2002
81
78.1
87
84.3
79
77.0
2001­
2003
80
77.2
83
80.4
79
77.0
At
the
January
22,
2004
meeting
concerns
were
expressed
over
using
EPA's
2007
Base
Case
Heavy
Duty
Diesel
Rulemaking
(
HDDR)
area
and
point
source
emissions
because
they
were
based
on
the
1996
NET
inventory
that
was
believe
to
be
much
poorer
quality
for
emissions
in
Oklahoma
than
the
1999
NEI
inventory.
An
analysis
of
the
EPA's
2007
HDDR
inventory
revealed
that
the
locations
and
stack
parameters
for
point
sources
in
2
Oklahoma
that
were
not
Electrical
Generating
Units
(
EGUs)
were
in
error.
In
fact,
the
locations
for
many
of
the
non­
EGU
point
sources
in
Oklahoma
were
the
center
of
the
county.
Thus,
new
projected
2007
Base
Case
area
and
point
source
emissions
were
generated
based
on
the
1999
NEI
inventory.

Revised
2007
Base
Case
Modeling
Results
Revised
2007
Base
Case
area
and
point
source
emissions
were
generated
by
projecting
the
1999
NEI
inventory
to
2007.
A
CAMx
revised
2007
Base
Case
simulation
was
conducted
and
2007
projected
8­
hour
ozone
Design
Values
calculated,
as
shown
in
Table
2.
Using
the
observed
1998­
2000
8­
hour
ozone
Design
Values,
attainment
is
not
demonstrated
at
the
Tulsa
(
85.2
ppb)
and
Skiatook
(
87.5
ppb).
Attainment
is
demonstrated
using
all
other
years
of
observed
Design
Values
and
at
all
other
monitoring
sites.

Table
2.
Projected
2007
8­
hour
ozone
Design
Values
(
DVs)
in
Oklahoma
using
the
Revised
2007
Base
Case
emissions
and
five
years
of
observed
DVs
from
1999
to
2003
(
attainment
demonstrated
when
project
DV
is
84.9
ppb
or
lower)
Tulsa
Skiatook
OSDH
Year
Obs
DV
2007
DV
Obs
DV
2007
DV
Obs
DV
2007
DV
1997­
1999
86
82.4
88
82.8
86
82.1
1998­
2000
89
85.2
93
87.5
84
80.2
1999­
2001
82
78.5
90
84.7
80
76.4
2000­
2002
81
78.6
87
83.9
79
76.7
2001­
2003
80
77.7
83
80.0
79
76.7
2007
Control
Strategies
Table
3
presents
the
2007
projected
8­
hour
ozone
Design
Values
at
the
key
Tulsa
and
Skiatook
monitors
in
Tulsa
and
OSDH
monitor
in
Oklahoma
City
for
the
2007
control
strategies
analyzed
as
part
of
the
Tulsa
and
Oklahoma
City
Clean
Air
Action
Plans
(
CAAP).
The
projected
8­
hour
ozone
Design
Values
at
the
Tulsa,
Skiatook
and
OSDH
monitors
for
the
Revised
2007
Base
Case
(
Run
0)
simulation
were
85.2,
87.5
and
80.2
ppb,
respectively.
When
permitted
sources
whose
permits
are
expiring
so
will
never
be
built
(
Run
5)
are
removed
from
the
2007
inventory
(
which
becomes
a
new
2007
Base
Case)
the
projected
8­
hour
ozone
Design
Values
for
the
three
sites
are
reduced
to
85.0,
87.0
and
80.0
ppb,
respectively.
Thus,
the
Tulsa
and
Skiatook
monitors
are
still
estimated
to
exceed
the
8­
hour
ozone
standard
under
the
2007
Base
Case
conditions.
Most
of
the
remaining
control
measures
reduce
the
projected
8­
hour
ozone
Design
Values
by
0.0
to
0.2
ppb.
Thus
the
Control
Strategy
17
that
combines
the
transportation
control
measures,
RVP
and
point
source
NOx
controls
reduces
the
2007
projected
8­
hour
ozone
Design
Values
at
the
Tulsa
and
Skiatook
monitors
by
0.3
and
0.5
ppb,
respectively,
to
84.8
and
87.1
ppb
thereby
demonstrating
modeled
attainment
at
the
Tulsa
but
not
the
Skiatook
monitor.
3
Table
3.
Projected
2007
8­
hour
ozone
Design
Values
for
various
2007
emission
scenarios
using
the
observed
1998­
2000
8­
hour
ozone
Design
Values.
2007/
2000
8­
Hr
O3
DV
(
ppb)
No.
Scenario
Tulsa
Skiatook
OSDH
Priority
Obs
1998­
2000
Observed
8­
Hr
O3
DVs
89
93
84
0.
Revised
2007
Base
Case
85.2
87.5
80.2
High
Sensitivity
Simulations
1a.
UCI
Oil&
Gas
VOC
Sens
(
High
Est)
Low
1b.
UCI
Oil&
Gas
VOC
Sens
(
Low
Est)
Low
2.
2007
5%
VOC
control
in
Tulsa
MSA
85.1
87.4
80.2
Low
3.
2007
5%
NOx
control
in
Tulsa
MSA
85.1
87.1
80.2
Low
4.
20075%
VOC&
NOxcontrolTulsaMSA
85.0
87.0
80.2
Low
2007
Emissions
Scenarios
5.
Remove
Expiring
Permitted
Sources1
85.0
87.3
80.0
Med
6.
7.8
RVP
in
Tulsa
TMA
84.9
87.2
80.0
Med
7
Stage
I
Controls
in
Tulsa
MSA
85.0
87.3
80.0
Med
8.
7.8
RVP
in
OKC
TMA
85.0
87.3
79.8
Low
9.
Stage
I
in
OKC
MSA
85.0
87.3
79.9
Low
10.
TCMs
in
OKC
TMA2
84.9
87.2
76.3
Med
11.
7.8
RVP
in
TTMA
85%
market
penetration
in
on­
road/
non­
road
84.9
87.2
80.0
High
12.
ITS/
Transportation
Congestion
Mitigation
in
TTMA3
84.9
87.1
80.0
High
13.
Combined
11.
and
12.2,3
84.8
87.1
80.0
High
14.
AEP­
PSO
Oolagah
1
Unit
Low
NOx
85.0
87.1
80.0
Low
15.
OG&
E
Muskogee
1
Unit
Low
NOx
84.9
87.2
80.0
Low
16.
GRDA
Chouteau
1
Unit
Low
NOx
85.0
87.2
80.0
Low
17.
Combine
13.­
16.
84.7
86.8
80.0
Low
18.
Stage
II
in
Tulsa
MSA
84.9
87.2
80.0
High
19.
Basic
I/
M
in
Tulsa
TMA
84.9
87.2
80.0
High
1.
Expiring
permitted
sources
will
stay
removed
in
all
runs
from
5.
on.
2.
Requires
link­
based
TDM
output
from
ACOG.
3.
Requires
link­
based
TDM
output
from
INCOG.

Run
10
Recalculation
DEQ,
ACOG,
and
Environ
agree
that
the
results
of
run
10
are
faulty,
and
will
not
be
used.
To
calculate
the
ozone
reductions
resulting
from
Transportation
Control
Measures
in
the
Oklahoma
City
Metropolitan
area,
ACOG
is
calculating
the
reductions
at
each
link
and
intersection,
and
those
results
will
be
used
to
develop
the
ozone
reduction
values.
4
PROPOSED
PROCEDURES
FOR
UPDATING
FUTURE­
YEAR
BOUNDARY
CONDITIONS
FOR
THE
OKLAHOMA
8­
HOUR
OZONE
EAC
MODELING
The
Oklahoma
8­
hour
ozone
Early
Action
Compact
(
EAC)
photochemical
modeling
of
the
August
13
through
September
1,
1999
episode
used
a
36/
12/
4
km
nested
grid
structure
as
shown
in
Figure
1.
Boundary
Conditions
(
BCs)
(
i.
e.,
the
assumed
concentrations
along
the
lateral
edges
of
the
36
km
grid)
were
generated
using
results
from
an
EPA
simulation
of
their
Models­
3
CMAQ
model
for
an
August
1999
Base
Case
simulation.

Figure
1.
Original
relationship
between
the
MM5
(
red)
and
CAMx
(
blue)
36/
12/
4­
km
modeling
domains.

Future­
year
(
2007)
modeling
was
performed
by
projecting
the
1999
emissions
to
2007.
However
since
EPA
only
performed
1999
Base
Case
modeling
using
CMAQ
for
the
August
1999
episode,
the
2007
BCs
for
the
Oklahoma
EAC
modeling
were
kept
at
1999
Base
Case
levels.
For
the
key
Skiatook
ozone
monitor
in
Tulsa,
CAMx
ozone
source
apportionment
modeling
demonstrated
that
BCs
and
emissions
from
the
Tulsa
MSA
are
by
far
the
two
highest
contributors
to
elevated
8­
hour
ozone
concentrations.
An
example
of
the
ozone
source
apportionment
modeling
for
August
21,
1999
is
shown
in
Figure
2.
The
use
of
1999
Base
Case
BCs
in
2007
will
overstate
the
ozone
contribution
to
BCs
in
the
Tulsa
area
because
several
EPA
regional
emission
control
rules
that
EPA
has
shown
to
effectively
reduce
regional
ozone
concentrations
and
ozone
transport
are
not
being
Abilene
Atlanta
Austin
Baton
Rouge
Beaumont
Birmingham
Chattanooga
Cincinnati
Corpus
Christi
Dallas
Fort
Worth
Galveston
Houston
Jackson
Kansas
City
Knoxville
Lake
Charles
Laredo
Little
Rock
Louisville
Memphis
Mobile
Monroe
Montgomery
Nashville
New
Orleans
Port
Arthur
San
Antonio
Shreveport
Springfield
St.
Louis
Victoria
Waco
Wichita
­
972
­
612
­
252
108
468
828
1188
1548
1908
LCP
Easting
(
km)
­
1728
­
1368
­
1008
­
648
­
288
72
432
LCP
Northing
(
km)

Tulsa
Oklahoma
City
Lawton
MM5
36km
CAMx
36km
MM5
12km
CAMx
12km
CAMx
4km
MM5
4km
5
accounted
for.
Rules
that
have
already
been
promulgated
include
the
NOx
SIP
Call,
Tier
2/
Low
Sulfur,
Heavy
Duty
Diesel
(
HDD)
and
Land
Based
Non­
Road
Engine
Rules.

Figure
2.
Ozone
source
apportionment
modeling
results
at
the
Skiatook
monitor
in
Tulsa
on
August
21,
1999.

The
development
of
more
representative
2007
BCs
for
the
Oklahoma
EAC
modeling
has
risen
in
importance.
Four
possible
approaches
have
been
identified:

1.
Perform
CMAQ
modeling
of
2007
using
EPA's
August
1999
CMAQ
database.
2.
Develop
BC
adjustment
factors
for
August
1999
based
on
current
to
future
year
regional
modeling
results
(
e.
g.,
from
HDD
or
IAQR
rules
or
RPO
modeling).
3.
Extend
the
current
Oklahoma
EAC
36
km
modeling
domain
to
include
the
Midwest
and
Southeast
(
and
other
areas
that
are
believed
contributing)
and
develop
1999
and
2007
Base
Case
emissions
inventories
for
the
new
areas.

Even
if
EPA's
August
1999
CMAQ
databases
could
be
obtained,
Option
1
would
involve
an
entirely
new
emissions
modeling
activity
for
a
new
grid
projection
and
a
modeling
grid
essentially
starting
from
scratch.
This
would
be
quite
a
resource­
intensive
activity.
Contributions
to
2PM
Ozone
on
990821
in
Skiatook
Total
O3
=
86.4
ppb.

0
5
10
15
20
25
30
35
40
45
Tulsa
BC
Central
US
SE
OK
NE
OK
Kansas
NW
OK
N
Texas
Oklahoma
City
SW
OK
S
Texas
SE
US
IC
O3
[
ppb]
BC
IC
Other
low­
level
Elevated
Points
Mobile
Biogenics
6
Option
2
has
been
considered,
and
discussions
with
EPA
have
been
held
in
regards
to
obtaining
some
of
their
current
and
future
year
CMAQ
modeling
output.
But
arbitrary
rules
would
have
to
be
developed
on
how
to
use
modeled
results
from
one
episode
to
adjust
boundary
conditions
for
another
that
we
believed
could
be
open
to
criticism.

Thus,
it
has
been
recommended
that
we
pursue
Option
3
and
extend
the
OK
EAC
36
km
to
include
the
Midwest
and
Southeast
(
and
elsewhere
as
needed)
and
then
develop
new
1999
Base
Case
and
2007
Base
Case
emission
inventories
for
the
new
areas
and
the
August
1999
episode.
The
1999
Base
Case
would
be
based
on
a
standard
application
using
the
NEI99
inventory;
whereas
the
2007
Base
Case
would
be
based
on
the
Heavy
Duty
Diesel
Rule
2007
control
case
emissions
inventory.
This
approach
can
be
pursued
provided
the
MM5
model
was
exercised
for
the
August
1999
episode
over
a
large
enough
area
to
capture
the
new
modeling
areas.
Figure
3
displays
the
MM5
modeling
domain
that
more
than
encompasses
the
new
areas
we
want
to
include.
The
following
steps
would
be
involved
in
this
approach:

 
Perform
ozone
source
apportionment
modeling
of
the
current
36/
12/
4
km
Oklahoma
EAC
database
with
seven
source
categories,
the
four
lateral
edges
of
the
BCs
(
N,
S,
E
and
W)
along
with
the
aloft
BCs
plus
all
anthropogenic
and
biogenic
emissions
to
identify
which
lateral
BCs
contribute
the
most
to
elevated
ozone
in
Tulsa.
 
Extend
the
OK
EAC
36
km
domain
to
include
the
new
areas.
For
example,
if
most
of
the
BC
contribution
comes
from
the
eastern
and
northern
boundaries
then
we
could
push
the
northern
boundary
to
the
top
of
the
US
and
the
eastern
boundary
to
include
as
far
east
as
New
York
City.
 
Develop
emissions
for
a
1999
Base
Case
and
the
new
expanded
36
km
domain
using
the
NEI99
inventory.
 
Run
the
new
1999
Base
Case
and
perform
the
model
performance
evaluation
to
assure
that
model
performance
is
acceptable.
 
Develop
emissions
for
a
2007
Base
Case
and
the
new
expanded
36
km
domain
using
the
2007
Control
Scenario
from
the
EPA
Heavy
Duty
Diesel
Rulemaking.
 
Run
the
2007
Base
Case
and
perform
Design
Value
projections
using
the
1998­
2000
observed
Design
Values
and
the
new
1999
and
2007
Base
Case
simulation.
 
Perform
additional
control
strategy
simulations
as
desired.
7
Figure
3.
MM5
domain
configuration
(
108/
36/
12/
4­
km)
for
the
August
1999
Oklahoma
episode.
8
Figure
4.
Revised
CAMx
domain
to
reduce
North
and
East
boundary
condition
influence.
0
180
360
540
720
900
1080
1260
1440
1620
1800
1980
2160
2340
­
1440
­
1260
­
1080
­
900
­
720
­
540
­
360
­
180
0
180
360
540
720
Austin
Corpus
Christi
Dallas
Fort
Worth
Houston
San
Antonio
Oklahoma
City
Tulsa
Kansas
City
St.
Louis
New
Orleans
Birmingham
Atlanta
Jacksonville
Miami
Tampa
Charlotte
Memphis
Nashville:
Louisville
Norfolk
Baltimore
Philadelphia
Pittsburgh
Buffalo
New
York
Cincinnati
Cleveland
Columbus
Toledo
Indianapolis
Chicago
Detroit
Milwaukee
Minneapolis
Omaha
Wichita
nx
x
ny
SW
to
NE
Corners
CAMx
36km
69
x
67
(­
108,
­
1584)
to
(
2376,
828)
CAMx
12km
83
x
101*
(
­
12,
­
1236)
to
(
984,
­
24)
CAMx
04km
101
x
101*
(
104,
­
688)
to
(
508,
­
284)

*
includes
buffer
cells
36km
12km
4km
CAMx
Modeling
Domain
for
ODEQ
