Appendix
B:
Illustrative
Control
Strategies
and
Projected
Improvement
in
2010
8­
hour
Ozone
Levels
in
16
Nonattainment
Areas
Prepared
By
Office
of
Air
Quality
Planning
and
Standards
and
Office
of
Policy
Analysis
and
Review
Office
of
Air
and
Radiation
of
the
United
States
Environmental
Protection
Agency
Research
Triangle
Park,
North
Carolina
27711
August
31,
2005
1
The
analysis
year
2010
was
selected
because
of
the
availability
of
2010
base
case
emissions
inventories
and
modeling
from
the
CAIR
rulemaking.

2
Illustrative
Control
Strategies
and
Projected
Improvement
in
2010
8
hour
Ozone
Levels
in
16
Nonattainment
Areas
Summary
This
report
summarizes
an
illustrative
analysis
of
two
scenarios
regarding
8­
hour
ozone
concentrations
in
2010
for
areas
in
the
eastern
United
States.
One
scenario
looks
at
predicted
ozone
concentrations
in
20101
without
additional
local
measures
or
the
Clean
Air
Interstate
Rule
(
CAIR).
The
other
scenario
looks
at
the
consequences
of
a
set
of
targeted
local
control
measures
without
CAIR.
While
there
are
significant
limitations,
this
illustrative
analysis
finds
that:

°
Current
state
&
federal
regulations
reducing
ozone
precursors
are
extremely
effective.
In
the
eastern
half
of
the
country,
ninety­
two
of
108
areas
with
design
values
above
the
standard
today
are
predicted
to
be
in
attainment
in
2010.

°
Additional
locally
targeted
controls
provide
significant
further
air
quality
gains.
In
particular,
10
of
the
remaining
16
non­
attainment
areas
in
the
eastern
half
of
the
country
are
projected
to
come
into
attainment
or
be
within
3ppb
of
the
standard
in
2010
with
a
simulated
strategy
consisting
of
stationary
and
area
source
control
measures.
Furthermore,
the
analyzed
strategy
yields
predicted
air
quality
improvements
in
all
16
nonattainment
areas.
Six
major
metropolitan
areas
are
projected
to
remain
above
the
standard
after
applying
these
targeted
local
control
measures.

°
Locally
targeted
control
measures
will
have
positive
effects
beyond
local
ozone
reductions
­­
that
is,
controls
in
one
non­
attainment
area
can
reduce
ozone
concentrations
in
downwind
ozone
non­
attainment
areas.
In
addition,
by
reducing
VOC
and
NOx
emissions,
ozone
control
measures
will
reduce
PM2.5
as
well
as
ozone
concentrations.

°
There
are
many
gaps
in
the
illustrative
analysis:

°
A
major
limitation
is
the
lack
of
a
more
complete
assessment
of
emission
reduction
possibilities
and
emission
reduction
targets.
The
control
measures
menu
in
this
analysis
was
not
comprehensive
(
e.
g.,
no
new
state
fuels
or
mobile
source
measures
were
considered).

°
The
analysis
does
not
reflect
emissions
reductions
that
will
occur
as
a
result
of
CAIR.
The
result
is
an
upward
bias
in
the
estimated
amount
of
additional
emission
reductions
required
for
attainment.
3
°
This
assessment
of
the
air
quality,
health
and
cost
impacts
of
a
set
of
targeted
local
control
measures
does
not
include
nonattainment
areas
in
the
West,
due
to
limitations
of
our
modeling
tools.

°
Other
gaps
include
no
consideration
of
the
effect
of
any
bottlenecks,
scale
effects,
learning
curves,
or
emerging
technologies
on
the
cost
of
the
analyzed
local
control
strategy.

°
The
assessment
of
estimated
benefits
is
not
comprehensive.
First,
the
assessment
does
not
include
health
effects
that
could
not
be
quantified
and
monetized.
Second,
welfare
benefits
(
e.
g.
improved
visual
range,
greater
crop
yield
and
tree
growth)
of
reducing
PM
and
O3
levels
are
not
quantified
or
monetized.
Third,
although
VOC
controls
may
result
in
lower
concentrations
of
particulate
matter
in
some
areas
(
via
reduced
formation
of
secondarily
formed
aerosols),
the
resultant
health
and
welfare
effects
benefits
are
not
quantified
or
monetized.

°
As
with
other
benefits
analyses,
the
benefits
estimates
are
subject
to
uncertainties
such
as
those
associated
with
estimates
of
future­
year
emissions
inventories
and
air
quality,
and
uncertainties
in
the
estimated
relationships
between
changes
in
pollutant
concentrations
and
resulting
changes
in
health
effects
(
e.
g.,
choice
of
study,
uncertain
biological
mechanism
for
health
endpoint).

This
limited
analysis
is
not
a
substitute
for
the
more
detailed
and
comprehensive
analyses
to
be
conducted
by
the
states
and
regional
planning
groups
as
part
of
the
state
implementation
planning
process.
Such
analyses
will
have
more
refined
local
emissions
inventory
projections
and
emissions
reduction
targets,
and
an
expanded
set
of
control
measure
possibilities.
Based
on
those
analyses,
the
states
will
make
judgments
concerning
control
needs,
control
options,
and
the
practicality
of
attainment
deadlines
and
reflect
them
in
their
state
implementation
plan
submittals
to
EPA.

If
states
ultimately
determine
that
certain
areas
cannot
practicably
attain
on
time,
the
Clean
Air
Act
provides
ways
for
an
area
to
receive
more
time
through
bump­
up
to
a
higher
classification
(
e.
g.
from
moderate
to
serious).
Bump­
up
would
mean
that
certain
specific
additional
requirements
apply
as
areas
put
in
place
measures
needed
to
attain
the
standard
as
expeditiously
as
practicable.
Due
to
1­
hour
requirements
previously
adopted
by
these
areas
and
projected
emission
reductions
from
federal
rules,
many
specific
requirements
of
a
higher
classification
might
already
be
satisfied.
Existing
federal
rules
are
projected
to
achieve
further
emissions
reductions
over
time.
In
addition,
the
history
of
the
Clean
Air
Act
has
repeatedly
demonstrated
that
new,
cleaner
technologies
are
developed
to
help
in
meeting
clean
air
goals,
and
the
large
number
of
emerging
technologies
with
potential
to
reduce
emissions
indicates
that
the
historical
trend
will
continue.

Methodology:
Emission
Reduction
and
Air
Quality
Impacts
This
illustrative
analysis
includes
information
on
the
designated
8­
hour
ozone
nonattainment
areas,
air
quality
simulations
of
base
case
conditions
for
the
CAIR
rule,
additional
air
quality
simulations
to
establish
emission
reduction
targets
and
verify
the
results
of
local
control
strategies,
and
application
of
an
air
quality
control
strategy
design
model
to
identify
emission
reduction
targets
for
each
non­
attainment
areas.
2The
original
designations
appeared
in
the
Federal
Register
on
April
30,
2004
(
Volume
69,
Number
84)
on
pages
23857­
23951.
There
were
some
subsequent
revisions
to
the
original
classifications
for
some
Subpart
2
areas.
The
current
classifications
can
be
found
at
http://
www.
epag.
gov/
air/
oaqps/
greenbook/
gntc.
html.
This
is
the
web
address
for
the
EPA
green
book
of
nonattainment
areas.

3Technical
Support
Document
for
the
Final
Clean
Air
Interstate
Rule:
Air
Quality
Modeling
Analysis;
Appendix
E:
8­
hour
Ozone:
Average
Ambient
and
Projected
2010
Base
and
CAIR.
The
document
can
be
found
at:
http://
www.
epa.
gov/
pdfs/
finaltechn07.
pdf
4
Designations
of
Areas
for
the
8­
hour
Ozone
NAAQS.
In
April
2004,
EPA
published
nonattainment
attainment,
and
unclassifiable
designations
for
the
8­
hour
ozone
national
ambient
air
quality
standard.
Nationwide,
there
were
126
areas
with
design
values
above
the
standard.
2
One
hundred
and
eight
of
those
areas
are
in
the
Eastern
United
States.
The
other
18
areas
are
in
California,
Arizona,
Colorado,
and
Nevada.

Table
1.
Summary
of
Designation
Information
for
Eastern
&
Western
Areas
with
8­
hour
Design
Values
above
the
NAAQS
Designation*
Eastern
Areas
Western
Areas
Subpart
1
(
non­
EAC)
62
9
Subpart
1
(
EAC)
12
1
Subpart
2­
Marginal
(
non­
EAC)
13
2
Subpart
2­
Marginal
(
EAC)
1
0
Subpart
2­
Moderate
20
2
Subpart
2­
Serious
0
3
Subpart
2­
Severe­
17
0
__
1_
108
18
*
Subpart
1
Clean
Air
Act
provisions
for
non­
attainment
areas
less
prescriptive.
Subpart
2
provisions
are
more
prescriptive.
EAC
areas
are
early
action
compact
areas.
The
effective
designation
date
for
these
areas
was
deferred.

Base
Case
Conditions.
As
part
of
the
development
of
the
CAIR
rule,
EPA
developed
a
base
case
projection
of
emissions
to
2010
and
a
simulation
using
CAM­
x
of
predicted
ozone
design
values.
3
That
base
case
projection
of
emissions
and
air
quality
simulation
indicated
that
92
of
108
non­
attainment
areas
were
expected
to
have
8­
hour
ozone
design
values
at
or
below
the
standard
in
2010.
By
using
that
base
case,
we
were
to
compare
the
impacts
of
the
local
control
strategy
with
a
base
case
that
was
well
understood.
However,
in
adopting
that
approach
we
precluded
consideration
of
air
quality
improvements
resulting
from
CAIR
in
formulating
a
hypothetical
local
control
strategy.

Selection
of
Emission
Reduction
Targets.
To
develop
illustrative
emission
reductions
targets
for
each
area
expected
to
be
in
non­
attainment
in
2010,
six
air
quality
simulations
were
conducted
using
the
CAM­
x
model.
These
predicted
the
impact
of
ozone
levels
of
reducing
nonattainment
area
emissions
by
10%,
25%,
and
50%
for
NOx
only
and
for
NOx
as
well
as
VOC
emissions,
relative
to
the
2010
base
case.
These
results
together
with
linear
interpolation
enabled
the
development
of
emission
reduction
targets
for
each
of
the
areas
projected
to
be
in
non­
attainment
in
2010
under
base
case
conditions.
4U.
S.
Environmental
Protection
Agency,
Office
of
Air
Quality
Planning
and
Standards.
AirControlNET
version
4.0.
Control
Measure
Documentation
Report.
Prepared
by
E.
H.
Pechan
and
Associates.
April
2005.

5These
areas
are
within
state
but
not
more
than
100
km
for
VOC
emissions
and
no
more
than
200
km
for
NOx
emissions.

5
These
emission
reduction
targets
were
quite
varied.
Eight
of
16
areas
had
emission
reduction
targets
ranging
from
5%
to
20%
reductions
in
VOC
or
NOx,
relative
to
2010
base
case
emissions.
The
other
8
areas
had
targets
ranging
from
25%
in
NOx
only
to
a
50%
reduction
in
both
VOC
and
NOx
emissions,
relative
to
the
2010
base
case.

Selection
of
Local
Controls
for
Analysis.
AirControlNET
is
a
PC
based
relational
data
base
tool
for
conducting
control
strategy
and
costing
analyses.
It
overlays
a
detailed
control
measures
data
base
of
limited
scope
on
EPA
emissions
inventories
to
compute
source
and
pollutant
specific
emissions
reductions
and
associated
costs
at
various
geographic
levels.
4
As
used
in
this
analysis,
AirControlNET
included
a
broad
menu
of
stationary
source
control
measures
(
e.
g.,
add­
on
control
technologies
and
combustion
controls)
but
did
not
provide
a
comprehensive
menu
of
existing
controls
known
to
be
available
for
all
types
of
sources.
For
example,
due
to
a
lack
of
data
on
the
percentage
of
emission
reductions
and
associated
costs,
the
analysis
did
not
consider
additional
state
strategies
to
reduce
emissions
from
fuels,
highway
vehicles,
or
non­
road
engines.
This
lack
of
data
also
limited
the
number
of
control
measures
considered
in
area
emissions
sectors.

Emission
reduction
possibilities
are
considered
within
and
nearby
the
non­
attainment
area.
5
The
controls
are
presented
on
a
cost
per
ton
of
VOC
or
NOx
reduced
and
put
in
ascending
order.
Controls
are
selected
until
the
emission
reduction
target
is
met
or
until
a
cost
per
ton
level
of
$
10,000
is
reached.
This
limitation
was
used
in
this
analysis
to
avoid
large
upward
biases
in
the
control
cost
in
view
of
the
limited
number
of
control
measures
that
could
be
considered
for
selection
in
the
local
control
strategy.

Examples
of
the
NOx
controls
selected
for
some
of
the
NOx
emissions
sources
in
some
of
the
analyzed
areas
include
combustion
optimization,
low
NOx
burners,
low
excess
air,
and
other
combustion
controls;
selective
catalytic
reduction;
selective
non­
catalytic
reduction;
combinations
of
SCR
and
SNCR
with
additional
control
methods,
steam
injection,
natural
gas
reburn,
and
other
measures.

Examples
of
the
VOC
measures
selected
for
some
of
the
VOC
sources
in
some
of
the
analyzed
areas
include
measures
to
reduce
emissions
from
solvents,
adhesives,
paints
and
coatings
used
by
various
sectors,
consumer
products,
landfill
gas
collection
and
other
measures.

After
the
slate
of
stationary
and
area
source
controls
was
developed,
EPA
conducted
an
air
quality
simulation
to
estimate
the
air
quality
impact
of
those
controls.

Emission
Reduction
and
Air
Quality
Impacts
of
Analyzed
Local
Control
Strategy
The
analyzed
local
control
strategy
results
in
annual
NOx
emission
reductions
of
nearly
485
thousand
tons
and
nearly
292
thousand
tons
VOC
emission
reductions.
In
the
2010
base
case,
16
areas
6
have
design
values
in
excess
of
the
standard.
With
the
local
control
strategy,
10
of
the
16
areas
have
2010
design
values
that
are
projected
to
meet
the
standard
or
come
to
within
3
ppb
of
the
standard.

Areas
Projected
to
Meet
or
Approach
the
Standard
Through
Analyzed
Local
Strategy.
Of
the
original
16
areas
in
nonattainment
in
the
2010
base
case,
the
areas
projected
to
meet
the
standard
with
the
local
control
strategy
include
Beaumont­
Port
Arthur,
Detroit­
Ann
Arbor,
Kent­
Queen
Anne
Cos.,
Providence,
and
Washington,
D.
C.
Interestingly,
the
analysis
makes
apparent
that
both
Beaumont­
Port
Arthur
and
Providence
benefit
from
the
local
control
strategy
emission
reductions
for
upwind
nonattainment
areas,
and
it
is
likely
that
other
cities
in
the
analysis
benefited
from
upwind
reductions
as
well.

In
addition,
another
5
areas
­­
Atlanta,
Buffalo­
Niagara,
Cleveland­
Akron­
Lorain,
Dallas­
Fort
Worth,
and
Sheboygan
 
come
within
1
to
3
ppb
of
the
standard
with
the
analyzed
local
controls.
Note
that
actual
SIP
demonstrations
typically
include
not
only
local
modeling
but
other
types
of
information
in
a
"
weight
of
evidence"
analysis
which
a
state
may
use
to
help
demonstrate
that
a
proposed
set
of
control
measures
is
sufficient
to
provide
for
attainment;
such
considerations
are
beyond
the
scope
of
this
analysis.
If
needed,
further
emissions
reductions
to
assist
attainment
in
these
areas
could
come
from
measures
in
AirControlNET
and
other
measures
not
in
AirControlNET.

Table
2
provides
examples
of
emission
reducing
measures
not
included
in
the
analysis.

Table
2.
Examples
of
Measures
Not
Included
in
the
Analysis:
Cleaner
Gasoline
Measures:
Use
of
lower
RVP
or
other
cleaner
gasolines
Cleaner
Vehicle
Measures:
Compressed
natural
gas
and
clean
diesel
buses
Use
of
hybrid
vehicles
or
clean­
fueled
vehicles
for
fleets
Incentives
for
purchase
of
new
cleaner
vehicles
Re­
powering
and
retrofit
for
on­
road
diesel
vehicles
In
use
highway
vehicle
emissions
Truck
idle
reduction/
truck
stop
electrification
OBD­
based
I/
M
for
motor
vehicles
I/
M
for
heavy­
duty
vehicles
Remote
emissions
sensing
Non­
road
engine
emissions
early
replacement,
repowering,
and
retrofit
of
non­
road
diesel
equipment
Non­
road
diesel
idle
reduction
(
e.
g.,
locomotives)
Cleaner
gasoline
for
non­
road
engines
Lawnmower
rebates
and
gas
can
replacement
programs
Other
mobile
measures
Smartway/
improve
truck
and
rail
freight
efficiency
Port
emissions
reductions
(
e.
g.,
diesel
engines)
Airports
(
e.
g.,
ground
support
equipment,
ground
transport)

Stationary
source
measures
Most
process
changes
Fuel
switching
Energy
efficiency
improvements
in
any
sector
Urban
heat
island
reduction
6For
example,
the
NOx
control
efficiency
of
combustion
controls
has
increased
substantially
over
time;
EPA
estimated
70
percent
control
for
certain
applications
in
the
CAIR
rulemaking.

7
Antibacksliding
provisions
of
the
Phase
1
Ozone
Implementation
Rule
do
not
identify
the
1­
hour
new
source
review
requirement
as
an
applicable
requirement
for
purposes
of
the
8­
hour
standard.
Thus,
States
are
not
required
to
retain
1­
hour
NSR
obligations
to
the
extent
they
are
more
stringent
than
what
would
apply
for
an
area's
8­
hour
classification.

7
Other
Areas.
The
remaining
6
areas
are
Baltimore,
Chicago­
Gary­
Lake
Co.,
Houston­
Galveston­
Brazoria,
Milwaukee­
Racine,
New
York
City­
Long
Island,
and
Philadelphia­
Wilmington­
Atlantic
City.
Based
on
2001­
2003
data,
these
six
areas
had
8­
hour
design
values
ranging
from
101
ppb
to
106
ppb.
In
the
base
case
air
quality
model
simulation
for
2010,
the
projected
design
values
improve,
ranging
from
91.3
to
100.5
ppb.
With
the
analyzed
local
control
strategy,
ozone
levels
further
improve
but
remain
above
the
standard,
ranging
from
89.7
ppb
to
97.7
ppb.

The
analysis
suggests
that
the
analyzed
set
of
controls
would
not
bring
these
areas
into
attainment
by
2010.
However,
as
noted
above,
the
strategies
considered
in
this
analysis
were
not
comprehensive,
and
the
impacts
of
CAIR
were
not
reflected.
Further
work
beyond
the
scope
of
this
analysis
is
needed
to
assess
attainment
strategies
for
8­
hour
ozone
nonattainment
areas.
As
part
of
the
state
implementation
planning
process,
the
states
and
regional
planning
groups
will
conduct
more
detailed
local
emissions
inventory
projections
and
modeling
to
more
precisely
estimate
emission
reduction
needs,
and
comprehensively
consider
potential
control
measures.
In
addition,
factoring
in
updated
improvements
in
control
technologies
could
foster
air
further
decreases
in
the
projected
post
strategy
design
values.
6
Based
on
those
analyses,
the
states
and
EPA
will
make
judgments
concerning
control
needs,
control
options,
and
practicable
attainment
deadlines
as
part
of
the
SIP
process.

In
the
event
an
area
cannot
practicably
attain
by
the
maximum
date
for
its
classification,
the
Clean
Air
Act
provides
the
opportunity
for
more
time.
An
area
regulated
under
subpart
2,
such
as
the
six
areas
discussed
above,
could
receive
a
later
attainment
date
through
a
state
request
to
bump­
up
to
a
higher
classification
(
e.
g.
from
moderate
to
serious).
The
Act
requires
EPA
to
a
grant
a
state
request
for
an
area
to
receive
a
higher
classification;
the
state
plan
still
must
have
an
attainment
date
that
is
as
expeditious
as
practicable.
Although
bump­
up
means
that
certain
additional
specified
requirements
apply,
an
area
may
already
be
meeting
most
or
all
of
these
specified
requirements
due
to
controls
previously
adopted
to
implement
the
1­
hour
ozone
standard.
This
is
because
these
areas
had
1­
hour
classifications
that
were
higher
than
their
8­
hour
classifications,
and
because
the
Phase
1
final
implementation
rule
for
the
8­
hour
O3
NAAQS
contains
anti­
backsliding
provisions
generally
requiring
continued
implementation
of
the
control
measures
required
for
the
1­
hour
classification.
Although
additional
statutorily
specified
control
requirements
may
be
satisfied
by
measures
already
in
place,
the
area's
plan
still
must
include
measures
sufficient
to
provide
for
attainment
by
its
new
attainment
date.
These
reductions
may
be
achieved
through
implementation
of
measures
that
are
necessary
to
demonstrate
reasonable
further
progress
(
RFP),
or
additional
reductions
beyond
RFP
may
be
needed
to
demonstrate
timely
attainment.
Preliminary
analyses
indicate
that
already
required
control
measures
(
e.
g.,
motor
vehicle
and
nonroad­
engine
rules,
CAIR,
etc.)
may
largely
or
fully
fulfill
reasonable
further
progress
requirements
for
many
areas.
Minimum
federal
new
source
review
requirements
would
be
more
stringent
due
to
bump­
up.
7
However,
because
States
may
choose
whether
to
retain
1­
hour
NSR
requirements
that
are
more
stringent
than
required
for
an
area's
8­
hour
classification,
an
area­
by­
area
examination
would
be
necessary
to
determine
whether
bump­
up
would
8
require
a
State
to
adopt
more
stringent
NSR
rules
than
already
exist
in
the
approved
SIP.

One
reason
that
more
time
could
help
such
an
area
attain
is
that
emissions
of
ozone
precursors,
and
ozone
levels,
are
projected
to
decline
in
the
eastern
United
States
for
the
next
decade
and
beyond
as
a
result
of
existing
federal
rules
(
e.
g.,
motor
vehicle
and
non­
road
engine
emission
standards,
CAIR).
In
addition,
more
time
could
allow
for
implementation
of
innovative
measures
and
emerging
technologies
that
could
achieve
additional
pollution
reductions.
Our
experience
over
the
past
30
years,
and
the
promise
of
numerous
cleaner
technologies
emerging
today,
strongly
suggest
that
technological
innovation
and
"
learning
by
doing"
will
continue
to
produce
new,
cleaner
processes
and
performance
improvements
that
reduce
air
pollution
at
reasonable
cost.
The
Clean
Air
Act
itself
has
spurred
such
advances,
as
innovative
companies
have
responded
to
the
challenges
of
the
Act
with
great
success,
producing
breakthroughs
such
as
alternatives
to
ozone­
depleting
chemicals
and
new
super­
performing
catalysts
for
automobile
emissions,
as
well
as
improvements
in
control
efficiency
and
cost
for
technologies
such
as
scrubbers
and
SCR.
EPA
is
compiling
many
examples
of
emerging
or
developing
technologies
that
may
help
in
meeting
air
quality
goals
in
the
short
or
medium
term.

Cost
Estimates
The
estimated
cost
of
the
analyzed
local
control
strategy
is
$
2.0
billion
in
year
2000
dollars.
However,
the
estimated
cost
of
control
is
not
distributed
evenly
across
the
16
areas.
This
geographic
variation
in
cost
is
the
result
of
different
emission
reduction
targets,
source
emissions
inventories,
and
control
measures
applicable
to
that
area.
See
Table
3
regarding
the
variations
in
total
cost
and
Table
4
for
the
variations
in
marginal
cost
(
the
cost
of
the
highest
dollar
control
measure
selected
in
the
local
strategy).

Table
3.
Variations
in
Total
Cost
Across
Non­
attainment
Areas
Range*
of
Annual
Cost
Areas
with
Total
Local
Control
Strategy
Cost
within
that
Range
$
20
thousand
to
$
3.3
million
Kent­
Queen
Anne
Cos.;
Beaumont­
Port
Arthur;
Sheboygan;
Providence
Atlanta
$
14
million
to
$
61
million
Cleveland­
Akron­
Lorain;
Washington
DC­
MD­
VA;
Buffalo­
Niagra;
Detroit­
Anne
Arbor;
Milwaukee­
Racine;
Dallas­
Fort
Worth;
Baltimore
$
290
million
to
$
650
million
Chicago­
Cook
&
Lake
Cos.;
Houston­
Galveston­
Brazoria;
Philadelphia­
Wilmington
Atlantic
City,
New
York
City­
Long
Island
*
The
ranges
are
specified
so
as
to
avoid
overstating
the
variations
in
cost.
For
example,
there
are
no
areas
that
have
estimated
local
control
strategy
costs
ranging
from
$
3.4
to
$
13
million.
Hence,
the
range
is
not
artificially
expanded
to
go
from
$
3.4
million
to
$
61
million.
9
Table
4.
Variations
in
the
Marginal
Cost
of
Control
Across
the
Non­
attainment
Areas
Range*
of
Marginal
Cost
Areas
with
Local
Control
Strategy
Marginal
Cost
within
that
$'
s
per
ton
Range
60
to
1000
Kent­
Queen
Anne
Cos.;
Beaumont­
Port
Arthur;
Sheboygan;
Atlanta;
Cleveland­
Akron­
Lorain;
Washington
DC­
MD­
VA
1700
to
3400
Buffalo­
Niagra
Falls;
Dallas­
Fort
Worth;
Detroit­
Ann
Arbor;
Providence
5600
to
7300
Baltimore,
Milwaukee­
Racine;
Chicago­
Cook
&
Lake
Cos.

8700
to
10,000
Philadelphia­
Wilmington­
Atlantic
City;
New
York
City­
Long
Island;
Houston­
Galveston­
Brazoria
*
The
ranges
are
specified
so
as
to
avoid
overstating
the
variations
in
marginal
cost.
For
example,
there
are
no
areas
with
marginal
cost
range
from
$
1100
to
$
1600
per
ton.
Hence,
that
range
is
not
artificially
expand
to
go
from
$
60
to
$
1600
per
ton.

With
consideration
of
available
local
control
measures
not
considered
in
this
analysis,
some
areas
may
be
able
to
achieve
the
standard
at
lower
costs
while
other
areas
may
achieve
further
emission
reductions
and
air
quality
improvements
for
additional
costs.

Benefits
The
simulated
local
control
strategy
resulted
in
485
thousand
tons
of
NOx
reductions
and
292
thousand
tons
of
VOC
reductions.
NOx
is
a
precursor
to
both
fine
particulate
matter
(
PM2.5)
as
well
as
ozone.
VOC
reductions
may
include
some
constituents
which
are
classified
as
toxic
air
pollutants
while
also
being
a
precursor
to
fine
particulate
matter
and
ozone.
The
benefits
of
these
reductions
should
include
the
net
gains
to
human
health
and
welfare
effects
as
a
result
of
changes
in
concentrations
of
PM2.5,
ozone,
and
air
toxics.

The
quantified
and
monetized
benefit
estimates
are
the
result
of
a
multi­
step
procedure.
For
example,
in
the
case
of
ozone,
ozone
concentrations
are
predicted
for
certain
episodes
in
2010
using
the
base
case
emissions
inventory
and
CAM­
x.
Those
resulting
values
are
then
extended
to
the
entire
ozone
season.
With
the
local
control
strategy
NOx
and
VOC
emission
reductions,
there
is
a
different
2010
emission
inventory
reflecting
the
controls.
The
control
strategy
inventory
is
used
in
another
simulation
of
the
CAM­
x
model.
The
outputs
of
the
two
simulations
are
differences
in
ozone
concentrations
between
the
base
case
and
local
control
strategy
emissions
inventories.
These
ozone
concentrations
changes
are
then
used
in
concentration
response
models
to
predict
changes
in
an
endpoint.
These
endpoints
are
then
monetized.

In
the
case
of
particulate
matter
related
benefits,
there
are
fewer
steps.
Because
of
the
similarity
in
8
The
document
can
be
found
at
http://
www.
epa.
gov/
interstateairquality/
pdfs/
finaltech08.
pdf.

10
the
geographic
area
covered
by
the
CAIR,
the
anticipated
reduction
as
a
result
of
NOx
emission
reductions
from
electrical
generating
units,
the
emissions
to
air
quality
relationships
developed
there
are
used
in
this
assessment
by
scaling
quantified
and
monetized
benefits
in
terms
of
tons
in
the
CAIR
rule
versus
tons
reduced
in
the
local
control
strategy
simulation.
This
is
used
in
term
to
estimate
the
benefits
of
reduced
NOx
from
the
non­
EGU
sector.

Scope
of
the
Benefit
Assessment.
The
benefit
estimates
provided
in
the
local
control
strategy
assessment
are
incomplete.
Additional
air
quality
analysis
would
be
needed
to
more
fully
characterize
the
benefits.
Even
then,
there
are
categories
of
benefits
which
EPA
is
not
yet
able
to
quantify
or
monetize.
See
Table
5.
Also,
as
with
other
benefits
analyses,
the
benefits
estimates
are
subject
to
uncertainties
such
as
those
associated
with
estimates
of
future­
year
emissions
inventories
and
air
quality,
and
uncertainties
in
the
estimated
relationships
between
changes
in
pollutant
concentrations
and
resulting
changes
in
health
effects
(
e.
g.,
choice
of
study,
uncertain
biological
mechamism).

Table
5.
Only
Some
Potential
Benefits
of
the
Local
Control
Strategy
Quantified
&
Monetized
Emission
Reductions
Potential
Categories
of
Benefits
from
the
Local
Strategy
Ozone
Health
Ozone
Welfare
PM2.5
Health
PM2.5
Welfare
Toxics
NOx
Reductions
From
EGUs*
Estimated
Not
Covered
Estimated
Not
Covered
Not
Applicable
From
non­
EGUs
Estimated
Not
Covered
Estimated
Not
Covered
Not
Applicable
VOC
Reductions
Estimated
Not
Covered
Not
Covered
Not
Covered
Not
Covered
*
EGU
are
electrical
generating
units.

Ozone
Non­
Mortality
Health
Benefits.
Ozone
non­
mortality
health
benefits
are
quantified
based
on
the
same
health
endpoints
and
methodologies
used
in
the
Regulatory
Impact
Assessment
for
the
Final
Clean
Air
Interstate
Rule
(
EPA­
452/
R­
05­
002,
March
2005).
8
We
estimate
that
the
analyzed
control
strategy
in
2010
would
result
in
$
55
million
in
non­
mortality
health
benefits
associated
with
ozone
reductions
(
year
2000
dollars):

°
570
fewer
incidences
of
adult
hospital
admissions
for
respiratory
causes,
with
monetized
value
of
$
10
million
°
530
fewer
incidences
of
child
hospital
admissions
for
respiratory
causes,
with
monetized
value
of
$
4.1
million
°
130
fewer
emergency
room
visits
related
to
asthma,
with
monetized
value
of
$
0.04
million
°
260,000
fewer
school
absences,
with
monetized
value
of
$
20
million
°
280,000
fewer
minor
restricted
activity
days,
with
monetized
value
of
$
15
million
°
reduced
outdoor
worker
productivity
with
monetized
value
of
$
5.7
million.
11
Ozone
Mortality
Health
Benefits.
Over
the
past
several
years,
EPA
has
consulted
with
the
Science
Advisory
Board
regarding
evidence
for
an
independent
ozone
mortality
effect.
Because
of
new
studies
and
the
recommendations
from
the
SAB,
EPA
sponsored
three
independent
meta­
analyses
of
the
ozone­
mortality
epidemiology
literature
to
inform
a
determination
on
including
this
important
health
endpoint.
The
three
meta­
analyses
were
published
in
the
journal
Epidemiology
in
July
2005.
(
Bell
et
al,
2005;
Ito
et
al,
2005;
Levy
et
al,
2005)
The
meta­
analyses,
as
well
as
a
major
study
in
the
the
Journal
of
the
American
Medical
Association
(
Bell
et
al,
2004),
reported
that
on
average,
short­
term
changes
in
ozone
are
significantly
associated
with
premature
mortality,
and
that
the
significance
of
the
association
is
robust
to
adjustment
for
particulate
matter.
The
JAMA
study
used
the
extensive
National
Morbidity,
Mortality,
and
Air
Pollution
Study
database
to
examine
associations
between
ozone
and
premature
mortality
in
95
U.
S.
urban
communities.
The
Agency
believes
that
publication
of
these
studies
significantly
enhances
the
scientific
defensibility
of
benefits
estimates
for
ozone
that
include
the
benefits
of
premature
mortality
reductions.
In
the
future
we
plan
to
examine
a
variety
of
ozone
mortality
quantification
methods,
including
approaches
that
provide
information
on
relative
probability
of
different
benefits
levels.
Using
effect
estimates
similar
to
those
found
in
these
new
studies,
EPA
estimates
that
the
monetary
value
of
the
ozone­
related
premature
mortality
benefits
could
be
substantial.
In
the
interim,
for
this
analysis,
ozone
mortality
health
benefits
are
quantified
based
on
the
same
ozone
mortality
studies
and
quantification
methodology
used
in
the
Regulatory
Impact
Assessment
for
the
Final
Clean
Air
Interstate
Rule
(
EPA­
452/
R­
05­
002,
March
2005,
pp.
C­
9
to
C­
11).
We
estimate
ozone
mortality
benefits
from
the
analyzed
local
controls
scenario
for
2010
at
roughly
180
reduced
premature
mortalities
per
year
with
corresponding
monetized
benefits
of
approximately
$
1.1
billion
annually
(
year
2000
dollars).
As
with
other
health
endpoints,
use
of
ozone
mortality
effects
estimates
from
different
studies
could
result
in
a
higher
or
lower
estimate.

PM
Non­
Mortality
Benefits.
We
estimate
that
the
reductions
in
EGU
NOx
emissions
which
are
part
of
the
analyzed
control
strategy
in
2010
would
result
in
$
19
million
in
non­
mortality
health
benefits
associated
with
PM
reductions
(
year
2000
dollars),
including:

°
37
fewer
incidences
of
chronic
bronchitis
in
adults
26
years
and
older;
monetized
value
of
$
13
million
°
91
fewer
incidences
of
non­
fatal
myocardial
infactions
in
adults
18
and
older;
monetized
value
of
$
7.5
million
(
3%
discount
rate)/
$
7.3
million
(
7%
discount
rate)
°
23
fewer
incidences
of
hospital
admissions
for
respiratory
causes
(
includes
chronic
obstructive
pulmonary
disease,
pneumonia,
and
asthma);
monetized
value
of
$
0.17
million
°
20
fewer
incidences
of
hospital
admissions
for
cardiovascular
causes
(
includes
total
cardiovascular
and
subcategories
for
ischemic
heart
disease,
dysrhythmias
and
heart
failure)
in
those
17
and
older;
monetized
value
of
$
0.43
million
°
53
fewer
emergency
room
visits
for
asthma
in
those
18
years
or
younger;
monetized
value
of
$
0.02
million
°
85
fewer
incidences
of
chronic
bronchitis
in
children
8­
12
years
old;
monetized
value
of
$
0.03
million
°
1,012
fewer
incidences
of
lower
respiratory
symptoms
in
children
7­
14
years;
monetized
value
of
$
0.02
million
°
799
fewer
incidences
of
upper
respiratory
symptoms
in
asthmatic
children
9­
11
years
old;
monetized
value
of
$
0.02
million
°
1,279
fewer
incidences
of
aggravation
of
asthma;
monetized
value
of
$
0.05
million
°
7,459
fewer
work
loss
days
in
adults
18­
65
years;
monetized
value
of
$
0.96
million.
12
°
43,157
fewer
minor
restricted
activity
days
in
adults
18­
65
years;
monetized
value
of
$
2.1
million.

PM
Mortality
Benefits.
We
estimate
that
EGU
NOx
reductions
in
the
analyzed
control
strategy
will
result
in
69
fewer
instances
of
premature
mortality
associated
with
long­
term
exposure
to
PM,
with
a
monetized
value
of
$
361
million
at
a
3%
discount
rate,
or
$
304
million
at
a
7%
discount
rate.
The
CAIR
RIA
provides
relevant
background
on
EPA's
treatment
of
PM
mortality
in
this
assessment.

Additional
PM
Health
Benefits.
There
are
also
monetized
PM
benefits
from
local
control
strategy
NOx
reductions
in
the
non­
EGU
sector.
Emission
reductions
from
this
sector
accounts
for
about
two
thirds
of
the
Nox
emission
reductions
from
the
local
control
strategy.
These
reductions
come
from
sources
such
as
industrial
boilers,
cement
kilns,
and
gas
turbines.
If
the
estimated
health
benefits
per
ton
of
NOx
reduction
were
half
as
much
as
that
for
the
EGU
reductions,
the
total
would
be
$
340
million.
If
the
benefits
were
equivalent,
the
estimate
would
be
$
680
million.
Without
corresponding
air
quality
modeling
to
undergird
these
estimates
there
is
uncertainty.
Clearly,
the
PM
health
benefits
for
these
non­
EGU
NOX
reductions
are
not
zero.
The
average
of
these
two
estimates
is
chosen
for
purposes
of
this
illustrative
analysis.
We
are
assuming
that
emission
reductions
from
these
boilers
are
on
a
benefit
per
ton
basis
75%
of
that
seen
with
reductions
from
EGU
on
a
per
ton
basis.

Limitations.
The
incomplete
coverage
of
benefits
categories
is
a
source
of
uncertainty.
However,
there
are
also
other
sources
of
uncertainty.
These
include
the
following:

°
Omissions
of
benefits
(
e.
g.,
individual
health
endpoints
that
could
not
be
monetized).
°
Lack
of
local
control
strategy
specific
air
quality
modeling
regarding
PM2.5
changes
from
NOX
and
VOC
emission
reductions.
°
Several
sources
of
uncertainty
which
may
affect
the
benefits
estimates
(
and
not
necessarily
the
costs)
°
baseline
incidence
rates
for
the
effects
°
mortality
function
and
associated
lags
°
economic
valuation
and
aggregation
of
endpoints
°
averting
and
mitigating
behavior
°
other
factors
noted
in
CAIR
RIA
(
choice
of
study,
potential
for
a
threshold,
etc.)

°
Emission
reductions
for
the
benefits
analysis
include
greater
reductions
 
4631
and
5378
more
tons
of
NOX
and
VOC
respectively
 
than
the
corresponding
cost
analysis.
This
would
cause
the
benefits
to
be
slightly
overestimated
relative
to
the
costs.
