Summary
of
Comments
and
Responses
on
the
2004
and
2001
Proposed
Guidelines
for
Best
Available
Retrofit
Technology
(
BART)
Determinations
Under
the
Regional
Haze
Regulations
Docket
Number
OAR­
2002­
0076
Table
of
Contents
1.0
General
Comments
in
Support
of
or
Opposition
to
BART
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
1
1.1
Support
of
the
Proposed
Guidelines
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
1
1.1.1
Pass
a
Strong
BART
Rule
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
1
1.1.2.
Reduce
Visibility
Impairment
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
1
1.1.3
Improve
the
Economy
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
2
1.1.4
Improve
Public
Health
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
3
1.1.5
Improve
Ecosystem
Health
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
3
1.1.6
Comments
in
Support
of
a
Comprehensive
Approach
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4
1.1.7
Comments
Supporting
Related
Programs
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5
1.1.8
Miscellaneous
Comments
Supporting
BART
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5
1.2
Opposition
to
the
Proposed
Guidelines
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6
1.2.1
Multi­
pollutant
Legislation
is
Preferable
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6
1.2.2
Guidelines
Restrict
State
Discretion
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6
1.2.3
Guidelines
Should
Only
Be
Applicable
to
Large
Utilities
.
.
.
.
.
.
.
.
.
.
.
.
.
8
1.2.4
Guidelines
Not
Needed
or
Premature
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
9
1.2.5
BART
Should
Take
Other
Programs
into
Account
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
10
1.2.6
Other
Comments
in
Opposition
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
10
1.3
The
Role
of
the
Proposed
Guidelines
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
11
1.3.1
States'
Role
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
11
1.3.2
Focus
of
Regional
Haze
Solutions
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
12
1.3.3
60­
year
Timeframe
for
Restoring
Natural
Conditions
.
.
.
.
.
.
.
.
.
.
.
.
.
.
13
1.3.4
Responsibility
for
Federal
Emissions
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
13
1.3.5
Relationship
to
the
Regional
Haze
Rule
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
13
1.3.6
Nationwide
Scope
of
the
Guidelines
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
14
2.0
Identifying
the
BART­
eligible
Sources
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
14
2.1
Identify
the
Emission
Units
in
the
Bart
Categories
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
14
2.1.1
Source
Category
Definitions
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
15
2.1.2
Guidance
on
Definition
of
Emission
Units
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
25
2.2
Start­
up
Dates
of
Emission
Units
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
26
2.2.1
Definitions
of
"
In
Existence"
and
"
In
Operation"
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
26
2.2.2
Applicability
Only
to
Boilers
at
a
Plant
Put
in
Place
between
1962
­
1977
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
26
2.2.3
Applicability
to
Individual
Emission
Units
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
26
2.2.4
Eligibility
Dates
of
Emission
Units
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
28
2.2.5
Modified
Units
and
BART
Eligibility
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
30
2.3
Compare
Potential
Emissions
to
the
250
tpy
Cutoff
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
31
2.3.1
List
of
Visibility
Impairing
Pollutants
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
31
2.3.2
Defining
"
Stationary
Source"
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
37
2.3.3
Use
of
Potential
Emissions
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
39
2.4
Identify
the
Emission
Units
and
Pollutants
that
Constitute
the
BART­
eligible
Source
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
40
2.4.1
Requiring
Analysis
of
All
Pollutants
When
Only
One
Exceeds
the
250
typ
Threshold
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
41
2.4.2
De
Minimis
Emission
Levels
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
42
2.5
States'
Burden
in
Identifying
the
BART­
eligible
Source
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
47
2.6
General
Comments
Regarding
Identifying
the
BART­
eligible
Source
.
.
.
.
.
.
.
.
47
3.0
Identify
Sources
Subject
to
BART
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
48
3.1
Methods
of
Determining
Whether
a
Source
is
Subject
to
BART
.
.
.
.
.
.
.
.
.
.
.
.
48
3.1.1
States
Assumes
That
All
Sources
are
Subject
to
BART
.
.
.
.
.
.
.
.
.
.
.
.
.
49
3.1.2
State
Determines
That
No
Sources
are
Subject
to
BART
Using
a
Cumulative
Approach
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
50
3.1.3
State
Makes
a
Source­
by­
Source
Determination
Using
an
Individualized
Exemption
Process
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
51
3.1.4
Other
Comments
Regarding
the
Methods
to
Determine
Whether
a
Source
is
Subject
to
BART
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
53
3.2
Source
Exemption
Process
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
54
3.2.1
Exemption
Process
for
Sources
More
Than
50
km
and
Less
Than
200
km
from
a
Class
I
Area
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
54
3.2.2
Exemption
Process
for
Sources
More
Than
200
km
from
a
Class
I
Area
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
54
3.2.3
Modeling
Sources
Less
Than
50
km
from
a
Class
I
Area
.
.
.
.
.
.
.
.
.
.
.
.
55
3.2.4
Zero­
out
Approach
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
56
3.2.5
General
Comments
Regarding
the
Source
Exemption
Process
.
.
.
.
.
.
.
.
58
3.3
Alternative
Approaches
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
66
3.3.1
Screening
Assessment
Using
CALPUFF
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
66
3.3.2
Look­
up
Tables
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
67
3.3.3
Source
Ranking
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
67
3.3.4
Emissions
Divided
by
Distance
(
Q/
D)
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
68
3.3.5
General
Comments
Regarding
Alternative
Approaches
.
.
.
.
.
.
.
.
.
.
.
.
.
69
3.4
Visibility
Thresholds
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
70
3.4.1
Use
of
Deciview
as
a
Metric
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
70
3.4.2
Compare
to
Visibility
Under
Natural
Conditions
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
70
3.4.3
Threshold
Level
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
74
3.4.4
General
Comments
Regarding
Visibility
Thresholds
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
79
3.5
General
Comments
Regarding
Identifying
Sources
Subject
to
BART
.
.
.
.
.
.
.
.
82
4.0
BART
Determination
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
83
4.1
Engineering
Analysis
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
83
4.1.1
Identify
Available
Retrofit
Emission
Control
Technologies
.
.
.
.
.
.
.
.
.
.
83
4.1.2
Determine
Whether
Options
are
Technically
Feasible
.
.
.
.
.
.
.
.
.
.
.
.
.
.
86
4.1.3
Evaluate
Impacts
and
Report
Results
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
88
4.1.4
Evaluate
Visibility
Impacts
of
Applying
Controls
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
97
4.1.5
Applicability
to
Plants
that
Have
Installed
BACT
or
LAER
or
are
Subject
to
MACT
or
NSPS
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
106
4.1.6
General
Comments
Regarding
the
Engineering
Analysis
.
.
.
.
.
.
.
.
.
.
.
107
4.2
Select
a
Best
Alternative
Using
the
Results
of
an
Engineering
Analysis
.
.
.
.
.
.
108
4.2.1
Select
a
"
Best"
Alternative
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
108
4.2.2
Consideration
of
the
Affordability
of
Controls
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
114
4.2.3
Presumptive
Emission
Limits
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
117
4.3
General
Comments
Regarding
BART
Determination
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
136
4.3.1
Comments
Regarding
Streamlining
the
BART
Determination
Process
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
136
4.3.2
Comments
Regarding
Reasonable
Progress
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
136
4.3.3
Other
General
Comments
Regarding
the
BART
Determination
.
.
.
.
.
.
137
5.0
Enforceable
Limits/
Compliance
Date
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
137
5.1
Proposed
Elimination
of
"
Committal"
SIP
Deadlines
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
137
5.2
Proposed
Harmonization
of
Deadlines
for
Section
308
and
309
SIPs
.
.
.
.
.
.
.
138
5.3
Compliance
Deadlines
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
139
5.4
Monitoring
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
139
5.5
Enforceable
Limits
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
140
6.0
Statutory
and
Executive
Order
Reviews
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
140
6.1.1
Administrative
Process
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
140
6.1.2
Executive
Order
12866
­
Regulatory
Impact
Analysis
.
.
.
.
.
.
.
.
.
.
.
.
.
140
6.1.3
Regulatory
Flexibility
(
SBREFA)
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
142
6.1.4
Paperwork
Reduction
Act
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
143
6.1.5
Unfunded
Mandates
Reform
Act
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
143
6.1.6
Federalism
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
143
6.1.7
Executive
Order
13211
­
Actions
Concerning
Regulations
that
Significantly
Affect
Energy
Supply,
Distribution
or
Use
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
144
7.0
Miscellaneous
Comments
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
145
7.1
Process
Related
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
145
7.2
Definition
of
Terms
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
147
7.3
Role
of
Federal
Land
Managers
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
148
7.4
Comments
about
Specific
Class
I
Areas
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
148
7.5
Other
Miscellaneous
Comments
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
149
8.0
Editorial
Comments
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
149
9.0
Comments
Unrelated
to
BART
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
149
10.0
Comment
Related
to
Trading
1
1.0
General
Comments
in
Support
of
or
Opposition
to
BART
1.1
Support
of
the
Proposed
Guidelines
1.1.1
Pass
a
Strong
BART
Rule
Comment:
Eighty­
three
commenters
urged
EPA
to
adopt
a
strong
BART
rule
to
bring
cleaner
air
to
our
national
parks,
improve
visibility,
improve
public
health,
improve
the
economy,
and/
or
protect
ecosystems
(
0152,
0153,
0154,
0155,
0156,
0159,
0162,
0164,
0165,
0167,
0203,
0204,
0221,
0227,
0234,
0235,
0260,
0261,
0262,
0263,
0264,
0284,
0292,
0305,
0313,
0318,
28­
3,
28­
14,
28­
15,
28­
16,
28­
42,
28­
47,
28­
98,
28­
127,
28­
221,
28­
239,
28­
247,
28­
248,
28­
249,
28­
250,
28­
263,
28­
271,
28­
327,
28­
328,
28­
334,
28­
335,
28­
338,
28­
241,
28­
262,
28­
265,
28­
266,
28­
285,
28­
286,
28­
290,
28­
291,
28­
292,
28­
295,
28­
307,
28­
362,
28­
377,
28­
380,
28­
394,
28­
412,
28­
428,
28­
433,
28­
435,
28­
477,
28­
490,
28­
501,
28­
505,
28­
507,
28­
513,
28­
526,
28­
534,
28­
535,
28­
536,
28­
538,
28­
562,
28­
592,
28­
7002,
28­
7006,
28­
7016,
28­
7048).
Action
is
long
overdue
(
0203,
28­
98,
28­
127)
and
a
strong
program
is
needed
to
meet
Congressional
intent
to
prevent
future
and
remedy
existing,
visibility
impairment
in
each
individual
Class
I
area
(
0227).
The
BART
requirements
are
the
key
mechanism
for
obtaining
emissions
reductions
and
visibility
improvement
in
the
regional
haze
program
(
0234).
Within
the
boundaries
of
the
court
May
2002
court
decision,
EPA
should
maintain
the
most
stringent
BART
provisions
possible
(
0234).

Commenter
28­
7048
said
BART
is
critical
guidance
that
helps
the
Forest
Service
meet
its
affirmative
responsibility
under
the
Act
and
their
Government
Performance
and
Results
Act
(
GPRA)
goals
for
cleaner
air.

One
commenter
(
0262)
said
a
weak
BART
rule
will
only
increase
volatility
and
regulatory
compliance
costs.
A
weak
rule
invites
revision
and
replacement
by
subsequent
administrations
and
litigation
by
environmental
and
public
health
groups.
In
contrast,
a
strong
BART
rule
will
be
inevitably
implemented
in
one
way
or
another
and
if
the
EPA
chooses
to
do
it
right
the
first
time,
it
will
help
to
dispel
uncertainty
and
multiple
rounds
of
compliance
costs
going
forward.
Commenter
28­
7006
said
EPA
should
take
regulatory
action
so
that
lawsuits
do
not
need
to
be
filed
to
ensure
a
basic
right
of
Americans
and
the
land.

Several
commenters
stressed
the
need
to
reduce
emissions
from
grandfathered
fossilfueled
electric
plants
and/
or
industrial
polluters
(
28­
16,
28­
127,
28­
250,
28­
257,
28­
271,
28­
327,
28­
328,
28­
334,
28­
296,
28­
252
28­
376,
28­
382,
28­
434,
28­
546,
28­
546,
28­
568,
28­
592).
Other
commenters
stressed
the
importance
of
including
all
states
in
the
final
program
(
28­
239,
28­
249,
28­
250,
28­
285).

Response:
2
We
agree
with
the
above
comments
and
believe
we
have
promulgated
a
strong
BART
program.

1.1.2.
Reduce
Visibility
Impairment
Comment:
Twenty­
two
commenters
stressed
the
need
to
protect
and
improve
visibility
in
Class
I
areas
(
0147,
0148,
0153,
0154,
0159,
0165,
0167,
0203,
0204,
0221,
0235,
0260,
0284,
28­
350,
28­
254,
28­
358,
28­
412,
28­
428,
28­
526,
28­
547,
28­
568,
28­
7002).
Commenters
said
that
states
must
have
the
necessary
tools
to
do
so
(
0147,
0148,
0235).
Many
commenters
cited
their
expectation
to
find
clean
air
and
clear
views
in
the
national
parks,
but
this
is
often
not
the
case
due
to
significant
visibility
reductions
(
0153,
0154,
0159,
0167,
0203,
0204,
0221,
0260,
0284,
28­
3,
28­
239,
28­
241,
28­
245,
28­
248,
28­
254,
28­
260,
28­
285,
28­
295,
28­
313,
28­
334,
28­
338,
28­
350,
28­
394,
28­
433,
28­
490,
28­
513,
28­
562).
Others
were
concerned
that
visibility
is
decreasing
in
many
areas,
citing
data
and
studies
as
confirmation
(
0167,
0221,
28­
338,
28­
352,
28­
358,
28­
362,
28­
399,
28­
526,
28­
538,
28­
7016
).
One
commenter
(
28­
400)
noted
support
for
the
clean
air
and
visibility
protection
programs.

Response:

We
agree,
and
believe
the
BART
program,
in
conjunction
with
the
overall
regional
haze
program,
will
provide
States
with
the
necessary
tools
to
significantly
improve
visibility
in
Class
I
areas.

Comment:
Seven
commenters
discussed
the
impacts
of
power
plant
emissions
and
pollution
from
other
states
as
a
cause
of
visibility
impairment
(
28­
257,
28­
260,
28­
286,
28­
302,
28­
358,
28­
399,
28­
536).
Some
commenters
provided
information
on
the
role
of
precursors
(
28­
302,
28­
358).

One
commenter
noted
that
the
existing
stationary
sources
affected
by
the
proposal
are
not
the
only
contributors
to
regional
haze
(
28­
278).

Response:

We
agree
and
note
that
the
BART
program
will
achieve
significant
emission
reductions
from
power
plants
and
will
generally
reduce
interstate
transport
of
particulate
emissions.

1.1.3
Improve
the
Economy
Comment:
Ten
commenters
(
0157,
0158,
1067,
0227,
0262,
0292,
28­
221,
28­
241,
28­
248,
28­
257)
described
the
economic
benefits
of
improving
visibility
in
Class
I
areas.
Tourism
results
in
enormous
economic
benefits,
which
are
threatened
by
hazy
vistas
(
0167,
0227,
0262,
0292,
28­
3
221,
28­
241,
28­
248,
28­
262,
28­
302,
28­
313,
28­
350,
28­
352,
28­
491,
28­
513,
28­
526,
28­
538).
Commenters
described
various
regional
economic
benefits
due
to
tourism
(
0167,
0227,
0292,
28­
352)
and
one
noted
a
study
by
the
Clean
Air
Task
Forces,
which
reported
the
cost
of
decreased
visitation
to
Acadia
National
Park
due
to
haze
as
$
328
million
lost
annually
(
0227).
Other
commenters
noted
results
of
a
survey
of
people
planning
to
visit
the
White
Mountains
of
New
Hampshire
that
64
to
68
percent
said
they
would
be
less
likely
to
visit
if
conditions
worsened
(
0227,
0292,
28­
352).

Commenters
also
cited
water
pollution
impacts
and
resulting
cleanup
costs
from
BART
sources
(
0157),
improving
public
health
(
0165,
0262),
and
acidification
of
the
soil
that
is
harming
wildlife
and
water
quality
(
0262).
Commenter
0165
said
that
the
costs
of
cleaning
our
air
are
more
than
repaid
in
better
health
and
tourism.
Commenters
(
28­
248,
28­
491.
28­
594)
said
NO
x
also
is
of
concern
from
its
impacts
on
agriculture,
silviculture,
and
other
commodity
production.

Commenter
28­
262
said
strong
BART
guidance
will
provide
cost­
effective
emissions
reductions
from
existing
sources
allowing
states,
tribes,
and
federal
land
managers
more
latitude
in
planning
for
future
economic
and
energy
growth.

Commenter
28­
313
cited
an
analysis
of
the
positive
impact
of
a
mountain
view
on
property
values,
which
supports
local
economies
with
ongoing
property
taxes.
Commenter
28­
350
said
economies
hundreds
of
miles
away
from
the
parks
will
benefit
from
requiring
older
power
plants
and
industrial
boilers
to
meet
modern
day
standards.
A
report
by
Abt
Associates
estimates
that
the
value
of
eliminating
power
plant
haze
is
over
seven
billion
dollars
a
year.

Response:

We
agree
that
regional
haze
affects
tourism
and
recognize
that
the
BART
rule
will
improve
conditions
in
Class
I
areas
that
affect
tourism.
We
believe
that
the
BART
program,
together
with
our
overall
approach
to
regional
haze,
will
help
address
the
above
cited
ecological
issues.
We
also
agree
that
strong
BART
guidance
will
provide
cost­
effective
reductions,
and
believe
that
today's
rule
represents
such
strong
guidance.
We
also
recognize
the
contribution
of
power
plant
emissions
to
regional
haze
and
believe
that
BART,
in
conjunction
with
other
EPA
rules,
will
help
address
emissions
from
older
power
plants.

1.1.4
Improve
Public
Health
Comment:
Fifty­
six
commenters
discussed
the
compelling
need
to
achieve
the
public
health
benefits
of
reducing
pollution
from
BART
sources
(
0153,
0154,
0156,
0158,
0159,
0163,
0165,
0203,
0204,
0227,
0260,
0262,
0263,
0284,
0292,
0313,
28­
13,
28­
14,
28­
42,
28­
98,
28­
127,
28­
221,
28­
239,
28­
247,
28­
248,
28­
250,
28­
257,
28­
263,
28­
265,
28­
271,
28­
286,
28­
290,
28­
302,
28­
313,
28­
328,
28­
334,
28­
338,
28­
346,
28­
352,
28­
358,
28­
362,
28­
376,
28­
394,
28­
399,
28­
412,
28­
438,
38­
477,
28­
490,
28­
491,
28­
513,
28­
526,
28­
529,
28­
535,
28­
562,
28­
7006,
28­
7016).
4
Many
of
these
commenters
were
specifically
concerned
about
health
impacts
from
fine
particulates.
Commenters
noted
that
because
many
of
the
sources
are
outside
of
park
boundaries,
the
air
quality
in
their
local
communities
will
be
improved
resulting
in
improved
public
health
(
0153,
0154,
0156,
0159,
0260,
0262,
0263,
0284,
28­
358).

Commenters
were
also
concerned
about
the
adverse
health
impacts
to
hikers
and
other
park
visitors
from
pollution
from
regional
haze
sources
(
0227,
0292,
28­
239,
28­
250,
28­
334,
28­
352,
28­
358,
28­
362,
28­
394,
28­
412,
28­
490).
One
commenter
cited
a
study,
Effects
of
Ozone
and
Other
Pollutants
on
the
Pulmonary
Function
of
Adult
Hikers,
published
in
Environmental
Health
Perspectives
1998,
that
demonstrated
that
ozone,
and
to
a
lesser
extent
fine
particulate
matter,
result
in
acute
respiratory
impacts
to
healthy,
active
adults
hiking
in
Eastern
parklands.

Response:

We
agree
that
the
BART
rule
will
have
significant
public
health
benefits.

1.1.5
Improve
Ecosystem
Health
Comment:
Three
commenters
said
that
improving
ecosystem
health
is
another
goal
of
the
program
(
0165,
0227,
0292).
One
commenter
(
0292)
said
that
smog
and
acid
rain
impair
natural
habitat
and
alter
fragile
ecosystems
in
mountain
environments.
In
addition
to
improved
visibility,
a
strong
BART
rule
will
provide
critical
environmental
benefits
by
reducing
the
emissions
of
pollutants
that
contribute
to
acid
deposition,
ozone
formation,
nitrogen
saturation,
and
estuary
eutrophication.
Another
commenter
(
0227)
provided
data
demonstrating
the
fragility
of
high
elevation
ecosystems.

Twenty­
six
commenters
described
the
extent
of
damage
to
ecosystems
from
air
pollution,
including
acid
rain
(
28­
13,
28­
14,
28­
15,
28­
98,
28­
221,
28­
241,
28­
248,
28­
250,
28­
262,
28­
263,
28­
286,
28­
287,
28­
302,
28­
328,
28­
334,
28­
338,
28­
352,
28­
358,
28­
362,
28­
376,
28­
394,
28­
526,
28­
535,
28­
562,
28­
7006,
28­
7048
)
and
the
impacts
on
resources,
tourism,
enjoyment
of
the
landscape,
etc..
Application
of
BART
will
benefit
these
resources
(
28­
262,
28,
263,
28­
302,
28­
328,
28­
352,
28­
526,
28­
7048).

Response:

We
believe
that
the
BART
program,
together
with
our
overall
approach
to
regional
haze,
will
help
address
the
above
cited
ecological
issues.
5
1.1.6
Comments
in
Support
of
a
Comprehensive
Approach
Comment:
Twenty­
five
commenters
supported
applying
the
BART
guidelines
for
all
of
the
affected
sources
and/
or
source
categories
listed
in
the
regional
haze
rule
to
ensure
visibility
goals
are
met
and
to
ensure
national
consistency
(
0159,
0167,
0200,
0232,
0284,
0313,
28­
127,
28­
255,
28­
256,
28­
258,
28­
263,
28­
326,
28­
266,
28­
290,
28­
302,
28­
325,
28­
327,
28­
335,
28­
350,
28­
358,
28­
362,
28­
389,
28­
392,
28­
404,
28­
7013).
One
commenter
supported
strategies
that
consider
many
sources
simultaneously
on
a
regional
basis,
consistent
with
National
Research
Council
recommendations
(
0167).
One
commenter
(
0232)
said
granting
discretionary
use
of
the
BART
guidelines
will
present
an
impediment
to
nationally
consistent
implementation
and
potentially
raise
issues
of
economic
competitiveness
between
neighboring
states
which
may
choose
to
apply
the
guidelines
differently.
The
commenter
urged
EPA
to
mandate
use
of
the
proposed
BART
guidelines
consistently
across
all
potentially
BART­
eligible
sectors.

Commenter
28­
358
added
that
EPA
has
clear
authority
and
is
obligated
under
the
Act
to
establish
regional
haze
BART
guidelines,
by
rulemaking,
for
all
26
categories
of
major
stationary
sources
set
forth
in
the
Act.
The
focus
on
large
fossil­
fuel
fired
powerplants
in
the
last
part
of
section
169A(
b)(
2)
and
in
the
exemption
process
at
section
169A(
c)
suggest
that
Congress
was
more
than
aware
of
the
disproportionate
contribution
of
those
large
power
plants'
emissions
to
haze­
producing
pollutants.
That
awareness
does
not,
however,
render
meaningless
those
portions
of
sections
169A(
b)(
2)(
A)
and
169A(
g)(
7)
which
clearly
demonstrate
that
section
169A
applies
to
a
wide
range
of
sources
and
not
just
powerplants.

Commenter
28­
254
said
the
guidelines
should
be
used
for
all
sources
contributing
to
visibility
impairment,
which
will
provide
a
more
equitable
application
of
the
requirement
among
the
states
and
Tribes
implementing
the
visibility
protection
program.

Response:

The
Clean
Air
Act
(
CAA)
and
the
relevant
legislative
history
make
clear
that
EPA
has
the
authority
and
obligation
to
publish
mandatory
guidelines
for
powerplants
exceeding
750
megawatts.
As
previously
noted,
Congress
in
section
169A(
b)
of
the
CAA
expressly
provided
that
emission
limitations
for
powerplants
larger
than
750
megawatts
"
shall
be
determined
pursuant
to
guidelines
promulgated
by
the
Administrator."
(
Emphasis
added).
This
unambiguous
language
leaves
little
room
to
dispute
that
the
guidelines
EPA
is
required
to
promulgate
must
be
used
by
States
when
making
BART
determinations
for
this
class
of
sources.
Having
carefully
considered
the
comments
and
further
reviewed
the
CAA
and
the
legislative
history,
we
have
concluded
that
it
would
not
be
appropriate
for
EPA
to
require
States
to
use
the
guidelines
in
making
BART
determinations
for
other
categories
of
sources.
The
better
reading
of
the
Act
indicates
that
Congress
intended
the
guidelines
to
be
mandatory
only
with
respect
to
750
megawatt
powerplants.
Thus,
while
we
agree
with
the
State
agency
comments
noting
the
strong
policy
reasons
in
support
of
covering
all
categories
in
order
to
establish
6
consistency
across
States
in
how
BART
determinations
are
made,
we
are
not
requiring
States
to
use
the
BART
guideline
for
these
other
categories.
In
response
to
State
concerns
about
equitable
application
of
the
BART
requirement
to
source
owners
with
similar
sources
in
different
states,
we
do
encourage
States
to
follow
the
guidelines
for
all
source
categories
but
are
not
requiring
States
to
do
so.
States
should
view
the
guidelines
as
guidance
for
these
other
categories.
We
disagree
with
comments
that
the
CAA
and
the
legislative
history
prohibit
us
from
issuing
guidance
for
other
source
categories.
As
the
guidelines
make
clear,
States
are
not
required
to
follow
the
approach
in
the
guidelines
for
sources
other
than
750
megawatt
powerplants.
As
such,
although
we
believe
that
the
guidelines
provide
useful
advice
in
implementing
the
BART
provisions
of
the
regional
haze
rule,
we
do
not
believe
that
they
hamper
State
discretion
in
making
BART
determinations.

1.1.7
Comments
Supporting
Related
Programs
Comment:
Five
commenters
suggested
additional
air
programs
that
should
be
supported
by
EPA
(
0165,
0175,
0203,
0318).
One
commenter
(
0203)
urged
EPA
to
finalize
a
stringent
Clean
Air
Interstate
Rule
(
CAIR)
that
achieves
significant
and
timely
reductions
in
air
pollution,
and
delivers
clean
air
to
many
more
millions
of
Americans.
Another
commenter
(
0318)
asked
EPA
to
strengthen
pending
power
plant
legislation
because
of
the
toll
on
parks.
The
commenter
added
that
mandatory
limits
on
carbon
dioxide
should
also
be
established
to
give
companies
regulatory
certainty.
Two
commenters
(
0165,
0175)
supported
all
clean
air
regulations.

One
commenter
(
0302)
said
the
state's
top
priority
is
attaining
health­
based
standards,
including
ozone
and
PM2.5.
The
commenter
supports
the
BART
controls,
but
to
the
extent
possible,
haze
reduction
strategies
should
be
coordinated
with
the
health­
based
standards
and
tailored
to
fit
with
existing
programs
and
future
programs
such
as
multi­
pollutant
control
initiatives.

Nine
commenters
said
that
the
reductions
should
be
permanent
and
above
and
beyond
the
emission
caps
established
for
the
Title
IV
and
NO
x
SIP
call
(
28­
247,
28­
248,
28­
286,
28­
302,
28­
338,
28­
350,
28­
358,
28­
399,
28­
7016).
Commenter
28­
302
added
that
failure
to
account
for
potential
interactions
between
different
regulatory
programs
may
significantly
diminish
the
BART
program's
ability
to
deliver
visibility
and
public
health
benefits
beyond
those
that
would
accrue
under
existing
programs.
EPA
should
revise
the
proposed
guidelines
to
address
the
potential
interaction
of
BART­
generated
emissions
allowances
and
existing
trading
programs
and
more
specifically
to
ensure
that
BART
implementation
provides
for
substantial
additional
emissions
reductions
that
will
be
needed
to
ensure
continued
progress
toward
national
visibility
goals.
Other
commenters
said
that
this
clarification
is
needed
for
EPA
to
fulfill
its
statutory
mandate
to
ensure
"
reasonable
progress"
(
28­
358,
28­
7016).

Response:
7
We
recognize
the
interrelated
nature
of
clean
air
programs
and
are
working
as
an
Agency
to
ensure
the
goals
of
the
Clean
Air
Act
(
CAA)
are
met.
As
noted
in
the
guidelines,
the
States
should
consider
coordinated
regulatory
strategies
when
developing
SIPs.
Note
that
Congress
amended
the
deadline
for
submission
of
regional
haze
SIPs
to
align
better
with
nonattainment
SIPs.
Congress
did
not
include
as
one
of
the
BART
factors
a
requirement
that
emission
reductions
be
surplus
to
those
required
by
other
provisions
of
the
Act.
Where
emission
controls
have
been
installed
to
meet
other
requirements,
they
may
be
considered
under
the
factor
for
existing
controls.
The
BART
provisions
establish
the
federally
required
minimum
controls
for
sources
subject
to
BART,
in
order
to
fulfill
their
obligation
towards
the
reasonable
progress
goals.
A
State
is
are
always
free,
as
a
matter
of
federal
law,
to
require
additional
reductions
if
the
State
believes
this
is
necessary
to
make
reasonable
progress.

1.1.8
Miscellaneous
Comments
Supporting
BART
Comment:
Three
commenters
pointed
to
BART's
flexibility
as
a
reason
to
implement
the
program
(
28­
286,
28­
338,
28­
7016).
In
particular,
commenter
28­
7016
said
EPA
should
reject
claims
by
power
plant
owners
that
they
cannot
achieve
the
best
clean
up
levels
and
provide
reliable
sources
of
energy.
Commenter
28­
256
supported
EPA
in
its
efforts
to
implement
Regional
Haze
regulations,
particularly
through
cost­
effective,
market­
based
approaches,
such
as
cap­
and­
trade
mechanisms.

One
commenter
(
28­
127)
supported
BART,
but
said
it
should
be
stricter.
The
time
frame
for
pollution
reductions
as
outlined
by
the
Haze
Rule
is
already
too
long
and
a
target
date
of
2038
should
be
set
with
no
exemptions.
A
3­
percent
annual
reduction
in
haze­
causing
air
pollution
is
reasonable,
practicable,
and
feasible.
Technologies
now
exist
that
could
achieve
this
rate
of
reduction.
Others
commenter
said
EPA
should
finalize
the
BART
provisions
in
their
current
form
as
soon
as
possible
(
28­
107,
28­
254,
28­
313,
28­
526,
28­
567,
28­
572,
28­
573,
28­
577,
28­
583,
28­
593,
28­
596).
Commenter
28­
254
added
that
the
process
of
a
strong
technology
review
followed
by
visibility
analysis
is
consistent
with
the
methodologies
used
by
the
WRAP.
Commenter
28­
290
supported
EPA's
proposed
guidelines.

One
commenter
said
EPA
should
take
steps
to
reduce
emissions
by
50
percent
(
28­
1).

Response:

We
agree
that
BART
is
needed
as
one
strong
component
of
a
strong
regional
haze
program.
The
target
date
for
a
return
to
natural
visibility
conditions
is
outside
the
scope
of
this
rulemaking.
In
addition,
the
CAA
sets
out
BART
as
a
technology
requirement
for
States
to
impose
on
individual
sources,
and
the
guidelines
assist
States
in
making
determinations
in
that
manner.

1.2
Opposition
to
the
Proposed
Guidelines
8
1.2.1
Multi­
pollutant
Legislation
is
Preferable
Comment:
Twelve
commenters
said
that
regional
haze
would
be
better
addressed
using
a
multipollutant
legislative
approach,
such
as
the
Clear
Skies
Act
of
2003
(
0185,
0188,
0206,
0208,
0217,
0244,
0246,
0252,
0254,
0279,
0303,
0319,
28­
335,
28­
404).
Commenters
said
that
power
plants
are
subject
to
numerous
rules,
which
is
duplicative
and
inefficient
and
creates
regulatory
uncertainty
and
disjointed
programs.
A
coordinated
strategy
would
provide
maximum
environmental
benefits
at
the
lowest
cost
and
eliminate
the
need
for
multiple
emission
control
programs
(
0206,
0208,
0303).
One
commenter
(
0188)
added
that
the
legislative
approach
would
provide
more
certainty
than
a
regulatory
approach,
which
is
susceptible
to
being
judicially
overturned
or
significantly
modified
by
future
Administrations.
Some
commenters
(
0244,
0252)
fear
that
the
Regional
Haze/
BART
rule
does
not
preclude
the
possibility
of
further
emission
reductions,
even
in
states
cover
by
the
Clean
Air
Interstate
Rule
(
CAIR),
because
of
visibility
requirements
related
to
"
reasonably
attributable"
BART
and
"
reasonable
progress."

Three
commenters
said
the
BART
guidelines
must
be
consistent
and
coordinated
with
the
multi­
pollutant
legislation
(
28­
255,
28­
272,
28­
342).
Commenter
28­
251
said
if
a
multi­
pollutant
bill
is
passed,
the
regional
haze
program
may
not
be
needed.

Response:

We
also
support
multi­
pollutant
legislation
and
recommend
that
States
adopt
coordinated
strategies
for
addressing
regional
haze,
PM
2.5,
and
other
obligations.

1.2.2
Guidelines
Restrict
State
Discretion
Comment:
Fifty­
two
commenters
said
that
the
proposed
BART
guidelines
restrict
state
discretion,
which
is
inconsistent
and/
or
illegal
under
the
Act
(
0189,
0197,
0207,
0208,
0209,
0214,
0217,
0238,
0241,
0242,
0245,
0246,
0247,
0270,
0321,
28­
243,
28­
329,
28­
331,
28­
342,
28­
243,
28­
244,
28­
354,
28­
270,
28­
304,
28­
336,
28­
337,
28­
339,
28­
341,
28­
342,
28­
244,
28­
345,
28­
347,
28­
349,
28­
351,
28­
353,
28­
354,
28­
355,
28­
356,
28­
357,
28­
359,
28­
360,
28­
361,
28­
364,
28­
383,
28­
385,
28­
386,
28­
395,
28­
397,
28­
403,
28­
747,
28­
788,
28­
7001).
Commenters
argued
that
visibility
plans,
not
BART
requirements,
are
the
regulatory
mechanism
by
which
progress
toward
visibility
improvement
will
be
best
assured
because
sources
other
than
BART
sources
are
the
more
important
contributors
to
visibility
impairment
(
0189,
0197).
One
commenter
(
0207)
said
that
by
focusing
primarily
on
point
sources,
EPA
misses
an
opportunity
to
more
appropriately
and
cost
effectively
address
visibility
impairment.
One
commenter
(
0241)
added
that
the
longterm
control
strategy
provisions,
coupled
with
the
BART
evaluation
factors
identified
in
the
Act,
provide
sufficient
guidance
for
states
to
develop
SIPs
that
comply
with
the
requirements
of
the
regional
haze
rule.
If
the
BART
levels
established
in
the
SIP
do
not
go
far
enough
to
demonstrate
reasonable
progress,
states
must
include
additional
requirements
in
their
SIPs.
This
grants
the
9
states
sufficient
discretion
to
determine
how
and
where
best
to
institute
controls
to
most
effectively
meet
the
reasonable
progress
requirements.
Commenter
0321
added
that
the
specificity
and
detail
in
section
169A(
b)(
2)(
A)
regarding
the
requirement
for
a
SIP
to
"
include"
a
BART
program
clearly
demonstrate
that
this
language
was
intended
to
set
forth
the
exclusive
program
for
controlling
the
BART­
eligible
sources
under
section
169A.

Commenters
(
28­
342,
28­
349,
28­
354,
28­
267)
said
the
guidelines
are
so
prescriptive
they
amount
to
micro­
management
of
the
states
and
that
a
one­
size­
fits­
all
strategy
is
inappropriate.
Other
commenters
said
the
while
the
detail
in
the
proposal
has
been
reduced
compared
to
2001,
it
is
still
too
burdensome
and
prescriptive
(
0189).
The
new
rules
still
expand
EPA's
limited
role
in
BART
at
the
expense
of
state
discretion,
in
contraction
of
the
Court's
decision
and
the
meaning
of
the
Act
(
0209,
0217,
0241,
0321).
Commenter
(
0247,
28­
267)
added
that
EPA
is
suggesting
a
"
one
size
fits
all"
approach
to
BART,
which
does
not
recognize
regional
differences.
The
Act
does
not
authorize
EPA
to
substitute
its
own
judgment
for
that
of
individual
states.
Commenter
28­
349
added
that
EPA's
regional
haze
program,
including
the
proposed
guidelines,
has
unduly
stressed
the
regulation
of
stationary
sources
that
are
potentially
subject
to
BART
requirements
rather
than
encouraging
states
to
pursue
a
broader
approach.

Furthermore,
EPA
has
no
legal
basis
for
imposing
BART
program
requirements
(
0189,
0197,
0207,
0241,
0270,
28­
244,
28­
304,
28­
305).
One
commenter
(
0189)
said
that
section
169A
does
not
grant
EPA
any
greater
authority
to
set
procedures
for
state
BART
decisions
than
it
does
to
prescribe
their
substance.
The
commenter
objected
to
EPA's
proposal
that
states
be
required
to
analyze
technical
feasibility,
cost,
and
nonair
quality
environmental
impact
in
a
manner
specified
by
EPA
and
subject
to
detailed
retrospective
EPA
review.
Section
169A
clearly
requires
a
state
evaluation
to
determine
BART
and
not
an
EPA
evaluation.
The
compulsory
nature
of
the
guidelines,
their
prescriptive
nature,
presumptive
levels
of
controls,
and
emissions
trading
criteria
are
all
clearly
beyond
the
Agency's
authority
under
the
Act
(
0207,
0209,
0217,
0238,
0241,
0246,
0247,
0270).
EPA
should
clarify
that
the
guidelines
are
advisory,
that
it
may
be
appropriate
for
a
state
or
tribe
to
pursue
an
alternative
path
in
making
BART
determinations,
and
eliminate
all
"
presumptive"
requirements
that
must
be
rebutted
by
states
in
order
to
implement
their
discretion
(
0217).
Commenter
28­
360
said
the
proposed
BART
guidelines
go
beyond
the
requirements
of
Section
169A
and
they
should
be
withdrawn,
revised,
and
presented
again
for
public
comment.

One
commenter
(
0270)
said
that
Congress'
use
of
the
term
"
guidelines"
is
significant
and
cannot
be
ignored.
The
commenter
cited
supporting
court
cases,
including
Appalachian
Power,
supporting
the
commenter's
contention
that
because
the
guidelines
are
couched
in
terms
of
command,
they
leave
the
states
very
little
discretion
as
to
how
to
implement
the
BART
provisions
of
the
Act.
According
to
the
commenter,
this
is
impermissible
under
the
Act.
Another
commenter
(
0321)
added
that
the
EOA
and
Federal
Land
Manager
(
FLM)
"
guidance
documents"
have
not
be
subject
to
scientific
peer
review
or
public
notice
and
comment
rulemaking
processes
under
the
Federal
Administrative
Procedures
Act.

One
state
commenter
(
0245)
as
concerned
that
EPA
exceeded
its
authority
when
it
10
requires
states
to
follow
its
BART
guidance
exclusively.
These
proposed
changes
may
create
more
litigation
concerning
these
rules,
creating
uncertainty
for
the
states
in
development
of
their
SIPs.

One
commenter
(
0197)
said
that
EPA's
reliance
on
sections
308(
e)(
1)
and
(
2)
of
the
Act
and
certain
associated
preamble
language
all
were
struck
down
by
the
Court's
decision.

Commenters
(
0208,
0209,
28­
329,
28­
331,
28­
360)
said
EPA
should
withdraw
the
guidelines,
revise
them
to
address
scientific
and
technological
limitations,
and
ensure
that
the
guidelines
are
consistent
with
its
authority
under
the
Act.

Response:

We
believe
that
the
BART
guidelines
are
helpful
to
States
while
providing
them
with
considerable
flexibility,
as
is
discussed
in
the
preamble,
and
as
States
have
indicated.
(
See
response
to
comment
1.1.6
to
address
the
issue
of
EPA's
authority
to
require
States
to
follow
the
guidelines.)

1.2.3
Guidelines
Should
Only
Be
Applicable
to
Large
Utilities
Comment:
Fifty
commenters
said
that
the
guidelines
should
only
be
applicable
to
large
utilities,
consistent
with
EPA's
authority
(
0182,
0184,
0186,
0189,
0201,
0226,
0237,
0242,
0243,
0270,
0275,
0294,
28­
243,
28­
251,
28­
259,
28­
267,
28­
273,
28­
275,
28­
304,
28­
306,
28­
325,
28­
333,
28­
337,
28­
339,
28­
340,
28­
342,
28­
343,
28­
344,
28­
345,
28­
349,
28­
351,
28­
353,
28­
354,
28­
357,
28­
363,
28­
364,
28­
383,
28­
390,
28­
391,
28­
395,
28­
405,
28­
406,
28­
747,
28­
748,
28­
770,
28­
778,
28­
7001,
28­
7003,
28­
7004,
28­
7008)
Many
commenters
said
that
the
clear
reading
of
section
169A
does
not
contain
a
requirement
for
states
to
utilize
the
BART
guidelines
when
evaluating
BART
controls
for
any
other
BART­
eligible
sources
except
for
BART­
eligible
EGUs
greater
than
750
MW.
They
also
cited
Conference
Reports,
Court
cases,
and
other
references.
Instead,
Congress
purposefully
and
intentionally
drew
a
distinction
between
EGUs
greater
than
MW
and
all
other
BART­
eligible
sources
for
purposes
of
the
applicability
of
the
BART
Guidelines.
As
such,
EPA's
proposal
to
mandate
that
states
apply
the
BART
Guidelines
to
all
BART­
eligible
sources
is
contrary
to
the
Act.
Another
commenter
(
0184)
added
that
if
EPA
were
given
broader
authority
over
all
sources,
there
would
be
no
need
to
make
a
specific
point
in
section
169A(
b)
to
large
utilities.

EPA
should
withdraw
the
guidelines
for
setting
BART
for
all
units
other
than
EGUs
greater
than
750
MW
and
allow
states
to
decide
which
sources
impair
visibility
(
0186,
0189,
0275,
0294).
This
approach
is
consistent
with
the
Act
and
with
the
Corn
Growers
remand.

Regarding
the
presumption
that
250
through
749
megawatt
EGU
control
levels
must
equal
the
control
levels
for
750
megawatt
and
greater
EGUs,
one
commenter
(
0201)
was
11
concerned
that
EPA
is
overstepping
its
authority
and
potential
litigation
could
again
impact
finalization
of
the
BART
rule,
thereby
significantly
delaying
the
guidance
needed
by
the
states
and
tribes
in
development
of
their
initial
SIPs
and
TIPs.

Commenter
28­
243
said
at
a
large
utility
with
at
least
one
BART­
affected
unit,
then
the
remaining
units
at
the
plant
would
be
exempted
from
the
required
application
of
the
guidelines
during
the
first
5­
year
SIP.

Two
commenters
(
28­
332,
28­
7008)
said
EPA
had
not
provided
enough
detail
on
EPA's
authority
to
require
the
use
of
BART
guidelines
for
all
26
major
source
categories.

Commenter
28­
273
said
the
Act
also
limits
BART
guidelines
to
requiring
controls,
rather
than
emission
limits,
except
for
large
utility
boilers.

Response:

We
believe
that
the
BART
guidelines
are
helpful
to
States
while
providing
them
with
considerable
flexibility,
as
is
discussed
in
the
preamble,
and
as
States
have
indicated.
(
See
response
to
comment
1.1.6
to
address
the
issue
of
EPA's
authority
to
require
States
to
follow
the
guidelines.)

1.2.4
Guidelines
Not
Needed
or
Premature
Comment:
Six
commenters
said
the
guidelines
are
not
needed
or
are
premature
because
EPA
has
not
accounted
for
the
effects
of
other
regulatory
programs
on
visibility
reductions
(
28­
251,
28­
333,
28­
353,
28­
354,
28­
355,
28­
748,
28­
275,
28­
333).
Commenters
said
EPA
should
account
for
the
effect
of
emission
reductions
of
other
provisions
of
the
Act
(
Title
IV,
NO
x
SIP
call,
NSR,
NSPS,
mobile
source
controls,
etc.)
in
order
to
demonstrate
a
need
for
the
regional
haze
rules
it
has
proposed.
Commenter
28­
355
said
it
is
poor
policy
and
a
misallocation
of
societal
resources
for
EPA,
through
the
BART
guidance,
to
regulate
these
pollutants
more
stringently
and
with
less
flexibility
than
under
programs
that
are
directed
toward
these
more
important
values.
Earlier
EPA
findings
stated
that
existing
programs
would
allow
the
majority
of
Class
I
areas
to
meet
visibility
goals
without
further
controls
and
the
extreme
stringency
and
broad
coverage
of
the
proposed
BART
guidance.

Commenter
28­
244
said
the
resources
that
EPA's
proposal
would
require
states
to
devote
to
this
enterprise
would
become
unavailable
for
far
more
beneficial
and
challenging
tasks
such
as
more
precise
analysis
of
the
causes
and
cures
of
visibility
impairment,
or
designing
practical
approaches
to
the
control
of
mobile
source
and
area
source
emissions.

Commenter
28­
264
said
the
relevance
BART
had
in
1977
to
address
sources
built
in
the
previous
15
years
was
past
and
this
tool
is
no
longer
the
best
choice
for
addressing
visibility
12
impairment.

Nine
commenters
said
the
guidelines
are
premature
until
EPA
has
completed
modeling
and
analysis
to
determine
if
control
strategies
will
work
and
the
needed
guidance
to
conduct
such
analyses
(
28­
270,
,
28­
331,
28­
329,
28­
342,
28­
353,
28­
354,
28­
357,
28­
360,
28­
385,
28­
386,
28­
392,
28­
401,
28­
402,
28­
7001,
28­
7009,
28­
7003,
28­
7004).
The
needed
guidelines
include
those
for
quantifying
natural
visibility
conditions,
visibility
monitoring,
and
air
quality
modeling.
EPA
should
also
refine
the
IMPROVE
protocol
to
ensure
it
applies
to
local
conditions.

Response:

We
believe
the
BART
guidelines
fulfill
the
CAA
directive
to
specifically
address
older
stationary
sources
that
cause
or
contribute
to
visibility
impairment,
and
that
Congress
recognized
and
indicated
that
a
successful
regional
haze
strategy
needs
to
include
a
strategy
to
address
large
older
stationary
sources.
We
also
believe
that
we
have
provided
the
other
guidance
requested
in
this
comment.

1.2.5
BART
Should
Take
Other
Programs
into
Account
Comment:
Two
commenters
said
EPA
should
allow
the
states
to
conserve
resources
to
address
the
health­
based
NAAQS
first
and
then
determine
how
to
integrate
regional
haze
with
the
NAAQS
(
28­
251,
28­
272).
Other
commenters
(
28­
244,
28­
266,
28­
340,
28­
748)
said
EPA
should
integrate
the
BART
requirements
(
such
as
emission
limits
and
compliance
averaging
times)
and/
or
allow
the
states
to
do
so
with
other
control
programs
such
as
PM2.5
SIPs,
the
acid
rain
program,
the
NO
x
Trading
Program,
SO
2
and
PM10
SIPs
and
8­
hour
standard
ozone
SIPs.
In
addition,
EPA
should
set
forth
an
integrated
schedule
for
the
BART
requirements,
PM2.5
requirements
and
remaining
regional
haze
requirements
(
28­
266).
Commenter
28­
244
said
states
should
be
free
to
conclude,
without
further
analysis,
that
no
additional
controls
on
pollutants
addressed
by
a
MACT
standard
or
under
the
NO
x
SIP
call
should
be
required
by
the
BART
program.

One
state
commenter
(
0299)
believes
that
the
proposed
BART
Guidelines
will
impose
a
tremendous
resource
burden
on
states
because
of
the
number
of
sources
involved
and
the
resource
intensive
determinations
of
BART
for
eligible
sources.
The
commenter
believes
that
an
easier
and
more
practical
approach
is
possible
(
e.
g.,
multi­
pollutant,
regional
or
source­
category
engineering
analyses).
EPA
must
recognize
the
simultaneous
burden
being
placed
on
states
to
develop
control
programs
to
comply
with
the
Regional
Haze
program
and
the
PM2.5
and
8­
hour
ozone
National
Ambient
Air
Quality
Standards
(
NAAQS),
and
other
multi­
pollutant
mandates.
Efforts
needed
to
comply
with
the
health­
based
standards
must
take
precedence,
and
State
resources,
which
are
limited,
should
be
conserved
for
that
purpose.
In
addition,
EPA
should
provide
the
states
with
additional
resources
(
e.
g.,
funding)
to
implement
the
BART
requirements.

Response:
13
We
recognize
the
inter­
related
nature
of
CAA
programs,
and
are
working
to
coordinate
with
other
air
progams
while
still
ensuring
that
the
specific
requirements
of
the
CAA
are
met.

1.2.6
Other
Comments
in
Opposition
One
commenter
(
0240)
was
is
concerned
that
the
proposed
rule
is
so
stringent
that
virtually
all
BART­
eligible
sources
will
be
required
to
install
BART
controls,
even
those
for
which
a
showing
of
meaningful
contribution
to
visibility
impacts
in
a
Class
I
area
has
not
been
demonstrated.
Such
an
outcome
conflicts
with
the
court
ruling
in
American
Corn
Growers.
Commenter
28­
251
said
all
plants
must
be
properly
classified
under
the
proposed
BART
rule
so
that
the
final
rule
does
not
unnecessarily
and
improperly
involve
facilities
in
an
expensive
emissions­
control
program.
Commenter
28­
345
added
that
EPA
should
carefully
consider
the
history
of
the
development
of
the
26
category
list
as
a
principle
for
defining
the
scope
of
each
source
on
the
list.

Commenters
(
28­
342,
28­
357,
28­
775)
said
EPA
must
withdraw
the
BART
guidelines
and
revise
them
to
be
consistent
with
the
Act.
Commenter
28­
342
added
that
many
of
the
legal
flaws
and
the
procedural
deficiencies
in
the
regional
haze
rule
are
reflected
in
the
Proposed
BART
Guidelines.
EPA
and
interested
parties,
including
UARG,
must
convene
a
collaborative
approach
to
develop
a
workable,
legally
sufficient
visibility
protection
program.

Commenters
(
28­
340,
28­
341,
28­
344,
28­
364)
said
the
guidelines
would
directly
and
seriously
harm
their
businesses.
Commenter
28­
355
said
EPA
failed
to
articulate
a
vision
of
the
public
good
that
would
be
served
by
the
guidance.

Commenter
28­
355
said
the
scope
and
stringency
of
the
BART
guidelines
are
inconsistent
with
regulatory
impact
analysis
(
RIA)
prepared
for
the
final
Regional
Haze
rule,
which
showed
that
in
large
parts
of
the
U.
S.
no
additional
control
will
be
required
to
meet
visibility
progress
goals.

Commenter
28­
747
said
while
the
regional
haze
rule
is
being
challenged,
the
purpose
of
any
BART
guideline
under
the
rule
should
be
designed
to
further
its
intent
and
the
proposed
BART
guidance
fails
to
meet
this
test
in
three
critical
areas:
(
1)
State
flexibility
in
setting
reasonable
progress
goals;
(
2)
State
flexibility
in
the
selection
of
mechanisms
for
achieving
reasonable
progress
goals;
and
(
3)
the
60­
year
timeframe
for
restoring
"
natural"
visibility
conditions.
By
prescribing
a
one­
size­
fits­
all
regulatory
program
that
is
disconnected
from
visibility
goals,
the
proposed
BART
guidance
would
not
only
work
a
wholesale
rewrite
of
the
regional
haze
rule,
but
would
drastically
increase
the
compliance
costs
of
EPA's
visibility
program.

Response:

We
believe
that
the
BART
guidelines
fulfill
the
CAA
requirements
in
section
169A,
while
14
providing
considerable
flexibility
to
States.
We
especially
note
that
the
guidelines
are
not
mandatory
for
most
source
categories.

1.3
The
Role
of
the
Proposed
Guidelines
1.3.1
States'
Role
Comment:
Fourteen
commenters
stressed
that
the
BART
guidelines
should
provide
tools
that
assist
states
in
conducting
BART
assessments
rather
than
being
overly
prescriptive
requirements
(
28­
243,
28­
264,
28­
266,
28­
270,
28­
305,
28­
342,
28­
251,
28­
357,
28­
391,
28­
392,
28­
401,
28­
402,
28­
748,
28­
7009).
Many
of
these
commenters
added
that
the
guidelines
are
contrary
to
the
Act
and
preclude
state
flexibility
(
see
section
1.2.2
for
more
discussion
of
this
topic.)
They
said
states
should
have
the
leadership
role
in
making
decisions
on
costs
and
benefits
and
the
guidelines
should
be
structured
to
provide
the
direction
they
need
in
the
process.

Some
commenters
stressed
that
states
should
have
the
authority
to
decide
how
to
apply
BART
to
sources
other
than
power
plants
(
28­
251,
28­
259,
28­
304,
28­
343,
28­
355,
28­
388,
28­
392,28­
403).
This
is
consistent
with
Act
and
Congressional
intent.

One
commenter
(
28­
241)
stressed
the
need
to
balance
improving
air
quality
with
the
costs
and
hardships
that
may
be
required.
The
commenter
saw
an
opportunity
for
a
collective
effort
to
develop
a
workable
solution
and
attain
air
quality
goals.
Commenter
28­
391
added
that
the
burden
on
states
and
industry
to
conduct
extensive
technology
evaluations
on
"
all"
pollutants
from
"
all"
sources
emitting
visibility­
impairing
pollutants
would
be
unreasonable
and
likewise
yield
no
practical
value.
States
are
already
burdened
with
implementation
plan
development
and
revisions
required
by
the
Regional
Haze
and
other
recent
Federal
requirements.

Response:

We
believe
that
the
final
BART
guidelines
are
helpful
to
States
while
providing
them
with
considerable
flexibility,
especially
noting
two
differences
between
the
proposed
and
final
guidelines:
(
1)
we
have
clarified
in
the
final
guidelines
that
they
are
not
mandatory
for
most
source
categories,
and
(
2)
that
the
final
guidelines
provide
more
discretion
to
States
than
was
indicated
in
the
proposed
guidelines
regarding
the
many
sources
of
VOC
emissions
in
many
States.
States
do
not
have
to
do
extensive
technology
evaluations
for
emissions
of
a
pollutant
if
they
do
not
find
those
emissions
to
be
"
reasonably
anticipated
to
cause
or
contribute"
to
visibility
impairment.

Comment:
Commenters
(
28­
333,
28­
275)
said
the
burden
of
proof
on
states
is
a
very
burdensome
and
precious
time
and
resource­
consuming
task,
and
fails
to
hold
EPA's
presumptive
decision
to
15
the
same
burden.
The
end
result
for
resource­
starved
State
agencies
will
likely
be
a
decision
to
over­
control
simply
because
the
burden
was
too
heavy
to
document
otherwise.
Guidance
should
be
what
its
name
implies
and
guide
not
direct.

Response:

We
have
taken
specific
steps
in
these
final
guidelines
to
reduce
the
burden
on
States.
For
example,
as
indicated
in
the
response
preceding
this
response,
the
final
guidelines
provide
considerably
more
discretion
to
States
regarding
the
many
sources
of
VOC
emissions
in
many
States.

Comment:
Two
commenters
said
EPA
must
specify
that
state
agencies,
not
sources
have
the
ultimate
responsibility
to
conduct
BART
eligibility
determinations,
engineering
analyses,
visibility
improvement
assessments
and
implementation
regulations
(
28­
348,
28­
387).
EPA'
s
statement
that
state
agencies
"
may
prefer
to
require
source
owners
to
assume
part
of
the
analytical
burden"
suggests
that
sources
are
at
least
partly
responsible
for
developing
and
implementing
SIPs.

However,
commenter
28­
255
said
EPA
should
include
in
the
rules
a
clear
statement
that
allows
states
to
adopt
requirements
placing
the
responsibility
on
source
owners
to
self
identify
if
they
meet
the
criteria
for
BART­
eligible
sources
and
to
require
source
owners
to
conduct
the
BART
analyses.

Response:

The
guidelines
recognize
the
responsibility
that
the
CAA
places
on
States
to
prepare
regional
haze
SIPs
and
to
conduct
BART
reviews,
and
the
authority
that
States
have
to
regulate
pollutant
emissions
from
sources,
and
to
require
specific
information
and
analyses
from
sources.

1.3.2
Focus
of
Regional
Haze
Solutions
Comment:
Two
commenters
said
the
guidelines
must
be
consistent
with
Congressional
intent
that
requires
EPA
to
arrest
and
reverse
the
trends
toward
impaired
visibility
in
Class
I
areas
caused
by
regional
haze
(
28­
358,
28­
397).
Commenter
28­
260
added
that
reductions
in
pollutants
is
not
enough;
improvements
in
visibility
must
be
quantifiable.
Commenter
28­
312
noted
that
existing
stationary
sources
affected
by
this
proposal
are
not
the
only
contributors
to
regional
haze.
However,
controlling
BART
sources
should
be
an
essential
element
of
a
national
regional
haze
program.

Two
commenters
stressed
the
need
to
look
at
air
quality
benefits
designing
control
strategies
(
28­
273,
28­
337).
Commenter
28­
273
added
that
EPA
should
find
regional
haze
solutions
that
are
aimed
at
addressing
the
most
significant
contributing
sources,
rather
than
16
targeting
only
major
stationary
sources.
As
so
often
occurs,
the
sources
targeted
for
emission
reductions
are
not
those
that
create
most
of
the
problem
(
i.
e.,
unpaved
roads),
but
those
that
are
easiest
to
regulate.
Stationary
sources
have
been
regulated
to
the
point
of
diminishing
returns.

Commenter
28­
356
said
instead
of
addressing
visibility,
the
agency
is
focused
on
limiting
emissions,
regardless
of
the
realized
environmental
benefit
which
may
(
or
may
not)
result.
The
BART
Guidance
ignores
specific
CAA
requirements
and
is
out
of
focus
with
the
original
intent
of
Congress.

Response:

We
believe
that
the
additional
visibility
modeling
provided
for
in
the
final
guidelines
will
help
focus
State
requirements
on
sources
that
contribute
to
visibility
impairment.
As
indicated
and
documented
in
the
preamble,
guidelines,
and
the
1999
Regional
Haze
rule,
while
regional
haze
is
caused
by
many
different
types
of
sources,
stationary
source
emissions
play
a
large
role.

1.3.3
60­
year
Timeframe
for
Restoring
Natural
Conditions
Comment:
Four
commenters
said
the
proposed
BART
guidance
would
undermine
critical
elements
of
the
regional
haze
rule,
especially
the
60­
year
timeframe
for
restoring
natural
visibility
conditions
by
imposing
the
most
stringent
technological
controls
in
a
12­
year
period
(
28­
304,
28­
339,
28­
344,
28­
383).
Commenter
28­
344
added
that
the
Regional
Haze
Rule
was
intended
to
distribute
the
costs
of
this
program
over
a
60
year
period.
EPA
has
failed
to
include
any
evidence
in
the
record
to
support
the
contention
in
its
proposal
that
all
big
sources
contribute
to
visibility
impairment
in
Class
I
areas.

Response:

We
believe
that
the
supporting
data
and
analyses
provided
for
the
final
guidelines
provides
considerable
evidence
for
the
contribution
of
large
sources
to
visibility
impairment.
In
addition,
the
CAA
requires
the
installation
of
BART
as
expeditiously
as
practicable.

1.3.4
Responsibility
for
Federal
Emissions
Comment:
Three
commenters
said
the
proposed
BART
rule
and
regional
haze
rule
carve
out
any
responsibility
for
substantial
federal
visibility­
impairing
emissions
(
28­
337,
28­
364,
28­
395).
Despite
this
carve­
out,
these
emissions
are
then
utilized
generally
to
develop
a
proposed
rule
that
attributes
regional
visibility
impacts
(
and
the
installation
of
costly
additional
subsequent
controls)
conversely
on
certain
stationary
sources.
17
Response:

We
believe
the
final
guidelines
carry
out
the
specific
mandate
of
the
CAA
to
address
older
stationary
sources
of
visibility
impairment.

1.3.5
Relationship
to
the
Regional
Haze
Rule
Comment:
Three
commenters
discussed
the
relationship
of
BART
requirements
to
the
regional
haze
rule
(
28­
337,
28­
347,
28­
364).
Commenters
(
28­
337,
28­
364)
said
the
proposed
BART
rule
would
sharpen
the
implementation
aspects
of
the
Agency's
existing
Regional
Haze
Rule.
Under
both
rules,
States
are
given
two
options
regarding
"
regional
haze
BART"
for
stationary
sources.
States
can
implement
EPA's
new
"
command
and
control"
BART
requirements
or
they
can
establish
an
"
alternative"
emissions
trading
program
which
must
be
more
stringent
than,
yet
based
on,
the
new
command
and
control
BART
requirements.

Commenter
28­
347
said
the
regional
haze
rule
commands
the
states
to
submit
SIPs
on
long­
term
control
strategies
to
demonstrate
reasonable
progress
toward
the
national
visibility
goal.
If
the
BART
levels
established
in
the
SIP
do
not
go
far
enough
to
demonstrate
reasonable
progress,
states
must
include
additional
requirements
in
their
SIPs.
This
grants
the
states
sufficient
discretion
to
determine
how
and
where
best
to
institute
controls
to
most
effectively
meet
the
reasonable
progress
requirements.
Further
delineation
of
the
BART
process
is
unnecessary
and
abrogates
the
states'
authority.

Response:

We
believe
the
BART
program
will
function,
together
with
other
provisions
of
the
regional
haze
rule,
to
provide
an
integrated
approach
to
addressing
visibility
impairment.
We
believe
these
guidelines
are
helpful
to
States
and
fulfill
the
CAA
directive
to
address
older
large
stationary
sources.

1.3.6
Nationwide
Scope
of
the
Guidelines
Comment:
Several
commenters
said
the
guidelines
should
apply
to
all
50
states
to
ensure
consistent
implementation,
to
balance
the
workload
across
the
state
and
sources,
and/
or
to
recognize
that
regional
haze
is
a
national
problem
(
28­
255,
28­
256,
28­
257,
28­
260,
28­
289,
28­
302,
28­
350,
28­
352,
28­
535,
28­
7002).
At
a
minimum
all
areas
with
potential
Class
I
designation
should
be
included
when
assessing
regional
haze
impacts
(
28­
260).
Commenter
28­
290
stressed
the
need
for
national
consistency
in
the
implementation
of
BART.
A
consistent
set
of
protocols
and
welldefined
"
starting
points"
must
be
established,
upon
which
BART
analyses
for
all
source
categories
should
be
based.
18
In
contrast,
commenters
(
28­
272)
said
EPA
should
further
support
its
assumption
that
all
states
and
all
BART
eligible
sources
are
causing
visibility
impairment.

Commenter
28­
303
added
that
EPA
should
provide
guidance
so
that
states
are
not
held
accountable
for
reducing
emissions
that
affect
visibility
degradation
in
Federal
Class
I
areas
through
BART
controls
when
such
emissions
are
transported
across
the
U.
S.
border.
Sources
include
Mexican
anthropogenic
or
biogenic
and
natural
sources
such
as
from
volcanoes,
agricultural
burning
as
far
away
as
Central
America,
and
dust
from
as
far
away
as
Africa.

Response:

We
believe
the
final
BART
guidelines
provide
a
reasonable
approach
to
addressing
visibility
impairment
at
all
Class
I
areas
affected
by
the
stationary
sources
covered
by
BART
under
the
CAA.
The
regional
haze
rule,
not
the
BART
guidelines,
requires
States
to
assess
the
causes
of
regional
haze
and
devise
appropriate
strategies.

2.0
Identifying
the
BART­
eligible
Sources
2.1
Identify
the
Emission
Units
in
the
Bart
Categories
2.1.1
Source
Category
Definitions
2.1.1.1
Proposal
to
Assess
Fossil
Fuel
Boilers
of
More
Than
250
Million
Btu/
hr
Heat
Input
Using
Individual
Unit
Rather
Than
Aggregate
Capacities
Comment:
Seventeen
commenters
supported
EPA
proposal
to
assess
fossil
fuel
boilers
of
more
than
250
million
Btu/
hr
heat
input"
using
individual
unit
rather
than
aggregate
capacities
(
0169,
0170,
0178,
0186,
0187,
0189,
0236,
0241,
0242,
0247,
0256,
0266,
0275,
0279,
0285,
0307,
28­
244,
28­
255,
28­
304,
28­
333,
28­
335,
28­
339,
28­
342,
28­
343,
28­
344,
28­
347,
28­
349,
28­
383,
28­
383,
28­
388,
28­
391,
28­
393,
28­
747,
28­
748,
28­
778,
28­
7003,
28­
8000).
Commenters
said
this
interpretation
is
consistent
with
the
Act,
PSD
is
not
a
relevant
model
and/
or
EPA
has
not
justified
the
proposal
(
0178,
0187,
0236,
0241,
0247,
0266,
0275,
28­
244,
28­
304,
28­
333,
28­
335,
28­
339,
28­
342,
28­
343,
28­
344,
28­
347,
28­
349,
28­
383,
28­
388,
28­
391,
28­
393,
28­
747,
28­
748,
28­
778,
28­
7003,
28­
8000).
In
particular,
the
PSD
program's
"
aggregation
rule"
was
intended
to
prevent
circumvention
of
the
program
by
constructing
several
smaller
boilers
as
opposed
to
one
large
one;
however,
this
logic
does
not
apply
to
the
BART
program,
since
BART
only
applies
to
existing
sources.
Furthermore,
the
"
aggregation"
rule
is
hard
to
reconcile
with
the
language
of
the
Act
and
difficult
to
reconcile
with
the
policy
of
BART.
PSD
is
an
air
quality
management
program
aimed
at
protecting
the
increments,
while
BART
is
focused
on
identifying
sources
that
can
be
cost­
effectively
retrofitted
with
control
technology.
There
might
be
a
case
for
aggregating
units
to
make
sure
their
aggregated
emissions
do
not
damage
the
increments,
but
there
is
no
similar
argument
for
using
aggregation
to
target
units
for
retrofit
that
are
smaller
than
19
the
units
that
Congress
specifically
identified
(
28­
8000).

Commenter
28­
349
added
that
Congress
directed
that
the
relevant
type
of
source
must
be
a
"
fossil
fuel
boiler
of
more
than
250
MMBtu/
hr
heat
input."
It
did
not
say
that
the
source
could
be
a
plant
containing
boilers
whose
cumulative
heat
input
happens
to
be
more
than
250
MMBtu/
hr.
In
addition,
EPA
itself
said
in
a
1987
applicability
review
for
PSD
that
the
category
of
"
fossil­
fuel
boilers
of
more
than
250
MMBtu/
hr
heat
input"
is
"
more
restrictive"
than
"
steam
electric
plants"
because
it
does
not
refer
to
a
"
plant"
but
instead
to
a
"
boiler."
Thus,
EPA
indicated
that
heat
inputs
for
such
"
boilers"
should
not
be
aggregated.

Comment:
Ten
commenters
specified
that
only
those
boilers
that
are
individually
greater
than
250
MMBtu/
hr
should
be
covered
(
0170,
28­
335,
28­
242,
28­
255,
28­
267,
28­
268,
28­
275,
28­
340,
28­
770,
28­
779,
).
These
boilers
are
already
in
existence
and
were
not
installed
to
circumvent
the
PSD
rules
(
28­
243,
28­
268,
28­
275,
28­
335).
Commenter
28­
340
added
that
to
consider
individual
small
boilers
for
BART
controls
simply
because
of
co­
location
needlessly
broadens
the
program.
In
many
instances,
while
on
the
same
plant
site,
these
boilers
are
in
scattered
locations
and
likely
exhaust
through
different
stacks,
meaning
no
single
control
device
could
be
employed.

Commenters
(
0170,
28­
779)
added
that
BART
should
not
be
required
on
those
individual
boilers
that
to
do
not
exceed
250
million
Btu/
hr
heat
input
because
these
have
a
low
potential
to
contribute
to
regional
haze.
Such
boilers
typically
burn
natural
gas
or
low
sulfur
fuel
oil.
Also,
this
interpretation
is
consistent
with
the
NSPS
and
NO
x
SIP
call
program
interpretation
of
the
term
and
avoids
inclusion
of
very
small
boilers
for
which
BART
controls
would
not
be
cost
effective
(
0187,
0236,
0285,
28­
255).
Commenter
(
28­
255)
were
concerned
that
the
proposal
might
include
institutional
boilers
for
which
BART
upgrades
may
not
be
feasible
and/
or
necessary.
Commenter
28­
390
said
the
BART
controls
are
intended
for
installation
on
an
individual
boiler,
not
a
group
of
boilers.
Commenter
28­
404
said
aggregating
all
units
together
would
result
in
over­
application
of
BART
controls
to
small
units
and
excessive
costs
to
implement
controls.
Commenters
(
0256,
0279,
0285,
28­
335)
agreed
with
EPA
that
this
reading
of
the
category
may
help
address
states'
concerns
over
the
implementation
burden
of
the
program.

Comment:
Sixteen
commenters
opposed
assessing
fossil
fuel
boilers
of
more
than
250
million
Btu/
hr
heat
input
using
individual
unit
rather
than
aggregate
capacities
(
0221,
0232,
0271,
0285,
0286,
28­
404,
28­
247,
28­
256,
28­
258,
28­
302,
28­
325,
28­
770,
28­
256,
28­
358,
28­
362,
28­
7013).
Commenters
(
0221)
said
that
states
must
aggregate
smaller
fossil
fuel­
fired
boilers
at
power
plants
that
generate
and
sell
electricity.
The
applicability
criteria
must
be
consistent
with
the
PSD
regulations
and
EPA's
proposed
departure
from
the
aggregate
methodology
under
the
PSD
program
is
contrary
to
law
(
0221,
0232,
0271,
28­
302,
28­
325,
28­
358,
28­
362).
As
under
PSD,
the
principle
of
treating
similarly­
situated
sources
similarly
so
as
to
meet
core
statutory
goals
and
requirements
should
apply
with
equal
force
under
the
2004
BART
guidelines
(
0221).
20
It
is
arbitrary
and
capricious
for
EPA,
in
determining
BART
eligibility,
not
to
aggregate
smaller
fossil­
fuel
boilers
within
their
own
category
(
e.
g.,
chemical
process
plant)
(
0221).
Under
the
regional
haze
rule,
all
recognized
emission
unites
(
i.
e.,
that
fall
within
one
of
the
26
listed
categories)
within
a
source
are
aggregated
for
purposed
of
determining
whether
the
source
meets
the
250
tpy
emission
threshold.
So
too
should
the
250
million
BTU/
hr
threshold
apply
to
the
sum
of
the
fossil­
fuel
boilers
within
a
source
as
defined
above.
This
would
ensure
that
all
sources
which
use
fossil­
fuel
boilers
to
generate
power
are
treated
similarly,
whether
the
250
million
BTU/
hr
heat
input
is
achieved
using
one
boiler
or
multiple
boilers.

Commenter
28­
7013
added
that
BART
eligibility
should
be
determined
using
the
facility's
aggregate
heat
input.
However,
this
assumes
the
adoption
of
appropriate
analysis
to
determine
if
an
individual
source
is
impacting
a
Class
I
area,
utilization
of
the
"
bottom
up"
BART
analysis
approach
and
no
presumptive
BART
emission
limitations.
If
these
three
prerequisites
are
not
adopted,
only
individual
heat
input
should
be
used.

One
commenter
(
0271)
said
the
proposed
provision
will
further
limit
the
effectiveness
of
the
BART
program
and
restrict
a
State
from
potentially
controlling
sources
that
could
adversely
affect
visibility
in
Class
I
areas.
EPA
has
not
adequately
explained
why
emissions
from
units
of
this
type
should
be
excluded
from
assessment
of
"
visibility
impairment"
in
Class
I
areas
as
the
aggregate
of
emissions
from
all
sources
at
a
facility
should
be
the
most
important
factor
in
this
assessment.
Nor
has
EPA
adequately
explained
why
they
believe
that
the
definition
of
"
fossil­
fuel
boilers"
only
pertains
to
units
at
a
"
power
plant"
(
69
FR
25191)
as
this
limitation
appears
within
the
text
of
the
Federal
Register.
Instead,
the
Act's
definition
of
"
fossil­
fuel
boilers"
includes
units
not
associated
with
electric
generation
and
thus
must
be
aggregated
to
determine
applicability.

One
commenter
(
0271)
said
the
before
EPA
uses
the
cost
effectiveness
of
controls
on
smaller
boilers
as
justification
for
not
summing
their
emissions
in
the
applicability
determination,
EPA
should
compare
the
cost
effectiveness
of
controls
on
smaller
boilers
to
the
control
costs
for
all
pollutants
and
source
types.
In
addition,
the
commenter
does
not
believe
that
this
summing
of
heat
input
will
lead
to
an
implementation
burden.

Comment:
One
commenter
(
0200)
said
EPA's
proposal
would
allow
multiple,
relatively
large
units
to
be
exempted.

Response
(
to
all
of
2.1.1
cited
above):

We
have
decided
to
retain
the
interpretation
that
this
category
should
be
read
as
including
only
those
boilers
individually
greater
than
250
million
Btu/
hour
heat
input.
We
agree
with
commenters
who
noted
that
including
smaller
boilers
would
pose
considerable
implementation
burden.
As
noted
in
the
2004
reproposal
notice,
we
do
not
believe
that
this
interpretation
is
likely
to
have
a
substantial
impact.
Because
smaller
boilers
are
generally
less
cost­
effective
to
control,
21
we
believe
that
BART
review
would
be
unlikely
to
result
in
a
significant
amount
of
control
on
these
boilers.
We
are
also
retaining
the
clarification
that
if
a
boiler
smaller
than
250
million
Btu/
hour
heat
input
is
an
integral
part
of
an
industrial
process
in
a
BART
source
category
other
than
electric
utilities,
then
it
would
be
reasonable
for
the
State
to
consider
the
boiler
part
of
the
BARTeligible
source
in
that
category.
(
By
"
integral
to
the
process",
we
mean
that
the
process
uses
any
by­
product
of
the
boiler,
or
vice­
versa.
We
have
added
this
clarification
to
the
definition
in
the
BART
guidelines.)
We
believe
that
if
a
State
is
already
considering
a
BART­
eligible
industrial
process
for
control,
and
a
boiler
is
integrated
into
that
process,
it
makes
common
sense
not
to
rule
out
control
options
for
that
process
as
a
whole.
Including
the
boiler
in
consideration
of
control
options
for
the
process
adds
minimal
additional
burden
while
leaving
maximum
discretion
to
the
State
in
determining
BART
for
the
process
as
a
whole.
For
example,
pulp
and
paper
mills
may
have
boilers
that
are
not
serving
the
energy
infrastructure
of
the
plant,
but
typically
are
serving
a
process
directly
by
using
the
waste
liquor
from
the
process.
Including
such
a
the
boiler
in
consideration
of
control
options
for
the
process
adds
minimal
additional
burden
while
leaving
maximum
discretion
to
the
State
in
determining
BART
for
the
process
as
a
whole.
We
are
also
clarifying
today
that
we
have
determined
that
this
category
should
include
all
individual
boilers
of
greater
than
250
million
Btu/
hour
heat
input
burning
any
amount
of
fossil
fuel,
as
opposed
to
only
those
boilers
that
burn
greater
than
50
percent
fossil
fuel.
We
believe
that
it
is
quite
possible
that
boilers
of
this
size
could
contribute
to
regional
haze
in
a
Class
I
area,
even
if
they
burn
less
than
50
percent
fossil
fuel.
Therefore
we
believe
that
each
fossil
fuel­
fired
boiler
merits
a
BART
review
and
should
not
be
prematurely
ruled
out.

Comment:
Three
commenters
(
0170,
0253,
0256)
noted
a
discrepancy
in
the
proposal
regarding
references
to
various
types
of
boilers.
Commenter
0170
said
all
sources
within
the
category
should
be
treated
the
same,
regardless
of
their
location.
Commenters
0253
and
0256
said
EPA
should
clarify
the
ambiguity.
In
particular,
EPA
at
one
point
indicates
that
its
decision
would
mean
that
"
only
individual
utility
boilers
greater
than
250
million
BTU/
hr"
would
be
BARTeligible
Id.
(
emphasis
added).
EPA's
use
in
this
passage
of
the
word
"
utility"
in
connection
with
the
word
"
boilers"
could
create
confusion
because
it
could
be
construed
(
or
misconstrued)
as
referring
to
utility
boilers
used
for
generating
electricity
from
steam,
i.
e.,
steam
electric
generating
units
that
constitute
a
section
169A(
g)(
7)
source
category
that
is
distinct
from
the
separate
source
category
that
section
169A(
g)(
7)
describes
by
using
the
word
"
boiler."
The
above­
quoted
statement
from
the
preamble
is
not
correct
if
the
phrase
"
utility
boilers"
is
meant
to
include
steam
electric
generating
units
because
EPA
makes
it
clear
that
all
steam
electric
generating
units
at
a
power
plant
must
be
considered
in
the
BART
eligibility
assessment
regardless
of
whether
those
units
individually
have
heat
input
above
or
below
250
MMBtu/
hour.
EPA's
preamble
statement
is
correct
only
for
those
"
utility"
boilers
(
greater
than
250
MMBtu/
hr
heat
input)
that
are
not
steam
electric
generating
units
at
power
plants.

Response
(
to
rest
of
2.1.1
cited
above):
22
We
have
corrected
this
discrepancy
in
the
final
BART
guidelines.

2.1.1.2
Boilers
Smaller
Than
250
Million
Btu/
hr
Integral
to
Source
Categories
Other
Than
Electric
Utilities
May
be
Subject
to
Control
Comment:
Three
commenters
opposed
the
idea
of
including
boilers
smaller
than
250
million
Btu/
hr
that
are
integral
to
source
categories
other
than
electric
utilities
as
potentially
subject
to
controls
(
0186,
0275,
0307).
Commenters
(
0186,
0275,
0307)
said
this
undermines
the
Congressional
intent
for
how
these
sources
are
to
be
regulated.
These
sources
were
not
intended
to
under
BART
analysis,
but
would
be
considered
with
all
other
sources
in
developing
state
visibility
plans.
One
commenter
(
0307)
said
this
approach
would
add
significantly
to
implementation
burden
and
is
arbitrary
and
capricious
that
small
boilers
or
other
support
facilities
included
in
one
of
the
26
named
BART
categories
must
be
considered,
but
a
like
facility
not
supporting
a
BART
source
category
need
not
be
considered.

Comment:
One
commenter
(
0285)
asked
EPA
to
clarify
the
situation
in
which
a
boiler
could
be
considered
an
integral
part
of
a
process.
In
an
industrial
setting,
basically
boilers
supply
hot
water
and/
or
steam
to
one
or
more
processes.
Is
that
considered
"
integral
to
the
process?"
Without
some
clarification,
it
will
be
difficult
to
determine
the
applicability.
Also,
in
the
interest
of
emission
reduction
benefits
and
resource
needs,
EPA
should
set
a
threshold
level,
above
which
a
small
boiler
that
is
an
integral
part
of
a
process
should
be
considered
for
control.

Response:

We
believe
that
if
a
boiler
smaller
than
250
million
Btu/
hour
heat
input
is
an
integral
part
of
an
industrial
process
in
a
BART
source
category
other
than
electric
utilities,
then
the
State
may
consider
the
boiler
part
of
the
BART­
eligible
source
in
that
category.
A
State
may
also
reasonably
conclude
that
an
industrial
boiler
smaller
than
250
million
Btu/
hour
heat
input
that
provides
only
heat
and
power
to
a
facility
is
not
integral
to
the
process.
We
have
added
this
clarification
to
the
definition
in
the
BART
guidelines.
If
a
State
is
already
considering
a
BARTeligible
industrial
process
for
control,
and
a
boiler
is
integrated
into
that
process,
it
makes
common
sense
not
to
prematurely
rule
out
control
options
for
any
of
the
emissions
from
that
process
as
a
whole.
Including
the
boiler
in
consideration
of
control
options
for
the
process
adds
minimal
additional
burden
while
leaving
maximum
discretion
to
the
State
in
determining
BART
for
the
process
as
a
whole.
For
example,
kraft
pulp
mills
may
have
boilers
that
are
not
serving
the
energy
infrastructure
of
the
plant,
but
typically
are
serving
a
process
directly
by
using
the
waste
liquor
from
the
process.
Including
such
a
the
boiler
in
consideration
of
control
options
for
the
process
adds
minimal
additional
burden
while
leaving
maximum
discretion
to
the
State
in
determining
BART
for
the
process
as
a
whole.

2.1.1.3
"
Fossil­
fuel
Fired
Boilers"
are
Those
that
Burn
More
Than
50%
Fossil
Fuels
23
Comment:
Twelve
commenters
agreed
with
EPA
that
a
boiler
has
to
burn
at
least
50
percent
fossil
fuels
to
be
considered
a
fossil­
fuel
boiler
(
0169,
0170,
0178,
0189,
0237,
0259,
0266,
0307,
28­
244,
28­
339,
28­
383,
28­
747).
Commenter
0237
added
that
this
determination
should
not
shift
year­
to­
year.
Commenters
(
0178,
0266)
also
said
that
Section
II
of
Appendix
Y
to
part
51
should
be
amended
to
read
as
follows:
"
Also,
you
should
consider
a
multi­
fuel
boiler
to
be
a
fossil­
fuel
boiler
if
it
burns
at
least
50
percent
fossil
fuels
on
an
annual
heat
input
basis."

Commenter
28­
244
said
EPA
has
never
defined
when
a
unit
that
burns
multiple
fuels
should
be
classified
as
"
fossil
fuel."
Interpreting
similar
language
contained
in
the
NO
x
SIP
call,
EPA
has
concluded
that
a
unit
that
burns
less
than
50%
fossil
fuel
should
not
be
classified
as
"
fossil
fuel."
EPA
should
adopt
the
same
approach
here.

Commenter
28­
779
added
that
EPA
should
also
provide
an
exemption
for
boilers
that
have
the
capacity
to
burn
fossil
fuel
of
greater
than
250
MMBtu/
hr
but
actually
have
an
annual
capacity
factor
for
fossil
fuels
of
10
percent
or
less.
The
BART
regulations
are
intended
for
"
fossil­
fuel"
boilers.
This
is
presumed
to
mean
boilers
that
utilize
fossil
fuel
as
their
"
primary"
fuel.
Many
boilers
have
fossil­
fuel
burning
capability
of
greater
than
250
MMBtu/
hr,
but
only
burn
such
fuels
during
startup
or
shutdown
or
as
a
supplemental
fuel.

Comment:
Two
commenters
(
0232,
0271)
opposed
this
definition.
Commenter
0232
said
there
is
no
justification
for
this
exclusion.
Commenter
0271
said
the
over­
riding
concern
should
be
the
ability
to
emit
pollutants
that
adversely
affect
visibility
levels.
Any
use
of
a
fossil­
fuel
in
the
boiler
should
enable
the
boiler
to
be
potentially
subject
to
BART
as
this
will
better
ensure
that
the
State
is
reviewing
and
potentially
controlling
all
large
sources
of
visibility
impairing
pollutants.
Promulgation
of
the
regulation
as
written
could
limit
the
states'
discretion
to
control
these
larger
sources.
Also,
enforceable
emission
limits
should
not
be
used
to
avoid
BART
controls
for
boilers
having
a
physical
design
capacity
to
emit
greater
than
250
tons
per
year
of
a
visibility
impairing
pollutant.
The
size
of
the
boiler
itself
should
be
the
only
determinant
as
this
will
prevent
intermittent
use
of
large
uncontrolled
units
which
may
adversely
affect
visibility
levels
in
Class
I
areas
albeit
for
shorter
intervals.

2.1.1.4
Timing
of
50
%
Threshold
Determination
Comment:
Three
commenters
raised
clarifications
concerning
the
timing
of
the
50%
threshold
determination
(
0178,
0266,
0237).
Two
commenters
(
0178,
0266)
said
that
the
threshold
should
be
applied
on
an
annual
average
basis.
A
third
commenter
(
0237)
noted
that
the
May
5
notice
appears
to
declare
that
a
non­
BART
source
can
in
2004
or
later
become
a
BART­
eligible
source
if
it
changes
its
future
fuel
source.
It
would
make
more
sense
to
determine
if
a
fossil
fuel
fired
source
met
the
EPA
definition
at
some
particular
point
in
time
 
preferably
on
August
7,
1977,
when
the
BART
provisions
were
enacted.
If
it
changed
to
an
alternative
non­
fossil­
fired
fuel
after
24
1977,
it
should
no
longer
be
considered
BART
eligible.
EPA
should
clarify
these
applicability
criteria
accordingly.

Response
(
for
2.1.1.3
and
2.1.1.4):

We
believe
that
it
would
be
reasonable
to
include
all
individual
boilers
of
greater
than
250
million
Btu/
hour
heat
input
burning
any
amount
of
fossil
fuel
within
this
category,
as
opposed
to
only
those
boilers
that
burn
greater
than
50
percent
fossil
fuel.
The
amount
of
fossil
fuel
burned
can
vary
year
to
year,
making
it
potentially
difficult
to
determine
if
a
boiler
burns
only
50
percent
fossil
fuel.
In
addition,
we
believe
that
it
is
quite
possible
that
boilers
of
this
size
could
contribute
to
regional
haze
in
a
Class
I
area,
even
if
they
burn
less
than
50
percent
fossil
fuel.
(
For
example,
boilers
that
burn
a
considerable
amount
of
wood
waste
may
have
significant
emissions
of
particulate
matter
or
precursors.)
Therefore
we
believe
that
each
fossil
fuel­
fired
boiler
merits
a
BART
review
and
should
not
be
prematurely
ruled
out.
It
is
likely
that
this
interpretation
would
affect
only
a
small
number
of
sources,
but
those
sources
may
have
significant
emissions
and
may
merit
review
under
BART.
The
guidelines
accordingly
adopt
an
interpretation
of
this
provision
that
a
source
which
is
permitted
to
burn
any
amount
of
fossil
fuel
at
the
time
the
BART
determination
is
made
would
be
BART­
eligible
if
it
meets
the
other
criteria
in
the
guidelines.

2.1.1.5
Relevancy
of
Enforceable
Operational
Limits
for
Multi­
Fuel
Boilers
Comment:
Three
commenters
addressed
enforceable
emissions
limits
as
they
may
apply
to
multi­
fuel
boilers
(
0169,
0189,
0271).
One
commenter
(
0169)
said
that
the
option
of
taking
a
federally
enforceable
operational
limit
to
stay
below
the
250
million
BTU/
hr
cutoff
should
be
available
to
all
affected
boilers
and
not
only
to
multi­
fuel
boilers.
In
contrast,
commenter
0271,
said
that
enforceable
emission
limits
should
not
be
used
to
avoid
BART
controls
for
boilers
having
a
physical
design
capacity
to
emit
greater
than
250
tons
per
year
of
a
visibility
impairing
pollutant.
The
size
of
the
boiler
itself
should
be
the
only
determinant
as
this
will
prevent
intermittent
use
of
large
uncontrolled
units
which
may
adversely
affect
visibility
levels
in
Class
I
areas
albeit
for
shorter
intervals.

One
commenter
(
0189)
agreed
with
EPA's
proposal
to
consider
emission
limits
in
determining
whether
sources
are
high
emitting
enough
to
be
BART
eligible.
However,
EPA
must
make
clear
in
its
final
rule,
as
the
courts
have
required,
that
any
emission
limit
that
is
practically
enforceable
may
be
considered
and
that
Federal
enforceability
is
not
required.

One
commenter
(
28­
8000)
said
multi­
fuel
(
coal
and
bark)
boilers
with
a
capacity
of
greater
than
or
equal
to
250
MMBtu/
hr
,
but
which
have
a
federally
enforceable
limit
of
less
than
250
MMBtu/
hr
on
coal,
should
not
be
considered
"
BART
eligible",
since
they
do
not
exceed
the
heat
input
threshold
for
fossil
fuels.
This
logic
should
apply
to
boilers
burning
any
combination
of
fuels:
as
long
as
the
fossil
fuel
contribution
is
held
to
less
than
250
MMBtu/
hr
via
some
form
of
enforceable
mechanism
(
Federal
or
State),
the
boiler
should
not
be
subject
to
BART.
25
Response:

We
agree
that
it
would
be
reasonable
for
States
to
consider
a
boiler
with
a
federally
enforceable
operating
limit
of
less
than
250
MMBtu/
hr
as
not
BART­
eligible
but
only
so
long
as
that
limit
is
taken
before
the
BART
determinations
are
made
by
the
State
and
are
permanent.
This
interpretation
would
apply
to
all
boilers,
not
just
multi­
fuel
boilers.
In
addition,
the
regional
haze
regulations
define
"
potential
to
emit."
The
guidelines
repeat
that
regulatory
definition
and
provide
an
example
illustrating
its
application.
EPA
did
not
propose
to
change
the
definition
in
2001
or
2004,
but
merely
highlighted
the
current
definition
in
40
CFR
51.301.
Although
we
noted
in
the
2001
proposed
guidelines
that
we
expected
to
undertake
a
rulemaking
to
determine
whether
only
federally
enforceable
limitations
should
be
taken
into
account
in
the
regional
haze
program
definition,
we
have
not
yet
begun
the
process
for
such
a
rulemaking.
For
the
time
being,
we
believe
that
States
may
consider
emissions
limitations
in
State
permits,
which
are
enforceable
under
State
law,
in
determining
a
source's
"
potential
to
emit."

2.1.1.6
Steam
Electric
Plants
that
Generate
Electricity
for
Sale
Comment:
Three
commenters
supported
a
definition
of
steam
electric
plants
should
apply
only
to
plants
that
generate
electricity
for
sale
(
0169,
0259,
0285).
Three
other
commenters
opposed
this
definition
(
0232,
0271,
0286).
Commenters
said
this
limitation
is
not
included
in
the
Act.
Commenter
0271
added
that
consistency
with
other
rules
should
not
be
the
driving
factor
in
limiting
this
category.
The
proposed
amendment
removes
a
State's
ability
to
"
be
certain
that
it
has
not
prematurely
ruled
out
any
sensible
control
options
for
that
process
as
a
whole",
a
factor
noted
to
be
important
by
the
EPA
in
making
a
BART
determination
for
fossil
fuel
boilers
(
69
FR
25191).
The
commenter
said
that
EPA,
therefore,
has
not
provided
an
adequate
basis
for
amending
a
definition
provided
in
the
Act
in
its
proposal.

2.1.1.7
Emission
Units
Included
Under
the
"
Steam
Electric
Plant"
Definition
Comment:
Ten
commenters
commented
on
the
emission
units
that
should
be
included
in
a
definition
of
"
steam
electric
plant"
(
0188,
0198,
0207,
0226,
0228,
0230,
0237,
0254,
0256,
0279).
Some
commenters
said
that
only
those
emission
units
that
are
listed
as
one
of
the
26
categorical
sources
must
be
considered
in
determining
BART
eligibility
for
the
fossil­
fuel
fired
steam
electric
plant
category
(
0188,
0198,
0226,
0228).

One
commenter
(
0207)
said
that
the
weight
of
evidence
(
Acid
Rain,
Hazardous
Air
Pollutants,
NO
x
SIP
Call,
CAIR)
supports
a
conclusion
that
the
Regional
Haze
Rule's
definition
of
"
steam
electric
plant"
does
not
include
a
cogeneration
unit
unless
it
provides
more
than
onethird
of
its
PEOC
and
more
than
25
MW
to
a
utility
power
distribution
system
for
sale.
However,
because
the
Agency
has
gone
on
record
several
times
with
its
opinion
that
the
exclusion
of
certain
cogenerators
from
the
definition
of
an
EGU
is
no
longer
appropriate,
the
commenter
requests
that
26
EPA
clarify
that
the
exclusion
does
apply
to
the
RHR's
definition
of
"
steam
electric
plant."
Commenters
0228
and
0279
agreed
with
this
clarification.

Commenter
0237
said
there
is
no
statutory
basis
to
include
"
cogeneration
units"
in
the
1977
definition.
While
the
1990
Clean
Air
Act
scheme
under
the
acid
rain
program
included
cogeneration
units
as
electric
generating
units,
there
is
no
linkage
in
1977
to
establish
that
cogeneration
units
were
contemplated
to
be
within
the
scope
of
section
169A's
regime
or
to
require
consideration
of
such
units
in
determining
if
sources
in
other
BART
categories
are
"
major
stationary
sources.
"
Moreover,
the
inclusion
of
cogeneration
units
in
the
BART
program
provides
yet
another
EPA­
created
regulatory
disincentive
to
such
units
in
the
face
of
government
policies
to
encourage
the
efficiencies
from
the
co­
production
of
steam
and
electricity.
Thus,
the
requirement
to
"
aggregate"
co­
fired
cogen
units,
regardless
of
their
size,
with
other
units
in
any
of
the
other
25
BART
categories,
is
burdensome
and
entirely
arbitrary.

In
the
preamble
to
the
re­
proposed
guidelines,
EPA
discusses
its
interpretation
of
two
of
the
BART­
eligible
source
categories:
(
1)
"
steam
electric
plants
of
more
than
250
million
BTU/
hr
heat
input,"
and
(
2)
"
fossil­
fuel
boilers
of
more
than
250
million
BTU/
hr
heat
input."
EPA
states
that
"
steam
electric
plants"
refers
to
sources
with
boilers
that,
when
aggregated,
fall
above
a
250
million
BTU/
hr
heat
input
threshold.
For
a
given
plant,
the
commenter
(
0207)
understands
this
definition
to
require
aggregation
of
all
heat
inputs
of
those
utility
units
that
generate
electricity
for
sale
as
well
as
heat
inputs
of
any
auxiliary
boilers
at
the
plant
which
are
fossil
fuel­
fired.
Conversely,
the
commenter
understands
this
definition
to
exclude
the
heat
inputs
of
any
other
types
of
combustion
units
at
the
plant,
such
as
combustion
turbines
or
internal
combustion
engines.
However,
with
regard
to
EPA's
definition
of
"
fossil­
fuel
boilers,"
the
commenter
understands
that
applicability
of
this
source
category
is
a
unit­
specific
determination,
i.
e.,
it
turns
only
on
the
heat
input
of
an
individual
boiler
on
a
case­
by­
case
basis.
This
definition
would
apply
only
to
a
boiler
that
does
not
generate
electricity
for
sale.
The
commenter
asked
EPA
to
clarify
and
incorporate
this
approach
into
the
final
rule
and
the
BART
guidelines.

In
the
guidelines,
EPA
proposes
to
classify
all
combined
cycle
and
cogenerating
plants
as
steam
electric
plants.
Commenters
0228
and
0256
said
EPA
should
clarify
that
combined
cycle
units
with
heat
recovery
steam
generators
(
HRSG's)
that
lack
auxiliary
firing
are
considered
simple
cycle
turbines
under
the
definition
of
a
simple
combustion
turbine
and,
since
EPA
makes
clear
that
simple
cycle
turbines
are
not
steam
electric
plants,
should
add
a
provision
in
the
guidelines
that
excludes
combined
cycle
units
without
auxiliary
firing
from
the
steam
electric
plant
category
and
exempts
such
units
from
a
BART
review.
Commenter
0230
added
that
emission
units,
such
as,
simple
cycle
combustion
turbines,
internal
reciprocating
combustion
engines,
and
auxiliary
boilers
with
heat
input
less
than
250
MMBtu/
hour
would
not
meet
the
threshold
for
BART­
eligibility
(
i.
e.,
fossil
fuel­
fired
steam
electric
plants
with
heat
inputs
of
greater
than
250
MMBtu/
hour).
Similarly,
commenter
0256
interprets
EPA's
approach
to
exclude
from
the
BART
eligibility
analysis
(
and
thus
from
the
BART­
eligible
source)
any
other
emission
units
located
at
a
power
plant,
e.
g.,
auxiliary
boilers
with
heat
input
equal
to
or
less
than
250
MMBtu/
hour,
simple
cycle
combustion
turbines,
and
internal
reciprocating
combustion
engines.
Those
types
of
27
emissions
units
are
not
units
"
meeting
one
or
more
of
the
source
category
descriptions."
In
other
words,
those
emissions
units
are
not
fossil­
fuel
fired
steam
electric
generating
units,
fossil­
fuel
fired
boilers
with
heat
input
above
250
MMBtu/
hour,
or
any
of
the
other
source
categories
listed
in
section
169A(
g)(
7).

Commenter
0254
said
the
identification
of
BART­
eligible
steam
electric
EGUs
ought
to
exclude
auxiliary
onsite
sources,
such
as
peaking
turbines,
and
emergency
generators,
that
are
not
a
part
of
the
steam
electric
source
category.
Another
group
of
auxiliary
sources,
heat
recovery
steam
generators
used
to
generate
process
steam
(
e.
g.,
for
industrial
customers)
should
also
be
exempt
from
BART
eligibility
under
the
steam
electric
source
category.

Commenter
0256
said
for
power
plants
that
the
"
stationary
source"
definition
could
be
read
to
require
inclusion,
in
the
BART­
eligible
source,
of
emissions
units
other
than
fossil
fuelfired
steam
electric
generating
units
and
greater­
than­
250­
MMBtu/
hour
fossil
fuel­
fired
boilers
located
at
power
plants.
The
commenter
did
not
believe
this
was
EPA's
intent
and
requested
explicit
clarification
of
that
point
in
the
final
BART
guidelines.
If
this
was
EPA's
intent,
the
commenter
said
that
such
an
approach
would
be
inconsistent
with
the
Act
and
would
be
bad
policy.

One
commenter
(
28­
405)
said
EPA's
proposal
to
cover
steam
electric
generating
plants
with
total
heat
input
of
250
MMBtu's
per
hour
would
cover
units
as
small
as
100
MMBtu/
hr,
or
the
equivalent
of
approximately
5
tons
of
coal/
hr
at
20
million
BTUs
per
ton
(
assuming
10,000
BTU/
lb).
In
comparison,
the
750
MW
plant
size
contemplated
by
Congress
would
be
expected
to
consume
roughly
200­
300
tons
of
coal
per
hour.
This
expansion
of
Congressional
intent
would
cast
a
broad
net
capturing
virtually
all
fossil­
fired
powerplants
constructed
between
1962
and
1977.
The
rule
would
expose
hundreds
of
generating
units
­
small,
medium
and
large
alike
­
and
regardless
of
location
or
actual
air
quality
impact
on
Class
I
areas,
to
the
potential
cost
of
BART
retrofits
for
multiple
pollutants.
Commenter
28­
263
agreed
that
if
the
final
rule
does
not
limit
application
of
the
guidelines
to
greater
than
750
MW
powerplants,
applicability
must
be
based
on
individual
boiler
capacities.

Response
for
2.1.1.6
and
2.1.1.7
(
above
comments):

We
have
retained
the
clarification
that
this
source
category
refers
only
to
those
fossil­
fuel
fired
steam
electric
plants
that
generate
electricity
for
sale.
We
believe
that
this
clarification
helps
to
distinguish
those
plants
that
are
electric
utilities
from
plants
in
other
industrial
categories,
and
as
such
is
a
reasonable
interpretation
of
the
source
category
title.
We
also
believe
that
while
large
co­
generators
would
be
excluded
from
the
fossil­
fuel
fired
steam
electric
plant
source
category,
most
large
co­
generators
will
be
BART­
eligible
under
the
fossil­
fuel
fired
boilers
source
category.

We
do
not
believe
it
makes
sense
for
this
category
to
include
only
those
steam
electric
plants
that
burn
greater
than
50
percent
fossil
fuel.
We
do
not
believe
that
a
boiler
should
be
excluded
from
BART
review
simply
because
it
is
located
at
a
plant
which
burns
less
than
50
28
percent
fossil
fuel.
Emissions
from
any
such
boiler
could
be
a
significant
contributor
to
regional
haze,
and
as
such,
we
believe
that
each
fossil­
fuel
fired
boiler
merits
a
BART
review.

We
do
wish
to
clarify
that
units
which
are
located
at
a
steam
electric
plant,
but
which
themselves
are
not
in
any
of
the
26
BART
source
categories,
should
not
be
considered
to
be
BART­
eligible
units.
We
believe
that
Congress
intended
that
BART
review
be
focused
on
units
in
the
source
categories
it
delineated.
This
interepretation
is
most
consistent
with
the
definition
of
BART­
eligible
source
as
we
have
explained
it
elsewhere
in
this
preamble
in
reference
to
whether
entire
plants
are
included
if
only
some
units
at
the
plant
meet
the
statutory
criteria.

Finally,
we
believe
that
all
combined
cycle
units
are
included
in
the
definition
of
fossil
fuel
fired
steam
electric
plant,
regardless
of
whether
the
combined
cycle
unit's
heat
recovery
steam
generator
lacks
auxilliary
firing.
Commenters
are
correct
that
some
EPA
programs
have
treated
combined
cycle
units
with
supplemental
firing
differently
from
combined
cycle
units
without
supplemental
firing.
However,
while
some
EPA
programs
do
not
consider
a
unit
to
be
a
combined
cycle
unit
unless
it
contains
supplemental
firing,
the
definition
at
issue
here
is
the
definition
of
fossil­
fuel
fired
steam
electric
plant,
not
fossil­
fuel
fired
unit.
The
CAA
defines
both
"
stationary
source"
(
for
visibility
purposes)
and
"
major
emitting
facility"
(
for
PSD
purposes)
to
include
"
fossil
fuel
fired
steam
electric
plants."
In
previous
guidance
for
PSD,
we
have
explained
that
combined
cycle
gas
turbines
do
fall
within
the
category
of
"
fossil­
fuel
fired
steam
electric
plants."
(
See
http://
www.
epa.
gov/
Region7/
programs/
artrd/
air/
nsr/
nsrmemos/
turbines.
pdf).

Comment:
Three
commenters
said
clarification
to
the
guidance
is
need
to
distinguish
between
utility
boilers
and
industrial
boilers
with
respect
to
presumptive
controls
(
28­
339,
28­
343,
28­
383).
Commenter
28­
343
added
that
the
impact
of
presumptive
BART
on
industrial
boilers
differs
from
the
impact
on
utilities.
Although
they
both
create
steam,
a
utility
boiler
and
an
industrial
boiler
are
significantly
different
based
on
1)
the
size
of
the
boiler,
2)
the
application
of
the
steam
the
boilers
generate
and
3)
the
design
of
the
boilers.
Because
of
these
differences,
the
control
costs
are
higher
and
the
cost
effectiveness
is
lower
for
industrial
boiler
control
versus
utility
boiler
control.
Also,
the
smaller
size
of
the
industrial
boiler
unit
makes
its
impact
on
regional
haze
far
less.

Response:

We
have
clarified
that
presumptive
limits
only
apply
to
electric
utility
boilers.

2.1.1.8
Other
Comments
on
Source
Category
Definitions
Comment:
Four
commenters
in
2004
provided
comments
on
the
non­
steam
electric
plant,
non­
boiler
categories
(
0231,
0259,
0278,
0285).
One
commenter
(
0231)
supported
the
proposed
changes
to
the
source
category
definitions.
One
commenter
(
0259)
asked
EPA
to
clarify
that
states
may
use
29
SIC
code
3341
to
identify
"
secondary
metal
production"
facilities
and
requests
that
EPA
also
identify
the
NAICS
code(
s)
for
this
category.

One
commenter
(
0278)
asked
EPA
to
clarify
that
the
proposed
BART
rules
source
category
(
13)
Phosphate
rock
processing
plants
(
69
FR
25190)
was
directed
at
thermal
furnace
plants
and
not
plants
using
a
chemical
process
for
producing
phosphorus.
If
EPA
intended
to
include
chemical
processes
for
producing
phosphorus,
then
the
commenter
requested
that
EPA
reconsider
and
exclude
these
plants
from
the
source
category
because
there
are
only
de
minimis
emissions
from
such
operations
and
they
do
not
contribute
to
the
regional
haze
problem.

One
commenter
(
0285)
said
lime
plants,
portland
cement
plants,
iron
and
steel
mill
plants
and
chemical
process
plants
have
emission
units
belonging
to
other
BART
source
categories,
such
as
petroleum
storage
tanks.
The
commenter
asked
EPA
to
confirm
that
these
emission
units
will
be
considered
for
BART
control
only
if
they
are
at
or
above
the
threshold
level
of
the
applicable
source
category.
For
example,
storage
tanks
at
any
of
these
plants
belonging
to
the
above
source
categories
would
be
considered
for
BART
control
only
if
their
aggregate
capacity
exceeds
300,000
barrels.

Five
commenters
in
2001
provided
comments
on
individual
source
category
definitions
(
28­
364,
28­
244,
28­
304,
28­
391,
28­
748).
Commenter
28­
364
discussed
the
types
of
facilities
that
might
be
considered
part
of
the
charcoal
production
facilities
(
CPF)
category
and
noted
that
it's
inclusion
on
the
list
resulted
from
the
1975
TRC
study
findings
related
to
PM
emission
rates
from
these
facilities
and
the
benefits
of
low­
emissions
facilities.
However,
EPA's
one­
size­
fits­
all
approach
would
arbitrarily
penalize
low­
emission
CPFs
by
making
them
eligible
for
BART
retrofits.
In
addition
to
distorting
what
Congress
intended,
making
these
sources
BART
eligible
places
them
at
a
competitive
disadvantage
vis­
a­
vis
the
sources
Congress
actually
intended
to
reach
by
way
of
its
regional
haze
legislation.
EPA
should
establish
provisions
to
ensure
that
technology­
setting
operations
are
not
arbitrarily
drawn
into
the
BART
program.

Two
commenters
discussed
the
chemical
process
plants
category
(
28­
244,
28­
304).
Commenter
28­
244
urged
EPA
to
make
clear
that
"
chemical
process
plants"
consist
only
of
those
facilities
that
fall
within
SIC
group
28.
In
contrast,
commenter
28­
304
said
use
of
SIC
group
28
is
erroneous
and
EPA
should
clarify
that
the
category
does
not
include
pharmaceutical
manufacturing,
because
the
statute,
relevant
regulations,
and
related
materials
plainly
show
that
it
is
neither
legally
nor
practically
appropriate
for
the
term
"
chemical
process
plant"
to
include
pharmaceutical
manufacturing.

Commenter
28­
391
said
EPA
should
use
the
definition
found
at
40
CFR
60.919
for
clarifying
the
types
of
equipment
to
be
considered
when
defining
Primary
Aluminum
Ore
Reduction
Plant.
The
commenter
also
noted
that
neither
the
Act
nor
the
federal
PSD
regulations
define
the
term
"
secondary
metal
production
plants."
The
commenter
suggested
using
the
SIC
Code
3341
definition
so
that
a
secondary
metal
production
facility
means
one
or
more
emission
30
units
that
derive
more
than
fifty
percent
(
50%)
of
the
metal(
s)
it
produces
from
purchased
scrap
and
dross."

Commenter
28­
748
the
guidelines
should
provide
additional
clarification
regarding
"
petroleum
storage
and
transfer
facilities
with
a
capacity
exceeding
300,000
barrels."
The
guidelines
must
make
sure
that,
in
order
to
be
BART­
eligible,
a
300,000
barrel
facility
must
have
been
placed
in
operation
as
a
single
construction
project
at
one
time
during
the
1962­
1977
period.
Thus,
even
if
tanks
placed
in
operation
at
different
times
during
the
1962­
1977
period
in
the
aggregate
had
a
total
capacity
exceeding
300,000
barrels,
they
would
not
be
BART­
eligible
because
they
were
not
part
of
a
single,
discrete
project.
Moreover,
if
a
petroleum
storage
and
transfer
facility
was
in
existence
at
a
plant
site
in
1962,
it
would
not
be
BART­
eligible,
even
if
a
300,000
barrel
facility
was
later
added
to
it
between
1962
and
1977.

Response:

We
have
clarified
the
above­
mentioned
definitions
in
the
final
BART
guidelines
in
response
to
these
comments.
See
the
preamble
for
detailed
explanations
of
each
definition.

Comment:
Commenter
28­
304
said
EPA
should
confirm
that
research
and
development
(
R&
D)
facilities
are
not
covered
by
the
BART
guidance,
even
when
R&
D
facilities
"
support"
an
activity
that
belongs
to
one
of
the
28
listed
major
source
categories.
R&
D
is
fundamentally
different
than
manufacturing
and
should
not
be
burdened
by
fixed
air
pollution
control
technology
requirements
that
can
easily
interfere
with
the
ever­
changing,
creative
process
of
research
and
development.
Moreover,
emissions
from
R&
D
facilities
are
inconsequential
as
compared
to
manufacturing
activities
belonging
to
the
28
listed
source
categories.

Response:

We
do
not
expect
most
R&
D
facilities
to
have
emissions
great
enough
to
qualify
as
BART­
eligible.
The
decision
as
to
whether
to
consider
R&
D
facilities
as
BART­
eligible
is
one
that
the
States
will
need
to
consider
on
a
case
by
case
basis.
If
States
would
like
assistance
on
this
issue,
they
may
request
guidance
from
EPA
on
this
issue.

2.1.1.9
Other
Comments
on
Identifying
the
Emission
Units
in
the
BART
Categories
Comment:
Two
commenters
addressed
implementation
issues
associated
with
identification
of
emission
units
in
the
BART
categories
(
0237,
0299).
Commenter
0237
said
by
applying
BART
to
certain
"
units"
underscores
how
broad
the
guideline
would
apply
far
more
broadly
than
the
Congress
intended.
The
logical
conclusion
of
EPA's
proposed
step­
by­
step
analysis
would
be,
if
BART
exists,
to
apply
BART
to
individual
pieces
of
equipment
down
to
a
piece
of
equipment
that
has
the
potential
to
emit
a
ton
or
less
of
a
regulated
BART
pollutant.
But
since
controls
generally
31
are
tied
to
equipment
groupings,
the
result
would
be
controlling
many
units
built
before
1962
and
potentially
some
installed
under
minor
NSR
after
1977.

Response:

We
believe
that
for
identifying
those
units
at
a
plant
which
are
BART­
eligible
in
part
as
those
units
constructed
between
1962­
77,
we
have
interpreted
the
CAA
in
a
logical,
reasonable
and
consistent
way,
as
explained
in
the
preamble
and
in
responses
to
comments
under
2.2
below.

Comment:
Commenter
0299
suggested
that
to
ease
the
resource
burden
imposed
on
the
states
by
this
rule,
EPA
should
phase­
in
the
program
whereby
the
states
would
be
allowed
to
determine
if
there
is
a
subset
of
the
twenty­
six
categories
of
sources,
accounting
for
the
bulk
of
the
targeted
emissions
reduction,
and
then
be
allowed
to
focus
on
those
source
categories
first.

Commenters
(
28­
266,
28­
278)
said
EPA
should
provide
a
list
of
SIC
and
SCC
codes
for
each
of
the
26
specific
source
categories,
and
a
list
of
the
visibility
impairing
pollutants.
Also,
is
this
list
fixed,
or
do
states
have
options
to
add
other
pollutants?
Commenter
28­
290
said
EPA
should
identify
the
number
of
sources
that
will
potentially
be
subject
to
BART
and
where
these
sources
are
located.
If
it
becomes
evident
that
implementation
of
a
case­
by­
case
BART
engineering
analysis
will
impose
an
undue
resource
burden
on
a
particular
area,
EPA
should
explore
ways
to
alleviate
this
burden
under
special
circumstances.

Other
commenters
noted
that
the
categories
are
too
ambiguous
(
28­
345,
28­
364,
28­
391).
Commenter
28­
264
said
EPA
must
properly
define
and
interpret
the
scope
of
BART­
eligible
sources
so
that
companies
and
individual
sources
are
not
improperly
placed
upon
the
BARTeligible
list.
Commenter
28­
391
said
EPA
should
identify
the
specific
emission
units
to
which
BART
applicability
applies
in
each
of
the
listed
categories.

Two
commenters
said
the
SIC
system
should
be
used
instead
of
the
NAICS
because
it
is
a
one­
time
activity
and
making
a
change
will
only
add
confusion
(
28­
255,
28­
256).
However,
one
commenter
said
the
two­
digit
SIC
system
is
too
broad
and
EPA
should
shift
to
the
NAICS.
In
the
absence
of
a
change
to
the
newer
classification
system,
clarification
to
the
guidelines
is
required
to
ensure
that
individuals
interpreting
these
guidelines
use
the
"
full
four
digit"
SIC
Code
to
distinguish
between
source
types.

Response:

We
expect
that
the
number
of
BART­
eligible
sources
in
each
State
will
be
such
that
States
can
complete
BART
determinations
in
the
time
alloted
by
the
CAA,
and
that
the
BART
determination
process
is
flexible
enough
that
States
can
require
greater
control
on
those
sources
which
contribute
the
most
to
visibility
impairment.
32
Also,
we
have
done
our
best
to
clarify
ambiguities
in
the
source
category
titles,
as
explained
above
and
in
the
preamble.
Because
the
source
category
list
for
BART
is
the
same
under
the
CAA
as
the
source
category
list
for
PSD/
NSR,
we
believe
there
is
a
common
understanding
across
States
of
most
of
the
source
category
titles
and
that
States
will
be
able
to
make
reasonable
determinations
as
to
whether
a
source
falls
within
a
listed
category.

Comment:
One
commenter
28­
272
said
EPA
should
provide
an
option
for
adoption
of
emission
standards
for
individual
source
categories.

Response:

The
BART
determination
process
is
targeted
at
individual
sources,
but
States
do
retain
considerable
discretion
in
how
they
approach
making
those
BART
determinations.

2.1.2
Guidance
on
Definition
of
Emission
Units
Comment:

Commenter
28­
266
said
EPA
should
identify
in
the
guidelines
the
emission
units
within
source
categories
to
which
BART
applies.
Commenter
28­
244
said
EPA
has
not
tailored
its
definition
of
"
unit."
Nothing
in
EPA's
proposed
approach
[
for
selecting
units
subject
to
BART]
does
anything
to
make
sure
that
the
particular
set
of
"
units"
at
a
facility
selected
will
provide
a
reasonable
and
logical
platform
for
the
installation
of
controls.
If
two
out
of
five
units
that
were
generally
"
controlled
jointly"
had
been
installed
during
the
"
window"
period,
it
might
not
make
much
sense
to
subject
them
to
BART
requirements.
For
example,
many
pulp
mills
have
multiple
digesters,
which
are
installed
at
various
times
but
may
be
controlled
together.
If
two
of
six
digesters
had
been
installed
during
the
"
window"
period,
BART
would
be
required
for
them,
but
not
for
the
other
four.
Given
that
digesters
are
controlled
together,
what
sense
does
this
make?
And
how
would
BART
for
the
eligible
sources
be
determined?

Response:

We
disagree
with
commenters
that
the
approach
set
forth
in
the
guidelines
for
identifying
BART­
eligible
sources
is
inappropriate
because
the
particular
set
of
units
identified
as
BARTeligible
will
not
necessarily
"
provide
a
reasonable
and
logical
platform
for
the
installation
of
controls."
We
do
not
agree
that
this
factor
is
relevant
to
the
identification
of
those
emissions
units
which
meet
the
definition
of
BART­
eligible
source.
Such
factors
are
important
in
the
States'
consideration
of
control
strategies
and
options
but
do
not
clearly
relate
to
the
first
step
of
identifying
those
sources
which
fall
within
one
of
26
source
categories,
were
built
during
the
15
year
window
of
time
from
1962
to
1977,
and
have
potential
emissions
of
greater
than
250
tons
per
year.
We
do
thus
agree
generally
with
the
commenter's
recommendation
of
allowing
States
33
to
consider
the
particular
history
and
control
potential
of
units
in
determining
BART,
but
do
not
agree
that
it
is
relevant
to
the
predicate
question
of
identifying
the
BART­
eligible
source.

2.2
Start­
up
Dates
of
Emission
Units
2.2.1
Definitions
of
"
In
Existence"
and
"
In
Operation"

Comment:
One
commenter
(
0231)
supports
retaining
the
same
definition
of
"
in
existence"
and
"
in
operation"
as
were
included
in
the
2001
proposal.
A
second
commenter
(
0237)
disagreed
with
the
April
2004
notice,
which
equates
EPA's
NSR
interpretation
of
"
commenced
construction"
with
the
phrase
"
in
existence
in
1977."
This
reinterpretation
of
unambiguous
statutory
language
in
Section
169A
expands
the
field
of
BART­
eligible
sources
with
no
justification
to
sources
for
which
owners
and
operators
had
begun
actual
construction
or
for
which
they
had
made
binding
commitments
to
construct.
Neither
type
of
source
was
"
in
existence,"
on
August
7,
1977.
There
is
no
supportable
policy
reason
for
broadening
the
statutory
intent
of
the
program.
In
the
NSR
program,
in
which
the
agency
created
a
definition
to
deal
with
the
issue
of
when
sources
were
grandfathered
from
the
1980
PSD
program
as
sources
with
knowledge
of
impending
permit
regulations
hurried
to
sign
contracts
before
the
regulations
themselves
became
final
(
an
interpretation
which
the
courts
have
determined
to
be
reasonable),
here
there
is
no
such
problem
because
the
Congress
was
attempting
to
bracket
a
group
of
sources
that
existed
before
the
1977
Amendments
were
passed,
that
would
be
otherwise
left
unregulated
by
the
Clean
Air
Act
of
1977,
and
which
Congress
deemed
reasonable
to
regulate
nonetheless.

2.2.2
Applicability
Only
to
Boilers
at
a
Plant
Put
in
Place
between
1962
­
1977
Comment:
Four
commenters
asked
EPA
to
clarify
in
the
guidelines
that
the
BART
requirements
should
not
apply
to
boilers
at
a
power
plant
that
were
not
built
between
the
1962
and
1977
timeframe,
even
if
one
or
more
other
boilers
at
the
plant
were
built
within
the
relevant
time
period
(
0210,
0241,
0247,
0253).
One
commenter
(
0253)
disagreed,
however,
that
the
heat
input
of
all
emissions
units
must
be
aggregated
even
if
some
of
those
units
fall
outside
the
1962
to
1977
date
range
required
for
BART­
eligibility.
Although
emissions
units
outside
the
date
range
are
subsequently
excluded
under
step
two
of
the
formula,
the
basic
approach
appears
to
be
inconsistent
with
the
provisions
of
the
statute
and
could
be
misconstrued
in
administration
of
the
BART
program.
Consequently,
it
would
be
preferable
to
exclude
emissions
units
outside
the
date
range
from
the
step
one
heat
input
calculation
in
order
to
avoid
any
confusion
that
might
arise
in
the
future.

2.2.3
Applicability
to
Individual
Emission
Units
Comment:
34
Eighteen
commenters
addressed
the
unit­
by­
unit
approach
to
determining
applicability
(
0169,
0189,
0207,
0237,
0307,
28­
244,
28­
258,
28­
325,
28­
331,
28­
337,
28­
345,
28­
349,
28­
350,
28­
352,
28­
355,
28­
364,
28­
386,
28­
395).
One
commenter
(
0307)
appreciated
the
clarification
that
one
or
more
BART­
eligible
units
in
a
facility
do
not
make
all
similar
units
within
that
facility
subject
to
BART.
Other
commenters
(
28­
258,
28­
325,
28­
350,
28­
352)
said
the
threshold
should
apply
to
all
units
at
a
facility.

Other
comments
identified
problems
with
the
approach.
One
commenter
(
0169)
said
EPA's
proposed
guidelines
extend
well
beyond
section
169(
b)(
2)(
A)
by
requiring
states
to
determine
BART
for
all
emission
units
constructed
or
reconstructed
between
8/
7/
1962
and
8/
7/
1977
at
all
of
the
listed
26
major
stationary
sources.
Commenters
(
28­
331,
28­
337,
28­
364,
28­
386,
28­
395)
objected
to
the
expansion
of
the
Regional
Haze
rule
definition
of
BART­
eligible
sources
by
requiring
states
to
find
that
all
sources
at
a
facility
are
subject
to
BART
if
one
source
at
the
facility
is
subject
to
BART.
Commenter
28­
244
said
this
policy
is
unjustified
and
far
too
inclusive
especially
when
one
considers
that
the
emission
amount
is
calculated
on
a
PTE
basis
which
will
always
be
greater
than
and
often
far
greater
than
actual
emissions.

Commenters
(
28­
345,
28­
349,
28­
355)
said
the
proposed
"
unit­
by­
unit"
analysis
for
BART
eligibility
is
contrary
to
Congressional
intent.
Commenter
28­
345
added
that
Congress
intended
that
BART
would
apply
only
if
the
emissions
from
an
entire
plant
exceeded
the
250
tpy
cut­
off.
The
statutory
BART
provisions
do
not
mention
individual
emissions
units,
any
type
of
aggregation
over
time,
or
aggregation
involving
parts
of
different
types
of
major
sources.
The
"
in
for
one,
in
for
all"
approach
is
not
consistent
with
other
CAA
programs
designed
to
reduce
source
emissions,
such
as
NSR
and
MACT.
EPA
should
revise
the
rule
so
that
meeting
the
threshold
of
one
pollutant
would
not
per
se
qualify
a
facility
as
a
major
source
of
other
pollutants.
Commenter
28­
349
said
the
proposal
would
create
a
burdensome,
confusing
process.
Under
the
emissionsunit
by­
emissions­
unit
approach,
emissions
from
"
linked"
units
should
not
be
considered
in
determining
BART­
eligibility.
A
unit
should
not
be
considered
BART­
eligible
unless
that
unit
itself
satisfies
all
the
statutory
criteria
for
BART
applicability.

Another
commenter
(
0189)
said
that
in
cases
where
different
units
at
a
source
were
constructed
or
reconstructed
at
different
times,
it
may
be
difficult
to
determine
the
exact
extent
of
BART
coverage.
EPA's
unit­
by­
unit
approach
to
address
this
problem
fails
to
ensure
that
the
particular
set
of
"
units"
at
a
facility
selected
by
this
test
will
provide
a
reasonable
and
logical
platform
for
the
installation
of
controls.
Citing
several
issues,
the
commenter
concluded
that
the
best
approach
to
this
problem,
consistent
with
the
overall
intent
of
Congress
in
enacting
section
169A,
would
be
to
allow
states
to
consider
the
particular
history
and
control
potential
of
units
in
determining
how
much
if
any
BART
to
require.
Congress
clearly
meant
to
give
states
this
authority
when
it
authorized
them
to
consider
cost
and
the
degree
of
control
already
installed
in
making
BART
determinations.
As
a
"
second
best"
alternative
(
but
not
endorsed
by
the
commenter),
EPA
should
make
BART
applicable
to
"
process
or
production
units"
as
defined
in
the
section
112(
g)
rule
as
units
that
perform
in
an
independent
function
within
the
facility.
35
Commenter
(
0207)
said
that
a
better
reading
of
section
169A
is
one
that
consistently
determines
BART
applicability
on
a
unit­
by­
unit
basis
regardless
of
the
source
category.
Interpreting
section
169A's
usage
of
the
term
"
major
stationary
source"
to
apply
on
an
individual
emissions
unit
basis
is
consistent
with
the
1977
Legislative
History.
The
commenter
said
that
EPA's
re­
proposed
guidelines
which
advance
the
concept
of
retrofitting
entire
plants
is
a
statutory
interpretation
with
more
weight
than
the
actual
words
of
the
Act
can
bear.
Furthermore,
in
that
case,
there
is
no
need
for
the
concept
of
a
de
minimis
emission
rate,
supra
at
12­
13.
The
unit's
potential
to
emit
will
either
trigger
the
250
tpy
threshold,
or
it
will
not.

One
commenter
(
0237)
said
EPA
has
failed
to
provide
any
reason
in
the
record
of
this
rulemaking
why
it
is
not
possible
for
states
to
determine
from
state
and/
or
local
tax
records,
or
some
other
source,
what
major
stationary
sources
existed
in
1977.
The
legislative
history
of
the
visibility
program,
however,
clearly
refers
to
major
emitting
facilities,
called
major
stationary
sources
in
the
BART
program,
to
avoid
confusions
with
the
broader
categories
of
major
emitting
facilities
covered
by
the
PSD
program.
The
1977
amendments
did
not
address
units.
Instead
Congress
referred
to
certain
"
major
stationary
sources"
in
categories
of
industry
defined
by
EPA
in
prior
PSD
and
NSPS
regulations.
EPA
has
failed
to
explain
in
the
proposal
why
it
has
departed
from
these
statutory
definitions
or
its
prior
regulations
by
focusing
on
"
units"
and
aggregation
concepts.

One
commenter
(
28­
344)
said
EPA
violates
the
plain
language
of
the
Act
by
determining
applicability
based
on
whether
sources
had
received
construction
permits
on
August
7,
1977
rather
than
whether
they
were
"
in
existence"
on
that
date.

Commenter
28­
244
said
EPA's
unit­
by­
unit
approach,
in
its
current
form,
does
not
make
much
of
a
positive
contribution
toward
a
workable
BART
program.
EPA
should
allow
States
to
consider
the
particular
history
and
control
potential
of
units
in
determining
how
much
if
any
BART
to
require.
Or,
instead
of
making
BART
applicable
to
undefined
"
units,"
instead
make
it
applicable
to
"
process
or
production."
"
Process
or
production
unit,"
would
have
the
same
meaning
that
it
does
in
EPA's
Section
112(
g)
rule.

Two
commenters
said
even
if
it
is
co­
located
with
an
eligible
unit,
a
unit
that
started
operation
before
August
7,
1962
is
not
BART­
eligible
(
28­
349,
28­
385).
Commenter
28­
349
added
that
the
proposed
approach
will
simply
create
a
burdensome,
wasteful
process
that
will
not
change
the
outcome
that
would
have
been
reached
under
the
simpler
approach
intended
by
Congress.
The
approach
will
engender
uncertainty
and
delay
without
providing
significant
environmental
benefits.

Response
to
2.2.1,
2.2.2,
and
2.2.3:

We
disagree
with
the
comments
recommending
that
we
interpret
the
term
"
in
existence"
to
refer
to
sources
that
are
in
actual
operation.
The
discussion
of
this
term
in
Step
2
is
based
on
the
regulatory
definition
which
has
been
in
place
since
1980.
The
guidelines
reiterate
this
definition
1
See
CAA
section
193.

36
and
provide
examples
of
its
application.
Interpreting
the
term
"
in
existence"
as
suggested
by
commenters
would
not
be
consistent
with
the
plain
language
of
the
regulations.
In
the
2001
and
2004
proposed
guidelines,
we
noted
that
"
the
term
`
in
existence'
means
the
same
thing
as
the
term
`
commence
construction'
as
that
term
is
used
in
the
PSD
regulations."
Commenters
were
critical
of
this
statement,
claiming
that
EPA
was
unlawfully
reinterpreting
section
169A
in
the
guidelines.
The
statement
in
Step
2
of
guidelines,
however,
is
not
a
reinterpretation
of
the
term
"
in
existence,"
but
merely
a
statement
noting
that
the
definitions
used
in
the
visibility
regulations
and
the
PSD
regulations
are
essentially
identical.
To
the
extent
that
commenters
are
claiming
that
the
existing
regulatory
definition
of
"
in
existence"
is
unlawful,
EPA's
interpretation
of
this
term
in
promulgating
the
1980
regulations
was
a
reasonable
one.
First,
it
is
worth
noting
that
the
regulations
adopting
this
interpretation
of
the
term
"
in
existence"
were
in
effect
in
1990
and
implicitly
endorsed
by
Congress
in
its
1990
amendments
to
the
CAA.
1
Moreover,
the
definition
at
issue
accurately
reflects
Congress'
intent
that
the
BART
provision
apply
to
sources
which
had
been
"
grandfathered"
from
the
new
source
review
permit
requirements
in
parts
C
and
D
of
title
I
of
the
CAA.
For
all
the
above
reasons,
we
are
neither
revising
the
regional
haze
regulations
to
change
the
definition
of
"
in
existence,"
nor
adopting
a
strained
interpretation
of
the
regulation
in
the
guidelines.
We
agree
with
commenters
that
the
definition
of
"
BART­
eligible
source"
does
not
require
States
to
find
that
all
emission
units
at
a
facility
are
subject
to
the
requirement
of
the
BART
provisions
if
only
one
part
of
the
facility
was
built
within
the
1962­
1977
time
period.
We
received
comments
on
this
issue
in
2001
and
clarified
in
2004
that
the
BART
guidelines
do
not
direct
States
to
find
that
all
boilers
at
a
facility
are
BART­
eligible
if
one
or
more
boilers
at
the
facility
were
put
in
place
during
the
relevant
time
period.
Under
Step
2
of
the
process
for
identifying
BART­
eligible
sources
set
out
in
the
guidelines,
States
are
required
to
identify
only
those
boilers
that
were
put
in
place
between
1962
and
1977.
As
explained
in
the
preamble
to
the
2004
reproposed
guidelines,
only
these
boilers
are
potentially
subject
to
BART.
We
do
not
agree
with
those
commenters
claiming
that
Congress
clearly
intended
to
apply
BART
only
if
an
"
entire
plant"
was
put
into
place
between
1962
and
1977.
Most
of
the
BART
source
categories
are
broad
descriptions
types
of
industrial
facilities
such
as
"
kraft
pulp
mills,"
"
petroleum
refineries"
or
"
primary
copper
smelters."
For
such
source
categories,
the
implication
of
commenters'
argument
would
that
if
any
portion
of
the
plant
was
in
operation
before
August
7,
1962,
then
Congress
intended
to
exempt
the
entire
plant
from
BART.
Such
an
interpretation
is
problematic
and
inequitable.
For
example,
under
this
approach
BART
would
not
apply
if
a
company
chose
to
expand
its
production
by
building
a
second
production
line
at
an
existing
line
in
1965,
but
would
apply
if
the
same
company
chose
to
build
the
same
equipment
at
a
greenfield
site.
Under
the
approach
set
forth
in
the
guidelines,
such
a
production
line
would
be
treated
similarly
under
either
set
of
facts.
We
do
not
believe
that
either
the
plain
language
of
the
statute
or
the
relevant
legislative
history
indicate
that
Congress
intended
for
major­
emitting
sources
of
visibility­
impairing
pollutants
to
be
exempted
from
the
BART
requirements
because
a
plant
contains
some
emission
units
that
began
operation
before
1962.
37
We
agree
with
the
comments
related
to
"
linked"
emission
units.
The
comment
appears
to
address
whether
emissions
from
the
"
linked"
units
are
considered
in
determining
BART
eligibility.
In
the
guidelines,
we
are
focusing
on
only
the
emissions
units
that
were
put
in
place
during
the
1962
to
1977
dates
and
the
emissions
from
those
units.
We
agree
that
even
if
changes
in
emissions
from
one
unit
could
affect
the
emissions
from
a
"
linked"
unit
that
was
not
put
in
place
within
the
1962­
1977
time
period,
this
would
not
affect
whether
the
"
linked"
unit
was
BARTeligible
The
approach
to
identifying
a
"
BART­
eligible
source"
in
the
guidelines
is
based
on
the
definitions
in
the
regional
haze
rule
of
the
relevant
terms.
For
750
MW
power
plants,
States
are
required
to
apply
the
definitions
as
set
forth
in
the
guidelines;
for
other
sources,
States
may
adopt
a
different
approach
to
the
task
of
identifying
BART­
eligible
sources,
so
long
as
that
approach
is
consistent
with
the
Act
and
the
implementing
regulations.
In
other
words,
while
the
guidelines
adopt
an
approach
for
large
power
plants
which
involves
the
aggregation
of
all
emissions
units
put
into
place
between
1962
and
1977,
States
have
the
flexibility
to
consider
other
reasonable
approaches
to
the
question
of
identifying
BART­
eligible
sources
for
other
source
categories.
This
also
applies
to
the
comment
on
States'
ability
to
use
the
"
process
or
production
unit"
definition
in
112(
g).
For
750
MW
power
plants,
many
of
the
issues
identified
by
commenters
with
the
approach
of
looking
at
a
facility
on
an
emission
unit
by
emission
unit
basis
do
not
exist.
Unlike
many
types
of
industrial
processes,
power
plants
consist
generally
of
a
discrete
number
of
very
large
emission
units.
For
other
types
of
facilities
such
as
kraft
pulp
mills
or
chemical
process
plants
which
may
have
many
small
emission
units
that
have
undergone
numerous
changes,
the
guidelines
do
not
limit
the
ability
of
the
States
to
approach
the
question
of
identifying
BART­
eligible
sources
in
ways
which
make
sense
for
the
particular
sources
given
their
design
and
history.
Finally,
we
believe
that
Congress
intended
the
BART
provision
to
apply
to
sources
which
had
been
"
grandfathered"
from
the
new
source
review
permit
requirements
in
parts
C
and
D
of
title
I
of
the
CAA.
We
do
not
believe
that
Congress
intended
that
the
term
"
in
existence"
should
be
defined
as
an
"
actual
operation"
test.
If
this
interpretation
were
correct,
there
would
be
sources
that
began
actual
operation
after
1977
that
would
not
be
subject
to
the
requirements
of
either
new
source
review
or
BART.
As
per
the
definition
of
"
commenced"
in
CAA
section
169(
2)(
A),
sources
are
grandfathered
from
new
source
permitting
requirements
if
they
have
"
entered
into
binding
agreements
or
contractual
obligations,
which
cannot
be
canceled
or
modified
without
substantial
loss
to
the
owner
or
operator."
The
definitions
in
the
regional
haze
rule,
as
reiterated
unchanged
by
the
BART
guidelines,
take
the
position
that
Congress
intended
for
any
source
that
is
not
considered
to
be
"
new"
as
of
August
7,
1977
should
be
considered
to
be
"
in
existence"
as
of
that
date.
Accordingly,
the
regional
haze
rule
uses
the
same
words
from
the
definition
of
"
commenced"
in
CAA
section
169(
2)(
A)
to
describe
those
sources
that
are
"
in
existence"
for
purposes
of
BART.
We
also
note
that
Congress
used
the
term
"
in
operation"
in
reference
to
1962,
and
could
have
used
the
term
"
in
operation"
for
1977
if
it
wanted
to
limit
the
application
of
the
BART
provision
in
the
way
that
some
commenters
have
suggested.

2.2.4
Eligibility
Dates
of
Emission
Units
38
Comment:
Thirteen
commenters
addressed
the
statutory
basis
and
practical
considerations
for
going
beyond
the
15­
year
BART
eligibility
window
(
0207,
0271,
0286,
28­
243,
28­
259,
28­
267,
28­
304,
28­
343,28­
345,
28­
349,
28­
364,
28­
391,
28­
748).
Commenters
(
0207,
28­
243,
28­
259,
28­
267,
28­
304,
28­
343,
28­
345,
28­
349,
28­
364,
28­
391
)
said
that
the
Act
clearly
limits
BART
to
units
that
were
in
existence
from
1962
­
1977
and
the
reconstruction
provisions
should
be
deleted.
Commenter
28­
304
said
section
169A(
h)(
2)(
A)
states
that
BART
is
potentially
required
for
major
stationary
sources
that
are
"
in
existence
on
August
7,
1977,
but
which
[
have]
not
been
in
operation
for
more
than
fifteen
years
as
of
such
date."
As
an
initial
matter,
the
words
"
in
existence"
are
abundantly
clear
and
unambiguous
­
if
the
major
stationary
source
is
physically
present
as
of
August
7,
1977,
then
BART
potentially
applies.
The
notion
that
physical
presence
depends
on
the
existence
or
non­
existence
of
major
facility
upgrades
is
simply
not
plausible.
Thus,
statutory
command
is
clear,
leaving
no
latitude
for
EPA's
proposal
that
reconstruction
somehow
changes
the
date
that
a
source
comes
into
existence.
This
conclusion
is
buttressed
by
the
statutory
provisions
that
underlie
the
existing
EPA
regulatory
programs
that
employ
the
concept
of
reconstruction.

Commenter
0207
added
that
the
legislative
history
of
Section
169A
of
the
1977
Amendments
to
the
Act
makes
clear
that
neither
EPA
nor
any
state
has
the
authority
at
this
time
to
require
BART
for
any
emission
unit
which
was
in
existence
on
August
7,
1977
but
was
not
in
operation
before
August
7,
1962.
The
95th
Congress
deliberately
restricted
the
application
of
BART
to
those
units
which
were
no
more
than
15
years
old
at
the
time
when
the
1977
Amendments
were
enacted.
Yet
almost
27
years
later,
EPA's
proposal
would
apply
BART
to
those
same
emission
units,
i.
e.,
ones
started
up
between
August
7,
1962
and
August
7,
1977.
However,
only
now
in
2004,
does
EPA
propose
to
require
BART
on
certain
emissions
units
as
old
as
27­
42
years
old.
EPA's
proposed
BART
requirement,
as
applied
to
emission
units
much
older
than
15
years
old,
is
clearly
inconsistent
with
the
specific
intention
of
the
Congress
at
the
time
that
section
169A
was
enacted.
That
regulatory
requirement
is
"
manifestly
contrary
to
the
statute."
Thus,
EPA's
proposed
BART
requirement
is
ultra
vires
and
cannot
be
legally
binding.

Commenters
(
28­
345,
28­
349,
28­
364,
28­
748)
that
one
of
the
principal
flaws
in
the
proposed
consideration
of
"
reconstructed"
units
is
the
attempt
to
treat
the
BART
program
as
if
it
were
analogous
to
the
NSPS
program.
The
concept
of
reconstruction
has
no
place
in
the
BART
program.
Serious
practical
problems
such
as
the
lack
of
records
would
make
the
proposed
approach
to
consider
"
reconstructed"
emissions
units
unworkable.
Other
commenters
(
28­
391,
28­
779)
also
raised
concerns
about
the
lack
of
records
to
make
an
applicability
determination.

In
contrast,
commenters
said
EPA
should
expand
the
range
of
dates
for
eligibility
of
units
under
the
BART
rule
to
all
units
in
operation
prior
to
August
7,
1997
(
0271)
or
to
all
large
and
previously
uncontrolled
units,
regardless
of
their
age
or
date
of
first
operation
(
0286).

Comment:
39
One
commenter
(
28­
398)
said
EPA
lacks
legal
authority
to
issue
guidelines
in
2001
requiring
states
to
perform
BART
analyses
on
stationary
sources
which
came
into
existence
between
1962
and
1977
because:
1)
As
of
2001,
these
sources
are
now
between
24
and
39
years
old
and
Congress
never
intended
for
sources
this
old
to
be
subjected
to
this
kind
of
continuing
retrofit
requirements;
(
2)
Sources
over
20
to
40
years
old
do
not
have
enough
"
remaining
useful
life"
to
allow
for
the
imposition
of
retrofitting
requirements;
(
3)
Since
1980
when
EPA
promulgated
visibility
rules,
the
states
have
already
had
the
opportunity
to
address,
and
have
addressed
many,
if
not
all,
of
those
1962
to
1977
sources
which
were
"
inadequately
controlled"
and
causing
significant
visibility
impacts
in
Class
I
areas
with
which
Congress
was
concerned.

Response:

The
discussion
of
"
reconstruction"
is
based
on
the
regulatory
definition
of
the
term
which
has
been
in
place
since
1980.
We
believe
it
makes
sense
for
States
to
have
the
flexibility
to
consistently
define
"
reconstruction"
across
programs,
and
that
a
major
reconstruction
should
reasonably
be
considered
as
"
new."

We
also
note
that
the
statutory
requirement
for
BART
is
still
in
effect,
and
nothing
in
the
CAA
allows
EPA
to
disregard
the
requirements
in
section
169A
because
sources
subject
to
it
have
aged
considerably.
We
do
not
believe
that
any
potential
lack
of
records
would
pose
a
problem
 
sources
can
easily
produce
records
to
show
when
they
were
constructed.
We
believe
that
the
statutory
language
of
section
169A
is
clear
enough
on
dates
and
that
we
cannot
preemptorily
change
those
dates.

In
addition,
many
of
these
very
old
sources
may
still
be
uncontrolled
and
have
considerable
emissions
of
visibility
impairing
pollution;
we
believe
they
merit
a
BART
review.
States
have
the
ability
to
judge,
at
the
BART
determination
step,
whether
or
not
sources
are
adequately
controlled.

2.2.5
Modified
Units
and
BART
Eligibility
2.2.5.1
Agree
with
EPA's
Proposal
to
Address
Modified
Units
Comment:
Nineteen
commenters
generally
agreed
with
EPA's
approach
to
modified
units
(
28­
244,
28­
255,
28­
256,
28­
258,
28­
263,
28­
267,
28­
273,
28­
302,
28­
335,
28­
345,
28­
349,
28­
358,
28­
362,
28­
364,
28­
388,
28­
404,
28­
748,
28­
770,
28­
779).
Commenters
(
28­
258,
28­
267)
specifically
supported
EPA's
proposal
to
consider
sources
that
began
operation
in
the
1962­
1977
window,
but
were
modified
after
1977
as
BART­
eligible
sources.
Other
commenters
(
28­
273,
28­
345,
28­
349,
28­
358,
28­
364,
28­
404,
28­
779)
also
agreed
that
plants
existing
before
1962
that
were
later
modified
should
be
exempt
from
the
BART
requirements.
Commenters
(
28­
345,
28­
349,
28­
358,
28­
364,
28­
404,
28­
748,
28­
779)
said
this
reading
is
consistent
with
the
statute
and
avoids
the
difficulty
of
considering
modifications
that
may
have
occurred
up
to
40
years
ago.
40
Commenter
28­
244
said
making
units
subject
to
BART
if
they
were
modified
during
the
1962­
1977
period
would
contradict
the
intent
of
Congress
to
exempt
pre­
1962
units
from
control.
It
would
be
very
difficult
if
not
impossible
to
determine
whether
modifications
occurred
during
a
period
38
to
24
years
ago,
at
a
time
when
no
legal
requirements
concerning
modifications
were
in
place.
Many
sources
may
no
longer
have
the
relevant
records.
Commenters
(
28­
255)
agreed
that
modifications
to
emissions
units
should
not
change
it's
eligibility
status.
Commenter
28­
256
said
the
best
interpretation
of
modifications
for
the
purposes
of
the
visibility
provisions
is
that
modified
units
are
still
"
existing."
Other
commenters
(
28­
267,
28­
362)
said
interpretation
of
"
modification"
in
the
BART
guidelines
should
be
consistent
with
the
interpretation
of
that
term
in
the
PSD
regulations,
i.
e.,
sources
that
propose
to
undergo
a
major
modification
are
to
be
treated
as
new
sources
and
must
undergo
NSR.
Commenter
28­
362
said
these
sources
in
operation
prior
to
1962
but
modified
within
the
1962­
1977
timeframe
would
be
BART­
eligible.

Response:

We
agree
generally
with
the
comments
and
are
retaining
the
approach
from
proposal.

2.2.5.2
Major
Modifications
Already
Subject
to
RACT/
BACT/
LAER
Should
be
Exempt
Comment:
Thirteen
commenters
said
that
major
modifications
that
occurred
after
1977
are
already
subject
to
RACT/
BACT/
LAER
and
should
be
considered
new
sources
and,
therefore,
not
BARTeligible
(
28­
264,
28­
273,
28­
275,
28­
304,
28­
342,
28­
343,
28­
362,
28­
363,
28­
388,
28­
403,
28­
404,
28­
779,
28­
7003)
for
several
reasons:

°
These
facilities
controls
would
already
meet
BART
requirements
(
28­
273,
28­
403)
°
It
would
just
be
a
paperwork
exercise
with
little
environmental
benefit
(
28­
264)
°
State
resources
are
better
spent
on
focusing
on
greater
sources
of
emission
reduction
(
28­
779)
°
Any
source
that
undergoes
NSR
is
by
definition
a
"
new
source"
(
28­
275)
°
EPA
should
better
integrate
and
coordinate
BART
control
requirements
with
other
controls
required
by
the
Act
(
28­
304)
°
The
fact
that
the
"
BART
requirements
in
the
Act
do
not
appear
to
provide
any
exemption
for
sources
which
were
modified
since
1977"
does
not
(
even
if
correct)
over
ride
the
applicability
of
the
PSD
program
and
the
policy
behind
the
August
7,
1977
part
of
the
age
criterion
(
28­
342).

Response:

We
disagree
that
modifications
after
1977
should
change
an
emissions'
unit
date
of
construction
for
purposes
of
BART
applicability.
The
commenter's
suggestion
that
such
sources
41
already
meet
BART
requirements
may
be
accurate,
but
does
not
provide
a
basis
for
exempting
the
source
from
review.
As
we
note
in
the
guideline,
the
review
process
will
take
into
account
the
controls
already
in
place
and
the
State
may
find
that
these
controls
are
consistent
with
BART,
or
they
may
find
that
the
source
may
or
may
not
have
been
controlled
to
an
adequate
extent
based
on
its
contribution
to
regional
haze.

2.3
Compare
Potential
Emissions
to
the
250
tpy
Cutoff
2.3.1
List
of
Visibility
Impairing
Pollutants
2.3.1.1
Exclusion
of
Ammonia
on
the
List
of
Visibility
Impairing
Pollutants
Comment:
Nineteen
commenters
agreed
that
ammonia
should
be
excluded
from
the
list
of
visibility
impairing
pollutants
(
0169,
0170,
0189,
0195,
0201,
0213,
0214,
0228,
0231,
0238,
0241,
0243,
0254,
0268,
0275,
0306,
0307,
0319,
0321,
28­
304).
Commenters
gave
several
reasons:

°
Ammonia
has
not
been
shown
to
be
a
significant
threat
to
visibility
in
Class
I
areas
(
0169)
°
EPA
has
provided
no
basis
for
its
tentative
determination
that
ammonia
should
be
regulated
as
a
visibility
impairing
pollutant
(
28­
304)
°
EPA
has
not
provided
any
analysis
of
the
indirect
legal
and
regulatory
implications
of
identifying
ammonia
as
a
visibility
impairing
pollutants
such
as
becoming
subject
to
PSD
(
28­
304)
°
Including
ammonia
would
discourage
use
of
SCR
(
28­
304)
°
Few
if
any
point
sources
emit
ammonia
in
amounts
that
exceed
the
proposed
BART
thresholds
(
0189,
0254,
0275)
°
Ammonia's
visibility
impact
is
too
complex
and/
or
not
enough
is
known
about
its
role
in
regional
haze
formation
to
justify
an
automatic
BART
requirement
(
0170,
0189,
0231,
0275,
0307,
0321)
°
"
Ammonia
slip"
is
a
necessary
condition
of
the
use
of
SCR
for
the
control
of
NO
x
(
0169,
0213,
0319)
°
More
work
is
needed
to
determine
whether
ammonia
emission
reductions
are
cost
effective
in
improving
visibility
in
Class
I
areas
(
0231)
°
Ammonia
does
not
impair
visibility
directly
and
reductions
in
SO
2
and
NO
x
emissions
will
tend
to
reduce
the
concentration
of
ammonia
particles
in
the
air
(
0243,
0321)
°
The
current
inventory
suffers
from
a
tremendous
amount
of
variability
and
uncertainty
(
0243,
0321)
°
There
would
be
limited
visibility
improvement
relative
to
SO
2
controls
(
0254)
42
Commenter
0231
added
that,
a
state
would
still
be
able
to
control
a
stationary
source
of
ammonia
as
part
of
the
reasonable
progress
requirement,
if
a
technical
analysis
indicated
such
controls
were
needed
to
achieve
reasonable
progress
goals.

If
EPA
nonetheless
decides
that
ammonia
should
be
subject
to
BART
controls,
one
commenter
(
021)
said
sources
with
potential
to
emit
less
than
250
tpy
of
ammonia
should
be
not
required
to
conduct
a
BART
analysis
for
ammonia.
Given
the
uncertainty
of
the
ammonia
emissions
inventory,
the
fact
that
ammonia
does
not
impair
visibility
directly,
and
the
fact
that
it
could
discourage
the
use
of
selective
catalytic
reduction
to
reduce
NO
x
emissions,
the
commenter
does
not
believe
that
BART
analyses
should
be
required
for
any
BART­
eligible
sources,
let
alone
those
that
have
the
potential
to
emit
less
than
250
tpy.

Commenter
(
0254)
said
it
is
premature
to
require
BART
controls
for
those
emissions
until
the
visibility
benefits
of
pending
SO2,
NO
x
and
mercury
controls
have
been
observed
as
a
result
of
CAIR,
the
Acid
Rain
Program,
the
NO
x
SIP
Call,
and
the
Mercury
Rule.
Conversely,
controlling
ammonia,
VOC
and
PM
from
other,
non­
EGU
emission
sources
will
lower
PM
levels
and
improve
visibility.
To
finalize
an
effective
rule
and
streamline
the
BART­
eligibility
process
for
state
environmental
agencies,
EPA
should
include
a
determination
in
the
final
rule
that
EGUs
are
excluded
from
eligibility
for
VOC,
primary
PM10
and
primary
PM2.5,
in
addition
to
ammonia.

Comment:
Eight
commenters
argued
that
ammonia
should
be
included
on
the
list
of
pollutants
(
0179,
0186,
0200,
0218,
0219,
0232,
0259,
0302).

Commenter
0259
is
confident
that
improvements
are
being
made
to
both
NH3
inventories
and
the
understanding
of
ammonia
in
the
haze
formation
process,
which
is
complex
and
variable
across
western
environments.
Although
EPA
appears
to
be
trying
to
strike
a
balance
between
the
significance
of
ammonia
and
VOC
emissions
from
BART­
eligible
sources
and
the
administrative
burden
of
implementing
BART
for
these
emissions,
the
commenter
is
concerned
about
using
current
uncertainties
and
general
trends
in
the
state
of
science
to
make
such
broad
relaxations
when
these
pollutants
are
known
to
contribute
to
haze.

One
commenter
(
0232)
suggested
that
a
de
minimis
threshold
similar
to
those
proposed
for
other
visibility
impairing
pollutants
may
be
appropriate
for
ammonia
to
address
concerns
that
certain
large
ammonia
emission
sources
may
have
significant
potential
contributions
to
visibility
degradation
but
would
be
exempt
from
installing
control
equipment
if
ammonia
is
not
included.

Other
commenters
(
0219,
0302)
described
the
benefits
and
need
for
a
comprehensive
approach,
especially
in
order
to
adequately
reduce
PM2.5.
Commenter
0219
added
that
ammonia
emissions
are
increasing,
deposition
of
ammonia
is
increasing,
and
ecosystem
damages
due
to
excess
nitrogen
are
already
being
identified.
Continued
increases
in
ammonia
emissions
will
continue
to
exacerbate
damages
resulting
from
excess
nitrogen
deposition.
43
Commenters
(
0218,
0302)
said
a
body
of
scientific
evidence
irrefutably
demonstrates
that
ammonia
is
a
known
precursor
to
visibility
impairing
pollution.
Commenter
0218
said
EPA's
proposed
omission
of
ammonia
directly
contravenes
Section
169A's
mandate
to
regulate
"
any
air
pollutant
which
may
reasonably
be
anticipated
to
cause
or
contribute
to
any
impairment
of
visibility"
and
the
Clean
Air
Act's
expansive
definition
of
"
air
pollutant"
under
section
302
of
the
statute.
The
commenter
referenced
the
National
Park
Service
scientist
John
Vimont's
presentation
at
the
July
28­
29,
2003
WRAP
meeting
on
NO
x
Issues
in
the
West,
available
on­
line
at
http://
www.
wrapair.
org/
forums/
ioc/
meetings/
030728/
index.
html.
The
commenter
disagreed
with
EPA's
assertion
that
reduction
of
SO
2
and
NO
x
through
BART
regulation
reduces
the
two
primary
compounds
with
which
ammonia
combines
to
form
visibility­
impairing
aerosols.
Instead,
the
BART
regulations,
even
if
vigorously
implemented,
will
only
reduce
by
a
small
fraction
the
total
SO
2
and
NO
x
concentrations
in
the
atmosphere
because
BART
only
applies
to
a
subset
of
major
stationary
sources
constructed
between
1962
and
1977,
and
56
percent
of
NO
x
emissions
are
from
transportation
sources,
which
are
not
subject
to
the
BART
requirements.
Therefore,
significant
amounts
of
SO
2
and
NO
x
from
other
sources
will
remain
and
be
available
to
form
visibility­
impairing
nitrate
aerosols
with
ammonia.

Commenter
0179
said
the
concern
about
discouraging
use
of
SCRs
should
be
remedied
by
allowing
the
facility
to
make
a
determination
showing
that
an
increase
in
ammonia
is
due
to
decreasing
NO
x
emissions,
rather
than
excluding
all
sources
of
ammonia
from
BART
eligibility
Comment:.
Commenters
(
0186,
0189,
0201,
0223,
0230,
0268,
0275,
0304,
0306,
0323)
said
EPA
should
encourage
and/
or
allow
States
to
consider
ammonia
emissions
in
visibility
protection
plans
and
continue
their
efforts
to
characterize
and
understand
the
role
of
ammonia
in
their
regions.
Commenter
0186
said
this
is
especially
an
issue
in
areas
that
are
ammonia­
limited,
and
secondary
aerosol
formation
may
not
be
significantly
reduced
by
traditional
precursor
emission
reductions.
In
ammonia­
limited
regions,
ammonia
reductions
may
be
the
most
cost­
effective
control
strategy
and
may
achieve
greater
visibility
improvements
than
other
precursor
reductions.

One
commenter
(
28­
7008)
urged
EPA
to
limit
ammonia
to
the
BART
guidelines
at
this
time.
If
ammonia
becomes
a
regulated
pollutant
under
the
title
V
program,
permitting
authorities
may
be
required
to
evaluate
operating
permits
for
thousands
of
ammonia
sources
not
previously
included
in
the
title
V
program,
including
animal
feeding
operations
and
wholesale/
retail
fertilizer
distribution
facilities.

Response:

In
regard
to
ammonia,
we
believe
there
is
sufficient
uncertainty
about
emission
inventories
and
about
the
potential
efficacy
of
control
measures
from
location
to
location
such
that
the
most
appropriate
approach
for
States
to
take
is
a
case­
by­
case
approach.
There
are
ample
scientific
data
illustrating
that
ammonia
in
the
atmosphere
can
be
a
precursor
to
the
formation
of
particles
2
See
Fine
particles:
Overview
of
Atmospheric
Chemistry,
Sources
of
Emissions,
and
Ambient
Monitoring
Data,
Memorandum
to
Docket
OAR
2002­
0076,
April
1,
2005.

44
such
as
ammonium
sulfate
and
ammonium
nitrate;
2
however,
it
is
less
clear
whether
a
reduction
in
ammonia
emissions
in
a
given
location
would
result
in
a
reduction
in
particles
in
the
atmosphere
and
a
concomitant
improvement
in
visibility.
In
other
words,
the
question
of
whether
ammonia
does
contribute
to
visibility
impairment
in
a
specific
instance
can
be
a
difficult
one.

It
may
be
that
States
will
not
be
faced
often
with
the
question
of
addressing
ammonia
in
making
BART
determinations.
As
noted
above,
States
are
required
to
make
BART
determinations
only
for
stationary
sources
that
fall
within
certain
industrial
categories.
The
types
of
sources
subject
to
the
BART
provisions
are
not
typically
significant
emitters
of
ammonia.
Because
of
this,
it
is
unlikely
that
including
ammonia
on
the
list
of
visibility­
impairing
pollutants
in
the
BART
guidelines
would
have
much
impact
on
the
States'
determinations
of
whether
a
source
is
BART­
eligible.
Thus,
while
we
believe
ammonia
can
contribute
to
visibility
impairment,
the
decision
whether
to
consider
ammonia
as
a
visibility­
impairing
pollutant
in
a
specific
case
will
be
left
up
to
the
State
to
make
a
reasoned
determination.

2.3.1.2
VOCs
as
Visibility
Impairing
Pollutants
Comment:
Five
commenters
said
VOCs
should
not
be
treated
differently
in
urban
and
rural
areas
(
0179,
0221,
0227,
0231,
0304).
Commenter
0231
said
differentiation
between
urban
and
rural
areas
for
VOC
control
purposes
assumes
there
is
more
known
about
the
chemistry
than
the
current
science
supports.
Commenter
0304
noted
that
while
distance
from
a
Class
I
area
certainly
factors
into
visibility
impact,
in
general
the
impact
a
source
has
does
not
depend
on
whether
it
is
located
in
an
urban
versus
rural
setting.

Commenter
0221
described
the
important
role
VOCs
play
in
organic
aerosol
formation,
which
are
a
major
constituent
of
visibility
reducing
aerosols.
The
commenter
said
EPA
must
include
the
contribution
VOC
emissions
in
implementing
the
BART
requirement.
Sources
that
are
emitting
more
than
one
of
the
pollutants
of
concern
(
i.
e.,
SO
2,
NO
x,
particulates,
and
VOC),
need
to
aggregate
the
emissions
of
all
of
these
pollutants
in
determining
BART­
eligibility
since
all
of
the
pollutants
are
significant
precursors
to
visibility
degradation.

Instead
of
assuming
that
NO
x
are
not
present,
EPA
should
require
state
agencies
to
demonstrate
that
there
are
not
significant
sources
of
NO
x
in
these
rural
areas
before
allowing
them
to
determine
that
BART
would
be
no
controls
for
VOCs
(
0179).

Comment:
Nine
commenters
said
more
focus
should
be
placed
on
controlling
VOCs
in
urban
rather
than
rural
areas
(
0186,
0189,
0207,
0213,
0223,
0228,
0254,
0276,
0307).
Commenters
agreed
with
the
approach
in
allowing
states
the
discretion
to
handle
VOC
emissions
in
rural
areas
because
45
anthropogenic
VOC
emissions
are
typically
dwarfed
by
biogenic
and/
or
nonpoint
VOC
emissions
(
0189,
0207,
0213).
Commenter
0186
said
rural
stationary
sources
subject
to
BART
with
VOC
emissions
of
greater
than
250
tons
per
year
should
undergo
a
BART
determination.
For
smaller
VOC
sources,
states
should
be
allowed
to
determine
that
no
BART
controls
are
required.

Comment:
Three
commenters
(
0228,
0254,
0276)
said
that
EPA
should
exclude
VOCs
for
power
plants
because
emissions
are
very
low,
reductions
would
be
minimal
and
costly,
and
would
provide
insignificant
visibility
improvements.
Commenter
0254
said
that
additional
primary
particulate
matter
reductions
should
also
be
excluded
for
the
same
reasons.
Commenter
0276
said
that
the
NO
x
SIP
call,
CAIR,
and
new
EGU
mercury
controls
will
result
in
major
SO
2
reductions
by
2010,
greatly
reducing
regional
haze
in
Class
I
areas.
Any
remaining
regional
haze
will
likely
be
dominated
by
organic
compounds.
Therefore,
these
BART
regulations
should
shift
emphasis
to
the
control
of
VOC
emissions
from
major
BART­
eligible
manufacturing
sources.

Comment:
One
commenter
(
0307)
urged
EPA
to
exclude
VOCs
from
the
required
list
of
BART
pollutants
because
automatic
inclusion
as
BART
pollutants
is
not
justified
based
on
current
knowledge.
States
may
include
specific
VOC
species,
or
all
VOCs,
in
their
visibility
protection
plans
if
they
believe
such
inclusion
is
justified.
Other
commenters
(
0223,
0230)
said
that,
like
ammonia,
states
should
be
encouraged
to
continue
their
investigations
of
these
pollutants
and
where
appropriate
include
controls
on
those
pollutants
in
their
regional
haze
SIPs.

Commenter
0302
said
VOC
should
not
be
included
as
a
pollutant
to
be
controlled
by
BART.
The
state
should
be
given
the
flexibility
of
determining
the
needs
to
meet
health­
based
standards
first
and
subsequently
determine
if
further
controls
are
needed
to
meet
regional
haze
regulations.
If
a
control
level
is
included,
states
should
be
given
discretion
in
determining
where
and
how
to
apply
VOC
controls
regardless
of
the
locale.

Comment:
One
commenter
(
0201)
said
EPA
needs
to
further
clarify
the
reasoning
behind
treatment
of
VOCs
as
either
visibility
impairing
or
not
visibility
impairing
in
urban
or
rural
settings.
To
assist
states
in
completing
BART
analyses,
EPA
should
describe
the
criteria
to
be
used
for
determining
whether
a
source
is
in
a
rural
or
urban
area
as
it
relates
to
BART
(
0201,
0304).

Response:

Organic
compounds
can
be
categorized
according
to
their
varying
degrees
of
volatility:
highly
reactive,
volatile
compounds
with
six
or
fewer
carbon
atoms
which
indirectly
contribute
to
PM
formation
through
the
formation
of
oxidizing
compounds
such
as
the
hydroxyl
radical
and
ozone;
semivolatile
compounds
with
between
seven
and
24
carbon
atoms
which
can
exist
in
particle
form
and
can
readily
be
oxidized
to
form
other
low
volatility
compounds;
and
high
molecular
weight
organic
compounds
 
those
with
25
carbon
atoms
or
more
and
low
vapor
3
Ibid.

4
Ibid.

46
pressure
­­
which
are
emitted
directly
as
primary
organic
particles
and
exist
primarily
in
the
condensed
phase
at
ambient
temperatures.
The
latter
organic
compounds
are
considered
to
be
primary
PM2.5
emissions
and
not
VOCs
for
BART
purposes.
Current
scientific
and
technical
information
clearly
shows
that
carbonaceous
material
is
a
significant
fraction
of
total
PM
2.5
mass
in
most
areas,
and
that
certain
aromatic
VOC
emissions
such
as
toluene,
xylene,
and
trimethyl­
benzene
are
precursors
to
the
formation
of
secondary
organic
aerosol.
3
However,
while
significant
progress
has
been
made
in
understanding
the
role
of
VOCs
in
the
formation
of
organic
PM,
this
relationship
remains
complex,
and
issues
such
as
the
relative
importance
of
biogenic
versus
anthropogenic
emissions
remain
unresolved.
Therefore
we
believe
that
the
best
approach
for
States
to
follow
in
considering
whether
VOC
emissions
are
precursors
to
PM
2.5
formation
is
a
case­
by­
case
approach.
States
should
consider,
in
particular,
whether
a
source's
VOC
emissions
are
those
higher­
carbon
VOCs
that
are
more
likely
to
form
secondary
organic
aerosols.
In
addition,
given
the
variable
contribution
of
a
given
amount
of
VOC
emissions
to
PM
2.5
formation,
States
may
also
wish
to
exercise
discetion
in
considering
only
relatively
larger
VOC
sources
to
be
BART­
eligible.
After
careful
consideration
of
the
comments,
we
agree
with
commenters
who
assert
that
EPA
should
not
suggest
a
general
distinction
between
the
relative
contributions
of
urban
and
rural
VOC
emissions
to
particle
formation.
The
state
of
knowledge
in
this
area
is
complex
and
rapidly
evolving.
Monitoring
data
in
the
East4
suggest
that
there
may
be
a
greater
contribution
to
particle
formation
in
urban
areas
from
VOCs
as
compared
to
rural
areas,
but
we
recognize
that
further
research
is
needed
to
better
determine
the
extent
of
the
contribution
of
specific
VOC
compounds
to
organic
PM
mass.
We
do
not
agree,
however,
with
commenters
who
make
the
blanket
assertion
that
rural
VOCs
are
not
a
significant
contributor
to
particle
formation,
as
it
is
possible
that
in
specific
areas,
such
as
where
NOx
emissions
are
high,
rural
anthropogenic
VOCs
could
potentially
play
a
significant
role.

2.3.1.3
PM
as
a
Visibility
Impairing
Pollutant
Comment:
Two
commenters
(
0199,
0254)
said
EPA
should
remove
PM
from
the
list
of
pollutants
because
additional
reductions
will
not
provide
meaningful
improvements
in
regional
haze
(
from
power
plants
at
least).
Commenter
0254
said
for
power
plants,
additional
primary
particulate
matter
reductions
and
new
VOC
reductions
likely
will
provide
insignificant
visibility
benefits
and
their
cost
effectiveness
in
producing
visibility
benefits
would
be
exceedingly
low.
Therefore,
these
emissions
should
not
be
considered
under
BART
for
EGUs.
Moreover,
it
is
premature
to
require
BART
controls
for
those
emissions
until
the
visibility
benefits
of
pending
SO
2,
NO
x
and
mercury
controls
have
been
observed
as
a
result
of
CAIR,
the
Acid
Rain
Program,
the
NO
x
SIP
Call,
and
the
Mercury
Rule.
Conversely,
controlling
ammonia,
VOC
and
PM
from
other,
non­
EGU
emission
sources
will
lower
PM
levels
and
improve
visibility.
To
finalize
an
effective
rule
and
5
See
Fine
particles:
Overview
of
Atmospheric
Chemistry,
Sources
of
Emissions,
and
Ambient
Monitoring
Data,
Memorandum
to
Docket
OAR
2002­
0076,
April
1,
2005.

6
These
methods
are
described
at
the
following
website:
http://
vista.
cira.
colostate.
edu/
improve/
Tools/
ReconBext/
reconBext.
htm.

47
streamline
the
BART­
eligibility
process
for
state
environmental
agencies,
EPA
should
include
a
determination
in
the
final
rule
that
EGUs
are
excluded
from
eligibility
for
VOC,
primary
PM10
and
primary
PM2.5,
in
addition
to
ammonia.

One
commenter
(
28­
243)
suggested
that
ground
level
or
near­
ground
level
emission
sources
of
PM10
could
be
excluded
from
a
BART
analysis.

Commenters
(
28­
275,
28­
333)
said
scientific
data
cited
by
EPA
shows
that
PM
coarse
(
PM10
­
PM2.5)
has
fundamentally
different
visibility
impacts
than
PM2.5,
their
interaction
with
light
and
other
elements
in
the
atmosphere
is
fundamentally
different,
and
PM­
coarse
has
little
impact
on
visibility
and
is
not
transported
long
distances.
BART
applicability
should
be
based
on
emissions
of
PM2.5
only
and
PM10
emissions
should
be
excluded.

Response:

The
guideline
continues
to
recommend
that
States
use
PM
10
as
the
indicator
for
particulate
matter
for
the
applicability
step,
i.
e.
in
determining
whether
a
source
is
BART­
eligible.
We
disagree
with
comments
that
the
coarse
fraction
of
PM
does
not
contribute
to
visibility
impairment;
while
we
agree
that
the
coarse
fraction
is
less
efficient
at
light
scattering
than
fine
particles,
there
is
ample
evidence
that
the
coarse
fraction
does
contribute
to
visibility
impairment.
5
For
example,
standard
methods
for
calculating
reconstructed
light
extinction
routinely
include
a
calculation
for
the
contribution
to
light
extinction
from
the
coarse
fraction,
an
implicit
recognition
that
these
particles
contribute
measurably
to
visibility
impairment.
6
Thus,
we
believe
that
States
will
need
to
take
the
coarse
fraction
of
particulate
matter
into
account
in
determining
whether
a
source
meets
the
threshold
for
BART
applicability.

However,
because
long­
range
transport
of
fine
particles
is
of
particular
concern
in
the
formation
of
regional
haze,
we
also
believe
that
it
is
very
important
to
estimate
the
PM
2.5
fraction
of
direct
particulate
emissions
as
correctly
as
possible.
In
addition,
we
believe
that
air
quality
modeling
results
will
be
more
meaningful
if
the
inputs
account
for
the
relative
particle
size
of
directly
emitted
particulate
matter
(
e.
g.
PM
10
vs.
PM
2.5).

The
acceptability
of
existing
stationary
source
test
methods
for
regional
haze
SIPs
depends
upon
what
is
measured
under
a
State's
current
test
methods
for
particulate
matter.
A
State's
current
test
methods
for
measuring
particulate
matter
emissions
from
stationary
sources
may
not
be
acceptable
for
regional
haze
SIP
purposes
if
these
methods
ignore
the
condensible
fraction
of
particulate
matter.
More
specifically,
for
purposed
of
makig
a
BART
determination,
a
7
Fine
particles:
Overview
of
Source
Testing
Approaches,
Memorandum
to
Docket
OAR
2002­
0076,
April
1,
2005.

48
test
method
that
ignores
a
significant
fraction
of
emissions
may
artificially
limit
the
control
measures
considered
in
the
BART
analysis.
Therefore
we
believe
States
should
revise
any
such
stationary
source
test
methods
to
ensure
that
you
account
for
the
condensible
fraction
of
particulate
emissions.
(
See
the
source
testing
technical
support
document
(
TSD)
in
the
docket
for
this
rule,
which
discusses
test
methods
for
particulate
matter
in
more
detail.
7)

2.3.1.4
General
Comments
on
the
List
of
Visibility
Impairing
Pollutants
Comment:
Three
commenters
said
there
is
no
explanation
of
scientific
basis
for
list
of
pollutants
and
the
presumption
that
reductions
in
NO
x,
VOC,
PM,
and
ammonia
will
aid
in
visibility
improvement
(
28­
243,
28­
266,
28­
346).
Commenters
(
28­
243,
28­
266)
said
further
analysis
is
needed
to
determine
the
impacts
of
improving
visibility.
Commenter
28­
346
said
control
levels
for
NO
x,
PM,
VOCs
and
ammonia
are
not
discussed,
but
the
presumption
that
these
controls
are
always
beneficial
to
visibility
in
Class
I
areas
is
unsupported.
Such
controls
would
either
provide
no
visibility
benefits
or
their
cost­
effectiveness
would
be
low.
Furthermore,
it
is
premature
to
require
BART
controls
for
these
pollutants
before
the
impact
of
SO
2
is
observed.

One
commenter
(
28­
390)
said
the
BART
analysis
for
steam
electric
plants
should
be
limited
to
SO
2
controls.
The
emission
reductions
that
result
from
the
application
of
controls
to
BART­
eligible
sources
should
be
intended
to
provide
the
necessary
reasonable
progress
in
visibility
improvement
for
the
initial
10­
year
regional
haze
planning
period.
Thus,
it
seems
reasonable
that
the
SO
2
reductions
associated
with
the
implementation
of
BART
should
be
sufficient
to
demonstrate
reasonable
progress
for
the
first
planning
period.
Consideration
of
additional
emission
reductions
for
ammonia,
PM,
NO
x
and
VOCs
from
steam
electric
plant
BART­
eligible
sources
should
be
postponed
until
the
second
10­
year
planning
period.

Another
commenter
(
28­
302)
said
nitrate
may
play
a
far
more
significant
role
in
visibility
impairment
if
SO
2
controls
are
successful
at
substantially
reducing
sulfate
precursor
species,
leaving
ammonia
available
to
form
ammonium
nitrate.
It
is
imperative
that
controls
for
SO
2
and
NO
x
be
implemented
simultaneously.
Commenter
28­
358
said
the
problem
of
regional
haze
is
caused
primarily
by
sulfates,
but
other
pollutants
including
nitrates,
organics,
light
absorbing
carbon
and
coarse
particulate
matter
also
play
a
significant
role
in
visibility
impairment.
It
was
suggested
during
the
public
hearings
on
this
BART
proposal
that
controlling
pollutants
other
than
sulfates
would
provide
little
or
no
visibility
improvements,
or
that
these
controls
would
not
prove
cost­
effective.
This
groundless
claim
is
refuted
by
an
overwhelming
body
of
technical
evidence.
Two
commenters
(
0221,
0235)
said,
with
respect
to
the
re­
proposed
BART
guidelines
themselves,
EPA
must
strengthen
them
to
consider
all
visibility­
impairing
contaminants.
49
Response:

The
final
BART
guidelines
and
supporting
information
provide
evidence
that
NOx,
PM,
VOC,
and
ammonia
can
impair
visibility.
See
especially
Fine
particles:
Overview
of
Atmospheric
Chemistry,
Sources
of
Emissions,
and
Ambient
Monitoring
Data,
Memorandum
to
Docket
OAR
2002­
0076,
April
1,
2005,
as
well
as
discussion
of
individual
pollutants
above
and
in
the
preamble.
We
also
disagree
that
the
available
evidence
supports
a
contention
that
BART
control
should
focus
solely
on
SO2
for
steam
electric
plants.
We
agree
with
the
commenter
who
noted
that
nitrates
may
be
increasingly
important
to
particle
formation
as
sulfate
emissions
are
reduced.
This
is
discussed
more
fully
in
the
memorandum
to
the
docket
cited
above.

Comment:
One
commenter
(
0259)
is
concerned
about
any
unintended
consequences
from
such
relaxations
in
the
list,
such
as
the
possibility
that
the
relaxations
might
limit
the
ability,
if
not
authority,
of
states
and
tribes
to
address
ammonia
emissions
through
other
regulatory
provisions,
such
as
reasonably
attributable
BART
and
reasonable
progress
from
non­
BART
sources.
The
commenter
recommends
that
EPA
consider
an
alternative
approach
to
balancing
administrative
and
compliance
burdens
against
the
likely
significance
of
various
pollutants
(
including
PM10)
from
BART­
eligible
sources.
Such
an
approach
should
be
scientifically­
sound,
consistent
across
all
pollutants,
and
not
create
a
poor
precedent
or
exemption
status
which
may
limit
state
discretion
as
more
is
learned
about
the
causes
of
haze
and
its
control
options.
If
the
available
science
and
technical
analyses
demonstrate
that
emissions
of
a
particular
pollutant
from
BARTeligible
sources
in
a
state
or
region
may
not
be
reasonably
anticipated
to
cause
or
contribute
to
regional
haze,
then
no
further
BART
analysis
for
this
pollutant
would
be
necessary.
Otherwise,
the
state
or
tribe
would
implement
regional
haze
BART,
an
alternative
program,
or
possibly
an
alternative
program
for
another
pollutant
(
e.
g.,
SO
2,
NO
x,
or
ammonia)
which
compensates
for
the
"
exemption"
in
an
equitable
way.
Since
BART­
eligible
sources
appear
to
emit
(
possibly
much)
less
than
5
percent
of
the
anthropogenic
PM10
inventory,
the
commenter
would
be
interested
in
applying
this
approach
to
PM10
as
well
as
ammonia
and
VOC.
The
commenter
would
also
be
willing
to
work
with
EPA
to
refine
such
an
approach.
One
commenter
(
28­
266)
asked
if
the
list
of
visibility
impairing
pollutants
is
fixed
or
if
states
have
the
option
to
add
other
pollutants.

Response:

Given
the
current
state
of
information
on
ammonia
and
VOC
as
discussed
above,
we
believe
we
have
fashioned
a
reasonable
approach
to
regulating
emissions
of
these
pollutants
for
the
BART
program.
States
have
considerable
flexibility
to
take
individual
or
regional
circumstances
into
account
in
determining
whether
these
pollutants
impair
visibility
in
a
region,
and
whether
as
a
result
BART
limits
or
other
measures
addressing
these
pollutants
are
appropriate.
For
PM10,
as
discussed
above,
we
believe
our
recommended
approach
will
aid
the
States
in
determining
how
best
to
conduct
the
BART
review,
including
assessment
of
visibility
impacts
from
these
pollutants.
See
also
Fine
particles:
Overview
of
Source
Testing
Approaches,
50
Memorandum
to
Docket
OAR
2002­
0076,
April
1,
2005,
and
Fine
particles:
Overview
of
Atmospheric
Chemistry,
Sources
of
Emissions,
and
Ambient
Monitoring
Data,
Memorandum
to
Docket
OAR
2002­
0076,
April
1,
2005.

2.3.2
Defining
"
Stationary
Source"

2.3.2.1
Facilities
Not
Part
of
a
BART
Source
Category
are
not
BART­
eligible
Comment:
Five
commenters
agreed
with
EPA's
proposed
treatment
of
support
facilities
and
conclusion
that
support
facilities
should
not
be
included
in
the
BART­
eligible
source
(
0170,
0189,
0237,
0275,
0307).
Commenter
0189
added
that
EPA's
proposal
is
ambiguous
because
while
the
preamble
correctly
refers
to
"
enforceable
operational
limitations",
the
draft
guidelines
refer,
incorrectly,
to
"
federally
enforceable
permit
limits."

Commenter
28­
748
said
the
"
support
facility
concept"
should
not
be
included
in
the
guidelines
because
Congress
did
not
intend
that
BART
applicability
be
determined
by
aggregation
on
an
emissions
unit­
by­
emissions
unit
basis.
Second,
the
"
support
facility
concept"
appears
nowhere
in
EPA's
regulations
but
is
instead
merely
a
collection
of
informal
interpretations
that
EPA
has
sought
to
engraft
onto
the
PSD
"
source"
definition.
Third,
even
if
the
"
support
facility
concept"
were
an
appropriate
part
of
the
BART­
eligibility
scheme,
the
proposal
fails
to
state
that
the
emissions
from
a
support
facility
could
be
included
only
if
the
support
facility
itself
was
placed
in
operation
between
1962
and
1977.
Fourth,
the
support
facility
concept
would
be
unworkable
in
practice.
The
problems
involved
in
attempting
to
accurately
allocate
as
of
the
time
of
construction
of
the
unit
to
be
"
supported"
the
support
facility's
emissions
among
all
the
different
emissions
units
being
"
supported"
would
likely
be
monumental
or
even
impossible.

Commenter
28­
349
added
that
the
proposal
fails
to
state
that
the
emissions
from
a
support
facility
could
be
included
only
if
the
support
facility
itself
was
placed
in
operation
between
1962
and
1977.

Response:

We
agree
that
emission
units
at
a
plant,
even
if
they
are
a
"
support
facility"
for
purposes
of
other
programs,
should
not
be
considered
for
BART­
eligibility
unless
they
are
in
one
of
the
listed
26
source
categories,
and
were
themselves
put
into
place
within
the
1962­
1977
time
period.
We
believe
this
represents
the
best
reading
of
the
statutory
language
in
section
169A
of
the
CAA.

2.3.2.2
Grouping
Individual
Units
Comment:
51
Eight
commenters
said
small
emissions
units
under
a
support
facility
concept
should
not
be
included
in
the
BART
determination
(
28­
242,
28­
244,
28­
345,
28­
349,
28­
355,
28­
357,
28­
364,
28­
7007)
for
several
reasons:

°
It
would
be
unnecessarily
time
consuming
(
28­
243
°
Sources
can't
accept
operating
limits
on
back­
up
units
and
it
would
be
unreasonable
to
subject
them
to
the
same
retrofit
requirements
as
base
load
units
(
28­
244)
°
Congress
did
not
intend
the
BART
applicability
be
determined
by
aggregation
on
an
emissions­
unit­
by­
emissions­
unit
basis
and/
or
is
unlawful
(
28­
345,
28­
349,
28­
357,
28­
364)
°
The
concept
is
unworkable
in
practice
(
28­
349)
°
BART
analysis
should
only
be
conducted
on
the
main
electric
generating
boilers
since
these
units
usually
make
up
more
than
95%
of
the
emissions
from
a
power
plant
(
28­
7007).

Comment:
Three
commenters
(
0169,
0189,
0237)
opposed
aggregation
as
exceeding
Clean
Air
Act
requirements
by
extending
BART
from
major
stationary
sources
to
emission
units.
Commenter
0237
described
other
regulatory
precedents
related
to
the
definitions
and
boilers
and
concluded
that
EPA
should
narrow
the
scope
of
the
BART
eligibility
criteria
to
equipment
that
began
operation
after
1962
in
the
distinct
industrial
source
categories
and
withdraw
step
4
of
the
BART
guidelines
that
commands
aggregation
of
emissions
from
different
categories.
The
commenter
added
that
the
proposed
guideline
erroneously
assumes
that
a
boiler
is
part
of
one
of
the
other
26
categories.
If
that
were
the
case,
there
would
have
been
no
category
for
fossil
fuel
fired
boilers.
Equally
true,
the
retrofit
control
for
a
boiler
would
be
different
than
a
BART
control
for
the
process
with
which
it
is
associated.
Therefore,
EPA's
rationale
for
aggregating
different
categories
is
not
reasonably
related
to
the
purpose
for
applying
BART
controls
to
"
major
stationary
sources"
in
the
listed
categories.

Response:

We
believe
that
the
suggested
approach
in
the
BART
guidelines
represents
the
best
reading
of
the
CAA
and
the
regulations.
As
explained
above,
a
State
may
reasonably
conclude
that
an
industrial
boiler
smaller
than
250
million
Btu/
hour
heat
input
that
provides
only
heat
and
power
to
a
facility
is
not
integral
to
the
process
and
should
not
be
included
in
the
BART­
eligible
source.
Where
the
industrial
boiler
is
integral
to
the
source,
it
is
reasonable
for
the
State
to
consider
it
as
part
of
the
BART­
eligible
source.
In
such
a
situation,
the
State
has
ample
opportunity
to
make
reasoned
judgements
about
the
level
of
control
to
require,
if
any,
on
particular
emission
units.

2.3.3
Use
of
Potential
Emissions
Comment:
52
Commenters
(
28­
345,
28­
349,
28­
364,
28­
748)
said
that
PTE
should
be
based
on
a
realistic
assessment
of
potential
emissions
when
the
source
was
put
in
operation,
rather
than
an
unrealistic
approach
based
on
today's
air
pollution
regulatory
system.
The
preamble
does
not
explain
why
the
"
current"
PTE
is
to
be
used
rather
than
the
PTE
of
the
source
when
it
was
placed
in
operation.
One
commenter
(
0237)
said
EPA's
proposal
should
address
whether
these
defined
types
of
facilities
were
major
stationary
sources
in
the
period
between
1962
and
1977.
The
commenter
added
that
it
may
be
more
reasonable
to
reconstruct
a
source's
potential
to
emit
during
that
period
by
determining
whether
a
source
in
one
of
the
26
categories
"
actually"
has
emissions
of
more
than
250
tpy
of
a
BART
pollutant,
than
trying
to
determine
from
current
data
what
a
source's
potential
to
emit
was
between
1962
and
1977.
Commenter
0238
supported
using
a
source's
actual
emissions.
Most
EGUs
operate
well
below
their
allowable
emissions,
so
an
analysis
of
visibility
impairment
based
on
the
allowable
emissions
rate
inflates
a
source's
actual
emissions.
One
commenter
(
28­
748)
said
the
guidelines
should
provide
that
a
source
that
constituted
a
"
major
stationary
source"
when
placed
in
operation
between
1962
and
1977
that
is
no
longer
a
"
major
stationary
source"
for
any
"
visibility­
impairing
pollutant"
is
not
subject
to
BART
requirements.
The
source's
PTE
may
be
limited
by
emissions
controls,
permit
limits,
or
other
constraints
that
did
not
exist
when
the
source
was
placed
in
operation.
One
commenter
(
0237)
said
for
units
such
as
boilers,
other
practical
limitations
on
a
source's
potential
to
emit
also
must
be
considered,
many
of
which
are
impossible
to
determine
or
if
possible,
certainly
requiring
an
outlay
of
resources
that
would
tax
most
plant
operating
budgets
without
any
assurance
of
environmental
benefit.
This
last
set
of
difficulties
apparently
led
EPA
to
rely
on
the
current
potential
to
emit
of
individual
units
that
were
put
into
operation
between
1962
and
1977.
This
is
equally
in
error,
because
there
is
no
reasonable
nexus
between
a
piece
of
equipment's
emissions
in
the
BART
eligibility
time
span
and
its
emissions
now,
which
are
related
to
current
plant
production
and
limitations
on
throughput
and
upstream/
downstream
equipment.
EPA's
failure
to
support
its
proposed
current
PTE
as
a
surrogate
for
1962­
1977
PTE
is
arbitrary
and
capricious,
and
undermines
the
reasonableness
of
the
re­
proposed
guideline
Response:

CAA
§
169A(
g)(
7)
defines
a
"
major
stationary
source"
as
a
source
with
the
potential
to
emit
250
tons
or
more
any
pollutant.
We
interpret
this
as
requiring
sources
with
a
current
potential
to
emit
250
tons
per
year
as
a
major
stationary
source.
We
also
agree
that
determining
a
source's
potential
to
emit
during
1962
to
1977
could
be
difficult.
We
also
agree
that
a
source's
current
potential
emissions
are
more
relevant
to
the
question
as
to
whether
it
is
still
a
major
stationary
source.
A
source
which
no
longer
has
the
potential
to
emit
250
tons
per
year
of
a
relevant
pollutant
need
not
be
considered
a
major
stationary
source
for
purposes
of
BART.

Comment:
Commenters
(
0184,
28­
273)
agreed
that
sources
with
federally
enforceable
emission
limits
can
use
the
limit
level
as
a
cap
for
the
purposes
of
calculating
potential
to
emit.
Another
53
commenter
(
28­
255)
said
EPA
must
expedite
the
rule
making
to
determine
whether
only
federally
enforceable
limits
should
be
taken
in
to
account
in
this
as
well
as
other
programs.
This
issue
must
be
settled
as
soon
as
possible.
Commenter
28­
385
said
EPA
should
either
go
out
with
guidance
that
allows
consideration
of
other
kinds
of
enforceable
limitations,
such
as
those
established
by
state
programs,
or
at
least
remain
neutral
on
the
issue
and
expressly
seek
comment
on
the
approach
to
be
used.
Also,
EPA
references
the
1990
draft
PSD
manual
for
support,
which
has
not
been
subjected
to
public
review
and
comment.
The
manual
has
been
in
draft
for
longer
than
10
years
without
being
put
in
final.
Thus,
references
to
the
manual
should
be
removed.
Nine
commenters
said
EPA's
definition
of
PTE
limits
that
are
federally
enforceable
rather
than
practicably
enforceable
is
illegal
because
the
courts
have
disapproved
it
(
28­
244,
28­
304,
28­
342,
28­
344,
28­
345,
28­
349,
28­
364,
28­
393,
28­
748
).
Five
commenters
said
state
enforceable
limits
should
be
considered
as
well
as
Federal
limits
(
28­
266,
28­
278,
28­
312,
28­
404,
28­
770)

Response:

As
explained
in
the
2001
and
2004
proposed
guidelines,
the
regional
haze
regulations
define
"
potential
to
emit."
The
guidelines
repeat
that
regulatory
definition
and
provide
an
example
illustrating
its
application.
EPA
did
not
propose
to
change
the
definition
in
2001
or
2004,
but
merely
highlighted
the
current
definition
in
40
CFR
51.301.
Although
we
noted
in
the
2001
proposed
guidelines
that
we
expected
to
undertake
a
rulemaking
to
determine
whether
only
federally
enforceable
limitations
should
be
taken
into
account
in
the
regional
haze
program
definition,
we
have
not
yet
begun
the
process
for
such
a
rulemaking.
For
the
time
being,
we
believe
that
States
may
consider
emissions
limitations
in
State
permits,
which
are
enforceable
under
State
law,
in
determining
a
source's
"
potential
to
emit."

Comment:
One
commenter
(
0189)
noted
that
EPA's
visibility
rules
provide
for
counting
"
fugitive"
emissions
toward
the
"
major
source"
determination.
However,
any
decision
to
actually
impose
controls
on
such
emissions
must
be
reasonable
in
terms
of
the
Congressional
factors
for
determining
BART.
There
is
no
practicable
control
technique
for
many
sources
of
"
fugitive"
emissions
at
forest
products
facilities,
such
as
chip
piles,
log
piles
or
material
handling
and
conveyance
systems.
In
addition,
the
emissions
from
such
operations
consist
overwhelmingly
of
large
particulates
or
naturally­
occurring
VOCs
with
little
or
no
impact
on
visibility.
The
commenter
urged
EPA
to
make
clear
that
control
of
sources
of
fugitive
emissions
should
be
required
only
when
it
is
technically
feasible,
cost
effective,
and
produces
visibility
improvement
benefits.
Similarly,
commenter
28­
244
urged
EPA
to
clarify
that
control
of
fugitive
emissions
from
sources
where
there
is
no
practicable
control
technique
or
whose
fugitive
emissions
have
no
impact
on
visibility
is
not
required.
Commenters
(
28­
266,
28­
278,
28­
312,
28­
404)
agreed
that
fugitive
emissions
should
not
count
toward
potential
emissions,
which
is
consistent
with
the
PSD
program.
54
Response:

The
visibility
regulations
require
the
counting
of
fugitive
emissions
for
purposes
of
determining
whether
a
source
is
a
"
major
stationary
source."
While
fugitive
emissions
do
count
toward
the
"
major
source"
definition,
it
is
also
true
that
when
States
conduct
BART
determinations
they
must
take
into
account
the
five
factors
set
out
in
the
CAA,
including,
as
the
guidelines
make
clear,
consideration
of
technical
feasibility
of
controls,
cost
effectiveness,
and
expected
visibility
improvement.

2.4
Identify
the
Emission
Units
and
Pollutants
that
Constitute
the
BART­
eligible
Source
2.4.1
Requiring
Analysis
of
All
Pollutants
When
Only
One
Exceeds
the
250
typ
Threshold
Comment:
Ten
commenters
said
the
total
amount
of
haze­
causing
pollutants
from
all
units
at
facility
should
be
considered
to
meet
the
250­
ton
criteria
(
0221,
28­
247,
28­
248,
28­
260,
28­
325,
28­
334,
28­
350,
28­
358,
28­
362,
28­
399).
Commenter
28­
257
specified
the
aggregate
approach
should
be
taken
for
power
plants..

One
commenter
(
0221)
said
the
relevant
definitions
in
the
Act,
including
the
definition
of
a
major
stationary
source
as
one
that
potentially
emits
250
tpy
of
any
pollutant
or
its
precursors,
and
the
fact
the
Administrator
has
identified
NO
x,
SO
2,
VOCs,
and
PM
as
either
precursors
to
or
part
of
the
combination
of
agents
which
form
visibility­
impairing
particulate
matter,
requires
the
aggregation
of
those
pollutants
in
determining
whether
a
source
potentially
emits
250
tons
or
more
of
particulate
matter
or
its
precursors.
The
commenter
provided
an
example
showing
why
such
aggregation
is
needed
to
avoid
"
absurd"
outcomes.

Three
other
commenters
agreed
that
a
facility
is
subject
to
BART
for
all
visibility
impairing
pollutants,
even
if
a
facility's
potential
to
emit
triggers
the
250
tons
per
year
threshold
for
only
one
pollutant
(
28­
258,
28­
266,
28­
362).
Commenter
28­
262
added
that
it
is
appropriate
that
all
visibility­
impairing
pollutants
be
subject
to
BART
once
the
PTE
threshold
has
been
crossed.
It
does
not
make
sense,
however,
to
conduct
the
PTE
analysis
on
a
pollutant­
bypollutant
basis.

One
commenter
(
0227)
said
EPA
should
aggregate
all
haze­
causing
pollutants
together
when
determining
if
a
source
exceeds
the
250­
ton
per
year
threshold
as
part
of
the
BART
eligibility
analysis.

Comment:
Forty­
nine
commenters
said
that
a
facility
should
be
BART­
eligible
only
for
each
specified
pollutant
with
emissions
exceeding
250
tpy
(
0169,
0181,
0184,
0186,
0187,
0202,
0206,
0214,
0215,
0228,
0229,
0236,
0237,
0238,
0240,
0241,
0242,
0247,
0275,
0294,
28­
259,
28­
264,
28­
55
273,
28­
275,
28­
304,
28­
306,
28­
332,
28­
339,
28­
341,
28­
342,
28­
344,
28­
345,
28­
347,
28­
348,
28­
349,
28­
355,
28­
360,
28­
383,
28­
385,
28­
390,
28­
392,
28­
403,
28­
404,
28­
747,
28­
778,
28­
7001,
28­
7007,
28­
7008,
28­
7013).
Commenters
(
0181,
0215,
0228,
0238,
0240,
0241,
0242,
0247,
0294)
said
the
BART
Guidelines
should
say
that
the
visibility
impairment
analysis
must
be
done
separately
for
each
visibility­
impairing
pollutant
emitted
by
the
source,
and
that
only
those
pollutants
that
have
been
demonstrated,
individually,
to
cause
or
contribute
to
visibility
impairment
from
that
source
are
required
to
go
through
a
BART
determination.
Commenters
(
240,
28­
273,
28­
7008)
said
that
this
approach
is
consistent
with
the
PSD
regulations,
one
of
the
objectives
of
which
is
to
protect
visibility
in
National
Parks.
At
a
minimum,
the
rule
should
require
that
states
adopt
de
minimis
levels
at
or
below
PSD
significance
thresholds,
as
suggested
on
page
25193
of
the
preamble.

This
proposal
is
inconsistent
with
the
statute
(
0186,
0187,
0202,
0206,
0214,
0215,
0236,
0237,
0275,
28­
304,
28­
306,
28­
342,
28­
344,
28­
345,
28­
349,
28­
355,
28­
747).
Commenter
0186
said
the
plain
language
of
the
statute
says
that
only
"
major
stationary
sources"
are
subject
to
BART,
and
there
is
no
suggestion
that
BART
is
intended
to
be
applied
to
pollutants
for
which
the
source
would
not
be
determined
under
section
169A
to
be
a
"
major
stationary
source."
Therefore,
there
is
no
policy
or
legal
justification
for
the
proposal,
and
EPA
should
limit
BART
determinations
to
pollutants
that
exceed
250
tons
per
year.
Commenters
(
0187,
0206,
0215,
0236,
0237,
28­
342,
28­
349)
added
that
the
proposal
contradicts
Congressional
intent.
Commenters
(
0202,
0294,
28­
345)
said
the
"
in
for
one,
in
for
all"
approach,
especially
on
a
cumulative
basis
for
all
affected
sources,
is
not
consistent
with
the
individual
facility
assessment
mandate
in
the
Act.
States
should
be
required
to
assess
available
individual
facility
information
before
determining
that
all
sources
within
the
facility
are
BART
eligible.
Commenters
(
0187,
0215,
0236,
0238,
28­
349)
said
the
statutory
language
requires
a
pollutant­
specific
visibility
impairment
analysis.

Commenter
28­
348
said
this
requirement
is
inconsistent
with
other
EPA
air
programs,
which
have
threshold
or
"
significant"
emissions
amounts
for
the
"
non­
triggering"
pollutants
below
which
assessment
and
controls
are
not
necessary.

Commenter
28­
7001
said
this
policy
is
unjustified
and
far
too
inclusive
especially
when
one
considers
that
the
emission
amount
is
calculated
on
a
"
potential
to
emit"
basis
which
will
always
be
greater
than
and
often
far
greater
than
actual
emissions.
Once
a
sources
is
"
subject
to
BART",
further
analysis
should
only
apply
to
those
EU's
that
trigger
the
250
tpy
applicability
limit.
Commenter
28­
349
added
that,
at
the
very
least,
BART
reviews
should
not
be
conducted
unless
the
PTE
of
the
pollutant
in
question
exceeds
a
reasonable
"
significance"
threshold
Commenters
(
0206,
0214,
0187,
0236,
0238,
28­
264,
28­
339,
28­
355,
28­
383,
28­
390,
28­
392,
28­
403,
28­
747,
28­
7007,
28­
7013)
said
that
applying
BART
to
non­
major
source
pollutants
would
not
produce
a
significant
deterioration
in
visibility
and
control
would
not
be
cost
effective.
56
Comment:
Commenters
(
0241,
28­
347,
28­
778)
said
states
would
not
be
precluded
from
requiring
controls
to
meet
the
regional
haze
SIP
requirements
for
pollutants
where
the
source's
PTE
is
less
than
250
tpy
but
would
grant
states
the
appropriate
authority
to
choose
whether
to
impose
such
controls.
By
requiring
a
BART
evaluation
for
all
visibility
impairing
pollutants
over
the
PSD
threshold
levels,
EPA
is
granting
no
deference
to
the
states'
decision­
making
authority
and
would
force
states
to
evaluate
BART
even
where
there
would
be
no
clear
environmental
benefit.
This
would
waste
state
resources
and
significantly
impede
the
states'
ability
to
complete
the
BART
evaluation
for
all
BART
eligible
sources
in
a
timely
fashion
so
as
to
meet
the
SIP
submittal
deadline.
Commenter
28­
385
added
that
said
the
requirements
should
allow
states
to
focus
their
limited
resources
on
completing
BART
reviews
for
pollutants
that
potentially
impact
visibility
and
regulated
sources
should
not
be
saddled
with
unreasonable
control
and
monitoring
costs.

Response:

We
disagree
with
the
comment
that
emissions
of
different
visibility­
impairing
pollutants
must
added
together
to
determine
whether
a
source
exceeds
the
250
ton
per
year
threshold.
The
CAA,
in
section
169A(
g)(
7),
defines
a
"
major
stationary
source"
as
one
with
the
potential
to
emit
250
tons
or
more
of
"
any
pollutant."
We
disagree
with
comments
that
the
BART
analysis
is
required
only
for
those
pollutants
that
individually
exceed
the
250
ton
per
year
threshold.
Section
169A(
b)(
2)(
A)
specifically
requires
States
to
submit
SIPs
that
include
a
requirement
that
a
major
stationary
source
which,
as
determined
by
the
State
.
.
.
emits
any
air
pollutant
which
may
reasonably
be
anticipated
to
cause
or
contribute
to
any
impairment
of
visibility
in
any
[
Class
I
area],
shall
procure,
install,
and
operate
.
.
.
the
best
available
retrofit
technology,
as
determined
by
the
State
.
.
.
for
controlling
emissions
from
such
source
for
the
purpose
of
eliminating
or
reducing
any
such
impairment.
The
regional
haze
regulations
similarly
require
that
the
States
submit
a
SIP
that
contains
A
determination
of
BART
for
each
BART­
eligible
source
in
the
State
that
emits
any
air
pollutant
which
may
reasonably
be
anticipated
to
cause
or
contribute
to
any
impairment
of
visibility
in
any
mandatory
Class
I
Federal
area.
40
CFR
51.308(
e)(
1)(
ii).
Nothing
in
these
statutory
or
regulatory
requirement
suggests
that
the
BART
analysis
is
limited
to
those
pollutants
for
which
a
source
is
considered
major.
At
best,
these
provisions
can
be
read
as
requiring
a
BART
determination
only
for
those
emissions
from
a
specific
source
which
do,
in
fact,
cause
or
contribute
to
visibility
impairment
in
a
particular
Class
I
area,
or
at
least
which
could
reasonably
be
anticipated
to
do
so.
Commenters,
however,
have
not
presented
any
evidence
that
as
a
general
matter
emissions
of
less
than
250
tons
per
year
of
PM
2.5,
SO
2,
or
other
visibility­
impairing
pollutants
from
potential
BART
sources
do
not
"
cause
of
contribute
to
any
impairment
of
visibility"
in
any
of
the
Class
I
areas
covered
by
the
regional
haze
rule.
As
there
is
no
such
evidence
currently
before
us,
there
is
no
basis
to
conclude
that
the
States
are
required
to
make
BART
determinations
only
for
those
pollutants
emitted
in
excess
of
250
tons
per
year.
57
At
the
same
time,
we
agree
with
certain
commenters
that
the
CAA
does
not
require
a
BART
determination
for
any
visibility
impairing
pollutant
emitted
by
a
source,
regardless
of
the
amount.
After
reviewing
the
language
of
the
Act
and
the
comments
received,
we
have
concluded
that
our
interpretation
of
the
relevant
language
in
section
169A(
b)(
2)(
A)
of
the
Act
in
the
2004
proposed
guidelines
is
not
necessarily
the
best
reading
of
the
BART
provisions.
Section
169A(
b)(
2)(
A)
of
the
Act
can
be
read
to
require
the
States
to
make
a
determination
as
to
the
appropriate
level
of
BART
controls,
if
any,
for
emissions
of
any
visibility
impairing
pollutant
from
a
source.
Given
the
overall
context
of
this
provision,
however,
and
that
the
purpose
of
the
BART
provision
is
to
eliminate
or
reduce
visibility
impairment,
it
is
reasonable
to
read
the
statute
as
requiring
a
BART
determination
only
for
those
emissions
from
a
source
which
are
first
determined
to
contribute
to
visibility
impairment
in
a
Class
I
area.
The
interpretation
of
the
requirements
of
the
regional
haze
program
reflected
in
the
discussion
above
does
not
necessitate
costly
and
time­
consuming
analyses.
Consistent
with
the
CAA
and
the
implementing
regulations,
States
can
adopt
a
more
streamlined
approach
to
making
BART
determinations
where
appropriate.
Although
BART
determinations
should
be
based
on
the
totality
of
circumstances
in
a
given
situation,
such
as
the
distance
of
the
source
from
a
Class
I
area,
the
type
and
amount
of
pollutant
at
issue,
and
the
availability
and
cost
of
controls,
it
is
clear
that
in
some
situations,
one
or
more
factors
will
clearly
suggest
an
outcome.
Thus,
for
example,
a
State
need
not
undertake
an
exhaustive
analysis
of
a
source's
impact
on
visibility
resulting
from
relatively
minor
emissions
of
a
pollutant
where
it
is
clear
that
controls
would
be
costly
and
any
improvements
in
visibility
resulting
from
reductions
in
emissions
of
that
pollutant
would
be
negligible.
In
a
scenario,
for
example,
where
a
source
emits
thousands
of
tons
of
SO2
but
less
than
one
hundred
tons
of
NOx,
the
State
could
easily
conclude
that
requiring
expensive
controls
to
reduce
NOx
would
not
be
appropriate.
In
another
situation,
however,
inexpensive
NOx
controls
might
be
available
and
a
State
might
reasonably
conclude
that
NOx
controls
were
justified
as
a
means
to
improve
visibility
despite
the
fact
that
the
source
emits
less
than
one
hundred
tons
of
the
pollutant.
Moreover,
as
discussed
below,
we
are
revising
the
regional
haze
regulations
to
allow
the
States
to
exempt
de
minimis
emissions
of
SO
2,
NO
x,
and
PM
2.5
from
the
BART
determination
process
which
should
help
to
address
the
concerns
of
certain
commenters
associated
with
the
burden
of
a
broad
BART
analysis.

2.4.2
De
Minimis
Emission
Levels
2.4.2.1
Overall
Support
or
Disapproval
for
the
Concept
Comment:
Forty­
one
commenters
said
there
should
be
de
minimis
levels
values
for
visibility­
impairing
pollutants
(
0169,
0170,
0178,
0184,
0188,
0195,
0198,
0199,
0201,
0206,
0207,
0210,
0215,
0226,
0228,
0230,
0231,
0241,
0242,
0243,
0253,
0254,
0256,
0266,
0268,
0275,
0279,
0299,
0304,
0306,
0307,
28­
244,
28­
266,
28­
267,
28­
304,
28­
332,
28­
347,
28­
391,
28­
397,
28­
403,
28­
779)
and
that
the
states
should
have
the
flexibility
to
identify
de
minimis
levels
of
pollutants
(
0178).
Commenters
provided
several
reasons
for
their
support:
58
°
BART
eligibility
for
every
visibility­
impairing
pollutant,
regardless
of
the
magnitude
of
emissions,
is
costly
and
inefficient
(
0170,
0178,
0188,
0199,
0201,
0207,
0253,
0254,
0256,
0266,
0275,
0279,
0299,
0304,
0307)
°
The
magnitude
of
emissions
at
low
levels
will
not
cause
or
contribute
to
visibility
impairment
(
0178,
0188,0199,
0206,
0207,
0210,
0266)
°
Controlling
such
pollutants
would
not
be
likely
to
produce
any
perceptible
visibility
improvement
(
0188,
0207,
0210,
0230,
0253,
0256)
°
Creating
de
minimus
levels
would
avoid
the
enormous
burden
and
costs
associated
with
determining
BART
for
countless
low­
emitting
units
and
establishing
and
tracking
compliance
with
site­
specific
permit
terms
and
conditions
for
insignificant
environmental
benefits
(
28­
304,
28­
397,
28­
779).
°
Reduces
resource
demands
on
states
(
0210,
0230,
0231,
0256,
0266,
0279)
°
Keeps
the
focus
on
relevant
pollutants
(
0242,
0254,
0275)
°
Similar
to
EPA's
approach
to
consider
only
boilers
emitting
more
than
250
tpy
as
eligible
sources
rather
than
aggregating
all
emissions
units
at
these
sources
(
0299)

Commenter
0199
specifically
supported
use
of
de
minimis
emission
thresholds
from
auxiliary
equipment
at
EUSGUs
because
there
is
no
discernable
improvement
in
Class
I
visibility
impairment
by
retrofitting
BART
controls
on
these
minor
sources.

Comment:
Three
commenters
opposed
the
use
of
de
minimis
exemptions
(
0179,
0200,
0221).
Commenter
0179
said
EPA
is
closing
the
wrong
loophole.
Instead
of
exempting
sources
that
are
clearly
BART
eligible,
EPA
should
close
the
loophole
that
allows
sources
that
emit
a
combined
total
of
more
than
250
TPY
of
haze
forming
pollutants
to
escape
BART.

Commenter
0200
said
the
proposed
rule
makes
clear
that
single
pollutant
de
minimis
exemptions
are
allowed.
The
commenter
believes
that
all
pollutants
should
be
tested
before
an
exemption
is
given.

Commenter
0221
gave
an
extensive
discussion
explaining
why
the
proposed
de
minimis
exemption
contravenes
the
judicial
limits
on
administrative
exemptions
established
in
Alabama
Power
Co.
v.
Costle,
exceeds
EPA's
authority
as
evidenced
by
the
language,
structure,
and
purpose
of
section
169A
of
the
Act,
and
is
arbitrary
and
capricious
in
setting
the
de
minimis
ceilings
equal
to
the
PSD
applicability
levels.
The
EPA's
proposed
de
minimis
exemption
provides
no
safeguards
that
would
prevent
a
state
from
setting
de
minimis
thresholds
at
levels
which
would
eliminate
or
significantly
reduce
the
benefits
of
applying
the
BART
requirement
to
particular
emissions
at
particular
sources.
When
comparing
the
Administrator's
exemption
authority
in
section
169
to
the
expansive
definition
of
what
constitutes
a
major
stationary
source,
it
is
clear
that
the
BART
applicability
determination
is
a
very
low
threshold
which
does
not
permit
de
minimis
exemptions
unless
the
operative
statutory
exemption
requirements
are
satisfied.

Response:
8
EDF
et
al.
v.
EPA,
82
F.
3d
451,
466
(
D.
C.
Cir.
1996)
citing
Alabama
Power
v.
Costle,
636
F.
2d
323
(
D.
C.
Cir.
1979).

9
Id.

59
As
proposed
in
2004,
we
believe
that
it
is
reasonable
to
give
States
the
flexibility
to
establish
de
minimis
levels
so
as
to
allow
them
to
exempt
from
the
BART
determination
process
pollutants
emitted
at
very
low
levels
from
BART­
eligible
sources.
As
explained
by
the
D.
C.
Circuit,
"
categorical
exemptions
from
the
requirements
of
a
statute
may
be
permissible
`
as
an
exercise
of
agency
power,
inherent
in
most
statutory
schemes,
to
overlook
circumstances
that
in
context
may
fairly
be
considered
de
minimis.'"
8
The
ability
to
create
de
minimis
exemptions
from
a
statute
is
a
tool
to
be
used
in
implementing
the
legislative
design.
9
The
intent
of
Congress
in
requiring
controls
on
emissions
from
certain
major
stationary
sources
was
to
eliminate
or
reduce
any
anticipated
contribution
to
visibility
impairment
from
these
sources.
This,
as
section
169A(
b)(
2)(
A)
states,
is
the
"
purpose"
of
BART.
In
making
a
determination
as
to
the
appropriate
level
of
controls,
however,
the
States
are
required
to
take
into
account
not
only
the
visibility
benefits
resulting
from
imposing
controls
on
these
sources
but
also
the
costs
of
complying
with
the
BART
provision.
The
BART
provision
is
accordingly
designed
to
ensure
that
the
States
take
into
consideration
all
emissions
of
certain
stationary
sources
in
making
a
BART
determination,
but
also
to
provide
States
with
the
flexibility
to
include
the
costs
and
benefits
of
controlling
these
sources
in
the
calculus
of
determining
the
appropriate
level
of
BART.
Although
as
discussed
above,
it
is
not
possible
based
on
the
information
before
us
to
assume
that
emissions
of
250
tons
per
year
of
a
visibility­
impairing
pollutant
do
not
cause
or
contribute
to
visibility
impairment,
we
do
believe
it
would
be
permissible
for
States
to
create
de
minimis
levels
at
a
much
lower
level.
If
a
State
were
to
undertake
a
BART
analysis
for
emissions
of
less
than
40
tons
of
SO
2
or
NO
x
or
15
tons
of
PM
10
from
a
source,
it
is
unlikely
to
result
in
anything
but
a
trivial
improvement
in
visibility.
This
is
because
reducing
emissions
at
these
levels
would
have
little
effect
on
regional
emissions
loadings
or
visibility
impairment.
We
believe
most
States
would
be
unlikely
to
find
that
the
costs
of
controlling
a
few
tons
of
emissions
were
justified.
Because
the
overall
benefits
to
visibility
of
requiring
BART
determinations
for
emissions
of
less
than
the
de
minimis
levels
would
be
trivial,
we
are
amending
the
regional
haze
rule
to
allow
the
States
this
flexibility.
The
de
minimis
levels
we
provide
for
today
apply
on
a
plant­
wide
basis.
Applying
de
minimis
levels
on
a
unit
by
unit
basis
as
suggested
by
certain
commenters
could
exempt
hundreds
of
tons
of
emissions
of
a
visibility­
impairing
pollutant
from
BART
analysis.
In
at
least
some
of
the
twenty­
six
source
categories
covered
by
the
BART
provisions,
a
single
control
device
can
be
used
to
control
emissions
from
multiple
units.
Thus,
it
is
possible
that
while
emissions
from
each
unit
are
relatively
trivial,
the
costs
of
controlling
emissions
from
multiple
units
might
be
costeffective
in
light
of
the
BART­
eligible
source's
total
emissions
of
the
pollutant
at
issue.
States
should
consider
the
control
options
in
such
situations
and
determine
the
appropriate
approach
for
the
specific
source.
60
2.4.2.2
PSD
Significance
Levels
Comment:
Two
commenters
opposed
the
use
of
the
PSD
levels
because
they
are
not
protective
enough.
One
commenter
(
0179)
said
the
de
minimis
levels
should
be
lower
than
the
PSD
significance
levels.
The
regional
haze
rule
is
not
intended
to
prevent
significant
deterioration
of
views;
it
is
intended
to
restore
views
to
their
natural
range.
Therefore,
PSD
is,
on
its
face,
not
appropriate
for
use
in
the
BART
rule.
Because
this
provision
does
not
appear
to
have
any
legal
or
scientific
basis
EPA
should
withdraw
it
from
the
final
rule.
Commenter
0221
said
EPA
has
made
no
showing
that
the
PSD
applicability
levels
are
an
appropriate
threshold
to
further
the
congressionally­
enunciated
goal
of
visibility
restoration
in
Class
I
areas.

Commenter
0306
said
the
PSD
significance
levels
are
a
good
starting
point;
however,
the
states
should
have
flexibility
to
establish
higher
or
lower
emission
rates
based
on
state
specific
factors
such
as
the
distance
to
Class
I
areas.

Comment:
Eighteen
commenters
supported
setting
the
de
minimis
levels
equal
to
the
PSD
significance
levels
(
0178,
0181,
0195,
0210,
0213,
0214,
0223,
0228,
0229,
0230,
0231,
0238,
0259,
0266,
0268,
0271,
0304,
0314,
28­
267,
28­
304,
28­
403).
Commenter
(
28­
304)
said
this
approach
would
create
consistency
between
the
closely
related
PSD
and
Regional
Haze
programs,
which
would
simplify
and
ease
the
implementation
of
regional
haze
requirements.
Commenters
(
0181)
said
the
application
of
such
emission
thresholds
will
limit
BART
applicability
to
pollutant
emissions
that
have
the
greatest
impact
on
visibility
impairment
in
Federal
Class
I
areas
and
result
in
the
greatest
air
quality
benefit
in
those
areas.
It
is
also
important
for
flexibility
in
a
state's
ability
to
implement
BART
on
a
case­
by­
case
basis
(
0229).
Commenter
0304
said
the
de
minimis
levels
used
for
BACT
are
familiar
and
have
served
the
states
well,
so
it
is
recommended
that
the
same
levels
be
used
in
the
application
of
BART.
The
commenter
prefers
that
these
levels
be
set
nationally
so
all
states
face
the
same
de
minimis
thresholds
and
the
program
is
equitably
applied
across
state
boundaries.

Commenters
(
0223,
0230,
0231,
0259,
0268,
0271)
supported
the
de
minimis
concept,
provided
that
the
level
could
not
be
higher
than
the
PSD
applicability
levels.

Comment:
Commenters
(
0170,
0178)
said
that
significant
emission
rates
should
be
established
for
each
pollutant
(
similar
in
concept
to
PSD
significant
emission
rates),
below
which
BART
would
not
be
required
for
that
pollutant.
Commenter
28­
391
suggested
establishing
the
de
minimus
emission
rates
at
10
tpy.
Commenters
(
0170,
28­
779)
said
these
thresholds
could
be
established
by
considering
the
relative
importance
of
each
pollutant
in
the
role
of
visibility
impairment.
This
would
allow
the
states
to
focus
on
the
larger
emitters.
61
Comment:
Ten
commenters
said
that
de
minimis
levels
should
be
higher
than
the
PSD
significance
levels
(
0170,
0186,
0189,
0206,
0237,
0241,
0247,
0254,
0275,
0307)
to
reduce
the
burden
on
states
and
sources
(
0170,
0241),
as
justified
by
the
statute
and
consistent
with
Congressional
intent
(
0186,
0189,
206),
and
to
avoid
applying
the
BART
review
to
sources
that
are
unlikely
to
cause
an
impact
in
a
Class
I
area
and
not
be
cost
effective
to
control
(
0237,
0241,
0254).
Commenters
noted
that
while
such
de
minimis
sources
would
not
be
subject
to
BART,
they
would
be
included
in
the
regional
visibility
plans
(
0186,
0189,
0275,
0307)

Commenters
suggested
alternative
levels:

°
100
tpy,
which
defines
a
major
source
for
title
V
and
a
large
emitting
source
under
compliance
assurance
monitoring
regulations
(
0170)
°
250
tpy,
consistent
with
the
statute
and
Congressional
intent
(
0186,
0189,
206,
0237,
0275)
and
consistent
with
the
PSD
definition
of
"
major"
(
0247).

Comment:
Commenter
0285
asked
EPA
to
clarify
the
significance
of
these
levels
to
the
overall
BART
determination
process.
For
example,
if
the
total
potential
emissions
of
a
source
for
any
BART
pollutant
are
at
or
below
the
de
minimis
level,
should
the
source
be
evaluated
for
BART
control
for
that
pollutant?

Response:

We
are
revising
the
regional
haze
rule
to
provide
States
with
the
ability
to
establish
de
minimis
levels
up
to
the
levels
proposed
in
2004.
States
can
accordingly
exclude
from
the
BART
determination
process
potential
emissions
from
a
source
of
less
than
forty
tons
per
year
of
SO
2
or
NO
x,
or
15
tons
per
year
for
PM
10.
In
the
guidelines,
we
include
this
as
part
of
the
BART
determination
in
section
IV
of
the
guidelines.
We
believe
that
this
approach
is
the
clearest
method
for
exempting
trivial
emissions
from
the
BART
determination
process.

We
are
not
requiring
that
States
use
these
levels
as
de
minimis
levels.
These
emission
levels
represent
the
maximum
allowable
de
minimis
thresholds,
because
we
believe
that
control
of
sources
smaller
than
these
levels
is
unlikely
to
result
in
anything
but
a
trivial
improvement
in
visibility.
States
retain
their
discretion
to
set
the
thresholds
at
lesser
amounts
of
each
pollutant,
if
they
judge
that
lower
levels
are
appropriate.
Alternatively,
States
may
find
it
useful
to
exclude
de
minimis
emissions
in
identifying
whether
a
source
is
subject
to
BART
in
section
III
of
the
guidelines.
Either
approach
is
consistent
with
the
regulation
issued
in
this
rule.

2.4.2.3
De
Minimis
Values
for
Ammonia
62
Commenter
0259
said
if
EPA
retains
the
classification
of
ammonia
as
a
visibility­
impairing
pollutant
for
BART
purposes,
then
its
de
minimis
level
be
given
the
same
degree
of
technical,
scientific,
and
administrative
consideration
as
provided
to
other
pollutants.

Response:

We
have
not
provided
a
de
minimis
value
for
ammonia,
as
we
have
accorded
States
considerable
discretion
in
determining
whether
ammonia
is
a
visibility
impairing
pollutant
in
an
area,
and
in
assessing
the
predicted
visibility
impacts
from
ammonia
emissions
in
areas
where
the
ammonia
is
a
problem.

2.4.2.4
Guidance
for
Use
of
De
Minimis
Values
Sixteen
commenters
described
the
need
for
additional
guidance
on
how
the
de
minimis
provisions
would
be
implemented
(
0210,
0202,
0214,
0215,
0223,
0226,
0228,
0229,
0230,
0231,
0237,
0238,
0242,
0256,
0268,
0279).

Comment:
Commenters
(
0201)
said
the
final
guidance
should
clarify
where
in
the
BART
determination
process
de
minimis
levels
may
be
used.
Commenters
(
0201)
noted
the
discussion
of
de
minimis
levels
appears
under
step
4
of
the
eligibility
determination
in
the
preamble
and
under
step
3
in
the
proposed
guidance,
leading
to
confusion
over
EPA's
intent.
Commenter
0202
said
that
the
determination
should
be
made
at
the
eligibility
level
to
reduce
regulatory
burdens
on
facilities.
Otherwise,
facilities
may
be
required
to
submit
visibility
assessments,
including
CALPUFF
modeling
for
de
minimis
levels
of
emissions.
Commenter
0229
agreed
that
elimination
of
de
minimis
amounts
of
pollutants
from
BART
determination
should
occur
in
the
initial
screening
to
avoid
States
expending
unnecessary
resources.

Other
commenters
(
0214,
0215,
0223,
0226,
0228,
0230,
0238,
0268,
0279)
agreed
that
the
de
minimis
provisions
should
be
contained
in
the
BART
determination
stage.
Commenters
(
0215,
0228,
0242,
0256)
said
the
de
minimis
provision
has
no
implication
in
the
BART­
eligibility
portion
of
the
Guidance
since
these
pollutants
would
be
emitted
in
amounts
less
than
the
250
tons
per
year
pollutant­
specific
BART
eligibility
threshold.
Commenters
(
0214,
0215,
0223,
0226,
0228,
0238,
0242,
0256)
said
EPA
should
move
these
provisions
to
the
BART
determination
section
of
the
rule
and
specifically
clarify
that
visibility­
impairing
pollutants
emitted
in
de
minimis
amounts
would
not
be
required
to
be
examined
in
the
BART
determination.
Commenter
0226
added
that
the
BART
determination
process
is
technically
involved
and
time
intensive,
and
ultimately
must
consider
visibility
improvement.
For
units
that
emit
less
than
the
proposed
de
minimis
amounts
of
a
pollutant,
even
complete
elimination
of
that
pollutant
will
not
likely
produce
a
perceptible
visibility
improvement,
therefore,
controls
could
not
be
considered
cost
effective.
It
is
important
from
a
public
policy
perspective
that
States
and
sources
be
allowed
to
avoid
BART
determinations
for
these
pollutants,
as
the
result
is
a
foregone
conclusion.
63
Response:

We
have
retained
use
of
de
miminis
in
step
3
of
the
BART­
eligibility
stage
in
the
final
guidelines
but
also
allow
for
its
use
in
step
4.
We
do
not
believe
moving
it
to
the
BART
determination
step
would
have
much
practical
significance,
other
than
to
include
those
minimal
additional
tons
in
the
total
tons
of
emissions
counted
when
modeling
to
determine
contribution
to
visibility
impairment.
In
any
case,
the
reduction
in
administrative
burden
on
States
would
be
virtually
the
same
at
either
stage.

Comment:
Commenter
0201
said
other
steps
in
the
process
that
would
benefit
from
the
establishment
of
pollutant
thresholds
include
1)
a
unit
level
threshold
for
quantification
of
emissions
in
the
eligibility
determination,
2)
a
threshold
for
inclusion
in
modeling
in
determining
which
sources
are
subject
to
BART
and
in
step
5
of
the
engineering
analysis,
and
3)
a
source
level
threshold
for
inclusion
in
regional
scale
model
demonstrations.

Commenters
(
0201)
said
it
would
be
useful
if
de
minimis
levels
would
apply
at
the
source
and
unit
level
to
help
determine
which
pollutants
must
be
considered
in
an
engineering
analysis.
EPA
should
clarify
whether
the
proposed
thresholds
apply
at
the
"
source"
or
"
unit"
level.
If
the
intent
is
to
apply
the
thresholds
at
the
source
level,
then
EPA
should
allow
states
to
also
establish
reasonable
thresholds
at
the
unit
level.
Commenter
0237
is
concerned
that
EPA's
intent
is
to
continue
with
a
flawed
aggregation
concept,
which
would
result
in
conducting
a
BART
analysis
for
many
individual
units
tied
to
other
non­
BART
units,
which
Congress
clearly
did
not
intend
to
occur.
Commenters
(
0214,
0237,
0238,
0256)
said
EPA
should
clarify
that
the
de
minimis
provision
applies
on
a
unit
basis.
Commenter
0214
said
these
changes
to
the
proposed
rule
would
ensure
that
smaller
units
less
able
to
economically
equip
with
emissions
controls
would
be
able
to
logically
forego
the
BART
process.
Commenters
(
0215,
0228,
0256)
also
agreed
that
a
unit
basis
makes
the
most
sense
for
steam
electric
plants.

Commenters
(
0214,
0215,
0238,
0256)
also
said
EPA
should
state
that
a
BART
determination
is
required
only
for
those
visibility­
impairing
pollutants
that
are
emitted
above
de
minimis
levels
from
each
individual
emissions
unit
that
is
part
of
a
BART­
eligible
source.

Comment:
Commenter
0231
said
EPA
should
clarify
whether
actual
or
potential
emissions
should
be
compared
to
the
de
minimis
levels.

Response:

The
rule
language
makes
clear
that
de
minimis
levels
apply
to
the
BART­
eligible
source,
not
to
each
individual
emission
unit.
If
de
minimis
levels
applied
at
the
unit
level,
then
emissions
from
a
large
number
of
units
at
a
source
could
be
excluded,
while
at
the
same
time
those
emissions
might
constitute
a
considerable
contribution
to
visibility
impairment
from
the
source.
64
Further,
we
do
not
believe
that
applying
de
minimis
levels
at
the
"
source"
level
would
result
in
undue
burden
on
the
State.
We
believe
that
States
retain
enough
discretion
that
they
will
be
able
to
focus,
at
the
BART
determination
stage,
the
greatest
effort
on
the
emissions
units
of
greatest
concern.
We
note
that
the
pollutant
for
which
sources
most
commonly
have
a
multiplicity
of
emission
units
is
VOCs,
and
the
BART
guidelines
give
States
considerable
flexibility
in
addressing
VOCs.
Finally,
we
have
clarified
that
de
minimis
levels
apply
to
a
source's
potential
to
emit
each
relevant
pollutant.

2.5
States'
Burden
in
Identifying
the
BART­
eligible
Source
Six
commenters
said
the
proposed
process
of
identifying
BART­
eligible
sources
is
far
too
resource
intensive
for
resource­
starved
state
agencies
(
28­
264,
28­
266,
28­
272,
28­
278,
28­
312,
28­
404).
States
lack
the
detailed
inventories
needed,
including
information
on
"
in
existence"
and
"
in
operation,"
as
well
as
information
on
"
reconstruction"
activities
and
potential
emissions
(
28­
264,
28­
266,
28­
278,
28­
312).
Other
commenters
were
concerned
about
the
overall
burden
of
the
program
on
states
as
well
(
28­
272,
28­
278,
28­
312).

Response:

As
stated
earlier,
we
have
included
additional
provisions
in
the
final
guidelines
aimed
at
reducing
the
burden
on
States
(
such
as
discretion
in
covering
VOC
emissions).
We
believe
that
States
will
be
able
to
identify
the
BART­
eligible
sources
and
make
reasoned
BART
determinations,
and
expect
that
they
will
be
able
to
obtain
needed
information
on
emissions
and
construction
and
operation
dates
from
sources.

2.6
General
Comments
Regarding
Identifying
the
BART­
eligible
Source
Several
commenters
provided
general
comments
regarding
identifying
the
BART­
eligible
source
(
0186,
0221,
0235,
0237,
0267,
0275,
0293,
0294,
28­
340,
28­
343,
28­
343,
28­
355,
28=
393).

Commenters
(
0186,
0237,
28­
340,
28­
343,
28­
748)
said
EPA's
approach
for
determining
BART­
eligibility
is
not
authorized
by
the
Act
and
must
be
substantially
rewritten
to
meet
statutory
intent.
Commenter
0186
said
the
Act
does
not
mention
individual
emission
units,
does
not
mention
any
type
of
aggregation
over
time,
and
does
not
mention
aggregation
involving
parts
of
different
types
of
major
sources.
In
contrast,
under
the
statutory
scheme,
if
a
new
plant
is
one
of
the
26
listed
(
e.
g.,
petroleum
refinery,
chemical
plant,
etc.)
and
was
placed
in
service
between
August
7,
1962
and
August
7,
1977,
then
all
a
State
needed
to
do
was
determine
whether
the
plant's
emissions
exceeded
the
threshold
emissions.
This
simple
approach
is
strongly
supported
by
the
fact
that
Congress
made
it
clear
that
federal
guidelines
were
to
only
be
established
for
large
power
plants.
Other
commenters
(
0237,
0275,
0294,
28­
340,
28­
343,
28­
748)
also
discussed
the
complexity
of
the
proposed
process,
its
inherent
infeasibility
and
impractibility,
extensive
costs,
65
and
inconsistency
with
the
Act.
Commenter
0237
added
that
in
analyzing
the
cost
of
the
proposal,
EPA
has
failed
to
consider
in
its
assessment
whether
such
quantification
procedures
are
either
practical
or
feasible,
and
if
so,
at
what
cost
per
company.
At
the
least,
the
Agency
must
carefully
address
the
costs
of
determining
BACT
eligibility
before
finalizing
this
rule.

Commenters
(
28­
342,
28­
393)
said
the
proposed
guidelines
exceed
the
scope
intended
by
Congress
by
addressing
BART­
applicability
issues
such
as
how
source
age,
source
heat
input
and
source
emission
strength
would
affect
BART
eligibility,
how
the
effect
of
a
source's
emissions
on
the
visual
air
quality
of
a
MCIF
Area
would
affect
BART
eligibility
and
which
emissions
at
affected
sources
would
be
subject
to
BART.
The
manner
in
which
the
states
decide
what
sources
and
emissions
are
subject
to
BART
is
left
to
the
discretion
of
the
states
as
they
apply
the
statutory
factors.

Commenter
0267
was
generally
supportive
of
the
guidelines.
Commenter
0293
appreciated
the
additional
flexibility
provided
in
the
revised
BART
guidelines.
Maintaining
flexibility
in
measures
to
achieve
reasonable
progress
goals
allows
states
to
develop
appropriate
strategies
according
to
the
contributions
to
regional
haze
at
each
Class
I
area.
The
proposed
rule
and
guidelines
support
state
discretion
in
the
process
for
determining
BART­
eligible
sources.

Commenter
28­
255
said
the
extremely
low
threshold
for
BART
eligibility
cannot
be
justified
given
the
limited
emissions
reductions
required
­
in
terms
of
geographic
distribution,
and
type
and
amount
of
pollution
reduced
­
to
meet
visibility
goals.
The
rule's
nationwide
coverage
cannot
be
justified
given
EPA's
own
conclusion
that
"
virtually
all"
counties
in
the
East,
Midwest,
Northeast
and
Southeastern
parts
of
the
U.
S.
with
Class
I
areas
will
meet
visibility
goals
as
a
result
of
emissions
reductions
due
to
other
programs
alone.

Commenters
(
0221,
0237)
said
with
respect
to
the
re­
proposed
BART
guidelines
themselves,
EPA
must
strengthen
them
to
properly
identify
all
BART­
eligible
sources.
Commenter
28­
267
noted
that
ambiguity
exists
in
the
interpretation
of
exactly
what
constitutes
a
"
BART­
eligible"
source
and
EPA
should
clearly
define
this
issue.
When
States
begin
the
BART
process
there
must
be
national
consistency
in
the
identification
of
sources
to
which
that
process
applies,
although
the
end
result
of
a
rigorous
application
of
the
process
may
appropriately
define
different
strategies
for
different
jurisdictions.

Response:

As
explained
earlier,
we
believe
that
States
will
be
able
to
identify
the
BART­
eligible
sources
and
make
reasoned
BART
determinations,
and
we
also
believe
that
the
BART
guidelines
as
finalized
are
consistent
the
Congressional
intent
and
should
result
in
considerable
reduction
of
regional
haze.

3.0
Identify
Sources
Subject
to
BART
66
3.1
Methods
of
Determining
Whether
a
Source
is
Subject
to
BART
3.1.1
States
Assumes
That
All
Sources
are
Subject
to
BART
3.1.1.1
States
Should
Have
the
Option
Comment:
Five
commenters
said
that
states
should
have
the
option
to
determine
that
all
potential
BART
sources
contribute
to
haze
in
a
Class
I
area
(
0179,
0200,
0232,
0286,
0323).
Commenters
(
0200,
0232)
said
this
should
be
the
starting
point
of
source
eligibility.
Commenter
0179
added
that
under
the
proposed
rule,
facilities
would
still
be
allowed
on
a
case­
by­
case
basis
to
determine
if
they
were
not
impacting
a
Class
I
area,
and
could
therefore
be
exempt
from
BART.

Response:

As
part
of
this
rulemaking,
States
do
have
the
option
to
determine
that
all
of
its
BARTeligible
sources
collectively
contribute
to
haze
in
a
Class
I
area,
and
are
thus
subject
to
BART;
but
the
State
is
not
required
to
start
at
this
point.
The
State
may
also
determine
that
its
BARTeligible
sources
collectively
do
not
cause
or
contribute
to
any
visibility
impairment
at
a
Class
I
area,
or
the
State
can
use
a
source
by
source
analysis
to
determine
BART
applicability.
In
no
case
is
a
State
required
to
start
at
one
or
any
of
these
steps.

3.1.1.2
States
Lack
the
Authority
Comment:
Sixteen
commenters
argued
that
the
Act
does
not
provide
the
authority
to
allow
the
aggregation
of
emissions
(
0185,
0186,
0188,
0215,
0223,
0226,
0228,
0229,
0230,
0242,
0253,
0254,
0256,
0279,
0319,
0321).
Commenters
said
EPA's
proposal
has
no
legal
support
on
under
section
169A(
b)(
2)(
A)
and
was
expressly
rejected
by
the
court.
Commenter
0125
said,
contrary
to
EPA's
suggestion,
the
point
of
the
statute
and
of
Corn
Growers
is
not
simply
to
provide
authority
to
the
states
irrespective
of
what
the
states
choose
to
do
with
that
authority.
Rather,
the
point
of
the
statute
and
the
court
decision
is
to
provide
authority
to
the
states
to
determine
­­
consistent
with
the
source­
specific
language
of
section
169A(
b)(
2)(
A)
­­
whether
the
emissions
from
a
particular
source
cause
or
contribute
to
a
visibility
impairment.

Commenter
0226
said
use
of
EPA's
proposed
Option
1
for
determining
whether
a
source
is
subject
to
BART
allows
States
to
magically
conclude
that
all
BART
eligible
sources
in
that
State
"
cause
or
contribute"
to
visibility
impairment
in
Federal
Class
I
area(
s)
without
any
analysis
whatsoever.
This
effectively
eliminates
the
second
step
and
deems
that
all
BART
eligible
sources
automatically
go
through
a
BART
determination.
Commenter
0253
added
that
neither
the
language
of
Section
169A
nor
its
legislative
history
suggest
that
BART
may
be
triggered
for
any
67
particular
source
by
anything
less
than
a
technical
analysis
establishing
that
the
source
individually
causes
or
contributes
to
visibility
impairment.

Commenter
0186
noted
that
as
a
new
alternative
to
continued
use
of
the
remanded
"
generic"
finding
approach,
EPA
would
require
States
to
make
an
individual
finding
of
impairment
(
or
non­
impairment)
through
modeling
of
individual
BART­
eligible
sources.
Such
an
extremely
burdensome,
costly
and
time­
consuming
process
will
discourage
resource­
limited
States,
already
burdened
by
development
of
new
State
Implementation
Plans
for
the
8­
hour
ozone
and
PM2.5
standard,
from
using
this
alternative
approach.
Instead,
States
would
resort
to
the
"
generic"
finding
approach
encouraged
by
EPA,
which
undoubtedly
would
lead
to
State
lawsuits
over
an
approach
that
has
already
been
found
to
be
inconsistent
with
statutory
intent.

Response:

Given
the
low
threshold
for
determining
whether
a
source
should
be
subject
to
BART,
we
do
not
believe
that
the
CAA
can
be
read
as
prohibiting
States
from
adopting
a
cumulative
approach
to
determining
applicability.
Congress
provided
an
exemption
process
for
sources
that
do
not
contribute
significantly
to
impairment.

3.1.1.3
States
Must
Use
Source­
Specific
Information
Six
commenters
were
concerned
with
the
provisions
of
the
proposed
rule
that
suggest
a
state's
ability
to
deny
sources
the
opportunity
to
do
a
source­
specific
analysis,
and
allows
states
to
deny
a
source
a
BART
exemption
when
a
source
demonstrates
that
its'
impact
on
visibility
is
less
than
the
established
threshold
(
0185,
0188,
0215,
0254,
0256,
0279).
Commenters
said
these
provisions
should
be
modified
to
ensure
that
individual
sources
are
afforded
the
opportunity
to
conduct
an
analysis
to
demonstrate
that
its
impact
does
not
reasonably
impair
visibility
in
any
Class
I
area.
Commenter
0279
added
that
source
assessments
also
must
be
made
on
a
pollutantby
pollutant
basis.

Commenter
0254
said
that
EPA
must
reference
the
"
1999
study"
referenced
in
the
preamble
(
69
FR
25193
col.
2)
as
the
basis
for
EPA
allowing
States
to
make
a
determination
that
all
BART
eligible
sources
are
subject
to
BART
and
provide
for
proper
stakeholder
comment.

Response:

A
sources
has
the
opportunity
to
demonstrate
to
EPA
(
and
the
relevant
FLM)
that
it
does
not
contribute
to
significant
visibility
impairment.
In
addition,
in
making
BART
determinations,
the
State
considers
the
degree
of
visibility
improvement
likely
to
result
from
the
installation
of
BART.
Thus,
a
source
has
the
opportunity
to
demonstrate
to
the
State
that
it
should
not
be
required
to
install
BART.
As
explained
in
the
preamble,
States
that
chose
to
determine
that
all
BART­
eligible
sources
within
the
State
are
subject
to
BART
will
need
to
show
that
its
emissions
68
cause
or
contribute
to
visibility
impairment
in
one
or
more
Class
I
areas
based
on
update
emissions
inventories.
Stakeholders
will
have
the
opportunity
to
comment
on
that
showing.

3.1.2
State
Determines
That
No
Sources
are
Subject
to
BART
Using
a
Cumulative
Approach
3.1.2.1
States
Should
Have
the
Option
Comment:
Nine
commenters
(
0169,
0188,
0215,
0228,
0230,
0231,
0242,
0256,
0271)
supported
the
second
proposed
option
that
allows
States
to
use
a
cumulative
approach
to
demonstrate
that
none
of
its
BART­
eligible
sources
contribute
to
visibility
impairment.
Commenters
(
0188,
0228,
0256)
added,
if
this
cumulative
analysis
shows
a
contribution,
the
State
must
allow
each
individual
source
an
opportunity
to
demonstrate
that
its
own
emissions
do
not
by
themselves
contribute
to
visibility
impairment
in
a
Class
I
area.
Commenters
(
0215,
0228,
0256)
said
this
option
creates
no
conflict
with
the
individual­
source
focus
of
the
statutory
language
and
Corn
Growers,
because
it
is
essentially
a
screening
mechanism
for
individual
sources.

Commenter
0231
suggested
that
EPA
might
want
to
provide
more
flexibility
and
expand
this
option
to
allow
states
to
analyze
smaller
groups
of
sources
(
less
than
the
full
universe
of
BART
eligible
sources)
together
to
understand
their
cumulative
impact.
If
the
cumulative
impact
is
below
the
threshold,
then
the
conclusion
would
be
that
none
of
the
individual
sources
contribute
to
visibility
impairment.

Commenter
28­
403
said
EPA
has
identified
individual
states
or
multi­
state
areas
that
do
not
possess
any
Class
I
areas
as
the
only
potential
geographic
areas
that
could
conceivably
demonstrate
trivial
contributions.
This
interpretation
ignores
the
fact
that
states,
particularly
large
Midwestern
states,
may
be
composed
of
many
distinct
and
diverse
geographical
areas
and
related
air
sheds.
EPA's
interpretation
would
limit
states
and
RPOs
from
making
fundamental
and
wellreasoned
decisions
based
upon
actual
source­
receptor
relationships
within
these
smaller
geographic
areas.

Response:
The
guidelines
describe
several
approaches
that
States
may
use
to
determine
whether
its
sources
are
reasonably
anticipated
to
cause
or
contribute
to
any
visibility
impairment.
States
have
the
authority
to
adopt
other,
well­
reasoned
approaches
to
this
determination.

3.1.2.1
Oppose
the
Option
Comment:
Three
commenters
(
0179,
0181,
0275)
opposed
the
option
to
justify
a
BART
exemption
by
submitting
an
air
quality
modeling
demonstration
to
EPA
that
shows
that
sources
do
not
cause
or
contribute
to
any
visibility
impairment
in
a
nearby
Class
I
area.
Commenter
0179
said,
given
the
wide
geographic
dispersal
of
Class
I
areas
it
seems
difficult
to
believe
that
any
state
could
69
make
a
blanket
declaration
that
NO
facilities
in
their
state
could
contribute
to
haze
in
a
Class
I
area.
The
commenter
was
concerned
that
this
provision
would
allow
for
"
gaming"
of
the
system
and
said
it
should
be
removed
from
the
final
rule.
Commenter
0181
opposed
a
cumulative
pollutant
impact
option
because
the
"
cause
or
contribute"
test
must
be
pollutant­
specific.
Only
those
pollutants
that
contribute
to
visibility
impairment
in
a
Federal
Class
I
area
are
required
to
undergo
a
BART
engineering
analysis.
The
commenter
recommended
that
the
final
BART
guidelines
include
an
exemption
provision
that
requires
an
air
quality
modeling
demonstration
to
determine
whether
individual
BART­
eligible
sources
"
cause
or
contribute"
to
visibility
impairment
in
a
nearby
Federal
Class
I
area
on
a
pollutant­
specific
basis.
Finally,
commenter
0275
said
the
proposed
test
for
determining
whether
a
BART­
eligible
source
is
subject
to
BART
requirements
conflicts
with
Congressional
intent.
In
determining
whether
a
source
"
may
reasonably
be
anticipated
to
cause
or
contribute
to
any
impairment
of
visibility
in
any
such
[
class
I
Federal]
area"
and
therefore
is
subject
to
BART,
the
states
are
to
consider
the
impact
of
the
specific
source
in
question,
not
the
collective
impact
of
all
sources
in
a
broad
geographic
area.

Response:

We
concluded
in
the
1999
regional
haze
rule
that
all
States
contribute
to
regional
haze,
but
it
is
possible
that
a
State's
may
conclude
that
its
subset
of
BART­
eligible
sources
by
themselves
are
not
reasonably
anticipated
to
cause
or
contribute
to
visibility
impairment
in
a
Class
I
area.
States
may
reach
such
a
conclusion
on
a
pollutant
specific
basis.

3.1.2.3
Method
of
Determination
Comment:
Commenter
0323
said
EPA
needs
to
clarify
option
2
that
the
0.5
deciview,
24­
hour
maximum
threshold
applies
to
the
cumulative
contribution.
Commenter
0231
recommended
regional
models
such
as
CMAQ
or
CAMx
be
used
for
analyses
of
multiple
source
impacts.

Response:

The
cumulative
options
in
the
guidelines
will
allow
for
a
State
to
determine
a
threshold
of
comparison.
CMAQ
or
CAMx
may
be
used
in
the
analyses
contingent
upon
an
acceptable
modeling
protocol.

3.1.3
State
Makes
a
Source­
by­
Source
Determination
Using
an
Individualized
Exemption
Process
3.1.3.1
States
Should
Use
a
Source­
by­
Source
Process
Comment:
Eighteen
commenters
supported
the
third
option
that
is
based
on
an
individual
exemption
process
(
0181,
0184,
0188,
0195,
0202,
0214,
0215,
0224,
0228,
0230,
0238,
0241,
0254,
0268,
70
0277,
0304,
0314,
0321).
Some
commenters
(
0188,
0228,
0230,
0254,
0277,
0314)
supported
the
idea
that
states
and/
or
sources
should
have
access
to
this
process.
Commenters
(
0181,
0195,
0202,
0224,
0256)
said
that
the
BART
identification
process
should
be
based
on
site­
specific
air
quality
modeling
demonstrations
and
that
the
Act
requires
individual
facility
evaluations.
Commenters
said
this
approach
was
reaffirmed
by
the
2002
Court
decision.

Response:

The
guidelines
recommend
an
individual
exemption
process
as
one
option
that
States
may
use.
We
do
not
agree
that
the
CAA
requires
individual
facility
evaluations
to
show
that
a
source
is
subject
to
BART.

Commenters
also
addressed
the
role
that
source­
driven
assessments
should
play.
Commenters
(
0184,
0197,
0214,
0215,
0228,
0238,
0241,
0254,
0268,
0321)
said
that
without
the
right
to
perform
source­
specific
assessments,
the
proposal
is
in
violation
of
American
Corn
Growers
Ass'n
v.
EPA.
Instead,
for
legal
and
policy
reasons,
EPA
should
specifically
direct
the
states
in
the
guidelines
to
allow
sources
to
conduct
individualized
source
assessments
in
the
sources'
discretion.
Further,
states
should
have
maximum
flexibility
in
deciding
whether
to
require
individualized
source
assessments.
EPA's
proposal
limits
a
state's
discretion
to
identify
which
sources
contribute
to
visibility
impairment.
Commenter
0241
added
that
EPA
claims
that
any
potential
inequities
resulting
from
its
collective
contribution
test
would
be
addressed
during
the
BART
determination
process.
However,
sources
should
not
be
forced
to
expend
resources
to
go
through
a
full
BART
analysis
to
demonstrate
they
do
not
cause
or
contribute
to
visibility
impairment
in
a
Class
I
area.
States
should
allow
sources
the
opportunity
early
in
the
process
to
demonstrate
that
they
should
not
be
subject
to
BART.
Commenters
(
0170,0171)
noted
that
air
quality
modeling
is
important
to
allow
individual
sources
to
demonstrate
they
do
not
contribute
to
visibility
impairment.

Response:

We
agree
that
States
have
the
right
to
perform
source­
specific
assessments
and
may
allow
sources
to
conduct
these
assessments.
We
do
not
agree
that
the
guidelines
limits
the
State's
discretion
in
this
area.

Commenter
0202
said
facilities
should
be
able
to
provide
information
on
expected
operations,
control
impacts
and
the
costs
to
address
BART
eligibility.
This
factor
is
especially
important
for
facilities
that
have
curtailed
operations,
such
as
a
number
of
primary
aluminum
facilities
in
the
northwest
United
States.
Commenter
304
said,
in
the
end,
most
states
will
likely
find
that
modeling
of
impacts
from
individual
sources
will
need
to
be
done
to
determine
what
sources
must
install
BART.

Response:
71
We
agree
and
the
rule
provides
for
this.

Comment:
One
commenter
(
0271)
expressed
concerns
that
States
using
a
source­
by­
source
determination
of
the
visibility
degradation
effects
of
BART
sources
will
not
achieve
the
national
goals
of
improving
visibility
in
Class
I
areas.

Response:

A
BART
review
allows
for
5
factors
to
be
considered
for
each
source
­
visibility
improvement
is
one
factor
and
influences
the
BART
decision
to
control
a
source.
Source
controls
provide
measurable
improvement
and
is
part
of
Reasonable
Progress
in
State
Implementation
Plans
toward
the
goal
of
natural
conditions.

3.1.3.2
Method
of
Determination
Comment:
Three
commenters
discussed
source­
specific
modeling
(
0234,
0254,
0323).
Commenter
0254
supported
using
air
quality
models
such
as
CALPUFF
or
any
other
state­
approved
model
to
show
no
contribution
of
visibility
impairment
in
a
given
Class
I
area,
and
to
make
a
determination
whether
a
given
source
qualifies
for
an
exemption.

Response:

EPA
allows
for
the
use
of
CALPUFF
or
other
appropriated
dispersion
model.
See
40
CFR
Part
51,
Appendix
W
in
April
2003.

Commenter
0323
said
that
because
source­
specific
modeling
is
resource­
intensive,
EPA
should
clarify
that
this
option
has
two
sub­
options:
(
1)
states
may
analyze
each
BART­
eligible
source
in
the
state
or,
(
2)
states
may
chose
to
presume
that
all
BART­
eligible
sources
within
the
state
meet
this
applicability
test,
but
provide(
s)
sources
with
the
ability
to
demonstrate
on
a
case
by
case
basis
that
this
is
not
the
case.

Response:
We
agree;
States
are
provided
these
options
in
the
rule.

Commenter
0234
said
for
sources
to
be
exempted,
EPA
should
hold
the
showing
of
no
contribution
to
a
very
high
standard
in
order
to
protect
the
integrity
of
the
regional
haze
program.
Exempting
BART­
eligible
sources
from
controls
should
be
a
very
high
hurdle
for
both
the
owners
and
operators
of
sources
and
states.

3.1.4
Other
Comments
Regarding
the
Methods
to
Determine
Whether
a
Source
is
Subject
to
BART
72
Comment:
Some
commenters
(
0170,
0171,
0187,
0267,
0293)
generally
supported
the
three­
option
approach
proposed
by
EPA.
Commenter
0293
added
that
maintaining
flexibility
in
measures
to
achieve
reasonable
progress
goals
allows
states
to
develop
appropriate
strategies
according
to
the
contributions
to
regional
haze
at
each
Class
I
area.
The
proposed
rule
and
Guidelines
support
state
discretion
in
evaluating
whether
BART
is
required.
In
contrast,
commenter
0307
said
EPA
should
not
direct
the
states
to
use
one
specific
method
for
comparison
of
alternatives,
as
this
properly
is
within
the
latitude
conferred
upon
the
states
by
Congress
in
the
Act
for
directing
their
visibility
protection
programs.

Response:

The
guidelines
provide
the
States
with
flexibility
and
do
not
direct
the
States
to
use
one
specific
method
for
comparison
of
alternatives.

Commenters
(
0247,
0256,
0275)
said
EPA
failed
to
provide
source
owners
and
operators
with
the
necessary
discretion
to
do
source­
specific
visibility
impairment
assessments
as
allowed
by
the
Act.
Commenter
0247
added
that
the
opportunity
for
a
source
to
demonstrate
that
it
should
not
be
subject
to
BART
should
be
provided
in
all
cases,
independent
of
which
option
a
state
uses
to
identify
affected
sources.

Response:

We
have
provided
a
source
by
source
visibility
assessment
methodology
in
this
final
rule
for
a
State
or
a
source
to
make
a
BART
applicability
assessment.

Commenters
said
that
the
statute
requires
a
source­
by­
source
review
said
the
statute
itself
restricts
BART
to
sources
that
"
may
reasonably
be
anticipated"
to
cause
or
contribute
to
visibility
impairment
(
0189,
0197,
0208,
0237,
0275,
0308.
0321).
Commenters
(
0189,
0275,
0294)
added
that
the
statute
articulates
a
requirement
for
fact­
based
state
decisions
on
the
reasons
for
visibility
impairment
before
determining
that
a
source
is
subject
to
BART.
EPA's
proposal
violates
that
core
requirement
both
by
strongly
endorsing
an
automatic
finding
that
all
BART­
eligible
sources
are
subject
to
BART,
and
by
requiring
burdensome
modeling
for
any
attempt
to
show
otherwise.
Commenter
0208
said
the
first
two
options
are
therefore
inconsistent
with
the
statute.
The
guidelines
in
their
current
form
recreate
the
"
Group
BART"
approach
dismissed
by
Corn
Growers.
Commenter
0237
said
the
Court's
holding
requires
that
if
a
source
shows
that
its
emissions
are
not
reasonably
anticipated
to
cause
or
contribute
to
a
significant
impairment
of
visibility
in
any
mandatory
Class
I
areas,
EPA
must
exempt
it
from
BART.
The
final
rule
must
clarify
that
such
an
exemption
is
not
permissive
but
mandatory.

Response:
73
Given
the
nature
of
regional
haze,
it
is
reasonable
to
conclude
that
a
major
source
located
in
an
area
upwind
of
a
Class
I
area
contributes
to
visibility
impairment.
The
regulations
provide
a
process
for
EPA
to
exempt
from
BART
a
source
that
can
show
that
its
emissions
are
not
reasonably
anticipated
to
cause
or
contribute
to
a
significant
impairment
of
visibility.

Commenters
(
0200,
0232)
said
the
variety
of
potential
BART
exemptions
identified
in
the
new
rule
could
create
a
great
disparity
in
the
level
of
control
applied.
The
success
of
a
retrofit
program
is
based
on
nationally
consistent,
stringent,
application
of
controls
to
similar
sources.
Commenter
0232
said
all
potentially
BART­
eligible
sources
should
be
considered
subject
to
BART
determinations.
The
commenter
added
that
EPA
should
provide
clear
and
uniform
guidelines
for
how
options
"
b"
and
"
c"
would
be
determined.

One
commenter
(
0221)
said
EPA's
proposed
options
go
too
far
in
allowing
sources
to
be
"
exempted"
from
BART
requirements.
Initially,
EPA
needs
to
clarify
that
states
may
not
permit
a
BART
eligible
source
to
avoid
BART
requirements
without
an
affirmative
demonstration,
either
by
the
state
or
the
source,
that
the
source
does
not
emit
any
air
pollutant
which
may
reasonably
be
anticipated
to
cause
or
contribute
to
any
impairment
of
visibility
in
a
Class
I
area.
While
it
appears
that
EPA
did
not
intend
for
sources
to
avoid
BART
in
the
absence
of
such
a
demonstration,
the
language
of
the
proposal
on
this
point
must
be
made
more
clear
so
as
to
ensure
meaningful
implementation
of
the
BART
requirements
of
the
Act.
Otherwise,
a
state
or
a
BART­
eligible
source
might
attempt
to
avoid
BART
requirements
simply
by
failing
to
conduct
the
proper
analysis
of
visibility
impairment
contribution.

Response:
We
agree
that
the
State
must
determine
that
a
source
is
not
reasonably
anticipated
to
cause
or
contribute
to
visibility
impairment
in
order
to
find
that
it
is
not
subject
to
BART.

3.2
Source
Exemption
Process
3.2.1
Exemption
Process
for
Sources
More
Than
50
km
and
Less
Than
200
km
from
a
Class
I
Area
Comment:
Commenter
0259
said
for
sources
located
more
than
50
km
from
a
Class
I
area,
EPA's
proposed
guidelines
would
require
the
use
of
the
CALPUFF
model
to
exempt
a
BART­
eligible
source
from
being
subject
to
BART.
It
is
the
WRAP's
position
that
EPA
should
allow
states
and
tribes
to
choose
an
alternative
to
CALPUFF
in
such
situations,
providing
they
can
demonstrate
to
EPA
that
it
meets
the
same
level
of
rigorous
analysis
as
CALPUFF.

Response:

The
rule
allows
for
the
use
of
alternate
models
to
CALPUFF.
74
3.2.2
Exemption
Process
for
Sources
More
Than
200
km
from
a
Class
I
Area
3.2.2.1
Development
of
a
Written
Modeling
Protocol
Comment:
Two
commenters
(
0185,
0221)
generally
supported
the
use
of
modeling
protocols.
Commenter
0185
agreed
with
EPA
that
states
or
individual
sources
may
perform
assessments
to
determine
whether
a
given
source
qualifies
for
an
exemption.
Commenter
0221
added
that
for
source­
receptor
distances
substantially
in
excess
of
200km,
states
must
determine
visibility
impact
using
an
EPA­
approved
modeling
protocol.

Response:

We
agree.
States
should
prepare
or
require
their
BART­
eligible
sources
to
prepare
a
modeling
protocol
prior
to
conducting
any
modeling,
regardless
of
the
distance
from
the
source
and
the
Class
I
area.
This
is
reflected
in
the
rule.

Comment:
Several
commenters
raised
issued
related
to
using
CALPUFF
for
source
receptor
distances
beyond
200
km
(
0183,
0230,
0245,
0319).
One
commenter
(
0319)
said
CALPUFF
has
not
proven
reliable
over
the
extended
distances
that
exist
between
Class
I
areas
and
power
plants
in
rural
areas
and
it
is
expensive
to
run.
Developing
alternative,
more
accurate
modeling
protocols
will
be
complex
and
expensive.
It
will
be
difficult
for
states
to
justify
such
expenses,
especially
when
the
threshold
limit
is
0.5
deciviews
and
resulting
improvements
in
haze
may
be
imperceptible.
The
states
are
in
a
better
position
to
determine
which
modeling
protocol
and
model
input
values
best
reflect
conditions
in
their
states.
Thus,
the
appropriate
role
for
EPA
is
advisory
to
the
states
and
tribes
when
states
and
tribes
are
approving
CALPUFF
and
other
modeling
protocols
and
conducting
assessments.

Response:

We
agree
that
we
should
have
only
an
advisory
role
in
development
of
the
protocol
as
the
States
better
understand
the
BART­
eligible
source
configurations
and
the
geophysical
and
meteorological
data
affecting
their
particular
Class
I
area(
s).
We
believe,
however,
that
EPA
should
have
final
approval
of
State
protocols
so
that
the
model,
and
the
BART
control
decision
which
depends,
in
part,
on
the
model
results,
is
applied
equitably
across
all
regions.
EPA
approval
will
help
ensure
that
vastly
different
modeling
results
are
not
produced
for
modeled
BART­
eligible
sources
that
have
similar
emission
configurations
and
operate
under
comparable
geophysical
and
meteorological
conditions.

One
commenter
(
0245)
requested
specific
guidance
regarding
the
use
of
CALPUFF
in
these
situations:
°
Guidance
for
providing
input
for
nonhomogeneous
ammonia
emissions
into
CALPUFF
75
°
Guidance
for
running
CALPUFF
over
300
to
1000
km
from
release
or
suggestions
to
offset
biases
observed
by
CAPTEX.

Response:

Emissions
of
ammonia
are
not
considered
in
the
CALPUFF
model.
Only
a
constant
background
concentration
of
ammonia
is
established
by
the
model
user.
The
majority
of
BARTeligible
sources
are
within
500
km
of
a
Class
I
area,
and
the
modeling
guideling
provides
suggestions
for
potentially
improving
the
performance
of
CALPUFF
beyond
200
km
(
e.
g.
puff
splitting).
A
modeling
protocol
for
any
visibility
impact
determination
for
the
purposes
of
BART
should
be
submitted
to
and
approved
by
EPA
prior
to
the
modeling.
Collaboration
among
interested
parties
is
suggested
as
an
appropriate
means
to
assess
how
to
run
the
model.

Comment:
Three
commenters
discussed
the
need
for
additional
guidance
regarding
modeling
protocols
(
0185,
0188,
0201).
Commenters
0188
and
0201
urged
EPA
to
develop
guidance
on
modeling
protocols
to
address
complex
cases
where
the
source
is
greater
than
200
km
away
from
the
nearest
Class
I
area,
to
help
focus
state
resources
on
approvable
protocols,
and
to
provide
flexibility.
Commenter
0321
added
that
the
purpose
of
the
modeling
is
to
determine
whether
a
source
passes
or
fails
EPA's
proposed
exemption
"
threshold."

Response:

We
have
provided
guidance
as
part
of
the
final
rule.
Modeling
can
also
be
used
to
determine
the
degree
of
visibility
improvement
expected
by
potentially
controlling
a
BARTeligible
source.

3.2.2.2
CALPUFF
Limitations
for
Distances
over
200
km
or
More
Comment:
One
commenter
(
28­
242)
said
EPA's
December
1998
Interagency
Workgroup
on
Air
Quality
Modeling
recommends
using
CALPUFF
for
transport
distances
of
200
km
or
less.
Oahu
emission
sources
are
more
than
200
km
from
both
Haleakala
and
Volcanoes
parks.
Thus,
based
on
the
most
current
EPA
guidance
for
modeling
long
range
transport,
it
would
not
be
appropriate
to
model
Oahu
emissions
impact
on
either
Class
I
area
using
CALPUFF.
The
proposed
rules
should
allow
for
BART
exemptions
for
sources
greater
than
200
km
from
the
Class
I
area.

Response:

The
rule
does
allow
for
exemptions
for
distances
beyond
200
km
from
a
Class
I
area.
We
disagree
that
CALPUFF
is
inappropriate
beyond
200
km.
There
are
model
options
available
which
may
improve
the
performance
beyond
this
distance.
A
modeling
protocol
should
be
76
submitted
prior
to
conducting
any
modeling
analyses.
The
rule
also
allows
for
alternative
methods
to
CALPUFF
for
determining
BART
exemptions.

3.2.3
Modeling
Sources
Less
Than
50
km
from
a
Class
I
Area
Comment:
Four
commenters
discussed
the
benefits
of
various
models
(
0221,
0225,
0307,
0323
).
Commenter
0221
urged
EPA
to
require
either
CALPUFF
or
some
other
air
quality
model
approved
by
EPA
for
more
local­
scale
use
such
as
PLUVUEII.
Commenter
0225
said
VISCREEN
or
PLUVEUII
modeling
results,
for
receptors
less
than
50
km
from
the
BARTeligible
source,
should
have
equal
priority
to
CALPUFF
modeling
results.
Commenter
0307
added
that
EPA
should
also
suggest
to
states
that
conventional
dispersion
models
such
as
ISC3
and
AERMOD
may
be
useful
in
determining
if
a
single
source
could
contribute
to
visibility
impairment
where
a
satisfactory
connection
between
concentration
of
pollutants
and
visibility
impairment
thresholds
can
be
demonstrated.
For
many
BART­
eligible
sources
this
type
of
conventional
modeling
may
already
have
been
done
for
purposes
such
as
PSD
or
NAAQS
demonstrations.
Being
able
to
capitalize
on
previous
modeling
work
would
save
resources
of
both
state
agencies
and
source
owners,
thereby
reducing
the
cost
and
complexity
of
the
BART
program.
Commenter
0323
noted
that
plume
impact
models
have
traditionally
been
applied
at
distances
less
than
50
km
and
encouraged
EPA
to
require
such
models
be
used.
For
such
sources,
a
long
range
transport
analysis
may
also
be
required
to
assess
haze
visibility
impacts
at
Class
I
lands
located
beyond
50
km
from
the
source.

Response:
We
agree
and
have
provided
for
alternate
models
to
be
used
in
the
analyses.

Comment:
Commenter
0225
made
specific
recommendations
regarding
implementation
of
the
various
models:

°
BART­
eligible
sources
should
be
exempt
from
having
to
evaluate
or
add
controls
when
VISCREEN
or
PLUVEUII
modeling
results
show
the
source
will
not
cause
a
plume
with
any
hourly
estimates
of
delta
E
greater
than
or
equal
to
2.0,
or
the
absolute
value
of
contrast
values
(|
C|)
greater
than
or
equal
to
0.05.
°
Due
to
its
conservative
nature,
CALPUFF
modeling
for
receptors
less
than
50
km
from
the
BART­
eligible
source
located
in
complex
terrain
without
a
full
CTDMPLUS
meteorological
data
should
be
allowed
a
higher
threshold,
i.
e.,
a
threshold
of
a
maximum
24­
hr
change
of
1.0
deciview.

Response:

The
final
rule
allows
for
alternative
models
and
methodologies
to
be
used
for
analyses.
A
modeling
protocol
should
be
submitted
prior
to
conducting
any
modeling.
Exemption
decisions
and
thresholds
are
at
State
discretion.
77
3.2.4
Zero­
out
Approach
3.2.4.1
Oppose
Using
Zero­
out
Approach
Comment:
Four
commenters
said
the
zero­
out
approach
is
invalid
because
it
unrealistically
excludes
emissions
from
an
entire
region
and
ignores
how
those
emissions
would
interact
and
change
the
conditions
of
the
complex
atmospheric
reactions
(
28­
405,
28­
275,
28­
333,
28­
385).
EPA
must
allow
the
States
to
exercise
a
much
higher
level
of
professional
judgment
and
not
rely
on
such
simplistic
analytical
mechanisms
to
make
these
important
determinations.
Other
commenters
(
28­
348,
28­
364,
28­
387)
said
the
approach
only
addresses
Statewide
or
area
wide
emissions
contributions
(
not
source
by
source
contributions)
to
visibility
impairment.

Commenter
28­
385
said
eliminating
all
man­
made
emissions
in
zero­
out
model
runs
will
likely
produce
effects
not
contemplated
by
the
model
used.
Hence,
the
results
of
zero­
out
model
runs
are
and
should
be
viewed
with
suspicion.
In
addition,
EPA
states
in
64
FR
35722
that
"
geographic
'
regions'
that
can
contribute
to
regional
haze
generally
extend
hundreds
or
thousands
of
kilometers."
TVA's
atmospheric
modelers
believe
that
making
BART­
type
determinations
for
distances
of
thousands
of
kilometers
is
beyond
current
modeling
capabilities.

Response:

States
have
discretion
in
models
and
methodologies
to
use
in
making
visibility
impact
determinations.
Although
most
BART
sources
are
within
300
km
to
500
km
from
a
Class
I
area,
there
are
models
available
that
are
capable
of
reasonably
estimating
visibility
impacts
at
much
greater
distances.

Comment:
One
commenter
(
28­
358)
noted
that
EPA
hypothesizes
that
a
state
may
demonstrate
that
it
makes
only
a
"
trivial"
contribution
to
visibility
impairment
in
Class
I
areas.
While
the
implications
of
a
state
asserting
that
it
makes
a
trivial
contribution
are
not
expressly
stated,
from
the
context
it
appears
that
this
would
functionally
operate
as
an
exclusion
to
the
BART
requirement,
either
by
exempting
a
state
from
the
obligation
to
adopt
a
regional
haze
SIP
as
a
threshold
matter,
or
by
exempting
major
stationary
sources
from
the
BART
requirement
through
the
implementation
of
a
state's
regional
haze
SIP.
Neither
of
those
outcomes
is
legally
permissible:
the
statute
does
not
admit
of
a
de
minimis
exception
for
states
from
the
BART
requirements.

One
commenter
said
EPA
must
revise
the
guidelines
to
require
that
states
conduct
a
modeling
demonstration
of
a
non­
trivial
contribution
to
visibility
impairment
prior
to
a
source,
or
group
of
sources
being
subject
to
further
BART
analysis.
78
Response:

The
guidelines
provide
an
option
for
States
to
show
that
the
emissions
from
its
BARTeligible
sources
do
not
cause
or
contribute
to
visibility
impairment.
EPA
determined
in
1999
that
all
States
must
submit
regional
haze
SIPs.

3.2.4.2
Support
Exempting
Individual
"
Trivial
Emissions"
Sources
Comment:
Twelve
commenters
said
sources
should
be
allowed
to
demonstrate
that
they
are
individually
"
trivial"
emissions
sources
(
28­
242,
28­
272,
28­
273,
28­
311,
28­
348,
28­
357,
28­
386,
28­
392,
28­
779,
28­
7004,
28­
7005,
28­
7007).
Commenter
28­
242
described
why
this
approach
is
relevant
in
Hawaii
and
some
commenters
(
28­
273,
28­
779)
provided
other
examples.
Other
commenters
said
a
case­
by­
case
exemption
process
is
required
by
the
Act
(
28­
272,
28­
357).
Commenters
stressed
the
states'
role
in
determining
sources
subject
to
BART
(
28­
311,
28­
348,
28­
386,
28­
392,
28­
7004,
28­
7005).
Commenters
(
28­
386,
28­
392)
particularly
objected
to
the
requirement
that
states
must
prove
the
combined
influence
of
all
BART­
affected
sources
(
every
power
generator,
industrial
boiler
etc.
from
each
of
the
26
source
categories)
is
negligible
for
each
of
the
156
Class
I
areas.
Commenter
28­
7004
stressed
that
an
engineering
analysis
should
not
be
required
of
any
sources
with
trivial
contributions
to
visibility
impairment
in
Class
I
areas.

Response:

The
guidelines
provide
the
States
with
considerable
discretion
to
adopt
reasonable
approaches
for
making
BART
determinations.

Commenter
28­
7005
said
additional
language
should
be
included
that
would
allow
states
with
large
discrepancies
in
the
distances
between
sources
and
the
nearest
Class
I
area
to
provide
for
an
exemption
for
non­
culpable
sources.
Modeling
showing
the
culpability
of
sources
and/
or
subgroups
of
sources
within
a
state
or
region
is
likely
to
show
that
the
implementation
of
BART,
or
any
other
types
of
controls
in
areas
great
distances
from
Class
I
areas
will
have
a
trivial
effect
on
the
visibility
in
those
areas.
Commenter
28­
7007
agreed
that
relative
contributions
based
on
emissions
and
the
distance
of
the
plant
from
the
Class
I
area
should
be
taken
into
consideration.

Response:

In
this
final
rule,
States
may
conduct
an
individual
exemption
analysis
for
BART­
eligible
sources.

3.2.4.3
Need
Guidance
and
Definitions
to
Use
Zero­
out
Approach
Comment:
79
Five
commenters
said
EPA
does
not
address
the
smallest
acceptable
geographic
area
to
be
"
zeroed­
out"
or
the
amount
of
difference
between
the
base
rule
and
"
zero­
out"
run
that
is
considered
to
be
"
trivial"
(
28­
266,
28­
273,
28­
312,
28­
357,
28­
7008).
Commenter
28­
273
added
that
while
it
might
be
challenging
to
establish
a
bright­
line
threshold
beyond
which
contribution
is
no
longer
trivial,
the
rule
should
at
least
attempt
to
face
this
challenge,
or
at
least
provide
some
factors
or
other
guidelines
to
assist
the
states
in
making
this
determination.
Commenter
28­
7008
added
that
EPA
must
also
further
delineate
the
process
and
requirements
that
must
be
followed
in
making
a
demonstration
that
emissions
only
constitute
a
"
trivial
contribution
to
visibility
impairment."

Response:

This
rule
establishes
that
States
can
develop
a
"
cause
or
contribute"
visibility
threshold
and
that
modeling
demonstrations
can
be
made
for
comparison
to
such
threshold
and
if
impacts
are
below
this
threshold,
a
source
or
group
of
BART­
eligible
sources
may
be
exempt
from
a
BART
analysis.

Commenter
28­
357
said
there
is
only
one
possible
definition
that
would
be
consistent
with
169A:
EPA
should
remind
the
states
that
they
may
exempt
geographic
areas
as
"
trivial"
contributors
to
visibility
impairment
if
the
BART­
eligible
sources
within
a
geographic
region
do
not
by
themselves
cause
humanly
perceptible
impairment
of
visibility
in
a
Class
I
area.
If
it
does
not
choose
such
a
definition,
EPA
is
not
only
reading
the
statutorily
required
cost/
benefit
analysis
out
of
the
BART
process,
it
is
robbing
the
word
"
visibility,"
the
heart
of
the
entire
program,
of
any
meaning.

Response:

A
source
or
group
of
sources
need
not
cause
perceptible
impairment
of
visibility
in
order
to
"
contribute"
to
visibility
impairment.
As
explained
in
the
rule,
the
number
of
sources
contributing
to
visibility
impairment
is
one
factor
to
consider
in
establishing
an
appropriate
threshold.

3.2.5
General
Comments
Regarding
the
Source
Exemption
Process
3.2.5.1
Flexibility
Over
the
Choice
of
Models
Comment:
Eight
commenters
discussed
the
limitations
or
other
technical
issues
associated
with
using
CALPUFF
(
0189,
0198,
0217,
0230,
0257,
0302,
0309,
0314).

Commenter
0189
cited
recent
experience
showing
that
appropriate
consideration
of
meteorological
events
such
as
precipitation
should
be
allowed
in
the
refined
BART
assessment
80
using
CALPUFF.
Enhancements
to
the
CALPUFF
code
soon
may
be
available
that
could
greatly
facilitate
the
incorporation
of
meteorological
interference
issues
in
regional
haze
modeling.

Response:

Any
revisions
to
the
CALPUFF
code,
including
adjustments
for
meteorological
influences,
that
have
been
approved
for
use
by
EPA
are
allowed
for
BART
assessments.
The
final
rule
also
allows
for
alternative
models
and
methodologies
to
be
used
for
analyses.

Commenters
pointed
out
several
limitations
to
the
CALPUFF
model
in
this
application
(
0198,
0217,
0230,
0275,
0302).
Commenters
(
0198,
0230,
0275)
supports
the
use
of
the
CALPUFF
model
as
a
screening
tool
only.
Limitations
described
include:
°
CALPUFF
simplifies
many
atmospheric
processes
that
are
important
for
estimating
visibility
impairment
at
Class
I
areas
due
to
emissions
from
potential
BART­
eligible
sources.
°
CALPUFF
sulfate
(
SO4)
and
nitrate
(
NO3)
formation
chemistry
modules
have
many
limitations:
a)
They
were
developed
for
conditions
not
representative
of
those
that
occur
in
the
locales
of
many
potential
BART­
eligible
sources
during
large
portions
of
the
year,
b)
They
are
inaccurate
compared
to
current
state­
of­
science
chemistry
modules,
and
c)
They
are
out­
of­
date
having
been
developed
over
two
decades
ago.
°
SO4
and
NO3
estimates
from
CALPUFF
are
likely
to
be
inaccurate
and
generally
overestimated.
°
The
CALPUFF
model
has
not
been
approved
for
visibility
evaluations.
°
CALPUFF
has
not
been
adequately
tested
or
evaluated
for
Class
I
type
air
quality
analyses.
°
There
is
very
little
information
regarding
CALPUFF's
performance
over
large
distances
and
complex
terrain.
°
There
has
been
no
testing
of
the
accuracy
of
the
CALPUFF
chemistry
models.
°
CALPUFF
is
time­
consuming
and
data
intensive.

Response:

As
discussed
in
the
preamble
to
the
rule,
we
understand
the
concerns
of
commenters
that
the
chemistry
modules
of
the
CALPUFF
model
are
less
advanced
than
some
of
the
more
recent
atmospheric
chemistry
simulations.
To
date,
no
other
modeling
applications
with
updated
chemistry
have
been
approved
by
EPA
to
estimate
single
source
pollutant
concentrations
from
long
range
transport.
In
its
next
review
of
the
Guideline
on
Air
Quality
Models,
EPA
will
evaluate
these
and
other
newer
approaches
and
determine
whether
they
are
sufficiently
documented,
technically
valid,
and
reliable
to
approve
for
general
use.
In
the
meantime,
as
the
Guideline
makes
clear,
States
are
free
to
make
individualized
showings
that
these
or
other
alternative
approaches
are
valid
and
appropriate
for
their
intended
applications.
81
Theoretically,
the
CALPUFF
chemistry
simulations,
in
total,
may
lead
to
model
predictions
that
are
generally
overestimated
at
distances
downwind
of
200
km.
Again,
States
can
make
judgements
concerning
the
conservativeness
or
overestimation,
if
any,
of
the
results.
The
use
of
other
models
and
techniques
to
estimate
if
a
source
causes
or
contributes
to
visibility
impairment
may
be
considered
by
the
State,
and
the
BART
guidelines
preserve
a
State's
ability
to
use
other
models.

One
commenter
(
0309)
had
serious
concerns
about
the
CALPUFF
model
used
by
the
EPA
to
assess
the
impact
of
the
proposed
BART
rule
on
Indian
country.
The
model
fails
to
independently
incorporate
tribal
lands
because
horizontal
grid
sizes
are
primarily
based
on
county
boundaries
which
ignore
overlapping
tribal
and
state
jurisdictions.
Such
modeling
therefore
has
absent
from
it
the
necessary
input
from
tribal
lands
that
insures
an
effective
analysis
of
air
dispersion
from
individual
sources,
and
the
ultimate
effect
of
such
source
emissions
on
Class
I
areas.
According
to
the
commenter,
a
CALPUFF
assessment
of
an
individual
source
cannot
be
used
as
the
preferred
approach
for
determining
whether
a
BART­
eligible
source
may
be
exempt
from
BART
until
air
dispersion
over
tribal
lands
is
properly
modeled,
taking
into
account
the
unique
and
vast
terrains
that
exist
on
such
lands.

Response:

We
disagree.
For
purposes
of
modeling,
a
BART­
eligible
source
is
independent
of
grid
size
and
can
be
located
anywhere
in
the
model
domain
which
may
include
tribal
lands
and
the
meteorological
and
terrain
effects
in
those
geographic
areas.
Air
quality
impacts
predicted
by
dispersion
models
is
not
geographically
specific.

Comment:
Ten
commenters
said
EPA
should
clarify
that
models
other
than
CALPUFF
may
be
used
subject
to
approval
by
the
state
agency
(
0188,
0214,
0217,
0228,
0231,
0238,
0256,
0268,
0270,
0302).
Some
commenters
noted
that
the
inconsistencies
in
the
preamble
compared
to
the
rules
needs
to
be
clarified
to
make
this
point.
Commenter
0256
added
that
EPA
should
revise
the
guidelines
to
remove
the
references
to
CALPUFF,
other
than
an
introductory
statement
that
indicates
that
CALPUFF
is
one
air
quality
model
that
may
be
used
and
that
other
models
may
be
used
subject
to
approval
by
the
state
air
agency.

Response:

We
agree
and
the
rule
provides
for
the
use
of
alternative
models.
However,
a
modeling
protocol,
including
a
proposal
to
use
alternative
models
to
CALPUFF,
should
be
submitted
prior
to
conducting
any
modeling.
This
will
ensure
consistency
of
model
use
across
states
and
regions.

Commenter
0270
said
prior
to
mandating
that
any
modeling
for
BART­
eligible
sources
located
at
least
50
km
from
a
Class
I
area
be
performed
using
CALPUFF,
EPA
must
explain
why
the
use
of
CALPUFF
is
discretionary
under
Part
51
and
mandatory
under
Part
308
and,
in
turn,
82
must
solicit
comment
on
the
propriety
of
specifying
CALPUFF
as
the
sole
modeling
technique
for
such
sources.
Until
EPA
solicits
such
comment,
it
is
unable
to
require
a
state
or
sources
within
the
state
and
located
at
least
50
km
from
a
Class
I
area
to
use
only
the
CALPUFF
model
when
trying
to
demonstrate
no
reasonably
attributable
haze
impairment
to
a
Class
I
area.

Response:

The
rule
does
not
mandate
the
use
of
CALPUFF
and
provides
States
the
flexibility
to
use
other
models
to
conduct
visibility
impact
assessments.

Commenter
0237
objected
to
the
unreasonable
modeling
regime
laid
out
in
the
proposed
BART
guidelines
for
proving
a
"
negative
determination.
For
all
practical
purposes,
CALPUFF
modeling
will
always
demonstrate
some
sort
of
impact
on
a
Class
I
area,
even
at
great
distances
from
Class
I
areas.
Further
if
exemption
is
predicated
on
this
type
of
modeling
alone
 
to
the
exclusion
of
other
simpler
methodologies
 
the
cost
of
the
modeling
will
discourage
most
states
from
choosing
to
allow
sources
to
try
to
show
the
impact,
even
if
this
could
be
done
successfully.
If
this
is
the
only
method
EPA
allows
in
the
final
rule,
it
would
be
tantamount
to
not
allowing
a
showing
at
all,
and
the
BART
guideline
clearly
would
fail
to
satisfy
the
Court's
decision
in
American
Corn
Growers
Association
v.
EPA.

Response:

We
disagree.
The
rule
allows
State
discretion
in
determining
an
appropriate
"
cause
or
contribute"
threshold
for
BART
applicability.
Also,
States
may
develop
their
own
methodologies,
including
non­
modeling
techniques
for
assessing
BART
applicability,
or
they
have
the
flexibility
to
use
models
other
than
CALPUFF.

Comment:
Commenters
discussed
alternatives
to
CALPUFF
(
0183,
0201,
0210,
0245).
Speaking
generally,
commenters
(
0201,
0210)
said
all
parties
will
acknowledge
that
CALPUFF
has
limits
and
encouraged
EPA
to
include
language
in
the
final
guidance
that
allows
the
use
of
other
analysis
tools
that
may
better
represent
actual
conditions,
including
grid
models
(
0210).
Commenter
0245
instead
encouraged
EPA
to
use
the
CAIR
program
as
the
more
appropriate
first
step
to
achieve
visibility
goals.
This
will
allow
sufficient
time
for
EPA
to
focus
its
resources
on
the
refinement
and
development
of
additional
tools
with
a
higher
degree
of
confidence
in
the
results
than
those
problematic
tools
relied
upon
in
this
proposal.
Commenters
0198
and
0230
described
the
important
role
that
photochemical
grid
models
can
play
and
that
their
use
should
be
allowed.
Commenter
0183
was
concerned
about
the
differences
in
the
chemistry
mechanisms
in
CALPUFF
compared
to
photochemical
models.
Use
of
photochemical
models
by
regional
planning
organizations
may
produce
results
inconsistent
with
results
obtained
from
using
CALPUFF
to
model
single
sources.

Response:
83
As
discussed
in
the
preamble
to
the
rule,
we
understand
the
concerns
of
commenters
that
the
chemistry
modules
of
the
CALPUFF
model
are
less
advanced
than
some
of
the
more
recent
atmospheric
chemistry
simulations.
To
date,
no
other
modeling
applications
with
updated
chemistry
have
been
approved
by
EPA
to
estimate
single
source
pollutant
concentrations
from
long
range
transport.
In
its
next
review
of
the
Guideline
on
Air
Quality
Models,
EPA
will
evaluate
these
and
other
newer
approaches
and
determine
whether
they
are
sufficiently
documented,
technically
valid,
and
reliable
to
approve
for
general
use.
In
the
meantime,
as
the
Guideline
makes
clear,
States
are
free
to
make
individualized
showings
that
these
or
other
alternative
approaches,
including
grid
models,
are
valid
and
appropriate
for
their
intended
applications.
With
regard
to
the
CAIR
program,
affected
states
have
the
option
to
use
the
CAIR
provisions
to
meet
the
BART
requirements
for
BART­
eligible
EGUs
or
other
source
types
covered
by
CAIR.

Commenter
0202
noted
that
research
is
progressing
on
developing
dilution
tunnel
methods
to
assess
aerosol
formation
from
stack
gas
testing,
relying
on
air
chamber
analysis
for
atmospheric
aerosol
formation.
Those
alternatives
may
become
a
viable
measurement
method
for
many
operations.
Commenter
0217
said
an
example
of
other
models
that
could
be
used
would
include
the
EPRI
plume
chemistry
model,
SCICHEM
(
which
is
technically
far
superior
to
CALPUFF).
Commenter
0275
said
that
additional
consideration
should
be
given
to
the
"
weight
of
evidence"
approach
(
as
adopted
in
ozone
demonstrations)
in
appropriate
circumstances.

Response:

We
agree.
States
have
the
option
to
use
alternatives
to
CALPUFF
in
making
visibility
assessments
for
BART.

Comment:
Four
commenters
said
that
the
choice
of
models
should
be
at
the
states'
(
or
RPOs)
discretion
(
0217,
0223,
0226,
0242).
Three
commenters
said
that
states
and
sources
should
both
be
allowed
this
flexibility
(
0228,
0230,
0230).
Commenter
0230
added
that
increasing
the
flexibility
in
the
potential
modeling
approaches
and
requiring
the
development
of
a
Modeling
Protocol
to
allow
EPA
and
others
to
review
the
proposed
modeling
approaches
would
alleviate
the
current
inconsistencies
and
contradictions
between
the
Proposed
BART
Rule
and
EPA's
air
quality
modeling
guidelines
and
draft
fine
particulate
and
regional
haze
modeling
guidance.

Response:

We
agree
that
a
modeling
protocol,
including
proposed
use
of
other
models
besides
CALPUFF,
should
be
required
for
all
modeling
demonstrations.
We
believe,
however,
that
EPA
should
have
final
approval
of
State
or
source
developed
protocols
so
that
the
model,
and
the
BART
control
decision
which
depends,
in
part,
on
the
model
results,
is
applied
equitably
across
all
regions.
EPA
approval
will
help
ensure
that
vastly
different
modeling
results
are
not
produced
84
for
modeled
BART­
eligible
sources
that
have
similar
emission
configurations
and
operate
under
comparable
geophysical
and
meteorological
conditions.

3.2.4.2
Suggested
Modeling
Methodology
Comment:

Thirteen
commenters
discussed
the
type
of
emissions
that
should
be
used
in
modeling
(
0169,
0198,
0214,
0226,
0228,
0230,
0231,
0242,
0254,
0256,
0259,
0276,
0306).

Commenters
supported
the
use
of
"
actual"
rather
than
"
allowable"
emissions
(
0198,
0214,
0226,
0228,
0242,
0254,
0256,
0276).
Commenters
said
that
it
was
Congress'
intent
to
use
actual
real
world
impacts
in
making
this
"
cause
or
contribute"
determination.
To
achieve
this
it
is
reasonable
to
use
the
sources
actual
emission
rate.
The
commenter
suggested
that
the
use
of
the
last
2
years
of
operation
is
appropriate
unless
another
2­
year
period
can
be
shown
to
be
more
representative.
Commenter
0256
added
that
a
source
also
should
be
allowed
to
model
using
its
actual
daily
activity
patterns
(
e.
g.,
lower
load
at
night
for
power
plants).
States
should
also
be
required
to
account
for
actual
emission
rates,
as
well
as
daily
and
annual
production
schedules,
which
are
representative
of
a
source's
actual
emissions
profile
(
0254).

The
CAA
requires
the
States
to
make
BART
determinations
for
any
source
that
the
State
determines
may
reasonably
be
anticipated
to
cause
or
contribute
to
visibility.
This
is
a
low
triggering
threshold
and
requires
the
State
to
look
forward
to
anticipate
those
sources
that
may
cause
or
contribute
to
visibility
impairment.
Accordingly,
conservative
emission
estimates
should
be
used.
We
agree,
however,
that
using
allowable
emission
rates
could
unduly
inflate
the
likely
impacts
of
a
source.
Although
a
source
could
emit
at
its
allowable
levels,
a
more
representative
emission
rate
based
on
its
highest
actual
emissions
over
the
past
three
to
five
years,
or
an
estimate
of
its
emissions
during
periods
of
high
capacity
utilization,
is
appropriate.
Emissions
from
a
source
can
vary
widely
on
a
day
to
day
basis;
during
peak
operating
days,
the
24­
hour
actual
emission
rate
could
be
more
than
double
the
daily
average.
On
the
other
hand,
in
the
long
term,
estimating
visibility
impacts
based
on
allowable
emission
rates
for
every
hour
of
the
year
may
unduly
inflate
the
maximum
24
hour
modeled
impairment
estimate
from
a
BART­
eligible
source.
.
Where
States
have
information
on
a
source's
daily
emissions,
an
emission
rate
based
on
the
maximum
actual
emissions
over
a
24
hour
period
for
the
most
recent
five
years
may
be
a
more
appropriate
gauge
of
a
source's
potential
impact
as
it
would
ensure
that
peak
emission
conditions
are
reflected,
but
would
likely
not
overestimate
a
source's
potential
impact
on
any
given
day.
We
have
accordingly
included
this
change
to
the
final
guidelines.
We
recommend
that
the
State
use
the
highes
24­
hour
average
actual
emission
rate,
for
the
three
or
five
year
period
of
meteorological
data,
to
assess
BART
applicability
and
to
characterize
the
maximum
potential
benefit
of
controls.

Commenter
0214
noted
that
most
EGU's
operate
well
below
their
allowable
emissions
rate
to
ensure
there
is
a
margin
of
safety
for
compliance
purposes.
Commenter
0276
said
sources
85
that
have
been
performing
well
below
their
Title
V
permitted
levels
should
not
be
penalized
during
these
screening
and
modeling
assessments.

Response:

As
discussed
above,
the
final
rule
provides
for
the
use
of
the
maximum
24­
hour
actual
emission
rate,
over
a
five
year
period,
rather
than
the
allowable
emission
rate.
A
source
is
encouraged
to
take
permit
limits
to
lower
potential
visibility
impacts.

Commenter
0228
said
the
use
of
an
average
emission
rate
over
a
multiple­
year
period
would
be
more
appropriate
and
urged
EPA
to
revise
the
guidelines
and
adopt
a
more
flexible
approach
that
would
provide
a
more
realistic
assessment
of
whether
a
source
is
causing
or
contributing
to
visibility
impairment
in
a
Class
I
area.

Response:

We
agree
and
have
provided
for
this
in
the
final
rule.

One
commenter
(
0230)
said
states
and
sources
should
have
the
option
of
using
maximum
actual
emissions,
rather
than
maximum
allowable
or
PTE
emissions
in
their
BART
analysis,
provided
that
the
source
is
willing
to
adopt
these
maximum
actual
emissions
as
a
maximum
permitted
emissions
rate.

Response:

In
the
final
rule,
the
maximum
24­
hour
actual
emission
rate
is
allowed
in
modeling
assessments.

Commenter
0242
said
if
EPA
is
unwilling
to
accept
the
actual
emission
rate
option,
EPA
should
provide
guidance
on
how
to
determine
the
allowable
emission
rate
for
nitrogen
oxide
emissions
for
plants
operating
under
seasonal
controls
resulting
from
the
NO
x
SIP
call.

Response:

In
the
final
rule,
the
maximum
24­
hour
actual
emission
rate
is
allowed
in
modeling
assessments.

Some
commenters
supported
the
use
of
annual
emissions
(
0169).
Commenter
0169
said
that
since
the
BART
rule
addresses
long­
term
visibility
impairment,
the
most
appropriate
emissions
to
consider
in
CALPUFF
modeling
would
be
annual
emissions.
In
visibility
modeling,
the
use
of
24­
hr
emissions
is
inappropriate
because
these
emissions
are
short
term
peaks
in
visibility
and
their
use
would
overestimate
both
the
likelihood
and
the
expected
magnitude
of
visibility
impacts.
86
Response:

We
disagree
with
the
use
of
annual
emissions.
Emissions
from
a
source
can
vary
widely
on
a
day
to
day
basis;
during
peak
operating
days,
the
24
actual
emission
rate
could
be
more
than
double
the
daily
average.
On
the
other
hand,
in
the
long
term,
estimating
visibility
impacts
based
on
allowable
emission
rates
for
every
hour
of
the
year
may
unduly
inflate
the
maximum
24
hour
modeled
impairment
estimate
from
a
BART­
eligible
source.
Where
States
have
information
on
a
source's
daily
emissions,
an
emission
rate
based
on
the
maximum
actual
emissions
over
a
24
hour
period
for
the
most
recent
five
years
may
be
a
more
appropriate
gauge
of
a
source's
potential
impact
as
it
would
ensure
that
peak
emission
conditions
are
reflected,
but
would
likely
not
overestimate
a
source's
potential
impact
on
any
given
day.
We
have
accordingly
included
this
change
to
the
final
guidelines.
We
recommend
that
the
State
use
the
worst
24­
hour
average
actual
emission
rate,
for
the
three
or
five
year
period
of
meteorological
data,
to
assess
BART
applicability
and
to
characterize
the
maximum
potential
benefit
of
controls.

Commenter
0231
noted
that
in
the
proposed
rule
EPA
uses
the
term
"
allowable
emissions"
in
reference
to
CALPUFF
and
the
emissions
over
distance
approach.
The
commenter
said
EPA
should
clarify
that
"
potential
emissions"
are
to
be
used.
Commenter
0259
asked
whether
the
24­
hour
period
should
be
determined
on
a
rolling
or
daily
basis.
Commenter
0306
said
EPA
needs
to
definitively
list
the
emission
rates
that
are
to
be
used
in
any
modeling
analysis.
This
would
include
whether
actual
or
allowable
emission
rates
are
to
be
used.
Many
sources
have
different
allowable
emission
rates
for
different
averaging
periods
(
e.
g.
1­
hour,
3­
hour,
24­
hour
or
annual
averages).
The
averaging
period
for
emission
rates
must
also
be
specified.

Response:

In
the
final
rule,
the
maximum
24­
hour
actual
emission
rate
is
allowed
in
modeling
assessments.

Comment:
Four
commenters
said
the
visibility
impairment
analysis
must
be
done
for
each
pollutant
separately
(
0226,
0242,
0256,
0268)
as
required
by
the
Act.
Commenters
(
0242,
0256)
anticipated
that,
for
most
fossil­
fuel
fired
steam
electric
plants,
emissions
of
VOCs
(
inherently
low
at
power
plants)
and
probably
PM
(
relatively
low
at
power
plants,
because
PM
is
already
well
controlled)
would
not,
individually,
"
cause
or
contribute
to"
visibility
impairment
in
any
Class
I
area.
Thus,
under
the
"
subject
to
BART"
step
of
the
Guidelines,
if
written
in
a
way
that
is
consistent
with
the
Act,
most
power
plant
owners
and
operators
should
be
able
to
show
that
neither
VOC
emissions
nor
PM
emissions
from
their
plants
are
"
subject
to
BART"
and
therefore
that
no
BART
determination
is
required
for
those
pollutants.

Comment:
Six
commenters
provided
additional
comments
on
the
modeling
methodology
(
0199,
0214,
0230,
0238,
0256,
0268).
Some
commenters
(
0214,
0230,
0238,
0256,
0268)
stressed
that
87
it
is
important
that
states
and
sources
retain
the
flexibility
to
decide
how
to
set
up
and
run
the
selected
model.
Commenter
0230
added
the
proposed
rule
should
allow
the
flexibility
to
use
the
latest
best
scientific
information
in
visibility
calculations
that
should
be
described
in
the
Modeling
Protocol
prior
to
performing
the
analysis
to
allow
review
and
comment.

Response:

We
agree.

Other
commenters
provided
specific
technical
comments:
°
Concerning
CALPUFF
model
receptors,
the
guidelines
should
specify
that
the
IMPROVE
monitor
is
the
receptor
by
which
modeled
visibility
impacts
should
be
evaluated
(
0199).
°
The
FLAG
(
2000)
report
specifies
extinction
coefficients
be
used
to
convert
particulate
matter
(
PM)
species
concentrations
to
visibility
extinction
that
include
a
value
of
3.0
to
convert
dry
particulate
SO4
and
NO3
concentrations
(
in
µ
g/
m3)
to
extinction
(
in
Mm­
1).
More
recent
scientific
evidence
suggests
that
a
SO4
and
NO3
dry
extinction
coefficient
of
2.5
is
more
appropriate
(
0230).
°
The
FLAG
(
2000)
document
recommends
a
cap
on
the
RH
of
98%;
the
SO4
and
NO3
extinction
coefficient
at
98%
RH
will
be
18
times
greater
than
the
"
dry"
value
at
50%
RH
or
less.
This
can
greatly
overstate
the
visibility
impacts
due
to
a
SO
x/
NO
x
source.
To
reduce
the
sensitivity
of
this
inaccuracy
in
the
visibility
calculations,
a
lower
cap
of
the
RH
such
as
90%
as
recommended
by
White
should
be
used
in
the
f(
RH)
calculations
(
0230).
Response:

Advances
in
science
and
their
incorporation
into
models
to
improve
model
performance
is
certainly
encouraged
by
EPA.
Once
model
parameter
revisions
have
been
demonstrated
to
be
acceptable
and
are
approvable
by
EPA,
we
may
revise
the
latest
guidance
and
EPA­
supported
modeling
updates.
States
should
contact
the
appropriate
FLM
and
EPA
regional
office
for
the
latest
guidance
and
modeling
updates.
We
disagree
that
hte
IMPROVE
monitor
should
be
the
receptor
used
in
visibility
impact
assessments.
The
IMPROVE
monitor
is
not
necessarily
representative
of
conditions
over
the
entire
Class
I
areas,
and
as
such,
maximum
visibility
impacts
may
occur
at
locations
other
than
the
improve
monitor.
One
of
the
modeling
goals
is
to
estimate
the
maximum
visibility
impact.

3.2.4.3
Need
for
Modeling
Guidance
Comment:
Five
commenters
(
0188,
0228,
0242,
0256,
0276)
discussed
how
to
identify
Class
I
areas
that
should
be
modeled.
Commenters
(
0188,
0228,
0242,
0256,
0276)
said
the
guidelines
should
require
sources
to
model
only
the
nearest
Class
I
area
(
or
possibly
the
two
closest
[
0242])
,
or,
in
the
alternative,
commenter
0228
added
to
provide
a
reasonable
methodology
to
minimize
in
as
much
as
possible
requirements
to
address
impacts
to
multiple
Class
I
areas.
88
Response:

As
stated
in
the
preamble,
a
common
sense
approach
should
be
taken,
particularly
where
an
analysis
may
add
a
significant
resource
burden
to
a
State.
For
example,
if
there
are
multiple
Class
I
areas
in
relatively
close
proximity
to
a
BART­
eligible
source,
a
State
may
model
a
full
field
of
receptors
at
the
closest
Class
I
area.
Then
a
few
strategic
receptors
may
be
added
at
the
other
Class
I
areas
(
perhaps
at
the
closest
point
to
the
source,
a
receptor
at
the
highest
and
lowest
elevation
in
the
Class
I
area,
a
receptor
at
the
IMPROVE
monitor,
and
a
few
receptors
that
are
expected
to
be
at
the
approximate
plume
release
height).
If
the
highest
modeled
impacts
are
observed
at
the
nearest
Class
I
area,
a
State
may
choose
not
to
analyze
the
other
Class
I
areas
any
further
and
additional
analyses
might
be
unwarranted.

Comment:
Ten
commenters
requested
additional
modeling
guidance
(
0201,
0221,
0230,
0242,
0267,
0304,
0306,
0314,
0321,
0323)

One
commenter
(
0221)
said
EPA
should
develop
a
CALPUFF
evaluation
protocol
that
must
be
followed
by
the
states
that
would
include
meteorological
data
(
i.
e.,
where
available
5
years
of
data
should
be
used),
emissions
reported
for
the
same
meteorological
years,
documented
source
parameters,
model
physical
parameters,
assumed
background
concentrations
for
ozone
and
ammonia,
etc.
See
comment
for
more
detail.
Also,
in
order
to
produce
consistent
BART
determinations
among
various
sources
and
areas,
EPA
should
develop
its
own
evaluation
protocol
for
both
shorter
and
longer
distances.
EPA
should
continuously
revise
modeling
protocols,
considering
new
models
for
use
such
as
CMAQ,
by
providing
a
modeling
clearinghouse
to
states.
Commenter
0267
said
that
consistent,
nationally
applicable
guidance
is
essential,
and
that
once
it
is
developed,
virtually
no
deviations
should
be
allowed.
The
commenter
added
that
the
CALPUFF
modeling
exercises
should
follow
IWAQM
Phase
II
recommendations,
but
EPA,
in
consultation
with
the
FLMs
and
States,
should
also
publish
additional
guidance
to
address
more
recent
issues
such
as
particle
speciation,
emission
rate
averaging
times,
and
"
natural
obscuration."
This
is
also
an
appropriate
time
to
create
regional
modeling
platforms
for
CALPUFF,
which
would
allow
states
and
sources
to
run
the
model
more
expeditiously,
and
consistently.

Response:

The
preamble
discussion
in
the
Modeling
Protocol
discussion
address
each
of
these
comments.

Commenter
0304
said
all
of
the
screening
approaches
mentioned
in
the
re­
proposal
are
problematic
and
should
be
dropped
because
important
factors
are
ignored
including
terrain,
local
meteorological
data
and
differences
in
release
parameters.

Response:
89
The
preamble
discussion
in
the
Alternatives
for
determining
visibility
impacts
from
individual
sources
section
addresses
these
comments.

Commenters
(
0230,
0267,
0304)
said
a
modeling
protocol
should
be
required
of
all
states
and
stakeholders
who
are
performing
the
BART
modeling
analysis
in
all
cases
in
order
for
all
interested
parties
(
e.
g.,
EPA,
States,
Federal
Land
Managers,
environmentalists,
stakeholders,
etc.)
to
have
a
chance
to
understand
the
modeling
approach,
how
the
results
will
be
used,
and
allow
comments
on
the
procedures
prior
to
the
publication
of
the
final
results.
Commenter
0267
said
modeling
protocols
should
be
developed
for
all
modeling
demonstrations,
but
agreed
with
the
exemption
proposed
for
sources
that
are
less
than
200
km
from
a
Class
I
area
if
a
detailed
generic
national
protocol
is
developed
and
sources
are
required
to
abide
by
it.
Commenter
0314
said
a
protocol
should
be
required
irrespective
of
the
distance
of
the
source
from
the
Class
I
area.
Commenter
0242
noted
a
concern
that
some
of
the
assumptions
build
into
the
default
CALPUFF
procedure
may
not
be
appropriate
for
the
Great
Lakes
area.
The
commenter
supports
the
use
of
modeling
protocols
to
work
out
acceptable
procedures
that
best
simulate
the
physical
and
chemical
processes
that
the
model
simulates.
Commenter
0304
added
it
would
be
helpful
for
EPA
to
establish
a
national
procedure
for
these
studies,
including
a
rnethodology
for
establishing
natural
background
conditions,
background
ammonia
concentrations,
and
determining
sulfuric
acid
emission
rates.

Response:

States
should
prepare
or
require
their
BART­
eligible
sources
to
prepare
a
modeling
protocol
prior
to
conducting
any
modeling,
regardless
of
the
distance
from
the
source
and
the
Class
I
area.
This
is
reflected
in
the
rule.
The
Guidance
for
Estimating
Natural
Visibility
Conditions
Under
the
Regional
Haze
Rule
(
U.
S.
EPA.
September
2003)
provides
methodologies
for
determining
natural
background
conditions.
The
Interagency
Workgroup
on
Air
Quality
Modelig
(
IWAQM)
Phase
2
Summary
Report
and
Recommendations
for
Modeling
Long
Range
Transport
Impacts,
U.
S.
Environmental
Protection
Agency,
EPA­
454/
R­
98­
019,
December
1998
has
estimates
of
background
ammonia
concentrations.
Sulfuric
acid
emission
rates
may
be
determined
using
EPA­
approved
emission
factors.

Commenter
0321
said
it
is
inappropriate
for
EPA
to
have
more
than
an
advisory
role
when
it
comes
to
the
State
approving
CALPUFF
modeling
protocols
and
conducting
CALPUFF
assessments
of
air
quality
impacts.
Commenter
0323
recommended
that
EPA
consider
a
multiagency
process
to
reach
agreement
on
appropriate
modeling
protocol
prior
to
allowing
BART
applicability
and
control
determinations
to
be
based
on
model
results.
Such
a
process
could
reasonably
be
engaged
in
prior
to
deadlines
for
state
implementation
plans,
and
would
not
delay
implementation
of
the
BART
guidelines.

Response:
90
We
agree
that
we
should
have
only
an
advisory
role
in
development
of
the
protocol
as
the
States
better
understand
the
BART­
eligible
source
configurations
and
the
geophysical
and
meteorological
data
affecting
their
particular
Class
I
area(
s).
We
believe,
however,
that
EPA
should
have
final
approval
of
State
protocols
so
that
the
model,
and
the
BART
control
decision
which
depends,
in
part,
on
the
model
results,
is
applied
equitably
across
all
regions.
EPA
approval
will
help
ensure
that
vastly
different
modeling
results
are
not
produced
for
modeled
BART­
eligible
sources
that
have
similar
emission
configurations
and
operate
under
comparable
geophysical
and
meteorological
conditions.
EPA
encourages
a
multi­
agency
process.

Commenter
0201
said
that
EPA
should
provide
data,
perhaps
using
example
facilities,
to
demonstrate
the
effect
of
the
process
so
that
states
can
get
a
better
feeling
for
sources
that
are
likely
to
be
exemptable
under
the
proposal.
This
would
help
states
understand
the
net
effect
of
all
of
the
parameters
chosen
in
the
exemption
process.

Response:

We
have
provided
example
model
runs
along
with
a
discussion
in
the
preamble
and
the
guidelines.
We
also
encourage
States
to
develop
their
own
modeling
studies.

Commenter
0306
said
that
Part
111,
Section
1.
b.
i
­
The
Guideline
on
Air
Quality
Models
(
CFR
Part
51,
Appendix
W)
should
be
included
along
with
IWAQM
Report
as
a
reference
for
CALPUFF
setup.

Response:

We
agree
and
have
provided
this
reference
in
the
final
rule.

Comment:
Six
commenters
said
EPA
should
finalize
the
draft
regional
modeling
guidance
as
soon
as
possible
(
28­
267,
28­
255,
28­
278,
28­
348,
28­
360,
28­
387).
The
source
apportionment
model
is
needed
as
well
(
28­
348,
28­
387).
Commenter
28­
360
said
EPA
must
subject
the
guidance
to
peer
review
and
notice
that
guidance
for
public
comment
since
the
air
quality
analysis
portion
of
the
BART
determination
process
is
critical
in
ultimately
determining
which
sources
must
install
BART.
Furthermore,
EPA
must
refrain
from
applying
a
"
weight­
of­
evidence"
approach
in
the
air
quality
analysis.

Response:

One
option
for
determining
which
sources
are
applicable
to
a
BART
review
will
allow
States
to
base
that
decision
upon
an
individual
source
analysis
in
determining
if
a
source
causes
or
contributes
to
visibility
impairment,
The
use
of
a
dispersion
model
may
be
more
appropriate
than
a
photochemical
grid
model
in
making
that
determination.
States,
however,
do
have
the
option
to
91
make
that
decision
to
conduct
a
BART
review
of
BART­
eligible
sources,
on
a
cumulative
basis,
and
a
photochemical
model
may
be
used
on
a
case
by
case
basis
in
consultaion
with
EPA.
Once
a
source
is
subject
to
a
BART
review,
a
State
must
consider
five
factors
together
in
establishing
an
appropriate
control
technology,
if
any,
for
BART;
one
of
those
factors
is
the
visibility
improvement
expected
from
an
individual
source.
Evaluation
of
five
factors
taken
together,
is
not
unlike
a
weight
of
evidence
determination,
and
for
purposes
of
BART,
is
a
statutory
requirement.
Again,
a
dispersion
model
is
probably
more
appropriate
in
making
the
visibility
improvement
determination.
The
Guidance
For
Demonstrating
Attainment
Of
Air
Quality
Goals
For
PM2.5
And
Regional
Haze
,
draft
2.1,
January
2001is
intended
to
address,
in
part,
long
term
strategies
for
States
to
demonstrate,
via
photochemical
grid
modeling,
a
return
to
natural
conditions
as
part
of
their
SIPs.
BART
is
part
of
the
long
term
strategy.
The
guidance
will
likely
be
final
in
early
2006
and
will
assist
States
in
making
that
determination.

Commenter
28­
393
said
EPA
must
develop
guidance
that
lets
the
States
determine
individual
source
culpability
to
visibility
impairment.
It
is
unacceptable
to
simply
state
that
a
source
apportionment
modeling
tool
is
under
development
and
until
such
time
all
BART­
eligible
sources
are
subject
to
BART.

Response:

This
rule
and
guideline
allow
States
to
determine
if
individual
sources
cause
or
contribute
to
visibility
impairment.

3.2.4.4
Resources
Required
for
Modeling
Comment:
Two
commenters
discussed
the
resources
required
for
modeling
(
0276,
0299).
Commenter
0276
asked
not
to
be
involved
in
the
intricacies
of
modeling
and
contribution
analysis,
etc.
based
on
the
extensive
capital
and
operating
expenditures
associated
with
the
emission
reduction
program
of
CAIR.
Commenter
0299
said
the
proposed
BART
guidelines
will
impose
a
tremendous
resource
burden
on
states
because
of
the
number
of
sources
involved
and
the
resource
intensive
determinations
of
BART
for
eligible
sources.
An
easier
and
more
practical
approach
is
possible
(
e.
g.,
multi­
pollutant,
regional
or
source­
category
engineering
analyses).
Even
though
EPA
intends
to
coordinate
the
PM2.5
program
with
the
Regional
Haze
program.
EPA
also
must
recognize
the
simultaneous
burden
being
placed
on
states
to
develop
control
programs
to
comply
with
the
Regional
Haze
program
and
the
PM2.5
and
8­
hour
ozone
NAAQS,
and
other
multi­
pollutant
mandates.
Complying
with
the
health­
based
standards
must
take
precedence,
and
limited
state
resources
should
be
conserved
for
that
purpose.
EPA
should
provide
the
states
with
additional
resources
(
e.
g.,
funding)
to
implement
the
BART
requirements.

Response:
The
final
rule
acknowledges
the
States'
substantial
authority
in
making
BART
determinations
and
provides
useful
guidance
for
making
these
determinations.
92
3.2.4.5
Other
General
Comments
Regarding
the
Source
Exemption
Process
Comment:
Three
commenters
raised
issues
regarding
the
proposed
use
of
the
FLAG
approach
in
the
BART
program
(
0209,
0230,
0270).
Commenters
(
0209,
0270)
said
EPA
has
improperly
pulled
FLAG
into
the
debate
through
use
of
CALPUFF
modeling.
The
FLAG
modeling
regime
has
farreaching
impacts
because
it
established
new
"
acceptability"
values
and
metrics
for
determining
"
adverse
impacts"
and
"
limits
of
acceptable
change"
for
AQRVs.
Further,
it
was
approved
through
a
simple
"
policy
directive,"
rather
than
through
the
federal
rulemaking
process
required
by
the
federal
Administrative
Procedure
Act.
This
is
the
type
of
government
action
(
a
substantive
legal
rule
in
the
guise
of
a
"
guidance
document")
that
the
D.
C.
Circuit
Court
of
Appeals
has
repeatedly
overturned
because
of
due
process
circumvention.
The
commenter
urged
EPA
not
to
rely
on
the
improper
precedent
of
FLAG
as
the
basis
for
its
BART
guidance.
Commenter
0230
raised
concerns
related
to
evaluating
sources
that
are
thirty
to
forty
years
old
and
will
not
be
along
much
longer
using
the
FLAG
procedures
that
provide
conservative
estimates
whether
a
proposed
new
source
has
the
potential
to
contribute
to
visibility
impairment
sometime
in
the
future.
Thus,
different
procedures
should
be
used
to
assess
the
visibility
impacts
of
a
potential
BART­
eligible
source
that
is
3­
4
decades
older
than
a
new
source
that
accounts
for
current
visibility
conditions,
not
the
hypothetical
future­
year
natural
conditions
in
the
FLAG
approach.

Response:

FLAG
recommends
that
sources
applying
for
a
PSD
permit
use
CALPUFF
to
assess
visibility
impacts
for
purposes
of
determining
whether
a
proposed
source
is
likely
to
have
an
adverse
impact
on
visibility.
EPA
proposed
that
States
use
CALPUFF
to
make
individual
visibility
assessments
for
purposes
of
implementing
the
BART
provisions
of
the
regional
haze
rule
as
we
believe
that
it
is
currently
the
best
tool
for
making
such
assessments.
Add
sentence
about
how
it's
the
only
model
approved
for
single
source
transport
assessments
(
it's
in
the
preamble
and
OMB
wanted
it
out).
We
have
taken
comment
on
the
use
of
CALPUFF
for
making
BART
determinations
and
have
responded
to
the
specific
issues
raised
by
commenters
elsewhere
in
the
RTC
and/
or
the
preamble.
Based
on
our
consideration
of
these
comments,
we
continue
to
believe
that
CALPUFF
is
currently
the
best
tool
for
making
such
assessments.
The
final
BART
Guidelines
recommend
that
States
use
CALPUFF
but
do
not
require
States
to
do
so.
In
addition,
the
BART
Guidelines
provide
specific
recommendations
for
how
States
should
use
CALPUFF
and
evaluate
the
results.

Commenter
0170
noted
that
seasonally
operated
sources,
although
meeting
the
BARTeligibility
criteria,
may
not
contribute
to
regional
haze
because
they
do
not
operate
at
times
when
regional
haze
is
a
problem.
The
BART
guidelines
should
provide
a
mechanism
to
exempt
such
sources
from
the
BART
engineering
analysis
and
control
requirements.

Response:
93
The
final
rule
provides
for
an
exemption
mechanism
by
which
a
State
determines
if
a
BART­
eligible
sources
causes
or
contributes
to
visibility
impairment.
A
source
which
contributes
to
visibility
impairment
seasonally
would
still
be
reasonably
anticipated
to
cause
or
contribute
to
"
any
visibility
impairment."
However,
the
State
may
take
into
account
such
factors
as
a
seasonal
source
in
making
its
BART
determination.
The
Guideline
also
suggests
that
States
may
adopt
a
more
streamlined
approach
to
the
BART
determination
when
one
or
more
factors
clearly
outweighs
others.

Commenter
0238
said
that
EPA
should
specify
in
the
final
guidelines
that,
in
their
individual
source
assessments,
Eastern
sources
should
be
able
to
model
for
visibility
impairment
assuming
that
the
CAIR
is
in
effect.

Response:

A
State
that
participates
in
the
CAIR
program
may
use
the
CAIR
provisions
to
satisfy
the
BART
requirements
for
EGUs
and
other
CAIR
affected
sources.
States
making
individual
source
assessments
for
BART
should
model
each
source's
contribution
to
visibility
as
compared
to
natural
background
conditions.

Comment:
Two
commenters
opposed
shifting
the
burden
to
states
to
prove
that
a
particular
source
does
not
contribute
to
visibility
impairment
(
28­
340,
28­
344).
The
states
do
not
have
the
resources
or
tools
to
prove
the
negative.

Response:

The
final
Guidelines
provide
that
States
may
conclude
that
all
BART
sources
within
a
state
are
reasonably
anticipated
to
cause
or
contribute
to
visibility
impairment
based
on
the
State's
assessment
of
the
collective
contribution
of
its
sources
to
regional
haze.
Thus,
the
BART
Guidelines
do
not
require
States
to
prove
the
negative.

3.3
Alternative
Approaches
3.3.1
Screening
Assessment
Using
CALPUFF
Comment:
Eleven
commenters
supported
performing
screening
assessments
using
CALPUFF
as
a
screening
model
(
0195,
0198,
0228,
0236,
0242,
0254,
0256,
0277,
0306,
0307,
0314).
Commenter
0195
added
that
the
following
conditions
should
apply:
(
1)
Sources
may
use
other
acceptable
models
including
updates/
variations
of
CALPUFF.
(
2)
Sources
must
have
the
ability
to
influence
the
inputs
to
CALPUFF
or
any
other
approved
model.
(
3)
Its
use
does
not
become
a
way
for
FLMs
to
implement
the
more
aggressive
FLAG
guidance.
Other
commenters
stressed
94
that
more
refined
modeling
can
be
done
at
the
source's
discretion
if
the
screening
model
indicates
impacts
above
the
impairment
threshold
level
(
0228,
0242,
0254,
0256,
0277,
0314).

Response:

The
BART
Guidelines
provide
that
other
appropriate
models
may
be
used
and
that
States
should
develop
modeling
protocols.
We
recommend
that
the
States
consult
with
EPA
on
the
use
of
other
models
­
and
on
the
modeling
protocols
used
­
before
any
modeling
demonstration.
CALPUFF
used
with
limited
meteorological
data
(
i.
e.
in
a
simpler
screening
mode)
may
not
be
as
conservative
as
CALPUFF
using
a
full
meteorological
domain,
and
should
not
be
used
to
exempt
a
source
from
BART
review
without
more
rigorous
evaluations
and
sensitivity
tests
showing
that
the
results
are
at
least
as
conservative
as
the
CALPUFF
model.
The
modeling
guidance
for
purposes
of
BART
is
not
related
to
the
implementation
of
FLAG.

Comment:
Three
commenters
were
concerned
about
using
CALPUFF
as
a
screening
model
(
0179,
0227,
0302).
Commenter
0179
said
that
in
order
to
ensure
accurate
assessment
of
potential
impacts,
EPA
should
use
the
complete
CALPUFF
assessment
model
(
or
equally
rigorous
models
that
are
appropriate
for
shorter
distances)
rather
than
less
rigorous
models
that
do
not
take
into
account
local
variations
in
meteorology
and
terrain.
Commenter
0227
cited
difficulties
with
the
flat
terrain
assumption
in
the
screening
mode,
the
limits
of
using
ISC
meteorology
(
instead
of
CALMET
refined
estimates),
and
underestimates
of
sulfate
concentrations
as
issues
in
using
the
screening
mode.
The
commenter
said
EPA
has
not
provided
enough
information
about
the
sensitivities
and
accuracy
of
the
screening
mode
of
CALPUFF
to
the
public.
Therefore,
until
further
analysis
has
been
done
and
the
approach
validated,
CALPUFF
screening
mode
should
not
be
used
to
determine
if
BART­
eligible
can
be
exempt
from
being
subject
to
BART.
Nor
should
other
alternatives
based
on
the
screening
mode
or
other
general
assumptions
be
used.

Response:

We
agree.

Comment:
Two
commenters
asked
EPA
to
develop
additional
screening
tools
(
0201,
0234).
Commenter
0201
encouraged
EPA
to
further
develop
the
screening
analysis
tools
proposed
to
maintain
national
consistency
in
their
use
before
the
guidance
is
finalized.
EPA
should
also
demonstrate
the
use
of
the
tools
through
application
to
hypothetical,
a
subset
of
actual
sources,
or
ideally
all
national
sources.
Commenter
0234
agreed
that
CALPUFF
screening
is
a
good
approach
to
the
exemption
process,
but
noted
that
it
could
be
resource
intensive
for
states
with
multiple
BART
eligible
sources.
A
method
for
modeling
units
with
multiple
emission
limits
should
be
developed.
EPA
should
also
include
up­
to­
date
references
in
the
actual
rule
and/
or
develop
modeling
guidance
for
the
states.
95
Response:

Modeling
alternatives
may
be
inappropriate
because,
in
their
existing
form,
they
do
not
account
for
important
factors
such
as
terrain,
local
meteorological
data,
prevailing
wind
directions
(
which
influence
pollutant
transport),
and
differences
in
stack
release
parameters.
There
is
no
direct
connection
between
emissions,
distance,
and
visibility
impairment.
We
believe
that
alternatives
should
not
be
used
to
exempt
a
source
from
BART
review
without
more
rigorous
evaluations
and
sensitivity
tests
showing
that
the
results
are
at
least
as
conservative
as
the
CALPUFF
model.
While
we
are
not
adopting
in
the
guideline
any
specific
alternative
to
modeling
for
large
power
plants,
a
State,
rather
than
EPA,
may
develop
its
own
alternative
approach
for
the
other
source
categories
to
determine
if
a
source
would
be
subject
to
BART,
provided
that
the
alternative
demonstrates
a
sufficient
basis
to
determine
clearly
that
the
source
causes
or
contributes
to
visibility
impairment,
or
that
more
refined
analysis
is
warranted.

3.3.2
Look­
up
Tables
Comment:
Seven
commenters
addressed
the
utility
of
look­
up
tables
(
0201,
0231,
0234,
0236,
0254,
0307,
0323).

Three
commenters
supported
the
use
of
look­
up
tables
as
a
screening
method
(
0201,
0307,
0323).
Commenter
0323
added
that
using
a
threshold
as
the
basis
for
look­
up
tables
applied
to
generic
screening
level
air
quality
model
runs
would
help
simplify
the
state
selection
process
and
be
a
useful
addition
to
the
guideline.
Three
commenters
urged
EPA
to
develop
lookup
tables
(
0231,
0234,
0307).
Commenter
0302
noted
that
EPA
would
need
to
commit
to
future
revisions
of
the
tables
as
data
and
technology
improve.

Two
commenters
said
that
look­
up
tables
are
inappropriate
(
0236,
0254)
because
they
may
exempt
sources
that
should
be
BART
eligible,
sacrifice
technical
accuracy
and
may
limit
cost
effectiveness
given
the
need
to
run
screening
models
for
the
varying
emission
characteristics
of
many
different
emission
units
within
each
source
category.

Response:

We
agree
that
look­
up
tables
may
be
inappropriate
for
the
reasons
stated
above.
Again,
we
emphasize
that
alternatives,
including
look­
up
tables
should
not
be
used
to
exempt
a
source
from
BART
review
without
more
rigorous
evaluations
and
sensitivity
tests
showing
that
the
results
are
at
least
as
conservative
as
the
CALPUFF
model.

3.3.3
Source
Ranking
Comment:
96
Four
commenters
opposed
using
a
source
ranking
approach
and/
or
the
Q/
D
approaches
(
0234,
0236,
0254,
0302)
because
of
the
following
concerns:
°
The
approaches
are
quite
arbitrary
and
do
not
take
transport
and
plume
characteristics
into
account
(
0234).
°
Both
techniques
could
potentially
be
applied
to
sources
beyond
300
km
of
a
Class
I
area.
This
possibility
poses
a
problem
because
there
is
no
process
to
confirm
that
an
individual
source
can
significantly
affect
visibility
at
this
distance
(
0234).
°
There
is
no
direct
connection
of
the
emission
metric
to
visibility
impairment
using
the
source
ranking
approach
(
0236).
°
The
source
ranking
approach
treats
SO
2
and
NO
X
equally
for
visibility
impairment
(
0236)
°
The
source
ranking
approach
treats
the
distance
from
a
Class
I
area
equally
and
does
not
consider
the
source­
receptor
geometry
and
prevailing
wind
directions
(
0236,
0254)

Response:

Modeling
alternatives
may
be
inappropriate
because,
in
their
existing
form,
they
do
not
account
for
important
factors
such
as
terrain,
local
meteorological
data,
prevailing
wind
directions
(
which
influence
pollutant
transport),
and
differences
in
stack
release
parameters.
There
is
no
direct
connection
between
emissions,
distance,
and
visibility
impairment.
We
believe
that
alternatives
should
not
be
used
to
exempt
a
source
from
BART
review
without
more
rigorous
evaluations
and
sensitivity
tests
showing
that
the
results
are
at
least
as
conservative
as
the
CALPUFF
model.
While
we
are
not
adopting
in
the
guideline
any
specific
alternative
to
modeling
for
large
power
plants,
a
State,
rather
than
EPA,
may
develop
its
own
alternative
approach
for
the
other
source
categories
to
determine
if
a
source
would
be
subject
to
BART,
provided
that
the
alternative
demonstrates
a
sufficient
basis
to
determine
clearly
that
the
source
causes
or
contributes
to
visibility
impairment,
or
that
more
refined
analysis
is
warranted.

Commenter
0254
added
that
EPA
should
include
guidelines
for
identifying
which
Class
I
area(
s)
ought
to
be
modeled
(
e.
g.,
the
nearest
downwind
area)
during
the
screening
and
refined
modeling
assessments.

Response:

We
include
guidelines
under
the
Modeling
Protocol
discussion.

Commenter
0302
objected
because
this
option
does
not
use
modeling.
It
appears
to
be
the
easiest
to
use
but
hardest
to
defend.
States,
in
essence,
"
draw
a
line"
so
that
sources
below
a
certain
level
are
considered
exempt.
Further
additional
information
is
necessary
on
the
proper
choice
of
a
common
metric
to
use.

Response:

See
the
discussion
above
regarding
the
use
of
alternatives.
97
Comment:
Regarding
the
"
source
ranking"
and
"
Q/
D"
criteria
ranking,
commenter
0323
urged
caution
in
applying
these
criteria
using
annual
emissions
only
because
this
may
mask
peak
or
seasonal
impacts
that
affect
shorter­
term
visibility
impairment.
The
commenter
encouraged
EPA
to
factor
variations
in
emissions
into
the
ranking
methods
if
they
choose
to
offer
such
methods.

Response:

See
the
discussion
above
regarding
the
use
of
alternatives.

3.3.4
Emissions
Divided
by
Distance
(
Q/
D)

Comment:
Two
commenters
supported
the
Q/
D
approach
(
0231,
0302).
Commenter
0231
said
an
advantage
of
the
method
is
that
it
doesn't
require
emissions
data
from
all
sources
in
other
areas
to
determine
whether
a
source
is
exempt.
It
is
also
far
less
resource
intensive
than
the
CALPUFF
modeling
options.
Based
upon
recent
comparison
work
completed
by
MANE­
VU,
it
appears
that
the
emissions
over
distance
method
produces
results
that
correlate
well
with
results
produced
using
the
CALPUFF
modeling
approach.
Commenter
0302
said
the
approach
could
work
very
well
as
a
first­
cut
determination
and,
in
concert
with
any
of
the
three
previous
methods,
could
further
determine
exemption
status.

Three
commenters
opposed
using
the
method
(
0234,
0236,
0254)
for
the
same
reasons
they
opposed
the
source
ranking
approach
described
in
section
3.3.3.,
above.

Response:

See
the
discussion
above
regarding
the
use
of
alternatives.

Comment:
One
commenter
(
0306)
noted
that
the
preamble
discussion
contained
some
errors
on
p.
25196.
It
states
"
The
Q/
D
method
determines
a
source
to
be
insignificant
if
the
allowable
emissions...
divided
by
a
constant
times
the
distance...
is
greater
than
a
value
of
1...
Therefore,
a
source
could
be
considered
insignificant
if
its
emissions
divided
by
20
times
its
distance...
is
greater
than
1.
The
commenter
said
"
greater"
should
be
replaced
with
'
less"
in
both
instances.

Regarding
the
"
source
ranking"
and
"
Q/
D"
criteria
ranking,
commenter
0323
urged
caution
in
applying
these
criteria
using
annual
emissions
only
because
this
may
mask
peak
or
seasonal
impacts
that
affect
shorter­
term
visibility
impairment.
The
commenter
encouraged
EPA
to
factor
variations
in
emissions
into
the
ranking
methods
if
they
choose
to
offer
such
methods.

Response:
98
See
the
discussion
above
regarding
the
use
of
alternatives.

3.3.5
General
Comments
Regarding
Alternative
Approaches
Comment:
Ten
commenters
supported
the
use
of
alternative
screen
approaches
(
0183,
0186,
0189,
0195,
0217,
0226,
0231,
0241,
0275,
0277).
Commenters
(
0186,
0195,
0275,
0277)
said
additional
screening
tools
should
be
developed.
Commenter
0241
said
a
state
should
have
the
authority
to
adopt
any
of
the
proposed
approaches
into
its
SIP
for
determining
whether
a
BARTeligible
source
may
be
exempt
from
BART.
Commenters
(
0195,
0226)
added
that
sources
should
not
be
precluded
from
the
use
of
an
intensive
detailed
model
by
the
affected
source.

Response:

See
the
discussion
above
regarding
the
use
of
alternatives.
Sources
are
not
precluded
from
using
a
detailed
model,
provided
they
submit
a
modeling
protocol.

Three
commenters
opposed
the
use
of
screening
models
(
0179,
0221,
0227).
Commenter
0179
said
the
proposed
approaches
would
give
a
less
accurate
assessment
of
potential
impacts
and
therefore
have
a
negative
impact
on
visibility
in
Class
I
areas.
Commenter
0221
said
EPA
has
not
demonstrated
that
any
of
the
screening
alternatives
is
at
least
as
protective
in
assessing
source­
receptor
relationships
as
CALPUFF.
Commenter
0227
added
an
individual
pollution
source
that
meets
the
initial
BART
criteria
should
not
be
given
an
exemption
from
BART
unless
a
rigorous
modeling
analysis
demonstrates
that
it
does
not
cause
or
contribute
to
haze
in
any
Class
I
area.
Adequate
air
quality
models
must
be
used
in
this
analysis.

Response:

See
the
discussion
above
regarding
the
use
of
alternatives.

Three
commenters
said
additional
information
is
needed
to
comment
on
the
alternative
approaches
(
0221,
0267,
0302).
For
example,
commenter
0221
was
seeking
an
analysis
of
the
sensitivity
of
results
to
assumptions
and
commenter
0302
wanted
more
detail,
with
examples,
of
how
each
proposal
would
work.

Response:

See
the
discussion
above
regarding
the
use
of
alternatives.

Comment:
Eight
commenters
provided
other
comments
on
the
use
of
alternative
approaches
(
0183,
0221,
0234,
0237,
0259,
0267,
0307).
Some
commenters
made
implementation
suggestions.
Commenter
0183
said
the
screening
techniques
should
be
designed
such
that
a
source
would
not
99
pass
one
screening
technique
and
fail
another.
Alternatively,
a
source
could
be
required
to
pass
at
least
two
of
the
suggested
screening
techniques
in
order
to
be
excluded
from
BART.
Since
the
screening
techniques
would
be
independent
tests,
passing
any
two
of
the
tests
would
lend
additional
credibility
for
exclusion.
Naturally,
a
source
could
choose
to
proceed
with
full­
blown
CALPUFF
to
determine
the
level
of
control
if
they
do
not
anticipate
screening
out.

Response:

See
the
discussion
above
regarding
the
use
of
alternatives.
More
detailed
modeling
is
allowed.

Commenter
0234
preferred
an
exemption
process
that
is
relatively
simple
and
definitive,
but
it
must
hold
sources
to
a
very
high
standard
in
order
to
declare
them
as
making
no
contribution
to
visibility
impairment.
Commenter
0237
also
preferred
a
simpler
methodology.
Commenter
0234
suggested
that
EPA
may
want
to
consider
combining
some
of
its
methods
and
creating
a
tiered
approach.

Response:

See
the
discussion
above
regarding
the
use
of
alternatives.

Commenter
0307
suggested
a
hierarchical
approach.
De
minimis
emission
levels
should
be
set
for
each
pollutant
at
as
high
a
value
as
can
be
established
with
confidence.
Then
individual
BART
eligible
sources
with
emissions
that
exceed
a
de
minimis
threshold
would
be
able
to
use
a
simple
tool
such
as
look
up
tables
to
see
if
those
emissions
may
contribute
to
visibility
impairment.
If
the
simple
tool
indicates
impairment
is
possible,
then
CALPUFF
screening
or
other
modeling
could
be
employed
as
the
next
step.
Finally,
if
CALPUFF
screening
or
other
modeling
indicates
the
source
may
be
contributing
to
visibility
impairment,
full
CALPUFF
modeling
could
be
used.

Response:

See
the
discussion
above
regarding
the
use
of
alternatives.

Commenter
0259
said
states
and
tribes
should
have
discretion
in
deciding
whether
the
alternative(
s)
would
be
used
in
lieu
of
or
as
a
first
step
in
a
full
CALPUFF
assessment.
If
used
in
lieu
of
a
full
CALPUFF
assessment,
the
simplified
alternative
should
be
no
more
likely
to
exempt
a
source
than
a
full
CALPUFF
assessment.

Response:

See
the
discussion
above
regarding
the
use
of
alternatives.
100
Commenter
0221
recognized
the
need
to
address
CALPUFF's
limitations
to
determine
the
contribution
of
VOC
sources.
The
commenter
recommended
the
Q/
D
method,
followed
by
the
source
ranking
method
until
better
modeling
techniques
for
VOC
are
developed.

Response:

See
the
discussion
above
regarding
the
use
of
alternatives.
We
believe
that
the
best
approach
for
States
to
follow
in
considering
whether
VOC
emissions
are
precursors
to
PM
2.5
formation
is
a
case­
by­
case
approach.
States
should
consider,
in
particular,
whether
a
source's
VOC
emissions
are
those
higher­
carbon
VOCs
that
are
more
likely
to
form
secondary
organic
aerosols.
In
addition,
given
the
variable
contribution
of
a
given
amount
of
VOC
emissions
to
PM
2.5
formation,
States
may
also
wish
to
exercise
discretion
in
considering
only
relatively
larger
VOC
sources
to
be
BART­
eligible.

Commenter
0267
said
maximum
short
term
emissions
should
be
used
in
these
approaches.

Response:

See
the
discussion
above
regarding
the
use
of
alternatives.

3.4
Visibility
Thresholds
3.4.1
Use
of
Deciview
as
a
Metric
Comment:
Two
commenters
opposed
using
deciview
as
the
measure
of
visibility
impairment
(
0197,
0214).
Commenter
0197
said
this
proposal
is
a
substantial
departure
from
EPA's
existing
visibility
regulations
that
are
based
on
state
decision
making
and
the
basis
for
the
change
is
not
explained.
Commenter
0214
said
the
0.5
deciview
value
is
inconsistent,
with
EPA's
finding
in
the
regional
haze
rule
that
one
deciview
is
the
appropriate
perceptibility
threshold.
Further,
the
proposal
would
substantially
change
the
"
deciview"
metric
by
converting
it
into
a
regulatory
0.5
deciview
standard
(
versus
a
"
goal")
for
defining
how
states
must
exercise
their
authority
and
discretion
in
determining
whether
an
individual
source
"
causes
or
contributes"
to
visibility
impairment
at
an
individual
Class
I
area.

Response:

We
believe
that
it
is
appropriate
for
States
to
use
the
deciview
in
establishing
a
threshold
for
BART
applicability.
CAA
169A(
b)(
2)(
A)
provides
for
a
BART
determination
for
any
BARTeligible
source
which
"
emits
any
air
pollutant
which
may
reasonably
be
anticipated
to
cause
or
contribute
to
any
impairment
of
visibility
in
any
[
Class
I]
area."
States
will
need
some
means
of
determining
whether
a
source
"
causes
or
contributes
to"
visibility
impairment.
The
guidelines
advise
States
to
use
a
deciview
metric
in
defining
"
cause
or
contribute",
as
explained
further
101
below.
The
fact
that
the
deciview
is
also
used
to
track
progress
toward
the
goal
of
natural
visibility
does
not
in
any
way
indicate
that
we
are
"
converting"
a
"
goal"
into
a
requirement.
Use
of
the
same
metric
in
the
"
cause
or
contribute"
context
as
used
for
establishing
reasonable
progress
goals,
tracking
changes
in
visibility
conditions,
and
defining
baseline,
current,
and
natural
conditions
simply
provides
for
a
consistent
approach
to
quantifying
visibility
impairment.
Specifically,
we
wish
to
clarify
here
that
we
have
concluded
that
for
purposes
of
determining
those
sources
subject
to
BART,
a
single
source
that
is
responsible
for
a
1.0
deciview
change
or
more
should
be
considered
to
"
cause"
visibility
impairment.
Additionally,
we
are
establishing
a
change
of
0.5
deciviews
as
the
maximum
visibility
change
we
believe
should
be
used
to
determine
whether
a
source
"
contributes"
to
impairment
for
the
following
reasons.
We
acknowledge
that
there
is
uncertainty
associated
with
identifying
an
absolute
deciview
change
as
"
just
noticeable,"
or
"
perceptible".
This
is
consistent
with
our
observations
in
the
preamble
to
the
1999
Regional
Haze
Rule,
where
we
noted
"
that
a
1.0
deciview
change
should
not
be
considered
the
threshold
of
perception
in
all
cases
for
all
scenes",
and
that
"
visibility
changes
of
less
than
1.0
deciview
are
likely
to
be
perceptible
in
some
cases,
especially
where
the
scene
being
viewed
is
highly
sensitive
to
small
amounts
of
pollution".
We
also
acknowledged
that
a
change
greater
than
1.0
deciview
may
be
required
for
perceptibility
under
certain
site­
specific
conditions.
However,
in
most
cases,
a
1.0
deciview
change
is
considered
to
be
perceptible.

One
commenter
(
0221)
said
the
deciview
metric
is
useful,
but
not
comprehensive,
in
capturing
the
full
range
of
human
judgement
in
perceiving
visual
air
quality.

Response:

We
agree.

3.4.2
Compare
to
Visibility
Under
Natural
Conditions
3.4.2.1
Support
Use
of
the
"
Natural
Conditions"
Baseline
Comment:
Three
commenters
agreed
that
visibility
improvements
should
be
measured
relative
to
natural
background
(
0221,
0232,
0234).
Commenter
0221
noted
that
the
designation
of
a
visibility
baseline
is
extremely
important
in
implementing
the
statutory
goals
and
requirements
because
of
EPA's
proposed
use
of
the
deciview
scale
to
measure
visibility
impacts.
Any
increase
in
the
baseline
beyond
natural
visibility
conditions
will
thus
unlawfully
distort
and
weaken
the
BART
requirement
by
effectively
raising
the
applicability
threshold
 
notwithstanding
the
fact
that
these
sources
clearly
are
contributing
to
the
very
manmade
visibility
impairment
that
the
Act
is
explicitly
designed
to
remedy.
Commenter
0232
said
measuring
against
natural
background
rather
than
relative
to
baseline
conditions
is
a
more
appropriate
approach
given
that
our
planning
goal
is
to
achieve
natural
background
by
the
end
of
the
program.
The
commenter
said
specific
statutory
language
related
to
this
issue
should
be
included
in
the
preamble.
10
See
also
our
explanation
of
the
CAA
goal
provided
in
the
regional
haze
rule
at
64
FR
126.

102
Response:

We
agree
and
have
discussed
the
language
in
the
preamble.

Two
commenters
said
a
baseline
of
natural
background
visibility
conditions
is
needed
to
assure
"
reasonable
progress"
in
attaining
natural
conditions
(
0221,
0271).
Commenter
0221
added
that
a
different
baseline
would
also
arbitrarily
and
illogically
make
the
applicability
threshold
less
protective
in
highly
polluted
Class
I
areas.
Not
only
would
this
create
inequities
among
the
regulated
sources
but
it
would
severely
undermine
EPA's
regulatory
responsibility
in
carrying
out
the
BART
requirement.

Response:

We
agree.

3.4.2.2
Oppose
Use
of
the
"
Natural
Conditions"
Baseline
Comment:
Nine
commenters
said
that
a
return
to
visibility
under
natural
conditions
is
unattainable
(
0185,
0198,
0228,
0250,
0253,
0254,
0256,
0279,
0319).
Commenters
said
that
changes
should
be
compared
to
currently
existing
conditions
because
true
"
natural
conditions"
cannot
be
verified,
do
not
account
for
manmade
emissions
from
other
countries,
and
are
not
a
realistic
target
for
improvement.
Commenters
(
0198,
0214,
0217,
0228,
0238,
0256)
said
natural
conditions
are
a
"
goal"
that
is
a
benchmark
that
is
relevant
to
the
states'
determination,
under
the
regional
haze
program,
of
the
level
of
"
reasonable
progress"
to
achieve
in
haze
levels,
there
is
no
legal
requirement
(
and
there
could
not
be
a
legal
requirement)
that
the
natural
conditions
goal
ultimately
must
be
achieved.

Response:

We
disagree
that
the
use
of
natural
conditions
as
the
baseline
for
making
visibility
impact
determinations
is
inappropriate.
The
visibility
goal
of
the
CAA
is
both
the
remedying
of
existing
impairment,
and
prevention
of
future
impairment.
The
court,
in
American
Corn
Growers,
upheld
our
interpretation
of
that
goal
as
the
return
to
natural
visibility
conditions.
10
Long­
term
regional
haze
strategies
are
developed
to
make
"
reasonable
progress"
towards
the
CAA
goal,
and
States
must
demonstrate
reasonable
progress
in
their
regional
haze
State
implementation
plans
(
SIPs).
Since
the
BART
program
is
one
component
of
that
demonstration,
visibility
changes
due
to
BART
are
appropriately
measured
against
the
target
of
natural
conditions.

Comment:
103
Three
commenters
said
that
current
visibility
conditions
make
more
sense
as
a
baseline
because
sources
subject
to
BART
today
will
likely
not
be
in
operation
in
the
2064
timeframe
(
0256,
0259,
0270).
Commenter
0256
added
that
using
current
visibility
conditions
for
the
analysis
will
give
a
more
realistic,
real
world
prediction
of
whether
controlling
the
source
pursuant
to
BART
will
actually
improve
visibility.
Congress
did
not
intend
sources
to
have
to
consider
retrofitting
controls
under
the
BART
provision
if
those
sources
currently
are
not
impacting
realworld
visibility.

Response:

See
the
discussion
above
regarding
the
use
of
natural
conditions
as
the
baseline
for
comparison.
In
addition,
using
current
conditions
would
not
provide
a
realistic
prediction
of
a
source's
impacts
at
the
time
BART
controls
would
be
installed
or
beyond.
The
sources
addressed
by
the
BART
rule
have
been
in
operation
for
more
than
28
years
and
many
are
likely
to
continuing
operating
for
many
years
to
come.
Over
that
time
period,
emissions
of
pollutants
are
likely
to
continue
to
fall
as
programs
such
as
CAIR
come
into
effect.
Thus,
current
conditions
would
not
be
an
appropriate
measure
in
any
event.

Comment:
Four
commenters
discussed
the
relation
between
the
baseline
conditions
and
the
use
of
the
deciview
metric
(
0217
0256,
0270,
0321).
Commenters
(
0217,
0256,
0270)
noted
that
because
of
the
logarithmic
nature
of
the
deciview
scale
(
which,
in
essence,
describes
percentage
change),
a
given
amount
of
pollution
(
and
its
corresponding
increment
of
light
extinction)
will
cause
a
much
larger
change
in
percent
or
deciviews
relative
to
natural
conditions
than
it
would
cause
when
compared
to
more­
polluted
current
conditions.
Compared
to
actual
current
conditions,
the
0.5
deciview
(
or
5%)
threshold
relative
to
natural
conditions
may
only
correspond
to
an
actual
change
of
at
most
3%,
but
more
likely
1%,
or
even
less
­­
hardly
a
perceptible
visibility
threshold.
Should
EPA
persist
with
using
natural
background,
it
is
critical
that
the
correct
value
for
defining
natural
conditions
be
used.

Response:

Using
existing
conditions
as
the
baseline
for
single
source
visibility
impact
determinations
would
be
fundamentally
at
odds
with
the
low
triggering
threshold
for
BART
review,
because
it
would
create
the
following
untenable
paradox:
the
dirtier
the
existing
air,
the
less
likely
it
would
be
that
any
control
is
required
from
any
one
source.
This
is
true
because
of
the
nonlinear
nature
of
visibility
impairment.
In
other
words,
as
a
Class
I
area
becomes
more
polluted,
any
individual
source's
contribution
to
changes
in
impairment
becomes
geometrically
less.
Therefore
the
more
polluted
the
Class
I
area
would
become,
the
less
control
would
seem
to
be
needed
from
an
individual
source.
We
agree
that
this
kind
of
calculation
would
essentially
raise
the
"
cause
or
contribute"
applicability
threshold
to
a
level
that
would
never
allow
enough
emission
control
to
significantly
improve
visibility.
Such
a
reading
would
render
the
visibility
provisions
meaningless,
as
EPA
and
the
States
would
be
prevented
from
assuring
"
reasonable
progress"
and
fulfilling
the
104
statutorily­
defined
goals
of
the
visibility
program.
Conversely,
measuring
improvement
against
clean
conditions
would
ensure
reasonable
progress
toward
those
clean
conditions.

Comment:
Five
commenters
said
that
a
comparison
to
natural
conditions
fails
to
consider
other
sources
of
visibility
impairment
(
0217,
0230,
0246,
0270,
0321).
Commenters
said
many
sources
causing
impairment
are
outside
the
regulatory
purview
of
the
BART
program.
Other
sources
of
impairment
include:

°
Mobile
sources.
°
The
effects
of
naturally
occurring
sea
salt
(
sodium
chloride)
from
the
ocean
and
wind
blown
dust
off
of
alkaline
flats
in
the
west.
°
The
effects
of
visibility
obscuration
due
to
naturally
occurring
wildfires
that
have
been
suppressed
over
the
last
several
decades
due
to
human
intervention.
°
The
effects
of
naturally
occurring
wind
blown
dust,
geogenic
activity,
seeps,
earthquakes,
severe
weather
phenomena
(
tornados,
hurricanes,
etc.),
and
non­
human
living
creature
emissions
(
e.
g.,
ammonia,
soil
disturbance,
etc.).
°
The
effects
of
naturally
occurring
meteorological
phenomena
of
fog,
clouds,
rain,
snow,
etc.
that
obscure
visibility.

Response:

As
provided
for
under
our
Guidance
for
Estimating
Natural
Visibility
Conditions
Under
the
Regional
Haze
Rule
(
U.
S.
EPA.
September
2003).
http://
www.
epa.
gov/
ttncaaa1/
t1/
memoranda/
rh_
envcurhr_
gd.
pdf,
proper
accounting
for
international
emissions
and
natural
phenomena
is
in
the
5
year
SIP
progress
report,
not
in
the
setting
of
natural
visibility
estimates.

Commenter
0321
said
the
states
and
EPA
have
acquired
substantial
ambient
monitoring
data
from
Interagency
Monitoring
of
Protected
Visual
Environments
(
IMPROVE)
monitors,
as
well
as
other
standard
monitoring
equipment
located
within
and
outside
of
Class
I
areas.
A
methodology
should
be
established
to
use
this
data
in
arriving
at
a
benchmark
more
reasonable
than
natural
conditions.
In
North
Dakota,
and
other
northern
tier
states,
emissions
from
Canadian
sources
not
subject
to
EPA
rules
contribute
to
the
impact
of
air
quality
measured
at
these
monitoring
sites.
It
is
not
reasonable
to
expect
North
Dakota
EGUs,
and
eventually
other
sources
through
reasonable
progress
goals
every
five
years,
to
be
reducing
emissions
in
an
attempt
to
achieve
unrealistic
natural
conditions,
when
our
air
quality
and
monitoring
network
will
continue
to
be
influenced
by
Canadian
man­
made
emissions
and
other
smaller
man­
made
contributions
(
agricultural
practices,
unpaved
roads,
controlled
burns,
etc.).

Response:
105
As
provided
for
under
our
Guidance
for
Estimating
Natural
Visibility
Conditions
Under
the
Regional
Haze
Rule
(
U.
S.
EPA.
September
2003).
http://
www.
epa.
gov/
ttncaaa1/
t1/
memoranda/
rh_
envcurhr_
gd.
pdf,
proper
accounting
for
international
emissions
and
natural
phenomena
is
in
the
5
year
SIP
progress
report,
not
in
the
setting
of
natural
visibility
estimates.

Comment:
Commenter
0254
said
given
that
these
background
estimates
are
just
that
­
estimates
­
and
they
will
be
revised
and
refined
over
time,
it
would
be
unwise
to
compare
impacts
and
improvements
to
a
moving
baseline,
whereby
depending
on
the
year
the
analysis
was
performed,
a
source
may
or
may
not
be
deemed
to
cause
or
contribute
to
visibility
impairment.

Response:

We
disagree.
EPA
has
provided
the
Guidance
for
Estimating
Natural
Visibility
Conditions
Under
the
Regional
Haze
Rule
(
U.
S.
EPA.
September
2003)
in
which
natural
background
conditions,
expressed
in
deciviews,
are
defined
for
each
Class
I
area.
Default
conditions
may
be
used
or
there
are
measures
to
develop
refined
estimates
of
natural
conditions.

Commenter
0270
said
EPA's
proposed
BART
Guidelines'
use
of
natural
background
visibility
conditions
(
projected
to
occur
by
2064)
as
the
baseline
for
BART
determinations
prior
to
2018
will
have
the
effect
of
inappropriately
accelerating
achieving
the
visibility
goal
of
the
Regional
Haze
program
to
2018
or
sooner.
The
desired
visibility
changes
emanating
from
application
of
BART
limitations
should
be
with
respect
to
current
conditions
and
not
the
natural
background
conditions
in
2064.
Even
if
additional
emission
reductions
were
to
be
required
from
those
sources
beyond
2018,
such
reductions
would
occur
through
subsequent
long
term
strategy
and
future
SIPS.
The
BART
process
is
required
to
compare
a
specific
source's
contribution
to
existing
visibility
impairment
and
the
costs
associated
with
lowering
that
contribution
to
a
desired
level;
in
other
words,
a
source
should
be
required
to
expend
resources
only
if
such
expenses
would
produce
a
perceptible
improvement
in
visibility
in
a
Class
I
area
under
current
conditions,
not
in
2064.

Response:

See
the
discussion
above
regarding
the
use
of
natural
conditions
as
the
baseline
for
comparison.
For
States
adopting
source
specific
BART
controls,
the
BART
provisions
will
be
a
part
of
the
State's
long­
term
strategy
for
achieving
reasonable
progress
toward
natural
conditions.
Nothing
in
the
CAA
suggests
that
a
State
should
only
require
a
BART
source
to
reduce
emissions
if
such
controls
by
themselves
would
produce
a
perceptible
improvement
in
visibility.

Commenter
0198
said
use
of
current
visibility
conditions
is
thus
more
appropriate
when
performing
the
significance
analysis
as
you
are
comparing
against
real
world
quantifiable
conditions.
106
Response:

See
the
discussion
above
regarding
the
use
of
natural
conditions
as
the
baseline
for
comparison.

Commenter
0254
said
EPA's
logic
regarding
natural
conditions
is
at
odds
with
the
BART
provisions.
While
the
goal
of
the
regional
haze
program
is
to
restore
natural
visibility
conditions,
a
feat
that
would
require
the
elimination
of
every
manmade
source
of
emissions,
BART
itself
is
just
one
step,
intended
to
make
progress
toward
visibility
improvement
by
focusing
on
specific
source
categories.

Response:

See
the
discussion
above
regarding
the
use
of
natural
conditions
as
the
baseline
for
comparison.

Response:

See
the
discussion
above
regarding
the
use
of
natural
conditions
as
the
baseline
for
comparison.

3.4.2.3
Incorrect
Assumptions
or
Factors
Not
Accounted
for
in
EPA's
Definition
of
Natural
Conditions
Comment:
Commenter
0217
said
the
default
values
do
not
vary
with
season
and
do
not
include
local
effects
such
as
sea
salt
along
the
coast
and
natural
organics
in
forests.
Nor
do
they
recognize
that
organic
particles
under
natural
conditions
will
have
a
different
composition
from
those
created
from
anthropogenic
pollution.
The
default
values
also
do
not
reflect
the
varying
contribution
of
natural
fire.
Also
missing
is
consideration
of
meteorological
factors
such
as
fog
and
precipitation.
There
is
official
sanction
(
via
a
Department
of
Interior
letter
of
January
2003)
for
considering
meteorological
impairment
as
a
component
of
natural
conditions
in
permit
applications
under
the
FLAG
guidance;
the
same
approach
should
be
allowed
for
BART
analyses.

Response:

EPA
has
provided
the
Guidance
for
Estimating
Natural
Visibility
Conditions
Under
the
Regional
Haze
Rule
(
U.
S.
EPA.
September
2003)
in
which
natural
background
conditions,
expressed
in
deciviews,
are
defined
for
each
Class
I
area.
Default
conditions
may
be
used
or
there
are
measures
to
develop
refined
estimates
of
natural
conditions
which
may
account
for
natural
phenomena.
The
modeling
guidance
for
purposes
of
BART
is
not
related
to
the
implementation
of
FLAG.
107
Commenter
0236
said
that
before
a
0.5
deciview
change
is
established
as
the
exemption
threshold,
the
technical
basis
for
the
natural
background
visibility
should
be
proposed
by
each
state
or
RPO
for
peer
review
and
technical
comments
from
the
public.
The
commenter
discussed
the
effects
of
hygroscopic
effects
of
some
organic
compounds
and
stressed
the
importance
of
addressing
these.
Until
these
deficiencies
are
corrected
and
a
more
refined
estimate
of
natural
visibility
is
proposed
by
either
a
state
or
RPO,
the
exemption
threshold
for
BART
should
be
1
deciview
to
account
for
the
uncertainties
inherent
in
the
current
natural
visibility
values.

Response:

See
the
response
above
regarding
the
estimate
of
natural
background
values.
Also,
States
have
discretion
in
determining
an
appropriate
cause
or
contribute
threshold.

Commenter
0225
noted
the
effects
that
volcanic
emissions
can
have
on
visibility
in
Hawaii.
Visibility
Conditions
Under
the
Regional
Haze
Program"
do
not
account
for
the
impact
of
the
Kilauea
Volcano,
and
therefore
are
not
representative
of
natural
visibility
conditions
in
Hawaii.
Individualized
natural
visibility
condition
values
should
be
developed
for
both
Haleakala
and
Volcanoes
National
Parks.

Response:

See
the
response
above
regarding
the
estimate
of
natural
background
values.

Commenter
0217
added
that
statistical
method
used
to
derive
the
natural
conditions
estimates
for
the
20%
clearest
and
20%
haziest
days
is
incorrect
and
causes
the
estimate
of
the
haziness
on
the
20%
haziest
natural
days
to
be
too
low.

Response:

See
the
response
above
regarding
the
estimate
of
natural
background
values.

Comment:
Two
commenters
raised
questions
regarding
the
definitions
(
0259,
0271).
Commenter
0259
said
by
natural
conditions,
does
EPA
mean
modeling
in
the
absence
of
all
other
anthropogenic
sources,
and
does
it
matter
what
the
assumed
natural
visibility
level
is
(
e.
g.,
seven
deciviews)?
Commenter
0271
said
it
is
unclear
from
the
proposed
guidance
what
natural
visibility
conditions
EPA
intends
the
calculated
changes
in
deciviews
from
a
sources
potential
control
are
to
be
compared
against.
Does
EPA
intend
for
a
state
to
compare
the
natural
visibility
conditions
on
the
20%
best,
worst
or
average
days?
The
commenter
assumed
EPA
intends
for
this
comparison
to
be
done
for
the
natural
background
conditions
on
the
20%
best
days,
but
that
is
not
stated
in
this
section.

Response:
108
The
definition
of
natural
conditions
is
that
background
that
exists
in
the
absence
of
human
caused
impairment.
The
Guidance
for
Estimating
Natural
Visibility
Conditions
Under
the
Regional
Haze
Rule
(
U.
S.
EPA.
September
2003)
has
estimates
of
natural
background
conditions
and
States
have
discretion
on
which
values
to
use.

3.4.3
Threshold
Level
3.4.3.1
0.5
Deciview
Threshold
is
Appropriate
Comment:
Five
commenters
said
that
the
0.5
deciview
threshold
is
appropriate
(
0179,
0221,
0275,
0306,
0307).
Commenter
0221
added
that
this
level
is
supported
in
the
literature
and
is
appropriate,
especially
given
the
extremely
low
triggering
threshold
for
visibility
impairment
established
by
Congress.
Given
that
other
factors,
not
captured
in
the
deciview
metric,
may
in
fact
effect
human
judgments
of
changes
in
visual
air
quality,
the
deciview
can
be
considered
to
be
a
useful
but
not
entirely
complete
measure
of
the
fundamental
visual
air
quality.
It
is
therefore
essential,
as
a
matter
of
science,
policy
and
law,
that
EPA
use
a
protective
threshold
 
no
more
lax
than
0.5
deciview
 
in
relying
on
the
deciview
metric
as
an
indicator
of
visual
air
quality.
Commenter
0275
said
that
since
no
visibility
benefits
can
be
discerned
below
that
level,
a
lower
screening
level
cannot
be
technically
or
legally
justified.
If
this
screening
level
were
adopted,
sources
with
a
lesser
impact
would
be
exempted
from
BART
control.
However,
under
the
reproposal
they
would
still
be
considered
along
with
all
other
sources
in
drawing
up
state
visibility
plans.

Response:

One
of
the
first
steps
in
determining
whether
sources
cause
or
contribute
to
visibility
impairment
for
purposes
of
BART
is
to
establish
a
threshold
against
which
to
measure
one
or
more
source's
impacts.
In
establishing
such
a
threshold,
you
should
keep
in
mind
that
in
virtually
all
situations,
a
1.0
deciview
change
in
visibility
is
perceptible
to
the
human
eye.
For
purposes
of
determining
those
sources
subject
to
BART,
a
single
source
that
is
responsible
for
a
1.0
deciview
change
or
more
should
likely
be
considered
to
"
cause"
visibility
impairment.
The
test
for
determining
whether
a
source
is
subject
to
BART,
however,
is
whether
the
source
causes
or
contributes
to
visibility
impairment.
If
"
causing"
visibility
impairment
means
causing
a
humanly
perceptible
change
in
visibility
in
virtually
all
situations,
then
"
contributing"
to
visibility
impairment
must
mean
having
some
lesser
impact
on
the
conditions
affecting
visibility
that
need
not
rise
to
the
level
of
human
perception.
The
question
is
how
much
change
which
is
not
always
humanly
perceptible
should
a
source
be
allowed
to
bring
about
before
it
is
considered
to
be
"
contributing"
to
impairment
for
purposes
of
BART.
In
considering
the
meaning
of
"
contribute
to,"
we
note
that
as
a
scientific
matter
the
National
Academy
of
Sciences
observed
that
the
effects
of
even
small
emissions
sources
on
visibility
are
cumulative
and
contribute
to
visibility
impairment.
109
Because
of
varying
circumstances
affecting
different
Class
I
areas,
the
appropriate
threshold
for
determining
whether
a
source
"
causes
or
contributes
to
any
visibility
impairment"
for
the
purposes
of
BART
may
reasonably
differ
across
States.
As
a
general
matter,
any
threshold
that
you
use
should
not
be
higher
than
0.5
deciviews,
and
could
be
as
low
as
0.1
deciviews
or
less.
One
possible
approach
in
establishing
a
threshold
is
to
consider
the
number
of
emissions
sources
affecting
the
Class
I
areas
at
issue
and
the
magnitude
of
the
individual
sources'
impacts.
In
general,
a
larger
number
of
sources
causing
impacts
in
a
class
I
area
should
warrant
a
lower
contribution
threshold.
The
low
threshold
suggested
by
the
language
of
the
CAA
supports
the
conclusion
that
a
source
with
any
measurable
predicted
impacts
 
for
example,
as
low
as
0.1
deciviews
or
less
 
could
be
considered
to
contribute
to
visibility
impairment.
It
would
be
particularly
reasonable
for
a
State
to
adopt
such
an
approach
in
establishing
a
contribution
threshold
where
a
large
number
of
sources
are
having
small
impacts
which,
when
combined,
would
produce
significant
visibility
impairment.
In
selecting
a
threshold,
you
may
wish
also
to
take
into
account
the
fact
that
different
individual
sources
will
have
varying
impacts
on
visibility
in
class
I
areas.
Depending
on
the
circumstances
and
number
of
sources
affecting
a
class
I
area
and
their
modeled
impacts,
a
threshold
could
be
set
that
captures
those
sources
responsible
for
the
vast
majority
of
the
total
visibility
impacts,
while
still
excluding
other
sources
with
relatively
small
impacts.

One
commenter
(
0231)
requested
that
EPA
review
the
data
available
in
the
scientific
community
and
reaffirm
that
0.5
deciviews
is
the
appropriate
threshold
value
and
is
technically
supportable.

Response:

See
the
discussion
above
regarding
the
establishment
of
which
threshold
is
appropriate.

3.4.3.2
0.5
Deciview
Threshold
is
Too
High
Comment:
Seven
commenters
said
that
the
0.5
deciview
threshold
is
too
high
(
0221,
0227,
0232,
0267,
0271,
0280,
0286).
Commenters
(
0221,
0227)
said
EPA's
proposed
approach
to
assessing
visibility
impairment
(
and
improvement)
is
flawed,
and
must
be
altered
to
comply
with
both
American
Corn
Growers
and
the
provisions
of
Section
169A
of
the
Act.
EPA's
proposed
approach
effectively
reads
the
"
contribution"
element
of
the
BART
applicability
test
out
of
Section
169A(
b)(
2)(
A)
of
the
Act.
A
single
BART­
eligible
source
can
contribute
to
visibility
impairment
at
much
lower
levels
than
0.5
deciview;
if
such
a
source
contributes
any
measurable
amount
to
visibility
impairment
(
impact
that
in
combination
from
contribution
from
other
such
sources
exceeds
0.5
deciview),
that
source
must
be
subject
to
BART
under
the
Act.
In
view
of
the
above,
EPA
must
set
the
minimum
threshold
for
individual
source
contribution
to
visibility
impairment
at
the
lowest
level
detectable
by
the
modeling
or
other
appropriate
analysis.
This
minimum
or
threshold
individual
contribution
level
must
in
any
event
be
set
no
greater
than
a
0.1
deciview
change
relative
to
natural
conditions,
a
clearly
measurable
level.
EPA
must
require
states
110
to
subject
all
BART­
eligible
sources
that
contribute
this
minimum
level
to
BART
requirements
unless
the
visibility
impairment
resulting
from
the
aggregate
of
the
contributions
of
all
such
BART­
eligible
sources
for
a
Class
I
area
are
less
than
0.5
deciview.

Response:

See
the
discussion
above
regarding
the
establishment
of
which
threshold
is
appropriate.

Commenter
0271
said
if
it
is
EPA's
goal
to
improve
visibility
to
natural
background
conditions,
a
logical
first
step
is
to
reduce
the
impact
from
the
oldest
sources
to
that
which
is
not
perceptible
to
the
human
eye.
Thus
a
significantly
lower
threshold
should
be
developed
and
implemented
for
the
Regional
Haze
rule.
The
"
collective
contribution"
approach
correctly
created
this
lower
threshold.
EPA
should
state
that
a
0.5
deciview
change
from
"
one
or
a
small
number"
of
sources
meets
the
definition
of
"
reasonably
attributable
visibility
impairment"
under
this
rule.
EPA
should
then
set
the
deciview
level
under
the
Regional
Haze
rule
to
a
level
adequate
to
protect
against
a
source's
contributory
effect.

Response:

See
the
discussion
above
regarding
the
establishment
of
which
threshold
is
appropriate.

Commenter
0280
applied
the
"
ranking"
and
Q/
D
method
on
BART­
eligible
sources
in
Wisconsin
as
a
first
step
in
determining
sources
subject
to
BART.
Both
ranking
and
Q/
D
method
showed
that
applying
these
methods
as
described
by
EPA
would
lead
to
exemption
of
most
non­
EGU
sources
in
Wisconsin.
These
preliminary
results
indicate
that
the
proposed
thresholds
are
probably
too
high
and
don't
reflect
the
real
aggregate
regional
air
quality
impact.
On
the
other
hand,
some
states
will
not
allow
any
exemptions
at
all,
which
corresponds
to
adopting
a
threshold
of
zero
deciview.
To
avoid
major
differences
between
States
we
think
a
threshold
between
zero
to
0.5
deciview
is
most
appropriate.
The
commenter
suggested
that
EPA
establish
a
de
minimis
threshold
between
0.1­
0.2
deciview.

Response:

See
the
discussion
above
regarding
the
establishment
of
which
threshold
is
appropriate.

Commenter
0267
said
EPA
should
provide
more
explanation
of
how
cumulative
effects
would
be
addressed
under
the
proposed
approach.

Response:

See
the
discussion
above
regarding
the
establishment
of
which
threshold
is
appropriate.

3.4.3.3
0.5
Deciview
Threshold
is
Too
Low
111
Comment:
Fifteen
commenters
said
that
the
0.5
threshold
level
is
not
supported
by
current
research
(
0188,
0189,
0197,
0198,
0207,
0217,
0224,
0230,
0250,
0252,
0253,
0254,
0256,
0270,
0321).
Commenters
said
that
the
19991
NAPAP
report
that
is
the
basis
for
EPA's
proposed
0.5
deciview
threshold
is
outdated
and
no
longer
reflects
the
best
science
regarding
what
is
a
"
just
noticeable"
difference.
Commenters
cited
more
recent
literature
justifying
a
threshold
of
1­
2
deciviews.
One
commenter
(
0207)
said
EPA
should
allow
States
and
regional
planning
organizations
the
flexibility
to
determine
appropriate
visibility­
impact
thresholds
in
light
of
current
knowledge
about
a
range
of
perceptibility
thresholds.

Response:

See
the
discussion
above
regarding
the
establishment
of
which
threshold
is
appropriate.

Comment:
Nineteen
commenters
said
that
the
0.5
standard
is
too
low
to
be
perceived
by
the
human
eye
and
the
threshold
should
be
set
at
1.0
deciview
or
higher
(
0185,
0209,
0214,
0217,
0226,
0228,
0229,
0242,
0244,
0246,
0247,
0252,
0256,
0268,
0276,
0277,
0319,
0321,
0324).
Commenters
cited
recent
literature
supporting
this
assertion.
Commenters
noted
that
EPA
acknowledged
in
the
preamble
that
at
lest
a
one
deciview
change
may
be
required
in
order
for
a
change
to
be
perceptible
and
that
human
perceptibility
should
be
the
basis
for
the
factor.

Response:

See
the
discussion
above
regarding
the
establishment
of
which
threshold
is
appropriate.

Comment:
Other
commenters
were
concerned
that
the
threshold
is
set
too
low
and
should
be
established
at
a
more
realistic
level
(
e.
g.,
1.0
deciviews
or
more)
(
0208,
0238,
0241).
Commenter
0208
was
concerned
that
EPA
has
established
a
visibility
significance
level
that
is
insignificant.
It
should
not
set
such
a
low
threshold
so
that
virtually
all
sources
are
burdened
with
BART
assessments
regardless
of
whether
they
cause
any
"
humanly
perceptible"
visibility
impairment.
Commenter
0319
said
EPA
should
begin
at
the
top
of
the
list
with
the
sources
with
the
greatest
emissions
closest
to
a
Class
I
area
and
then
provide
the
resources
to
the
states
to
prove
the
contribution
and
determine
and
implement
improvements
at
these
sources
first
and
incrementally
before
requiring
all
plants
to
prove
their
innocence.

Response:

See
the
discussion
above
regarding
the
establishment
of
which
threshold
is
appropriate.

Comment:
112
Commenter
0321
said
the
perception
discussion
does
not
even
consider
the
uncertainty
of
CALPUFF
modeling
over
hundreds
of
kilometers.
It
will
be
difficult
for
States
to
justify
the
cost/
benefit
of
emissions
reductions
to
achieve
visibility
improvement
using
such
a
small
criterion.

Response:

See
the
discussion
above
regarding
the
establishment
of
which
threshold
is
appropriate.
States
have
the
discretion
in
setting
an
appropriate
threshold
and
may
consider
the
CALPUFF
results
in
establishing
the
threshold.

Comment:
Two
commenters
addressed
the
role
that
natural
conditions
play
in
the
discussion
of
a
visibility
threshold.
Commenter
0236
said
the
metric
for
visibility
degradation
to
exempt
a
source
from
being
BART
eligible
should
be
a
0.5
deciview
change
only
if
consideration
for
the
inherent
problems
and
limitation
of
using
estimated
natural
visibility
conditions
as
the
basis
for
comparison
are
accounted
for.
To
assess
the
change
of
visibility
from
natural
background
conditions,
the
visibility
impairment
during
periods
of
fog,
rain,
snow
and
other
weather
related
events
and
periods
of
visibility
impairment
during
nighttime
periods
must
be
eliminated
from
the
24­
hr
change
in
visibility
calculations.
Otherwise,
the
exemption
metric
should
be
1
deciview.

Response:

EPA
has
provided
the
Guidance
for
Estimating
Natural
Visibility
Conditions
Under
the
Regional
Haze
Rule
(
U.
S.
EPA.
September
2003)
in
which
natural
background
conditions,
expressed
in
deciviews,
are
defined
for
each
Class
I
area.
Default
conditions
may
be
used
or
there
are
measures
to
develop
refined
estimates
of
natural
conditions
which
may
account
for
natural
phenomena.

Commenter
0304
supports
the
0.5
deciview
metric,
but
notes
it
may
be
clearer
to
propose
this
metric
as
simply
a
5
percent
light
extinction
change
from
natural
background
conditions.
The
commenter
recommended
that
EPA
work
to
better
define
a
process
by
which
natural
conditions
are
determined
and
by
which
the
0.5
deciview
metric
is
calculated.

Response:

The
deciview
metric
is
a
function
of
light
extinction
and
its
derivation
is
such
that
a
unit
change
in
visibility
expressed
in
deciviews,
perceived
or
measured,
is
the
same
regardless
of
geography.
This
is
not
the
case
for
light
extinction.
Also,
see
the
discussion
above
regarding
the
determination
of
natural
conditions.

3.4.3.4
0.5
Deciview
Threshold
as
a
Standard
Comment:
113
Three
commenters
opposed
converting
a
deciview
"
goal"
into
a
"
standard
(
0195,
0209,
0238).
Commenter
0209
said
EPA
is
violating
the
intent
of
the
Act
by
forcing
reductions
where
there
are
no
"
humanly
perceptible,"
visibility­
impairing
emissions.
The
new
BART
rule
would
require
that
sources
that,
according
to
a
conservative
computer
model,
cause
greater
than
5%
change
in
visibility,
could
not
be
excluded
from
BART.
In
effect,
this
proposal
would
change
the
regulatory
role
of
the
"
deciview"
metric
by
converting
it
into
a
regulatory
0.5
deciview
standard
(
versus
a
"
goal")
for
defining
how
states
must
exercise
their
authority
and
discretion
in
determining
whether
an
individual
source
"
causes
or
contributes"
to
visibility
impairment
within
an
individual
Class
I
area.

Response:

We
believe
that
it
is
appropriate
for
States
to
use
the
deciview
in
establishing
a
threshold
for
BART
applicability.
CAA
169A(
b)(
2)(
A)
provides
for
a
BART
determination
for
any
BARTeligible
source
which
"
emits
any
air
pollutant
which
may
reasonably
be
anticipated
to
cause
or
contribute
to
any
impairment
of
visibility
in
any
[
Class
I]
area."
States
will
need
some
means
of
determining
whether
a
source
"
causes
or
contributes
to"
visibility
impairment.
The
guidelines
advise
States
to
use
a
deciview
metric
in
defining
"
cause
or
contribute".
The
fact
that
the
deciview
is
also
used
to
track
progress
toward
the
goal
of
natural
visibility
does
not
in
any
way
indicate
that
we
are
"
converting"
a
"
goal"
into
a
requirement.
Use
of
the
same
metric
in
the
"
cause
or
contribute"
context
as
used
for
establishing
reasonable
progress
goals,
tracking
changes
in
visibility
conditions,
and
defining
baseline,
current,
and
natural
conditions
simply
provides
for
a
consistent
approach
to
quantifying
visibility
impairment.
Specifically,
we
wish
to
clarify
here
that
we
have
concluded
that
for
purposes
of
determining
those
sources
subject
to
BART,
a
single
source
that
is
responsible
for
a
1.0
deciview
change
or
more
should
be
considered
to
"
cause"
visibility
impairment.
States
have
discretion
in
establishing
which
"
cause
or
contribute"
threshold
is
appropriate,
but
in
no
case
should
that
threshold
be
greater
than
0.5
dv.

Comment:
Two
commenters
said
EPA
should
establish
a
visibility
improvement
threshold
that
needs
to
be
met
for
a
BART­
eligible,
"
subject
to
BART"
source
to
actually
be
required
to
implement
BART
28­
348,
28­
387).
There
must
be
evidence
that
an
actual
visibility
improvement
is
likely
to
occur.

Response:

We
agree
and
the
rule
establishes
thresholds
for
BART­
eligible
sources.

3.4.3.4
Cumulative
Impact
of
Sources
Comment:
Three
commenters
said
that
due
to
cumulative
impacts,
EPA
should
require
that
for
any
combination
of
BART
sources
that
results
in
an
impact
on
visibility
in
any
Class
1
area
of
more
114
than
0.5
deciviews
(
by
CALPUFF
modeling
for
any
24­
hour
period)
each
individual
source
must
be
BART
eligible
(
0189,
0221,
0234).
Commenter
0234
said
in
a
case
where
multiple
sources
each
have
an
impact
on
visibility
impairment
of
less
than
0.5
deciview
but
together
have
an
impact
of
more
than
0.5
deciview,
each
of
these
sources
makes
a
contribution
to
the
impairment
of
visibility
that
results
from
their
emissions.
BART
guidelines
that
result
in
exemptions
for
these
sources
would
subvert
the
goals
of
the
regional
haze
program.

Response:

See
the
discussion
above
regarding
the
establishment
of
which
threshold
is
appropriate.
States
have
the
discretion
in
setting
an
appropriate
"
cause
or
contribute"
threshold.

Commenter
0221
provided
extensive
discussion
of
why
this
approach
is
consistent
with
the
Act
and
the
Corn
Growers
decision.
An
alternative
method
to
give
meaning
to
the
statutory
text
and
intent
of
Congress
would
be
to
apply
a
two­
step
analysis
that
first
assesses
the
cumulative
impact
of
BART­
eligible
sources
on
a
particular
Class
I
area
and
then
exempts
individual
sources
based
upon
their
percentage
contribution
to
that
cumulative
impact.
Under
this
approach,
a
state
would
run
a
CALPUFF
analysis
to
determine
if
the
cumulative
visibility
impact
from
all
BART­
eligible
sources
exceeded
the
0.5
deciview
visibility
impairment
threshold
as
measured
from
natural
visibility
conditions.
If
so,
the
state
would
run
individual
CALPUFF
analyses
to
determine
which
sources
contributed
more
than
a
certain
percentage
to
the
cumulative
visibility
impairment.
The
FLMs
use
a
similar
method
in
evaluating
the
visibility
impacts
associated
with
proposed
new
sources
under
the
PSD
program.
EPA
must,
however,
require
states
to
apply
BART
requirements
to
any
BART­
eligible
source
unless
the
source
or
the
regulating
state
performs
the
analysis
necessary
to
demonstrate
that
pollution
from
the
source
neither
causes
visibility
impairment
nor
makes
a
measurable
or
small
percentage
(
4­
10%)
contribution
to
noticeable
impairment.

Response:

States
have
the
option
to
demonstrate
that
all
BART­
eligible
sources
in
the
State
collectively
cause
or
contribute
to
visibility
impairment
at
a
Class
I
area.
Therefore,
all
sources
in
the
State
may
be
deemed
subject
to
BART.
Further,
a
State
has
the
option
of
conducting
an
individual
source
attribution
approach
for
the
cause
or
contribute
determination.
States
also
have
discretion
in
establishing
the
appropriate
threshold
in
determining
if
a
source
should
be
subject
to
BART.

Two
commenters
noted
that
the
BART
guidelines
allow
states
to
demonstrate,
using
a
cumulative
approach,
that
none
of
its
BART­
eligible
sources
contribute
to
visibility
impairment,
but
with
no
specific
guidance
as
to
the
deciview
threshold
from
the
cumulative
impact
of
multiple
BART­
eligible
sources
(
0217,
0246).
Commenter
0217
said
under
some
circumstances
where
two
or
more
individual
BART­
eligible
sources,
for
example,
may
show
a
visibility
impact
of
0.60
deciview
on
a
Class
I
area,
but
with
individual
sources
showing
each
at
less
than
0.50
deciview,
it
115
would
appear
that
the
cumulative
modeling
would
become
a
redundant
exercise
since
under
the
Corngrowers
decision,
each
BART­
eligible
source
has
the
prerogative
to
model
its
individual
visibility
impact.
Perhaps,
a
cumulative
BART­
eligible
demonstration
should
be
allowed
to
use
a
higher
deciview
threshold
than
for
individual
sources.

Response:

See
the
discussion
above
regarding
the
establishment
of
which
threshold
is
appropriate.
States
have
the
discretion
in
setting
an
appropriate
"
cause
or
contribute"
threshold.

Commenter
0267
was
uncomfortable
with
the
individual
source
threshold
of
0.5
deciview
or
5%
change
to
exempt
sources
from
BART.
EPA
needs
to
specify
how
the
cumulative
effects
of
many
sources,
each
with
impacts
just
below
the
threshold,
will
be
addressed.
Commenter
0323
said
EPA
may
want
to
explore
an
approach
combining
BART
requirements
with
overall
Long
Term
Strategy
requirements
by
stressing
that
the
cumulative
effects
from
a
multitude
of
sources
(
BART
and
non­
BART)
will
need
to
be
addressed
under
regional
haze
plans
submitted
by
the
states
and
Tribes.
Commenter
0307
said
situations
where
multiple
BART­
eligible
sources
impact
visibility
at
the
same
Class
I
receptor
are
expected
to
be
rare
and
states
should
be
allowed
the
flexibility
to
deal
with
these
circumstances
as
they
arise.

Response:

See
the
discussion
above
regarding
the
establishment
of
which
threshold
is
appropriate.
States
have
the
discretion
in
setting
an
appropriate
"
cause
or
contribute"
threshold.
Also,
the
long
term
strategy
assessments,
which
are
part
of
the
Reasonable
Progress
determinations
required
in
the
SIPs,
provide
an
overall
determination
of
visibility
in
a
particular
Class
I
area(
s)
in
a
State.
These
strategies
effectively
deal
with
all
sources
contribution
to
visibility
impairment
in
the
Class
I
area.

3.4.3.5
Other
Comments
Regarding
the
Threshold
Level
One
commenter
(
0221)
urged
EPA
to
calculate
that
0.5
deciview
impairment
threshold
(
and
the
lower
individual
contribution
threshold)
on
the
basis
of
a
1­
hour
maximum
reading
or
an
average
no
longer
than
24
hours.
Using
a
longer
average
will
simply
serve
to
weaken
the
robustness
of
the
threshold.
Furthermore,
national
park
visitors
experience
instantaneous,
not
average,
visibility
conditions;
the
form
of
the
threshold
should
reflect
this
actual
experience
as
closely
as
practicable.

Response:
116
We
believe
that
a
24­
hour
average
calculated
from
modeled
hourly
visibility
impacts,
effectively
reduces
the
impact
of
any
one
particular
hour
that
could
be
higher
due
to
a
number
of
meteorological
effects.

3.4.4
General
Comments
Regarding
Visibility
Thresholds
3.4.4.1
Setting
One
Threshold
Level
is
Inappropriate
Comment:
Seven
commenters
said
that
issues
of
human
perceptibility
are
very
Class
I
are
site­
specific
(
0197,
0209,
0214,
0238,
0268,
0277,
0321).
Commenter
0209
said
Congress
has
recognized
that
states
need
the
ability
to
use
their
discretion
in
determining
which
sources
contribute
to
visibility
impairment.
EPA's
suggested
BART
standard
would
violate
this
tenet
by
dictating
a
single
threshold
that
humans
could
not
perceive.
Commenter
0277
said
threshold
levels
should
be
established
separately
for
the
eastern
and
western
regions
of
the
United
States.
The
"
natural
visibility"
conditions
are
established
separately
for
eastern
and
western
regions
in
the
guidance.
The
threshold
levels
should
be
established
consistent
with
establishment
of
the
natural
visibility.
The
rule
also
should
allow
for
development
of
refined
threshold
levels
for
individual
Class
I
areas,
consistent
with
establishing
refined
natural
visibility
conditions
in
the
guidance.
Citing
Congressional
intent,
commenter
0321
added
that
EPA
should
allow
states
to
use
site
specific
path
lengths
within
each
respective
Class
I
area.

Response:

See
the
discussion
above
regarding
the
establishment
of
which
threshold
is
appropriate.
States
have
the
discretion
in
setting
an
appropriate
"
cause
or
contribute"
threshold.
Also,
we
disagree
that
separate
threshold
levels
should
be
established
based
on
geography
since
a
unit
change
in
visibility
expressed
in
deciviews,
perceived
or
measured,
is
the
same
regardless
of
geography.
The
change
in
deciviews,
which
is
compared
to
the
threshold,
is
based
on
the
percentage
change
that
occurs
in
existing
light
extinction
at
a
Class
I
area,
measured
relative
to
the
natural
conditions
estimate
at
the
Class
I
area.
Natural
conditions
may
be
different
at
each
Class
I
area,
and
so
a
given
change
in
light
extinction
may
result
in
a
higher
or
lower
percentage.
As
such,
a
source
may
"
pass"
the
threshold
test
in
one
Class
I
area,
but
"
fail"
in
another.
We
disagree
that
Congressional
intent
was
to
allow
site
specific
path
lengths
to
be
used.
On
the
contrary,
haze
should
be
assessed
uniformly
across
Class
I
areas
which
is
not
related
to
path
lengths
for
specific
vistas.

3.4.4.2
States
Should
Have
Authority
to
Set
a
Threshold
Level
Comment:
Nine
commenters
said
states
should
have
the
discretion
to
designate
the
threshold
for
deciview
change
that
is
perceptible
rather
than
dictating
a
stringent
one
size
fits
all
threshold
(
0197,
0214,
0217,
0226,
0238,
0246,
0268,
0270,
0321).
Commenter
0270
said,
consistent
with
117
the
overriding
purpose
of
the
proposed
rulemaking
(
as
directed
by
the
Court
in
the
Corn
Growers
litigation
affirming
states'
broad
authority
over
BART
determinations),
determining
the
appropriate
threshold
level
should
be
left
to
the
states'
discretion.
It
is
appropriate
for
the
Guidelines
to
provide
the
basis
for
a
proposed
threshold
including
the
most
recent
advances
in
visibility
science.
The
states
should
then
have
the
unfettered
ability
to
determine
an
appropriate
threshold.

Response:

See
the
discussion
above
regarding
the
establishment
of
which
threshold
is
appropriate.
States
have
the
discretion
in
setting
an
appropriate
"
cause
or
contribute"
threshold.

Commenter
0321
note
that
the
proposal's
BART
source
attribution
proposal
is
also
a
substantial
departure
from
EPA's
own
existing
visibility
regulations.
For
example,
the
phrase
"
visibility
impairment"
is
defined
under
EPA's
Phase
I
regulations
as
"
any
humanly
perceptible
change
in
visibility
(
visual
range,
contrast,
coloration)
from
that
which
would
have
existed
under
natural
conditions.
Visibility
impairment
is
"
reasonably
attributable"
to
a
source
if
the
impairment
is
"
attributable
by
visual
observation
or
any
other
technique
the
State
deems
appropriate."
Nowhere
in
EPA's
proposal
does
EPA
explain
the
basis
for
its
dramatic
departure
with
respect
to
state
decision
making
authority.

Response:

See
the
discussion
above
regarding
the
establishment
of
which
threshold
is
appropriate.
States
have
the
discretion
in
setting
an
appropriate
"
cause
or
contribute"
threshold.

3.4.4.3
Setting
a
"
Very
Low"
Triggering
Threshold
Comment:
Two
commenters
supported
setting
a
very
low
triggering
threshold
for
determining
whether
a
source
is
subject
to
BART
(
28­
258,
28­
302).
Commenter
28­
258
added
that
the
source
could
be
hundreds
or
thousands
of
kilometers
away
from
a
Class
I
areas
and
still
emit
pollutants
that
could
be
transported
to
the
Class
I
area.

Response:

See
the
discussion
above
regarding
the
establishment
of
which
threshold
is
appropriate.
States
have
the
discretion
in
setting
an
appropriate
"
cause
or
contribute"
threshold.

Comment:
Seventeen
commenters
opposed
setting
a
very
low
triggering
threshold
(
28­
304,
28­
337,
28­
344,
28­
345,
28­
349,
28­
346,
28­
353,
28­
354,
28­
355,
28­
356,
28­
364,
28­
393,
28­
397,
28­
404,
28­
406,
28­
748,
28­
7003).
Commenters
said
this
geographic
threshold
would
result
in
most,
118
if
not
all,
BART­
eligible
sources
being
located
in
geographic
regions
that
contribute
to
regional
haze
(
28­
304,
28­
337,
28­
353,
28­
354,
28­
356).
Such
a
broad
presumption
is
contrary
to
the
Act,
which
requires
the
states
to
conduct
a
source­
by­
source
analysis,
and
limits
state
discretion
(
28­
345,
28­
349,
28­
253,
28­
254,
28­
353,
28­
354,
28­
355,
28­
364,
28­
393,
28­
748,
28­
7003).
It
would
also
result
in
limited
visibility
improvements
using
costly
controls
and
would
be
unnecessarily
broad
(
e.
g.,
include
home
furnaces)
(
28­
304,
28­
345,
28­
353,
28­
397,
28­
404).

Response:

See
the
discussion
above
regarding
the
establishment
of
which
threshold
is
appropriate.
States
have
the
discretion
in
setting
an
appropriate
"
cause
or
contribute"
threshold
as
part
of
their
source­
by­
source
analysis.

The
guidelines
fail
to
account
for
differing
source
impacts
on
visibility
(
28­
345,
28­
349).
The
"
zero­
out"
approach
is
not
an
acceptable
alternative
(
28­
302,
28­
356,
28­
7003).
Commenter
28­
302
added
that
allowing
the
"
zero­
out"
approach
for
exemption
demonstrations
on
a
state­
bystate
basis
will
significantly
reduce
the
effectiveness
of
the
BART
program
by
allowing
individual
states
or
geographical
regions
to
claim
"
trivial"
impacts.
The
aim
of
the
regional
haze
rule
is
to
"
address
visibility
impairment
due
to
the
cumulative
air
pollutant
emissions
from
numerous
sources
over
a
wide
geographic
area."

Response:

The
goal
of
the
regional
haze
program
is
a
return
to
natural
conditions.
The
rule
provides
States
several
options,
some
of
which
are
cumulative
determinations,
for
assessing
if
BARTeligible
sources
cause
or
contribute
to
visibility
impairment.
Discretion
and
flexibility
is
afforded
to
the
States
in
how
to
make
a
successful
demonstration
as
well
as
setting
an
appropriate
threshold
for
BART
exemptions.

Comment:
One
commenter
said
guidance
is
needed
from
EPA
to
implement
such
a
threshold
(
28­
266).
EPA
should
what
is
meant
by:
"
can
be
emitted
and
transported"
­
i.
e.,
Is
it
necessary
to
show
that,
say,
at
least
one
back
trajectory
on
the
20%
worst
visibility
days
from
a
Class
I
area
pass
through
the
county
[
or
state]
in
which
a
given
source
is
located?

Response:

Discretion
and
flexibility
is
afforded
to
the
States
in
how
to
make
a
successful
demonstration
as
well
as
setting
an
appropriate
threshold
for
BART
exemptions.

Three
commenters
requested
clarification
for
what
is
meant
by
"
geographic
region"
and
what
methods
can
be
used
to
determine
"
spatial
extent"
(
28­
258,
28­
266,
28­
393).
Commenter
28­
258
added
that
EPA's
BART
guidelines
should
make
clear
that
a
geographic
region
would
119
likely
extend
beyond
a
state's
boundaries
in
which
a
Class
I
area
is
located,
and
may
even
extend
beyond
adjacent
states'
boundaries.

Response:

As
part
of
their
regional
haze
SIPs,
States
may
make
an
acceptable
demonstration
of
which
BART­
eligible
sources
within
its
boundary
do
or
do
not
cause
or
contribute
to
visibility
impairment
at
a
Class
I
area.
Likewise,
the
determination
can
be
made
for
sources
outside
of
the
State
boundary.
All
source
contributions
should
be
accounted
for
as
part
of
the
State's
reasonable
progress
demonstration,
and
thus
geographic
region
and
spatial
extent
are
less
important.

3.4.4.4
Other
Comments
on
Setting
Visibility
Thresholds
Comment:
One
commenter
(
0323)
said
whatever
the
final
threshold
for
a
single­
source
impact
for
BART
sources,
EPA
should
clarify
that
the
purpose
of
this
modeling
assessment
is
to
evaluate
a
source's
anticipated
contribution
to
uniform
regional
haze
over
the
Class
I
area.
EPA
should
state
that
the
assumption
of
a
uniform
haze
contribution
based
on
CALPUFF
modeling
eliminates
the
need
to
assess
issues
related
to
the
size
of
the
Class
I
area,
views
within
a
Class
I
area,
and
weather
impact
interactions.

Response:

We
believe
that
the
guidelines,
which
allow
States
to
set
the
cause
or
contribute
threshold
(
which
in
no
case
should
be
larger
than
0.5
dv),
adequately
address
this
comment.
Also,
the
modeling
assess
conditions
across
the
entire
Class
I
area,
regardless
of
size,
and
has
features
to
account
for
weather
impacts,
such
as
the
use
of
monthly
average
relative
humidity.

One
commenter
(
0245)
discussed
the
impact
sulfate
errors
can
have
on
deciview
change
calculations,
along
with
other
possibly
typical
uncertainties
in
pollutant
concentration
predictions
using
CALPUFF,
suggest
that
CALPUFF
screening
predictions
may
be
unreliable.
Furthermore,
visibility
impacts
predicted
as
part
of
BART
Engineering
Analysis
may
also
be
unreliable.
EPA
needs
to
provide
guidance
for
adjusting
the
deciview
threshold
to
reflect
uncertainty
in
modeled
particulate
concentrations.
Alternatively,
guidance
for
determining
adequate
CALPUFF
modeling
performance
for
visibility
impact
calculations
will
be
needed.

Response:

As
discussed
in
the
preamble
to
the
rule,
we
understand
the
concerns
of
commenters
that
the
chemistry
modules
of
the
CALPUFF
model
are
less
advanced
than
some
of
the
more
recent
atmospheric
chemistry
simulations.
To
date,
no
other
modeling
applications
with
updated
chemistry
have
been
approved
by
EPA
to
estimate
single
source
pollutant
concentrations
from
120
long
range
transport.
In
its
next
review
of
the
Guideline
on
Air
Quality
Models,
EPA
will
evaluate
these
and
other
newer
approaches
and
determine
whether
they
are
sufficiently
documented,
technically
valid,
and
reliable
to
approve
for
general
use.
In
the
meantime,
as
the
Guideline
makes
clear,
States
are
free
to
make
individualized
showings
that
these
or
other
alternative
approaches,
including
grid
models,
are
valid
and
appropriate
for
their
intended
applications,
as
well
as
the
appropriate
threshold
to
use
for
the
cause
or
contribute
determination..

Commenter
0201
said
EPA
should
clarify
and
detail
the
technical
analysis
and
regulatory
support
for
establishing
a
one­
time
impact
of
greater
than
0.5
deciview
for
the
purposes
of
determining
BART
source
culpability
in
light
of
the
form
of
the
regional
haze
goals.
EPA
should
also
identify
and
discuss
the
expected
affect
of
the
0.5
deciview
threshold
in
relation
to
the
inclusion
or
exclusion
of
sources
from
the
application
of
BART.

Response:

See
the
discussion
above
regarding
the
establishment
of
which
threshold
is
appropriate.
We
disagree
that
the
use
of
a
one­
time
modeled
visibility
impact
over
five
years
of
meteorological
data
is
overly
strict,
and
we
believe
that
the
comparison
of
this
impact,
to
the
cause
or
contribute
threshold,
to
determine
if
a
source
is
subject
to
BART
is
appropriate
because
of
the
low
triggering
threshold
established
by
Congress.
Where
a
modeling
demonstration
may
use
more
conservative
measures
than
we
used
in
our
example
analyses
in
developing
the
guidelines
(
e.
g.
hourly
relative
humidity)
and
the
State
believes
that
the
maximum
daily
results
may
not
be
truly
representative
of
a
source's
anticipated
impacts,
a
State
may
consider
the
99th
percentile
of
the
predicted
visibility
impacts.
Also,
a
conclusion
that
a
source
is
subject
to
BART
requires
only
that
the
source
undergo
review
by
the
State.
In
establishing
the
appropriate
level
of
controls,
if
any,
that
an
existing
source
should
install,
the
degree
of
visibility
improvement
resulting
from
the
use
of
controls
is
one
of
the
factors
to
be
taken
into
consideration
by
the
State.

Commenter
0302
said
information
on
a
threshold
level
for
deciview
usage
is
unclear,
and
EPA
must
clarify
how
this
value
translates
to
level
of
control
for
each
BART
source.
The
commenter
does
not
fully
comprehend
the
use
of
a
visibility
threshold
as
a
determination
method
for
control
options.
EPA
must
make
appropriate
mathematical
correlations
between
the
emissions
from
a
specific
source
and
the
0.5
deciview
value.
If
it
is
unable
to
do
so,
the
threshold
level
should
be
removed
from
the
guidance.

Response:

A
deciview
threshold
is
not
a
required
metric
in
a
State's
evaluation
of
the
five
factor
analysis
conducted
for
BART
control
decisions
(
although
a
State
has
the
option
of
using
that
metric).
If
expected
visibility
improvement
is
shown
from
the
various
control
choices,
the
State
can
weigh
the
results
with
the
other
four
BART
determination
factors
when
establishing
BART
for
a
particular
source.
If
expected
visibility
improvements
are
significantly
different
under
one
121
control
scenario
than
under
another,
then
a
State
may
use
that
information,
along
with
information
on
the
other
BART
factors,
to
inform
its
BART
determination;
an
absolute
threshold
is
not
dictated.

3.5
General
Comments
Regarding
Identifying
Sources
Subject
to
BART
Comment:
Three
commenters
provided
general
comments
regarding
identifying
sources
subject
to
BART
(
0167,
0221,
0253).
Commenter
0167
was
deeply
concerned
with
EPA's
proposal
to
allow
states
to
categorically
exempt
large
industrial
sources
and
power
plants
from
the
requirement
to
install
BART
or
to
allow
similar
exclusions
for
sources
causing
less
than
a
0.5
deciview
(
deciview)
visibility
impact.
The
Act
explicitly
requires
the
application
of
BART
to
any
source
discharging
pollutants
that
`
cause
or
contribute'
to
visibility
impairment
in
Class
I
areas.
EPA's
proposed
regimen
for
applying
BART
is
fatally
flawed
in
reading
the
contribution
test
out
of
the
statute.
It
also
fails
to
satisfy
EPA's
overarching
responsibility
under
the
Act's
visibility
protection
program
to
`
assure'
reasonable
progress
toward
the
national
visibility
goal.
Even
though
EPA
is
grappling
with
its
response
to
the
American
Corn
Growers
decision,
EPA
has
missed
the
mark
and
must
put
in
place
a
system
to
apply
BART
to
all
industrial
sources
that
not
only
cause
visibility
impairment
but
also
contribute
to
the
pollution
in
specially
protected
national
parks
and
wilderness
areas.

Response:

While
we
do
not
necessarily
agree
with
commenters
who
said
we
conflated
the
"
cause
or
contribute"
test,
we
recognize
that
some
clarification
of
our
treatment
of
this
test
would
be
helpful.
Specifically,
we
wish
to
clarify
here
that
we
have
concluded
that
for
purposes
of
determining
those
sources
subject
to
BART,
a
single
source
that
is
responsible
for
a
1.0
deciview
change
or
more
should
be
considered
to
"
cause"
visibility
impairment.
Additionally,
we
are
establishing
a
change
of
0.5
deciviews
as
the
maximum
visibility
change
we
believe
should
be
used
to
determine
whether
a
source
"
contributes"
to
impairment
for
the
following
reasons.
In
proposing
a
"
cause
or
contribute"
threshold
of
0.5
deciviews,
we
acknowledged
that
there
is
uncertainty
associated
with
identifying
an
absolute
deciview
change
as
"
just
noticeable,"
or
"
perceptible".
This
is
consistent
with
our
observations
in
the
preamble
to
the
1999
Regional
Haze
Rule,
where
we
noted
"
that
a
1.0
deciview
change
should
not
be
considered
the
threshold
of
perception
in
all
cases
for
all
scenes",
and
that
"
visibility
changes
of
less
than
1.0
deciview
are
likely
to
be
perceptible
in
some
cases,
especially
where
the
scene
being
viewed
is
highly
sensitive
to
small
amounts
of
pollution".
We
also
acknowledged
that
a
change
greater
than
1.0
deciview
may
be
required
for
perceptibility
under
certain
site­
specific
conditions.
However,
in
most
cases,
a
1.0
deciview
change
is
considered
to
be
perceptible.
The
test
for
determining
whether
a
source
is
subject
to
BART
is
whether
it
can
reasonably
be
anticipated
to
cause
or
contribute
to
visibility
impairment.
If
"
causing"
visibility
impairment
means
causing
a
humanly
perceptible
change
in
visibility
in
virtually
all
situations,
then
"
contributing"
to
visibility
impairment
must
mean
having
some
lesser
impact
on
the
conditions
11
All
states
are
working
together
in
regional
planning
organizations,
and
we
expect
that
states
will
have
modeling
information
that
identifies
sources
affecting
visibility
in
individual
class
I
areas,
and
the
magnitude
of
their
impacts.

122
affecting
visibility
that
need
not
rise
to
the
level
of
human
perception.
The
question
is
how
much
change
which
is
not
always
humanly
perceptible
should
a
source
be
allowed
to
bring
about
before
it
is
considered
to
be
"
contributing"
to
impairment
for
purposes
of
BART.
In
considering
the
meaning
of
"
contribute
to,"
we
note
that
as
a
scientific
matter
the
National
Academy
of
Sciences
observed
that
the
effects
of
even
small
emissions
sources
on
visibility
are
cumulative
and
contribute
to
visibility
impairment.
In
a
regulatory
context,
we
believe
that
a
State's
decision
as
to
an
appropriate
threshold
for
contribution
could
depend
upon
the
number
of
sources
affecting
a
class
I
area.
To
illustrate,
in
an
unrealistic
hypothetical
situation
in
which
only
one
emissions
source
was
affecting
visibility
in
a
class
I
area,
that
source
could
have
a
deciview
impact
only
slightly
below
the
perceptibility
threshold
without
contributing
to
noticeable
impairment.
However,
if
there
were
100
sources
each
changing
visibility
by
0.1
deciviews,
the
total
impact
would
be
a
10­
deciview
change
in
visibility.
Clearly
in
this
hypothetical
example
all
sources
would
be
contributing,
in
equal
amounts,
to
substantial
visibility
impairment.
Because
circumstances
will
vary
in
different
locations,
EPA
believes
that
states
should
have
discretion
to
set
an
appropriate
threshold
depending
on
the
facts
of
the
situation.
EPA
believes
that
it
would
be
difficult
for
a
state
to
justify
a
threshold
higher
than
0.5
deciviews
for
two
reasons.
First,
studies
suggest
that
0.5
deciviews
represents
a
level
of
visibility
impairment
which
can
be
perceived
in
some
but
not
all
situations.
A
source
whose
emissions
are
predicted
to
result
in
perceptible
or
nearly
perceptible
results
can
definitely
be
considered
to
contribute
to
impairment.
Second,
0.5
deciviews
represents
one
half
of
the
1
deciview
level
that
we
are
equating
with
a
single
source
"
causing"
visibility
degradation.
Multiple
sources
typically
affect
visibility
in
class
I
areas,
so
a
source
causing
a
0.5
deciview
change
can
be
expected
to
be
contributing
to
noticeable
visibility
impairment.
In
determining
whether
the
maximum
threshold
of
0.5
deciviews
or
a
lower
threshold
is
appropriate
for
purposes
of
BART,
EPA
believes
that
states
should
consider
the
number
of
emissions
sources
affecting
the
class
I
area
and
the
magnitude
of
the
individual
sources'
impacts.
11
In
general,
a
larger
number
of
sources
causing
impacts
in
a
class
I
area
will
warrant
a
lower
contribution
threshold.
Given
the
low
threshold
suggested
by
the
language
of
the
CAA,
it
would
be
reasonable
to
conclude
that
a
source
with
any
predicted
impacts
 
for
example,
as
low
as
0.1
deciviews
or
less
 
could
be
considered
to
contribute
to
visibility
impairment
if
a
large
number
of
sources
were
having
small
impacts
that,
combined,
would
produce
visibility
impairment.
In
selecting
a
threshold,
states
may
take
into
account
the
fact
that
individual
sources
have
varying
amounts
of
impact
on
visibility
in
class
I
areas.
Depending
on
facts
regarding
the
number
of
sources
affecting
a
class
I
area
and
their
modeled
impacts,
a
threshold
could
be
set
that
captures
those
sources
responsible
for
virtually
all
of
the
total
visibility
impacts,
while
still
excluding
other
sources
with
very
small
impacts.
For
example,
assume
that
a
class
I
area
is
affected
by
10
sources
123
from
various
industries,
and
that
all
the
sources
are
BART­
eligible.
Eight
sources
have
impacts
of
0.2
or
0.3
deciviews,
and
two
sources
have
impacts
of
0.1
deciviews
or
less.
The
state
could
reasonably
choose
to
set
a
threshold
that
excludes
the
two
sources
with
impacts
of
0.1
deciviews
or
less.
Note
in
this
example
that
only
the
sources
with
the
smallest
impacts
are
excluded,
and
that
the
combined
visibility
impact
of
all
the
sources
that
affect
the
class
I
area,
but
are
not
subject
to
BART,
is
not
perceptible.
We
would
note
that
under
this
guidance,
states
would
have
discretion
in
setting
the
threshold
for
"
contributes
to"
based
on
modeled
impacts
of
sources.
Consistent
with
the
court's
concern
in
the
American
Corn
Growers
decision,
we
are
not
requiring
states
to
find
sources
subject
to
BART
regardless
of
their
of
their
impact
on
Class
I
areas.
We
are
suggesting
that
in
establishing
a
threshold
for
assessing
contribution
for
BART,
it
may
be
logical
to
draw
a
line
between
"
contribution"
and
"
non­
contribution"
based
on
the
number
and
magnitude
of
the
various
sources
affecting
the
Class
I
areas
at
issue.
Such
an
approach
gives
States
the
ability
to
assess
the
empirical
evidence
showing
contribution
and
to
design
an
appropriate
regulatory
regime
in
light
of
the
nature
of
the
problem.
We
note
that
for
750
MW
power
plants,
such
a
line
drawing
exercise
is
likely
to
be
unnecessary,
as
such
sources
will
in
most
or
all
cases
have
impacts
far
exceeding
1.0
deciviews.

Commenter
0221
said,
in
order
to
foster
consistency
among
states
in
the
same
region
and
integrity,
EPA
should
work
through
the
RPOs
to
reconcile
all
BART
modeling
results
among
the
affected
states.
This
could
be
readily
accomplished
by
the
regional
modeling
centers
sponsored
by
the
RPOs,
utilizing
available
regional
modeling
data
(
i.
e.,
CMAQ)
A
periodic
formal
multi­
state
review
process
would
be
a
core
function
of
an
EPA
clearinghouse
and
would
involve
the
following
stakeholders:
state
representatives,
regional
planning
organizations,
federal
land
managers,
EPA
representatives,
industry
representatives,
environmental
groups,
and
other
relevant
parties.
The
review
process
would
be
used
to
ensure
the
development
and
use
of
screening
techniques
in
the
BART
determination
process
are
scientifically
defensible.
It
would
also
ensure
that
the
EPA's
modeling
criteria
are
utilized
as
set
forth
in
the
BART
Guidelines.
Lastly,
the
multi­
state
review
process
would
guarantee
that
BART
determinations
are
being
applied
consistently
across
BART­
eligible
sources
and
that
the
modeling
techniques
incorporate
the
latest
regional
modeling
data.

Response:

We
favor
coordination
among
States,
EPA
regions,
RPOs,
and
other
federal
agencies
to
agree
on
a
modeling
protocol(
s)
which
would
provide
consistent
application
of
the
model.
Such
coordination
may
relieve
the
administrative
burden
of
examining
and
approving
a
protocol
for
each
separate
BART­
eligible
source.
A
protocol
developed
by
and
acceptable
to
all
stakeholders
would
expedite
the
modeling
process
and
further
ensure
nationwide
consistency
in
application
of
the
model.
Coordinated
efforts
could
help
develop
standardized
data
sets
to
provide
the
added
benefit
of
reducing
the
cost
and
time
needed
for
a
modeling
analysis
and
facilitating
comparability
between
individual
analyses.
A
centralized
regional
database,
accessible
on
a
web
site,
may
provide
quality­
assured
upper
air,
surface,
precipitation,
monitoring,
land
use,
and
terrain
data
124
already
in
model­
ready
format.
Some
of
the
RPOs,
for
instance,
have
developed
data
sets
for
model
inputs
that
can
be
used
by
States,
further
reducing
time
and
cost
constraints
while
standardizing
the
model
input
requirements.
The
use
of
predefined
receptor
networks
developed
by
the
FLMs
can
also
relieve
resource
constraints.

Commenter
0253
said
the
collective
impact
of
the
parameters
proposed
by
EPA
will
likely
result
in
BART
being
triggered
on
a
wholesale
basis
for
sources
that
have
no
actual
impact
on
visibility
conditions
in
a
Class
I
area.
Stacking
the
deck
in
that
manner
could
result
in
significant
expenditures
by
electric
utilities
that
would
result
in
no
real
improvement
in
visibility
conditions.
Consequently,
the
commenter
urged
EPA
to
revise
its
proposed
test
to
incorporate
parameters
which
more
accurately
reflect
a
source's
impact
on
visibility,
including:
(
1)
a
threshold
of
no
less
than
one
deciview;
(
2)
a
comparison
with
current
visibility
conditions
which
provides
a
more
accurate
benchmark;
and
(
3)
using
the
20%
worst
days
as
the
primary
metric
in
evaluating
changes
in
visibility
to
be
more
consistent
with
the
approach
used
elsewhere
by
EPA
under
the
regional
haze
program.

Response:

See
several
discussions
above
regarding
State
discretion
in
setting
an
appropriate
cause
or
contribute
threshold,
the
comparison
of
modeling
results
to
natural
background
conditions,
and
the
use
of
the
maximum
24­
hour
impact
or
the
99th
percentile
value
in
making
a
cause
or
contribute
assessment.
States
also
have
discretion
in
the
use
of
a
metric
or
combination
of
metrics
in
making
the
degree
of
visibility
improvement
determinations
used
in
their
BART
control
decisions.

Comment:
Four
commenters
said
the
guidelines
are
deficient
in
providing
clear
criteria
for
states
to
use
in
determining
which
BART­
eligible
sources
become
"
subject
to
BART"
(
28­
343,
28­
345,
28­
348,
28­
387).
Commenters
(
28­
348,
28­
387)
said
there
are
no
objective
or
clearly
defined
criteria
for
determining
which
"
BART­
eligible"
sources
become
sources
that
are
"
subject
to
BART."
The
regulation
strongly
implies
that
all
BART­
eligible
sources
should
be
deemed
to
be
"
subject
to
BART."
For
example,
the
emissions
trading
proposal
refers
to
BART­
eligible
rather
than
"
subject
to
BART"
sources
as
potential
participants
in
the
trading
program.
Commenter
28­
387
added
that
EPA
should
require
that
states
objectively
assess
each
BART­
eligible
source,
taking
into
account
the
levels
of
the
various
types
of
BART­
related
emissions,
distances
from
Class
I
areas,
or
other
appropriate
thresholds,
to
insure
that
eligible
sources
not
impacting
Class
I
areas
do
not
become
"
subject
to
BART."

Response:

See
discussions
above
regarding
State
discretion
in
setting
an
appropriate
cause
or
contribute
threshold
125
Commenter
28­
343
said
the
proposed
guidelines
collapse
the
two
distinct
steps
of
BARTeligibility
and
BART­
subjectivity
into
one
by
making
it
difficult
for
states
to
prove
"
trivial
contribution".
In
this
way,
all
BART­
eligible
sources
become
subject
to
BART
and
the
state's
responsibility
is
usurped.
Commenter
28­
345
added
that
under
EPA's
proposal,
at
least
three
of
the
eight
clauses
of
the
statutory
matrix­­
namely
technical
feasibility,
cost,
and
non­
air
quality
environmental
impact­­
would
become
instead
a
source
of
burdens
on
the
states,
which
would
be
required
to
analyze
these
factors
in
a
manner
specified
by
the
EPA
and
subject
to
detailed
retrospective
EPA
review
before
they
could
make
any
decision
of
their
own.
The
result,
quite
clearly,
would
be
immense
pressure
on
states
to
adopt
substantive
BART
standards
acceptable
to
EPA.

Response:

The
guidelines
acknowledge
the
State's
authority
in
the
BART
determination
process.

Comment:
One
commenter
(
28­
306)
said
clearer
guidance
that
includes
exemption
for
sources
employing
large­
scale
emission
control
equipment
is
essential
to
ensure
cost­
effectiveness
in
implementing
BART
requirements
on
industries
currently
investing
in
heavy
pollution
controls.
Commenter
28­
392
said
the
actual
costs
and
benefits
of
controls
should
be
considered
in
determining
whether
insignificant
contributors
to
visibility
impairment
need
to
revise
their
emission
limitations.

Response:

Once
a
State
has
determined
that
a
BART­
eligible
source
causes
or
contributes
to
visibility
impairment,
that
State
is
required
to
determine
BART
controls
based
on
the
five
factor
analysis.
Those
factors
are:
the
technology
available,
the
costs
of
compliance,
the
energy
and
nonair
quality
environmental
impacts
of
compliance,
any
pollution
control
equipment
in
use
at
the
source,
and
the
remaining
useful
life
of
the
source.
The
guidance
discusses
each
of
these
factors
and
should
adequately
address
the
above
comment.

4.0
BART
Determination
4.1
Engineering
Analysis
4.1.1
Identify
Available
Retrofit
Emission
Control
Technologies
4.1.1.1
Technologies
Not
Yet
Applied
in
Full
Scale
Operation
Comment:
126
Five
commenters
disagreed
that
a
control
technology
is
applicable
if
it
had
been
specified
in
a
permit
(
0184,
0240,
28­
267,
28­
273,
28­
385).
They
said
this
should
not
be
the
sole
basis
for
control
consideration.
Controls
are
sometime
part
of
a
construction
permit
and
are
functioning
in
a
"
first­
of­
its­
kind
application,"
which
means
the
use
and
emission
limits
are
based
upon
the
results
from
using
that
control.
Until
the
control
is
tested
in
a
full­
scale
practice
it
should
not
be
considered
applicable.
Commenter
28­
273
asked
for
clarification
by
what
is
meant
by
"
commercial
construction."

Response:

We
believe
that
if
a
control
technology
has
been
specified
in
a
permit,
it
provides
evidence
that
the
technology
is
being
successfully
deployed
in
practice,
which
is
an
important
common
sense
indicator
that
it
is
available.

4.1.1.2
Technologies
as
Stringent
as
NSPS/
PSD
Comment:
Ten
commenters
(
0208,
0217,
0253,
0270,
28­
345,
28­
349,
28­
364,
28­
385,
28­
7004,
28­
7007)
addressed
EPA's
proposal
to
eliminate
the
presumption
that
any
facility
that
complies
with
NSPS
is
also
in
compliance
with
BART.
Commenters
(
0208,
0217,
0270)
said
BART
was
created
within
the
Act
to
require
the
retrofit
of
facilities
that
were
not
subject
to
NSPS.
Also
they
state
that
NSPS
was
to
be
used
as
a
ceiling
for
BART
as
specified
by
Congress
and
recognized
by
EPA
in
the
1980
Guidelines.
They
do
not
feel
that
it
was
the
intent
of
Congress
to
make
BART
more
stringent
than
NSPS.
EPA
should
not
attempt
to
require
more
from
BART­
eligible
facilities
than
from
new
facilities
built
just
after
BART
requirement
was
placed
in
the
Act.
Commenter
28­
7004
added
that
analyzing
for
scenarios
more
stringent
than
NSPS
standards
is
totally
unwarranted
for
sources
25
to
40
years
old,
however,
an
economic
study
of
upgrading
to
NSPS
standards
has
some
merit.

Five
commenters
said
NSPS
should
be
the
presumptive
control
level
(
28­
345,
28­
349,
28­
364,
28­
385,
28­
7007).
In
other
words,
the
state
should
deem
the
NSPS
control
level
to
be
BART
unless
it
determines
based
on
its
application
of
the
six
statutory
factors
that
the
technology
is
not
appropriate
for
a
particular
source
(
28­
345,
28­
349).
When
BART
provisions
were
passed
in
1977,
Congress
expected
that
BART
sources
would
be
retrofitted
with
the
control
technologies
available
in
1977
or
within
the
next
few
years­­
not
those
developed
25
or
more
years
after
that
(
28­
349,
28­
364).
Commenter
28­
385
said
the
Act
expressly
requires
that
EPA
periodically
update
promulgated
NSPS
(
no
less
frequently
than
at
least
every
8
years).
EPA's
failure
to
do
this
should
not
be
used
as
a
reason
to
jettison
the
opportunity
to
use
applicable
NSPS
as
presumptive
BART
levels.
Rather,
EPA
should
fulfill
its
statutory
responsibility
under
the
NSPS
program,
allowing
NSPS
to
be
the
presumptive
control
level.

Response:
127
We
believe
there
are
many
cases
where
the
NSPS
or
other
required
control
level
is
out
of
date
 
for
example,
it
may
be
20
or
more
years
old,
and
may
have
been
outstripped
by
advances
in
control
technology.
We
believe
that
common
sense
indicates
that
such
advances
in
control
technology
should
not
be
summarily
ignored,
but
should
be
seriously
considered
in
a
BART
review.
(
This
is
also
why
we
are
finalizing
the
specific
change
to
the
1980
BART
guidelines
discussed
in
the
proposal
preamble).

Comment:
Three
commenters
(
0237,
28­
275,
28­
333)
said
BACT
and
LAER
can
only
be
applied
to
new
sources.
However,
for
sources
that
already
have
BACT
or
LAER,
these
controls
should
be
considered
to
satisfy
BART
(
0253).

Three
commenters
(
0237,
0275,
0299)
supported
EPA's
proposal
to
allow
MACT
compliance
as
BART
compliance
because
for
several
sources
MACT
is
more
stringent
than
BART
for
VOCs,
SO
X,
and
PM.
They
are
concerned
however
that
the
states
may
have
more
costs
due
to
additional
analysis
of
new
technologies.
The
states
should
not
be
required
to
do
a
BART
analysis
for
sources
subject
to
MACT.

Comment:
Thirteen
commenters
said
all
control
technologies,
including
those
more
stringent
than
NSPS,
should
be
considered
(
0221,
0255,
28­
255,
28­
263,
28­
266,
28­
267,
28­
278,
28­
290,
28­
302,
28­
312,
28­
325,
28­
327,
28­
358).

One
commenter
(
0221)
said
that
the
NSPS
should
never
have
been
considered
the
ceiling
for
BART.
Commenters
(
0255,
28­
263,
28­
266,
28­
302)
agreed
that
NSPS
do
not
represent
the
best
control
strategies
for
existing
sources.
Commenter
28­
327
said
while
NSPS
served
as
a
starting
point,
implementation
of
BART
should
consider
controls
from
a
top­
down
approach,
with
NSPS
serving
as
the
floor
only
to
prevent
severe
erosion
of
use
of
effective
and
available
control
requirements.
Commenter
28­
358
added,
however,
that
they
fundamentally
disagree
that
the
1980
guidelines
ever
precluded
a
state
from
establishing
as
BART
for
purposes
of
remedying
reasonably
attributable
impairment,
an
emission
limitation
that
is
more
stringent
than
the
NSPS
level
of
control.

Response:

While
we
do
not
believe
it
is
appropriate
to
prejudge
specific
MACT
determinations,
we
note
that
we
expect
that
in
many
cases
MACT
control,
specifically
for
PM
and
VOC,
is
likely
to
be
judged
equivalent
to
BART.
This
may
also
be
true
for
sources
that
have
had
modifications
subject
to
BACT
or
LAER.
As
States
array
control
options,
an
older
NSPS
would
represent
a
floor
level
of
control,
but
in
many
cases
would
likely
be
an
inappropriate
level
of
control
for
BART,
given
advancements
in
control
technology
over
time.

4.1.1.3
Consideration
of
Controls
Already
in
Place
128
Comment:
Five
commenters
(
0171,
0189,
0202,
0214,
0245)
stated
that
EPA
has
not
indicated
what
controlled
sources
must
do
to
comply
with
the
BART
rule,
while
EPA
has
determined
what
controls
are
necessary
for
uncontrolled
sources.
Commenter
0214
was
concerned
that
the
strictness
of
the
guidelines
will
lead
states
to
require
controlled
sources
to
either
remove
the
controls
for
even
newer
controls
at
great
cost
and
very
little,
if
any,
improvement
in
emission
levels
and
visibility
in
Class
I
areas.
States
should
be
able
to
use
their
discretion
to
determine
whether
additional
controls
are
needed.

Commenters
(
0189,
28­
390)
noted
that
sources
already
subject
to
controls
under
the
"
NO
X
SIP
call"
should
not
be
subject
to
the
BART
rule,
at
least
during
the
first
10­
year
planning
period
(
28­
390).
Commenter
28­
390
added
that
otherwise,
the
NO
x
controls
currently
being
installed
on
non­
BART­
eligible
sources
to
comply
with
the
NO
x
SIP
call
will
be
rendered
unnecessary
and
therefore
stranded
investments
since
the
combined
NO
x
emission
reduction
from
BART
and
non­
BART
sources
would
exceed
NO
x
SIP
call
requirements.
Others
(
0202,
0245,
28­
306)
indicated
that
those
sources
using
MACT
should
not
be
subject
to
the
BART
rule.
Commenter
0245
added
that
existing
nonattainment
strategies
in
states
may
also
satisfy
BART.

Response:

The
BART
review
is
the
appropriate
stage
for
States
to
take
into
account
controls
applied
under
the
NOx
SIP
call,
which
we
do
not
believe
should
be
considered
unnecessary
for
all
purposes.
We
do
note
that
controls
such
as
SCR
installed
under
the
NOx
SIP
call
should
satisfy
BART
with
just
the
relatively
smaller
added
expense
of
running
year­
round.

Comment:
Commenters
(
28­
403,
28­
779)
agreed
that
BART
does
not
require
the
redesign
of
a
source
when
considering
available
control
alternatives.
Commenter
28­
779
said
this
is
especially
true
for
BART,
where
existing
sources
are
being
affected.
EPA
states
that
"
design
changes"
may
be
available
for
a
source.
However,
there
may
be
considerable
confusion
between
what
constitutes
"
redesign"
of
a
source
as
opposed
to
merely
"
design
changes".
Commenter
28­
403
noted
that
while
design
changes,
combustion
technology
changes
and
fuel
substitutions
may
be
options
for
source
owners
to
consider
when
evaluating
least­
cost
alternatives
and
other
compliance
strategies
under
an
emission
trading
program,
such
options
are
clearly
beyond
the
purview
of
the
states
and
EPA
in
the
course
of
evaluating
of
appropriate
BART
emission
control
options.

Commenter
28­
267
said
fuel
switching
from
coal
to
natural
gas
should
not
be
a
BART
requirement
for
a
coal­
fired
boiler
meeting
BACT
control
requirements.

Response:
129
In
some
cases,
retrofit
design
changes
(
for
example,
use
of
low
NO
X
burners)
may
be
available
for
making
a
given
production
process
or
emissions
unit
inherently
less
polluting
(
although
because
BART
applies
to
existing
sources,
we
recognize
that
there
will
probably
be
far
fewer
opportunities
to
consider
inherently
lower­
emitting
processes
than
may
be
available
for
NSR).
In
such
cases,
the
ability
of
design
considerations
to
make
the
process
inherently
less
polluting
is
appropriate
for
inclusion
as
a
control
alternative
for
the
source
at
this
first
step.
Costs
and
other
factors
will
be
evaluated
in
a
later
step.

We
note
that
we
have
added
language
to
the
guidelines
to
note
that
we
do
not
intend
switching
fuel
form
(
ie.
from
coal
to
gas)
to
be
a
recommended
BART
option.

4.1.1.4
Identify
All
Available
Technologies
Comment:
Five
commenters
addressed
the
need
to
identify
all
available
retrofit
control
technologies
(
28­
302,
28­
327,
28­
266,
28­
278,
28­
312).
Commenter
28­
327
noted
that
the
control
technologies
needed
are
available,
proven,
and
continually
being
improved
on.

Commenter
28­
302
said
the
guidelines
should
be
clarified
to
ensure
that
combinations
of
control
technologies
can
greatly
increase
control
efficiencies
relative
to
the
use
of
a
single
control
technology
in
isolation.
In
contrast,
other
commenters
stressed
that
the
guidelines
should
make
it
clear
that
BART
does
not
include
installation
of
multiple
add­
on
control
equipment
at
a
single
emissions
unit
(
28­
266,
28­
278,
28­
312)

Response:

Combinations
of
inherently
lower­
polluting
processes/
practices
(
or
a
process
made
to
be
inherently
less
polluting)
and
add­
on
controls
could
possibly
yield
more
effective
means
of
emissions
control
than
either
approach
alone.
Therefore,
the
option
to
use
an
inherently
lower­
polluting
process
does
not,
in
and
of
itself,
mean
that
no
additional
add­
on
controls
need
to
be
included
in
the
BART
analysis.
These
combinations
would
be
included
in
step
1
and
evaluated
later
in
the
BART
process.

4.1.2
Determine
Whether
Options
are
Technically
Feasible
Comment:
Four
commenters
(
0184,
0207,
0229,
0237,
28­
273,
28­
358,
28­
779)
discussed
the
definition
of
available
options.

Two
commenters
agree
with
EPA
stating
that
technologies
considered
in
the
"
pilot
stage
should
not
be
considered
available
(
0184,
0237),
however
both
voiced
concerns
that
the
proposed
rule
did
not
clearly
differentiate
between
"
commercial
demonstration
and
licensing"
phases
and
130
when
the
technology
would
be
considered
available
for
each
phase.
Commenter
0229
agrees
with
EPA
and
supports
the
clarification
as
it
applies
to
coal­
fired
generators
not
being
replaced
by
natural
gas
turbines.
Commenter
0207
agreed
with
the
definition
as
set
by
EPA,
but
in
the
event
that
the
technology
becomes
unavailable,
due
to
increased
demand
resulting
from
implementation
of
new
regulations,
thinks
that
the
states
should
have
the
flexibility
in
scheduling
retrofit
projects.
In
contrast,
other
commenters
(
28­
358)
said
because
these
guidelines
are
being
considered
almost
a
decade
before
the
emissions
controls
requirements
must
be
in
place,
the
concept
of
"
availability"
of
a
technology
for
regional
haze
BART
purposes
should
be
expanded
to
include
those
that
are
in
the
pilot
scale
testing
phase,
not
simply
those
that
have
been
licensed
and
demonstrated
as
EPA
proposes.

One
commenter
(
0237)
did
not
agree
with
EPA
allowing
technology
to
be
considered
available
if
it
became
available
before
the
end
of
public
comment.
They
indicated
that
the
states
have
used
a
lot
of
resources
in
determining
appropriate
technologies,
and
that
time
and
effort
will
be
wasted.
They
suggest
that
it
may
be
prudent
to
wait
until
the
end
of
the
public
comment
period
when
defining
available
options.
Additionally,
they
do
not
think
that
foreign
technology
should
be
considered
without
sufficient
performance
experience.
Foreign
technologies
often
have
less
stringent
emissions
requirements
than
those
imposed
by
EPA.
Lastly,
they
disagree
with
EPA
stating
that
process
and
raw
material
changes
must
be
analyzed
as
BART
alternatives.
Retrofit
technology
should
not
include
process
and
rat
material
changes
given
its
definition;
also,
HAP
emissions
from
existing
sources
have
been
adequately
considered
as
indicated
by
the
Act.

One
commenter
(
28­
385)
said
EPA
should
modify
the
final
guidelines
to
expressly
read:
(
1)
Only
a
reasonable
number
of
technologies
should
be
considered
such
that
an
appropriate
range
of
technologies
with
varying
reduction
severities
is
assessed;
(
2)
Technologies
that
have
not
been
used
by
a
pertinent
source
category
or
that
have
not
been
used
within
the
United
States
on
domestic
fuels
should
be
considered
only
if
the
range
of
technologies
otherwise
evaluated
is
inordinately
small
in
number.
In
such
an
event,
the
states
should
be
cautioned
about
potential
transfer
problems
of
such
technologies
and
directed
to
appropriately
account
for
such
uncertainties
when
weighing
the
various
BART
criteria.

Response:

While
States
have
discretion
over
these
considerations,
the
guidelines
generally
try
to
use
and
interpret
these
concepts
in
a
common
sense
manner.
If
a
control
technology
is
available
for
use,
generally
it
means
that
other
sources
with
similar
(
but
not
necessarily
identical)
design
have
used
it
successfully.
In
addition,
we
believe
it
is
reasonable
and
an
exercise
of
good
government,
for
States
to
consider
information
that
is
made
available
during
the
public
comment
period.
Finally,
at
the
discretion
of
the
State,
available
air
pollution
control
technologies
can
reasonably
include
technologies
employed
outside
of
the
United
States
that
have
been
successfully
demonstrated
in
practice
on
full
scale
operations,
particularly
those
that
have
been
demonstrated
as
retrofits
to
existing
sources.
131
Generally
for
the
BART
review,
States
should
consider
lower­
polluting
processes
based
on
demonstrations
from
facilities
manufacturing
identical
or
similar
products
using
identical
or
similar
raw
materials
or
fuels.
Add­
on
controls,
on
the
other
hand,
should
be
considered
based
on
the
physical
and
chemical
characteristics
of
the
pollutant­
bearing
emission
stream.
Thus,
candidate
add­
on
controls
may
have
been
applied
to
a
broad
range
of
emission
unit
types
that
are
similar,
insofar
as
emissions
characteristics,
to
the
emissions
unit
undergoing
BART
review.

Comment:
Two
commenters
(
28­
273,
28­
779)
said
the
cut­
off
for
incorporating
available
technologies
should
be
at
the
initial
SIP
proposal
date
and
not
before
the
closure
of
the
comment
period.
Commenter
28­
779
added
that
this
will
avoid
debate
at
a
later
date
as
to
technologies
that
should
be
considered
in
the
engineering
analysis.

One
commenter
(
0184)
states
that
the
definition
of
applicable
technology
is
too
broad
and
that
the
definition
needs
to
be
redefined
to
make
technology
not
only
applicable
but
efficient
too.

Other
commenters
said
EPA
should
revise
the
provision
to
include
a
vendor
guarantee
because
without
a
guarantee
facilities
have
no
means
of
determining
what
emissions
reductions
can
be
expected
(
0184,
0240,
28­
273).
Commenter
0255
provided
specific
information
regarding
SO
2
control
technologies
for
furnace
sorbent
injection,
dry
sorbent
injection,
and
wet
sorbent
injection,
as
well
as
NO
X
control
in
combustion
and
post
combustion
controls
(
See
Table
4
Appendix
A)
for
BART­
eligible
emission
units
from
26
different
industrial
categories.

Comment:
Two
commenters
said
the
guidelines
should
more
clearly
address
technically
infeasible
options
(
28­
244,
28­
385).
Commenter
28­
244
said
Congress
clearly
intended
to
leave
judgments
of
technological
feasibility
to
the
individual
States
subject
to
the
overall
requirement
to
make
reasonable
progress
toward
the
visibility
goal.
EPA's
proposal,
however,
would
compel
States
to
regard
proposed
technologies
as
feasible
unless
they
could,
in
effect,
prove
a
negative
by
showing
infeasibility
to
EPA's
satisfaction.
Commenter
28­
385
said
the
guidelines
should
more
affirmatively
acknowledge
that
with
retrofit
applications,
source­
specific
considerations
can
preclude
application
of
some
otherwise
feasible
control
technologies.

Response:

As
noted
above,
we
believe
that
considerations
of
good
government
make
it
incumbent
on
States
to
consider
information
that
is
made
available
during
the
public
comment
period.
In
addition,
including
controls
for
considerations
at
step
1
of
the
BART
process
does
not
preclude
their
being
evaluated
critically
later
in
the
BART
process.
The
States
have
considerable
discretion
in
undertaking
BART
determinations
and
have
the
ability
to
make
reasoned
judgments
as
to
technical
feasibility.
132
4.1.3
Evaluate
Impacts
and
Report
Results
4.1.3.1
Estimate
the
Cost
of
Control
to
Account
for
Remaining
Useful
Life
Comment:
Twenty
commenters
(
0184,
0189,
0207,
0221,
0232,
0237,
0271,
0286,
28­
247,
28­
248,
28­
254,
28­
258,
28­
260,
28­
338,
28­
352,
28­
358,
28­
362,
28­
389,
28­
399,
28­
7016)
addressed
the
proposed
closure
provisions
for
sources
with
limited
remaining
useful
life.

Commenters
(
28­
358,
28­
362,
28­
389)
agreed
with
EPA's
proposal
to
set
compulsory
retirement
dates
5
years
after
the
date
of
the
SIP
revision.
Commenter
28­
358
added
that
if
the
source
has
not
been
permanently
retired
by
the
specified
retirement
date,
the
owner
or
operator
of
that
source
must
either
(
a)
shut
down
temporarily,
make
changes
to
ensure
that
BART
emissions
levels
can
be
achieved,
without
taking
remaining
useful
life
into
consideration,
and
pay
a
preset
fine;
or
(
b)
immediately
and
permanently
shut
down.

Although
several
commenters
agreed
with
EPA
that
a
source
should
be
able
to
extend
the
useful
life
when
applying
BART
controls,
they
said
there
should
be
a
federally
enforceable
closure
date
when
there
is
little
remaining
useful
life
included
in
the
operating
permit
(
0189,
0221,
0232,
0237,
0271,
0286,
28­
247,
28­
248,
28­
254,
28­
258,
28­
260,
28­
338,
28­
352,
28­
358,
28­
362,
28­
399,
28­
7016).
Commenters
(
28­
248,
28­
338,
28­
399,
28­
7016)
added
that
these
facilities
also
must
be
required
to
abate
pollution
levels
in
the
interim.

Commenter
0207
said
in
determining
BART
the
states
should
be
allowed
subjective
judgement
regarding
other
statutory
factors.
Commenter
28­
244
said
many
sources
will
quite
clearly
have
limited
useful
lives
even
if
their
owners
are
unwilling
to
commit
to
a
precise
shutdown
date.
Nothing
in
the
statute
forbids
States
to
consider
the
likelihood
that
these
sources,
too,
will
have
a
"
limited
useful
life"
when
they
make
BART
decisions.
Commenter
0237
notes
that
there
is
no
provision
for
decision
changes
when
a
shut
down
is
apparent.
The
Title
V
permit
could
have
a
small
amendment
which
would
indicate
that
the
change
is
not
considered
a
change
in
the
method
of
operation.

Seven
commenters
discussed
the
flexibility
given
to
sources
to
adopt
either
option
or
to
change
their
mind
at
a
future
point
(
0184,
28­
255,
28­
306,
28­
273,
28­
770,
28­
243,
28­
404,
28­
779).
Commenters
(
28­
404,
28­
779)
said
sources
should
have
the
option
to
choose
either
option.
Commenter
28­
243
stressed
that
flexibility
should
be
combined
with
a
provision
for
a
date
certain
component.

A
facility
should
also
be
given
the
opportunity
to
install
BART
on
sources
to
prolong
their
useful
life
if
they
had
not
during
determination,
or
if
the
unit
was
idle
during
the
determination
period
(
0184).
Commenter
28­
255
said
if
a
facility
decides
to
stop
operation
instead
of
installing
BART,
it
must
do
so
no
later
than
it
would
have
had
to
have
had
BART
installed
and
operational.
If
it
wants
to
keep
operating
after
that
date,
it
must
install
BART
by
that
date
or
get
offsets
from
133
other
sources
until
BART
can
be
installed.
Commenter
28­
273
said
a
source
that
maintains
continuous
operation
after
BART
becomes
effective
and
changes
its
mind
about
closure
should
be
allowed
the
opportunity
to
install
BART.
If
the
source
believes
it
may
continue
beyond
the
remaining
useful
life
predicted
date,
it
could
take
a
permit
that
requires
back­
applying
the
appropriate
BART
by
that
date,
or
ceasing
operations
by
that
date
until
that
BART
is
installed.
Similar
considerations
should
be
given
to
idle
units.
Commenter
28­
770
added
that
if
an
operator
intends
to
shut
down
a
source
by
a
given
date,
but
then
wishes
to
continue
operating
beyond
the
5
years,
the
source
would
be
required
to
install
the
level
of
controls
that
would
have
been
considered
BART
if
the
BART
analysis
had
not
assumed
a
reduced
remaining
useful
life.

Commenter
28­
306
said
EPA
should
provide
for
deferral
of
applicable
BART
requirements
for
facilities
that
have
curtailed
operations.
Flexibility
with
respect
to
maintaining
consistency
with
the
statutory
requirement
to
install
BART
within
5
years
of
EPA's
approval
of
the
relevant
SIP
is
of
potential
concern
for
a
large
segment
of
the
aluminum
industry
for
whom
40%
of
domestic
primary
aluminum
production
capacity
is
fully
curtailed
for
an
indefinite
period
of
time
due
to
the
lack
of
affordable
electricity
in
the
Northwest.

Two
commenters
said
the
states
should
be
given
flexibility
(
28­
342,
28­
361).
Commenter
28­
342
said
the
states
were
limited
to
case­
by­
case
assessments
of
control
costs
.
Commenter
28­
361
said
EPA's
top­
down
approach
would
require
the
most
stringent
controls
as
the
"
best"
alternative
if
a
facility
wants
to
operate
more
than
5
years
beyond
the
date
of
the
SIP
approval.
Neither
option
provides
true
State
flexibility.

Response:

We
have
retained
the
provision
that
intent
to
shut
down
should
be
accompanied
by
a
permit
provision
to
that
effect.
We
note
that
without
such
a
permit
provision,
it
would
be
difficult
for
a
State
to
judge
that
a
source
will
clearly
have
a
limited
useful
life.
In
addition,
the
CAA
requires
that
controls
be
installed
within
five
years
of
a
BART
determination.
Therefore
in
order
to
be
consistent
with
the
statute,
we
include
the
provision
that
a
source
would
not
operate
more
than
five
years
beyond
the
statutory
compliance
date.

Comment:
Five
commenters
said
remaining
useful
life
should
be
considered
in
the
economic
analysis
(
28­
344,
28­
385,
28­
387,
28­
392,
28­
7013).
Commenter
28­
344
said
EPA's
approach
seems
to
imply
that
a
source
which
has
a
10­
year
remaining
useful
life
could
not
take
credit
for
it
in
the
BART
analysis
because
it
would
by
definition
be
in
operation
in
5
years.
In
addition,
Congress
did
not
require
enforceable
restrictions
in
order
to
take
into
account
the
remaining
useful
life
of
a
source
and
EPA
cannot
do
so
in
this
Guideline
document.
This
would
be
an
unprecedented
attempt
to
regulate
commercial
and
business
activity
as
well
as
problematic
for
sources
that
might
later
decide
in
lieu
of
shutdown
to
modify
specific
units
and
trigger
NSR.
Commenter
28­
387
agreed
that
EPA's
proposal
exceeded
statutory
authority
and
suggested
that
states
should
be
encouraged
to
use
reasonable
estimates
of
remaining
life
for
each
facility
using
objective
criteria
134
or
indicators
such
as
engineering
estimates,
publicly
available
documents,
information
provided
by
the
affected
source,
and
the
existence
of
newly­
installed
pollution
control
equipment
(
e.
g.
selective
catalytic
reduction
to
meet
ozone
SIP
requirements).
Commenter
29­
385
recommended
that
BART
determinations
where
remaining
useful
life
is
a
significant
criterion
be
re­
evaluated
each
of
the
3
years
prior
to
the
statutory
deadline
for
installing
controls.
Commenters
(
28­
392,
28­
7013)
said
the
remaining
useful
life
should
be
used
to
amortize
the
cost
of
emission
control
projects.

Six
commenters
said
remaining
useful
life
should
not
be
considered
when
calculating
control
costs
(
28­
127,
28­
258,
28­
263,
28­
302,
28­
355,
28­
358).

Some
commenters
(
28­
258,
28­
358)
opposed
providing
flexibility
to
adjust
facility
shutdown
according
to
market
conditions,
as
it
will
only
delay
implementation
of
the
best
available
technology.
If
a
facility
does
not
want
to
accept
an
enforceable
restriction
for
ending
operations,
then
the
remaining
useful
life
should
have
no
impact
on
control
costs
and
the
BART
determinations.
Commenter
28­
358
added
that
if
a
source
is
deemed
to
have
a
short
remaining
useful
life,
its
operator
should
not
be
permitted
under
any
circumstances
to
speed
up
the
amortization
of
BART
controls
or
avoid
fully
applying
BART
at
the
end
of
that
agreed
upon
time
period.
If
an
operator
wants
flexibility,
the
operator
must
pay
for
it.
Similarly,
the
suggested
option
of
requiring
a
BART­
eligible
source
that
continues
in
operation
beyond
its
declared
useful
life
to
install
full
BART
controls
at
the
point
of
life
extension
makes
no
environmental
or
economic
sense
 
but
simply
creates
a
loophole
for
older
sources
to
avoid
BART
for
a
period
of
years.
Should
EPA
decide
to
allow
BART­
eligible
sources
with
limited
remaining
useful
lives
to
be
partially
exempted
from
the
requirement
that
they
install
BART,
the
retirement
date
must
be
included
as
an
operating
condition
in
the
sources'
Title
V
permits.

Response:

We
have
amended
the
calculation
of
remaining
useful
life
to
be
more
consistent
with
generally
accepted
accounting
practices
and
with
the
guidance
in
EPA's
Control
Cost
Manual.
The
calculation
now
begins
at
the
date
controls
will
be
put
in
place,
and
upfront
construction
costs
should
be
rolled
into
the
cost
calculation
for
the
first
year
of
operation
of
controls.
We
believe
this
is
the
fairest
and
most
justifiable
way
to
account
for
all
relevant
costs
and
for
the
entire
time
period
during
which
the
source
will
operate
with
controls.

Comment:
Commenter
28­
127
objected
to
applying
this
provision
to
power
plants
because
of
their
history
of
continuing
to
operate
old
plants
and
avoiding
installation
of
controls.
Commenter
28­
335
said
it
is
unfair
for
a
source
to
take
the
benefit
of
an
assumption
that
a
source
has
a
short
remaining
useful
life,
only
to
claim
later
that
changed
circumstances
have
proved
its
assumption
wrong.
However,
that
the
ability
to
seek
a
permit
modification
with
appropriate
conditions
would
provide
companies
with
some
flexibility.
The
initial
permit
would
have
to
contain
a
mechanism
to
135
address
the
benefit
to
the
source
as
a
result
of
the
projection
of
a
short
remaining
life,
in
the
event
that
circumstances
change
Commenter
28­
302
was
concerned
that
if
operators
retain
flexibility
and
decide
to
extend
the
life
of
their
facilities
after
BART
determinations
have
been
made,
then
these
facilities
may
continue
to
operate
at
a
higher
emission
rate
than
if
BART
had
been
installed.
There
appears
to
be
no
precedent
for
"
federally
enforceable
restrictions
preventing
further
operation".
Any
provisions
allowing
for
consideration
of
"
remaining
useful
lifetime"
in
calculating
control
costs
would
reduce
the
effectiveness
of
the
BART
program.
If
a
facility
subject
to
BART
is
in
operation
at
the
time
of
the
SIP
submittal,
it
must
have
plans
to
install
controls.
Commenter
28­
263
agreed
that
remaining
useful
life
has
no
bearing
on
calculating
control
costs,
or
any
other
aspect
of
BART.
If
a
source
is
operating
and
subject
to
BART,
it
must
control
its
emissions
using
BART.
No
flexibility
should
be
provided.

Response:

As
indicated
above,
the
amortization
of
remaining
useful
life
for
BART
purposes
should
provide
clear
information
relevant
to
control
costs
that
is
consistent
with
the
statutory
compliance
requirements.
This
information
will
help
States
assess
the
cost­
effectiveness
of
controls
and
the
likelihood
that
sources
will
or
will
not
shut
down.

4.1.3.2
Estimate
the
Cost­
effectiveness
of
Control
Comment:
Three
commenters
(
0187,
0189,
0236)
made
comments
regarding
the
baseline
emissions
of
a
source
as
it
applies
to
BART
determination.
Two
commenters
suggested
that
EPA
use
an
approach
similar
to
that
employed
by
NSR,
where
the
pre­
modification
emissions
are
averaged
over
24
months
as
determined
by
the
owner/
operator
over
a
5­
year
time
period
(
0187,
0236).
One
commenter
(
0189)
suggested
the
states
use
historical
emissions
data
instead
of
potential
to
emit.

Commenter
28­
244
said
EPA
should
make
clear
that
in
calculating
the
cost­
effectiveness
of
BART
controls,
states
should
consider
the
actual
historical
emissions
of
a
source,
not
its
"
potential
to
emit."
Clearly,
determining
whether
sources
and
parts
of
sources
are
"
major"
and
subject
to
BART
will
often
require
complex
judgments.
The
status
of
individual
units
in
the
1962­
1977
period
must
be
determined.
The
proper
approach,
from
the
standpoint
of
both
law
and
policy,
would
be
to
address
any
need
for
additional
controls
after
the
first
round
of
BART
as
part
of
the
"
reasonable
progress"
obligation.

Response:
136
The
guidelines
note
that
actual
emissions
should
generally
be
used
to
calculate
the
baseline
period
for
the
cost­
effectiveness
calculation.
While
the
guidelines
recognize
that
specific
circumstance
will
dictate
the
specific
period
over
which
to
calculate
actual
emissions,
the
guidelines
do
make
clear
that
the
time
period
should
be
a
realistic
depiction
of
anticipated
annual
emissions
 
in
other
words
it
should
be
a
period
representative
of
emissions
expected
in
the
future.

Comment:
Nine
commenters
(
0189,
0195,
0206,
0214,
0237,
0238,
0303,
0321,
0323)
made
comments
regarding
the
acceptability
of
WRAP
criteria
for
the
cost
of
control
of
a
source
as
it
applies
to
BART
determination.
Five
commenters
(
189,
0206,
0237,
0303,
0323)
agreed
with
EPA
stating
that
control
costs
as
identified
by
the
WRAP
with
the
average
removal
costs
less
than
$
500/
ton
as
low,
between
$
500
and
$
3,000/
ton
as
medium
and
over
$
3,000/
ton
as
high
were
helpful
criteria.
Commenter
0189
also
noted
that
an
upper
bound
of
$
2,000/
ton
would
be
more
reasonable
for
the
high
costs,
instead
of
those
determined
by
the
WRAP.
Commenter
0323
urges
EPA
to
use
caution
when
establishing
absolute
ranges
for
emission
reductions
noting
that
the
costs
of
controls
decreases
over
time.
Also
any
criteria
should
also
include
a
discussion
of
the
economic
benefits
of
improved
visibility.
Four
commenters
(
0195,
0214,
0238,
0321)
indicated
that
the
WRAP
criteria
for
the
cost
of
control
constrains
the
states'
flexibility
in
making
BART
determinations.

Ten
commenters
(
0187,
0189,
0207,
0229,
0236,
0237,
0275,
0277,
0285,
28­
273)
made
comments
regarding
cost­
effectiveness
as
a
screening
criteria
for
controls
of
a
source
as
it
applies
to
BART
determination.
Six
commenters
(
0187,
0189,
0236,
0237,
0275,
0277)
agreed
that
EPA's
assistance
with
cost­
effectiveness
analysis
would
be
helpful.
Commenters
0189
and
0275
gave
only
limited
support,
however,
noting
that
states
should
be
allowed
to
do
their
own
analysis
using
CALPUFF
and
chose
controls
based
on
the
most
cost­
effective
visibility
improvements.
Commenter
0275
added
that
EPA
should
establish
an
upper
bound
of
cost
effectiveness
beyond
which
controls
should
not
be
considered
for
BART,
setting
high
cost
thresholds
of
$
2,500/
ton
for
NO
X
and
$
2,000/
ton
for
SO
2.
Commenter
0237
noted
that
BACT/
LAER
include
expensive
costs
of
technology
when
making
major
modifications
which
are
not
justified
in
BART,
therefore
the
guidance
should
be
established
with
RACT­
type
of
approach
limiting
costs
to
$
1,500/
ton
thereby
streamlining
the
BART
inquiry.

Commenter
28­
273
said
EPA
should
provide
guidance
on
what
constitutes
cost
effectiveness.
The
BART
guidelines
should
not
follow
the
same
vague
cost­
effectiveness
criteria
as
BACT
determinations.

Response:

Although
we
did
not
explicitly
include
these
or
any
other
cost
criteria
in
the
guidelines,
we
agree
that
the
WRAP
criteria
are
helpful
cost
criteria
and
could
reasonably
be
used.
States
137
have
considerable
flexibility
to
exercise
judgement
and
take
specific
circumstances
into
account
when
weighing
cost­
effectiveness
as
one
of
the
statutory
BART
factors.

Comment:
Three
commenters
(
0207,
0229,
0285)
disagreed
with
EPA
assisting
states
in
their
costeffectiveness
analysis
for
BART
determination.
Commenter
0207
stated
that
the
proposal
ignores
the
visibility
benefits
of
emissions
reductions
for
$/
ton
cost­
effectiveness
approach
thus
ignoring
the
goal
of
the
Regional
Haze
program
which
is
to
improve
air
quality
and
decrease
visibility
impairment.
In
the
sense
that
the
Regional
Haze
program
goals
were
not
addressed,
commenter
0277
also
indicated
that
it
may
be
more
appropriate
to
establish
a
cost
effectiveness
based
on
visibility
improvements
rather
than
cost/
ton
reduction.
Commenter
0285
requests
that
EPA
make
visibility
impact
analysis,
as
it
applies
to
controls
alternatives,
part
of
the
cost
estimation
process
in
BART.
Cost
is
an
important
consideration
in
determining
technical
feasibility
and
should
not
be
discounted
by
EPA
set
costs
(
0229).

The
rule
should
allow
states
to
weight
visibility
improvements
benefits
of
BART
against
the
costs.
Conducting
air
quality
analysis
should
only
be
required
to
justify
a
decision
on
a
control
level.
Also,
EPA
should
clarify
that
the
cost
of
emission
reduction
is
balanced
again
visibility
improvement
benefits
as
emphasized
in
the
Corn
Growers
court
(
0275).
States
should
have
the
authority
to
determine
what
is
cost
effective
when
determining
BART
(
0184).

Commenter
28­
398
noted
that
the
proposed
Guidance
provides
that
the
"
costs
of
compliance"
cost­
benefit
factor
is
the
cost
in
dollars
per
ton
of
emissions
avoided
rather
than
the
cost
in
dollars
per
increment
of
visibility
improved.
Commenter
28­
405
said
EPA
should
reject
simplistic
"
dollar
per
ton
reduced"
measures
of
cost­
effectiveness
in
favor
of
end­
result
measures
such
as
visibility
benefits
per
dollar
of
control
cost,
as
tools
to
assess
the
benefits
of
alternative
control
strategies.

Response:

We
believe
it
is
not
possible
to
come
up
with
a
simple
numeric
cost/
visibility
metric
that
acounts
for
the
interaction
of
PM
precursors
in
the
atmosphere,
geographic
variability
in
visibility
improvement
from
specific
levels
of
emissions
reductions,
and
many
other
factors.
Since
we
do
not
have
information
on
how
such
a
metric
could
be
developed,
we
do
not
believe
it
is
appropriate
for
us
to
recommend
such
a
metric
to
States.

4.1.3.3
Other
Comments
on
Cost
Effectiveness
and
Estimating
Cost
of
Control
Comment:
Seven
commenters
(
0184,
0199,
0237,
0253,
0275,
0302,
0304)
had
additional
statements
regarding
the
cost
effectiveness
of
controls.
138
Retrofitting
existing
controls,
that
are
already
providing
low
emissions
rates,
is
not
cost
effective,
especially
when
the
controls
could
serve
as
BART
(
0199).
Further
it
is
necessary
to
know
the
cost
of
existing
controls
if
the
average
costs/
ton
are
considered,
and
not
assume
the
baseline
is
"
no
control"
(
0237).

EPA
cost
estimates
significantly
underestimate
the
actual
costs
of
developing
and
implementing
pollution
control
technology
as
part
of
BART
analysis
(
0253).
EPA
defined
cost
of
$
900/
ton
emissions
reduction
is
under
half
of
the
cost
considered
effective
in
recent
regulations
like
CAIR,
and
the
costs
for
the
NO
X
SIP
Call
were
also
underestimated
(
Table
3,
Appendix
A,
Attachment
13)
(
0255).

EPA
air
pollution
control
rates
and
emissions
levels
seem
to
be
specified
for
electric
generating
units
(
EGU)
without
consideration
for
other
sources
that
may
be
subject
to
BART.
Those
non­
EGU
sources
may
be
subject
to
MACT
regulations,
therefore
there
should
be
sufficient
data
for
EPA
to
determine
cost
effectiveness
for
BART.
EPA
needs
to
provide
clarification
for
non­
EGU
sources
regarding
the
most
cost
effective
emissions
reductions
under
BART
(
0302).

"
Cost
effective
controls
are
seldom
cost
effective,
not
because
they
are
inexpensive,
but
because
a
given
technology
controls
a
lot
of
emissions"
(
0304).

Response:

See
the
RIA
for
this
rulemaking
for
estimates
of
non­
EGU
source
control
costs,
and
discussion
of
the
range
of
cost­
effectiveness
of
controls
that
could
be
expected
for
different
source
categories.

Comment:
Four
commenters
addressed
the
idea
of
using
average
costs
(
0189,
28­
244,
28­
398,
28­
405).
Commenter
0189
said
the
use
of
average
cost,
instead
of
marginal
cost,
may
promote
technologies
that
would
otherwise
be
disqualified
because
average
costs
are
always
lower
than
marginal
costs.
In
contrast,
commenter
28­
244
said
EPA's
requirements
that
states
use
average
cost
contradicts
basic
economic
principles,
which
require
a
marginal
cost
approach
to
evaluating
the
costs
and
benefits
of
different
control
approaches.
Considering
average
cost
will
make
BART
candidates
that
would
flunk
an
intellectually
respectable
cost­
effectiveness
analysis
look
like
good
candidates
for
installation.
States
should
have
the
freedom
to
decide
to
pursue
visibility
improvement
in
a
manner
that
does
not
create
an
avoidable
risk
of
local
unemployment
and
economic
loss.

One
commenter
(
28­
362)
said
it
is
suggested
in
the
proposed
guidelines
that
the
potential
for
plant
shutdown
as
a
result
of
installing
control
technology
should
be
considered
in
the
BART
analysis.
This
suggestion
has
no
place
in
the
BART
guidelines.
Once
BART
for
the
source
is
139
determined,
market
forces
should
determine
whether
the
source
installs
BART
or
ceases
operation.

Five
commenters
(
0185,
0206,
0207,
0302,
0303)
had
additional
statements
regarding
estimating
the
cost
of
controls.
The
emissions
reductions
of
SO
2
and
NO
X
should
be
balanced
with
(
1)
minimizing
costs
to
residential,
commercial,
and
industrial
electric
consumers,
(
2)
easing
strains
on
the
natural
gas
supply
and
associated
pricing,
and
(
3)
continued
availability
of
reliable
electric
generation
(
0185).
States
should
also
weigh
all
of
the
statutory
factors
when
determining
BART
without
being
required
by
EPA
to
make
cost
the
most
important
and
deciding
factor
when
choosing
controls
(
0207).
Additionally,
smaller
facilities
should
be
able
to
consider
the
age
of
the
facility
and
improved
visibility
from
applying
BART,
and
not
just
cost
criteria
when
determining
controls
(
0206,
0303).
States
need
guidelines
that
promote
consistency
across
the
nation.
The
impact
analysis
of
cost
of
controls
is
almost
identical
to
BACT
determinations
for
NSR,
which
seem
to
be
very
subjective
(
0302).

Response:

We
note
that
once
costs
are
calculated,
States
have
considerable
flexibility
in
determining
how
those
costs
should
be
weighed.
In
addition,
we
do
believe
that
average
cost
provides
useful
information,
especially
when
tempered
by
comparison
with
incremental
costs
of
control.

4.1.3.4
Consideration
of
Non­
Air
Quality
Environmental
Benefits
Comment:
Seven
commenters
(
0221,
0232,
28­
290,
28­
302,
28­
358,
28­
387,
28­
404,
28­
7008)
support
EPA's
requirement
that
states
must
take
into
consideration
non­
air
quality
impacts
of
compliance
with
BART
to
include
environmental
benefits.
Also,
BART
guidelines
should
require
an
assessment
of
positive
impacts
that
will
result
from
applying
BART.
In
addition
to
decreasing
nitrogen
compounds
effective
BART
controls
would
also
(
1)
reduce
SO2­
related
acid
deposition,
(
2)
reduce
biomass
loss
as
associated
with
elevated
ozone
levels,
and
(
3)
reduce
mercury
emissions
(
0221).
Commenter
0221
adds
that
they
are
concerned
that
the
analysis
promoted
by
EPA
is
overly
compartmentalized
and
ignores
positive
consequences.
Commenter
28­
404
said
the
net
benefit
in
overall
pollution
reductions
should
be
compared
to
the
incremental
improvement
in
regional
haze
on
a
source­
by­
source
basis.

Commenter
28­
358
said
the
final
guidelines
should
explicitly
require
an
assessment
of
the
positive
non­
air
quality
environmental
impacts
 
i.
e.,
an
estimate
of
the
environmental
benefits
 
that
will
result
from
applying
BART.
It
is
possible
to
estimate
the
ecosystem
benefits
in
Class
I
areas
associated
with
certain
levels
of
lowered
emissions
of
sulfur
dioxide,
nitrogen
oxides,
and
particulate
matter
in
upwind
regions.
Such
an
estimate,
prepared
as
a
qualitative
assessment,
must
be
included
in
the
BART
determination
process.
140
The
guidelines
to
not
appear
to
have
any
criteria
outlines
for
consideration
of
non­
air
quality
environmental
impacts
and
one
commenter
(
0232,
28­
302)
proposed
that
language
be
revised
to
include
documentation
for
the
public
record
when
alternative
controls
are
used
due
to
extraordinarily
high
costs,
energy
or
non­
air
quality
environmental
impacts.
Commenter
28­
387
suggested
that
the
guidelines
specify
or
confirm
that
states
have
the
discretion
to
reasonably
consider
all
three
types
(
i.
e.,
cost
of
compliance,
energy
impacts,
and
non­
air
quality
impacts)
of
analysis.
Commenters
(
28­
290,
28­
302,
28­
7008)
said
EPA
should
clarify
the
role
and
weight
non­
air
quality
impacts
should
have
in
determining
BART
and
reasonable
further
progress.

Commenter
28­
352
said
the
impact
analysis
should
also
examine
other
air
impacts
beyond
the
precursors
to
regional
haze
(
i.
e.,
air
quality
impacts,
water
quality
impacts
etc.).
For
example,
some
NO
x
controls
might
increase
emissions
of
carbon
monoxide,
which
is
not
a
precursor
to
regional
haze.
This
could
represent
a
significant
adverse
impact
posed
by
the
application
of
the
control
that
would
not
be
addressed
under
the
current
guidance.

Eleven
commenters
(
0195,
0214,
0223,
0226,
0229,
0230,
0238,
0254,
0256,
0268,
0321)
oppose
EPA
requirement
that
non­
air
quality
impacts
include
environmental
benefits
when
determining
BART.
Nine
commenters
stated
that
the
requirement
that
non­
air
quality
impacts
to
include
benefits
to
the
environment
is
inconsistent
with
the
Act.
Section
§
169A(
g)(
2)
directs
States
to
consider
"
energy
and
non­
air
quality
environmental
impacts"
of
compliance
meaning
only
adverse
effects
may
be
considered
in
determining
BART
(
0214,
0223,
0226,
0229,
0230,
0238,
0254,
0256,
0321).
Further,
three
commenters
stated
that
EPA
should
be
consistent
with
Section
112
MACT
regulations
which
define
non­
air
impacts
as
adverse
impacts
(
0195,
0268,
0321).

Commenter
0256
included
extensive
remarks
stating
EPA
gives
no
explanation
to
why
states'
consideration
of
beneficial
non­
air
quality
impacts
of
control
technologies
is
consistent
with
Section
§
169A(
g)(
2).
They
state
that
this
is
not
consistent
with
Section
§
169A(
g)(
2)
for
the
same
reasons
that
EPA
gave
when
interpreting
Section
§
112(
d)(
2).
Section
§
112(
d)(
2)
provides
MACT
guidance
which
is
applicable
for
new
and
existing
sources
where
cost
has
been
taken
into
consideration
with
any
non­
air
quality
health
and
environmental
impacts
and
energy
requirements.
Further,
in
Sierra
v
EPA
(
D.
C.
Cir.
2004),
the
court
ruled
that
non­
air
quality
impacts
exclude
environmental
benefits
for
Section
§
112(
d)(
2)
from
emissions
reductions
rejecting
Sierra
Club's
position
that
impacts
of
"
deposition,
persistence,
toxicity
and
bioaccumulation
of
metal
HAP"
by
deposition
be
considered.
The
commenter
recommends
that
the
subsection
in
the
BART
rule
entitled
"(
5)
Benefits
to
the
Environment"
be
removed.
Consideration
of
environmental
benefits
would
increase
the
burden
on
States
when
making
BART
determination
especially
when
statutory
provision
is
focused
on
improvement
of
visibility.

Response:

We
agree
that
positive
environmental
impacts
should
be
considered,
as
well
as
impacts
to
other
media
such
as
water
quality
impacts.
The
guidelines
contains
a
discussion
to
that
effect.
141
We
believe
that
States
have
considerable
flexibility
for
how
to
weigh
these
factors.
We
do
not
believe
that
a
cost­
benefit
analysis
is
required
or
appropriate,
as
many
of
these
factors
are
not
easily
quantified
and
may
be
weighed
qualitatively.
In
addition,
we
do
not
believe
that
it
is
reasonable
to
assume
that
section
169A
necessarily
has
the
same
basis
as
section
112(
d)(
2)
of
the
CAA,
as
these
provisions
were
written
for
wholly
different
purposes.

Comment:
There
were
three
other
comments
regarding
the
consideration
of
non­
air
quality
environmental
benefits
in
determining
BART.
Commenter
0254
thinks
EPA
should
examine
the
characteristics
and
toxicity
of
PM
2.5
components
and
develop
regulatory
strategies
for
control
technologies
that
would
provide
the
greatest
health
benefits.
This
was
also
the
conclusion
of
a
NSA
National
Research
council
report
entitled
"
Research
Priorities
for
Airborne
Particulate
Matter,
IV:
Continuing
Research
Progress"
(
March
2004).
Focusing
only
on
two
constituents
of
PM
2.5,
nitrates
and
sulfates,
limits
health
benefits
the
agency
has
projected.
There
is
substantial
toxicological
evidence
that
SO
2
and
NO
X
emissions
from
EGUs
do
not
have
a
causal
association
with
health
effects.
The
cost­
benefit
analysis
in
CART
makes
a
blanket
assumption
that
all
the
constituents
on
PM
2.5
will
have
the
same
health
effects
as
nitrates
and
sulfates,
thus
excluding
many
Americans
who
are
exposed
to
PM
2.5
and
experience
health
effects.
EPA's
assumptions
are
also
inconsistent
with
EPRI's
and
Emory
University's
Atmospheric
Research
Inhalation
Epidemiology
Study
in
which
no
statistically
significant
association
between
sulfate
levels
and
health
effects
was
found.
EPA's
health
benefits
of
the
proposed
rule
are
over
stated
and
BART
should
contain
a
high
degree
of
uncertainty
regarding
health
benefits
of
control
technology.

Commenter
0276
indicates
that
controls
implemented
by
BART
will
have
much
smaller
health
benefits
than
those
derived
from
controls
put
in
place
before
BART.
They
recommend
dropping
this
entire
section,
or
structure
the
section
to
focus
on
the
incremental
health
benefits.

Commenter
28­
352
noted
that
stream
water
sampling
in
the
Great
Gulf
Wilderness
at
high
elevations
found
some
of
the
highest
nitrate
levels
in
the
region.

Comment:
Regarding
non­
air
quality
environmental
impacts
the
commenter
indicated
that
EPA's
statement
that
a
state
does
not
have
to
perform
an
analysis
if
the
most
stringent
controls
are
adopted
is
incorrect.
States
must
assess
all
five
statutory
BART
determination
factors
for
each
potential
control
regardless
of
environmental
impacts.
Comparing
the
environmental
impacts
of
two
control
technologies,
one
more
stringent
than
the
other,
may
result
in
the
more
stringent
control
producing
greater
environmental
harm.
This
comparison
is
important
given
the
location
of
a
facility
(
0256).
142
Response:

We
believe
that
this
provision
reasonably
balances
the
goals
of
section
169A
with
the
analytic
burden
it
imposes
on
States.

Comment:
Regarding
evaluating
impacts
and
reporting
results
commenter
0304
states
modeling
impacts
should
not
be
considered
to
moderate
control
technology
when
determining
BART
because
they
are
not
a
consideration
when
determining
BACT.

Response:

Assessment
of
impacts
on
regional
haze
is
specifically
required
by
the
CAA
for
BART.

4.1.3.5
Incompatibility
with
a
Control
for
Another
Pollutant
Comment:
Five
commenters
addressed
incompatibility
issues
with
controlling
some
pollutants
(
28­
266,
28­
278,
23­
309,
28­
312,
28­
332).
Commenters
(
28­
266,
28­
278,
28­
312)
said
incompatibility
with
a
control
for
another
pollutant,
such
as
mercury,
should
be
a
criterion
for
rejecting
(
or
modifying)
a
BART
control
option.
Commenter
28­
266
added
that
EPA
must
provide
guidance
on
how
to
select
the
best
alternative
if
a
source
must
control
several
pollutants
but
the
available
technologies
have
overlapping
or
conflicting
requirements.
Commenter
28­
332
said
language
should
be
added
concerning
how
increases
in
visibility
impairing
pollutants
from
control
technologies
should
be
included
in
the
BART
engineering
analysis.

Commenter
28­
309
said
when
a
particular
pollutant
is
reduced,
consideration
should
be
given
to
the
increases
which
may
occur
in
other
pollutants.
For
example,
if
utilities
install
SCRs
downstream
rather
than
upstream
from
ESPs
and
wet
scrubbers,
this
might
result
in
higher
levels
of
ammonia
or
sulfur
trioxide
emissions
from
the
stack.
If
the
ESP
and
wet
scrubber
is
located
after
the
SCR
with
high
efficiency
mist
eliminators
after
the
scrubber
and
liquid
collectors),
a
portion
of
the
ammonia
and
converted
sulfur
trioxide
will
be
collected
on
the
ash.
Carbon
monoxide,
and
possibly
benzo[
a]
pyrene,
is
increased
with
the
use
of
some
low
NO
x
techniques.

Response:

We
note
that
the
guidelines
addresses
this
possibility
and
suggests
that
when
the
result
of
a
BART
determination
would
be
two
different
BART
technologies
that
do
not
work
well
together,
States
have
flexibility
to
substitute
a
different
technology
or
combination
of
technologies,
provided
that
they
achieve
at
least
the
same
emissions
reductions
for
each
pollutant.

4.1.4
Evaluate
Visibility
Impacts
of
Applying
Controls
143
4.1.4.1
Pollutant­
by­
Pollutant
Basis
Comment:
Seventeen
commenters
addressed
the
concept
of
ranking
reductions
on
a
one
pollutant
versus
a
combination
of
pollutants
(
0169,
0228,
28­
243,
28­
255,
28­
256,
28­
262,
28­
267,
28­
272,
28­
273,
28­
286,
28­
302,
28­
332,
28­
335,
28­
350,
28­
362,
28­
779,
28­
7048).

Commenters
(
28­
255,
28­
256,
28­
302,
28­
332,
28­
335,
28­
350)
said
the
guidelines
should
address
the
process
for
ranking
control
technologies
that
reduce
more
than
one
visibility
impairing
pollutant.
If
the
total
visibility
impairing
pollutant
emission
reduction
is
greater
than
a
technology
reducing
emissions
of
only
one
pollutant,
it
should
be
ranked
higher
and
given
preference.
Commenters
(
28­
302,
28­
362)
added,
however,
that
considerations
of
simultaneous
pollutant
reductions
should
not,
justify
reduced
stringency
of
controls
for
the
pollutant
which
is
specifically
targeted
under
the
BART
review.
Commenter
28­
779
said
control
technology
ranking
should
account
for
the
relative
importance
of
the
pollutants
in
reducing
visibility
impairment.
Commenter
28­
272
said
there
should
be
flexibility
to
account
for
the
different
roles
pollutants
have
in
different
regions
of
the
country.
Commenter
28­
243
said
the
options
selection
process
would
result
in
choosing
options
that
produce
a
multiple
benefit
for
the
same
dollar
as
options
that
have
more
limited
benefits.

Response:

The
States
are
not
required
to
rank
the
control
technologies
and
have
discretion
in
weighing
the
five
factors
for
control
decisions.

Commenters
(
28­
262,
28­
7048)
suggested
two
analytical
approaches:
(
1)
Calculate
an
annualized
pollution
emitted
for
each
pollutant
for
each
technology
alternative,
creating
a
matrix,
and
rank
pollutant
one,
then
pollutant
two,
then
pollutant
three,
sum
and
rank.
(
2)
Convert
the
emissions
to
extinction
values
and
sum
the
extinction
and
rank.
This
would
rank
by
weighting
the
visibility
impacts.
This
would
be
a
scattering
total
that
could
be
used
for
ranking
multiple
pollutants.
It
would
not
replace
an
analysis
of
the
impacts
on
visibility
in
Class
I
areas.

Response:

The
States
are
not
required
to
rank
the
control
technologies
and
have
discretion
in
weighing
the
five
factors
for
control
decisions.

Commenter
28­
286
said
technology
that
provides
lower
emissions
should
be
adopted.
States
should
strongly
consider
the
amount
of
visibility
improvement
that
would
result
when
determining
which
technology
is
the
BACT.

Response:
144
Lower
emissions
is
not
one
of
the
five
factors
that
are
statutorily
required
as
part
of
the
BART
review.
However,
when
making
the
control
determination,
emissions
from
the
source
would
be
lower
if
controls
are
put
in
place.

One
commenter
requests
EPA
specify
that
the
visibility
analysis
be
performed
per
pollutant
(
0228).
Another
commenter
(
28­
273)
would
limit
the
consideration
of
multiple
benefits
to
tie
breaker
situations.
Commenter
28­
362
said
the
identification
of
a
control
technique
that
will
be
effective
in
addressing
multiple
pollutants
should
not
serve
as
a
substitute
for
a
pollutant­
bypollutant
BART
analysis.
Likewise,
a
control
technique
that
reduces
multiple
pollutants
should
not
be
considered
BART
for
any
individual
pollutant
if
the
BART
analysis
for
that
pollutant
identifies
a
technically
feasible
control
technology
that
is
more
effective
in
reducing
emissions
of
that
pollutant.

Response:

States
have
the
discretion
in
making
the
BART
determination
and
are
encouraged
to
make
the
most
effective
assessment
based
on
the
five
statutory
factors.
A
pollutant
by
pollutant
basis
or
a
multi­
pollutant
approach
is
acceptable.

Other
commenters
(
0169,
28­
267)
said
the
visibility
impacts
analysis
should
be
resolved
by
the
state's
discretion
in
adopting
alternative
controls
for
each
source
when
determining
BART.

Response:
We
agree
and
have
provided
discretion
in
the
final
rule.

Comment:
XX
commenters
said
the
guidelines
should
provide
that
cost
effectiveness
is
to
be
measured
on
the
basis
of
a
source's
impact
on
visibility
impairment,
not
reductions
in
the
amount
of
pollutants
emitted
(
28268,
28­
337,
28­
349,
28­
353).

Response:

States
have
discretion
in
how
they
evaluate
all
of
the
five
factors,
including
cost
effectiveness,
when
making
the
BART
determination.
One
possible
alternative
in
determining
cost
effectiveness
is
to
make
an
acceptable
demonstration
based
on
a
visibility
assessment.

Comment:
Three
commenters
said
more
guidance
is
needed
on
the
role
of
visibility
impacts
(
28­
268,
28­
364,
28­
395).
Commenter
28­
268
said
EPA
should
publish
guidelines
that
index
each
pollutant's
average
impact
on
visibility,
to
allow
aggregation
of
both
costs
and
benefits
across
all
five
pollutants.
Commenters
(
28­
364,
28­
395)
said
the
Act
does
not
suggest
that
certain
of
the
BACT
criteria
­
energy,
environmental
and
economic
impacts
and
other
costs
­
should
be
emphasized
over
others.
12CALPUFF
Analysis
in
Support
of
the
June
2005
Changes
to
the
Regional
Haze
Rule,
U.
S.
Environmental
Protection
Agency,
June
15,
2005,
Docket
No.
OAR­
2002­
0076.

145
Response:

States
have
discretion
to
decide
how
to
weigh
each
of
the
five
statutory
factors,
including
the
choice
to
whether
or
not
to
assess
pollutants
in
aggregate
or
separately,
when
making
a
BART
determination,
and
one
factor
need
not
be
emphasized
over
another.
Our
sample
modeling
runs12
may
help
inform
the
determination
of
pollutant
impacts
on
particular
Class
I
areas,
but
since
each
Class
I
area
is
unique,
average
impacts
are
difficult
to
determine.
We
do
not
believe
that
a
generalization
of
impacts
and
costs
would
provide
realistic
estimates.

4.1.4.2
Use
of
CALPUFF
or
EPA­
Approved
Model
Comment:
Eight
comments
were
received
on
evaluating
visibility
impacts
of
applying
controls
using
CALPUFF
or
another
EPA­
approved
model
(
0207,
0217,
0226,
0242,
0246,
0256,
0275,
0277).
Some
commenters
indicated
that
the
States
should
have
flexibility
to
use
models
other
than
CALPUFF.
Additionally,
the
guidelines
seem
to
mandate
the
use
of
CALPUFF
while
saying
States
can
use
other
approved
models
as
well,
making
them
unclear
(
0207,
0217,
0226,
0242,
0246,
0275,
0277).
Specifically
there
need
to
be
clarification
on
which
modeling
systems
are
allowed
for
evaluating
visibility
impacts
(
0217,
0246).
One
commenter
stated
that
CALPUFF
is
not
a
preferred
model
for
determining
visibility
impairment
(
0246).
Another
commenter
indicated
that
the
screening
tool
capabilities
of
CALPUFF
are
acceptable,
however
source
owner
and
operators
should
have
the
discretion
to
use
all
the
capabilities
of
CALPUFF
where
they
deem
necessary
(
0256).

Response:
Other
EPA­
approved
models
may
be
used
and
an
EPA­
approved
modeling
protocol,
which
allows
for
determining
the
model
inputs
and
model
selection
should
be
submitted
to
the
State
prior
to
any
modeling
demonstration.
CALPUFF
used
with
limited
meteorological
data
(
i.
e.
in
a
simpler
screening
mode)
may
not
be
as
conservative
as
CALPUFF
using
a
full
meteorological
domain,
and
should
not
be
used
to
exempt
a
source
from
BART
review
without
more
rigorous
evaluations
and
sensitivity
tests
showing
that
the
results
are
at
least
as
conservative
as
the
CALPUFF
model.
The
State
or
the
source
may
perform
the
modeling
for
BART
assesssments.
Although
we
focused
on
the
use
of
CALPUFF
for
primary
pollutants
in
revising
the
Guideline
of
Air
Quality
Modeling,
section
7.2.1.
e.
of
the
Guideline
states:
e.
CALPUFF
(
Section
A.
3)
may
be
applied
when
assessment
is
needed
of
reasonably
attributable
haze
impairment
or
atmospheric
deposition
due
to
one
or
a
small
group
of
sources.
This
situation
may
involve
more
sources
and
larger
modeling
domains
than
that
to
which
VISCREEN
ideally
may
be
applied.
The
procedures
and
analyses
should
be
determined
in
consultation
with
the
appropriate
reviewing
authority
(
paragraph
3.0(
b)
and
the
affected
FLM(
s).
146
We
believe
that
our
proposed
use
of
CALPUFF
is
thus
fully
in
keeping
with
the
Guideline
on
Air
Quality
Models,
especially
in
light
of
the
low
triggering
threshold
for
determining
whether
a
source
is
reasonably
anticipated
to
cause
or
contribute
to
visibility
impairment
in
a
Class
I
area,
and
the
fact
that
the
modeling
results
are
used
as
only
one
of
five
statutory
criteria
evaluated
to
determine
BART
emission
limits.

4.1.4.3
Use
of
Allowable
Emissions
Comment:
Five
comments
were
received
on
evaluating
visibility
impacts
of
applying
controls
using
allowable
emissions.
Three
commenters
state
that
EPA
must
make
clear
that
States
should
use
emission
levels
that
will
be
permissible
at
the
time
BART
controls
take
effect,
and
not
use
current
emissions
levels
(
0189,
0238,
0275).
One
commenter
supports
the
use
of
allowable
emissions
(
0256).
One
commenter
indicates
that
actual
emissions,
rather
than
allowable
emissions,
should
be
considered
(
0268).

Response:
For
BART
applicability,
an
emission
rate
that
applies
at
the
time
the
State
submits
the
SIP
should
be
used.
Emissions
from
a
source
can
vary
widely
on
a
day
to
day
basis;
during
peak
operating
days,
the
24­
hour
actual
emission
rate
could
be
more
than
double
the
daily
average.
On
the
other
hand,
in
the
long
term,
estimating
visibility
impacts
based
on
allowable
emission
rates
for
every
hour
of
the
year
may
unduly
inflate
the
maximum
impairment
from
a
BART­
eligible
source.
Where
States
have
information
on
a
source's
daily
emissions,
an
emission
rate
based
on
the
maximum
actual
emissions
over
a
24­
hour
period
for
the
most
recent
five
years
is
a
more
appropriate
gauge
of
a
source's
potential
impact
as
it
would
ensure
that
peak
emission
conditions
are
reflected,
but
would
likely
not
overestimate
a
source's
potential
impact
on
any
given
day.
We
have
accordingly
included
this
change
to
the
final
guidelines.
We
recommend
that
the
State
use
the
highest
24­
hour
average
actual
emission
rate,
for
the
three
or
five
year
period
of
meteorological
data,
to
characterize
the
maximum
impact
for
the
BART
applicability
test.

4.1.4.4
Use
of
Results
of
the
20%
Worst
Modeled
Days
Comment:
Nine
comments
were
received
supporting
evaluating
visibility
impacts
of
applying
controls
using
the
results
of
the
20%
worst
modeled
days
(
0205,
0226,
0228,
0240,
0250,
0254,
0256,
0276,
0314).
Several
commenters
(
0228,
0240,
0250,
0254,
0256)
stated
that
the
use
of
using
a
maximum
24­
hr
emissions
over
5
years
to
evaluate
the
thresholds
is
unreasonable
and
inappropriate
and
departs
from
the
previous
methodologies
for
the
Regional
Haze
Program.
Additionally
this
threshold
is
restrictive
because
the
highest
24­
hr
impact
over
a
three
or
five
year
period
may
be
influenced
by
short­
term
weather
conditions,
like
high
humidity
(
0240).

Response:
See
the
discussion
above
regarding
emission
rates.
Also,
the
final
rule
provides
several
options
for
making
impact
determinations
for
purposes
of
evaluating
visibility
improvement.
We
have
pointed
out,
however,
that
evaluating
the
fifth
statutory
factor
is
not
necessarily
based
on
a
13CALPUFF
Analysis
in
Support
of
the
Regional
Haze
Rule,
U.
S.
Environmental
Protection
Agency,
April
15,
2005,
Docket
No.
OAR­
2002­
0076.

14
Ibid.

147
one­
time
test
for
a
maximum
24­
hour
impact.
A
State
is
encouraged
to
account
for
the
magnitude,
frequency,
and
duration
of
the
contributions
to
visibility
impairment
caused
by
the
source
based
on
the
natural
variability
of
meteorology.
These
are
important
elements
to
consider
as
they
would
provide
useful
information
on
both
the
short
term
peak
impact
and
long
term
average
assessments
which
are
critical
in
making
the
visibility
assessment.
Additionally,
the
use
of
a
comparison
threshold,
as
is
done
for
determining
if
BARTeligible
sources
should
be
subject
to
a
BART
determination,
may
be
an
appropriate
way
to
evaluate
visibility
improvement.
However,
we
believe
the
States
should
have
flexibility
in
setting
absolute
thresholds,
target
levels
of
improvement,
or
de
minimis
levels
since
the
deciview
improvement
must
be
weighed
among
the
five
factors,
and
States
are
free
to
determine
the
weight
and
significance
to
be
assigned
to
each
factor.
For
example,
a
0.3
deciview
improvement
may
merit
a
stronger
weighting
in
one
case
versus
another,
so
one
"
bright
line"
may
not
be
appropriate.
In
addition,
comparison
thresholds
can
be
used
in
a
number
of
ways
in
evaluating
visibility
improvement
(
e.
g.
the
number
of
days
or
hours
that
the
threshold
was
exceeded,
a
single
threshold
for
determining
whether
a
change
in
impacts
is
significant,
a
threshold
representing
an
x
percent
change
in
improvement,
etc.).
In
our
example
modeling
analysis
of
a
hypothetical
source13,
we
used
three
different
24­
hour
thresholds
(
1.0,
0.5,
and
0.1
deciviews)
and
examined
the
number
of
days
that
those
thresholds
were
exceeded
for
a
source
with
a
90
percent
change
in
SO2
emissions
(
i.
e.
10,000
TPY
and
1,000
TPY).
The
number
of
days
that
the
thresholds
were
exceeded
in
the
10,000
TPY
case
was
substantial,
and
the
visibility
improvement
due
to
the
reduction
in
emissions
was
dramatic
(
i.
e.
the
number
of
days
exceeding
the
thresholds
dropped
considerably)
14.
Other
ways
that
visibility
improvement
may
be
assessed
to
inform
the
control
decisions
would
be
to
examine
distributions
of
the
daily
impacts,
determine
if
the
time
of
year
is
important
(
e.
g.
high
impacts
are
occurring
during
tourist
season),
consideration
of
the
cost­
effectiveness
of
visibility
improvements
(
i.
e.
the
cost
per
change
in
deciview),
using
the
measures
of
deciview
improvement
identified
by
the
State,
or
simply
compare
the
worst
case
days
for
the
pre­
and
postcontrol
runs.
States
may
develop
other
methods
as
well.

Comment:
Six
comments
were
received
opposing
evaluating
visibility
impacts
of
applying
controls
using
the
results
of
the
20%
worst
modeled
days
(
0198,
0225,
0242,
0249,
0251,
0319).
Two
commenters
added
that
the
criteria
should
use
the
20%
worst
day
based
on
monitored
data,
and
not
modeled
data
(
0198,
0242).
One
commenter
stated
that
by
using
the
20%
worst
modeled
days,
natural
source
contributions
are
not
taken
into
consideration
(
0225).
Another
commenter
stated
that
sources
should
not
be
limited
to
just
the
worst
days,
but
the
improvements
should
be
based
upon
controls
reducing
visibility
impairments
on
any
day
(
0249).
Two
commenters
(
0249,
0251)
stated
that
the
requirement
that
improvement
be
made
on
the
20%
worst
days
and
no
148
degradation
on
the
20%
best
day
does
not
constitute
a
criterion
for
measuring
progress.
This
rationale
ignores
the
middle
60%
of
days
in
which
visibility
can
worsen
because
sources
may
be
allowed
to
increase
emissions
on
these
day
as
a
trade­
off
for
cutting
emissions
on
the
worst
days
(
0251).
Commenter
0251
went
on
to
say
that
there
is
no
data
or
evidence
to
support
EPA's
assertion
that
progress
on
the
worst
days
means
progress
on
other
days.
One
commenter
(
0319)
also
opposes
the
use
of
the
maximum
modeled
24­
hr
change
over
a
5
year
period
as
a
criteria
for
considering
BART.

Response:
See
the
discussion
above
for
evaluating
visibility
impacts
for
control
options.
We
encourage
States
to
account
for
the
magnitude,
frequency,
and
duration
of
the
visibility
impacts.

Comment:
Two
other
comments
were
received
regarding
evaluating
visibility
impacts
of
applying
controls
using
the
results
of
the
20%
worst
modeled
days
(
0249,
0256)
Although
the
rules
require
a
state
to
submit
visibility
SIPs
for
the
10­
year
period
of
the
SIP
specifying
reasonable
progress
goals,
the
rules
do
not
contain
a
requirement
that
states
determine
the
rate
of
improvement
need
to
achieve
natural
conditions
in
60
years.
In
fact
a
state
can
reject
the
60
year
schedule
but
demonstrating
that
the
schedule
is
unreasonable,
in
the
state's
view
(
0249).
Additionally,
EPA
should
clarify
that
the
20%
worst
modeled
days
are
the
observed
20%
worst
day
as
determined
from
IMPROVE
(
0256).

Response:
See
the
discussion
above
for
evaluating
visibility
impacts
for
control
options.
States
have
the
option
in
how
to
evaluate
visibility
for
the
fifth
factor
assessment,
but
as
a
general
matter,
evaluations
should
be
made
using
the
worst
modeled
as
the
IMPROVE
receptor
may
not
represent
the
maximum
impairment
across
the
entire
Class
I
area.

4.1.4.5
Use
of
Hourly
Modeled
Impacts
Comment:
Eight
commenters
opposed
using
the
maximum
24­
hour
value
in
evaluating
the
"
cause
or
contribute
threshold
(
0188,
0205,
0214,
0226,
0238,
0253,
0321,
0323).
Commenters
said
that
using
a
single
day
in
a
5­
year
period
represents
a
single
spike
that
may
be
unduly
influenced
by
short­
term
abnormal
conditions.
Commenter
0205
added
that
the
proposed
analysis
is
so
stringent
that
few
if
any
sources
will
be
able
to
demonstrate
no
contribution
and
EPA
has
circumvented
the
court
ruling
in
American
Corn
Growers
by
essentially
requiring
all
BART­
eligible
sources
to
install
BART
controls
without
a
meaningful
showing
of
real
contribution
to
visibility
impacts
in
a
Class
I
area.
Given
EPA's
recognition
in
other
recent
standards
of
the
need
for
a
more
stable
statistical
form,
and
the
earlier
rule's
reliance
on
the
average
20%
worst
modeled
days,
PCA
requests
that
EPA
use
the
average
of
the
20%
worst­
modeled
days
as
in
the
1999
RHR
to
determine
if
a
source
reasonably
contributes
to
regional
haze.
Following
the
analysis
provided
in
the
1999
RHR
best
meets
the
1999
RHR's
original
intent.
149
Response:
The
final
rules
allows
a
State
to
use
the
98th
percentile
impact
over
the
meteorological
period
modeled,
whichever
they
deem
is
most
appropriate.
Such
an
approach
affords
added
protection
against
the
possibility
that
poor
visibility
might
be
caused
by
unusual
meteorology
and
thus,
in
our
view,
represents
a
reasonable
lower­
bound
for
comparison.
If
the
98th
percentile
value
from
your
modeling
is
less
than
your
threshold,
then
you
may
conclude
that
the
source
does
not
contribute
to
visibility
impairment
and
is
not
subject
to
BART.

Commenters
0214
and
0238
said
EPA's
requirement
to
use
the
maximum
24­
hour
value
over
the
5­
year
period
of
meteorological
data
in
the
modeling,
as
proposed,
is
too
stringent.
In
contrast,
EPA
uses
an
"
average
of
the
20%
worst
days"
metric
in
the
"
visibility
improvement"
step
in
the
BART
determination
evaluation
and
in
the
regional
haze
program.
The
default
"
natural
condition"
deciview
values
for
Class
I
areas
in
EPA's
natural
conditions
guidance,
which
EPA
recommends
states
use
for
the
visibility
impairment
analysis,
exist
only
for
the
average
of
the
20%
best
and
worst
days.
There
are
no
default
"
maximum
24­
hour"
values
in
EPA's
guidance.
The
accuracy
of
these
measurements
will
be
very
dubious
for
some
cooperative
owned
or
operated
plants
that
do
not
even
have
meteorological
data
that
is
site­
specific,
and
they
are
hundreds
of
miles
from
the
nearest
airport,
which
is
the
default
value
for
such
data.

Response:
See
the
discussion
above
concerning
which
values
may
be
used.

Commenter
0323
responded
to
EPA's
request
for
comment
on
the
use
of
hourly
modeled
impacts
from
CALPUFF
to
assess
the
number
of
hours
above
a
visibility
threshold.
While
hourly
model
data
are,
by
their
nature,
less
reliable
in
predicting
actual
conditions,
a
measure
that
reports
the
total
number
of
hours
above
a
given
level
would
still
be
a
useful
measure
of
the
long­
term
effect
of
BART
control.
EPA
should
require
states
to
report
a
combination
of
measures
regarding
the
improvement
in
visibility
expected
from
implementing
BART.
This
combination
should
include
the
EPA's
proposed
metric
of
the
change
in
the
20
percent
worst
days
and
a
metric
that
examines
the
amount
of
time
during
a
year
the
source's
visibility
impact
would
exceed
a
threshold
with
and
without
implementation
of
BART
control.

Response:
See
the
discussion
above
concerning
which
values
may
be
used.

Comment:
Two
commenters
suggested
alternatives
to
the
24­
hour
value
(
0205,
0256).

Commenter
0205
said
a
threshold
comparison
is
more
appropriate
than
the
overall
comparison
of
the
20%
worst
case
days.
However,
the
threshold
comparison
should
be
on
at
least
a
daily
average
(
or
longer),
not
an
hourly
average
due
to
the
possibility
of
short­
term
spikes
based
on
certain
meteorological
conditions.
A
comparison
of
number
of
days
above
or
below
a
certain
threshold
is
preferable
since
below
a
certain
threshold,
the
impacts
of
visibility
are
not
perceptible.
Unlike
concentration
levels
of
certain
pollutants
(
i.
e.,
ozone)
which
do
not
have
a
threshold
below
which
there
are
no
effects,
there
are
pollutant
levels
below
which
there
is
no
visibility
impact.
The
comparison
of
number
of
days
will
also
allow
for
a
more
complete
picture
of
how
the
control
150
device
will
impact
visibility.
As
noted
previously,
a
small
number
of
unusual
meteorological
conditions
can
produce
significant
spikes
on
a
single
day
or
days.
Since
the
overall
goal
of
the
regional
haze
rule
is
long­
term
visibility
improvement,
a
comparison
of
the
total
number
of
days
exceeding
a
threshold
over
multiple
years
will
provide
a
better
overall
indicator
of
visibility
improvement.

Response:
See
the
discussion
above
concerning
which
values
may
be
used.

Commenter
0256
said
if
EPA
retains
the
maximum
24­
hour
value
for
the
visibility
impairment
analysis,
EPA
should
at
least
allow
the
use
of
only
1
year,
rather
than
5
years,
of
meteorological
data.
That
would
simplify
the
modeling
and
would
lessen
the
chance
that
one
day
with
atypical,
extreme
conditions
would
dictate
the
result.
Moreover,
if
the
24­
hour
approach
is
used,
EPA
should
clarify
that
block
averaging
is
to
be
used
to
determine
24­
hour
values.

Response:
See
the
discussion
above
concerning
which
values
may
be
used.
A
State
has
discretion
to
use
block
averaging
if
they
so
choose.
Three
or
five
years
of
meteorological
data
may
be
used
per
40
CFR
51,
Appendix
W,
April
2003.

4.1.4.6
Levels
of
Visibility
Improvement
Comment:
Nine
commenters
commented
on
the
utility
and
implementation
of
visibility
improvement
thresholds
(
0189,
0228,
0230,
0238,
0253,
0256,
0271,
0275,
0306,
28­
7048).
Commenter
0228
said
EPA
should
clarify
that
visibility
improvement
differences
among
BART
control
options
should
be
considered
insignificant
if
the
differences
differ
by
less
than
the
perceptibility
threshold
level,
which
should
be
set
in
excess
of
1
deciview
and
not
the
0.5
deciview
level
EPA
is
proposing.
Other
commenters
said
the
minimum
threshold
should
be
1
deciview
(
0238,
0256).

Response:
See
the
discussion
above
regarding
how
to
evaluate
visibility
improvement
for
the
BART
determination.

Commenters
(
0230,
0275)
said
the
Proposed
BART
Rule
proposes
a
visibility
threshold
to
determine
when
a
source
causes
or
contributes
to
visibility
impairment
at
a
Class
I
area;
however,
no
threshold
is
proposed
to
determine
whether
BART
controls
on
a
specific
BART­
eligible
source
results
in
a
visibility
improvement.
Commenter
0306
said
it
is
unclear
how
the
modeled
net
visibility
improvement
would
be
specifically
utilized
in
the
BART
analysis.
Is
there
a
target
level
of
improvement?
Is
there
a
de
minimis
level?

Response:
As
discussed
above,
States
have
discretion
in
how
to
evaluate
visibility
improvement
and
how
to
weigh
that
improvement
with
the
other
four
factors
to
make
the
BART
control
decision.
We
are
not
dictating
a
target
or
de
minimis
level
as
visibility
improvement
may
be
weighed
differently
for
each
Class
I
area.
151
Commenter
0230
added
that,
if
the
BART
controls
result
in
no
perceptible
improvement
in
visibility
at
a
Class
I
area,
then
it
is
not
a
cost­
effective
strategy
for
improving
visibility
at
Class
I
areas
because
it
will
have
no
perceived
impact.
Based
on
Pitchford
and
Malm
(
1994)
and
Henry
(
2002)
a
2
deciview
threshold
of
perception
would
be
appropriate,
with
a
1
deciview
threshold
providing
a
margin
of
safety.
Commenter
0253
agreed
that
EPA
should
revise
its
methodology
to
screen
out
predicted
visibility
improvements
that
are
irrelevant
to
determining
the
level
of
control
that
constitutes
BART,
provide
for
calculation
of
visibility
improvement
associated
with
each
potential
level
of
BART
control,
and
measure
visibility
improvement
based
on
the
overall
improvement
in
general
visibility.

Response:
Perceptibility
may
not
be
a
factor
in
determining
controls
at
an
individual
source.
For
example,
if
there
are
multiple
sources
contributing
to
impairment
at
a
Class
I
area,
the
collective
visibility
benefit
of
controlling
each
of
these
sources
could
be
perceptible
(
assuming
that
a
control
level
is
set
based
on
weighing
the
five
factors,
including
cost,
for
each
BART­
eligible
source),
in
spite
of
the
fact
that
the
individual
improvement
expected
from
each
source
by
itself
may
not
be
perceptible.
Our
hypothetical
modeling
analyses
(
provided
in
a
technical
memo
to
the
final
rule)
may
inform
estimates
of
visibility
improvement
associated
with
a
potential
control
efficiency.

Commenter
0271
noted
EPA
is
proposing
to
add
a
fifth
criteria
for
inclusion
in
the
BART
engineering
analyses
that
must
be
conducted
on
a
source­
by­
source
basis;
namely
the
degree
of
visibility
improvement
that
would
result
from
the
addition
of
one
control
option
over
another.
This
analysis
is
unnecessary,
confusing
and
without
adequate
standards
or
guidance
for
implementation.
In
general,
the
analysis
of
the
level
of
controls
in
the
BART
process
should
yield
the
greatest,
most
cost­
effective
control
efficiency
for
NO
x
and
SO
2
at
or
above
the
presumptive
level
of
control
for
these
air
contaminants.
An
analysis
of
the
degree
of
visibility
improvement
within
this
range
may
result
in
very
small
increments
of
visibility
improvements
within
Class
I
areas
from
a
single
source,
making
consideration
of
this
very
small
increment
meaningless
or
tilted
to
the
selection
of
the
lower
control
efficiency
option.
EPA
should
remove
this
criterion
from
the
analysis
to
ensure
that
the
highest,
most
cost
effective
controls
will
result.

Response:

The
BART
guidelines
recommend
a
way
for
States
to
take
into
account
the
degree
of
visibility
improvement
that
may
result
from
the
use
of
BART.
We
agree
that
an
analysis
of
the
degree
of
visibility
improvement
resulting
from
the
use
of
one
technology,
as
compared
with
another,
may
show
very
small
increments
of
visibility
improvement.
The
States
have
the
discretion
to
consider
the
degree
of
visibility
improvement
in
a
reasonable
manner.
As
the
States
will
be
making
BART
determinations
at
the
same
time
that
they
are
establishing
reasonable
progress
goals,
they
should
consider
this.

Commenter
0189
said,
as
the
Corn
Growers
case
emphasized,
the
statute
clearly
provides
for
states
setting
BART
to
balance
the
visibility
benefits
of
BART
controls
comprehensively
against
their
burdens.
There
is
not
a
word
of
that
in
EPA's
proposal.
Although
it
would
allow
152
States
setting
BART
to
run
CALPUFF
modeling,
it
completely
fails
to
specify
how
they
might
consider
the
results.

Response:

The
court
in
American
Corn
Growers'
invalidated
EPA's
1999
provisions
on
BART
determinations
because
they
bifurcated
visibility
from
the
other
BART
factors
and
precluded
States
from
weighing
visibility
impacts
on
a
source
specific
basis.
EPA
has
remedied
that
defect
in
this
rulemaking.
Nothing
in
the
case
suggested
that
EPA
must
dictate
the
manner
in
which
States
weigh
visibility
against
other
factors.
Consistent
with
the
court's
emphasis
on
the
discretion
afforded
to
States
by
CAA
169A,
for
sources
other
than
750
MW
power
plants,
States
have
considerable
latitude
under
the
guidelines
to
determine
how
this
weighing
should
be
done.

Comment:
Commenters
(
28­
266,
28­
312,
28­
352)
said
EPA
states
that,
for
any
Class
I
area,
the
cumulative
visibility
improvement
is
a
"
substantial
fraction"
­
how
much
is
a
"
substantial
fraction"?
Note
this
criteria
is
confusing
in
light
of
the
statement
in
the
preamble
that
EPA
"
calculated
relatively
small
impacts
for
BART,
in
part
because
the
baseline
for
the
analysis
assumed
a
substantial
degree
of
emissions
for
BART­
eligible
sources
in
response
to
the
NAAQS
for
PM2.5".
This
might
suggest
that
it
may
be
difficult
for
states
to
show
that
BART
emission
reductions
will
provide
"
a
sufficient
visibility
improvement
to
justify
their
installation".
Commenter
28­
352
said
the
BART
guidelines
should
require
that
"
any"
prevention
of
future
deterioration
or
existing
impairment
in
visibility
associated
with
a
Class
I
area
be
sufficient
basis
for
control
for
all
eligible
sources
in
the
region
under
the
cumulative
visibility
assessment.

Response:

In
this
final
rule,
visibility
improvement
as
a
result
of
controls
is
determined
on
a
source
by
source
basis
(
per
the
American
Corn
Growers
court
decision).
An
analysis
of
all
five
statutory
factors
is
required
when
assessing
controls
for
BART
and
visibility
alone
can
not
be
sufficient
in
determining
controls.

Commenter
28­
266
did
not
agree
that
application
of
BART
must
represent
a
"
substantial
fraction"
of
the
achievable
visibility
improvement
from
all
measures
in
the
SIP.
As
long
as
such
emission
reductions
will
improve
visibility
in
a
Class
I
area,
then
they
are
justified
and
must
be
required.

Response:

An
analysis
of
all
five
statutory
factors
is
required
when
assessing
controls
for
BART
and
visibility
improvement
(
assumed
to
result
from
emission
reductions)
alone
can
not
be
sufficient
in
determining
controls.
153
Commenter
28­
7048
said
depending
on
the
metric
used
for
reasonable
progress,
the
early
emission
reductions
from
BART
may
be
more
important
and
economically
valuable
than
the
guidance
will
estimate.
In
particular,
the
deciView
(
dV)
is
a
logarithmic
function.
Linear
progress
in
dV
in
the
long­
term
requires
a
more
rapid
reduction
in
emissions
in
the
short­
term.
The
guidance
does
not
mention
this
and
it
isn't
considered
in
the
establishment
of
the
budget
or
in
the
analysis
of
whether
the
BART
reductions
are
needed.

Response:

BART
must
be
installed
as
expeditiously
as
possible.
States
can
take
into
account
the
timing
of
BART
reductions
in
establishing
reasonable
progress
goal.

4.1.4.7
Compare
to
Natural
Conditions
Comment:
Three
commenters
addressed
the
role
a
natural
conditions
baseline
should
play
in
the
analysis
(
0189,
0225,
0268).
Commenters
(
0189,
0268)
stressed
that
in
assessing
the
visibility
benefits
of
BART
controls,
states
should
assess
their
impact
on
the
visibility
levels
projected
to
exist
during
their
operating
life,
not
on
pristine
visibility
levels
that
do
not
currently
exist
and
are
not
projected
to
be
achieved
until
2064
under
EPA's
visibility
rules.
Commenter
0225
was
concerned
that
the
degree
of
visibility
improvements
evaluated
from
additional
controls
should
include
the
evaluation
of
the
quality
of
visibility
that
natural
conditions
represent.
Sources
impacting
Class
I
areas
where
natural
sources
are
the
main
cause
of
poor
visibility
should
not
be
required
to
add
controls
because
the
emission
reductions
will
not
result
in
good
visibility
conditions.

Response:
The
visibility
goal
of
the
CAA
is
both
the
remedying
of
existing
impairment,
and
prevention
of
future
impairment.
The
court,
in
American
Corn
Growers,
upheld
our
interpretation
of
that
goal
as
the
return
to
natural
visibility
conditions.
Long­
term
regional
haze
strategies
are
developed
to
make
"
reasonable
progress"
towards
the
CAA
goal,
and
States
must
demonstrate
reasonable
progress
in
their
regional
haze
State
implementation
plans
(
SIPs).
Since
the
BART
program
is
one
component
of
that
demonstration,
visibility
changes
due
to
BART
are
appropriately
measured
against
the
target
of
natural
conditions.
With
regard
to
BART­
eligible
sources
not
being
in
operation
for
the
duration
of
the
program,
a
State,
in
making
BART
determinations,
is
explicitly
directed
by
the
CAA
to
account
for
the
remaining
useful
life
of
a
source.
Thus,
States
may
factor
into
their
reasonable
progress
estimates
those
shut­
downs
that
are
required
and
effected
in
permit
or
SIP
provisions.
In
addition,
15
Guidance
for
Estimating
Natural
Visibility
Conditions
Under
the
Regional
Haze
Rule
(
U.
S.
EPA.
September
2003).
http://
www.
epa.
gov/
ttncaaa1/
t1/
memoranda/
rh_
envcurhr_
gd.
pdf
Natural
background
conditions,
expressed
in
deciviews,
are
defined
for
each
Class
I
area.
EPA
has
issued
guidance
for
estimating
natural
background
conditions
which
has
estimates
of
default
conditions
as
well
as
measures
to
develop
refined
estimates
of
natural
conditions.

154
as
provided
for
under
our
guidance15,
proper
accounting
for
natural
phenomena
is
in
the
5
year
SIP
progress
report,
not
in
the
setting
of
natural
visibility
estimates.

4.1.4.8
Modeling
Domain
Comment:
Four
commenters
said
the
guidelines
must
clarify
which
BART
eligible
sources
need
to
be
included
in
making
this
demonstration
for
a
given
Class
I
Area.
(
28­
266,
28­
272,
28­
278,
28­
312).
Commenters
(
28­
266,
28­
312)
asked
which
BART
sources
in
the
large,
uniform
modeling
domain
covering
at
least
the
eastern
half
of
the
U.
S.
need
to
be
considered
in
making
this
demonstration
for
a
given
Class
I
area?
Commenter
28­
278
noted
that
there
is
no
provision
for
modeling
of
subgroupings
of
BART
sources
within
a
region,
nor
for
the
determination
that
some,
but
not
all,
of
the
BART
controls
are
needed.
However,
40
CFR
51.308(
e)(
1)(
ii)(
B)
states
that
this
analysis
is
to
consider
all
sources
subject
to
BART
"
located
within
the
region
that
contributes
to
visibility
impairment
in
the
Class
I
area."
This
latter
requirement
suggests
that
only
those
which
contribute
to
visibility
impairment
in
the
Class
I
area
of
interest
need
to
be
considered.
Therefore,
it
is
confusing
which
BART
sources
in
a
large
modeling
domain
need
to
be
considered
in
a
cumulative
analysis.

Response:
The
final
rule
provides
States
the
option
to
demonstrate,
cumulatively,
that
all
its
BART­
eligible
sources
are
subject
to
or
exempt
from
BART.
There
is
also
an
option
to
evaluate
BART­
eligible
sources
individually.
In
either
case,
the
State
need
only
analyze
sources
within
the
State
boundary
for
BART
applicability
determinations.

4.1.4.9
Engineering
Analysis
Must
Include
Source­
by­
Source
Visibility
Impact
Analysis
Comment:
Seven
commenters
said
the
BART
engineering
analysis
needs
to
include
an
evaluation
of
the
improvement
in
visibility
which
may
result
from
the
implementation
of
the
various
BART
technology
options
(
28­
302,
28­
303,
28­
305,
28­
342,
28­
346,
28­
357,
28­
390).
Some
commenters
(
28­
302)
said
EPA
must
strengthen
the
justification
for
their
proposed
treatment
of
visibility
improvements
on
a
cumulative
(
as
opposed
to
source­
by­
source)
basis.
As
the
proposed
BART
rule
states,
EPA
should
require
a
cumulative
air
quality
impacts
analysis
given
the
regional
nature
of
the
haze
problem.
Such
an
analysis
is
required
by
the
Act
(
28­
305).
Commenters
(
28­
155
390)
said
the
cumulative
air
quality
analysis
does
not
serve
this
purpose
since
it
is
conducted
after
the
BART
engineering
analysis
has
been
completed
and
the
BART
emission
limitation
established.
The
BART
engineering
analysis
should
first
require
a
baseline
comparison
of
the
BART
factors
for
all
the
various
control
alternatives,
from
the
status
quo
to
the
most
stringent
control
alternatives
(
28­
346,
28­
357).
The
BART
process
must
require
the
balancing
of
the
costs
or
a
proposed
emission
control
strategy
against
its
visibility
benefit.

Response:

The
guidelines
provide
the
States
with
an
approach
for
making
visibility
assessments.
We
do
not
agree
that
the
CAA
requires
a
cumulative
air
quality
impacts
analysis,
but
as
explained
in
the
preamble,
we
believe
States
have
the
discretion
to
pursue
such
an
approach.
Nothing
in
the
CAA
suggests
that
States
must
balance
the
costs
of
a
proposed
emission
control
strategy
against
its
visibility
benefit.

Comment:
Twenty­
nine
commenters
said
emission
control
decisions
must
be
made
on
a
case­
by­
case
basis
(
28­
243,
28­
275,
28­
329,
28­
331,
28­
333,
28­
337,
28­
340,
28­
341,
28­
342,
28­
346,
28­
353,
28­
354,
28­
356,
28­
357,
28­
360,
28­
361,
28­
364,
28­
389,
28­
395,
28­
398,
28­
401,
28­
402,
28­
405,
28­
406,
28­
7001,
28­
7003,
28­
7004,
28­
7007,
28­
7009)
for
several
reasons:

°
The
statute
and/
or
Congressional
intent
clearly
require
a
case­
by­
case
analysis
(
28­
243,
28­
337,
28­
341,
28­
346,
28­
353,
28­
354,
28­
356,
28­
359,
28­
360,
28­
361,
28­
364,
28­
389,
28­
395,
28­
405,
28­
406,
28­
7001,
28­
7003,
28­
7009)
°
Limits
states'
ability
to
design
a
program
that
more
cost
effectively
tailors
BART
to
impacts
of
individual
sources
(
28­
275,
28­
329,
28­
331,
28­
340,
28­
342,
28­
346,
28­
354,
28­
357,
28­
360,
28­
398,
28­
401,
28­
402,
28­
406,
28­
7004,
28­
7009)
°
Without
knowing
the
extent
of
a
source's
actual
contribution
to
visibility
impairment
and
the
degree
to
which
that
contribution
would
be
mitigated
by
a
given
set
of
controls,
there
is
no
basis
to
make
a
cost­
effectiveness
BART
determination
as
to
that
source
(
28­
333,
28­
346)
°
Sources
with
imperceptible
visibility
impacts
may
be
forced
to
install
costly
emission
control
equipment
even
though
little
if
any
visibility
improvement
would
be
obtained
(
28­
359,
28­
360,
28­
364,
28­
395)
°
Inconsistency
with
other
BART
criteria,
i.
e.,
one
of
the
BART
criteria
required
for
determining
which
retrofit
technology
is
best
for
an
individual
source
is
the
amount
of
visibility
improvement
which
can
be
anticipated
from
the
installation
and
use
of
that
technology
(
28­
361)
°
Inconsistency
with
prior
EPA
interpretations
(
28­
395,
28­
398).

Response:
The
final
rule
now
establishes
that
BART
controls
are
determined
on
a
source
by
source
basis.
156
Comment:
Ten
commenters
opposed
the
use
of
cumulative
air
quality
analysis
(
28­
273,
28­
302,
28­
304,
28­
337,
28­
349,
28­
356,
28­
357,
28­
358,
28­
404,
28­
7001).
Commenter
28­
273
said
if
the
emission
limit
is
determined
by
best
available
retrofit
technology,
then
the
air
quality
assessment
should
be
used
to
estimate
the
cumulative
effect
of
implementation
and
assess
achievement
of
the
goal,
not
for
further
reduction
of
limits.
Commenters
(
28­
337,
28­
349,
28­
356)
by
directing
that
visibility
benefits
be
considered
solely
on
a
collective
basis,
the
proposed
process
would
prevent
the
states
from
undertaking
the
source­
by­
source
weighing
of
costs
and
benefits
intended
by
Congress.
Commenter
28­
357
said
the
approach
is
inconsistent
with
the
cost/
benefit
approach
required
by
169A.

Response:
The
final
rule
now
establishes
that
BART
controls
are
determined
on
a
source
by
source
basis.

Commenter
28­
304
said
nothing
in
the
Act
or
EPA's
implementing
regulations
allows
the
determination
as
to
the
adequacy
of
BART
controls
to
be
based
on
an
anti­
degradation
analysis
that
encompasses
all
sources
of
visibility
impairing
pollutants.
Second,
EPA
has
offered
absolutely
no
factual
support
or
analysis
in
support
of
this
aspect
of
the
proposed
BART
guidance.
For
example,
EPA
should
have
investigated
whether
it
is
even
possible
for
BART
controls
on
BART
eligible
sources
to
achieve
this
anti­
degradation
criterion.
Lastly,
this
aspect
of
the
BART
guidance
is
plainly
unreasonable.
Simply
put,
there
is
no
rational
basis
for
EPA
to
conclude
that
the
level
of
control
that
should
be
applied
to
BART­
eligible
sources
can
or
should
be
dependent
on
construction
or
modification
of
sources
not
subject
to
BART.

Response:

The
BART
determination
is
evaluated
on
a
source
by
source
basis
and
requires
the
state
to
weigh
each
of
the
five
statutory
factors
including
the
visibility
benefits
for
various
control
technologies.
The
BART
determination
involves
an
analysis
of
only
BART
eligible
sources
and
does
not
consider
construction
or
modification
of
other
non
BART­
eligible
sources.

Commenter
28­
358
opposed
any
suggestion
that
the
cumulative
visibility
analysis
could
be
used
to
exempt
a
state
or
region
from
the
BART
requirements
or
to
create
exemptions
for
specific
sources
on
de
minimis
contribution
grounds.
EPA's
proposed
formulation
is
complicated,
unnecessary,
and
will
yield
results
contrary
to
the
statute's
purposes.
The
statute
requires
that
modeling
results
from
a
Cumulative
Air
Quality
Analysis
justify
the
application
of
BART
on
all
sources
subject
to
BART
in
a
region
if
they
demonstrate,
in
the
associated
Class
I
area,
that
any
existing
visibility
impairment
will
be
remedied
or
any
future
deterioration
in
visibility
will
be
prevented.
The
final
BART
guidelines
should
reflect
that
requirement.

Response:
157
The
commenter
suggests
a
reading
of
the
statute
that
was
the
basis
of
the
1999
regional
haze
rule
and
was
invalidated
by
the
D.
C.
Circuit.
In
the
final
guidelines,
the
option
of
using
cumulative
analysis
to
show
that
no
source
is
subject
to
BART
simply
reflects
the
logical
conclusion
that
if
a
group
of
sources
taken
together
do
not
contribute
to
impairment,
than
no
single
source
can
either.
States
also
have
the
ability
to
use
a
cumulative
analysis
to
show
that
a
individual
members
of
a
group
of
sources
are
reasonably
anticipated
to
contribute
to
impairment,
but
we
cannot
require
States
to
make
such
a
finding.

Commenter
28­
404
said
the
proposed
guidance
should
also
include
an
alternative
BARTevaluation
approach
that
would
start
with
the
stringent
control
alternative
to
increase
rule
flexibility.
In
all
cases,
BART­
evaluations
should
iterate
emissions
reductions
from
controls
compared
to
incremental
improvement
in
regional
haze
impacts
on
a
unit­
specific
basis
or
possibly
provide
for
alternative
performance­
based
emissions
standards.

Response:

The
guidelines
describe
a
reasonable
approach
for
making
BART
determinations.
States
generally
have
the
authority
to
adopt
reasonable
alternative
approaches.

4.1.4.9
Other
Comments
Regarding
the
Visibility
Improvement
Analysis
Comment:
Eight
commenters
(
0207,
0221,
0226,
0253,
0256,
0277,
0321,
28­
332)
provided
several
areas
where
the
guidelines
should
be
improved
or
clarified:
°
EPA
should
clarify
that
the
analysis
is
pollutant­
specific
(
e.
g.,
the
modeling
evaluation
of
a
BART
control
option
for
SO
2
reduction
should
not
be
combined
with
the
modeling
evaluation
of
a
BART
control
option
for
Nox)
(
0226,
0256).
°
EPA
should
clarify
that
only
the
closest
Class
I
area
must
be
modeled
(
0226,
0256).
°
EPA
should
describe
CALPUFF
as
one
possible
model
to
use,
rather
than
as
the
only
model
that
may
be
used
(
0256).
°
States
and
sources
should
have
the
flexibility
to
perform
multiple
modeling
runs
based
on
different
levels
of
available
control
(
0207,
0226,
0256,
0277).
°
Predicted
visibility
improvements
that
are
imperceptible
should
be
given
no
weight
in
determining
the
level
of
control
that
constitutes
BART
(
0253).
°
States
should
be
allowed
to
establish
the
required
degree
of
visibility
improvement
factor
(
0321).

°
EPA
should
clarify
how
the
cumulative
air
quality
analysis
can
be
reconciled
with
the
geographic
area
of
influence
determination
discussed
in
Section
II(
A)
(
28­
332).

Response:
The
final
rule
has
provided
clarification
of
each.
The
analysis
may
be
pollutant­
specific
at
the
discretion
of
the
State.
The
closest
Class
I
area
is
modeled,
but
some
strategic
receptors
158
should
be
placed
at
a
nearby
Class
I
area
to
ensure
that
the
highest
visibility
impacts
are
occurring
at
the
closet
park.
See
further
discussion
of
the
Modeling
Protocol
in
the
preamble
to
the
rule.
There
is
an
option
to
use
alternative
models
to
CALPUFF.
States
and/
or
sources
may
perform
multiple
modeling
runs.
Even
though
the
visibility
improvement
from
an
individual
source
may
not
be
perceptible
(
i.
e.
the
expected
improvement
may
be
less
than
or
equal
to
0.5
deciviews),
it
should
still
be
considered
in
setting
BART
because
the
contribution
to
haze
may
be
significant
relative
to
other
source
contributions
in
the
Class
I
area.
Thus,
we
disagree
that
the
degree
of
improvement
should
be
contingent
upon
perceptibility.
Failing
to
consider
less­
than­
perceptible
contributions
to
visibility
impairment
would
ignore
the
CAA's
intent
to
have
BART
requirements
apply
to
sources
that
contribute
to,
as
well
as
cause,
such
impairment.
States
can
assess
how
to
use
the
expected
visibility
improvement
due
to
various
control
technologies
in
setting
a
BART
emission
limit.
Finally,
there
is
no
requirement
to
reconcile
the
geographic
area
of
influence
determination
to
the
cumulative
air
quality
analysis,
and
it
is
no
longer
relevant
to
this
rule.

Commenter
0221
said
that
issues
relating
to
the
evaluation
of
the
visibility
improvement
in
determining
BART
controls
are
in
many
ways
the
same
as
issues
relating
to
the
evaluation
of
visibility
impairment
in
determining
which
BART­
eligible
sources
are
subject
to
BART.
Thus,
comments
relative
to
EPA's:
use
of
the
0.5
deciview
threshold
for
aggregate
visibility
impairment
applied
on
a
24­
hour
average
basis,
use
of
a
natural
visibility
baseline,
use
of
a
substantially
lower
deciview
threshold
to
determine
the
contribution
by
an
individual
source
to
visibility
impairment,
and
demonstration
of
those
thresholds
by
means
of
appropriate
modeling
rather
than
other
less
reliable
and
more
subjective
techniques
are
all
equally
applicable
here.

Response:
We
agree
and
have
addressed
these
comments
elsewhere
in
this
document.

Commenter
0267
supported
EPA's
proposed
method
to
determine
visibility
improvement
associated
with
installation
of
BART.

Commenter
28­
7001
said
location,
in
and
of
itself,
is
not
a
means
of
determining
impact.
By
EPA's
own
admission,
the
atmospheric
processes
that
determine
downwind
visibility
impact
on
a
Class
I
area
are
part
of
a
process
including
prevailing
met
conditions,
actual
emissions
from
a
particular
source,
interactions
with
other
source
emissions
and
distance
and
directions
of
transport.

Response:
We
agree
and
models
account
for
the
various
meteorological
and
atmospheric
chemistry
influences
when
predicting
visibility
impacts
­
the
impact
may
not
be
determined
simply
by
a
source's
location.

4.1.5
Applicability
to
Plants
that
Have
Installed
BACT
or
LAER
or
are
Subject
to
MACT
or
NSPS
Comment:
Eight
commenters
supported
the
proposal
that
MACT
compliance
should
count
as
BART
compliance
because
MACT
controls
are
far
more
stringent
than
the
requirements
for
BART
(
0169,
159
0184,
0186,
0189,
0201,
0207,
0304,
0308).
Commenter
0201
added
that
EPA
may
want
to
clarify
that
the
MACT
standard
must
have
regulated
the
subject
visibility­
impairing
pollutant.

Response:

We
do
not
agree
that
MACT
is
necessarily
more
stringent
than
BART.
States
will
be
able
to
judge
whether
this
is
true
in
particular
instances.
We
do
agree
that
the
MACT
standard,
to
be
relevant,
must
have
regulated
the
same
pollutant
as
is
regulated
by
BART.

Comment:
Several
commenters
asked
EPA
to
extend
this
provision
to
sources
that
have
recently
implemented
NSPS
(
0169,
0184)
and/
or
BACT
or
LAER
(
0184,
0207,
0214,
0215,
0294).
Commenters
added
that
sources
that
are
modified
after
1977
should
not
be
included
in
the
BART
pool
because
they
would
already
meet
BACT,
LAER,
and/
or
NSPS
controls
(
0184,
0214,
0215,
0228,
0237,
0238,
0253,
0254,
0256,
0268).
Commenter
0256
added
that
any
requirement
of
a
BART
review
for
new,
post­
1977
emissions
units
or
modified
emissions
units
would
be
inconsistent
with
Congress'
intent
to
limit
BART
to
sources
in
existence
on
August
7,
1977.
Congress
chose
that
date
as
the
cutoff
for
BART
because
units
added
after
that
date
typically
would
have
been
subject
to
NSR
(
unless
they
avoided
NSR
by
netting
out).

Commenter
0237
distinguished
between
sources
reconstructed
after
1977
and
subject
to
NSPS
and
sources
subject
to
minor
or
major
PSD/
NSR
analysis
after
1980
as
not
being
BARTeligible
Commenter
0215
added
existing
units
that
began
operation
before
August
7,
1962
are
not
BART
eligible,
regardless
of
any
changes
made
to
those
units
in
the
1962­
1977
time
frame
(
or
thereafter).
This
would
be
an
unlawful
retroactive
application
of
the
Act.
Commenter
0242
agreed.

Response:

We
have
added
some
discussion
of
the
interaction
of
BART
and
these
other
provisions
to
the
guidelines.
We
have
also
noted
that
sources
which
have
reconstructed
after
1977
and
been
subject
to
new
source
standards
are
not
BART­
eligible.

4.1.6
General
Comments
Regarding
the
Engineering
Analysis
Comment:
Six
commenters
suggested
ways
EPA
should
provide
flexibility
and
resources
to
the
states
in
conducting
the
engineering
analyses
(
0197,
0201,
0237,
0245,
0275,
0302).
Some
commenters
(
0197,
0302)
said
states
do
not
have
the
resources
to
perform
many
site­
specific
engineering
analyses
and
flexibility
is
needed
to
allow
them
to
cost­
effectively
find
the
necessary
emission
reductions
(
0302).
Commenter
0197
said
EPA
should
not
limit
a
state's
freedom
to
identify
which
160
sources
contribute
to
visibility
impairment
and
by
directing
how
states
must
apply
cost­
benefit
considerations
in
deciding
whether
additional
controls
may
be
needed.

Commenters
said
EPA
has
ignored
the
costs
of
engineering
analyses
(
0237)
and
should
help
states
to
address
the
resource
needs
including:
°
Provide
funding
for
states
to
accomplish
the
SIP
development
and
engineering
analysis
activities
that
the
proposal
mandates
(
0201,
0245)
°
Provide
written
guidance
for
industry
in
accomplishing
the
required
engineering
analysis
(
0201,
0245)
°
Provide
or
support
RPO
development
of
potential
emission
control
options
by
major
source
category
(
0201,
0245)
°
Develop
and
maintain
a
clearinghouse
of
BART
determinations
(
0201,
0245).

Three
commenters
said
EPA
should
establish
a
"
BART
Clearinghouse"
to
provide
states
with
information
regarding
retrofit
control
information
to
help
them
make
technology
determinations
(
28­
254,
28­
264,
28­
290).
Commenter
28­
290
added
that
a
process
is
needed
that
will
allow
decision­
makers
to
pool
information
and
coordinate
efforts.
Another
commenter
(
28­
244)
said
individual
states
or
groups
of
states
should
adopt
a
generic
approach
to
at
least
of
the
BART
issues
in
the
interest
of
consistency
and
efficiency,
while
reserving
some
others
for
case­
bycase
"
balancing."

Response:

Commenters
have
made
many
good
suggestions
above
for
the
implementation
phase
of
BART,
and
we
will
consider
them.

Comment:
Commenter
0275
said
if
states
find
that
additional
controls
are
needed
to
achieve
continued
visibility
improvements,
states
should
select
the
sources
to
be
controlled
on
the
basis
of
visibility
impact
and
cost­
effectiveness
as
the
primary
factors,
with
other
secondary
factors
such
as
state
of
development
of
control
technology
and
administrative
costs
also
considered.
Sources
should
not
be
exempted
from
additional
control
or
review
simply
because
they
were
controlled
under
CAIR
or
BART.

Response:

States
have
considerable
flexibility
in
determining
how
best
to
weigh
the
BART
factors.

Comment:
Two
commenters
said
the
required
analyses
will
require
substantial
time
and
effort
to
perform
(
28­
266,
28­
312).
Given
the
potentially
large
number
of
sources
subject
to
BART
in
some
regions
(
and
the
need
to
review
each
visibility­
impairing
pollutant,
even
if
the
potential
emissions
for
only
one
of
these
pollutants
exceeds
the
250
TPY
threshold),
this
requirement
will
161
pose
a
tremendous
burden
on
the
states.
Another
commenter
(
28­
345)
noted
that
the
engineering
and
impact
analyses
are
far
too
prescriptive
and
exceed
EPA's
authority.

Response:

We
believe
that
the
BART
guidelines
appropriately
balance
Congressional
directive
and
implementation
issues,
and
we
believe
we
have
provided
as
much
flexibility
as
possible
to
States
in
order
to
facilitate
implementation.
We
believe
States
will
be
able
to
implement
BART
as
it
has
been
finalized.

4.2
Select
a
Best
Alternative
Using
the
Results
of
an
Engineering
Analysis
4.2.1
Select
a
"
Best"
Alternative
4.2.1.1
Top­
down
Analysis
Comment:
Forty
commenters
supported
using
a
top­
down
analysis
for
performing
BART
engineering
analyses
(
0159,
0200,
0221,
0232,
0234,
0267,
0284,
0304,
0314,
0323,
28­
247,
28­
248,
28­
254,
28­
255,
28­
256,
28­
257,
28­
258,
28­
260,
28­
262,
28­
263,
28­
265,
28­
267,
28­
289,
28­
290,
28­
302,
28­
332,
28­
334,
28­
335,
28­
338,
28­
350,
28­
352,
28­
358,
28­
362,
28­
376,
28­
389,
28­
399,
28­
770,
28­
7008,
28­
7016,
28­
7048).
Commenters
(
0221,
28­
335,
28­
358)
said
the
idea
that
the
BART
engineering
analysis
should
begin
with
the
most
environmentally
stringent
control
option
and
proceed
from
there
to
less
stringent
controls
is
reinforced
by
the
similarities
between
BART
and
the
NSR
BACT
standard,
which
uses
the
top­
down
approach
and
is
consistent
with
Congressional
intent.
Commenters
(
28­
254,
28­
260,
28­
263,
28­
265,
28­
7016)
added
that
the
plain
terminology
of
BART
is
best
available
retrofit
technology.
If
the
most
effective
removal
technology
is
not
the
starting
point
of
the
analysis,
then
the
other
BART
factors
may
not
be
applied
to
a
consistent,
established
benchmark.
In
addition,
this
approach
is
consistent
with
other
programs
such
as
PSD
(
28­
254,
28­
255,
28­
335,
28­
350,
28­
362,
28­
389).

Commenter
0234
said
that
to
consider
less
stringent
controls
first
would
limit
the
effectiveness
of
the
BART
program
to
reduce
emissions.
Commenter
0304
said
using
a
top­
down
analysis
places
the
burden
of
proof
on
the
individual
source
to
show
that
a
certain
level
of
control
is
not
technologically
feasible
or
not
cost
effective,
but
using
a
bottom­
up
approach
places
the
burden
of
proof
on
the
agency
to
show
that
the
next
higher
level
of
control
is
appropriate.
Commenters
(
0323,
28­
350)
said
a
top­
down
approach
is
more
straightforward
to
implement
than
the
alternative
and
would
tend
to
more
thoroughly
consider
stringent
control
alternatives.
The
final
selection
of
the
control
technology
should
be
based
on
a
review
of
all
the
BART
factors,
to
assure
a
reasonable
balancing
of
dollar
per
ton
removal
and
degree
of
improvement
in
visibility.
Commenter
28­
267
said
the
process
should
include
due
consideration
of
the
differences
between
the
BACT
and
BART
processes.
162
Commenters
(
28­
332,
28­
358)
said
the
top­
down
approach
would
provide
greater
consistency
with
other
regulations
(
BACT)
and
reduce
administrative
requirements.

Comment:
Fifty­
nine
commenters
opposed
using
a
top­
down
analysis
(
0182,
0186,
0188,
0189,
0195,
0208,
0210,
0214,
0217,
0223,
0226,
0228,
0237,
0238,
0241,
0242,
0253,
0256,
0268,
0270,
0275,
0294,
28­
259,
28­
270,
28­
275,
28­
306,
28­
331,
28­
337,
28­
337,
28­
339,
28­
340,
28­
343,
28­
344,
28­
345,
28­
347,
28­
349,
28­
355,
28­
356,
28­
357,
28­
361,
28­
364,
28­
383,
28­
384,
28­
385,
28­
386,
28­
392,
28­
395,
28­
401,
28­
402,
28­
405,
28­
747,
28­
748,
28­
778,
28­
779,
28­
7001,
28­
7003,
28­
6004,
28­
7009,
28­
7013).
Commenters
(
0182,
0208,
0217,
0241,
0270,
0275,
28­
33,
28­
345,
28­
347,
28­
357,
28­
392,
28­
778,
28­
7003)
said
this
approach
is
contrary
to
the
Act
and
that
had
Congress
wanted
the
states
to
perform
a
BACT­
like
analysis
on
BART
eligible
sources,
it
would
have
done
so
explicitly.
Commenters
(
0186,
0208,
0270,
0275,
0294,
28­
339,
28­
340,
28­
343,
28­
349,
28­
361,
28­
364,
28­
383,
28­
747,
28­
748,
28­
7003)
said
EPA's
prescriptive
guidelines
for
making
BART
determinations
undermine
the
statutory
authority
given
to
the
states
in
the
Act.
Commenter
0237
said
EPA
lacks
authority
to
disapprove
a
state's
criteria
for
determining
BART.
Commenter
0189
provided
a
legal
analysis
supporting
the
Congressional
grant
of
substantive
discretion
to
states
to
choose
their
own
BART.
This
approach
also
circumvents
the
60­
year
plan
for
attaining
visibility
goals
by
defining
BART
to
be
so
stringent
that
States
will
be
far
exceeding
reasonable
progress
expectations
at
the
outset
of
the
program.
This
approach
is
inconsistent
with
congressional
intent
to
spread
the
cost
of
the
regional
haze
program
over
a
60­
year
period
(
0186).
Commenter
29­
275
added
that
assuming
a
starting
point
at
the
most
stringent
level
and
subjecting
the
analysis
to
"
correct
until
demonstrated
wrong"
methodology
is
clearly
inappropriate
for
a
long
range
program
with
as
many
technical
uncertainties
as
exists
in
the
regional
haze
program.

Commenters
(
0188,
0195,
0210,
0214,
0223,
0226,
0238,
0253,
0256,
0268,
28­
259,
28­
270,
28­
306,
28­
331,
28­
337,
28­
355,
28­
357,
28­
385,
28­
779,
28­
7001)
said
that
the
top­
down
methodology
allows
a
BART
determination
to
be
made
without
taking
into
account
the
statutory
factors
that
Congress
indicated
must
be
taken
into
account.
The
only
satisfactory
approach
is
one
in
which
all
the
BART
control
options
are
considered
and
compared
in
light
of
the
statutory
factors
(
cost,
visibility
improvement,
and
so
on).
Commenters
(
0223,
0230)
said
EPA
should
clarify
that
all
five
factors
listed
in
Act
must
be
addressed
regardless
of
the
approach
adopted.

Some
commenters
stressed
the
need
to
conduct
the
cost­
benefit
assessment
required
by
the
Act
instead
of
shifting
the
process
to
a
"
top
down"
BACT
program
that
is
only
applicable
under
the
NSR
provisions
(
28­
259,
28­
270,
28­
306,
28­
331,
28­
384,
28­
386,
28­
401,
28­
402,
28­
7009).
Other
commenters
(
0188,
0242,
0268,
28­
357)
added
that
this
approach
does
not
take
into
account
visibility
improvement
among
potential
control
options.
Commenters
(
0253,
0256)
said
EPA's
proposed
methodology
inflates
the
visibility
improvements
associated
with
a
given
level
of
control
in
a
manner
that
skews
the
BART
analysis.
Visibility
improvement
is
essentially
overweighted
and
the
other
statutory
factors
are
underweighted.
163
Commenter
0195
said
without
the
consideration
of
any
existing
pollution
controls
in
place,
a
strict
top­
down
methodology
would
potentially
require
upgrades
to
an
already
controlled
facility.
Commenter
28­
344
said
the
proposed
guidelines
treat
BART
as
the
equivalent
of
"
top
down
BACT",
thus
failing
to
take
into
account
that
BART
was
intended
to
be
retrofitted
onto
existing
facilities
that
are
not
otherwise
undertaking
modifications.
Commenters
(
28­
392,
28­
7013)
added
that
the
"
top
down"
BACT
approach
hides
the
high
incremental
costs
associated
with
very
expensive
emission
reductions
by
averaging
those
costs
with
the
more
cost
effective
emissions
reductions.
The
"
top
down"
approach
also
creates
a
false
presumption
that
sources
with
existing
controls
cannot
meet
BART
without
additional
emission
controls.

One
commenter
(
28­
247)
said
the
presumption
of
maximum
stringency
would
be
particularly
burdensome
as
applied
to
smaller
utility
and
industrial
boilers.
Industrial
boilers
generally
differ
from
the
750
MW+
utility
boilers
contemplated
for
BART
regulation
in
the
statute
in
three
principal
respects:
size,
load
variability,
and
design
of
the
boilers.

Response:

We
believe
that
the
best
approach
is
to
provide
States
with
flexibility
to
use
the
process
tht
works
best
in
their
decision­
making
structure.
Therefore
we
are
not
recommending
a
specific
method
of
arraying
control
options.

4.2.1.2
Bottom­
up
Analysis
Comment:
Thirteen
commenters
supported
a
"
bottom­
up"
approach
by
evaluating
the
lest
stringent
technically
feasible
control
option
(
0184,
0202,
0241,
0247,
0302,
28­
259,
28­
268,
28­
273,
28­
306,
28­
345,
28­
385,
28­
390,
28­
392,
28­
405,
28­
747,
28­
778,
28­
7013).
Commenters
(
0184,
0202,
28­
273,
28­
306,
28­
345,
28­
385)
said
BACT,
LAER,
NSPS,
or
other
requirements
should
serve
as
a
BART
ceiling.
Commenter
0247
said
the
Act
does
not
mandate
that
sources
installing
BART
meet
the
maximal
level
of
control
required
for
BACT.
BART
is
intended
to
apply
to
sources
that
were
never
subject
to
PSD,
and
represents
a
different
control
requirement.

Commenter
0241
added
that
states
should
not
be
forced
to
justify
why
more
stringent
control
alternatives
were
not
selected.
If
a
control
technology
results
in
an
improvement
in
visibility
that
the
state
determines
is
appropriate
to
meet
its
regional
haze
SIP
requirements,
then
no
such
justification
is
needed.
Commenter
0302
would
require
more
documentation
however.
According
to
the
commenter
under
this
process,
states
would
still
be
required
to
(
1)
display
and
rank
all
of
the
options
in
order
of
control
effectiveness
and
identify
the
average
and
incremental
costs
of
each
option;
(
2)
consider
the
energy
and
non­
air
quality
environmental
impacts
of
each
option;
and
(
3)
provide
a
justification
for
adopting
the
technology
that
the
MDEQ
selects
as
the
"
best"
level
of
control,
including
an
explanation
as
to
why
other
more
stringent
control
technologies
were
rejected.
The
level
of
documentation
required
under
this
method
would
give
164
necessary
transparency
for
industry
and
the
general
public.
However,
this
needs
a
firm
national
policy
and
guidance
in
order
to
implement
the
BART
regulations.

One
commenter
(
0221)
opposed
the
use
of
a
bottom­
up
approach
because,
by
its
very
nature,
less
likely
to
result
in
the
"
best
available"
controls
for
limiting
haze­
forming
pollution.
For
the
most
part,
the
most
stringent
control
technologies
for
reducing
visibility
impairing
pollutants
such
as
SO
2
and
NO
x
tend
to
be
widely
"
available."
Thus,
by
allowing
a
bottom­
up
approach,
EPA
would
be
increasing
the
complexity
of
the
process,
the
likelihood
of
analytical
mistakes,
and
the
opportunities
for
"
gaming."
Because
a
bottom­
up
approach
will
require
sources
to
consider
the
full
range
of
available
control
technologies,
it
significantly
increases
the
administrative
burden
on
states
involved
in
the
BART
process.
Finally,
there
simply
is
no
substantive
justification
supporting
the
adoption
of
a
bottom­
up
(
or
middle­
both
ways)
approach
to
the
selection
of
BART
controls.

Response:

We
are
not
endorsing
a
bottom­
up
approach
as
the
most
useful
approach
for
weighing
control
options.

4.2.1.4
State
Determines
Approach
Comment:
Four
commenters
supported
an
approach
that
would
allow
a
state
more
options
in
how
it
structures
its
BART
analysis
(
28­
347,
28­
385,
28­
779,
28­
7005).
Commenter
28­
347
added
that
a
state
could
start
with
the
least
stringent
control
alternative
or
somewhere
in
the
middle
of
the
range
and
then
consider
the
additional
emission
reductions,
costs
and
other
impacts
of
successively
more
stringent
control
alternatives.
The
Act
grants
states
the
primary
role
in
determining
BART
and,
thus,
the
state
should
be
free
to
select
a
"
bottom
up"
or
intermediate
approach
in
conducting
the
engineering
analysis.

Two
commenters
said
the
starting
point
of
the
analysis
does
not
matter,
because
the
result
will
be
the
same
(
28­
243,
28­
337).
Commenter
28­
343
added
that
when
choosing
the
best
BART
option,
EPA
still
requires
that
the
most
stringent
alternative
be
picked
unless
another
one
can
be
justified.

Seven
commenters
opposed
an
approach
where
the
state
could
consider
alternative
approaches
for
several
reasons
(
28­
254,
28­
255,
28­
302,
28­
338,
28­
343,
28­
358,
28­
7016):

°
Administrative
burden
for
states
and
tribes
(
28­
254)
°
Inconsistency
with
parallel
reviews
of
BART
determinations
(
28­
254)
°
Inconsistent
analyses
across
the
country
(
28­
255)
°
Inconsistent
with
the
statutory
objective
of
identifying
the
best
control
options
(
28­
302,
28­
338,
28­
358,
28­
7016).
165
Commenter
28­
343
added
that
the
state
is
still
required
to
rank
all
options
in
order
of
control
effectiveness
and
consider
the
energy
and
nonair
quality
environmental
impacts
of
each
option.
This
approach
is
overly
prescriptive
and
usurps
the
responsibilities
of
the
states.

Response:

We
have
provided
a
more
flexible
approach
to
arraying
control
options
in
the
final
BART
guidelines,
as
we
believe
States
can
and
should
exercise
their
discretion
under
the
CAA
in
choosing
the
method
that
works
best
for
them.

4.2.1.5
Cost­
effective
Option
Comment:
Two
commenters
supported
an
alternative
approach
proposed
by
EPA
to
allow
the
analysis
to
begin
with
the
most
cost­
effective
control
option
(
28­
259,
28­
268).
Commenter
28­
268
added
that
it
is
more
appropriate
to
conduct
marginal
engineering
cost
analysis
for
determining
the
most
cost­
effective
BART
options.
The
true
cost­
effectiveness
of
each
option
is
not
clear
until
a
marginal
analysis
is
done.

Two
commenters
said
the
process
should
apply
"
middle
of
the
road
technologies"
that
accomplishes
reductions
at
more
reasonable
costs
(
28­
275,
28­
333).

Response:

We
believe
cost­
effectiveness
need
not
be
given
greater
weight
than
other
options,
although
we
recognize
its
importance.

4.2.1.6
Other
Comments
Regarding
Selecting
a
"
Best
Alternative"

Comment:
Fifteen
commenters
provided
other
comments
regarding
selecting
a
best
alternative
(
0197,
0207,
0229,
0234,
0254,
0256,
0293,
0299,
0319,
0321,
0319,
28­
266,
28­
272,
28­
273,
28­
344,
28­
312).

Commenters
said
that
EPA's
proposals
are
too
prescriptive
and
limit
a
state's
freedom
to
identify
which
sources
contribute
to
visibility
impairment
and
by
directing
how
states
must
apply
cost­
benefit
considerations
in
deciding
whether
additional
controls
may
be
needed
(
0197,
0207,
0321).
Commenter
0207
particularly
objected
to
the
prescriptive
nature
of
Appendix
Y
and
stressed
that
the
Corn
Growers
decision
explained
that
all
five
169A(
g)(
2)
factors
are
meant
to
be
considered
by
the
states.
Commenters
(
0229,
0256,
0319)
said
that
so
long
as
all
five
factors
are
evaluated,
the
order
of
analysis
should
be
left
to
the
states.
166
Response:

We
agree,
and
have
left
the
order
of
analysis
to
the
States.

Comment:
Commenter
0234
said
BART
controls
and
exemption
policy
must
be
consistently
applied
to
sources
across
a
wide
region.
Consistency
between
regional
planning
organizations
and
EPA
regions
in
implementing
the
BART
guidelines
is
critically
important
to
the
success
of
the
regional
haze
program.
Commenter
0299
was
concerned
that
the
BART
Guidelines,
insofar
as
they
establish
a
case­
by­
case
review
of
individual
sources,
creates
a
regulatory
framework
that
will
result
in
inconsistent
control
determinations.
This
is
especially
true
in
the
absence
of
clear
criteria
for
making
determinations
and
where
more
than
50
different
regulatory
agencies
will
be
making
the
decisions.
Commenter
28­
272
stressed
the
need
to
encourage
consistency
between
states
in
the
implementation
of
BART
on
a
least
a
regional
basis.

Response:

We
believe
the
final
guidelines
implement
Congressional
intent,
and
we
also
strongly
encourage
States
to
work
together
toward
as
much
consistency
as
possible.

Comment:
Commenter
0293
appreciated
the
additional
flexibility
provided
in
the
revised
BART
guidelines.
Maintaining
flexibility
in
measures
to
achieve
reasonable
progress
goals
allows
states
to
develop
appropriate
strategies
according
to
the
contributions
to
regional
haze
at
each
Class
1
area.
The
proposed
rule
and
Guidelines
support
state
discretion
in
determining
which
BART
controls
will
be
most
effective
in
each
of
the
respective
source
categories.

Response:

We
agree
that
flexibility
and
State
discretion
are
important.

Comment:
Commenter
0256
said
the
procedures
described
in
the
guidelines
for
choosing
BART
are
hard
to
follow
and
do
not
consistently
seem
to
include
consideration
of
the
visibility
improvement
factor.
The
commenter
suggested
specific
deficiencies
that
should
be
corrected.

Response:
We
believe
we
have
provided
a
reasonable
process
for
visibility
assessment
and
we
encourage
States
to
follow
it.

Comment:
Commenters
(
0391,
28­
266)
said
the
proposed
BART
rule
may
cause
states
and
companies
to
incur
substantial
costs
undertaking
modeling
and
evaluating
impacts.
The
rule
appears
to
be
a
167
mandate
without
providing
the
funding
necessary
to
implement
the
rule
at
the
state
level.
This
is
a
significant
burden
for
states
already
struggling
with
insufficient
revenues
to
support
imposing
administrative
burdens
and
costs
without
funding
existing
services,
let
alone
assume
additional
responsibilities.
Commenters
(
28­
266,
28­
312)
added
that
EPA
should
consider
alternative
approaches
to
a
source­
by­
source
analysis
for
each
pollutant,
such
as
multi­
pollutant
regional
or
source­
category
analyses.

Response:

We
recognize
the
critically
important
role
of
the
regional
planning
organizations
(
RPOs)
in
help
with
the
technical
analysis
and
planning
for
BART
implementation.
We
believe
that
with
the
help
of
the
RPOs
the
States
are
certain
to
be
able
to
implement
BART
in
a
timely
and
efficient
way.

Comment:
Commenter
28­
273
said
use
of
the
most
stringent
achievable
reduction
is
unreasonable
in
evaluating
a
control
for
the
most
stringent
level
that
the
control
is
"
capable
of
achieving."
It
would
be
more
reasonable
to
take
a
statistical
approach.
The
commenter
added
that
whatever
averaging
basis
is
used
for
the
BART
evaluation
needs
to
be
the
averaging
period
to
which
the
unit
is
regulated.

Response:

We
believe
States
have
the
flexibility
to
consider
statistical
approaches
where
relevant.

Comment:
Four
commenters
opposed
both
of
the
alternatives
proposed
by
EPA
(
28­
244,
28­
333,
28­
344,
28­
391).
Both
approaches
involve
more
than
justifiable
limitations
on
the
regional
or
state
authority
to
determine
BART
on
a
case­
by­
case
basis.
Commenters
(
28­
333,
28­
391)
added
that
these
approaches
do
not
determine
the
degree
of
visibility
improvement
that
would
result
from
a
source
installing
a
particular
level
of
BART
controls.

Response:

We
have
left
the
order
of
analysis
to
the
States,
and
believe
they
will
weigh
the
visibility
improvement
factor
along
with
other
factors
in
ultimately
determining
BART
control
for
specific
sources.

Comment:
Commenter
28­
344
said
the
proposed
guideline
requests
comment
on
whether
BART
should
be
"
top­
down"
or
"
bottom­
up,"
without
first
questioning
the
departure
from
NSPS
as
the
statutory
reference
point
for
BART
controls
or
providing
an
adequate
basis
for
that
departure.
168
The
EPA
must
revise
the
NSPS
(
as
was
intended
every
5
years)
and
these
standards
must
be
the
reference
for
BART
decisions.

Response:

We
noted
earlier
that
we
believe
NSPS
is
a
point
of
departure
but
not
always
an
appropriate
end
point
for
the
BART
determination.

4.2.2
Consideration
of
the
Affordability
of
Controls
Comment:
Seventeen
commenters
supported
considering
the
affordability
of
controls
(
0189,
0199,
0241,
0270,
28­
329,
28­
337,
28­
349,
28­
353,
28­
354,
28­
356,
28­
363,
28­
364,
28­
390,
28­
395,
28­
405,
28­
779,
28­
7001).
Commenters
added
that
a
strict
$/
ton
decision
point
would
be
too
limited
and
factors
such
as
visibility
improvement
(
28­
329,
28­
337,
28­
349,
28­
353,
28­
354,
28­
363,
28­
364,
28­
390,
28­
395,
28­
405,
28­
7001),
effects
on
secondary­
formed
particles
(
28­
353,
28­
354),
and
viability
of
continuing
plant
operation
(
28­
779)
must
be
considered.

Commenters
(
28­
364,
28­
395)
said
the
proposed
BART
rule,
like
the
RHR,
nullifies
the
individual
source
cost­
effectiveness
analysis
set
forth
in
Section
169A(
g)(
2)
by
bifurcating
the
implementation
of
the
statutory
BART
cost­
effectiveness
factors.
T
he
RHR
and
proposed
BART
rule
provide
for
an
analysis
of
all
but
the
visibility
improvement
factor
on
an
individual
source
basis.
The
last
factor
is
required
to
be
analyzed
on
a
collective
basis;
that
is,
states
are
prohibited
from
examining
the
degree
of
improvement
that
would
result
from
installing
BART
on
individual
sources
but
are
instead
required
to
examine
the
degree
of
improvement
that
would
result
from
installing
BART
on
all
BART­
eligible
sources.
Such
a
regime
is
not
the
cost­
benefit
analysis
required
by
Congress
in
Section
169A(
g)(
2).

Two
commenters
opposed
consideration
of
any
compliance
costs
in
the
BART
guidelines
because
they
will
not
improve
implementation
and
will
vary
widely
based
on
individual
sources
(
28­
362,
28­
770).
Commenter
28­
7005
said
a
strict
dollar­
per­
ton
approach
would
affect
sources
that
have
little
or
no
impact
on
Class
I
areas
exactly
the
same
as
sources
in
close
proximity
to
Class
I
areas.
As
a
result
this
type
of
focus
is
unwarranted.

Response:

We
believe
this
measure
of
cost­
effectiveness
provide
useful
information
and
warranted.

Comment:
Ten
commenters
noted
additional
impacts
that
should
be
considered
(
0189,
0199,
0202,
0275,
0300,
0310,
28­
243,
28­
306,
28­
346,
28­
361).
Commenter
0189
said
that
potential
"
shutdown
costs
should
be
considered
by
a
state
in
reducing
the
BART
controls
on
a
given
source.
169
Commenter
0275
said
EPA's
analysis
of
the
costs
and
benefits
of
complying
with
the
proposed
BART
program
takes
credit
for
benefits
realized
under
other
regulatory
programs,
but
not
the
costs
of
those
programs.
As
such,
the
analysis
is
arbitrary
and
unsupportable.

Commenter
28­
243
said
an
estimate
with
an
accuracy
of
up
to
plus
or
minus
30%
can
lead
to
erroneous
outcomes.
A
more
accurate
detailed
cost
estimate
will
produce
more
credible
estimates.
Commenter
28­
346
added
that
by
focusing
attention
on
the
cost
per
ton
of
removal,
the
guidelines
would
divert
attention
away
from
what
the
visibility
protection
program
is
all
about
(
visibility
improvements)
and
toward
emission
management
for
the
sake
of
emission
management.

Commenters
(
0199,
0202,
0300,
0310,
0202,
28­
306,
28­
361)
were
concerned
about
impacts
on
specific
industries.
Commenters
(
0202,
28­
306)
said
the
potential
impact
of
the
guidelines
and
BART
requirements
is
of
special
concern
for
the
aluminum
industry,
which
is
facing
ongoing
energy
supply
and
cost
problems.
The
impact
of
BART
could
not
only
create
higher
direct
costs
for
pollution
controls
on
aluminum
facilities,
but
would
also
create
higher
electricity
prices
due
to
the
costs
of
implementation
to
the
electric
utility
industry.
Commenter
28­
361
said
the
mining
industry
would
be
adversely
affected
by
the
Proposed
Guidelines
for
BART
determinations
if
they
exceed
statutory
authority,
thereby
resulting
in
excessive
costs
to
retrofit
power
plants
and
other
industrial
sources,
or
result
in
a
switch
to
natural
gas
as
a
fuel
source.

Commenters
(
0300,
0310)
noted
the
important
role
lignite
fuel
plays
in
Texas
and
said
consideration
of
energy
affordability
in
the
regulatory
program
is
essential.
Commenter
0319
added
that
BART
as
proposed,
will
be
another
air
quality
rule
that,
in
conjunction
with
recently
proposed
rules
governing
mercury,
SO
2
and
NO
x
emissions,
could
make
lignite
and
tribal
coals
uncompetitive
and
further
concentrate
U.
S.
coal
supply
in
the
Southern
Powder
River
Basin.
(
SPRB).

Commenter
0199
was
concerned
that
any
further
emission
reductions
applicable
to
electric
generating
units
(
EGU)
be
implemented
in
a
cost
effective
manner,
with
reasonable
time­
frames
to
minimize
the
impact
to
residential
and
industrial
customers
alike.

Response:

We
note
that
commenters
from
different
industries
have
different
views
of
affordability
analysis.
This
points
out
that
flexibility
is
needed
for
States
to
take
into
account
site­
specific
issues.

Comment:
One
commenter
(
28­
358)
said
the
six
statutory
BART
factors
must
be
considered,
not
weighed
 
a
cost­
benefit
approach
may
not
become
the
basis
for
BART
determinations.
The
six
BART
factors
are
simply
listed
as
those
Congress
directs
EPA
and
the
states
to
consider
in
determining
BART.
Congress
has
delegated
to
EPA
and
the
states
the
responsibility
for
developing
a
process
by
which
the
BART
factors
are
considered.
Congress
has
not,
however,
170
required
EPA
or
the
states
to
use
any
specific
method
such
as
a
balancing
test
in
assessing
the
consideration
factors
listed
at
section
169A(
g)(
2),
nor
are
EPA
and
the
states
required
to
give
each
consideration
factor
any
specific
weight.
EPA,
therefore,
should
clarify,
in
the
final
rule,
that
a
cost­
benefit
analysis
cannot
be
used
to
define
BART,
nor
should
the
cost­
effectiveness
analysis
(
either
average
or
incremental
cost­
effectiveness128)
become
the
fulcrum
of
the
analysis,
effectively
emasculating
the
other
BART
consideration
factors.

Commenter
28­
258
said
one
concern
with
including
costs
is
that
the
control
may
be
rejected
solely
on
costs
rather
than
looking
at
all
of
the
source­
specific
factors
and/
or
if
it
is
necessary
to
ensure
reasonable
progress
for
improving
visibility
in
Class
1
areas.
Even
if
the
costs
are
in
the
high
range,
the
guidelines
should
include
provisions
for
requiring
presumptive
BART.
The
cost­
benefit
approach
will
just
serve
as
a
mechanism
to
further
delay
implementation
of
BART
through
a
complicated
analysis
of
impacts
on
visibility
from
all
SIP
measures.

Response:

We
are
not
advocating
a
cost­
benefit
test
in
the
final
BART
guidelines.

Comment:
Commenters
(
0241,
0270)
said
on
the
issue
of
the
costs
of
control,
EPA
clearly
is
usurping
the
states'
authority
under
Section
169A,
which
requires
states
to
look
at
the
costs
of
compliance
in
determining
BART.

Response:

We
believe
States
have
the
flexibility
to
consider
costs
of
compliance
in
different
ways.

Comment:
Seven
commenters
addressed
the
use
of
incremental
cost
effectiveness
analysis
(
28­
255,
28­
256,
28­
264,
28­
273,
28­
302,
28­
306,
28­
404,
28­
7008).
Some
commenters
(
28­
273,
28­
306,
28­
404)
said
the
rule
should
be
based
on
consideration
of
incremental
costs.
Commenter
28­
273
said
this
approach
is
the
most
logical
and
equitable
approach,
and
the
exclusive
use
of
incremental
cost
effectiveness
is
also
consistent
with
NSR/
PSD
requirements.
Other
commenters
argued
for
consideration
of
these
costs,
but
would
place
limits
on
their
use.
For
example
commenter
28­
256
said
high
incremental
costs
should
not
be
the
sole
criteria
for
rejection
of
a
particular
control
technology.
In
order
to
justify
elimination
of
a
control
option,
a
demonstration
that
both
the
incremental
and
total
costs
are
excessive
should
be
made.
Commenter
28­
302
said
consideration
of
incremental
cost
effectiveness
should
only
be
allowed
when
it
does
not
result
in
reduced
stringency
of
controls
over
those
determined
using
absolute
cost
effectiveness
alone.
Commenter
28­
7008
rejected
the
use
of
incremental
cost
analysis
saying
that
it
is
most
appropriate
to
consider
the
cost
effectiveness
numbers
alone.
Incremental
cost
effectiveness
can
vary
greatly
depending
on
the
number
of
interim
options
defined,
thereby
creating
the
potential
for
inconsistencies
in
BART
determinations
regionally.
171
Commenter
28­
264
said
the
incremental
analysis
must
include
a
mechanism
to
evaluate
the
visibility
benefits
that
would
be
achieved
by
the
installation
of
control
technology,
and
then
adjust
the
expectations
for
control
based
on
the
expected
benefits.

Commenter
28­
255
said
EPA
should
publish
and
update
as
necessary
upper
limits
for
cost
effectiveness
and
incremental
costs
for
all
pollutants.
This
will
ensure
consistency
in
BART
analyses.

Response:

We
believe
incremental
cost
analysis
is
an
important
and
useful
measure
and
the
guidelines
indicated
it
should
be
used.
We
do
not
dictate
specific
cost
cutoffs
that
States
should
set
that
would
be
helpful
in
all
circumstances.

Comment:
Two
commenters
addressed
the
role
of
the
OAQPS
Cost
Manual
(
28­
266,
28­
391).
Commenter
28­
266
said
to
ensure
regional
and
national
consistency,
EPA
should
update
various
guidance,
such
as
the
1980
BART
Guidance
and
the
1996
OAQPS
Control
Cost
Manual.
Commenter
28­
391
noted
that
the
manual
does
not
account
for
all
of
the
direct
and
collateral
costs
that
will
be
associated
with
BART
and
should
not
be
used
as
the
sole
basis
for
cost
estimations.
The
commenter
recommended
that
rather
than
specifying
a
specific
cost
estimation
methodology,
the
EPA
should
identify
the
elements
of
an
acceptable
cost
analysis,
recognizing
that
the
added
costs
for
retrofit
will
differ
at
each
location.

Response:

Other
commenters
(
28­
255,
28­
258)
cited
the
need
for
specific
criteria
in
the
guidelines
for
evaluating
the
costs
of
controls,
compliance
to
improve
implementation
and
consistency.
Commenter
28­
358
added
that
the
guidelines
should
specify
the
level
of
inflation­
adjusted
costs
for
which
the
presumptive
BART
must
be
met.

Response:

We
believe
Congress
intended
that
States
have
flexibility
in
weighing
the
factors.

4.2.3
Presumptive
Emission
Limits
4.2.3.1
SO
2
Limits
for
Utility
Boilers
Comment:
Thirty­
four
commenters
said
the
proposed
SO
2
levels
are
reasonable
(
0167,
0171,
0227,
0231,
0232,
0234,
0255,
0256,
0267,
0271,
0314,
0323,
28­
239,
28­
345,
28­
247,
28­
248,
28­
249,
28­
250,
28­
258,
28­
260,
28­
262,
28­
286,
28­
289,
28­
302,
28­
334,
28­
338,
28­
350,
28­
352,
28­
172
362,
28­
399,
28­
433,
28­
7002,
28­
7016,
28­
7048)
citing
examples
and
references
in
support
of
these
levels
(
0171,
0255,
0256,
28­
302).
Commenter
0323
also
agreed
with
EPA's
approach
of
either
a
specific
emission
rate
limit
or
level
of
control
for
sulfur
dioxide
that
helps
to
provide
equity
among
sources
which
use
higher
sulfur
coal
and
those
using
coal
with
inherently
lower
sulfur
content.

Response:
The
EPA
appreciates
these
comments.

Comment:
Commenters
(
28­
248,
28­
399,
28­
7016)
said
EPA
should
expand
this
presumption
by
applying
it
to
electric
generating
facilities
that
are
under
controlled
for
purposes
of
SO
x.
Commenter
28­
258
said
the
presumptive
BART
level
of
SO
2
controls
should
be
required
at
all
utility
boilers,
not
just
those
with
no
SO
2
controls.

Response:
The
final
BART
presumption
for
coal­
fired
EGUs
with
a
pre­
existing
SO2
controls
are
to
be
determined
by
States
based
on
an
analysis
of
cost­
effective
FGD
upgrades
designed
to
improve
the
average
SO2
removal
efficiency
of
the
existing
system.
The
EPA
is
strongly
encouraging
States
to
the
implement
the
BART
presumptive
limits
on
coal­
fired
EGUs
greater
than
200
MW
and
on
all
oil­
fired
EGUs
regardless
of
capacity.
See
TSD
"
Setting
BART
Presumptive
SO2
Limits
for
Electric
Generating
Units"
for
a
more
detailed
discussion
on
how
EPA
determined
its
final
recommendations.

Comment:
Nine
commenters
said
the
proposed
levels
are
too
stringent
(
0176,
0181,
0217,
0246,
0253,
28­
264,
28­
341,
28­
357,
28­
7007).
Commenter
28­
264
said
the
level
is
too
high
for
western
utilities
that
are
burning
low­
sulfur
coal
where
a
level
of
85­
90%
would
be
more
appropriate.
One
commenter
(
0176)
cited
his
experience
with
eastern
bituminous
waste
and
did
not
believe
the
proposed
emissions
reductions
were
possible.
Commenter
0253
said
any
default
SO
2
control
level
mandating
greater
than
90%
control
are
inappropriate
for
units
utilizing
wet
scrubbers
and
high
sulfur
coals.
Commenter
28­
7007
suggested
a
range
of
70
to
90%
would
be
appropriate.

Response:
The
EPA
is
finalizing
the
SO2
BART
presumptive
limits
based
on
current
FGD
capabilities
given
the
various
combinations
of
coal
types
and
sulfur
contents
inherent
in
the
BART­
eligible
population
of
units.
The
EPA
is
maintaining
both
an
emission
limit
and
percent
removal
in
the
final
BART
presumptive
limits
to
provide
equity
among
various
unit
specific
characteristics
and
the
removal
capabilities
of
different
FGD
retrofit
systems.
We
also
believe
that
units
burning
higher
sulfur
coals
can
utilize
wet
scrubbers
systems
to
achieve
removal
efficiencies
of
at
least
95
percent
and
units
burning
lower
sulfur
coals
can
utilize
semi­
dry
FGD
systems
to
achieve
an
emission
rate
limit
of
0.15
lb
SO2/
mmBtu
or
less.
173
Comment:
Commenters
(
0217,
0246,
28­
357)
said
if
presumptive
levels
are
included,
consideration
should
be
given
to
the
levels
reflected
in
the
WRAP
agreement
(
85%
control
efficiency
at
85%
capacity
factor.)
EPA's
presumption
of
90
to
95%
control
efficiency
contradicts
the
consensus
based
level
of
SO
2
reduction
determined
by
the
WRAP
found
to
be
necessary
for
visibility
improvement
in
western
states.
Moreover,
given
that
the
WRAP
stakeholder
agreement
was
determined
to
be
better
than
BART,
the
proposed
guidelines
should
not
be
interpreted
to
require
changes
in
Section
309
SIPs
already
submitted
under
the
regional
haze
rules.

Response:
In
assessing
whether
the
WRAP
agreement
was
better
than
BART,
EPA
noted
that
for
utility
boilers
that
are
currently
uncontrolled,
emissions
reductions
of
90
percent
or
better
are
readily
achievable.
EPA's
conclusion
that
the
WRAP
agreement
(
which
assumed
85%
control
at
sources
that
are
currently
uncontrolled)
would
achieve
greater
reasonable
progress
than
would
installation
of
BART
was
based
on
the
consideration
of
a
number
of
factors
as
explained
in
that
rulemaking.

Comment:
Commenter
0181
asked
EPA
to
remove
the
presumptive
SO
2
control
level
from
the
final
BART
guidelines.
Instead,
there
should
be
a
BART
determination
process
that
promotes
costeffective
emission
reductions
based
on
specific
boiler
design
features
and
site­
specific
constraints.
The
proposed
presumptive
BART
approach,
as
proposed,
does
not
provide
affected
states
and
sources
with
the
flexibility
needed
to
design
and
install
SO
2
emission
control
systems
in
a
costeffective
manner
to
meet
the
BART
requirements.
Furthermore,
the
presumptive
SO
2
emissions
control
approach
inappropriately
ignores
the
need
for
a
visibility
impact
evaluation
which
is
required
in
step
5
of
the
proposed
case­
by­
case
BART
engineering
analysis.

Response:
The
EPA
believes
the
final
BART
presumptive
limits
promote
highly
cost­
effective
SO2
emission
reductions.
The
EPA
is
maintaining
both
an
emission
limit
and
percent
removal
in
the
final
SO2
BART
presumptive
limits
to
provide
equity
among
unit
specific
characteristics
(
i.
e.
boiler
design,
coal
types,
sulfur
content,
unit
size,
etc.)
and
potential
FGD
retrofit
systems.
We
believe
our
SO2
presumptive
limits
provide
States
enough
flexibility
to
take
into
account
unit
specific
circumstances.
States
have
the
ability
to
establish
alternate
BART
presumptive
limits
based
on
a
five­
step
BART
analysis
given
the
presence
of
persuasive
evidence.
See
Section
7
part
a,
Visibility
Analysis
for
SO2
and
NOx
emissions
from
EGUs,
in
BART
Preamble.

Comment:
Commenter
28­
341
said
through
switching
to
Powder
River
Basin
(
PRB)
coal,
many
of
Ameren's
facilities
have
met
and
exceeded
the
SO
2
reduction
requirements
of
the
Acid
Rain
program.
Should
BART
guidelines
be
adopted,
however,
they
will
be
forced
to
install
scrubbers.
174
at
each
one
of
their
facilities
that
employs
the
use
of
low­
sulfur
coal
to
meet
the
proposed
90­
95%
control
levels.

Response:
The
EPA
anticipates
States
to
adopt
the
BART
presumptions
on
all
coal­
fired
BARTeligible
EGUs
greater
than
200
MW.
Analysis
conducted
by
EPA
suggest
that
units
burning
lower
sulfur
coals
and
PRB
coal
can
cost­
effectively
retrofit
with
FGD
systems
to
meet
the
presumptive
limits.

Comment:
Four
commenters
said
the
proposed
presumptive
control
range
is
not
achievable
for
existing
boilers,
and
is
even
more
stringent
than
NSPS
limits
for
new
sources
(
28­
403,
28­
404,
28­
405,
28­
7013).
Commenter
28­
403
said
EPA's
data
do
not
support
such
a
limit
nor
does
EPA
explain
why
the
presumption
is
needed
and
why
the
states'
discretion
should
be
preempted.
Commenter
28­
404
said
this
presumption
will
result
in
excessive
costs
for
older
equipment.
Commenter
28­
405
said
if
presumptive
control
levels
were
to
be
established,
they
should
start
at
the
minimum
level
of
emission
controls,
with
additional
controls
justified
on
the
basis
of
the
source­
specific
balancing
process
Congress
set
forth
in
169A.
Commenter
28­
7013
added
that
the
presumptive
limits
are
more
consistent
with
BACT,
which
is
contrary
is
contrary
to
the
Act's
clear
distinction
between
BART
and
BACT.

Response:
The
EPA
has
recently
proposed
to
revise
the
NSPS
to
represent
the
SO2
control
capabilities
which
are
available
today.
In
the
proposed
NSPS
amendments
EPA
proposed
a
single
limit
without
a
compliance
alternative,
such
as
a
percent
removal
as
allowed
in
BART.
In
setting
the
NSPS
EPA
set
a
higher
emission
limit
than
proposed
in
BART
to
account
for
account
for
the
need
of
future
fuel
diversity,
in
particular
higher
sulfur
coals.
Conversely,
the
BART
presumptions
were
established
based
on
the
defined
population
of
eligible
units
and
were
determined
to
be
achievable
by
all
eligible
units.
The
EPA
would
also
like
to
point
out
the
BART
SO2
limits
are
more
stringent
than
the
NSPS
for
about
90
percent
of
coals.

Comment:
Eleven
commenters
said
that
the
proposed
presumptive
limits
clearly
exceed
the
statutory
authority
(
28­
270,
28­
342,
28­
353,
28­
354,
28­
356,
28­
360,
28­
386,
28­
390,
28­
392,
28­
778,
28­
7004).
Commenters
(
28­
270,
28­
386,
28­
392,
28­
778)
said
that
the
evaluation
must
be
conducted
according
to
the
statutory
process
and
BART
must
be
determined
on
a
case­
by­
case
basis.
Commenters
(
28­
342,
28­
354,
28­
356,
28­
360,
28­
390)
objected
to
EPA's
proposed
presumptive
limits
for
SO
2
controls
as
a
flagrant
example
of
EPA
usurping
State
authority
and/
or
as
being
arbitrary
and
economically
counterproductive.
Commenter
28­
7004
added
that
the
Act
consciously
addresses
the
need
for
SO
2
scrubbing
in
Title
IV.
The
Act
does
not
envision
or
authorize
a
much
more
onerous
and
expensive
nationwide
SO
2
scrubbing
program
as
part
of
Section
169A.
The
"
guidance"
which
presumes
90
to
95%
SO
2
removal
as
BART
for
uncontrolled
utility
boilers
is
not
supported
by
the
law.
175
Response:
The
CAA
requires
BART
limits
for
certain
power
plants
to
be
determined
pursuant
to
guidelines
issued
by
EPA.
States
have
the
ability
to
take
into
account
the
specific
characteristics
of
individual
sources
in
making
final
BART
determinations.

Comment:
Three
commenters
said
the
proposed
levels
are
too
lenient
(
0221,
28­
263,
28­
7008).
Commenters
(
0221,
28­
263)
said
because
of
the
extended
SIP
timeline,
BART
retrofits
will
not
be
applied
until
2013
and
therefore,
it
is
expected
that
the
"
best"
available
sulfur
removal
technologies
will
routinely
be
able
to
achieve
in
excess
of
95%
removal
efficiencies
and
emission
rates
of
lower
than
0.10
lb/
MMBTU
by
that
time.
Therefore,
the
presumptive
level
of
95%
reduction
or
0.10­
0.15
lb/
MMBTU
SO2
emission
rate
that
is
stipulated
in
the
proposal
should
be
strengthened
to
an
SO
2
removal
efficiency
of
95%
or
better
or
an
emission
rate
of
0.10
lb/
MMBTU
or
better.
Commenter
28­
7008
was
concerned
with
the
effect
that
setting
presumptive
minimums
would
have
in
a
true
top­
down
control
technology
evaluation.
Using
an
artificially
low
presumptive
minimum
could
eliminate
consideration
of
more
effective
control
options
that
are
still
achievable
by
a
given
source.

Response:
The
EPA
believes
the
final
BART
SO2
presumptive
limits
represent
the
removal
efficiencies
of
current
FGD
systems
given
the
identified
universe
of
BART­
eligible
sources.
The
final
BART
presumptive
limits
require
a
removal
efficiency
of
at
least
95
percent
or
control
to
an
emission
rate
of
0.15
lb
SO2/
mmBtu
or
less.
States
have
discretion
to
set
more
stringent
presumptive
limits
based
on
the
findings
of
the
five­
step
BART
analysis
process.

Comment:
Thirteen
commenters
specified
that
proposed
control
level
must
account
for
fuel
type
and/
or
coal
sulfur
content
(
0171,
0187,
0206,
0213,
0224,
0236,
0241,
0255,
0303,
0314,
28­
335,
28­
347,
28­
770).
Commenter
0213
presented
historical
data
showing
that
the
proposed
emissions
limit
could
not
be
attained
in
certain
cases
depending
upon
the
sulfur
content
of
the
fuel
being
burned.
A
statistical
approach
should
be
utilized
analyzing
the
performance
of
existing
FGD
installations
to
ensure
that
the
worst
case
performers
can
reasonably
be
expected
to
comply
with
the
limit.
Commenters
(
28­
335,
28­
347,
28­
770)
said
the
presumption
should
not
apply
to
oil
or
gas­
fired
utility
boilers.

Response:
The
EPA
agrees
with
the
commenters
and
reiterates
that
in
establishing
the
BART
presumptive
limits
EPA
accounted
for
variations
in
fuel
type,
fuel
sulfur
content,
existing
SO2
control
systems,
anticipated
unit
cost­
effectiveness
values
and
the
removal
capabilities
of
current
FGD
systems.
The
EPA
is
finalizing
BART
presumptive
limits
which
include
an
emission
rate
limit
and
a
percent
removal
requirement
which
it
feels
to
be
reasonably
achievable
by
all
coal­
fired
BART­
eligible
units
absent
extreme
circumstance.
The
EPA
is
finalizing
SO2
BART
presumptive
limits
for
oil­
fired
EGUs
but
not
gas­
fired
EGUs.
176
Comment:
Commenters
(
0187,
0226,
0224,
0236,
0241,
0303)
said
the
rationale
used
to
justify
the
default
and
presumptive
BART
SO
2
emission
limitations
for
coal­
fired
units
can
be
applied
to
oilfired
EGUs.
In
contrast
to
coal­
fired
electric
generating
units,
flue
gas
de­
sulfurization
has
not
historically
been
used
at
oil­
fired
facilities
to
comply
with
SO
2
emission
limitations.
To
the
commenter's
knowledge,
there
are
no
existing
oil­
fired
steam
electric
generating
units
in
the
U.
S.
equipped
with
flue
gas
de­
sulfurization
for
control
of
SO
2
emissions.
The
primary
reason
for
this
is
that
the
vast
majority,
if
not
all,
of
the
states
limit
the
sulfur
content
of
the
oil
that
may
be
burned
at
EGUs
as
part
of
their
SIP
strategy
for
complying
with
the
SO
2
NAAQS.

Response:
Analyses
conducted
by
the
EPA
in
determining
the
presumptive
BART
limitations
for
coaland
oil­
fired
EGUs
were
based
on
emissions
data,
anticipated
unit
cost­
effectiveness
values
and
current
capabilities
of
SO2
control
technologies
or
methods
and
their
industry
applicability.
The
EPA
agrees
with
the
commenters
suggestion
pertaining
oil­
fired
units
and
notes
the
final
BART
presumptions
for
oil­
fired
units
is
a
limitation
on
fuel
oil
sulfur
content.

Comment:
Commenter
0314
suggested
setting
the
presumptive
BART
as
a
range
(
0.1
to
0.15
lb/
MMBTU)
allowing
the
State
to
establish
a
specific
level
for
each
plant.

Response:
The
final
BART
presumptive
limits
are
to
be
established
on
a
unit­
by­
unit
basis
by
the
States.
States
have
the
ability
to
establish
a
SO2
removal
efficiency
of
at
least
95
percent
or
an
emission
rate
limit
of
0.15
lb
SO2/
mmBtu
or
less.

Comment:
Three
commenters
said
that
presumptive
controls
may
need
to
differ
depending
on
the
location
of
the
source
(
0244,
0252,
0306).
Two
commenters
(
0244,
0252)
said
the
presumptive
control
level
of
90%
for
western
units
is
or
may
be
unrealistic.
In
contrast
to
the
presumptive
level
proposed
by
EPA
of
90%
SO
2
reduction,
WRAP
has
conducted
analysis
that
the
presumptive
level
of
control
for
a
scrubber
in
the
west
should
be
set
at
85%.
Commenter
0306
said
EPA
needs
to
consider
the
cold
climate
of
certain
northern
states
when
determining
presumptive
BART
for
SO
2.
Cold
weather
can
cause
icing
of
very
wet
stacks
which
can
lead
to
safety
hazards
and
functional
problems.
Reheat
of
the
stack
may
be
necessary,
to
reduce
or
eliminate
icing
problems.
Reheating
the
flue
gas
by
bypassing
some
of
the
gas
around
the
scrubber
will
reduce
overall
SO
2
removal.
The
proposed
presumptive
BART
may
not
be
achievable
in
cold
climates
without
overly
burdensome
costs.

Response:

Investigations
conducted
by
EPA
show
that
a
large
number
of
power
plants
in
the
US
are
equipped
with
flue
gas
desulfurization
(
FGD)
systems
without
any
flue
gas
reheat.
A
1996
EIA
database
177
shows
over
100
FGD
system
installations
that
have
been
operating
for
many
years
with
no
flue
gas
reheat.
(
EIA­
767
Database
Annual
Steam­
Electric
Plant
Operation
and
Design
Data
for
1996,
http://
www.
eia.
doe.
gov/
cneaf/
electricity/
page/
eia767hist.
html.)
Several
of
these
installations
are
located
in
states
with
a
cold
climate,
such
as
in
Utah,
North
Dakota,
Pennsylvania,
Ohio,
Indiana,
and
Illinois.
Also,
many
of
these
installations
are
removing
SO
2
at
rates
exceeding
90
percent.
It
should
also
be
noted
that
flue
gas
bypass
is
not
the
only
means
available
to
provide
flue
gas
reheat
at
the
outlet
of
a
scrubber.
Steam
or
electric
heaters
can
also
be
used
for
this
purpose,
without
a
need
to
bypass
a
portion
of
the
flue
gas
at
the
inlet
of
a
scrubber.
Therefore,
EPA
cannot
agree
with
the
commenters
that
the
FGD
installations
in
the
west
would
have
to
be
designed
with
a
gas
bypass
to
provide
flue
gas
reheat.
.

Comment:
Nine
commenters
said
that
the
proposed
control
levels
must
account
for
the
type
of
scrubber
in
place
(
0171,
0210,
0228,
0229,
0241,
0250,
0253,
0254,
0256).
Commenter
described
the
need
for
site­
specific
analyses
to
address
scrubbers
in
place
(
0210,
250).
Commenters
(
0241,
0254)
added
that
EGUs
that
currently
have
SO
2
controls
should
not
be
forced
to
meet
the
presumptive
level
unless
a
full
BART
analysis
is
undertaken
which
shows
that
incremental
visibility
and
cost­
effectiveness
justify
forcing
controlled
EGUs
to
meet
the
more
stringent
limit.
Others
(
0229)
said
that
the
recent
addition
of
controls,
even
if
reductions
fall
below
the
presumptive
levels,
should
be
cause
for
allowing
states
to
screen
such
a
BART­
eligible
source
out
early
in
the
evaluation
process.
Commenter
0254
said
the
ability
to
upgrade
existing
scrubber
units
will
be
limited
and
for
many,
particularly
dry
scrubbers,
the
presumptive
95%
reduction
level
will
not
be
achievable.
In
addition,
many
of
these
units
have
been
installed
recently,
e.
g.,
since
the
1990
Clean
Air
Act
Amendments,
and
requiring
these
facilities
to
re­
visit
large
investments
in
controls,
after
such
a
short
interval,
imparts
an
excessive
burden
on
those
businesses.

Response:
The
BART
presumptions
for
EGUs
with
pre­
existing
SO2
controls
are
to
be
determined
by
the
States,
as
either
an
emission
rate
limit
or
a
percent
removal,
based
on
a
unit­
specific
analysis
of
potential
system
upgrades
designed
to
increase
the
average
SO2
removal
capability
of
the
preexisting
system.
As
the
CAA
identifies
"
any
existing
pollution
control
technology
in
use"
at
a
source
as
one
of
the
factors
to
consider
in
a
BART
determination,
States
are
likely
to
account
for
existing
controls
in
establishing
limits.
In
addition,
the
guidelines
suggest
that
States
may
be
able
to
adopt
more
streamlined
approaches
to
making
BART
determinations
where
consideration
of
one
of
the
factors
clearly
outweighs
the
others.
The
EPA
recognizes
that
semi­
dry
FGD
systems
are
not
typically
designed
to
achieve
removal
efficiencies
of
at
least
95
percent.
For
this
reason,
the
EPA
anticipates
smaller
units
burning
low
sulfur
coals
could
utilize
the
semi­
dry
FGD
technologies
to
achieve
an
emission
rate
of
0.15
lb
SO2/
mmBtu.
The
EPA
also
suggests
there
are
cost­
effective
upgrade
options
for
EGUs
with
pre­
existing
semi­
dry
scrubber
systems.

Comment:
178
Commenter
0253
said
that
in
largely
focusing
on
the
capabilities
of
state­
of­
the­
art
scrubbers
installed
on
new
generating
units
and
ignoring
the
feasibility
or,
in
many
cases,
infeasibility
of
retrofitting
such
technology
on
existing
units,
EPA's
analysis
is
substantially
flawed.
EPA
overlooks
the
fact
that,
in
many
instances,
it
is
technically
infeasible
to
upgrade
early
generation
scrubbers
to
meet
the
95%
control
level
that
may
be
achievable
with
current
state­
ofthe
art
scrubbers.

Response:
The
EPA
established
the
BART
presumptive
limits
for
coal­
fired
EGUs
without
existing
SO2
controls
based
on
the
control
levels
achievable
with
current
state­
of­
the­
are
new
and
retrofit
scrubber
installations.
The
BART
presumption
for
EGUs
with
pre­
existing
SO2
controls
are
to
be
established
by
the
States
(
not
necessarily
95
percent
control)
based
on
an
evaluation
of
technically
feasible
upgrade
options
available
for
early
generation
scrubbers
in
order
to
improve
the
average
SO2
removal
of
existing
controls.

Comment:
Commenters
(
0228,
0256)
said
the
SO
2
presumption
should
not
apply
to
sources
that
have
already
installed
and
are
operating
scrubbers,
or
to
sources
that
have
committed
to
specific
levels
of
scrubbing
through
consent
orders
or
agreements
with
EPA
or
a
state
regulatory
agency.
If
a
source
can
demonstrate
a
reduction
in
visibility
impairment
below
the
specified
threshold
(
whether
that
threshold
is
EPA's
currently
proposed
0.5
deciview
or
an
alternative
level)
with
less
stringent
controls,
EPA
and
the
states
should
not
impose
by
default
more
stringent
reduction
requirements.

Response:
The
presumptions
in
the
guidelines
do
not
apply
to
sources
that
have
existing
scrubbers.
Nonetheless,
EPA's
analysis
suggest
cost­
effective
FGD
system
upgrades
designed
to
improve
average
SO2
removal
efficiencies
are
available
for
some
BART­
eligible
units
with
currently
operational
scrubber
systems.

Comment:
In
contrast,
one
commenter
(
0221)
said
that
under­
controlled
sources
(
i.
e.,
those
previously
retrofitted
to
meet
the
1979
NSP
of
70
percent)
can
meet
removal
efficiencies
of
95%
or
more.
The
performance
of
boilers
already
using
the
wet
limestone
process
can
be
improved
through
the
application
of
"
once­
through
wet
FGD
technology,"
such
as
increasing
sorbent,
reactivity
and
other
upgrades.

Response:
EPA
analyses
suggest
cost­
effective
FGD
system
upgrades
designed
to
improve
average
SO2
removal
efficiencies
are
available
for
BART­
eligible
units
with
currently
operational
scrubber
systems.
See
TSD
"
Setting
BART
Presumptive
SO2
Limits
for
Electric
Generating
Units"
for
a
more
detailed
discussion
on
EPA's
list
of
recommended
upgrade
options
for
wet
FGD
systems.
179
Comment:
Eight
commenters
discussed
the
role
that
plant
size
and/
or
capacity
should
play
in
applying
a
default
SO
2
control
level
(
0206,
0227,
0241,
0253,
0256,
0267,
0271,
0303).
Some
commenters
said
that
the
presumptive
limits
should
not
apply
to
units
between
250
and
750
MW
(
0206,
0241,
0303).
Commenter
0253
said
EPA
should
clarify
that
only
BART­
eligible
sources
would
be
counted
toward
the
750­
MW
cut­
off.
Commenters
(
0206,
0303)
added
that
Congress
did
not
specify
similar
requirements
for
smaller
plants
(
i.
e.
facilities
below
750
MWs).
The
installation
of
controls
necessary
to
reduce
SO
2
emissions
by
90
to
95
percent
is
more
likely
to
be
uneconomical
at
smaller
facilities
(
0206,
0253,
0303).
The
BART
process
must
be
followed
for
such
a
determination
to
be
made
(
0241).

Response:
The
BART
presumptions
are
binding
for
units
greater
than
750
MW.
EPA
strongly
encourages
States
to
apply
the
BART
presumptive
limits
to
units
greater
than
200
MW
regardless
of
plant
size.
The
EPA's
unit­
specific
economic
analysis
for
BART
eligible
coal­
fired
units
suggests
on
average
it
is
less
economical
for
units
less
than
200
MW
to
retrofit
with
FGD
systems
than
those
units
greater
than
200
MW.
However,
the
analysis
did
demonstrate
that
some
BART­
eligible
units
less
than
200
MW
were
modeled
to
have
cost
effectiveness
values
comparable
to
those
unit
greater
than
200
MW
and
that
States
are
encouraged
to
apply
the
BART
presumptions
to
such
sources.

Comment:
Commenter
0256
said
EPA's
proposed
distinction
between
the
SO
2
"
default
requirement"
and
the
"
rebuttable
presumption"
based
on
total
power
plant
capacity
is
unsupportable.
Because
BART
applies
on
a
unit
basis,
the
fact
that
the
total
capacity
of
the
plant
is
higher
or
lower
than
750
MW
has
no
obvious
bearing
on
whether
an
individual
unit
can
meet
the
default/
presumption.
Indeed,
if
anything,
units
at
a
larger
plant
would
perhaps
have
more
problems
meeting
the
presumption
than
would
units
at
a
smaller
plant,
because
a
larger
plant
would
tend
to
have
more
units
and
thus
more
potential
for
space
constraints
in
retrofitting
add­
on
controls.
If
EPA
retains
the
SO
2
presumption,
it
should
be
a
rebuttable
presumption
for
all
units
greater
than
250
MW,
regardless
of
total
power
plant
capacity.

In
contrast,
commenters
(
0227,
0271)
said
EPA
should
require
presumptive
controls
for
all
units
at
750
MW
plants,
including
those
between
250
and
750
MW.
Commenter
0271
said
this
conforms
to
the
specific
wording
of
the
Act
(
Sec.
169A(
b)(
2)(
A))
for
controls
on
these
smaller
than
750
MW
units.
Commenter
0267
supported
applying
the
proposed
presumptive
level
to
all
EGUs
greater
than
250
MW.

Response:
The
BART
guidelines
establish
a
presumption
for
uncontrolled
units
200
MW
or
greater
at
power
plants
in
excess
of
750
MW.
The
rationale
for
addressing
these
units
is
set
forth
in
the
relevant
TSD.
180
Comment:
Commenter
0268
noted
that
Arizona
has
chosen
to
address
visibility
impairment
through
Section
309
of
the
1999
RHR.
In
order
to
pursue
the
309
pathway,
Arizona
has
made
a
demonstration
to
EPA
that
the
state
will
realize
overall
SO
2
reductions
greater
than
the
reductions
that
would
be
achieved
solely
through
imposition
of
BART
controls
at
the
state's
BART­
eligible
sources.
EPA
should
clearly
and
definitively
recognize
that
presumptive
BART
limits
for
SO
2
do
not
apply
to
BART­
eligible
sources
located
in
states
that
have
already
committed
to
"
better
than
BART"
SO
2
reductions
through
Section
309
SIP
submittals.

Response:
The
regional
haze
rule
allows
States
to
adopt
alternative
measures
in
lieu
of
BART.

Comment:
Twelve
commenters
suggested
alternatives
to
presumptive
control
levels
(
0213,
0215,
0256,
0302,
28­
255,
28­
256,
28­
264,
28­
266,
28­
278,
28­
312,
28­
355,
28­
404).
Commenter
0213
supports
an
emissions
trading
program
to
implement
BART
controls
as
proposed
in
the
June
10,
2004
Clean
Air
Interstate
Rule
(
CAIR).
However
if
an
individual
unit
program
is
prescribed,
the
commenter
supports
a
percent
removal
limit
as
opposed
to
an
emission
limit.

Response:
EPA
appreciates
these
comments
and
supports
BART­
eligible
units
option
to
opt
into
the
CAIR
in
applicable
states.

Comment:
Commenter
0215
said
EPA
should
retain
the
option
of
complying
with
the
SO
2
default/
presumption
for
EGUs,
either
by
adhering
to
the
straight
percentage
reduction
or
a
set
emission
rate.
Due
to
the
variability
in
coal
and
equipment
types
(
including
vintages),
EPA
must
retain
flexibility
for
those
EGUs
that
must
comply
with
these
BART
requirements.

Response:
The
EPA
appreciates
this
comment.

Comment:
Commenter
0256
agrees
with
EPA
that,
for
units
burning
lower
sulfur
coal,
achieving
95%
reduction
could
be
problematic
and
that
an
emission
limitation
should
be
used.
Other
commenters
(
28­
256,
28­
264,
28­
266,
28­
278,
28­
312,
28­
355,
28­
404)
agreed
if
a
presumptive
approach
is
used,
then
the
presumptive
norms
should
be
explicitly
stated
in
the
rulemaking
and
should
be
no
higher
than
3.0
lb/
MWh
for
SO
2
(
28­
256),
based
on
a
weight
rate
limit
or
other
fuel­
neutral
basis
to
account
for
facilities
burning
higher
sulfur
coal
(
28­
264,
28­
335),
on
an
emissions
standard
to
allow
sources
to
take
credit
for
improvements
they
have
already
made
(
28­
266,
28­
278,
28­
312,
28­
404).

Response:
181
The
EPA
conducted
a
unit­
specific
analysis
in
determining
the
BART
presumptive
limits
and
is
finalizing
presumptions
which
it
feels
to
be
achievable
by
the
BART­
eligible
population,
absent
extreme
circumstances.
The
final
BART
presumptive
limits
for
coal­
fired
EGUs
without
existing
SO2
controls
includes
both
an
emission
rate
limit
and
percent
reduction
requirement
to
provide
equity
among
the
units
and
their
various
characteristics
Comment:
Commenter
0302
prefers
the
flexibility
of
a
percentage
control
level
comparable
to
a
performance
level
of
0.1
to
0.15
lbs/
MMBTU
SO
2.
A
percentage
control
level
would
offer
the
non­
EGU
BART­
subject
sources
the
best
controls
that
are
achievable,
for
all
types
of
sources,
under
reasonable
cost
effectiveness.
Without
the
proper
guidance
for
BART­
subject
sources
that
are
EGUs,
the
commenter
needs
flexibility
in
determining
the
actual
emission
rates
at
which
these
sources
will
be
required
to
demonstrate
compliance.

Response:
States
will
make
BART
determinations
based
on
a
consideration
of
the
five
factors
set
forth
in
the
CAA.
The
presumptions
in
the
guidelines
apply
only
to
EGUs.

Comment:
Two
commenters
(
28­
255,
28­
264)
said
the
presumptive
levels
should
not
be
established
as
absolute
values.
Commenter
28­
255
said
the
values
should
be
presented
as
a
range.
Commenter
28­
264
said
the
presumption
should
not
interfere
with
the
states'
discretion
and
statutory
process.
If
EPA
intends
to
apply
this
"
presumptive
level"
as
a
mandate,
then
EPA
should
establish
it
as
a
national
rule
rather
than
continuing
the
pretense
that
states
are
making
a
case­
by­
case
determination.

Response:
Congress
stated
that
BART
emission
limits
should
be
determined
pursuant
to
guidelines
issued
by
EPA.
As
such,
the
guidelines
do
not
interfere
with
the
discretion
given
to
States
by
Congress.
The
guidelines
are
being
promulgated
through
rulemaking
Comment:
Three
commenters
discussed
the
role
that
the
averaging
period
plays
in
determining
performance
(
0213,
0256,
0314).
Commenter
0213
noted
the
difference
in
performance
over
longer
averaging
times
between
state­
of­
the­
art
technology
and
existing
FGD
installations.
EPA
should
review
existing
installations
to
determine
an
appropriate
level
of
performance
for
determining
an
SO
2
removal
efficiency
limit.
Commenter
0256
said
the
choice
of
0.15
lb/
MMBtu
is
predicated
on
the
use
of
at
least
an
averaging
time
of
30­
day
rolling
or
longer.
The
0.15
lb/
MMBtu
limit
could
not
be
achieved
with
the
use
of
a
shorter
averaging
time.
Commenter
0314
asked
that
an
averaging
period,
i.
e.,
a
30­
day
rolling
average,
for
the
control
levels
be
established
in
the
rule.
A
longer
averaging
period
will
allow
for
lower
annual
emissions.
182
Response:
The
EPA
recognizes
the
correlation
between
sustainable
FGD
performance
and
length
of
averaging
times.
For
this
reason
it
is
establishing
as
part
of
the
BART
rulemaking
a
30­
day
rolling
averaging
period
for
the
SO2
presumptive
limits.

Comment:
Four
commenters
said
that
EPA
has
underestimated
the
cost
per
ton
of
emissions
removed
(
0187,
0206,
0236,
0303).
Commenters
(
0187,
0236)
provided
an
example
of
the
costeffectiveness
range
for
reducing
SO
2
emissions
to
0.1
to
0.15
lbs/
MMBTU
estimated
to
be
between
$
1800
and
$
3400
per
ton
of
SO
2
removed
when
burning
Powder
River
Basin
coals.
This
is
much
higher
than
the
$
200­$
1000
per
ton
cost­
effectiveness
range
cited
by
EPA
in
support
of
the
proposed
presumptive
BART
SO
2
limit.
Commenter
0187
gave
an
example
of
an
oil­
fired
unit
exceeding
$
3,000
per
ton
of
SO
2
removed,
which
supports
a
separate
presumptive
SO
2
emission
limit
for
oil­
based
units
based
on
the
maximum
fuel
sulfur
content
in
oil.
Commenters
(
0206,
0303)
were
concerned
about
additional
costs
faced
by
smaller
plants
(
upgrading
fans,
piping,
etc.)
that
can
drive
the
cost
into
the
$
3,000
range.
The
commenters
said
EPA
should
treat
smaller
plants
like
other
major
sources
and
allow
states
greater
flexibility
in
determining
BART
for
such
facilities.

Response:
EPA's
unit
specific
economic
analysis
suggests
an
average
cost
effectiveness,
for
all
BART
units
greater
than
200
MW,
of
$
919
per
ton
SO2
removed,
with
a
majority
of
the
units
anticipated
to
have
cost
effectiveness
values
ranging
from
$
400
to
$
2000
per
ton
SO2
removed,
see
TSD
"
Setting
BART
Presumptive
SO2
Limits
for
Electric
Generating
Units"
for
a
more
detailed
discussion
on
how
EPA
determined
its
final
recommendations.
EPA's
economic
analysis
also
suggested
that
some
unit
greater
than
200
MW
could
experience
expensive
cost
effectiveness
values,
however
States
have
the
flexibility
to
determine
alternate
BART
presumptive
limits
based
on
the
results
of
the
5
part
BART
analysis.
The
EPA
strongly
encourages
States
to
apply
the
BART
presumption
at
units
greater
than
200
MW
because
our
economic
analysis
suggested
higher
average
cost
effectiveness
values
for
units
less
than
200
MW.
However,
EPA
also
recognizes
that
some
units
less
than
200
MW
were
modeled
to
have
reasonable
cost
effectiveness
values.
The
EPA
determined
the
final
BART
presumption
for
oil­
fired
units
to
be
a
limitation
of
fuel
oil
sulfur
content.

Comment:
In
contrast,
commenters
(
0221,
0267)
said
achieving
the
control
levels
are
cost
effective.

Response:
The
EPA
appreciates
these
comments.

Comment:
Six
commenters
provided
other
comments
regarding
SO
2
limits
for
utility
boilers
(
0215,
0224,
0241,
0250,
0255,
0319).
Commenter
0215
said
if
EPA
retains
the
SO
2
183
default/
presumption,
it
should
make
it
clear
that
if
any
source
is
willing
to
achieve
95%
control
or
somewhere
along
the
scale
of
0.10
to
0.15
lb/
MMBtu,
it
should
automatically
be
exempt
from
the
requirement
to
perform
a
BART
analysis
for
SO
2.
Commenter
0224
said
EPA
might
specify
flue
gas
desulfurization
as
a
presumptive
technology.
Commenter
0255
described
several
options
that
are
available
to
further
reduce
SO
2
emissions
and
improve
the
reliability
of
existing
FGD
systems
to
above
95
%
removal
efficiencies.

Response:
We
agree
that
under
some
circumstances,
a
State
may
reasonably
find
that
if
a
source
is
willing
to
meet
the
presumptive
limit,
it
need
not
undertake
a
BART
analysis
for
SO
2.
Our
presumptive
limits
for
controlling
SO
2
from
coal­
fired
power
plants
was
based
on
flue
gas
desulfurization,
but
States
may
determine
that
other
technologies
are
appropriate.

Comment:
Commenters
(
0241,
0250)
opposed
establishing
any
SO
2
control
level
for
EGUs
as
presumptive
BART
because
this
usurps
state
authority
and
completely
ignores
the
BART
evaluation
process
required
by
the
Act.
Commenter
0250
said
this
is
especially
inappropriate
for
units
already
equipped
with
scrubbers.

Response:
The
guidelines
do
not
establish
presumptions
for
units
already
equipped
with
scrubbers.

Comment:
Commenter
0319
was
concerned
that
setting
an
emission
rate
goal
incorporating
cap
and
trade
mechanisms
rather
than
a
percent
control
goal
would
encourage
fuel
switching
to
SPRB
coal.
Even
plants
that
meet
the
BART
standards
would
be
encouraged
to
switch
to
SPRB
coal
to
avoid
capital
investment
or
to
profit
from
the
sale
of
surplus
allowances.
Tribal
coal
and
lignite
production
could
be
put
at
risk.
The
commenter
provided
a
proposal
to
address
the
impacts
on
tribal
coal
as
Attachment
A.

Response:
The
final
BART
presumptions
do
not
include
a
cap
and
trade
program
for
EGUs.
EPA
believes
the
BART
presumptive
limits
for
EGUs
are
stringent
enough
that
fuel
switching
alone
will
not
bring
a
unit
into
compliance
with
the
0.15
lb
SO2/
mmBtu
emission
rate
limit.
The
final
BART
presumptions
for
coal­
fired
EGUs
without
pre­
existing
SO2
controls
were
finalized
based
on
what
EPA
believes
are
the
current
capabilities
of
wet
and
semi­
dry
FGD
systems.

4.2.3.2
NO
x
Limits
for
Utility
Boilers
Comment:
184
Eighteen
commenters
on
the
2001
notice
supported
presumptive
limits
for
NO
x
set
at
90
to
95%
(
28­
245,
28­
247,
28­
248,
28­
255,
28­
260,
28­
262,
28­
302,
28­
325,
28­
334,
28­
338,
28­
350,
28­
352,
28­
358,
28­
362,
28­
390,
28­
399,
28­
433,
28­
7016).
Commenter
28­
302
said
the
combination
of
low­
NOx
burner
(
LNB)
technology
and
SCR
controls
can
routinely
reduce
NO
x
emissions
by
over
90%
and
other
commenters
cited
research
documenting
the
performance
of
controls
(
28­
325,
28­
390)
and
EPA's
own
experience
and
data
(
28­
28­
352,
28­
362,
28­
7016).

Commenter
28­
256
said
if
a
presumptive
norm
approach
is
used,
then
the
presumptive
norms
should
be
explicitly
stated
in
the
rulemaking
and
should
be
no
higher
than
1.5
lb/
MWh
for
NO
x..

Comment:
Nine
commenters
addressed
whether
the
NO
x
levels
proposed
in
2004
are
too
stringent
(
0167,
0171,
0207,
0227,
0231,
0232,
0255,
0271,
0302).

Two
commenters
supported
the
proposed
NO
x
limit
of
0.2
as
a
reasonable
level
of
control
(
0231,
0302).
Commenter
0232
supported
the
setting
the
presumptive
control
levels
of
NO
x
at
90%.

Three
commenters
said
the
proposed
levels
are
too
stringent
(
0171,
0207,
0227).
Commenter
0171
described
recent
plant
retrofits
using
low­
NO
x
burners
and
low­
NO
x
concentric
firing
systems
that
could
not
practically
achieve
the
0.2
lb/
MMBTU
limit.
Commenter
0207
said
the
presumptive
limit
is
counter
to
available
evidence
and
unreasonable,
especially
for
T­
fired
units
equipped
with
low­
NO
x
burners.
EPA's
experience
in
the
Acid
Rain
program
is
consistent
with
a
higher
emission
rate
and
should
be
reflected
in
setting
the
BART
levels.
The
commenter
concluded
that
the
experience
of
some
of
its
members
with
low­
NO
x
burners,
the
background
information
for
development
of
Group
1,
Phase
II
NO
x
limits,
and,
most
likely,
the
voluminous
NO
x
x
emission
data
reported
to
EPA
from
all
Phase
I
and
Phase
II
units,
all
demonstrate
that
a
presumptive
0.2
lb/
MMBTU
NO
x
emission
limit
for
utility
boilers
is
a
very
unreasonable,
representation
of
BART
for
such
units.
Commenter
0227
was
concerned
that
the
NO
x
presumptive
control
level
EPA
proposed
is
higher
than
the
level
current
technology
can
provide.
EPA
should
review
currently
proposed
NO
x
retrofits
and
consider
lowering
it
presumptive
level
of
0.2
lbs/
MMBTU.

Three
commenters
said
the
proposed
NO
x
levels
are
too
lenient
(
0167,
0255,
0271).
Commenter
0167
urged
EPA
to
require
electric
generating
units
in
the
West
to
meet
rigorous
NO
x
pollution
control
standards.
Commenter
0255
said
the
current
suite
of
best
available
NO
x
technologies,
post
combustion
as
well
as
combustion
controls,
can
achieve
significantly
lower
emission
rates
than
the
proposed
0.2
lb/
MMBtu.
In
2002
ozone
season,
the
air
pollution
control
industry
had
100
units
achieving
a
NO
x
emission
rate
of
less
than
0.2
lb/
MMBtu.
Additionally,
cost
effective
emission
reductions
from
the
industrial
sector
are
feasible
but
have
not
been
adequately
considered
in
the
BART
rule
considering
the
quantity
of
visibility
impairing
emissions.
For
industrial
sources,
setting
an
appropriate
level
of
reduction
with
some
flexibility
such
as
185
providing
the
option
of
meeting
an
emission
rate
limit
or
percent
reduction
is
one
way
to
assure
significant
emission
reductions.
Commenter
0271
recommended
the
presumptive
control
level
of
0.15
lbs.
of
NO
x
per
MMBTU.
This
level
of
control
has
been
demonstrated
in
the
NO
x
budget
program
in
the
eastern
United
States.

Comment:
Six
commenters
said
that
the
presumption
is
overly
broad
(
0189,
0206,
0217,
0246,
0303,
0323).
Commenters
(
0189)
said
that
EPA
should
make
clear
(
a)
that
add­
on
NO
x
controls
should
not
be
part
of
BART
for
non­
EGUs
because
of
the
costs
of
the
these
controls
and
(
b)
that
its
presumptive
control
level
of
.2
lbs/
MMBTU
does
not
apply
to
non­
EGUs
because
they
are
unable
to
meet
these
levels
without
the
use
of
add­
on
controls.

Other
commenters
(
0206,
0303)
said
EPA
should
not
establish
a
prescribed
method
of
level
of
control
because
variations
between
small
and
large
facilities,
e.
g.,
between
coal­,
oil­,
and
gasfired
units,
and
the
age
of
the
units
make
application
of
such
a
presumption
overbroad.
Commenters
(
0217,
0246)
also
said
disagreed
with
the
presumptive
NO
x
limit
at
0.27
lb/
MMBTU
because
of
the
many
differences
in
coal­
fired
boilers.
BART
determinations
should
be
made
by
individual
states
on
a
case­
by­
case
basis,
consistent
with
state
discretion.
Commenter
0323
said
EPA
should
consider
a
range
of
emission
rates
that
reflect
recent
advances
in
technology,
particularly
with
respect
to
tangentially
fired
boilers
at
EGUs.

Three
commenters
said
EPA's
characterization
of
the
capabilities
of
control
technology
is
inaccurate.
Commenters
(
0217,
0268)
provided
results
of
a
survey
showing
the
effect
of
adding
advance
combustion
controls
to
existing
units
that
currently
do
not
have
the
most
advanced
controls.
Expected
NO
x
emissions
rates
ranged
from
0.17
to
0.49
lbs/
MMBTU.
The
average
emissions
rate
achievable,
weighted
by
heat
input,
was
0.27
lbs/
MMBtu,
well
in
excess
of
the
presumed
limit
of
0.2
lbs/
MMBTU.
In
contrast,
commenter
0255
described
the
effectiveness
of
SNCR
as
a
candidate
to
meet
or
better
the
proposed
emission
rates.

Comment:
Thirteen
commenters
said
it
is
important
to
consider
the
type
of
coal
or
other
fuel
in
setting
a
presumptive
control
level
(
0186,
0188,
0213,
0214,
0217,
0224,
0238,
0254,
0255,
0275,
0306,
0319,
0321).

Commenters
(
0186,
0275)
said
the
0.2
lb/
MMBTU
presumptive
level
of
NO
x
control
should
not
apply
to
gas
fired
power
plant
boilers
because
combustion
controls
on
gas­
fired
units
can
achieve
a
much
better
(
i.
e.,
lower)
level
of
performance.

Other
commenters
(
0188,
0214,
0238,
0319,
0321)
said
that
the
0.2lb/
MMBTU
level
of
control
has
not
been
demonstrated
using
SCR
technology
on
lignite­
fired
EGUs
and
cited
current
research
supporting
this
claim
(
0188,
0321).
Commenter
0238
added
that
over­
fired
air
will
not
reduce
NO
x
emissions
from
cyclones
to
that
level.
Commenter
0321
said
the
proposed
emission
limit
is
not
appropriate
for
lignite­
fired
cyclone
units
either.
Commenters
(
0214,
0238)
added
that
186
EPA
should
give
the
states
maximum
flexibility
to
use
the
five
statutory
factors
in
their
BART
NO
x
determinations.
Commenter
0306
said
EPA
should
reexamine
the
presumptive
BART
for
cyclone
boilers
firing
North
Dakota
lignite.
Commenters
(
0319,
0321)
supported
a
recommendation
for
a
NO
x
limit
of
approximately
0.4
lb/
MMBTU,
or
evaluated
on
a
case­
by­
case
basis.
Commenter
0321
added
that
NO
x
emissions
from
lignite
unite
appear
to
have
less
impact
on
visibility
in
that
region
of
the
country.

Commenter
0217
said
coal
rank
also
accounts
for
differences
in
NO
x
levels.
For
example,
subbituminous
coal
from
the
Powder
River
Basin
has
been
shown
to
yield
lower
NO
x
in
some
units
than
other
coals.
However,
this
is
not
always
true.
In
addition,
unit
operators
must
balance
oxygen
levels
to
attain
proper
NO
x
levels
while
still
maintaining
appropriate
carbon
monoxide
(
CO)
levels.
A
unit
with
a
stringent
CO
emission
limit
may
not
allow
a
plant
operator
to
achieve
the
lowest
NO
x
level
possible
with
existing
controls.

Commenter
0213
said
that
cyclone
boilers
that
burn
bituminous
coal
may
require
a
higher
presumptive
emission
rate
even
after
OFA
retrofits.
Commenter
0254
said
pre­
NSPS
boilers
firing
bituminous
coal
cannot
meet
the
presumptive
limit
with
the
prescribed
control
systems
(
combustion
controls
or
low­
NO
x
burners).
Commenter
0224
described
different
levels
of
performance
between
Western
plants
(
which
could
meet
the
limit)
and
Eastern
plants
(
which
could
not
meet
the
limit).
The
commenter
said
any
presumptive
performance
limits
must
reflect
different
types
of
coal
and
boiler
configurations.

Commenter
0255
said
the
emission
rate
achieved
from
a
particular
unit
depends
on
several
factors,
including
coal
type
and
boiler
configuration.
The
commenter
described
several
examples
of
high
performing
units.

Comment:
One
commenter
(
0270)
cited
studies
showing
the
nitrates
contribute
only
a
few
percent
to
total
light
extinction
except
in
Class
I
areas
downwind
of
Southern
California.
Accordingly,
NO
x
emissions
from
stationary
sources
are
unlikely
to
lead
to
any
perceptible
changes
in
visibility.
Therefore,
it
is
unlikely
that
NO
x
emission
sources
in
the
West
will
be
good
candidates
of
application
of
BART.
The
proposed
presumptive
NO
x
emission
level
is
not
consistent
with
the
Bush
Administration's
publicly
announced
policy
of
"
no
SCR
in
the
West"
under
the
CSA.
EPA
should
conclude
that
application
of
combustion­
based
controls
would
constitute
BART
for
NO
x
emissions
in
the
West,
irrespective
of
the
resulting
emission
level.

Comment:
Seven
commenters
said
that
the
boiler
configuration
should
be
considered
(
0195,
0210,
0217,
0224,
0241,
0256,
0270).
Commenters
(
0210)
said,
if
retained,
the
NO
x
presumption
should
be
modified
to
address
the
appropriate
control
levels
for
all
boiler
types.
EPA's
identified
presumptive
level,
0.2
lbs/
MMBTU,
is
not
reasonable
for
most
boiler
types
because
it
cannot
be
achieved
without
using
combustion
controls
(
0241,
0256,
0270).
EPA
could
establish
more
reasonable
levels
by
reviewing
the
emissions
data
readily
available
from
the
Acid
Rain
program
187
(
0210).
Commenter
0241
said
states
should
undertake
a
full
BART
analysis,
as
required
by
the
Act,
to
determine
the
appropriate
NO
x
control
level
that
should
be
established
as
BART
for
EGUs.
If
a
presumption
is
retained,
EPA
should
change
the
presumption
from
a
lb/
MMBtu
emission
limitation
to
a
statement
indicating
that
combustion
controls
installed
after
enactment
of
Title
IV
of
the
Act
should
be
presumed
to
satisfy
the
BART
requirement
for
NO
x.
States
could
then
evaluate
on
a
boiler­
specific
basis
the
emission
limitation
that
could
be
met
using
combustion
controls.
Any
other
approach
is
simply
not
justified
by
the
data
and
would
be
arbitrary
and
capricious.

Commenter
0195
said
NO
x
control
on
cyclone
type
boilers
represent
unique
circumstances
that
make
it
virtually
improbable
to
control
on
a
long
term
basis
below
0.4
lb/
MMBTU.

Commenter
0224
described
different
levels
of
performance
between
Western
plants
(
which
could
meet
the
limit)
and
Eastern
plants
(
which
could
not
meet
the
limit).
The
commenter
said
any
presumptive
performance
limits
must
reflect
different
types
of
coal
and
boiler
configurations.

Commenter
0270
said
it
is
well
known
that
wall­
fired
boilers
can't
achieve
the
same
level
of
NO
x
controls
as
tangentially­
fired
boilers
with
these
types
of
control
technologies.
The
commenter
does
not
believe
that
a
limit
of
0.2
lbs/
MMBTU
can
be
achieved
on
wall­
fired
units
via
retrofit
installation
of
advance
combustion
controls.

Comment:
Three
commenters
(
0195,
0252,
0271)
provided
other
comments
regarding
applicability.
Commenter
0195
supported
the
definitions
for
Combustion
controls
as
defined
in
40
CFR
76.2
Acid
Rain
Nitrogen
Oxides
Emission
Reduction
Program
in
order
to
provide
consistency
across
EPA
programs.
Commenter
0252
analyzed
data
for
eastern
units
and
western
units
with
different
configurations
and
for
units
burning
different
types
of
coal.
The
commenter
concluded
that
EPA
should
not
set
a
presumptive
level
of
control,
since
it
should
be
determined
on
a
case­
by­
case
basis.
Commenter
0271
said
the
presumptive
level
of
control
must
be
applied
to
all
units
at
a
"
powerplant
having
a
total
generating
capacity
in
excess
of
750
megawatts"
even
those
units
less
than
750
megawatts.
This
conforms
to
the
specific
wording
of
the
Clean
Air
Act
(
Sec.
169A(
b)(
2)(
A))
for
controls
on
these
smaller
than
750
megawatt
units.

Comment:
Nineteen
commenters
discussed
the
use
of
post­
combustion
controls
(
0171,
0179,
0199,
0206,
0217,
0221,
0224,
0232,
0234,
0244,
0246,
0252,
0253,
0255,
0256,
0267,
0271,
0280,
0303).
Commenters
supported
the
presumption
that
BART
can
be
achieved
without
the
use
of
post­
combustion
controls
(
0199).
However
some
types
of
units
or
units
burning
some
types
of
fuel
cannot
meet
the
presumptive
limits
without
installing
controls
(
0199,
0217,
0244,
0252,
0253).
Commenters
described
several
examples.

Some
commenters
were
concerned
that
the
proposed
rule
would
require
some
plants
to
install
SCR
to
meet
the
NO
x
control
level
proposed
(
0171,
0206,
0217).
Commenter
0171
said
the
188
implementation
of
the
regional
haze
rule,
including
any
potential
retrofit
of
SCR
technology,
must
consider
the
site
specific
data
of
the
Class
I
area,
the
use
of
current
emission
controls
on
potential
sources,
and
the
degree
of
visibility
improvements.
Commenters
(
0217,
0246)
added
that
individual
EGU
differences
need
to
be
accounted
for
by
states
in
making
their
BART
determination.
Also,
the
guidelines
indicate
that
if
a
state
finds
that
a
source's
visibility
contribution
warrants
the
installation
of
SCR,
it
may
be
allowed.
However,
the
guidelines
need
to
also
provide
for
instances
where
the
visibility
condition
warrant
a
lesser
control
level
than
what
would
be
achieved
by
advanced
combustion
control.
There
is
reference
to
this
concept
in
the
preamble
that
is
not
reflected
in
the
guidelines.
Commenters
(
0206,
0303)
said
the
costs
associated
with
SCR
operations
at
coal­
fired
power
plants
are
not
insignificant.
Annual
operation,
rather
than
only
during
ozone
season,
would
significantly
reduce
catalyst
life,
requiring
more
frequent
replacement.

Commenters
agreed
that
low­
NO
x
burners
should
be
the
presumptive
technology
because
they
are
cost
effective
and
the
incremental
cost
of
adding
SCR
or
SNCR
would
be
substantially
less
cost
effective
(
0224).

Commenter
0255
said
post­
combustion
controls
(
e.
g.
SCR
and
SNCR)
as
well
as
advanced
(
or
second
generation)
low­
NO
x
burners
are
significantly
more
efficient
and
are
currently
the
best
available
control
technologies
for
NO
x.
The
majority
of
the
low­
NOx
burners
currently
on
coalfired
units
were
installed
to
meet
the
compliance
deadline
for
the
start
of
Phase
I
and
II
of
the
Acid
Rain
Program
in
1995
and
2000,
respectively.
Units
that
installed
their
low­
NO
x
burners
for
Phase
I
of
the
Acid
Rain
Program
will
have
had
their
air
pollution
control
equipment
in
place
for
almost
20
years
by
the
time
the
BART
rule
is
implemented
in
2013.
Even
the
units
that
installed
low­
NO
x
burners
by
2000
will
have
had
their
equipment
in
place
for
over
10
years
by
2013.
Exemptions
for
existing
low­
NO
x
burner
equipment
that
is
not
meeting
the
0.2
lb/
MMBtu
emission
rate
should
not
be
given
as
technology
performance
has
significantly
improved
since
1995
and
will
continue
to
improve
out
to
2013.
For
these
reasons,
EPA
should
remove
the
provision
that
permits
sources
with
first
generation
low­
NO
x
burners
presumptive
NO
x
emission
rate
and
mandate
that
BARTeligible
sources
meet
the
0.2
lb/
MMBTU
or
lower
emission
rate.
Commenter
0255
also
provided
information
showing
that
good
performance
of
SCR
on
lignite
burning
units.

Some
commenters
approved
of
including
year­
round
SCR
operation
for
removal
of
NO
x
(
0179,
0221,
0232,
0234,
0271).
Commenters
(
0232,
0234)
would
limit
this
application
to
sources
already
operating
SCR.
Commenter
0221
described
how
the
technology
has
advanced,
is
cost
effective,
and
will
meet
levels
of
0.15
lb/
MMBTU
(
NO
x
SIP
call)
to
0.125
(
IAQR/
CAIR)
at
costs
estimated
by
EPA
to
be
less
than
$
2,000/
ton.
The
commenter
said
that
SCR
represents
the
best
available
technology
for
removal
of
NO
x.
Commenter
0232
said
a
greater
control
efficiency
is
possible
for
these
sources.

Commenter
0267
said
the
NO
x
level
should
be
lowered
to
require
application
of
postcombustion
controls.
At
a
minimum,
application
of
post­
combustion
controls
for
NO
x
should
not
be
ruled
out
as
BART.
Commenter
0280
agreed
and
said
the
presumptive
NO
x
level
of
0.2
189
lbs/
MMBTU
should
be
considered
as
the
maximum,
not
minimum,
allowable
emission.
States
should
require
the
lowest
emission
rate
that
can
be
achieved
using
all
available
and
applicable
control
technologies
as
in
BACT.
The
economic
feasibility
should
be
the
limiting
criteria.

Comment:
Six
commenters
provided
other
comments
on
controls
(
0221,
0241,
0244,
0252,
0255,
0256).

Commenter
0241
said
if
EPA
intends
to
retain
a
presumption
regarding
NOx
controls
in
the
final
Guidelines,
EPA
should
specify
only
that
EGUs
should
be
equipped
with
combustion
controls
to
limit
NO
x
emissions
without
identifying
a
presumed
NO
x
emission
rate.
Commenters
(
0244,
0252)
agreed
that
EPA
should
not
set
a
presumptive
level
of
control,
since
it
should
be
determined
on
a
case­
by­
case
basis.
However,
if
a
limit
must
be
set
as
guidance
for
the
states
it
should
be
higher
than
0.20
lb/
MMBTU,
in
order
to
avoid
forcing
the
use
of
post
combustion
controls
(
e.
g.
SCR
and
SNCR.).

Commenter
0256
said
if
EPA
does
retain
the
presumptions,
at
a
minimum,
the
presumptions
should
not
apply
to
units
that
already
are
controlled
with,
for
SO
2,
scrubbers,
and,
for
NO
x,
combustion
controls
that
were
added
after
enactment
of
Title
IV
of
the
Act.
Units
with
combustion
controls
that
were
added
after
enactment
of
Title
IV
should
be
deemed
to
already
have
the
best
combustion
controls
available,
and
thus
should
not
be
subject
to
the
presumption.
Again,
the
state
should
be
allowed
to
analyze
BART
for
those
units
applying
the
statutory
factors
in
light
of
the
relatively
recent
NO
x
controls
that
were
retrofit
at
the
unit
Commenter
0221
said
that
because
power
plants
account
for
about
one­
quarter
of
the
nitrogen
oxide
emissions
nationally,
and
because
the
Agency
also
has
and
will
have
considerable
"
experience
in
evaluating
NO
x
control
options
for
utility
boilers,"
a
presumptive
control
level
for
NO
x
should
be
presumed
as
"
best
available"
at
the
outset
of
the
required
top­
down
BART
analysis
for
both
controlled
and
uncontrolled
power
plants.
In
order
to
provide
support
for
such
a
presumption,
EPA
should
undertake
a
similar
analysis
of
NO
x
removal
technologies
as
was
done
by
ORD
for
sulfur
dioxide.

Commenter
0255
provided
data
on
100
coal­
fired
units
and
the
majority
are
very
high
performing.
The
incentive
to
control
emissions
under
the
NO
x
SIP
Call
will
become
stronger
over
time,
assuming
electric
demand
continues
to
grow,
as
sources
utilize
their
banked
emissions.
As
the
bank
of
emission
credits
is
reduced,
additional
NO
x
reductions
will
be
required.
Some
sources
will
be
required
to
install
control
technologies
while
those
with
controls
will
have
an
incentive
to
operate
their
control
technologies
at
their
highest
efficiency
so
they
can
sell
the
emission
credits
that
they
offset.

Comment:
190
Four
commenters
discussed
the
averaging
period
for
the
emission
limits
(
0207,
0214,
0238,
0314).
Commenters
said
the
averaging
period
is
unclear
and
should
be
annual
vs.
a
short­
term
limit
(
0207,
0214,
0238).
Commenter
0314
suggested
a
30­
day
rolling
average.

Comment:
Eight
commenters
made
general
comments
on
the
proposed
NO
x
levels
(
0181,
0186,
0215,
0225,
0241,
0268,
0275,
0314).
Commenter
0181
said
EPA
should
remove
the
presumptive
NO
x
control
levels
from
the
final
BART
guidelines
and
instead
replace
it
with
a
site­
specific
cost
analysis.
Once
again,
the
presumptive
NO
x
emission
control
approach
inappropriately
ignores
the
need
for
a
visibility
impact
evaluation
which
is
required
in
Step
5
of
the
proposed
case­
by­
case
BART
engineering
analysis.
Commenter
0268
such
a
presumptive
limit
infringes
on
a
state's
authority
to
establish
BART
on
a
case­
by­
case
basis
considering
not
only
visibility
improvement,
but
other
factors
as
well.

Commenter
0186
said
EPA's
proposal
for
no
post­
combustion
NO
x
controls
on
power
plants
should
apply
equally
across
all
source
categories;
otherwise,
there
should
be
no
presumptive
control
technology
prohibition
for
any
source
category.

Commenter
0215
said
if
an
EGU
is
achieving
the
default/
presumption
for
NO
x,
it
should
not
be
required
to
do
a
BART
determination
for
that
pollutant.
Such
presumptions
would
streamline
the
BART
process
for
those
sources
willing
to
reach
these
control
levels.

Commenter
0225
said
for
sources
located
within
50
km
of
a
Class
I
area,
the
source
could
be
required
to
perform
an
evaluation
of
control
technologies
that
compare
their
cost
effectiveness.
A
number
of
the
large
power
companies
have
completed
these
types
of
evaluations
and
have
a
good
understanding
of
the
cost
of
the
various
control
scenarios
for
their
units.
States
may
want
to
request
cost
effectiveness
evaluations
from
sources
to
determine
if
a
source
can
cost
effectively
reduce
their
emissions.
Because
a
larger
number
of
sources
outside
of
the
NO
x
SIP
Call
region
have
not
been
required
to
install
many
advanced
NO
x
controls,
cost
effective
reductions
can
be
achieved.

Commenter
0241
said,
for
the
same
reasons
identified
above
with
respect
to
the
presumptive
SO
2
control
level
for
EGUs,
the
commenter
objects
to
EPA'
s
proposal
to
establish
a
presumptive
NO
x
control
level
for
EGUs.

Commenter
0275
supported
EPA's
proposed
finding
that
add­
on
NO
x
controls
would
be
too
costly
to
be
considered
BART
for
EGUs,
in
most
cases.
Even
more
so,
this
would
hold
true
for
non­
EGU
boilers,
"
smaller
boilers
are
generally
less
cost
effective
to
control."
However,
this
finding
cannot
be
made
unless
a
demonstration
is
made
that
visibility
impairments
do
not
remain.
That
is,
NO
x
add­
on
controls
for
EGUs
and
non­
EGUs
may
be
appropriate
if
modeling
indicates
that
2018
visibility
goals
will
otherwise
not
be
met.
191
Commenter
0314
expects
continuing
technological
improvement
in
the
future
and
that
a
specific
presumptive
BART
level
set
by
rule,
will
make
it
difficult
for
a
state
to
establish
a
lower
level.
As
proposed
for
SO
2,
the
commenter
suggested
a
range
for
NO
x
that
will
account
for
the
variability
in
the
boilers
as
well
as
technology
improvements
in
the
future.
A
range
of
0.13
to
0.17
lb/
MMBTU
was
suggested.

4.2.3.3
General
Comments
Regarding
Presumptive
Limits
Comment:
Seven
commenters
supported
the
use
of
presumptive
emission
limits
in
the
BART
program
(
0179,
0221,
0255,
0285,
0299,
0323,
28­
266).
Commenters
said
the
limits
are
achievable
(
0179,
0255)
and
are
one
of
the
most
essential
components
of
the
BART
guidelines
because
they
establish
a
starting
point
representing
the
"
best
available
technology"
in
the
required
engineering
analysis
(
0221).
Some
commenters
requested
that
EPA
establish
more
stringent
limits
and
harmonize
the
presumptive
BART
emission
limits
for
EGUs
under
the
regional
haze
program
with
the
BART
guidelines
(
0221).
Commenter
0323
approved
including
a
specific
BART
determination
for
large,
bituminous
coalfired
electric
generating
stations.
These
facilities
constitute
the
largest
emitters
of
sulfur
dioxide
and
nitrogen
oxides
among
all
26
source
categories
subject
to
BART.

Commenters
(
0285,
28­
266)
urged
EPA
to
develop
national
guidelines
and
presumptive
BART
requirements
for
source
types
common
to
several
states
to
make
the
process
more
efficient
and
workable
for
the
states.
Commenters
(
0299,
28­
266)
added
that
the
guidelines
should
encourage
consistency
between
states
and
said
EPA
should
establish
a
consistent
set
of
protocols
upon
which
a
BART
analysis
for
each
source
category
should
be
based.
Commenter
0299
added
that
such
protocols
should
not
eliminate
the
flexibility
each
state
may
need
to
address
unusual
circumstances
at
individual
sources,
but
these
protocols
could
potentially
eliminate
the
need
for
case­
by­
case
review
for
many
sources.
In
the
absence
of
a
presumptive
standard,
the
proposed
BART
guidelines
must
contain
explicit
criteria
for
determining
BART.
In
addition,
the
BART
Guidelines
should
include
defined
criteria
for
"
engineering",
"
useful
life",
and
"
cost".

Comment:
XX
commenters
said
EPA
should
establish
presumptive
limits
for
other
pollutants
and/
or
source
categories
(
28­
248,
28­
255,
28­
358).
Commenter
28­
255
would
also
seek
to
establish
limits
for
source
categories
in
addition
to
utility
boilers
to
make
the
process
more
uniform
and
consistent
across
the
country
and
easier
for
the
states.
Commenter
28­
358
said
the
SO
2
limits
should
be
expanded
for
all
BART­
eligible
industrial
boilers
in
all
26
categories
and
set
at
95%
to
reflect
control
technology
performance.
Commenter
28­
248
proposed
setting
a
limit
for
particulate
matter
at
90%.

Comment:
Forty­
nine
commenters
opposed
establishing
presumptive
emission
limits
(
0171,
0181,
0184,
0195,
0197,
0198,
0208,
0210,
0211,
0214,
0215,
0217,
0224,
0226,
0228,
0238,
0244,
192
0246,
0247,
0252,
0253,
0254,
0256,
0268,
0270,
0275,
0276,
0308,
0314,
0321,
28­
243,
28­
264,
28­
267,
28­
305,
28­
333,
28­
340,
28­
342,
28­
347,
28­
354,
28­
356,
28­
357,
28­
360,
28­
385,
28­
390,
28­
401,
28­
402,
28­
748,
28­
7003,
28­
7009).

Commenters
said
this
approach
restricts
the
flexibility
of
the
states
and/
or
sources
to
make
the
control
technology
evaluations
on
a
case­
by­
case
basis
(
0171,
1081,
0195,
0197,
0211,
0217,
0247,
0253,
0256,
0268,
0275,
0276,
0321,
28­
243,
28­
305,
28­
333,
28­
340,
28­
342,
28­
349,
28­
354,
28­
357,
28­
360,
28­
385,
28­
390,
28­
748)
without
resulting
in
a
notable
degree
of
improvement
to
visibility
(
0171,
0197).

The
Act
requires
states
to
use
the
statutory
factors
in
making
case­
by­
case
BART
determinations
(
0195,
0197,
0210,
0211,
0217,
0228,
0238,
0244,
0246,
0247,
0252,
0253,
0256,
0268.
0308,
0314,
0321,
28­
264,
28­
267,
28­
347,
28­
355,
28­
357,
28­
359,
28­
360,
28­
385,
28­
390,
28­
401,
28­
402,
28­
747,
28­
748,
28­
7001,
28­
7003,
28­
7009).
Commenters
(
0211,
0217,
0246)
added
that,
rather
than
assist
states
by
providing
them
with
resources,
tools
and
guidance
under
BART
guidelines,
EPA's
use
of
rebuttable
presumptions
will
tend
to
require
significant
levels
of
state
resources
for
states
to
exercise
discretion
in
their
BART
determination
processes.
Most
states
have
limited
administrative
resources
to
address
these
issues
and
therefore,
EPA's
use
of
presumptive
requirements
will
discourage
state
level
discretion
­­
opposite
the
intent
conveyed
by
the
Coin
Growers
decision.

Commenters
(
0198,
0226,
0254)
said
the
presumptive
levels
are
not
achievable
across
the
range
of
boiler
types,
vintage
of
existing
controls
and
coal
composition.
Commenter
0198
preferred
the
use
of
trading
under
CAIR
to
meet
BART
requirements
in
the
most
efficient
and
cost­
effective
manner.
Commenter
0226
said
if
a
presumptive
level
of
control/
emission
rate
is
retained
in
the
final
rule,
the
guidelines
should,
at
a
minimum,
make
it
clear
that
states
have
the
authority
to
select
an
alternative
level
of
control/
emission
rate,
without
an
undue
evidentiary
burden,
given
appropriate
consideration
of
all
the
statutory
BART
factors.
Commenter
0253
asked
that
EPA
either
delete
the
provisions
for
mandatory
default
control
levels
or
provide
a
source
with
the
option
of
using
the
default
control
levels,
rather
than
making
them
mandatory.
Commenter
0276
said
if
EPA
establishes
presumptive
control
levels
for
EGUs,
average
annual
values
are
more
appropriate
than
those
in
the
proposed
regulations.
In
addition,
BART
should
not
be
structured
in
a
way
that
results
in
rigid
technology­
based
limits
that
interfere
with
established
NO
x
and
SO
2
emission
trading
programs.
Commenter
0314
said
if
EPA
needs
to
define
a
target,
it
should
use
a
range
to
prevent
diminishing
states'
flexibility
in
case­
by­
case
BART
reviews.

Commenters
(
0214,
0215,
0238,
0254)
said
EPA
fails
to
make
clear
how
the
proposed
default
values
for
SO
2
and
NO
x
relate
to
visibility
impairment.
Visibility
is
both
Class
I
area
and
source
specific,
which
is
the
reason
Congress
gave
the
states
the
lead
role
and
discretion
in
the
BART
program
to
determine
which
sources
need
to
install
or
upgrade
controls.
Through
the
use
of
presumptions
and
default
values,
however,
EPA's
prescriptive
process,
as
proposed,
would
make
the
installation
of
maximum
controls
more
likely
without
regard
to
visibility
benefits.
Instead,
as
required
by
statute
and
rules,
EPA
should
give
the
states
maximum
flexibility
to
use
the
193
five
statutory
factors
in
their
BART
determinations.
Commenters
(
0228,
0247)
said
sources
must
be
allowed
to
assess
the
visibility
improvements
of
a
variety
of
control
options.
Commenter
0308
said
states
should
be
allowed
to
exempt
BART­
eligible
sources
from
BART
requirements
if
these
sources
do
not
impact
or
have
very
little
impact
on
visibility
in
a
Class
I
area.

Commenter
28­
342
said
even
if
a
presumption
were
appropriate
under
the
visibility
protection
program
(
and
it
is
not),
EPA
cannot
adopt
one
now
because
the
Agency
has
not
satisfied
the
conditions
for
creating
a
presumption.
Federal
presumptions
can
be
promulgated
only
when
the
Agency
has
established
(
based
on
generic
application
of
law
to
fact)
that
the
presumption
is
valid.

Commenters
(
0208,
0270)
said
EPA's
default
and
presumptive
control
levels
for
SO
2
and
NO
x
are
inconsistent
with
the
BART
control
efficiencies
proposed
by
the
WRAP
and
should
be
removed
from
the
proposed
guidelines.
In
no
event,
however,
should
BART
force
plants
in
the
West
to
reduce
emissions
to
levels
greater
achievable
using
dry
scrubbing
technology
for
SO
2
and
advanced
low­
NO
x
combustion
technology
for
NO
x.
As
EPA
has
pointed
out
in
its
discussion
of
Clear
Skies
and
CAIR,
wet
scrubbers
and
SCR
are
inappropriate
for
the
arid
West.
In
fact,
the
statute
itself
makes
the
NSPS
a
ceiling
for
BART.
Commenter
0270
said
the
proposed
presumptive
BART
emission
limitations
are
inconsistent
with
the
Regional
Haze
Rule
for
the
GCVTC
states
(
WRAP
Annex
and
SIPs
under
40
CFR
51.309),
the
Administration's
proposed
legislation
for
the
power
generation
sector,
the
CSA,
and
the
proposed
IAQR.

Commenters
(
0184)
said
limits
should
not
be
set
for
those
pollutants
for
which
a
facility
has
no
known
control.

Commenter
0246
said
EPA
should
clarify
that
greater
reasonable
progress
need
not
be
demonstrated
by
achieving
emission
reductions
greater
than
those
that
would
occur
by
applying
the
presumptive
control
levels
proposed
by
EPA
to
all
BART­
eligible
sources.
Such
an
approach
would
undermine
the
fundamental
role
of
the
states
in
determining
BART
or
what
constitutes
greater
reasonable
progress
than
BART.
The
WRAP,
states,
tribes
and
stakeholders
conducted
a
thorough
analysis
and
balancing
of
the
BART
factors
to
arrive
at
the
WRAP
milestones
 
all
of
this
was
conducted
without
imposing
presumptive
control
levels
during
the
analysis
and
discussion.

If
EPA
does
utilize
the
presumptive
emission
limits,
commenter
0197
said
,
if
a
source
agrees
to
95%
control
or
the
range
of
0.1­
0.15
lb/
MMBTU
for
SO
2,
they
should
be
automatically
exempt
from
the
BART
process.
The
same
goes
for
the
0.2
lb/
MMBtu
for
NO
x.

Comment:
Commenter
0268
questioned
the
legal
authority
of
EPA
to
impose
presumptive
limits
for
sources
between
250
MW
and
750
MW
in
size
because
the
visibility
provisions
of
the
Act
only
specify
evaluation
by
the
states
of
units
with
a
capacity
in
excess
of
750
MW.
Commenter
0184
said
the
statute
requires
state
to
take
a
case­
by­
case
approach
for
non­
EGU
sources.
Discussion
of
such
emission
limits
should
be
explicitly
limited
to
those
large
EGU
facilities
for
which
EPA
has
194
the
statutory
authority
to
define
BART
and
to
set
such
limits.
Commenter
0237
said
the
presumptive
BART
level
for
utility
boilers
would
impose
hardship
on
industrial
sources
because
there
are
many
practical
differences
in
the
way
industrial
boilers
and
electrical
generating
units
are
operated,
including
short
term
swings
in
demand,
that
make
such
BART
controls
unreasonable.
Therefore,
EPA
must
underscore
in
its
final
guideline
that
these
control
levels
are
not
presumptive
BART
for
industrial
boilers.

Five
commenters
said
EPA
should
consider
the
type
of
emission
unit
(
e.
g.,
coal
type,
boiler
type,
and/
or
emission
controls)
when
making
BART
determinations
(
0195,
0198,
0214,
0238).
Commenters
(
0214,
0238)
added
that
states
and
sources
should
be
focusing
on
the
incremental
costs
and
benefits
from
the
installation
of
additional
controls
or
the
upgrading
of
existing
controls
in
the
BART
determination.

Comment:
Two
commenters
supported
the
use
of
optional
presumptive
limits
as
a
choice
for
source
owners/
operators
to
avoid
a
BART
analysis
(
0210,
0256).
This
approach
would
limit
the
resources
expended
for
many
determinations,
but
would
allow
states
to
consider
all
of
the
statutory
factors
where
appropriate.

Comment:
Six
commenters
said
EPA
needed
to
address
the
averaging
period
associated
with
the
presumptive
limits
(
0195,
0227,
0228,
0268,
0306,
0323).
Some
commenters
suggested
annual
average
limits
(
0195,
0268)
while
others
said
they
should
be
no
less
than
a
30­
day
or
monthly
average
(
0228).
Other
commenters
asked
EPA
to
specify
an
averaging
time
similar
to
the
averaging
time
of
visibility
impact
assessments,
e.
g.,
24­
hour
average
(
0227,
0306).

Commenter
28­
385
said
EPA
indicates
that
BART
determinations
should
specify
an
averaging
time
"
consistent
with
established
reference
methods."
EPA
is
again
skewing
BART
determinations
to
a
"
top­
down"
BACT
assessment,
which
is
not
consistent
with
the
statute.
Short
averaging
times
would
significantly
impact
cost
by
requiring
controls
with
redundant
systems,
but
would
provide
only
minimal
additional
annual
tons
reduction
and
negligible
visibility
benefits.

Comment:
One
commenter
said
EPA
should
establish
emission
limits
for
PM
because
of
the
magnitude
of
emissions
and
the
availability
of
proven
controls
(
0255).

Comment:
Seven
commenters
made
other
general
comments
regarding
presumptive
emission
limits
(
0195,
0214,
0232,
0238,
0255,
0299,
0306).

Commenters
(
0195,
0214,
0238)
said
exceptions
should
be
made
for
startups,
shutdowns
and
malfunctions
of
control
equipment.
195
Commenters
(
0214,
0238)
said
EPA
should
make
it
clear
that
if
any
source
is
willing
to
achieve
95%
control
or
somewhere
along
the
scale
of
0.1
to
0.15
lb/
MMBTU,
it
should
automatically
be
exempt
from
the
requirement
to
perform
a
BART
analysis
for
SO
2.
The
same
should
also
be
true
regarding
a
BART
NO
x
analysis
for
sources
willing
to
achieve
the
0.2
lb/
MMBTU
NO
x
x
presumption
for
EGUs.
Such
presumptions
would
streamline
the
BART
process
for
those
sources
willing
to
reach
these
control
levels.
Commenter
0255
said
EPA
should
set
a
high
threshold
for
sources
to
justify
being
exempted
because
the
current
suite
of
best
available
technologies
can
readily
meet
and
go
beyond
the
mandated
limits.
EPA
should
encourage
the
states
to
consider
these
factors
as
well.

Commenter
0232
said
there
should
be
a
provision
to
revisit
appropriate
BART
control
levels
at
a
date
certain.
While
it
is
appropriate
to
provide
clarity
and
certainty
to
industry
for
the
installation
of
retrofit
controls,
it
is
also
important
to
provide
for
application
of
new
controls
with
increased
control
efficiency
due
to
technology
improvements
in
order
to
meet
the
express
requirements
of
the
Act
to
prevent
any
future
and
remedy
any
current
impairment
of
visibility
in
Class
I
areas.

Commenter
0255
noted
that
EPA
did
not
address
emission
of
SO
3
from
power
plants.
This
has
become
an
increase
area
of
concern
for
power
companies
as
states
have
set
emission
standards
for
SO
3
in
the
operating
permits
for
power
plants.
A
number
of
control
technologies
have
been
developed
to
address
this
pollutant
and
are
able
to
achieve
a
high
level
of
removal.

Commenter
0299
said
the
BART
presumptive
level
of
control
proposed
for
EGUs
boilers
for
SO
2
emissions
and
for
NO
x
emissions
should
be
made
consistent
with
other
clean
air
programs.
Specifically,
the
presumptive
level
of
controls
for
EGU
boilers
should
be
consistent
with
the
requirements
adopted
in
the
CAIR,
and
consistent
with
the
requirements
contained
in
the
federal
NO
x
SIP
Call.

Commenter
0306
said
BART
should
be
no
more
stringent
than
a
current
BACT
assessment
for
an
existing
facility.
EPA
should
define
presumptive
BART
in
a
manner
consistent
with
recent
NSR
enforcement
actions.
Presumptive
BART
should
not
be
more
stringent
than
requirements
in
recent
agreements
(
or
proposed
agreements)
that
settled
NSR
violations.

4.3
General
Comments
Regarding
BART
Determination
4.3.1
Comments
Regarding
Streamlining
the
BART
Determination
Process
Comment:
Four
commenters
made
suggestions
to
streamline
the
BART
determination
process
(
0189,
0285,
0302,
0304).
Commenter
0189
said
EPA
should
allow
an
individual
state
 
or
a
group
of
states
in
a
region
 
to
adopt
a
generic
approach
to
at
least
some
BART
issues
in
the
interest
of
consistency
and
efficiency,
while
reserving
some
others
for
case­
by­
case
"
balancing".
It
would
also
be
efficient
for
EPA
itself
to
first
accept,
and
then
facilitate,
such
use
of
streamlined
state
196
procedures
for
making
BART
decisions.
To
the
extent
EPA
believes
states
making
BART
decisions
should
consider
certain
factors,
EPA
could
encourage
States
to
consider
these
factors,
and
make
that
consideration
more
efficient,
by
issuing
"
guidance"
stating
its
view
of
those
issues
for
the
states
to
consider.
EPA's
failure
to
even
consider
these
clearly
superior
alternative
approaches
is
emblematic
of
its
lack
of
attention
both
to
the
freedom
of
action
that
the
Act
clearly
grants
to
states
making
BART
decisions,
and
to
the
paperwork
burdens
that
its
preferred
approach
would
impose
on
states.
Other
commenters
requested
guidance,
such
as
control
technology
guidelines,
to
provide
a
starting
point
for
the
BART
analysis
(
0285,
0304).

Commenter
0302
stated
a
preference
for
the
"
alternative
approach
to
control
engineering
analysis
under
the
BART"
because
of
limits
on
state
resources
to
perform
site­
specific
engineering
analyses.
Commenter
0304
said
the
BART
approach
should
mirror
BACT
where
possible.
The
BACT
approach
has
served
the
PSD
program
well.
It
is
well
understood
by
both
agency
staff
and
industry.

Response:

We
believe
we
have
provided
as
streamlined
a
process
as
possible,
given
the
directives
in
the
CAA.
As
explained
in
the
preamble,
we
intend
to
propose
revised
alternative
provisions
to
allow
for
emissions
trading
programs
in
response
to
recent
litigation.
In
regard
to
the
BACT
process,
we
have
structured
the
BART
review
similarly
to
BACT
reviews
in
many
ways,
in
order
to
be
consistent
with
process
with
which
States
are
familiar.

4.3.2
Comments
Regarding
Reasonable
Progress
Comment:
Five
commenters
addressed
reasonable
progress
issues.
Commenter
0246
interprets
"
greater
reasonable
progress"
as
encouraging
consideration
of
all
the
BART
factors
including
visibility,
as
well
as
the
comparative
levels
of
emissions
reductions.
Whether
done
on
a
sourcespecific
or
category­
wide
basis,
states
have
the
responsibility
to
consider
all
of
the
factors
and
the
latitude
to
do
so
in
a
manner
they
deem
appropriate.
This
is
similar
to
the
approach
contained
in
the
WRAP
Annex.

Commenters
(
0249,
0251)
noted
that
the
rules
lack
criteria
for
measuring
reasonable
progress,
nor
do
they
provide
an
objective
basis
for
determining
whether
a
state's
plan
assures
reasonable
progress.

Commenter
0251
provided
extensive
comments
on
this
issue.
The
commenter
concluded
that
the
absence
of
criteria
for
determining
whether
progress
is
"
reasonable"
also
renders
the
haze
rules
inadequate
to
meet
the
statutory
mandate
to
"
assure"
reasonable
progress.
Also,
EPA's
failure
to
define
"
reasonable"
progress
also
violates
the
basic
prohibition
under
administrative
law
against
arbitrary
and
capricious
administrative
action.
197
Response:

We
note
that
this
rulemaking
is
not
intended
to
address
criteria
for
reasonable
progress,
but
only
to
implement
the
BART
provisions
of
CAA
section
169A.

4.3.3
Other
General
Comments
Regarding
the
BART
Determination
Comment:
Two
commenters
stressed
the
need
for
states
to
consider
the
statutory
factors
including
visibility
benefits,
energy
impacts,
or
the
remaining
life
of
control
equipment
(
0189,
0275).
Commenter
0275
added
that
the
proposal
also
fails
to
specify
how
states
would
address
the
visibility
analysis
results.
One
commenter
(
0232)
said
the
guidelines
fail
to
clearly
define
what
cost
would
be
considered
unreasonable,
what
energy
impact
would
be
considered
unacceptable
or
what
threshold
environmental
impact
would
result
in
the
elimination
of
a
particular
technology
option.
For
the
sake
of
a
nationally
consistent
program,
additional
specificity
is
warranted
in
the
guidelines
as
to
what
levels
would
prevent
the
most
stringent
controls
from
being
considered.
Two
commenters
said
nowhere
in
the
Act
is
there
any
presumption
that
some
technology
is
BACT
simply
because
it
has
been
determined
to
be
BACT
for
a
given
type
of
emission
source
in
another
location
(
28­
364,
28­
395).
Congress
recognized
that
the
balancing
test
is
mandatory
simply
because
site
specific
considerations
will
warrant
emphasis
on
different
considerations.

Response:

The
BART
guidelines
is
intended
to
aid
States
in
assessing
the
BART
factors.
States
have
considerable
latitude
in
weighing
the
BART
factors
and
taking
specific
circumstances
into
account.

Comment:
One
commenter
(
0207)
asked
EPA
to
clarify
in
the
final
rule
that
projects
undertaken
to
comply
with
the
regional
haze
rule
are
the
types
of
projects
excluded
from
NSR
regulatory
schemes
by
the
Pollution
Control
Project
provisions.
Commenter
28­
255
said
if
NSR
regulations
are
amended
to
change
the
definition
of
baseline
period,
it
should
be
done
as
soon
as
possible
so
it
won't
disrupt
the
BART
process.
Commenter
0184
said
the
final
rule
should
exempt
BART
compliance
from
triggering
NSR/
PSD
There
are
several
examples
in
the
proposal
that
indicate
that
a
physical
modification
or
change
in
the
method
of
operation
could
be
necessary
to
meet
the
BART
requirements.
The
commenter
added
that
EPA
should
make
an
effort
to
find
regional
haze
solutions
that
are
aimed
at
addressing
the
most
significant
contributing
sources,
rather
than
targeting
only
major
stationary
sources.
A
more
sensible
approach
would
be
to
examine
all
sources
and
the
associated
cost­
effectiveness
of
controls
and
then
make
more
holistic
determinations
as
to
which
sources
should
be
addressed.
While
this
issue
is
beyond
the
scope
of
the
BART
rule
 
or
even
the
underlying
regional
haze
rule
 
EPA
should
use
the
issue
of
regional
haze
to
develop
policies
that
seek
reductions
from
the
most
significant
sources
of
a
given
emission
in
the
most
cost­
effective
manner.
This
might
require
16
Consolidated
Appropriations
Act
for
Fiscal
Year
2004.
Pub.
L.
108­
199,
January
23,
2004.

198
creative
technologies
and
decision
making,
as
well
as
politically
tough
approaches,
but
it
is
necessary
to
truly
address
most
of
the
remaining
air
quality
issues.
Commenter
28­
273
said
the
final
rule
should
explicitly
exempt
BART
compliance
from
triggering
NSR/
PSD.

Response:

These
comments
are
outside
the
scope
of
the
BART
rulemaking.

5.0
Enforceable
Limits/
Compliance
Date
5.1
Proposed
Elimination
of
"
Committal"
SIP
Deadlines
Comment:
Two
commenters
agreed
with
the
proposal
to
eliminate
the
requirement
for
the
commttal
SIP
(
0231,
0259).
One
commenter
(
0245)
was
concerned
about
the
level
of
resources
required
by
the
BART
guidelines
combined
with
the
CAIR,
PM2.5
and
8­
hour
ozone
rules
impairing
the
states'
ability
to
complete
the
regional
haze
SIP
by
2008.

One
commenter
(
0251)
noted
that
the
Transportation
Equity
Act
(
TEA­
21)
set
explicit
deadlines
for
submittal
of
haze
SIPs
by
the
states.
The
commenter
discounted
EPA's
"
speculation"
about
Congressional
intent,
and
said
that
the
fact
that
Congress
allowed
an
exception
for
the
Grand
Canyon
regional
planning
process
is
a
strong
indication
that
it
did
not
intend
to
do
so
for
other
regional
efforts.
The
nature
of
visibility
impairment
in
a
given
region
and
the
measures
needed
to
address
it
do
not
change
depending
on
whether
an
area
is
designated
"
attainment"
or
"
nonattainment"
in
the
Code
of
Federal
Regulations.
There
is
no
reason
that
states
cannot
begin
working
together
well
before
the
haze
SIP
submittal
deadline
to
address
these
issues
and
develop
coordinated
strategies.
EPA's
illegal
extension
of
the
deadline
simply
encourages
states
to
needlessly
delay
that
process.
Two
commenters
(
28­
266,
28­
278)
said
EPA
should
restate
the
SIP
timelines
in
the
guidance.
Commenter
28­
278
added
that
it
would
facilitate
multi­
pollutant
strategy
development,
by
addressing
the
coordination
of
the
SIP
timelines
for
regional
haze,
fine
particles
and
8­
hour
ozone.

Response:

As
explained
in
the
2004
reproposal,
Congress
has
now
reset
the
deadlines
for
regional
haze
SIP
submittals.
16
SIPs
will
be
due
3
years
from
promulgation
of
PM2.5
designations.
Therefore
SIPs
will
be
due
December
17,
2007.
199
5.2
Proposed
Harmonization
of
Deadlines
for
Section
308
and
309
SIPs
Comment:
Commenters
discussed
the
proposed
harmonization
of
deadlines
for
section
308
and
309
SIPs.
Some
commenters
supported
the
revised
changes
to
the
due
dates
for
submitting
SIPs
and
SIP
revisions
under
section
308
and
309
(
0231,
0259,
0293).

Commenters
(
0231,
0259,
0306,
28­
389)
urged
EPA
to
finalize
the
BART
guidelines
as
soon
as
possible
to
allow
states
with
an
extended
regulatory
time
frame
to
complete
the
technical
analysis
and
SIP
development
process
in
time
to
meet
the
SIP
submittal
deadline.
Additional
delay
will
only
further
complicate
state
work
efforts
as
assumptions
will
have
to
be
made
in
order
to
move
forward
toward
SIP
completion
absent
final
decisions
on
the
BART
guidelines.
Commenter
28­
389
said
that
some
of
the
operational
details
of
the
backstop
trading
program
established
in
the
Annex
of
the
GCVTC
are
tied
to
existing
BART
guidelines
and
the
Regional
Haze
rule.
We
believe
that
the
approach
taken
in
the
Annex
is
consistent
with
the
proposed
BART
rule
and
the
Guidelines.

Commenter
0169
said
in
implementing
the
SIP
requirements
for
BART,
consideration
should
be
given
to
facilities
which
have
recently
implemented
regulations
such
as
MACT
or
NSPS
which
also
reduce
SO
x,
NO
x,
VOC,
and
PM2.5.
If
a
facility
has
just
made
significant
capital
investments
in
new
technologies
or
state
of
the
art
emission
controls
which
also
reduce
regional
haze,
they
should
be
given
credit
for
such
actions
and
any
new
SIP
requirements
should
be
harmonized
with
respect
to
additional
controls
and
compliance
dates.

Commenter
0217
noted
that
several
WEST
Associates
members
operate
in
the
five
states
(
Arizona,
New
Mexico,
Oregon,
Utah
and
Wyoming)
that
have
selected
a
Section
309
course
of
action
under
the
regional
haze
program.
Those
states
have
submitted
SIP
proposals
on
or
before
December
31,
2003.
EPA
should
acknowledge
that
nothing
in
the
re­
proposed
regional
haze
rule
provisions
or
the
BART
guidelines
would
necessitate
changes
to
the
essential
program
elements
upon
which
those
SIP
submittals
relied
(
e.
g.,
the
quantitative
emission
milestones
for
SO
2).
Commenter
0259
urged
EPA
to
ensure
that
the
revised
rule
and
guidelines
will
have
no
impact
on
the
five
state
implementation
plans
already
submitted
under
section
309
of
the
Regional
Haze
Rule.

Response:

See
the
discussion
of
CEED
v.
EPA
in
the
preamble
to
this
rulemaking.

Comment:
Commenter
28­
255
said
the
year
2002
should
be
established
as
the
base
year
for
308
states'
emission
inventories
for
consistency
between
RPOs
and
states.

Response:
17
70
FR
9705,
February
28,
2005.

200
The
comment
is
outside
the
scope
of
the
BART
rulemaking.

5.3
Compliance
Deadlines
Comment:
Three
commenters
said
the
proposed
compliance
date
set
at
5
years
after
EPA
approves
the
SIP
is
highly
uncertain
given
the
unknown
amount
of
time
for
EPA
rulemaking
(
28­
266,
28­
278,
28­
312).
The
commenters
asked
EPA
to
provide
guidance
to
states
on
the
estimated
times
for
review
of
the
SIP
submittals.

One
commenter
(
28­
303)
said
the
compliance
deadline
of
2018
allows
ample
time
for
both
program
development
and
compliance.

Response:

The
5­
year
compliance
deadline
was
set
by
Congress
in
CAA
section
169A.

5.4
Monitoring
Comment:
Commenter
0207
said,
with
respect
to
emission
monitoring,
EPA
appears
to
recognize
that
BART
itself
should
not
necessarily
include
a
particular
monitoring
requirement.
For
example,
EPA
states
in
the
re­
proposed
guidelines
that
a
requirement
that
BART
emission
limits
be
met
on
a
continuous
basis
"
does
not
necessarily
require
the
use
of
continuous
emissions
monitoring
(
CEMs)."
While
most
if
not
all
BART­
eligible
EGUs
already
employ
"
Part
75"
CEMs,
non­
EGU
sources
may
not.
The
commenter
is
aware
that
various
periodic
monitoring
methods
are
sufficient
to
ensure
compliance
on
a
continuous
basis.
Expending
resources
on
installation
and
operation
of
new
continuous
emission
monitoring
equipment
does
not
have
an
impact
on
visibility,
as
long
as
other
appropriate
methods
are
available
to
demonstrate
compliance.
Moreover,
certain
pollutants,
such
as
VOCs,
do
not
readily
lend
themselves
to
a
continuous
monitoring
scheme.

Response:

States
will
need
to
set
specific
monitoring
requirements
for
BART
sources.
Section
302(
k)
of
the
CAA
requires
emissions
limits
such
as
BART
to
be
met
on
a
continuous
basis,
and
the
guidelines
discusses
specific
issues
related
to
this
requirement.
In
order
to
provide
consistency
across
programs,
as
part
of
the
presumptive
limits
for
large
EGUs
we
have
added
a
clarifying
provision
recommending
an
averaging
time
of
a
30­
day
rolling
average.
We
have
also
included
a
recommended
definition
of
"
boiler
operating
day"
that
is
consistent
with
the
definition
in
the
proposed
revisions
to
the
NSPS
for
utility
boilers
in
40
CFR
Part
60,
subpart
Da.
17
This
would
allow
30­
day
rolling
average
emission
rates
to
be
calculated
consistently
across
sources.
201
5.5
Enforceable
Limits
Comment:
Three
commenters
noted
that
the
proposed
guideline
also
suggests
that
an
enforceable
limit
should
be
established
for
each
pollutant
subject
to
control
(
28­
266,
28­
278,
28­
312).
Does
this
mean
that
states
must
establish
ammonia
limits
for
boilers,
as
well
as
SO
2,
NO
x,
VOC
and
PM?

Response:

As
noted
above
and
in
the
preamble,
States
have
considerable
discretion
in
regard
to
ammonia.

6.0
Statutory
and
Executive
Order
Reviews
6.1.1
Administrative
Process
Comment:
One
commenter
(
0276)
said
EPA
is
pursuing
an
inappropriate
administrative
process
in
this
regulatory
development
that
will
result
in
litigation.
Affected
parties
are
not
being
provided
an
opportunity
to
comment
on
the
specific
regulatory
language
that
might
drive
significant
financial
investments
in
capital
equipment.
For
example,
individual
source
exemptions
might
be
subject
to
"
Look­
up
Tables"
or
"
Source
Rankings"
or
the
"
Q/
D
Method."
However,
these
tables,
rankings,
etc.
have
not
been
provided
and
cannot
be
evaluated.
Further,
there
may
or
may
not
be
a
trading
program.
This
proposal
should
be
designated
an
advance
notice
of
proposed
rule
making
and
provide
another
chance
to
comment
on
something
more
than
a
menu
of
possible
approaches.

Commenter
28­
391
said
the
proposed
guidance
is
contrary
to
the
EPA's
earlier
efforts
to
reduce
the
complexity
and
costs
of
part
70
applications
and
permits,
and
conflicts
with
provisions
for
treatment
of
insignificant
emission
units
and
activities
in
the
major
source
operating
permit
regulations.
Issuance
of
the
BART
guidelines,
as
proposed,
will
negate
the
previous
progress
made
to
remove
regulatory
barriers,
minimize
administrative
costs,
and
would
ultimately
require
BART­
eligible
sources
and
permitting
authorities
to
expend
the
very
resources
previously
determined
to
be
unnecessary
because
of
the
trivial
environmental
benefits
to
be
realized.

Commenter
28­
294
said
EPA
should
extend
the
public
hearing
schedule
to
include
evening
hours
to
encourage
public
participation.

Response:

We
believe
that
at
this
point,
ample
opportunity
for
public
comment
has
been
provided
on
all
the
provisions
of
the
final
BART
Guidelines,
including
public
hearings
in
different
locations.
In
addition,
the
BART
Guidelines
implement
specific
provisions
of
the
CAA,
and
the
final
Guidelines
202
incorporates
a
number
of
new
provisions
aimed
at
streamlining
BART
determinations,
particularly
with
regard
to
flexibility
in
conducting
determinations
for
VOC
and
ammonia
sources.

6.1.2
Executive
Order
12866
­
Regulatory
Impact
Analysis
Comment:
Thirty­
one
commenters
said
the
regulatory
impact
analysis
is
flawed
and
should
be
redone
(
28­
243,
28­
278,
28­
304,
28­
305,
28­
312,
28­
333,
28­
336,
28­
337,
28­
340,
28­
341,
28­
342,
28­
344,
28­
345,
28­
347,
28­
348,
28­
349,
28­
355,
28­
356,
28­
357,
28­
358,
28­
361,
28­
364,
28­
387,
28­
395,
28­
398,
28­
405,
28­
406,
28­
747,
28­
748,
28­
778,
28­
7013)
for
several
reasons:

°
BART
controls
will
produce
undue
economic
hardship
to
the
largest
industrial
customers
and
the
communities
they
support
(
28­
243)
°
Analysis
should
address
estimated
emission
reduction
amounts,
changes
in
visibility
levels,
and
costs
associated
with
BART
controls
(
28­
278,
28­
312,
28­
342)
°
The
costs
of
the
program,
including
control
costs
and
compliance
costs,
are
vastly
understated
(
28­
304,
28­
333,
28­
336,
28­
348,
28­
355,
28­
356,
28­
357,
28­
358,
28­
364,
28­
387,
28­
398,
28­
405,
28­
406,
28­
747,28­
748)
and
several
commenters
provided
examples
°
The
accuracy
of
EPA's
renewed
economic
cost
projection
is
questionable,
given
that
EPA's
proposed
BART
rule
appears
to
greatly
limit
the
"
flexibility"
the
states
were
provided
by
Congress
in
1977
for
performing
BART
cost­
benefit
analysis
(
28­
337)
°
Analysis
ignores
impacts
of
delays
in
implementation
of
NAAQS
implementation
(
particularly
PM2.5
standards)
on
compliance
cost
assumptions
(
28­
341,
28­
342,
28­
344,
28­
345,
28­
347,
28­
348,
28­
358,
28­
387,
28­
398,
28­
405,
28­
778).
In
addition,
most
of
the
utility
plants
in
the
western
states
are
not
located
in
areas
expected
to
be
designated
non­
attainment
for
PM2.5
or
ozone
(
28­
357)
°
A
new
RIA
is
needed
to
address
the
goal
of
"
natural
conditions"
(
28­
342)
°
Neither
the
proposed
guidelines
nor
the
RIA
specifically
addresses
the
amount
of
compliance
costs
that
will
be
imposed
on
sources
in
the
25
source
categories
other
than
electric
utilities
in
the
"
fossil­
fuel
boilers"
category.
It
is
arbitrary
to
ignore
the
substantial
costs
that
would
be
imposed
on
other
types
of
industrial
sources
under
the
EPA's
proposed
emissions
unit­
by­
emissions
unit
approach
(
28­
345,
28­
349,
28­
364,
28­
748)
°
Because
BART
is
a
technology­
driven
regulation,
EPA
should
presume
for
its
economic
impact
assessment
that
all
BART­
eligible
sources
would
be
required
to
install
flue­
gas
desulfurization
(
for
coal
and
oil
combustion
sources)
to
control
SO
2,
Selective
Catalytic
Reduction
to
control
NOx,
ultra­
efficient
ESPs
or
baghouses
to
control
PM,
along
with
appropriate
controls
for
VOCs
and
ammonia
(
for
sources
using
ammonia
for
NO
x
control)
(
28­
348)
°
Neither
the
RIA
nor
the
guidelines
attempts
to
determine
how
many
sources
will
be
found
subject
to
BART
because
of
the
expansive
BART­
eligibility
scheme
set
forth
for
the
first
time
in
the
guidelines
(
28­
349)
203
°
The
RIA
reduces
the
total
costs
by
67%
to
take
account
of
the
trading
program
(
similar
to
the
Title
IV
program).
However,
the
BART
trading
program
is
complex,
and
it
is
unlikely
states
will
use
it
(
28­
349,
28­
748)
°
There
is
no
basis
for
concluding
that
only
35
utility
units
will
be
required
to
install
additional
controls
to
meet
BART.
That
conclusion
is
completely
at
odds
with
conclusions
set
forth
in
a
report
on
BART
by
NESCAUM
(
A
Basis
for
Control
of
BART­
Eligible
Sources),
which
states
at
least
365
units
will
be
required
to
set
BART
in
29
eastern
and
midwestern
states
(
28­
349)
°
By
seeking
to
apply
stringent
presumptive
controls
on
these
de
minimis
sources
of
visibility­
impairing
pollutants
(
i.
e.
emergency
generators),
the
proposed
guidelines
would
impose
significant
control
costs
without
discernible
environmental
benefits
(
28­
355)
°
EPA
determined
that
because
the
guidelines
clarify,
and
do
not
change,
existing
requirements
of
the
regional
haze
rule,
they
have
no
effect
on
the
RIA
prepared
for
the
regional
haze
rule.
However,
to
the
extent
that
the
"
clarification"
set
forth
in
the
guidelines
establishes
the
scope
of
the
costly
and
perhaps
lengthy
engineering
analysis
and
public
participation
process
to
determine
BART
for
numerous
facilities,
the
regulatory
impact
on
implementing
states
can
be
significant.
Now
that
the
proposed
guidelines
spell
out
more
clearly
EPA's
position
on
the
details
of
a
BART
analysis,
a
new
RIA
is
warranted
(
28­
361)
°
EPA
continues
to
rely
upon
its
out­
of­
date
RIA
­
a
document
prepared
under
very
different
economic
and
energy
related
circumstances.
For
example,
the
RIA
has
a
more
than
fivefold
error
with
respect
to
natural
gas
costs.
Without
better
substantiation
and
more
accurate
and
up­
to­
date
information,
EPA's
outdated
economic
impact
projections
cause
significant
concern
as
to
whether
the
EPA
conclusion
of
"
no
adverse
effect"
complies
with
the
analysis
required
by
the
President's
Executive
Order
(
28­
264,
28­
395)
°
EPA
should
use
current­
year
dollars
for
estimating
impacts
rather
than
1990
dollars
(
28­
387)
°
The
BART
guidance
purports
to
require
the
most
stringent
control
technology
not
just
for
SO
2,
but
for
all
five
visibility­
impairing
pollutants
as
well,
each
of
which
requires
distinct
and
separate
controls.
While
$
72
million
may
be
a
reasonable
total
compliance
cost
for
a
properly
designed
BART
program,
it
is
orders
of
magnitude
less
than
what
EPA's
proposal
would
cost
(
28­
747)
°
The
RIA
should
consider
the
costs
of
other
regulatory
requirements
only
if
the
sources
which
have
implemented
controls
for
those
requirements
are
provided
credit
for
the
associated
emission
reductions
relative
to
a
BART
analysis
for
regional
haze.
Consistent
with
a
"
bottom
up"
analysis,
only
incremental
reductions
and
costs
should
be
considered.
This
could
lead
to
the
determination
that
the
current
controls
actually
satisfy
BART
(
28­
7013.

Response:

As
part
of
the
regulatory
package
for
the
final
BART
rule,
EPA
has
prepared
a
revised
Regulatory
Impacts
Analysis
(
RIA).
This
analysis
has
broadened
the
scope
of
the
assessment,
204
taking
into
account
similar
issues
as
raised
by
the
commenters.
However,
as
explained
in
the
proposal
materials,
since
BART
is
guidance
to
the
States,
this
EPA
assessment
is
illustrative
of
the
costs
and
benefits
expected
to
the
nation
as
a
result
of
this
final
rulemaking.
The
details
of
the
assessment
are
found
in
the
RIA.
In
addition,
we
do
not
agree
that
the
BART
rule
limits
the
flexibility
States
were
given
in
the
CAA
for
performing
cost­
benefit
analysis.
The
CAA
clearly
lays
out
the
criteria
for
BART
determinations,
including
costs
and
visibility
improvement,
and
the
BART
guidelines
provide
much
flexibility.
See
the
preamble
for
detailed
discussion.

6.1.3
Regulatory
Flexibility
(
SBREFA)

Comment:
Five
commenters
said
EPA
had
not
fulled
its
obligations
regarding
small
entities
(
28­
251,
28­
342,
28­
345,
28­
349,
28­
364).
Commenters
said
there
will
be
substantial
impacts
on
small
businesses,
in
the
aggregates
industry
for
example
(
28­
251).
Commenter
28­
342
said
EPA
must
make
a
showing
that
the
"
large
industrial
plants"
that
are
subject
to
BART
are
in
fact
not
owned
by
small
entities.
Commenters
(
28­
345,
28­
349,
28­
364)
agreed
that
many
small
businesses
might
easily
be
major
sources
subject
to
the
rule,
especially
due
to
the
expansive
BART­
eligibility
scheme.

Commenters
(
28­
345,
28­
364)
noted
that
EPA
concludes
that
its
proposal
would
impose
requirements
on
states,
not
sources,
and
therefore
presents
no
special
small
business
issues
to
assess
and
include
in
its
RIA.
The
commenter
disagreed
with
this
reading
of
the
SBREFA
because
it
exempts
EPA
from
SBREFA
for
many
of
its
most
costly
and
significant
rules,
since
so
many
of
the
agency's
regulatory
programs
operate
by
imposing
requirements
on
the
states
that
inevitably
lead
to
controls
on
small
businesses.
Clearly,
a
substantial
number
of
small
entities
will
be
paying
higher
electricity
bills
as
a
result
of
the
impacts
of
BART
pursuant
to
these
guidelines.
Commenter
28­
242
said
the
guidelines
do
not
defer
to
state
discretion
as
EPA
claims
(
28­
342).

Response:

The
BART
rule
and
guidelines
would
not
establish
requirements
applicable
to
small
entities.
The
rule
would
apply
to
States,
not
to
small
entities.
The
BART
requirements
in
the
regional
haze
rule
require
BART
determinations
for
a
select
list
of
major
stationary
sources
defined
by
section
169A(
g)(
7)
of
the
CAA.
However,
as
noted
in
the
proposed
and
final
regional
haze
rules,
the
State's
determination
of
BART
for
regional
haze
involves
some
State
discretion
in
considering
a
number
of
factors
set
forth
in
section
169A(
g)(
2),
including
the
costs
of
compliance.
Further,
the
final
regional
haze
rule
allows
States
to
adopt
alternative
measures
in
lieu
of
requiring
the
installation
and
operation
of
BART
at
these
major
stationary
sources.
As
a
result,
the
potential
consequences
of
the
BART
provisions
of
the
regional
haze
rule
(
as
clarified
in
today's
rule)
at
specific
sources
are
speculative.
Any
requirements
for
BART
will
be
established
by
State
rulemakings.
The
States
would
accordingly
exercise
substantial
intervening
discretion
in
implementing
the
BART
requirements
of
the
regional
haze
rule
and
today's
guidelines.
205
Nonetheless,
the
EPA
has
undertaken
an
illustrative
analysis
to
assess
the
potential
small
business
impacts
of
BART
based
upon
EPA's
assessment
of
the
actions
States
may
take
to
comply
with
the
BART
rule
and
guidelines.
This
assessment
is
in
the
final
Regulatory
Impact
Analysis
(
RIA)
that
accompanies
the
final
rule.

6.1.4
Paperwork
Reduction
Act
Comment:
Two
commenters
said
a
new
and
complete
analysis
of
the
paperwork
burden
is
needed
(
28­
244,
28­
342).
Commenter
28­
342
said
EPA
seems
to
dismiss
its
need
to
comply
with
this
Act
because
the
information
collection
requirements
clarify,
but
do
not
modify,
the
information
collection
requirements
for
BART.
In
fact,
the
proposed
BART
guidelines
do
add
to
the
requirements
of
the
regional
haze
rule
so
EPA
should
comply
with
the
Paperwork
Reduction
Act.
Commenter
28­
244
said
nothing
in
the
original
visibility
rule
could
be
read
to
require
a
"
top
down"
approach
under
which
a
state
would
have
to
analyze
all
more
stringent
technologies
in
exhaustive
detail
in
order
to
justify
its
choice
of
a
technology
of
moderate
expense,
to
evaluate
all
control
alternatives
using
six
separate
charts,
to
require
states
to
survey
fourteen
different
databases
listed
by
EPA
to
determine
which
technologies
were
"
available,"
etc.

Response:

EPA
has
reviewed
its
conclusion
as
requested
by
the
commenters.
Based
on
that
review,
EPA
is
retaining
it's
conclusions
that
BART
clarifies,
but
does
not
modify
the
information
collection
requirements
for
BART.
Therefore,
this
action
does
not
impose
any
new
information
collection
burden.
However,
the
OMB
has
previously
approved
the
information
collection
requirements
contained
in
the
existing
regulations
[
40
CFR
Part
51]
under
the
provisions
of
the
Paperwork
Reduction
Act,
44
U.
S.
C.
3501
et
seq.
and
has
assigned
OMB
control
number
2060­
0421,
EPA
ICR
number
1813.04.

6.1.5
Unfunded
Mandates
Reform
Act
Comment:
Two
commenters
said
the
provisions
of
the
UMRA
do
apply
to
the
proposal
(
28­
244,
28­
342)
because
of
the
extraordinary
analytical
burdens
of
the
proposal,
and
its
bias
toward
tight
controls
(
28­
244).
Commenter
28­
342
said
EPA
should
comply
with
the
UMRA
because
the
regional
haze
rule
and
the
proposed
BART
guidelines
place
what
is
tantamount
to
a
direct
regulatory
burden
on
States
to
conduct
massive
regulatory
analyses
and
on
governmental
owners
of
fossil
fueled
electric
generating
units
to
install
90
to
95%
scrubbers
to
reduce
their
SO
2
emissions.
EPA's
arguments
that
it
has
complied
with
the
UMRA
are
unpersuasive.

Response:
206
The
EPA
is
not
reaching
a
final
conclusion
as
to
the
applicability
of
UMRA
to
today's
rulemaking
action.
The
reasons
for
this
are
discussed
in
the
1999
regional
haze
rule
(
64
FR
35762)
and
in
the
2001
BART
guidelines
proposal
(
66
FR
38111­
38112).
Notwithstanding
this,
the
discussion
in
chapter
9
of
the
RIA
constitutes
the
UMRA
statement
that
would
be
required
by
UMRA
if
its
statutory
provisions
applied.
Consequently,
we
continue
to
believe
that
it
is
not
necessary
to
reach
a
conclusion
as
to
the
applicability
of
the
UMRA
requirements.

6.1.6
Federalism
Comment:
Three
commenters
said
EPA
has
not
complied
with
the
Federalism
requirements
(
28­
342,
28­
354,
28­
364)
because
EPA
has
failed
to
obtain
meaningful
and
timely
input
from
state
and
local
officials.
Commenter
28­
342
said
that
in
order
to
be
"
meaningful,"
EPA's
consultation
with
the
State
and
local
associations
must
occur
in
connection
with
a
proposal
that
is
consistent
with
the
broad
authority
vested
in
the
States
to
implement
the
visibility
protection
program.
Commenter
28­
354
said
EPA's
proposed
BART
guidelines
impermissibly
limit
State
authority
and
unlawfully
exceed
the
Agency's
authority.
In
addition,
commenters
(
28­
354,
28­
364)
said,
the
guidelines
violate
the
"
division
of
governmental
responsibilities
between
the
national
government
and
the
States"
as
set
forth
in
the
visibility
provisions
of
the
Act
and
clearly
have
"
substantial
direct
effects
on
the
States,
on
the
relationship
between
the
national
government
and
the
States,
[
and]
on
the
distribution
of
power
and
responsibilities
among
the
various
levels
of
government."
Commenter
28­
342
agreed
that,
contrary
to
EPA's
assertions,
the
guidelines
do
impose
a
significant
analytical
burden
on
the
states
and
compliance
burdens
on
government­
owned
sources.

Response:

EPA
has
concluded
that
today's
action,
promulgating
the
BART
guidelines,
will
not
have
federalism
implications,
as
specified
in
section
6
of
the
Executive
Order
13132
(
64
FR
43255,
August
10,
1999),
because
it
will
not
have
substantial
direct
effects
on
the
States,
nor
substantially
alter
the
relationship
or
the
distribution
of
power
and
responsibilities
between
the
States
and
the
Federal
government.
Nonetheless,
we
consulted
with
a
wide
scope
of
State
and
local
officials,
including
the
National
Governors
Association,
National
League
of
Cities,
National
Conference
of
State
Legislatures,
U.
S.
Conference
of
Mayors,
National
Association
of
Counties,
Council
of
State
Governments,
International
City/
County
Management
Association,
and
National
Association
of
Towns
and
Townships,
during
the
course
of
developing
this
rule.

6.1.7
Executive
Order
13211
­
Actions
Concerning
Regulations
that
Significantly
Affect
Energy
Supply,
Distribution
or
Use
Comment:
207
XX
commenters
said
EPA's
analysis
was
inadequate
for
several
reasons
(
28­
337,
28­
343,
28­
345,
28­
346,
28­
348,
28­
354,
28­
356,
28­
357,
28­
361,
28­
364,
28­
387,
28­
395,
28­
406,
28­
7001):

°
It
was
erroneously
based
on
the
impacts
of
SO
2
controls
alone
(
28­
337,
28­
354,
28­
361,
28­
364,
28­
395,
28­
7001)
°
The
analysis
relies
on
a
flawed
RIA
including
assumptions
that
costs
will
have
already
been
undertaken
for
the
PM2.5
standard
and
the
energy
costs
are
substantially
understated
(
28­
346,
28­
354,
28­
364,
28­
387)
°
It
is
unclear
whether
the
analysis
took
into
account
the
price
spikes
for
energy
experienced
in
late
2000
(
28­
361)
°
EPA
did
not
complete
the
analysis
correctly
and
negative
energy
impacts
may
result
from
implementing
BART
as
required
(
28­
7001)
°
It
ignored
the
effect
on
the
supply
of
energy,
could
result
in
shut­
downs
and
further
decreases
in
supply
and
should
evaluate
effects
on
the
reserve
margin
and
the
ability
of
electricity
providers
to
supply
enough
energy
to
meet
demand
(
28­
342,
28­
346,
28­
354)
°
EPA
should
conduct
a
more
meaningful
and
thorough
analysis
of
the
proposed
rule's
impact
on
energy
prices,
supply,
distribution
and
use
before
issuing
the
final
rule
(
28­
345,
28­
364,
28­
406).
Commenters
were
concerned
that
EPA
rushed
its
analysis
and
that
the
significant
energy
impacts
to
companies
are
yet
to
be
understood.
°
In
meeting
the
requirement
of
EO
13211,
EPA
concludes
that
the
"
rulemaking
is
not
a
significant
energy
action"
and
that
"[
the
BART
rules]
clarify,
and
do
not
change,
the
existing
requirements
of
the
Regional
Haze
Rule".
However,
in
other
parts
of
the
proposed
rulemaking,
EPA
alternatively
concludes
that
the
BART
rulemaking
is
a
"
significant
regulatory
action",
which
clearly
is
the
case.
(
28­
356)
°
The
economic
rationale
that
EPA
uses
to
justify
the
guidelines
should
be
re­
done
so
that
it
more
accurately
reflects
the
actual
costs
that
will
be
borne
by
western
ratepayers
as
well
as
better
illuminating
the
potential
impacts
to
the
western
electricity
grid
consistent
with
Executive
Order
13211
(
28­
357)
°
The
environmental
impact
and
the
disposal
cost
of
that
solid
waste
should
also
be
addressed
in
cost
calculations
as
should
CO
2
impacts
(
28­
406).

Response:

As
part
of
the
regulatory
package
for
the
final
CAVR
rule,
EPA
has
prepared
a
revised
Regulatory
Impacts
Analysis
(
RIA).
This
analysis
has
broadened
the
scope
of
the
assessment,
taking
into
account
similar
issues
as
raised
by
the
commenters.
However,
as
explained
in
the
proposal
materials,
since
CAVR
is
guidance
to
the
States,
this
EPA
assessment
is
illustrative
of
the
costs
and
benefits
expected
to
the
nation
as
a
result
of
this
final
rulemaking.
The
details
of
the
assessment
are
found
in
the
RIA.

7.0
Miscellaneous
Comments
208
7.1
Process
Related
Comment:
Five
comments
discussed
process­
related
issues
(
0189,.
0208,
0259,
0285,
0302).

Commenter
0189
said
the
implementing
mechanism
for
BART
is
the
SIP
process.
Congress'
choice
of
the
SIP
vehicle,
therefore,
demonstrates
a
legislative
intent
that
EPA
should
get
at
most
one
chance
to
review
state
judgments
of
BART
applicability,
namely
when
it
reviews
the
SIP
for
approval
or
disapproval.
Such
a
reading
is
also
the
only
one
consistent
with
sound
policy.
Installation
of
BART
is
a
one­
time
event
that
must
take
place
within
5
years
of
SIP
approval
under
the
statutory
framework.
Congress
intended
BART
as
a
down
payment
on
the
"
reasonable
progress"
target.
The
proper
approach,
from
the
standpoint
of
both
law
and
policy,
would
be
to
address
any
need
for
additional
controls
after
the
first
round
of
BART
as
part
of
the
"
reasonable
progress"
obligation.

Response:

The
BART
process
will
follow
the
established
process
for
SIP
review.
We
are
not
establishing
a
new
process
in
this
rulemaking.

Comment:
Commenter
0208
said
EPA
should
encourage
voluntary
emission
reduction
projects
by
specifically
exempting
such
plants
from
additional
BART
analysis,
specially
in
instances
where
controls
were
voluntarily
installed
at
significantly
more
stringent
levels
than
that
represented
by
a
BART
limit.
The
commenter
provided
examples
where
state
legislatures
have
made
findings
that
voluntary
measures
met
BART.

Response:

We
have
included
a
provision
providing
additional
flexibility
along
these
lines
for
sources
who
adopt
the
most
stringent
controls
available.

Comment:
Commenter
0259
noted
that
the
WRAP
is
developing
an
alternative
and
more
efficient
approach
to
the
one
described
in
the
proposed
BART
guidelines
to
address
sources
in
California.
Specifically,
the
WRAP
is
identifying
which
types
of
BART
sources
(
by
category)
are
located
in
which
air
pollution
control
districts,
and
then
examining
the
rules
and/
or
permit
levels
applicable
to
those
categories
to
determine
if
BART
is
being
satisfied
.
In
most
if
not
all
cases,
this
approach
should
obviate
the
need
to
explicitly
identify
BART­
eligible
sources
and
to
conduct
further
BART
analyses.
If,
however,
a
BART
category
is
present
in
a
district
whose
rules
may
not
satisfy
BART
for
one
or
more
pollutants,
then
the
standard
approach
in
the
guidelines
would
be
followed.
This
approach
could
also
be
used
in
other
states
when
addressing
BART­
eligible
source
emissions
which
are
well
controlled.
The
commenter
urged
EPA
to
ensure
that
the
final
guidelines
allow
209
states
and
tribes
to
use
an
approach
such
as
the
one
described
above
to
demonstrate
compliance
with
the
BART
provisions
of
the
Regional
Haze
Rule.

Response:

We
note
that
any
such
approach
would
still
need
to
meet
the
requirements
of
CAA
section
169A.

Comment:
Commenter
0285
asked
EPA
to
prioritize
sources
for
control
and
consider
the
"
resource
vs.
environmental
benefit"
ratio
when
prioritizing.
In
particular,
EPA
should
examine
how
resources
should
be
devoted
to
sources
with
fugitive
dust,
for
example,
plant
roads,
storage
piles
and
material
handling
activities.
These
sources
emit
only
PM,
the
emissions
levels
are
generally
low,
and
the
fraction
of
fine
PM
is
small.
Priority
should
be
given
to
the
high
level
emitters
of
SO
2
and
NO
X.,
using
air
quality
modeling
or
an
alternative
analysis,
which
may
determine
within
what
distance
of
a
Class
I
area
these
smaller
sources
may
significantly
affect
the
visibility
and
may
require
that
the
fugitive
dust
sources
within
that
distance
should
be
considered
for
control.
EPA
also
should
prioritize
for
control
the
emissions
units
in
the
general
source
categories
such
as
the
lime
plants,
portland
cement
plants,
iron
and
steel
mill
plants
and
chemical
process
plants.
These
plants
have
a
large
number
of
emission
units,
many
of
which
are
considered
insignificant.
For
example,
cement
plants
contain
numerous
operations,
but
it
is
the
kilns
that
are
typically
the
sole
source
of
PM,
SO
2,
NO
X
and
VOC
emissions
and
account
for
a
large
percentage
of
the
total
PM
emissions.
Therefore,
at
cement
plants,
unless
they
are
close
to
Class
I
area,
the
priority
for
control
should
be
focused
on
kilns
only.
Emission
units
at
these
plants,
identified
by
source
classification
codes
contained
in
EPA's
Factor
Information
Retrieval
(
FIRE)
Data
System,
should
be
prioritized
for
BART
control
so
that
the
largest
ones
are
the
focus.

Response:

This
comment
is
best
directed
to
States
implementing
the
BART
provisions.

Comment:
Commenter
0302
noted
that
in
the
proposed
BART
guidelines,
EPA
presents
alternative
methods
for
determining
which
sources
would
be
BART­
subject
sources,
which
sources
should
be
controlled,
and
how
to
assess
the
impact
from
individual
sources.
More
in­
depth
criteria
and
information
for
options
is
needed
to
make
a
judgment
on
the
best
approach.

Response:

As
explained
earlier,
States
have
flexibility
to
make
these
judgements
in
a
variety
of
ways,
so
long
as
they
consider
the
BART
factors.

7.2
Definition
of
Terms
210
Comment:
Commenter
0267
said
anytime
emission
limits
are
discussed
in
the
rule,
the
averaging
period
should
be
specified
so
that
comparisons
with
other
limits
are
made
simpler,
and
compliance
demonstration
methods
are
transparent.

Commenters
(
28­
267,
28­
272,
28­
278,
28­
332,
28­
349,
28­
255,
28­
266,
28­
273,
28­
306)
said
EPA
should
define
or
clarify
the
following
terms
°
"
substantial
fraction"
°
"
significant
contribution"
°
"
sufficient
visibility
improvement"
°
"
region"
°
"
geographic
region"
(
what
methods
should
RPOs
use
to
determine
the
spatial
extent
of
the
region,
how
much
flexibility
do
RPOs
have,
how
boundaries
of
the
geographic
regions
that
"
generally
extend
for
hundreds
or
thousands
of
miles"
would
be
reasonably
limited.)
°
"
can
be
emitted
and
transported"
(
is
it
necessary
to
show
that
at
least
one
back
trajectory
on
the
20%
worst
visibility
days
from
a
Class
I
pass
through
the
county
or
state
in
which
a
given
source
is
located)
°
"
trivial
contribution"
(
what
is
the
smallest
acceptable
geographic
area
to
be
"
zeroed
out,"
or
the
amount
of
difference
between
the
base
run
and
"
zero
out"
run
that
is
considered
to
be
"
trivial",
is
it
the
same
as
"
significant
contribution"
or
different?
Commenter
28­
255
supports
0.1
deciview
as
an
appropriate
value
and
only
one
term
should
be
used
to
denote
"
significant"
impact)
°
"
non­
trivial
contribution"
°
"
contiguous
and
adjacent
properties"
in
section
II(
A)
°
"
emissions
unit"
(
the
range
of
possible
items
that
EPA
has
treated
as
emissions
units
at
a
plant
site
has
varied
from
a
single
emissions
point
to
a
large,
integrated
collection
of
equipment
consisting
of
numerous
emissions
points,
moving
from
the
known
definition
of
"
major
stationary
source"
to
the
undefined,
yet
proposed,
"
emissions
unit"
basis
will
prove
confusing
for
states
in
the
implementation
process.)
°
"
installation"
when
it
refers
to
"
building,
structure,
facility,
or
installation"

Response:

Many
of
these
require
State­
specific
judgement;
others
are
defined
in
other
EPA
programs
or
explained
earlier
in
this
document.

Comment:
Commenter
28­
306
said
the
proposed
rule
encourages
states
to
use
the
NSR
definitions
of
"
potential
to
emit"
and
"
physical
or
operational
change".
These
NSR
concepts
have
led
to
extensive
compliance
problems,
and
resulted
in
serious
controversy
and
deliberations
in
NSR
revisions.
As
a
result
of
this
debate,
definitions
are
likely
to
be
changed
in
NSR
revisions,
and
would
be
out
of
date
in
BART
guidelines
perhaps
in
the
near
future.
211
Response:

We
have
explained
above
that
these
are
the
current
operative
definitions,
which
are
the
only
definitions
it
is
reasonable
for
us
to
use.

7.3
Role
of
Federal
Land
Managers
Comment:
Sixteen
commenters
discussed
the
role
of
FLMs
(
0147,
0148,
0150,
0159,
0162,
0163,
0164,
0214,
0221,
0227,
0238,
0267,
0284,
0292,
0304,
0318).

Some
commenters
said
that
the
FLMs
should
be
active
participants
in
the
BART
process
(
0147,
0148,
0150,
0159,
0164,
0221,
0227,
0267,
0284,
0292,
0304,
0318).
Commenters
(
0221,
0227)
added
that
this
is
consistent
with
Congress'
intent,
but
was
concerned
that
EPA's
proposed
rule
deprives
the
FLMs
of
any
meaningful
participation
in
the
BART
process.
Given
the
almost
unlimited
exemption
authority
granted
to
states
under
the
proposed
rule,
the
consultation
with
FLMs
during
the
SIP
process
will
occur
after
the
critically
important
decisions
about
the
scope
and
rigor
of
the
state's
visibility
protection
program
have
already
been
made.
This
is
a
reversal
in
EPA's
long­
standing
interpretation
of
the
FLMs'
role
is
without
reasoned
basis
and
contrary
to
law.
Commenter
0267
noted
that
FLMs
should
retain
the
ability
to
review
the
decisions
for
all
BART­
eligible
source
exemptions
under
any
of
the
approaches.
Commenter
0304
said
FLMs
and
states
should
retain
the
authority
to
certify
impairment
at
specific
Class
I
areas
which
may
be
"
reasonably
attributable"
to
sources
subject
to
the
IAQR.
This
is
necessary
because
some
sources
will
likely
have
large
local
impacts,
and
regional
controls
may
prove
to
be
less
effective
for
some
Class
I
areas.

Two
commenters
(
0214,
0238)
were
concerned
that
the
BART
Guidelines
will
be
used
by
FLMs
to
implement
the
more
aggressive
FLAG
guidance
instead
of
the
more
flexible
BART
guidelines.
The
guidelines
must
stress
that
the
states
retain
their
flexibility
to
implement
the
BART
program
in
their
discretion
and
that
the
FLMs
have
only
an
input
role
in
the
process.

Response:

FLMs
retain
the
above­
mentioned
authorities
and
we
believe
they
play
an
important
role
in
protecting
Class
I
areas.

7.4
Comments
about
Specific
Class
I
Areas
Comment:
Nine
commenters
provided
information
on
specific
Class
I
areas,
illustrating
the
need
to
improve
visibility,
protect
plant
life,
protect
aquatic
life,
result
in
a
healthy
park
experience,
protect
the
tourism
economy,
etc.
(
0157,
0158,
0162,
0164,
0166,
0227,
0305,
0313,
0318).
212
Commenters
(
0225,
28­
242)
questioned
whether
Volcanoes
National
Park
should
even
be
considered
part
of
the
Regional
Haze
program
given
that
the
Kilauea
Volcano
is
such
an
enormous
known
source
of
naturally
occurring
SO
2.
Commenter
28­
242
said
natural
emissions
are
responsible
for
the
visibility
impacts
and
the
regional
haze
rules
should
not
apply
in
this
situation.

Commenter
28­
399
said
applicable
public
lands
should
include
those
designated
as
national
parks
and
wilderness
areas
after
1977,
including
PA's
Class
II
Valley
Forge
National
Park.

Response:

We
appreciate
the
additional
information
on
specific
Class
I
areas.
We
note
that
changing
the
list
of
Class
I
areas
is
outside
the
scope
of
the
BART
rulemaking.

7.5
Other
Miscellaneous
Comments
Comment:
Three
commenters
made
other
comments
on
the
proposal
(
0169,
0173,
0256).
Commenter
0169
said
states
should
consider
unintended
consequences
on
regional
haze
during
the
evaluation
of
BART
requirements
and
provided
an
example
of
development
of
a
detached
NO
x
plume
resulting
from
changes
in
a
power
plant's
chemistry
due
to
the
installation
of
scrubbers.
The
plume
created
haze
over
the
Grand
Canyon.
Commenter
0173
said
biodiesel
blends
would
be
a
beneficial
and
applicable
retrofit
control
option
that
could
be
included
in
the
final
regulations
and
guidelines
and
described
their
uses
and
benefits.
Commenter
0256
asked
EPA
to
neither
consider
nor
respond
to
comments
submitted
in
2001
by
the
commenter
because
many
of
these
are
no
longer
relevant.

Response:

We
appreciate
this
additional
information.

Comment:

Commenter
28­
242
said
EPA
should
assist
Hawaii's
utilities
in
maintaining
air
quality
while
providing
reliable
power
at
a
reasonable
cost
to
its
customers.

Response:

This
comment
is
outside
the
scope
of
the
BART
rulemaking.

Comment:
Commenter
28­
322
said
emissions
reductions
may
not
be
realized
in
a
timely
manner
if
the
legal
foundation
of
BART
is
open
to
challenge.
EPA
should
minimize
the
possibility
of
legal
challenges
late
in
the
process.
EC/
R
Incorporated
BART
Public
Comment
Summary
Draft
 
do
not
quote
or
cite
March
18,
2005
213
Response:

We
believe
the
BART
guidelines
meet
the
legal
requirements
of
the
CAA.

8.0
Editorial
Comments
Comment:
Commenter
0285
suggested
changes
to
conform
with
the
table
of
contents.

Response:
We
have
tried
to
ensure
consistency
in
drafting
this
final
rulemaking.

9.0
Comments
Unrelated
to
BART
Comment:
Two
commenters
provided
comments
unrelated
to
BART
in
that
they
discussed
EPA's
New
Source
Review
policy
(
0149)
and
the
need
to
finalize
mercury
pollution
rules
for
power
plants
(
0318).
Commenter
28­
564
was
concerned
about
EPA
enforcement
of
power
plants
who
upgrade
old
facilities
without
installing
pollution
control
measures
as
well.

Response:

This
comment
is
outside
the
scope
of
the
BART
rulemaking.

4.2.3
Presumptive
Emission
Limits
4.2.3.1
SO
2
Limits
for
Utility
Boilers
Comment:
Thirty­
four
commenters
said
the
proposed
SO
2
levels
are
reasonable
(
0167,
0171,
0227,
0231,
0232,
0234,
0255,
0256,
0267,
0271,
0314,
0323,
28­
239,
28­
345,
28­
247,
28­
248,
28­
249,
28­
250,
28­
258,
28­
260,
28­
262,
28­
286,
28­
289,
28­
302,
28­
334,
28­
338,
28­
350,
28­
352,
28­
362,
28­
399,
28­
433,
28­
7002,
28­
7016,
28­
7048)
citing
examples
and
references
in
support
of
these
levels
(
0171,
0255,
0256,
28­
302).
Commenter
0323
also
agreed
with
EPA's
approach
of
either
a
specific
emission
rate
limit
or
level
of
control
for
sulfur
dioxide
that
helps
to
provide
equity
among
sources
which
use
higher
sulfur
coal
and
those
using
coal
with
inherently
lower
sulfur
content.

Response:
The
EPA
appreciates
these
comments.
214
Comment:
Commenters
(
28­
248,
28­
399,
28­
7016)
said
EPA
should
expand
this
presumption
by
applying
it
to
electric
generating
facilities
that
are
under
controlled
for
purposes
of
SO
x.
Commenter
28­
258
said
the
presumptive
BART
level
of
SO
2
controls
should
be
required
at
all
utility
boilers,
not
just
those
with
no
SO
2
controls.

Response:
The
final
BART
presumption
for
coal­
fired
EGUs
with
a
pre­
existing
SO2
controls
are
to
be
determined
by
States
based
on
an
analysis
of
cost­
effective
FGD
upgrades
designed
to
improve
the
average
SO2
removal
efficiency
of
the
existing
system.
The
EPA
is
strongly
encouraging
States
to
the
implement
the
BART
presumptive
limits
on
coal­
fired
EGUs
greater
than
200
MW
and
on
all
oil­
fired
EGUs
regardless
of
capacity.
See
TSD
"
Setting
BART
Presumptive
SO2
Limits
for
Electric
Generating
Units"
for
a
more
detailed
discussion
on
how
EPA
determined
its
final
recommendations.

Comment:
Nine
commenters
said
the
proposed
levels
are
too
stringent
(
0176,
0181,
0217,
0246,
0253,
28­
264,
28­
341,
28­
357,
28­
7007).
Commenter
28­
264
said
the
level
is
too
high
for
western
utilities
that
are
burning
low­
sulfur
coal
where
a
level
of
85­
90%
would
be
more
appropriate.
One
commenter
(
0176)
cited
his
experience
with
eastern
bituminous
waste
and
did
not
believe
the
proposed
emissions
reductions
were
possible.
Commenter
0253
said
any
default
SO
2
control
level
mandating
greater
than
90%
control
are
inappropriate
for
units
utilizing
wet
scrubbers
and
high
sulfur
coals.
Commenter
28­
7007
suggested
a
range
of
70
to
90%
would
be
appropriate.

Response:
The
EPA
is
finalizing
the
SO2
BART
presumptive
limits
based
on
current
FGD
capabilities
given
the
various
combinations
of
coal
types
and
sulfur
contents
inherent
in
the
BART­
eligible
population
of
units.
The
EPA
is
maintaining
both
an
emission
limit
and
percent
removal
in
the
final
BART
presumptive
limits
to
provide
equity
among
various
unit
specific
characteristics
and
the
removal
capabilities
of
different
FGD
retrofit
systems.
We
also
believe
that
units
burning
higher
sulfur
coals
can
utilize
wet
scrubbers
systems
to
achieve
removal
efficiencies
of
at
least
95
percent
and
units
burning
lower
sulfur
coals
can
utilize
semi­
dry
FGD
systems
to
achieve
an
emission
rate
limit
of
0.15
lb
SO2/
mmBtu
or
less.

Comment:
Commenters
(
0217,
0246,
28­
357)
said
if
presumptive
levels
are
included,
consideration
should
be
given
to
the
levels
reflected
in
the
WRAP
agreement
(
85%
control
efficiency
at
85%
capacity
factor.)
EPA's
presumption
of
90
to
95%
control
efficiency
contradicts
the
consensus
based
level
of
SO
2
reduction
determined
by
the
WRAP
found
to
be
necessary
for
visibility
improvement
in
western
states.
Moreover,
given
that
the
WRAP
stakeholder
agreement
was
determined
to
be
better
than
BART,
the
proposed
guidelines
should
not
be
interpreted
to
require
changes
in
Section
309
SIPs
already
submitted
under
the
regional
haze
rules.
215
Response:
The
WRAP
agreement
covers
a
much
broader
universe
of
emission
sources
than
these
presumptive
limits,
including
many
smaller
sources.
We
believe
the
presumptive
limits
we
have
provided
are
strongly
supportable
as
BART
limits
for
the
large
EGUs
covered.
See
the
preamble
for
further
justification.

Comment:
Commenter
0181
asked
EPA
to
remove
the
presumptive
SO
2
control
level
from
the
final
BART
guidelines.
Instead,
there
should
be
a
BART
determination
process
that
promotes
costeffective
emission
reductions
based
on
specific
boiler
design
features
and
site­
specific
constraints.
The
proposed
presumptive
BART
approach,
as
proposed,
does
not
provide
affected
states
and
sources
with
the
flexibility
needed
to
design
and
install
SO
2
emission
control
systems
in
a
costeffective
manner
to
meet
the
BART
requirements.
Furthermore,
the
presumptive
SO
2
emissions
control
approach
inappropriately
ignores
the
need
for
a
visibility
impact
evaluation
which
is
required
in
step
5
of
the
proposed
case­
by­
case
BART
engineering
analysis.

Response:
The
EPA
believes
the
final
BART
presumptive
limits
promote
highly
cost­
effective
SO2
emission
reductions.
The
EPA
is
maintaining
both
an
emission
limit
and
percent
removal
in
the
final
SO2
BART
presumptive
limits
to
provide
equity
among
unit
specific
characteristics
(
i.
e.
boiler
design,
coal
types,
sulfur
content,
unit
size,
etc.)
and
potential
FGD
retrofit
systems.
We
believe
our
SO2
presumptive
limits
provide
States
enough
flexibility
to
take
into
account
unit
specific
circumstances.
States
have
the
ability
to
establish
alternate
BART
presumptive
limits
based
on
a
five­
step
BART
analysis
given
the
presence
of
persuasive
evidence.
See
Section
7
part
a,
Visibility
Analysis
for
SO2
and
NOx
emissions
from
EGUs,
in
BART
Preamble.

Comment:
Commenter
28­
341
said
through
switching
to
Powder
River
Basin
(
PRB)
coal,
many
of
Ameren's
facilities
have
met
and
exceeded
the
SO
2
reduction
requirements
of
the
Acid
Rain
program.
Should
BART
guidelines
be
adopted,
however,
they
will
be
forced
to
install
scrubbers.
at
each
one
of
their
facilities
that
employs
the
use
of
low­
sulfur
coal
to
meet
the
proposed
90­
95%
control
levels.

Response:
The
EPA
anticipates
States
to
adopt
the
BART
presumptions
on
all
coal­
fired
BARTeligible
EGUs
greater
than
200
MW.
Analysis
conducted
by
EPA
suggest
that
units
burning
lower
sulfur
coals
and
PRB
coal
can
cost­
effectively
retrofit
with
FGD
systems
to
meet
the
presumptive
limits.

Comment:
Four
commenters
said
the
proposed
presumptive
control
range
is
not
achievable
for
existing
boilers,
and
is
even
more
stringent
than
NSPS
limits
for
new
sources
(
28­
403,
28­
404,
28­
405,
28­
7013).
Commenter
28­
403
said
EPA's
data
do
not
support
such
a
limit
nor
does
EPA
216
explain
why
the
presumption
is
needed
and
why
the
states'
discretion
should
be
preempted.
Commenter
28­
404
said
this
presumption
will
result
in
excessive
costs
for
older
equipment.
Commenter
28­
405
said
if
presumptive
control
levels
were
to
be
established,
they
should
start
at
the
minimum
level
of
emission
controls,
with
additional
controls
justified
on
the
basis
of
the
source­
specific
balancing
process
Congress
set
forth
in
169A.
Commenter
28­
7013
added
that
the
presumptive
limits
are
more
consistent
with
BACT,
which
is
contrary
is
contrary
to
the
Act's
clear
distinction
between
BART
and
BACT.

Response:
The
EPA
has
recently
proposed
to
revise
the
NSPS
to
represent
the
SO2
control
capabilities
which
are
available
today.
In
the
proposed
NSPS
amendments
EPA
proposed
a
single
limit
without
a
compliance
alternative,
such
as
a
percent
removal
as
allowed
in
BART.
In
setting
the
NSPS
EPA
set
a
higher
emission
limit
than
proposed
in
BART
to
account
for
account
for
the
need
of
future
fuel
diversity,
in
particular
higher
sulfur
coals.
Conversely,
the
BART
presumptions
were
established
based
on
the
defined
population
of
eligible
units
and
were
determined
to
be
achievable
by
all
eligible
units.
See
the
preamble
for
further
justification.

Comment:
Eleven
commenters
said
that
the
proposed
presumptive
limits
clearly
exceed
the
statutory
authority
(
28­
270,
28­
342,
28­
353,
28­
354,
28­
356,
28­
360,
28­
386,
28­
390,
28­
392,
28­
778,
28­
7004).
Commenters
(
28­
270,
28­
386,
28­
392,
28­
778)
said
that
the
evaluation
must
be
conducted
according
to
the
statutory
process
and
BART
must
be
determined
on
a
case­
by­
case
basis.
Commenters
(
28­
342,
28­
354,
28­
356,
28­
360,
28­
390)
objected
to
EPA's
proposed
presumptive
limits
for
SO
2
controls
as
a
flagrant
example
of
EPA
usurping
State
authority
and/
or
as
being
arbitrary
and
economically
counterproductive.
Commenter
28­
7004
added
that
the
Act
consciously
addresses
the
need
for
SO
2
scrubbing
in
Title
IV.
The
Act
does
not
envision
or
authorize
a
much
more
onerous
and
expensive
nationwide
SO
2
scrubbing
program
as
part
of
Section
169A.
The
"
guidance"
which
presumes
90
to
95%
SO
2
removal
as
BART
for
uncontrolled
utility
boilers
is
not
supported
by
the
law.

Response:
See
the
preamble.

Comment:
Three
commenters
said
the
proposed
levels
are
too
lenient
(
0221,
28­
263,
28­
7008).
Commenters
(
0221,
28­
263)
said
because
of
the
extended
SIP
timeline,
BART
retrofits
will
not
be
applied
until
2013
and
therefore,
it
is
expected
that
the
"
best"
available
sulfur
removal
technologies
will
routinely
be
able
to
achieve
in
excess
of
95%
removal
efficiencies
and
emission
rates
of
lower
than
0.10
lb/
MMBTU
by
that
time.
Therefore,
the
presumptive
level
of
95%
reduction
or
0.10­
0.15
lb/
MMBTU
SO2
emission
rate
that
is
stipulated
in
the
proposal
should
be
strengthened
to
an
SO
2
removal
efficiency
of
95%
or
better
or
an
emission
rate
of
0.10
lb/
MMBTU
or
better.
Commenter
28­
7008
was
concerned
with
the
effect
that
setting
presumptive
minimums
would
have
in
a
true
top­
down
control
technology
evaluation.
Using
an
artificially
low
presumptive
minimum
217
could
eliminate
consideration
of
more
effective
control
options
that
are
still
achievable
by
a
given
source.

Response:
The
EPA
believes
the
final
BART
SO2
presumptive
limits
represent
the
removal
efficiencies
of
current
FGD
systems
given
the
identified
universe
of
BART­
eligible
sources.
The
final
BART
presumptive
limits
require
a
removal
efficiency
of
at
least
95
percent
or
control
to
an
emission
rate
of
0.15
lb
SO2/
mmBtu
or
less.
States
have
discretion
to
set
more
stringent
presumptive
limits
based
on
the
findings
of
the
five­
step
BART
analysis
process.

Comment:
Thirteen
commenters
specified
that
proposed
control
level
must
account
for
fuel
type
and/
or
coal
sulfur
content
(
0171,
0187,
0206,
0213,
0224,
0236,
0241,
0255,
0303,
0314,
28­
335,
28­
347,
28­
770).
Commenter
0213
presented
historical
data
showing
that
the
proposed
emissions
limit
could
not
be
attained
in
certain
cases
depending
upon
the
sulfur
content
of
the
fuel
being
burned.
A
statistical
approach
should
be
utilized
analyzing
the
performance
of
existing
FGD
installations
to
ensure
that
the
worst
case
performers
can
reasonably
be
expected
to
comply
with
the
limit.
Commenters
(
28­
335,
28­
347,
28­
770)
said
the
presumption
should
not
apply
to
oil
or
gas­
fired
utility
boilers.

Response:
The
EPA
agrees
with
the
commenters
and
reiterates
that
in
establishing
the
BART
presumptive
limits
EPA
accounted
for
variations
in
fuel
type,
fuel
sulfur
content,
existing
SO2
control
systems,
anticipated
unit
cost­
effectiveness
values
and
the
removal
capabilities
of
current
FGD
systems.
The
EPA
is
finalizing
BART
presumptive
limits
which
include
an
emission
rate
limit
and
a
percent
removal
requirement
which
it
feels
to
be
reasonably
achievable
by
all
coal­
fired
BART­
eligible
units
absent
extreme
circumstance.
The
EPA
is
finalizing
SO2
BART
presumptive
limits
for
oil­
fired
EGUs
but
not
gas­
fired
EGUs.

Comment:
Commenters
(
0187,
0226,
0224,
0236,
0241,
0303)
said
the
rationale
used
to
justify
the
default
and
presumptive
BART
SO
2
emission
limitations
for
coal­
fired
units
can
be
applied
to
oilfired
EGUs.
In
contrast
to
coal­
fired
electric
generating
units,
flue
gas
de­
sulfurization
has
not
historically
been
used
at
oil­
fired
facilities
to
comply
with
SO
2
emission
limitations.
To
the
commenter's
knowledge,
there
are
no
existing
oil­
fired
steam
electric
generating
units
in
the
U.
S.
equipped
with
flue
gas
de­
sulfurization
for
control
of
SO
2
emissions.
The
primary
reason
for
this
is
that
the
vast
majority,
if
not
all,
of
the
states
limit
the
sulfur
content
of
the
oil
that
may
be
burned
at
EGUs
as
part
of
their
SIP
strategy
for
complying
with
the
SO
2
NAAQS.

Response:
Analyses
conducted
by
the
EPA
in
determining
the
presumptive
BART
limitations
for
coaland
oil­
fired
EGUs
were
based
on
emissions
data,
anticipated
unit
cost­
effectiveness
values
and
current
capabilities
of
SO2
control
technologies
or
methods
and
their
industry
applicability.
218
The
EPA
agrees
with
the
commenters
suggestion
pertaining
oil­
fired
units
and
notes
the
final
BART
presumptions
for
oil­
fired
units
is
a
limitation
on
fuel
oil
sulfur
content.

Comment:
Commenter
0314
suggested
setting
the
presumptive
BART
as
a
range
(
0.1
to
0.15
lb/
MMBTU)
allowing
the
State
to
establish
a
specific
level
for
each
plant.

Response:
The
final
BART
presumptive
limits
are
to
be
established
on
a
unit­
by­
unit
basis
by
the
States.
States
have
the
ability
to
establish
a
SO2
removal
efficiency
of
at
least
95
percent
or
an
emission
rate
limit
of
0.15
lb
SO2/
mmBtu
or
less.

Comment:
Three
commenters
said
that
presumptive
controls
may
need
to
differ
depending
on
the
location
of
the
source
(
0244,
0252,
0306).
Two
commenters
(
0244,
0252)
said
the
presumptive
control
level
of
90%
for
western
units
is
or
may
be
unrealistic.
In
contrast
to
the
presumptive
level
proposed
by
EPA
of
90%
SO
2
reduction,
WRAP
has
conducted
analysis
that
the
presumptive
level
of
control
for
a
scrubber
in
the
west
should
be
set
at
85%.
Commenter
0306
said
EPA
needs
to
consider
the
cold
climate
of
certain
northern
states
when
determining
presumptive
BART
for
SO
2.
Cold
weather
can
cause
icing
of
very
wet
stacks
which
can
lead
to
safety
hazards
and
functional
problems.
Reheat
of
the
stack
may
be
necessary,
to
reduce
or
eliminate
icing
problems.
Reheating
the
flue
gas
by
bypassing
some
of
the
gas
around
the
scrubber
will
reduce
overall
SO
2
removal.
The
proposed
presumptive
BART
may
not
be
achievable
in
cold
climates
without
overly
burdensome
costs.

Response:
The
WRAP
agreement
covers
a
much
broader
universe
of
emission
sources
than
these
presumptive
limits,
including
many
smaller
sources.
We
believe
the
presumptive
limits
we
have
provided
are
strongly
supportable
as
BART
limits
for
the
large
EGUs
covered.
In
addition,
a
State's
analysis
of
the
BART
factors
can
take
specific
circumstances
into
account.
See
the
preamble
for
further
explanation.

Comment:
Nine
commenters
said
that
the
proposed
control
levels
must
account
for
the
type
of
scrubber
in
place
(
0171,
0210,
0228,
0229,
0241,
0250,
0253,
0254,
0256).
Commenter
described
the
need
for
site­
specific
analyses
to
address
scrubbers
in
place
(
0210,
250).
Commenters
(
0241,
0254)
added
that
EGUs
that
currently
have
SO
2
controls
should
not
be
forced
to
meet
the
presumptive
level
unless
a
full
BART
analysis
is
undertaken
which
shows
that
incremental
visibility
and
cost­
effectiveness
justify
forcing
controlled
EGUs
to
meet
the
more
stringent
limit.
Others
(
0229)
said
that
the
recent
addition
of
controls,
even
if
reductions
fall
below
the
presumptive
levels,
should
be
cause
for
allowing
states
to
screen
such
a
BART­
eligible
source
out
early
in
the
evaluation
process.
Commenter
0254
said
the
ability
to
upgrade
existing
scrubber
units
will
be
limited
and
for
many,
particularly
dry
scrubbers,
the
presumptive
95%
reduction
level
will
not
be
achievable.
In
addition,
many
of
these
units
have
been
installed
recently,
e.
g.,
since
the
1990
Clean
219
Air
Act
Amendments,
and
requiring
these
facilities
to
re­
visit
large
investments
in
controls,
after
such
a
short
interval,
imparts
an
excessive
burden
on
those
businesses.

Response:
The
BART
presumptions
for
EGUs
with
pre­
existing
SO2
controls
are
to
be
determined
by
the
States,
as
either
an
emission
rate
limit
or
a
percent
removal,
based
on
a
unit­
specific
analysis
of
potential
system
upgrades
designed
to
increase
the
average
SO2
removal
capability
of
the
preexisting
system.
See
the
preamble
for
further
explanation.
The
EPA
recognizes
that
semi­
dry
FGD
systems
are
not
typically
designed
to
achieve
removal
efficiencies
of
at
least
95
percent.
For
this
reason,
the
EPA
anticipates
smaller
units
burning
low
sulfur
coals
could
utilize
the
semi­
dry
FGD
technologies
to
achieve
an
emission
rate
of
0.15
lb
SO2/
mmBtu.
The
EPA
also
suggests
there
are
cost­
effective
upgrade
options
for
EGUs
with
pre­
existing
semi­
dry
scrubber
systems.

Comment:
Commenter
0253
said
that
in
largely
focusing
on
the
capabilities
of
state­
of­
the­
art
scrubbers
installed
on
new
generating
units
and
ignoring
the
feasibility
or,
in
many
cases,
infeasibility
of
retrofitting
such
technology
on
existing
units,
EPA's
analysis
is
substantially
flawed.
EPA
overlooks
the
fact
that,
in
many
instances,
it
is
technically
infeasible
to
upgrade
early
generation
scrubbers
to
meet
the
95%
control
level
that
may
be
achievable
with
current
state­
ofthe
art
scrubbers.

Response:
The
EPA
established
the
BART
presumptive
limits
for
coal­
fired
EGUs
without
existing
SO2
controls
based
on
the
control
levels
achievable
with
current
state­
of­
the­
are
new
and
retrofit
scrubber
installations.
The
BART
presumption
for
EGUs
with
pre­
existing
SO2
controls
are
to
be
established
by
the
States
(
not
necessarily
95
percent
control)
based
on
an
evaluation
of
technically
feasible
upgrade
options
available
for
early
generation
scrubbers
in
order
to
improve
the
average
SO2
removal
of
existing
controls.

Comment:
Commenters
(
0228,
0256)
said
the
SO
2
presumption
should
not
apply
to
sources
that
have
already
installed
and
are
operating
scrubbers,
or
to
sources
that
have
committed
to
specific
levels
of
scrubbing
through
consent
orders
or
agreements
with
EPA
or
a
state
regulatory
agency.
If
a
source
can
demonstrate
a
reduction
in
visibility
impairment
below
the
specified
threshold
(
whether
that
threshold
is
EPA's
currently
proposed
0.5
deciview
or
an
alternative
level)
with
less
stringent
controls,
EPA
and
the
states
should
not
impose
by
default
more
stringent
reduction
requirements.

Response:
EPA's
analysis
suggest
cost­
effective
FGD
system
upgrades
designed
to
improve
average
SO2
removal
efficiencies
are
available
for
some
BART­
eligible
units
with
currently
operational
scrubber
systems.
See
the
preamble.

Comment:
220
In
contrast,
one
commenter
(
0221)
said
that
under­
controlled
sources
(
i.
e.,
those
previously
retrofitted
to
meet
the
1979
NSP
of
70
percent)
can
meet
removal
efficiencies
of
95%
or
more.
The
performance
of
boilers
already
using
the
wet
limestone
process
can
be
improved
through
the
application
of
"
once­
through
wet
FGD
technology,"
such
as
increasing
sorbent,
reactivity
and
other
upgrades.

Response:
EPA
analyses
suggest
cost­
effective
FGD
system
upgrades
designed
to
improve
average
SO2
removal
efficiencies
are
available
for
BART­
eligible
units
with
currently
operational
scrubber
systems.
See
TSD
"
Setting
BART
Presumptive
SO2
Limits
for
Electric
Generating
Units"
for
a
more
detailed
discussion
on
EPA's
list
of
recommended
upgrade
options
for
wet
FGD
systems.

Comment:
Eight
commenters
discussed
the
role
that
plant
size
and/
or
capacity
should
play
in
applying
a
default
SO
2
control
level
(
0206,
0227,
0241,
0253,
0256,
0267,
0271,
0303).
Some
commenters
said
that
the
presumptive
limits
should
not
apply
to
units
between
250
and
750
MW
(
0206,
0241,
0303).
Commenter
0253
said
EPA
should
clarify
that
only
BART­
eligible
sources
would
be
counted
toward
the
750­
MW
cut­
off.
Commenters
(
0206,
0303)
added
that
Congress
did
not
specify
similar
requirements
for
smaller
plants
(
i.
e.
facilities
below
750
MWs).
The
installation
of
controls
necessary
to
reduce
SO
2
emissions
by
90
to
95
percent
is
more
likely
to
be
uneconomical
at
smaller
facilities
(
0206,
0253,
0303).
The
BART
process
must
be
followed
for
such
a
determination
to
be
made
(
0241).

Response:
The
BART
presumptions
are
binding
for
units
greater
than
750
MW.
EPA
strongly
encourages
States
to
apply
the
BART
presumptive
limits
to
units
greater
than
200
MW
regardless
of
plant
size.
The
EPA's
unit­
specific
economic
analysis
for
BART
eligible
coal­
fired
units
suggests
on
average
it
is
less
economical
for
units
less
than
200
MW
to
retrofit
with
FGD
systems
than
those
units
greater
than
200
MW.
However,
the
analysis
did
demonstrate
that
some
BART­
eligible
units
less
than
200
MW
were
modeled
to
have
cost
effectiveness
values
comparable
to
those
unit
greater
than
200
MW
and
that
States
are
encouraged
to
apply
the
BART
presumptions
to
such
sources.

Comment:
Commenter
0256
said
EPA's
proposed
distinction
between
the
SO
2
"
default
requirement"
and
the
"
rebuttable
presumption"
based
on
total
power
plant
capacity
is
unsupportable.
Because
BART
applies
on
a
unit
basis,
the
fact
that
the
total
capacity
of
the
plant
is
higher
or
lower
than
750
MW
has
no
obvious
bearing
on
whether
an
individual
unit
can
meet
the
default/
presumption.
Indeed,
if
anything,
units
at
a
larger
plant
would
perhaps
have
more
problems
meeting
the
presumption
than
would
units
at
a
smaller
plant,
because
a
larger
plant
would
tend
to
have
more
units
and
thus
more
potential
for
space
constraints
in
retrofitting
add­
on
controls.
If
EPA
retains
the
SO
2
presumption,
it
should
be
a
rebuttable
presumption
for
all
units
greater
than
250
MW,
regardless
of
total
power
plant
capacity.
221
In
contrast,
commenters
(
0227,
0271)
said
EPA
should
require
presumptive
controls
for
all
units
at
750
MW
plants,
including
those
between
250
and
750
MW.
Commenter
0271
said
this
conforms
to
the
specific
wording
of
the
Act
(
Sec.
169A(
b)(
2)(
A))
for
controls
on
these
smaller
than
750
MW
units.
Commenter
0267
supported
applying
the
proposed
presumptive
level
to
all
EGUs
greater
than
250
MW.

Response:
See
the
preamble.

Comment:
Commenter
0268
noted
that
Arizona
has
chosen
to
address
visibility
impairment
through
Section
309
of
the
1999
RHR.
In
order
to
pursue
the
309
pathway,
Arizona
has
made
a
demonstration
to
EPA
that
the
state
will
realize
overall
SO
2
reductions
greater
than
the
reductions
that
would
be
achieved
solely
through
imposition
of
BART
controls
at
the
state's
BART­
eligible
sources.
EPA
should
clearly
and
definitively
recognize
that
presumptive
BART
limits
for
SO
2
do
not
apply
to
BART­
eligible
sources
located
in
states
that
have
already
committed
to
"
better
than
BART"
SO
2
reductions
through
Section
309
SIP
submittals.

Response:
We
agree.

Comment:
Twelve
commenters
suggested
alternatives
to
presumptive
control
levels
(
0213,
0215,
0256,
0302,
28­
255,
28­
256,
28­
264,
28­
266,
28­
278,
28­
312,
28­
355,
28­
404).
Commenter
0213
supports
an
emissions
trading
program
to
implement
BART
controls
as
proposed
in
the
June
10,
2004
Clean
Air
Interstate
Rule
(
CAIR).
However
if
an
individual
unit
program
is
prescribed,
the
commenter
supports
a
percent
removal
limit
as
opposed
to
an
emission
limit.

Response:
EPA
appreciates
these
comments
and
supports
BART­
eligible
units
option
to
opt
into
the
CAIR
in
applicable
states.

Comment:
Commenter
0215
said
EPA
should
retain
the
option
of
complying
with
the
SO
2
default/
presumption
for
EGUs,
either
by
adhering
to
the
straight
percentage
reduction
or
a
set
emission
rate.
Due
to
the
variability
in
coal
and
equipment
types
(
including
vintages),
EPA
must
retain
flexibility
for
those
EGUs
that
must
comply
with
these
BART
requirements.

Response:
The
EPA
appreciates
this
comment.

Comment:
Commenter
0256
agrees
with
EPA
that,
for
units
burning
lower
sulfur
coal,
achieving
95%
reduction
could
be
problematic
and
that
an
emission
limitation
should
be
used.
Other
commenters
222
(
28­
256,
28­
264,
28­
266,
28­
278,
28­
312,
28­
355,
28­
404)
agreed
if
a
presumptive
approach
is
used,
then
the
presumptive
norms
should
be
explicitly
stated
in
the
rulemaking
and
should
be
no
higher
than
3.0
lb/
MWh
for
SO
2
(
28­
256),
based
on
a
weight
rate
limit
or
other
fuel­
neutral
basis
to
account
for
facilities
burning
higher
sulfur
coal
(
28­
264,
28­
335),
on
an
emissions
standard
to
allow
sources
to
take
credit
for
improvements
they
have
already
made
(
28­
266,
28­
278,
28­
312,
28­
404).

Response:
The
EPA
conducted
a
unit­
specific
analysis
in
determining
the
BART
presumptive
limits
and
is
finalizing
presumptions
which
it
feels
to
be
achievable
by
the
BART­
eligible
population,
absent
extreme
circumstances.
The
final
BART
presumptive
limits
for
coal­
fired
EGUs
without
existing
SO2
controls
includes
both
an
emission
rate
limit
and
percent
reduction
requirement
to
provide
equity
among
the
units
and
their
various
characteristics.
See
the
preamble
for
further
explanation.

Comment:
Commenter
0302
prefers
the
flexibility
of
a
percentage
control
level
comparable
to
a
performance
level
of
0.1
to
0.15
lbs/
MMBTU
SO
2.
A
percentage
control
level
would
offer
the
non­
EGU
BART­
subject
sources
the
best
controls
that
are
achievable,
for
all
types
of
sources,
under
reasonable
cost
effectiveness.
Without
the
proper
guidance
for
BART­
subject
sources
that
are
EGUs,
the
commenter
needs
flexibility
in
determining
the
actual
emission
rates
at
which
these
sources
will
be
required
to
demonstrate
compliance.

Response:
See
the
preamble.

Comment:
Two
commenters
(
28­
255,
28­
264)
said
the
presumptive
levels
should
not
be
established
as
absolute
values.
Commenter
28­
255
said
the
values
should
be
presented
as
a
range.
Commenter
28­
264
said
the
presumption
should
not
interfere
with
the
states'
discretion
and
statutory
process.
If
EPA
intends
to
apply
this
"
presumptive
level"
as
a
mandate,
then
EPA
should
establish
it
as
a
national
rule
rather
than
continuing
the
pretense
that
states
are
making
a
case­
by­
case
determination.

Response:
See
the
preamble
for
explanation
of
the
circumstances
under
which
the
presumptive
limits
apply.

Comment:
Three
commenters
discussed
the
role
that
the
averaging
period
plays
in
determining
performance
(
0213,
0256,
0314).
Commenter
0213
noted
the
difference
in
performance
over
longer
averaging
times
between
state­
of­
the­
art
technology
and
existing
FGD
installations.
EPA
should
review
existing
installations
to
determine
an
appropriate
level
of
performance
for
223
determining
an
SO
2
removal
efficiency
limit.
Commenter
0256
said
the
choice
of
0.15
lb/
MMBtu
is
predicated
on
the
use
of
at
least
an
averaging
time
of
30­
day
rolling
or
longer.
The
0.15
lb/
MMBtu
limit
could
not
be
achieved
with
the
use
of
a
shorter
averaging
time.
Commenter
0314
asked
that
an
averaging
period,
i.
e.,
a
30­
day
rolling
average,
for
the
control
levels
be
established
in
the
rule.
A
longer
averaging
period
will
allow
for
lower
annual
emissions.

Response:
The
EPA
recognizes
the
correlation
between
sustainable
FGD
performance
and
length
of
averaging
times.
For
this
reason
it
is
establishing
as
part
of
the
BART
rulemaking
a
30­
day
rolling
averaging
period
for
the
SO2
presumptive
limits.

Comment:
Four
commenters
said
that
EPA
has
underestimated
the
cost
per
ton
of
emissions
removed
(
0187,
0206,
0236,
0303).
Commenters
(
0187,
0236)
provided
an
example
of
the
costeffectiveness
range
for
reducing
SO
2
emissions
to
0.1
to
0.15
lbs/
MMBTU
estimated
to
be
between
$
1800
and
$
3400
per
ton
of
SO
2
removed
when
burning
Powder
River
Basin
coals.
This
is
much
higher
than
the
$
200­$
1000
per
ton
cost­
effectiveness
range
cited
by
EPA
in
support
of
the
proposed
presumptive
BART
SO
2
limit.
Commenter
0187
gave
an
example
of
an
oil­
fired
unit
exceeding
$
3,000
per
ton
of
SO
2
removed,
which
supports
a
separate
presumptive
SO
2
emission
limit
for
oil­
based
units
based
on
the
maximum
fuel
sulfur
content
in
oil.
Commenters
(
0206,
0303)
were
concerned
about
additional
costs
faced
by
smaller
plants
(
upgrading
fans,
piping,
etc.)
that
can
drive
the
cost
into
the
$
3,000
range.
The
commenters
said
EPA
should
treat
smaller
plants
like
other
major
sources
and
allow
states
greater
flexibility
in
determining
BART
for
such
facilities.

Response:
EPA's
unit
specific
economic
analysis
suggests
an
average
cost
effectiveness,
for
all
BART
units
greater
than
200
MW,
of
$
919
per
ton
SO2
removed,
with
a
majority
of
the
units
anticipated
to
have
cost
effectiveness
values
ranging
from
$
400
to
$
2000
per
ton
SO2
removed,
see
TSD
"
Setting
BART
Presumptive
SO2
Limits
for
Electric
Generating
Units"
for
a
more
detailed
discussion
on
how
EPA
determined
its
final
recommendations.
EPA's
economic
analysis
also
suggested
that
some
unit
greater
than
200
MW
could
experience
expensive
cost
effectiveness
values,
however
States
have
the
flexibility
to
determine
alternate
BART
presumptive
limits
based
on
the
results
of
the
5
part
BART
analysis.
The
EPA
strongly
encourages
States
to
apply
the
BART
presumption
at
units
greater
than
200
MW
because
our
economic
analysis
suggested
higher
average
cost
effectiveness
values
for
units
less
than
200
MW.
However,
EPA
also
recognizes
that
some
units
less
than
200
MW
were
modeled
to
have
reasonable
cost
effectiveness
values.
The
EPA
determined
the
final
BART
presumption
for
oil­
fired
units
to
be
a
limitation
of
fuel
oil
sulfur
content.

Comment:
In
contrast,
commenters
(
0221,
0267)
said
achieving
the
control
levels
are
cost
effective.

Response:
224
The
EPA
appreciates
these
comments.

Comment:
Six
commenters
provided
other
comments
regarding
SO
2
limits
for
utility
boilers
(
0215,
0224,
0241,
0250,
0255,
0319).
Commenter
0215
said
if
EPA
retains
the
SO
2
default/
presumption,
it
should
make
it
clear
that
if
any
source
is
willing
to
achieve
95%
control
or
somewhere
along
the
scale
of
0.10
to
0.15
lb/
MMBtu,
it
should
automatically
be
exempt
from
the
requirement
to
perform
a
BART
analysis
for
SO
2.
Commenter
0224
said
EPA
might
specify
flue
gas
desulfurization
as
a
presumptive
technology.
Commenter
0255
described
several
options
that
are
available
to
further
reduce
SO
2
emissions
and
improve
the
reliability
of
existing
FGD
systems
to
above
95
%
removal
efficiencies.

Response:
See
the
preamble
and
the
technical
support
document:
"
Setting
BART
Presumptive
SO2
Limits
for
Electric
Generating
Units".

Comment:
Commenters
(
0241,
0250)
opposed
establishing
any
SO
2
control
level
for
EGUs
as
presumptive
BART
because
this
usurps
state
authority
and
completely
ignores
the
BART
evaluation
process
required
by
the
Act.
Commenter
0250
said
this
is
especially
inappropriate
for
units
already
equipped
with
scrubbers.

Response:
See
the
preamble.

Comment:
Commenter
0319
was
concerned
that
setting
an
emission
rate
goal
incorporating
cap
and
trade
mechanisms
rather
than
a
percent
control
goal
would
encourage
fuel
switching
to
SPRB
coal.
Even
plants
that
meet
the
BART
standards
would
be
encouraged
to
switch
to
SPRB
coal
to
avoid
capital
investment
or
to
profit
from
the
sale
of
surplus
allowances.
Tribal
coal
and
lignite
production
could
be
put
at
risk.
The
commenter
provided
a
proposal
to
address
the
impacts
on
tribal
coal
as
Attachment
A.

Response:
The
final
BART
presumptions
do
not
include
a
cap
and
trade
program
for
EGUs.
EPA
believes
the
BART
presumptive
limits
for
EGUs
are
stringent
enough
that
fuel
switching
alone
will
not
bring
a
unit
into
compliance
with
the
0.15
lb
SO2/
mmBtu
emission
rate
limit.
The
final
BART
presumptions
for
coal­
fired
EGUs
without
pre­
existing
SO2
controls
were
finalized
based
on
what
EPA
believes
are
the
current
capabilities
of
wet
and
semi­
dry
FGD
systems.

4.2.3.3
General
Comments
Regarding
Presumptive
Limits
225
Comment:
Seven
commenters
supported
the
use
of
presumptive
emission
limits
in
the
BART
program
(
0179,
0221,
0255,
0285,
0299,
0323,
28­
266).
Commenters
said
the
limits
are
achievable
(
0179,
0255)
and
are
one
of
the
most
essential
components
of
the
BART
guidelines
because
they
establish
a
starting
point
representing
the
"
best
available
technology"
in
the
required
engineering
analysis
(
0221).
Some
commenters
requested
that
EPA
establish
more
stringent
limits
and
harmonize
the
presumptive
BART
emission
limits
for
EGUs
under
the
regional
haze
program
with
the
BART
guidelines
(
0221).
Commenter
0323
approved
including
a
specific
BART
determination
for
large,
bituminous
coalfired
electric
generating
stations.
These
facilities
constitute
the
largest
emitters
of
sulfur
dioxide
and
nitrogen
oxides
among
all
26
source
categories
subject
to
BART.

Commenters
(
0285,
28­
266)
urged
EPA
to
develop
national
guidelines
and
presumptive
BART
requirements
for
source
types
common
to
several
states
to
make
the
process
more
efficient
and
workable
for
the
states.
Commenters
(
0299,
28­
266)
added
that
the
guidelines
should
encourage
consistency
between
states
and
said
EPA
should
establish
a
consistent
set
of
protocols
upon
which
a
BART
analysis
for
each
source
category
should
be
based.
Commenter
0299
added
that
such
protocols
should
not
eliminate
the
flexibility
each
state
may
need
to
address
unusual
circumstances
at
individual
sources,
but
these
protocols
could
potentially
eliminate
the
need
for
case­
by­
case
review
for
many
sources.
In
the
absence
of
a
presumptive
standard,
the
proposed
BART
guidelines
must
contain
explicit
criteria
for
determining
BART.
In
addition,
the
BART
Guidelines
should
include
defined
criteria
for
"
engineering",
"
useful
life",
and
"
cost".

Response:
We
agree
with
these
comments;
see
the
preamble
for
justification
of
the
specific
presumptive
limits.

Comment:
Several
commenters
said
EPA
should
establish
presumptive
limits
for
other
pollutants
and/
or
source
categories
(
28­
248,
28­
255,
28­
358).
Commenter
28­
255
would
also
seek
to
establish
limits
for
source
categories
in
addition
to
utility
boilers
to
make
the
process
more
uniform
and
consistent
across
the
country
and
easier
for
the
states.
Commenter
28­
358
said
the
SO
2
limits
should
be
expanded
for
all
BART­
eligible
industrial
boilers
in
all
26
categories
and
set
at
95%
to
reflect
control
technology
performance.
Commenter
28­
248
proposed
setting
a
limit
for
particulate
matter
at
90%.

Response:
We
believe
BART
determinations
are
to
be
conducted
by
States
under
the
CAA;
we
further
believe
presumptive
limits
for
NOx
and
SO2
for
specific
large
EGUs
are
justified
based
upon
our
analysis
discussed
in
the
preamble
and
technical
support
documents
(
TSDs).

Comment:
Forty­
nine
commenters
opposed
establishing
presumptive
emission
limits
(
0171,
0181,
0184,
0195,
0197,
0198,
0208,
0210,
0211,
0214,
0215,
0217,
0224,
0226,
0228,
0238,
0244,
226
0246,
0247,
0252,
0253,
0254,
0256,
0268,
0270,
0275,
0276,
0308,
0314,
0321,
28­
243,
28­
264,
28­
267,
28­
305,
28­
333,
28­
340,
28­
342,
28­
347,
28­
354,
28­
356,
28­
357,
28­
360,
28­
385,
28­
390,
28­
401,
28­
402,
28­
748,
28­
7003,
28­
7009).

Commenters
said
this
approach
restricts
the
flexibility
of
the
states
and/
or
sources
to
make
the
control
technology
evaluations
on
a
case­
by­
case
basis
(
0171,
1081,
0195,
0197,
0211,
0217,
0247,
0253,
0256,
0268,
0275,
0276,
0321,
28­
243,
28­
305,
28­
333,
28­
340,
28­
342,
28­
349,
28­
354,
28­
357,
28­
360,
28­
385,
28­
390,
28­
748)
without
resulting
in
a
notable
degree
of
improvement
to
visibility
(
0171,
0197).

The
Act
requires
states
to
use
the
statutory
factors
in
making
case­
by­
case
BART
determinations
(
0195,
0197,
0210,
0211,
0217,
0228,
0238,
0244,
0246,
0247,
0252,
0253,
0256,
0268.
0308,
0314,
0321,
28­
264,
28­
267,
28­
347,
28­
355,
28­
357,
28­
359,
28­
360,
28­
385,
28­
390,
28­
401,
28­
402,
28­
747,
28­
748,
28­
7001,
28­
7003,
28­
7009).
Commenters
(
0211,
0217,
0246)
added
that,
rather
than
assist
states
by
providing
them
with
resources,
tools
and
guidance
under
BART
guidelines,
EPA's
use
of
rebuttable
presumptions
will
tend
to
require
significant
levels
of
state
resources
for
states
to
exercise
discretion
in
their
BART
determination
processes.
Most
states
have
limited
administrative
resources
to
address
these
issues
and
therefore,
EPA's
use
of
presumptive
requirements
will
discourage
state
level
discretion
­­
opposite
the
intent
conveyed
by
the
Coin
Growers
decision.

Commenters
(
0198,
0226,
0254)
said
the
presumptive
levels
are
not
achievable
across
the
range
of
boiler
types,
vintage
of
existing
controls
and
coal
composition.
Commenter
0198
preferred
the
use
of
trading
under
CAIR
to
meet
BART
requirements
in
the
most
efficient
and
cost­
effective
manner.
Commenter
0226
said
if
a
presumptive
level
of
control/
emission
rate
is
retained
in
the
final
rule,
the
guidelines
should,
at
a
minimum,
make
it
clear
that
states
have
the
authority
to
select
an
alternative
level
of
control/
emission
rate,
without
an
undue
evidentiary
burden,
given
appropriate
consideration
of
all
the
statutory
BART
factors.
Commenter
0253
asked
that
EPA
either
delete
the
provisions
for
mandatory
default
control
levels
or
provide
a
source
with
the
option
of
using
the
default
control
levels,
rather
than
making
them
mandatory.
Commenter
0276
said
if
EPA
establishes
presumptive
control
levels
for
EGUs,
average
annual
values
are
more
appropriate
than
those
in
the
proposed
regulations.
In
addition,
BART
should
not
be
structured
in
a
way
that
results
in
rigid
technology­
based
limits
that
interfere
with
established
NO
x
and
SO
2
emission
trading
programs.
Commenter
0314
said
if
EPA
needs
to
define
a
target,
it
should
use
a
range
to
prevent
diminishing
states'
flexibility
in
case­
by­
case
BART
reviews.

Commenters
(
0214,
0215,
0238,
0254)
said
EPA
fails
to
make
clear
how
the
proposed
default
values
for
SO
2
and
NO
x
relate
to
visibility
impairment.
Visibility
is
both
Class
I
area
and
source
specific,
which
is
the
reason
Congress
gave
the
states
the
lead
role
and
discretion
in
the
BART
program
to
determine
which
sources
need
to
install
or
upgrade
controls.
Through
the
use
of
presumptions
and
default
values,
however,
EPA's
prescriptive
process,
as
proposed,
would
make
the
installation
of
maximum
controls
more
likely
without
regard
to
visibility
benefits.
Instead,
as
required
by
statute
and
rules,
EPA
should
give
the
states
maximum
flexibility
to
use
the
five
statutory
factors
in
their
BART
determinations.
Commenters
(
0228,
0247)
said
sources
must
227
be
allowed
to
assess
the
visibility
improvements
of
a
variety
of
control
options.
Commenter
0308
said
states
should
be
allowed
to
exempt
BART­
eligible
sources
from
BART
requirements
if
these
sources
do
not
impact
or
have
very
little
impact
on
visibility
in
a
Class
I
area.

Commenter
28­
342
said
even
if
a
presumption
were
appropriate
under
the
visibility
protection
program
(
and
it
is
not),
EPA
cannot
adopt
one
now
because
the
Agency
has
not
satisfied
the
conditions
for
creating
a
presumption.
Federal
presumptions
can
be
promulgated
only
when
the
Agency
has
established
(
based
on
generic
application
of
law
to
fact)
that
the
presumption
is
valid.

Response:
See
the
preamble.

Comment:
Commenters
(
0208,
0270)
said
EPA's
default
and
presumptive
control
levels
for
SO
2
and
NO
x
are
inconsistent
with
the
BART
control
efficiencies
proposed
by
the
WRAP
and
should
be
removed
from
the
proposed
guidelines.
In
no
event,
however,
should
BART
force
plants
in
the
West
to
reduce
emissions
to
levels
greater
than
achievable
using
dry
scrubbing
technology
for
SO
2
and
advanced
low­
NO
x
combustion
technology
for
NO
x.
As
EPA
has
pointed
out
in
its
discussion
of
Clear
Skies
and
CAIR,
wet
scrubbers
and
SCR
are
inappropriate
for
the
arid
West.
In
fact,
the
statute
itself
makes
the
NSPS
a
ceiling
for
BART.
Commenter
0270
said
the
proposed
presumptive
BART
emission
limitations
are
inconsistent
with
the
Regional
Haze
Rule
for
the
GCVTC
states
(
WRAP
Annex
and
SIPs
under
40
CFR
51.309),
the
Administration's
proposed
legislation
for
the
power
generation
sector,
the
CSA,
and
the
proposed
IAQR.

Response:
See
the
preamble
and
TSDs
for
discussion
of
specific
SO2
and
NOx
limits.
In
addition,
we
do
not
believe
the
BART
presumptive
limits
are
inconsistent
with
other
EPA
rules
or
with
the
proposed
Clear
Skies
legislation.
The
preamble
contains
a
discussion
of
the
relationship
between
BART
and
the
final
CAIR
rule
(
which
was
proposed
as
the
IAQR).

Comment:
Commenters
(
0184)
said
limits
should
not
be
set
for
those
pollutants
for
which
a
facility
has
no
known
control.

Response:
We
agree.

Comment:
Commenter
0246
said
EPA
should
clarify
that
greater
reasonable
progress
need
not
be
demonstrated
by
achieving
emission
reductions
greater
than
those
that
would
occur
by
applying
the
presumptive
control
levels
proposed
by
EPA
to
all
BART­
eligible
sources.
Such
an
approach
would
undermine
the
fundamental
role
of
the
states
in
determining
BART
or
what
constitutes
greater
reasonable
progress
than
BART.
The
WRAP,
states,
tribes
and
stakeholders
conducted
a
228
thorough
analysis
and
balancing
of
the
BART
factors
to
arrive
at
the
WRAP
milestones
 
all
of
this
was
conducted
without
imposing
presumptive
control
levels
during
the
analysis
and
discussion.

If
EPA
does
utilize
the
presumptive
emission
limits,
commenter
0197
said
,
if
a
source
agrees
to
95%
control
or
the
range
of
0.1­
0.15
lb/
MMBTU
for
SO
2,
they
should
be
automatically
exempt
from
the
BART
process.
The
same
goes
for
the
0.2
lb/
MMBtu
for
NOx.

Response:
Reasonable
progress
toward
visibility
goals
are
to
be
assessed
in
the
regional
haze
State
implementation
plan
(
SIP)
under
the
regional
haze
rule.
In
addition,
the
CAA
requires
BART
determinations
on
specific
sources
in
regional
haze
SIPs.
These
are
separate
processes
that
the
State
must
conduct
and
evaluate
taking
into
account
regional
visibility
conditions;
BART
may
be
just
one
component
of
a
regional
haze
SIP.
In
addition,
the
WRAP
milestones
were
developed
under
a
separate
process
from
source­
specific
BART;
the
relationship
between
the
WRAP
milestones
under
section
309
of
the
regional
haze
rule
with
reasonable
progress
is
different
from
the
relationship
of
source­
specific
BART
under
section
308
of
the
regional
haze
rule.

In
regard
to
commenter
197,
see
the
preamble
for
discussion
of
BART
determinations
for
sources
meeting
the
presumptive
limits.

Comment:
Commenter
0268
questioned
the
legal
authority
of
EPA
to
impose
presumptive
limits
for
sources
between
250
MW
and
750
MW
in
size
because
the
visibility
provisions
of
the
Act
only
specify
evaluation
by
the
states
of
units
with
a
capacity
in
excess
of
750
MW.
Commenter
0184
said
the
statute
requires
state
to
take
a
case­
by­
case
approach
for
non­
EGU
sources.
Discussion
of
such
emission
limits
should
be
explicitly
limited
to
those
large
EGU
facilities
for
which
EPA
has
the
statutory
authority
to
define
BART
and
to
set
such
limits.
Commenter
0237
said
the
presumptive
BART
level
for
utility
boilers
would
impose
hardship
on
industrial
sources
because
there
are
many
practical
differences
in
the
way
industrial
boilers
and
electrical
generating
units
are
operated,
including
short
term
swings
in
demand,
that
make
such
BART
controls
unreasonable.
Therefore,
EPA
must
underscore
in
its
final
guideline
that
these
control
levels
are
not
presumptive
BART
for
industrial
boilers.

Five
commenters
said
EPA
should
consider
the
type
of
emission
unit
(
e.
g.,
coal
type,
boiler
type,
and/
or
emission
controls)
when
making
BART
determinations
(
0195,
0198,
0214,
0238).
Commenters
(
0214,
0238)
added
that
states
and
sources
should
be
focusing
on
the
incremental
costs
and
benefits
from
the
installation
of
additional
controls
or
the
upgrading
of
existing
controls
in
the
BART
determination.

Response:
We
agree
that
presumptive
limits
do
not
apply
to
industrial
boilers.
States
have
discretion
to
assess
the
BART
factors
and
apply
appropriate
limits
to
these
and
other
non­
EGU
sources.
See
the
preamble
for
discussion
of
our
legal
authority
to
set
presumptive
limits
for
certain
EGUs.
229
We
agree
that
coal
type,
boiler
type
and/
or
emission
controls
should
be
considered,
and
the
preamble
discusses
how
the
presumptive
limits
have
considered
them.

Comment:
Two
commenters
supported
the
use
of
optional
presumptive
limits
as
a
choice
for
source
owners/
operators
to
avoid
a
BART
analysis
(
0210,
0256).
This
approach
would
limit
the
resources
expended
for
many
determinations,
but
would
allow
states
to
consider
all
of
the
statutory
factors
where
appropriate.

Response:
We
believe
States
have
the
discretion
they
need
to
assess
the
BART
factors.
See
the
preamble.

Comment:
Six
commenters
said
EPA
needed
to
address
the
averaging
period
associated
with
the
presumptive
limits
(
0195,
0227,
0228,
0268,
0306,
0323).
Some
commenters
suggested
annual
average
limits
(
0195,
0268)
while
others
said
they
should
be
no
less
than
a
30­
day
or
monthly
average
(
0228).
Other
commenters
asked
EPA
to
specify
an
averaging
time
similar
to
the
averaging
time
of
visibility
impact
assessments,
e.
g.,
24­
hour
average
(
0227,
0306).

Response:
We
agree,
and
have
specified
a
30­
day
rolling
average
in
the
final
Guidelines.

Comment:
One
commenter
said
EPA
should
establish
emission
limits
for
PM
because
of
the
magnitude
of
emissions
and
the
availability
of
proven
controls
(
0255).

Response:
We
agree
that
controls
are
readily
available
to
address
PM
emissions,
and
we
encourage
States
to
strongly
consider
those
controls
and
set
limits
appropriate
to
each
source
subject
to
BART.

Comment:
Seven
commenters
made
other
general
comments
regarding
presumptive
emission
limits
(
0195,
0214,
0232,
0238,
0255,
0299,
0306).

Commenters
(
0195,
0214,
0238)
said
exceptions
should
be
made
for
startups,
shutdowns
and
malfunctions
of
control
equipment.

Response:
States
have
discretion
to
appropriately
address
startup,
shutdowns
and
malfunctions
of
control
equipment
in
Title
V
permits.

Comment:
230
Commenters
(
0214,
0238)
said
EPA
should
make
it
clear
that
if
any
source
is
willing
to
achieve
95%
control
or
somewhere
along
the
scale
of
0.1
to
0.15
lb/
MMBTU,
it
should
automatically
be
exempt
from
the
requirement
to
perform
a
BART
analysis
for
SO
2.
The
same
should
also
be
true
regarding
a
BART
NO
x
analysis
for
sources
willing
to
achieve
the
0.2
lb/
MMBTU
NO
x
x
presumption
for
EGUs.
Such
presumptions
would
streamline
the
BART
process
for
those
sources
willing
to
reach
these
control
levels.
Commenter
0255
said
EPA
should
set
a
high
threshold
for
sources
to
justify
being
exempted
because
the
current
suite
of
best
available
technologies
can
readily
meet
and
go
beyond
the
mandated
limits.
EPA
should
encourage
the
states
to
consider
these
factors
as
well.

Response:
See
the
preamble.

Comment:
Commenter
0232
said
there
should
be
a
provision
to
revisit
appropriate
BART
control
levels
at
a
date
certain.
While
it
is
appropriate
to
provide
clarity
and
certainty
to
industry
for
the
installation
of
retrofit
controls,
it
is
also
important
to
provide
for
application
of
new
controls
with
increased
control
efficiency
due
to
technology
improvements
in
order
to
meet
the
express
requirements
of
the
Act
to
prevent
any
future
and
remedy
any
current
impairment
of
visibility
in
Class
I
areas.

Response:
Under
the
CAA,
BART
is
specifically
an
element
of
the
first
regional
haze
SIP.

Comment:
Commenter
0255
noted
that
EPA
did
not
address
emission
of
SO
3
from
power
plants.
This
has
become
an
increase
area
of
concern
for
power
companies
as
states
have
set
emission
standards
for
SO
3
in
the
operating
permits
for
power
plants.
A
number
of
control
technologies
have
been
developed
to
address
this
pollutant
and
are
able
to
achieve
a
high
level
of
removal.

Response:
States
have
discretion
to
address
pollutants
under
BART
to
the
extent
that
they
are
visibility­
impairing.

Comment:
Commenter
0299
said
the
BART
presumptive
level
of
control
proposed
for
EGUs
boilers
for
SO
2
emissions
and
for
NO
x
emissions
should
be
made
consistent
with
other
clean
air
programs.
Specifically,
the
presumptive
level
of
controls
for
EGU
boilers
should
be
consistent
with
the
requirements
adopted
in
the
CAIR,
and
consistent
with
the
requirements
contained
in
the
federal
NO
x
SIP
Call.

Response:
231
See
the
preamble
for
a
discussion
of
the
relationship
of
BART
requirements
and
CAIR
requirements.
In
addition,
BART
is
consistent
with
the
NOx
SIP
Call;
for
EGUs
required
to
install
SCR
under
the
NOx
SIP
Call,
the
BART
presumption
is
simply
that
those
sources
should
run
the
SCRs
year­
round.

Comment:
Commenter
0306
said
BART
should
be
no
more
stringent
than
a
current
BACT
assessment
for
an
existing
facility.
EPA
should
define
presumptive
BART
in
a
manner
consistent
with
recent
NSR
enforcement
actions.
Presumptive
BART
should
not
be
more
stringent
than
requirements
in
recent
agreements
(
or
proposed
agreements)
that
settled
NSR
violations.

Response:
BART
is
determined
under
a
separate
process
from
specific
NSR
settlement
agreements.

4.2.3.2
NOx
Limits
for
Utility
Boilers
Comment:
Eighteen
commenters
on
the
2001
notice
supported
presumptive
limits
for
NOx
set
at
90
to
95%
(
28­
245,
28­
247,
28­
248,
28­
255,
28­
260,
28­
262,
28­
302,
28­
325,
28­
334,
28­
338,
28­
350,
28­
352,
28­
358,
28­
362,
28­
390,
28­
399,
28­
433,
28­
7016).
Commenter
28­
302
said
the
combination
of
low­
NOx
burner
(
LNB)
technology
and
SCR
controls
can
routinely
reduce
NOx
emissions
by
over
90%
and
other
commenters
cited
research
documenting
the
performance
of
controls
(
28­
325,
28­
390)
and
EPA's
own
experience
and
data
(
28­
28­
352,
28­
362,
28­
7016).

Response:
EPA
is
finalizing
presumptive
Nox
limits
for
some
boiler
types
based
on
using
SCRs
with
90%
reduction
efficiency.
See
preamble.

Comment:
Commenter
28­
256
said
if
a
presumptive
norm
approach
is
used,
then
the
presumptive
norms
should
be
explicitly
stated
in
the
rulemaking
and
should
be
no
higher
than
1.5
lb/
MWh
for
NOx..

Response:
EPA
has
explicitly
stated
its
presumptive
Nox
limits
for
coal­
fired
EGUs
in
the
rulemaking.
The
presumptive
Nox
limits
EPA
is
finalizing
vary
by
unit
type
and
coal
type.
For
tangentially­
fired
boilers
burning
sub­
bituminous
coal,
EPA
is
finalizing
a
presumptive
Nox
limit
of
0.15
lb/
mmbtu
as
this
commenter
suggests.
For
other
boiler
and
coal
types,
EPA's
final
presumptive
Nox
limits
are
higher
than
0.15
lb/
mmbtu.
See
preamble
for
discussion
of
the
rationale
EPA
used
for
setting
presumptive
Nox
limits.
232
Comment:
Two
commenters
supported
the
proposed
NOx
limit
of
0.2
as
a
reasonable
level
of
control
(
0231,
0302).
Commenter
0232
supported
the
setting
the
presumptive
control
levels
of
NOx
at
90%.
Three
commenters
said
the
proposed
levels
are
too
stringent
(
0171,
0207,
0227).
Commenter
0171
described
recent
plant
retrofits
using
low­
NOx
burners
and
low­
NOx
concentric
firing
systems
that
could
not
practically
achieve
the
0.2
lb/
MMBTU
limit.
Commenter
0207
said
the
presumptive
limit
is
counter
to
available
evidence
and
unreasonable,
especially
for
T­
fired
units
equipped
with
low­
NOx
burners.
EPA's
experience
in
the
Acid
Rain
program
is
consistent
with
a
higher
emission
rate
and
should
be
reflected
in
setting
the
BART
levels.
The
commenter
concluded
that
the
experience
of
some
of
its
members
with
low­
NOx
burners,
the
background
information
for
development
of
Group
1,
Phase
II
NOx
limits,
and,
most
likely,
the
voluminous
NOx
emission
data
reported
to
EPA
from
all
Phase
I
and
Phase
II
units,
all
demonstrate
that
a
presumptive
0.2
lb/
MMBTU
NOx
emission
limit
for
utility
boilers
is
a
very
unreasonable,
representation
of
BART
for
such
units.
Commenter
0227
was
concerned
that
the
NOx
presumptive
control
level
EPA
proposed
is
higher
than
the
level
current
technology
can
provide.
EPA
should
review
currently
proposed
NOx
retrofits
and
consider
lowering
it
presumptive
level
of
0.2
lbs/
MMBTU.

Three
commenters
said
the
proposed
NOx
levels
are
too
lenient
(
0167,
0255,
0271).
Commenter
0167
urged
EPA
to
require
electric
generating
units
in
the
West
to
meet
rigorous
NOx
pollution
control
standards.
Commenter
0255
said
the
current
suite
of
best
available
NOx
technologies,
post
combustion
as
well
as
combustion
controls,
can
achieve
significantly
lower
emission
rates
than
the
proposed
0.2
lb/
MMBtu.
In
2002
ozone
season,
the
air
pollution
control
industry
had
100
units
achieving
a
NOx
emission
rate
of
less
than
0.2
lb/
MMBtu.
Commenter
0271
recommended
the
presumptive
control
level
of
0.15
lbs.
of
NOx
per
MMBTU.
This
level
of
control
has
been
demonstrated
in
the
NOx
budget
program
in
the
eastern
United
States.

Response:
In
response
to
comments,
EPA
has
revised
its
proposed
Nox
presumptive
limit
and
has
set
final
presumptive
limits
based
on
a
thorough
unit
by
unit
analysis
of
BART­
eligible
EGUs
and
what
each
unit
would
be
capable
of
achieving
with
current
combustion
control
technology.
EPA's
final
presumptive
Nox
limits
for
coal­
fired
EGUs
are
established
by
unit
type
and
coal
type
to
more
accurately
reflect
the
capability
of
each
BART­
eligible
EGU.
See
preamble
for
further
discussion.

Comment:
Six
commenters
said
that
the
presumption
is
overly
broad
(
0189,
0206,
0217,
0246,
0303,
0323).
Other
commenters
(
0206,
0303)
said
EPA
should
not
establish
a
prescribed
method
of
level
of
control
because
variations
between
small
and
large
facilities,
e.
g.,
between
coal­,
oil­,
and
gas­
fired
units,
and
the
age
of
the
units
make
application
of
such
a
presumption
overbroad.
Commenters
(
0217,
0246)
also
said
they
disagreed
with
the
presumptive
NOx
limit
at
0.27
233
lb/
MMBTU
because
of
the
many
differences
in
coal­
fired
boilers.
Commenter
0323
said
EPA
should
consider
a
range
of
emission
rates
that
reflect
recent
advances
in
technology,
particularly
with
respect
to
tangentially
fired
boilers
at
EGUs.

Three
commenters
said
EPA's
characterization
of
the
capabilities
of
control
technology
is
inaccurate.
Commenters
(
0217,
0268)
provided
results
of
a
survey
showing
the
effect
of
adding
advance
combustion
controls
to
existing
units
that
currently
do
not
have
the
most
advanced
controls.
Expected
NOx
emissions
rates
ranged
from
0.17
to
0.49
lbs/
MMBTU.
The
average
emissions
rate
achievable,
weighted
by
heat
input,
was
0.27
lbs/
MMBtu,
well
in
excess
of
the
presumed
limit
of
0.2
lbs/
MMBTU.
In
contrast,
commenter
0255
described
the
effectiveness
of
SNCR
as
a
candidate
to
meet
or
better
the
proposed
emission
rates.

Response:
See
preamble.
EPA's
analysis
for
the
final
BART
rulemaking
did
include
analysis
of
advanced
combustion
controls.

Comment:
Thirteen
commenters
said
it
is
important
to
consider
the
type
of
coal
or
other
fuel
in
setting
a
presumptive
control
level
(
0186,
0188,
0213,
0214,
0217,
0224,
0238,
0254,
0255,
0275,
0306,
0319,
0321).

Response:
See
preamble.

Comment:
Commenters
(
0186,
0275)
said
the
0.2
lb/
MMBTU
presumptive
level
of
NOx
control
should
not
apply
to
gas
fired
power
plant
boilers
because
combustion
controls
on
gas­
fired
units
can
achieve
a
much
better
(
i.
e.,
lower)
level
of
performance.

Response:
The
presumptive
limits
EPA
is
setting
only
apply
to
coal­
fired
EGUs
that
are
BART­
eligible.
See
preamble.

Comment:
Some
commenters
have
argued
that
SCR
technology
is
not
feasible
for
application
on
lignite­
fired
boilers.
These
commenters
indicated
that
SCR
had
not
been
demonstrated
to
effectively
control
NOx
emissions
from
lignite­
fired
boilers.
They
also
cited
experience
from
an
SCR
pilot
test
on
a
slipstream
from
a
boiler
firing
lignite
which
showed
a
problem
with
the
blinding
of
the
SCR
catalyst.
This
catalyst
fouling
was
believed
to
be
caused
by
sulfates
of
calcium
and
sodium
present
in
the
flue
gas
generated
from
the
firing
of
lignite.
18
Impacts
of
Texas
Lignite
Coal
on
SCR
Catalyst
Life
and
Performance:
Field
Data
from
TXU's
Martin
Lake
Plant,"
EPRI
Website,
www.
epri.
com,
accessed
on
December
11,
2003
(
docket
no.
OAR­
2003­
0053­
1944)

19
Email
from
J.
Staudt
to
S.
Khan,
Lignite
Installation
List,
July
5,
2004
(
docket
no.
OAR­
2003­
0053­
1946)

20
Email
from
J.
Staudt
to
S.
Khan,
Response
to
Comments
­
IAQR,
July
2,
2004
(
docket
no.
OAR­
2003­
0053­
1945)

21
Email
from
J.
Staudt
to
S.
Khan,
SCR
on
ND
Lignite,
August
20,
2004
(
docket
no.
OAR­
2003­
0053­
1947)

234
Other
commenters
(
0188,
0214,
0238,
0319,
0321)
said
that
the
0.2lb/
MMBTU
level
of
control
has
not
been
demonstrated
using
SCR
technology
on
lignite­
fired
EGUs
and
cited
current
research
supporting
this
claim
(
0188,
0321).
Commenter
0238
added
that
over­
fired
air
will
not
reduce
NOx
emissions
from
cyclones
to
that
level.
Commenter
0321
said
the
proposed
emission
limit
is
not
appropriate
for
lignite­
fired
cyclone
units
either.
Commenter
0306
said
EPA
should
reexamine
the
presumptive
BART
for
cyclone
boilers
firing
North
Dakota
lignite.

Response:

EPA
assumed
the
use
of
SCR
in
setting
presumptive
Nox
limits
for
coal­
fired
cyclone
units.
See
preamble.
Based
on
the
investigations
conducted
by
the
Agency,
there
is
considerable
experience
in
the
industry
to
show
that
use
of
SCR
on
lignite­
fired
boilers
is
technically
feasible.
A
slipstream
SCR
test
conducted
by
EPRI
at
the
750­
MW
Martin
Lake
Plant
in
Texas
has
shown
acceptable
results
from
Texas
Lignite.
18
Furthermore,
one
technology
supplier
reports
that
SCR
has
been
successfully
used
on
lignite
and
brown
coal
boilers
in
Europe.
19
The
SCR
catalyst
suppliers
have
also
indicated
that
they
will
offer
performance
guarantees
for
SCR
applications
on
lignite­
fired
boilers
in
the
U.
S.
20
These
suppliers
believe
that
the
overall
SCR
experience
from
facilities
firing
different
types
of
coals
is
applicable
to
the
lignite­
fired
facilities.
Therefore,
the
Agency
believes
that
application
of
the
SCR
technology
on
lignite
boilers
in
the
U.
S.
is
technically
feasible.

The
EPA
also
notes
that
the
commenters
have
not
provided
any
details
of
the
pilot
test
that
has
been
mentioned
with
regards
to
the
blinding
of
a
catalyst
from
North
Dakota
lignite­
generated
flue
gases.
Without
this
information,
this
part
of
the
comment
cannot
be
addressed
properly.
The
EPA's
own
investigations
show
that
the
main
reason
for
this
problem
was
ash
build
up
on
the
catalyst,
which
is
a
problem
commonly
experienced
in
pilot
test
facilities.
21
In
the
full­
scale
installations,
this
concern
is
addressed
during
the
SCR
design
stage.
The
methods
used
to
avoid
this
problem
include
duct
design
to
promote
ash
fallout
prior
to
the
SCR,
catalyst
reactor
design
to
avoid
ash
build
up,
and
on­
line
cleaning
methods
(
sootblowers
and
sonic
horns).
235
Comment:
Commenter
0217
said
coal
rank
also
accounts
for
differences
in
NOx
levels.
For
example,
subbituminous
coal
from
the
Powder
River
Basin
has
been
shown
to
yield
lower
NOx
in
some
units
than
other
coals.
However,
this
is
not
always
true.
In
addition,
unit
operators
must
balance
oxygen
levels
to
attain
proper
NOx
levels
while
still
maintaining
appropriate
carbon
monoxide
(
CO)
levels.
A
unit
with
a
stringent
CO
emission
limit
may
not
allow
a
plant
operator
to
achieve
the
lowest
NOx
level
possible
with
existing
controls.

Response:
In
response
to
comments,
EPA
revised
its
proposed
approach
for
setting
presumptive
Nox
limits
and
is
finalizing
presumptive
limits
based
on
unit
and
coal
type.
If
a
particular
unit
is
unable
to
meet
the
presumptive
limits,
the
State
may
consider
several
factors
in
determining
whether
the
source
must
meet
the
presumptive
limit.
Limitations
such
as
the
one
this
commenter
suggests
could
be
one
of
the
factors
States
consider.

Comment:
Commenter
0213
said
that
cyclone
boilers
that
burn
bituminous
coal
may
require
a
higher
presumptive
emission
rate
even
after
OFA
retrofits.
Commenter
0254
said
pre­
NSPS
boilers
firing
bituminous
coal
cannot
meet
the
presumptive
limit
with
the
prescribed
control
systems
(
combustion
controls
or
low­
NOx
burners).
Commenter
0224
described
different
levels
of
performance
between
Western
plants
(
which
could
meet
the
limit)
and
Eastern
plants
(
which
could
not
meet
the
limit).
The
commenter
said
any
presumptive
performance
limits
must
reflect
different
types
of
coal
and
boiler
configurations.
Commenter
0255
said
the
emission
rate
achieved
from
a
particular
unit
depends
on
several
factors,
including
coal
type
and
boiler
configuration.
The
commenter
described
several
examples
of
high
performing
units.

Response:
EPA
used
SCR
in
setting
presumptive
Nox
limits
for
coal­
fired
cyclone
units.
See
preamble.
Also
see
preamble
for
discussion
of
how
EPA
set
presumptive
limits
by
unit
and
coal
type.

Comment:
The
proposed
presumptive
NOx
emission
level
is
not
consistent
with
the
Bush
Administration's
publicly
announced
policy
of
"
no
SCR
in
the
West"
under
the
CSA.
EPA
should
conclude
that
application
of
combustion­
based
controls
would
constitute
BART
for
NOx
emissions
in
the
West,
irrespective
of
the
resulting
emission
level.

Response:
EPA
has
used
combustion
controls
for
setting
presumptive
BART
Nox
limits
except
for
one
boiler
type
(
cyclones).
See
preamble
for
discussion
of
why
EPA
used
SCR
on
cyclone
units.
236
Comment:
Seven
commenters
said
that
the
boiler
configuration
should
be
considered
(
0195,
0210,
0217,
0224,
0241,
0256,
0270).
Commenters
(
0210)
said,
if
retained,
the
NOx
presumption
should
be
modified
to
address
the
appropriate
control
levels
for
all
boiler
types.
EPA's
identified
presumptive
level,
0.2
lbs/
MMBTU,
is
not
reasonable
for
most
boiler
types
because
it
cannot
be
achieved
without
using
combustion
controls
(
0241,
0256,
0270).
EPA
could
establish
more
reasonable
levels
by
reviewing
the
emissions
data
readily
available
from
the
Acid
Rain
program
(
0210).
Commenter
0195
said
NOx
control
on
cyclone
type
boilers
represent
unique
circumstances
that
make
it
virtually
improbable
to
control
on
a
long
term
basis
below
0.4
lb/
MMBTU.

Response:
See
preamble.
EPA
reviewed
data
available
from
the
Acid
Rain
program
as
well
as
the
NOx
SIP
Call
to
determine
presumptive
Nox
limits.
EPA
used
SCR
in
setting
presumptive
limits
for
cyclone
units.
See
preamble
for
further
discussion.

Comment:
Commenter
0270
said
it
is
well
known
that
wall­
fired
boilers
can't
achieve
the
same
level
of
NOx
controls
as
tangentially­
fired
boilers
with
these
types
of
control
technologies.
The
commenter
does
not
believe
that
a
limit
of
0.2
lbs/
MMBTU
can
be
achieved
on
wall­
fired
units
via
retrofit
installation
of
advance
combustion
controls.

Response:
In
response
to
comments,
EPA
revised
its
approach
for
setting
presumptive
Nox
limits
for
coal­
fired
EGUs.
See
preamble.

Comment:
Commenter
0195
supported
the
definitions
for
Combustion
controls
as
defined
in
40
CFR
76.2
Acid
Rain
Nitrogen
Oxides
Emission
Reduction
Program
in
order
to
provide
consistency
across
EPA
programs.

Response:
EPA's
analysis
assumed
current
combustion
control
technology
means
low
Nox
burners,
over­
fire
air,
and
coal
reburning.
See
preamble.
This
is
very
similar
to
the
definition
of
combustion
controls
in
40
CFR
76.2.

Comment:
Commenters
supported
the
presumption
that
BART
can
be
achieved
without
the
use
of
post­
combustion
controls
(
0199).
However
some
types
of
units
or
units
burning
some
types
of
237
fuel
cannot
meet
the
presumptive
limits
without
installing
controls
(
0199,
0217,
0244,
0252,
0253).
Commenters
described
several
examples.

Response:
In
response
to
comments,
EPA
has
revised
its
proposed
approach
for
setting
presumptive
NOx
limits
and
is
finalizing
an
approach
for
setting
presumptive
NOx
limits
that
reflects
differences
in
boiler
type
and
coal
type.
See
preamble.
The
presumptive
NOx
limits
EPA
is
finalizing
do
not
assume
that
all
BART­
eligible
units
can
meet
the
presumptive
NOx
limits.
See
preamble.
EPA's
analysis
shows
that
most
units
that
cannot
meet
the
final
presumptive
NOx
limits
could
meet
the
presumptive
limits
using
advanced
combustion
controls.
Furthermore,
the
State
may
consider
the
appropriate
BART
factors
in
deciding
what
NOx
limits
a
particular
BART­
eligible
unit
must
meet.

Comment:
Commenters
(
0206,
0303)
said
the
costs
associated
with
SCR
operations
at
coal­
fired
power
plants
are
not
insignificant.
Annual
operation,
rather
than
only
during
ozone
season,
would
significantly
reduce
catalyst
life,
requiring
more
frequent
replacement.

Response:
EPA
estimated
the
costs
of
operating
SCRs
installed
at
existing
BART­
eligible
coal­
fired
EGUs
an
additional
seven­
months
each
year,
i.
e.,
year­
round
operation
as
opposed
to
operation
during
the
five­
month
ozone
season
only.
The
average
cost
of
operating
existing
SCRs
year­
round
is
approximately
$
170/
ton.
EPA
considers
$
170/
ton
to
be
highly
cost­
effective
reductions
and
therefore
is
requiring
units
with
existing
SCRs
to
operate
year­
round
to
comply
with
BART.
See
excel
spreadsheet
in
the
docket
entitled
"
Estimated
cost
of
operating
existing
SCRs
and
SNCRs
on
coal­
fired
BART­
eligible
EGUs
year­
round
as
opposed
to
during
ozone
season
only."

Comment:
Commenters
agreed
that
low­
NOx
burners
should
be
the
presumptive
technology
because
they
are
cost
effective
and
the
incremental
cost
of
adding
SCR
or
SNCR
would
be
substantially
less
cost
effective
(
0224).

Response:
See
preamble
for
a
discussion
of
why
EPA
used
SCR
for
cyclone
units.
EPA
used
current
combustion
controls
for
all
other
boiler
types
in
setting
presumptive
limits.

Comment:
Commenter
0255
said
post­
combustion
controls
(
e.
g.
SCR
and
SNCR)
as
well
as
advanced
(
or
second
generation)
low­
NOx
burners
are
significantly
more
efficient
and
are
currently
the
best
available
control
technologies
for
NOx.
The
majority
of
the
low­
NOx
burners
currently
on
coal­
fired
units
were
installed
to
meet
the
compliance
deadline
for
the
start
of
Phase
I
and
II
of
the
238
Acid
Rain
Program
in
1995
and
2000,
respectively.
Units
that
installed
their
low­
NOx
burners
for
Phase
I
of
the
Acid
Rain
Program
will
have
had
their
air
pollution
control
equipment
in
place
for
almost
20
years
by
the
time
the
BART
rule
is
implemented
in
2013.
Even
the
units
that
installed
low­
NOx
burners
by
2000
will
have
had
their
equipment
in
place
for
over
10
years
by
2013.
Exemptions
for
existing
low­
NOx
burner
equipment
that
is
not
meeting
the
0.2
lb/
MMBtu
emission
rate
should
not
be
given
as
technology
performance
has
significantly
improved
since
1995
and
will
continue
to
improve
out
to
2013.
For
these
reasons,
EPA
should
remove
the
provision
that
permits
sources
with
first
generation
low­
NOx
burners
presumptive
NOx
emission
rate
and
mandate
that
BART­
eligible
sources
meet
the
0.2
lb/
MMBTU
or
lower
emission
rate.
Commenter
0255
also
provided
information
showing
that
good
performance
of
SCR
on
lignite
burning
units.

Response:
EPA's
analysis
assumed
units
that
installed
combustion
controls
(
low
NOx
burners
and/
or
over­
fire
air)
prior
to
1997
would
have
to
replace
their
existing
combustion
controls
with
current
combustion
control
technology.
See
preamble
and
Technical
Support
Documents
referenced
in
the
preamble
for
further
discussion.
Combustion
controls
installed
in
1997
or
later
were
considered
current
combustion
control
technology
and
therefore
were
not
replaced.
EPA
chose
1997
as
the
cutoff
because
EPA
believes
it
is
reasonable
to
assume
that
the
generation
of
combustion
controls
developed
for
the
Acid
Rain
program
NOx
requirements
became
widely
operational
by
1997.
EPA
considered
a
later
cutoff
date
but
chose
1997
in
order
to
be
conservative
and
minimize
the
number
of
units
that
may
have
to
replace
existing
combustion
controls.

Comment:
Some
commenters
approved
of
including
year­
round
SCR
operation
for
removal
of
NOx
(
0179,
0221,
0232,
0234,
0271).
Commenters
(
0232,
0234)
would
limit
this
application
to
sources
already
operating
SCR.
Commenter
0221
described
how
the
technology
has
advanced,
is
cost
effective,
and
will
meet
levels
of
0.15
lb/
MMBTU
(
NOx
SIP
call)
to
0.125
(
IAQR/
CAIR)
at
costs
estimated
by
EPA
to
be
less
than
$
2,000/
ton.
The
commenter
said
that
SCR
represents
the
best
available
technology
for
removal
of
NOx.
Commenter
0232
said
a
greater
control
efficiency
is
possible
for
these
sources.

Response:
EPA
agrees
and
is
establishing
year­
round
operation
of
SCR
or
SNCR
as
BART
for
units
with
existing
SCR
or
SNCR.
See
preamble.

Comment:
Commenter
0267
said
the
NOx
level
should
be
lowered
to
require
application
of
post­
combustion
controls.
At
a
minimum,
application
of
post­
combustion
controls
for
NOx
should
not
be
ruled
out
as
BART.
Commenter
0280
agreed
and
said
the
presumptive
NOx
level
of
0.2
lbs/
MMBTU
should
be
considered
as
the
maximum,
not
minimum,
allowable
emission.
States
should
require
the
lowest
emission
rate
that
can
be
achieved
using
all
available
and
applicable
control
technologies
as
in
BACT.
The
economic
feasibility
should
be
the
limiting
criteria.
239
Response:
EPA
has
used
SCR
in
setting
presumptive
limits
for
cyclone
units.
See
preamble.
For
all
other
unit
types,
EPA
has
used
current
combustion
control
technology
for
setting
presumptive
NOx
limits.
EPA
used
cost­
effectiveness
as
a
guide
in
determining
whether
to
use
SCR
or
current
combustion
control
technology
in
setting
presumptive
limits.
See
preamble.

Comment:
Commenter
0256
said
if
EPA
does
retain
the
presumptions,
at
a
minimum,
the
presumptions
should
not
apply
to
units
that
already
are
controlled
with,
for
SO2,
scrubbers,
and,
for
NOx,
combustion
controls
that
were
added
after
enactment
of
Title
IV
of
the
Act.
Units
with
combustion
controls
that
were
added
after
enactment
of
Title
IV
should
be
deemed
to
already
have
the
best
combustion
controls
available,
and
thus
should
not
be
subject
to
the
presumption.
Again,
the
state
should
be
allowed
to
analyze
BART
for
those
units
applying
the
statutory
factors
in
light
of
the
relatively
recent
NOx
controls
that
were
retrofit
at
the
unit.

Response:
EPA's
analysis
assumed
units
that
installed
combustion
controls
(
low
NOx
burners
and/
or
over­
fire
air)
prior
to
1997
would
have
to
replace
their
existing
combustion
controls
with
current
combustion
control
technology.
See
preamble
and
Technical
Support
Documents
referenced
in
the
preamble
for
further
discussion.
Combustion
controls
installed
in
1997
or
later
were
considered
current
combustion
control
technology
and
therefore
were
not
replaced.
EPA
chose
1997
as
the
cutoff
because
EPA
believes
it
is
reasonable
to
assume
that
the
generation
of
combustion
controls
developed
for
the
Acid
Rain
program
NOx
requirements
became
widely
operational
by
1997.
EPA
considered
a
later
cutoff
date
but
chose
1997
in
order
to
be
conservative
and
minimize
the
number
of
units
that
may
have
to
replace
existing
combustion
controls.

Comment:
Commenter
0221
said
that
because
power
plants
account
for
about
one­
quarter
of
the
nitrogen
oxide
emissions
nationally,
and
because
the
Agency
also
has
and
will
have
considerable
"
experience
in
evaluating
NOx
control
options
for
utility
boilers,"
a
presumptive
control
level
for
NOx
should
be
presumed
as
"
best
available"
at
the
outset
of
the
required
top­
down
BART
analysis
for
both
controlled
and
uncontrolled
power
plants.
In
order
to
provide
support
for
such
a
presumption,
EPA
should
undertake
a
similar
analysis
of
NOx
removal
technologies
as
was
done
by
ORD
for
sulfur
dioxide.

Response:
EPA
has
completed
a
similar
analysis
of
NOx
removal
technologies.
See
Technical
Support
Document
entitled
"
Methodology
for
Developing
BART
NOx
Presumptive
Limits,"
footnotes
1
and
2.
240
Comment:
Commenter
0255
provided
data
on
100
coal­
fired
units
and
the
majority
are
very
high
performing.
The
incentive
to
control
emissions
under
the
NOx
SIP
Call
will
become
stronger
over
time,
assuming
electric
demand
continues
to
grow,
as
sources
utilize
their
banked
emissions.
As
the
bank
of
emission
credits
is
reduced,
additional
NOx
reductions
will
be
required.
Some
sources
will
be
required
to
install
control
technologies
while
those
with
controls
will
have
an
incentive
to
operate
their
control
technologies
at
their
highest
efficiency
so
they
can
sell
the
emission
credits
that
they
offset.

Response:
EPA's
analysis
included
all
BART­
eligible
coal­
fired
EGUs
that
EPA
could
identify
(
491
total
coal­
fired
EGUs).
See
preamble
and
Technical
Support
Documents
referenced
in
the
preamble
for
further
discussion.
EPA
used
its
analysis
in
setting
presumptive
NOx
limits
for
coal­
fired
units
greater
than
200
MW
at
plants
in
excess
of
750
MW.
In
setting
presumptive
limits,
it
was
necessary
to
consider
all
BART­
eligible
coal­
fired
EGUs.
The
list
of
100
coal­
fired
units
the
commenter
references
is
not
as
inclusive
as
the
population
of
units
EPA
used
in
setting
presumptive
NOx
limits.
See
preamble
for
further
discussion.

Comment:
Commenter
0314
expects
continuing
technological
improvement
in
the
future
and
that
a
specific
presumptive
BART
level
set
by
rule,
will
make
it
difficult
for
a
state
to
establish
a
lower
level.
As
proposed
for
SO2,
the
commenter
suggested
a
range
for
NOx
that
will
account
for
the
variability
in
the
boilers
as
well
as
technology
improvements
in
the
future.
A
range
of
0.13
to
0.17
lb/
MMBTU
was
suggested.

Response:
EPA
acknowledges
the
likelihood
of
future
improvements
in
NOx
control
technology.
EPA
considered
both
current
and
advanced
combustion
control
technology
in
setting
presumptive
NOx
limits
for
coal­
fired
EGUs
and,
in
response
to
comments,
revised
its
proposed
approach
for
setting
presumptive
limits
and
is
finalizing
an
approach
that
recognizes
the
differences
in
boiler
type
and
coal
type
in
setting
presumptive
limits.
States
have
flexibility
to
impose
limits
more
stringent
than
the
presumptive
limits
EPA
is
finalizing
should
the
State
determine
that
EPA's
presumptive
limits
no
longer
reflect
current
combustion
control
technology,
or
if
the
State
determines
a
more
stringent
limit
is
required
at
a
particular
source.

Comment:
Additionally,
cost
effective
emission
reductions
from
the
industrial
sector
are
feasible
but
have
not
been
adequately
considered
in
the
BART
rule
considering
the
quantity
of
visibility
impairing
emissions.
For
industrial
sources,
setting
an
appropriate
level
of
reduction
with
some
flexibility
such
as
providing
the
option
of
meeting
an
emission
rate
limit
or
percent
reduction
is
one
way
to
assure
significant
emission
reductions.
241
Response:
We
agree
and
believe
we
have
done
this.

Comment:
Commenters
(
0189)
said
that
EPA
should
make
clear
(
a)
that
add­
on
NOx
controls
should
not
be
part
of
BART
for
non­
EGUs
because
of
the
costs
of
the
these
controls
and
(
b)
that
its
presumptive
control
level
of
.2
lbs/
MMBTU
does
not
apply
to
non­
EGUs
because
they
are
unable
to
meet
these
levels
without
the
use
of
add­
on
controls.

Response:
See
the
preamble.

Comment:
BART
determinations
should
be
made
by
individual
states
on
a
case­
by­
case
basis,
consistent
with
state
discretion.

Response:
We
agree.
See
the
preamble.

Comment:
Commenters
(
0214,
0238)
added
that
EPA
should
give
the
states
maximum
flexibility
to
use
the
five
statutory
factors
in
their
BART
NOx
determinations.

Response:
See
the
preamble.
We
believe
these
limits
are
reasonable
and
that
States
have
discretion
to
make
different
determinations,
if
warranted
based
on
the
BART
factors.

Comment:
Commenter
0321
added
that
NOx
emissions
from
lignite
unite
appear
to
have
less
impact
on
visibility
in
that
region
of
the
country.

Response:
If
States
find
that
specific
plants
in
specific
locations
have
minimial
effect
on
visibility
impairment,
they
have
discretion
as
to
the
appropriate
BART
determination.

Comment:
Commenters
(
0319,
0321)
supported
a
recommendation
for
a
NOx
limit
of
approximately
0.4
lb/
MMBTU,
or
evaluated
on
a
case­
by­
case
basis.
242
Response:
See
the
preamble.

Comment:
Commenter
0252
analyzed
data
for
eastern
units
and
western
units
with
different
configurations
and
for
units
burning
different
types
of
coal.
The
commenter
concluded
that
EPA
should
not
set
a
presumptive
level
of
control,
since
it
should
be
determined
on
a
case­
by­
case
basis.

Response:
See
the
preamble
and
technical
support
document.

Comment:
Some
commenters
were
concerned
that
the
proposed
rule
would
require
some
plants
to
install
SCR
to
meet
the
NOx
control
level
proposed
(
0171,
0206,
0217).
Commenter
0171
said
the
implementation
of
the
regional
haze
rule,
including
any
potential
retrofit
of
SCR
technology,
must
consider
the
site
specific
data
of
the
Class
I
area,
the
use
of
current
emission
controls
on
potential
sources,
and
the
degree
of
visibility
improvements.
Commenters
(
0217,
0246)
added
that
individual
EGU
differences
need
to
be
accounted
for
by
states
in
making
their
BART
determination.
Also,
the
guidelines
indicate
that
if
a
state
finds
that
a
source's
visibility
contribution
warrants
the
installation
of
SCR,
it
may
be
allowed.
However,
the
guidelines
need
to
also
provide
for
instances
where
the
visibility
condition
warrant
a
lesser
control
level
than
what
would
be
achieved
by
advanced
combustion
control.
There
is
reference
to
this
concept
in
the
preamble
that
is
not
reflected
in
the
guidelines.

Response:
See
the
preamble.
We
agree
that
States
have
the
discretion
whether
to
require
SCR
and
that
States
have
discretion
to
determine
whether
visibility
conditions
particular
control
levels.

Comment:
One
commenter
(
0270)
cited
studies
showing
the
nitrates
contribute
only
a
few
percent
to
total
light
extinction
except
in
Class
I
areas
downwind
of
Southern
California.
Accordingly,
NOx
emissions
from
stationary
sources
are
unlikely
to
lead
to
any
perceptible
changes
in
visibility.
Therefore,
it
is
unlikely
that
NOx
emission
sources
in
the
West
will
be
good
candidates
of
application
of
BART.

Response:
See
the
preamble
for
reference
to
studies
discussing
nitrate
concentrations
in
the
West,
as
well
as
additional
studies
in
the
docket.

Comment:
Commenter
0241
said
states
should
undertake
a
full
BART
analysis,
as
required
by
the
Act,
to
determine
the
appropriate
NOx
control
level
that
should
be
established
as
BART
for
EGUs.
If
a
presumption
is
retained,
EPA
should
change
the
presumption
from
a
lb/
MMBtu
emission
limitation
243
to
a
statement
indicating
that
combustion
controls
installed
after
enactment
of
Title
IV
of
the
Act
should
be
presumed
to
satisfy
the
BART
requirement
for
NOx.
States
could
then
evaluate
on
a
boiler­
specific
basis
the
emission
limitation
that
could
be
met
using
combustion
controls.
Any
other
approach
is
simply
not
justified
by
the
data
and
would
be
arbitrary
and
capricious.

Response:

The
presumptive
emission
limits
are
based
on
combustion
controls,
as
the
preamble
discusses.

Comment:
Commenter
0271
said
the
presumptive
level
of
control
must
be
applied
to
all
units
at
a
"
powerplant
having
a
total
generating
capacity
in
excess
of
750
megawatts"
even
those
units
less
than
750
megawatts.
This
conforms
to
the
specific
wording
of
the
Clean
Air
Act
(
Sec.
169A(
b)(
2)(
A))
for
controls
on
these
smaller
than
750
megawatt
units.

Response:
We
agree
that
the
presumptive
limits
should
not
just
apply
to
750
MW
units.
See
the
preamble.

Comment:
Eight
commenters
made
general
comments
on
the
proposed
NOx
levels
(
0181,
0186,
0215,
0225,
0241,
0268,
0275,
0314).
Commenter
0181
said
EPA
should
remove
the
presumptive
NOx
control
levels
from
the
final
BART
guidelines
and
instead
replace
it
with
a
site­
specific
cost
analysis.
Once
again,
the
presumptive
NOx
emission
control
approach
inappropriately
ignores
the
need
for
a
visibility
impact
evaluation
which
is
required
in
Step
5
of
the
proposed
case­
by­
case
BART
engineering
analysis.
Commenter
0268
such
a
presumptive
limit
infringes
on
a
state's
authority
to
establish
BART
on
a
case­
by­
case
basis
considering
not
only
visibility
improvement,
but
other
factors
as
well.
Commenter
0186
said
EPA's
proposal
for
no
post­
combustion
NOx
controls
on
power
plants
should
apply
equally
across
all
source
categories;
otherwise,
there
should
be
no
presumptive
control
technology
prohibition
for
any
source
category.

Response:
See
the
preamble.
States
have
discretion
to
exercise
judgment
in
considering
visibility
and
the
other
BART
factors.

Comment:
Commenter
0215
said
if
an
EGU
is
achieving
the
default/
presumption
for
NOx,
it
should
not
be
required
to
do
a
BART
determination
for
that
pollutant.
Such
presumptions
would
streamline
the
BART
process
for
those
sources
willing
to
reach
these
control
levels.

Response:
We
agree
that
a
streamlined
BART
process
is
warranted
for
such
sources.

Comment:
244
Commenter
0225
said
for
sources
located
within
50
km
of
a
Class
I
area,
the
source
could
be
required
to
perform
an
evaluation
of
control
technologies
that
compare
their
cost
effectiveness.
A
number
of
the
large
power
companies
have
completed
these
types
of
evaluations
and
have
a
good
understanding
of
the
cost
of
the
various
control
scenarios
for
their
units.
States
may
want
to
request
cost
effectiveness
evaluations
from
sources
to
determine
if
a
source
can
cost
effectively
reduce
their
emissions.
Because
a
larger
number
of
sources
outside
of
the
NOx
SIP
Call
region
have
not
been
required
to
install
many
advanced
NOx
controls,
cost
effective
reductions
can
be
achieved.

Response:
States
have
discretion
under
the
Guidelines
to
require
such
an
analysis.

Comment:
Commenter
0241
said,
for
the
same
reasons
identified
above
with
respect
to
the
presumptive
SO2
control
level
for
EGUs,
the
commenter
objects
to
EPA'
s
proposal
to
establish
a
presumptive
NOx
control
level
for
EGUs.

Response:
See
the
preamble.

Comment:
Commenter
0275
supported
EPA's
proposed
finding
that
add­
on
NOx
controls
would
be
too
costly
to
be
considered
BART
for
EGUs,
in
most
cases.
Even
more
so,
this
would
hold
true
for
non­
EGU
boilers,
"
smaller
boilers
are
generally
less
cost
effective
to
control."
However,
this
finding
cannot
be
made
unless
a
demonstration
is
made
that
visibility
impairments
do
not
remain.
That
is,
NOx
add­
on
controls
for
EGUs
and
non­
EGUs
may
be
appropriate
if
modeling
indicates
that
2018
visibility
goals
will
otherwise
not
be
met.

Response:
2018
visibility
goals
are
to
be
assessed
in
the
regional
haze
State
implementation
plan
(
SIP)
under
the
regional
haze
rule.
In
addition,
the
CAA
requires
BART
determinations
on
specific
sources.

Comment:
Four
commenters
discussed
the
averaging
period
for
the
emission
limits
(
0207,
0214,
0238,
0314).
Commenters
said
the
averaging
period
is
unclear
and
should
be
annual
vs.
a
short­
term
limit
(
0207,
0214,
0238).
Commenter
0314
suggested
a
30­
day
rolling
average.

Response:
We
agree
and
have
specified
a
30­
day
averaging
period
in
the
Guidelines.

Comment:
Commenter
0241
said
if
EPA
intends
to
retain
a
presumption
regarding
NOx
controls
in
the
final
Guidelines,
EPA
should
specify
only
that
EGUs
should
be
equipped
with
combustion
245
controls
to
limit
NO
x
emissions
without
identifying
a
presumed
NO
x
emission
rate.
Commenters
(
0244,
0252)
agreed
that
EPA
should
not
set
a
presumptive
level
of
control,
since
it
should
be
determined
on
a
case­
by­
case
basis.
However,
if
a
limit
must
be
set
as
guidance
for
the
states
it
should
be
higher
than
0.20
lb/
MMBTU,
in
order
to
avoid
forcing
the
use
of
post
combustion
controls
(
e.
g.
SCR
and
SNCR.).

Response:
See
the
preamble.

10.0
BART
and
Trading
Programs
Note:
Numerous
comments
were
received
on
the
portion
of
the
proposed
BART
guidelines
which
would
have
provided
guidance
for
trading
programs
developed
in
lieu
of
BART.
The
final
BART
guidelines
do
not
contain
such
a
section.
Instead,
issues
related
to
trading
programs
or
other
alternative
programs
in­
lieu
of
BART
will
be
addressed
in
a
separate
rulemaking,
in
order
to
respond
to
the
recent
decision
of
the
D.
C.
Circuit
in
Center
for
Energy
and
Economic
Development
v.
EPA,
No.
03­
1222,(
D.
C.
Cir.
Feb.
18,
2005).
Comments
relevant
to
trading
programs
in
general
will
be
addressed
in
that
rule
making.
This
section
of
the
document
responds
to
comments
regarding
the
criteria
for
evaluating
whether
an
alternative
program
makes
greater
reasonable
progress
than
BART,
and
on
the
fundamental
legal
and
policy
issues
raised
by
the
substitution
for
BART
of
emission
trading
programs,
particularly
as
related
to
the
CAIR.

10.1
Whether
a
Trading
Program
Can
be
Better
than
BART
Comment:
Nine
commenters
supported
use
of
market­
based
trading
programs
(
0187,
0199,
0207,
0209,
0211,
0216,
0236,
0237,
0275)
and
cited
their
many
benefits
in
terms
of
emission
reductions,
efficient
use
of
states'
staffing
and
enforcement
resources,
incentives
for
early
reductions,
incentives
for
development
of
new,
better
technologies,
etc.
Some
commenters
specifically
supported
the
CAIR,
Clear
Skies
Initiative,
implementation
of
8­
hour
ozone
and
fine
PM
controls,
clean
fuels
and
engines,
and
MACT
in
their
ability
to
satisfy
the
BART
requirements
(
0199,
0209,
0211,
0237,
0275).
Commenter
0237
said
EPA
should
not
require
States
to
require
BART
reductions
if
they
are
able
to
demonstrate
that
federal
measures,
including
CAIR,
meet
the
regional
haze
reasonable
progress
milestone
of
a
10%
deciview
visibility
improvement
on
the
20%
most
impaired
days
by
2018.
EPA
has
the
authority
to
avoid
the
effort
and
cost
on
the
part
of
states
and
affected
sources
for
BART
determinations
where
they
are
not
needed
to
achieve
the
2018
goals
and
there
is
no
requirement
to
impose
additional
controls
as
a
result
of
that
analysis.

Similarly,
Commenters
(
0186,
0275)
said
EPA
should
allow
states
to
consider
the
entire
mix
of
federal
controls
being
implemented
(
e.
g.,
cleaner
fuels
and
engines,
interstate
transport
controls,
the
8­
hour
ozone
and
PM2.5
standards)
to
determine
if
additional
local
controls
are
need
to
meet
national
visibility
goals
in
2018
and
beyond.
In
cases
where
emission
reductions
from
federal
controls
alone
will
be
sufficient
to
achieve
the
2018
visibility
improvement
goals,
EPA
should
regard
the
BART
obligation
as
satisfied
for
all
sources.
Commenter
0310
discussed
the
246
progress
that
the
electric
utility
industry
has
already
made
in
achieving
massive
reductions
in
SO
2
and
NO
x
emissions.

Commenters
(
0169,
0299)
said
states
should
be
free
to
implement
an
emissions
trading
program
or
other
alternative
measure
instead
of
requiring
BART
and
endorsed
the
options
that
states
could
include
non­
BART­
eligible
sources
in
a
trading
program,
and
have
the
flexibility
to
implement
an
emissions
trading
program
in
conjunction
with
other
states.
Commenter
0277
approved
of
the
option
of
allowing
non­
BART
sources
to
participate
in
a
trading
program.
Commenter
0271
said
the
flexibility
of
a
trading
program
could
be
an
available
control
option
to
all
visibility
impairing
sources
in
2012
and
to
BART
sources
to
meet
additional
control
requirements
beyond
BART
levels.

Response:
The
EPA
also
supports
the
use
of
market
based
trading
programs
as
cost
effective
means
to
meet
air
quality
objectives,
where
appropriate,
as
evidenced
by
our
determination
that
CAIR
may
substitute
for
BART
for
EGUs
in
CAIR
affected
States.
We
also
believe
it
is
appropriate
and
desirable
to
include
non­
BART
sources
in
such
trading
programs,
in
order
to
increase
market
viability.
However,
for
the
following
reasons,
it
premature
to
make
any
judgements
as
to
what
is
needed
to
make
reasonable
progress.

The
Regional
Haze
Rule,
at
40
CFR
51.308(
d)(
1),
provides
the
process
for
States
to
follow
in
setting
reasonable
progress
goals.
In
summary,
this
process
entails:
(
a)
determining
the
uniform
rate
of
progress
needed
to
reach
natural
background
conditions
in
60
years,
(
b)
determining
the
emission
reductions
needed
to
achieve
the
rate
for
the
first
10­
year
strategy
period,
(
c)
considering
certain
statutory
factors
relative
to
potentially
affected
sources
(
the
factors
are
the
costs
of
compliance,
time
necessary
for
compliance,
energy
and
non­
air
environmental
impacts,
and
remaining
useful
life).
After
conducting
this
analysis,
the
State
sets
the
reasonable
progress
goals
for
each
Class
I
area.
If
a
reasonable
progress
goal
is
set
at
a
point
representing
less
than
the
uniform
rate
of
progress
to
natural
background
conditions,
the
state
must
explain
why
the
slower
rate
of
progress
is
reasonable.
Conversely,
if
the
State
determines
that
a
faster
rate
of
progress
is
reasonable,
the
State
should
adopt
goals
reflecting
that
rate.
See
64
FR
35714,
35732
(
July
1,
1999).

After
determining
the
reasonable
progress
goals
for
each
Class
I
area,
the
State
then
incorporates
the
measures
needed
to
achieve
those
goals
in
its
long
term
strategy.
See
40
CFR
51.308(
d)(
3)(
v).
Long
term
strategies
must
also
requires
states
to
consider
all
anthropogenic
sources
of
visibility
impairment,
including
major
and
minor
stationary
sources,
mobile,
and
area
sources,
smoke
management
techniques
for
agricultural
and
forestry
management.
40
CFR
51.308(
d)(
3)(
iv).

Thus,
the
process
of
setting
reasonable
progress
goals
and
appropriate
long
term
strategies
involves
the
examination
of
all
source
categories,
includes
the
application
of
statutory
247
factors
to
determine
reasonableness,
and
is
not
terminated
by
a
finding
that
reductions
from
any
particular
source
category
are
sufficient
to
achieve
the
uniform
rate
to
natural
conditions.

Therefore,
the
proper
question
is
not
whether
federal
programs
"
achieve"
the
reasonable
progress
goals,
but
rather
what
should
those
goals
be,
in
light
of
CAIR
and
other
new
and
ongoing
measures;
and
what
additional
measures,
if
any,
are
needed
to
achieve
those
goals.

Comment:
Sixteen
commenters
said
other
proposed
rules
such
as
the
Clean
Skies
Initiative
and
the
CAIR
provide
no
guarantee
that
air
quality
in
the
parks
will
improve
and
EPA
should
not
allow
these
other
programs
to
substitute
for
BART
requirements
for
power
plants
(
0148,
0153,
0154,
0159,
0162,
0163,
0164,
0203,
0203,
0204,
0221,
0232,
0260,
0284,
0292,
0318).
Commenters
(
0203,
0204,
0221,
0232)
added
that
the
overdue
effort
to
improve
air
quality
in
public
lands
must
be
harmonized
with
continuing
attempts
to
reduce
the
interstate
transport
of
fine
particle
and
ozone
pollution.
A
fully
protective
federal
approach
that
targets
park­
specific
pollution
sources
and
interstate
transport
of
air
pollution
will
help
ensure
that
the
RHRs
reasonable
progress
goals
are
met
during
the
planning
period.

Response:
Although
no
predictions
of
the
future
can
be
guaranteed,
the
EPA
believes
that
the
demonstration,
based
on
emissions
projections
and
air
quality
modeling,
that
CAIR
makes
greater
reasonable
progress
than
BART,
is
very
robust.
We
have
sought
to
harmonize
the
interstate
transport
requirements
of
the
CAIR
with
the
visibility
provisions
of
the
CAA
through
the
determination
that
CAIR
may
be
considered
a
BART
substitute
for
EGUs
in
CAIR
affected
States,
while
maintaining
park
specific
protections
through
the
reasonably
attributable
BART
provisions.

Comment:
Commenter
0237
was
concerned
that
by
promulgating
its
interpretations
for
BART,
EPA
will
argue
that
they
have
now
been
codified
for
BACT
and
LAER.
The
commenter
asked
EPA
to
clarify
its
intent
so
that
there
will
be
a
final
record
of
review
if
the
commenter
believes
that
EPA's
final
determinations
on
the
various
definitions
are
necessary.

Response:
It
is
not
clear
what
"
interpretations"
applicable
to
BART
the
commenter
believes
might
be
extended
to
the
BACT
and
LAER
context.
Although
there
may
be
similarities
between
the
programs,
there
are
of
course
also
differences
in
terms
of
the
statutory
provisions
involved
and
the
respective
purposes
of
the
programs.
Therefore
it
is
not
possible
to
say
in
the
abstract
whether
or
not
an
interpretation
under
BART
will
necessarily
apply
under
BACT
or
LAER.

10.2
Criteria
for
Demonstrating
that
an
alternative
program
makes
greater
reasonable
progress
than
BART
10.2.1
Criteria
in
General
248
Comment:
One
commenter
(
A200028­
335)
noted
that
EPA
proposes
to
require
that
any
state
emission
trading
program
achieve
greater
visibility
improvement
than
would
be
achieved
via
source­
specific
BART.
However,
in
considering
greater
visibility
improvement
that
sourcespecific
BART
itself
achieves,
it
will
be
important
to
keep
in
mind
that
there
will
surely
be
exceptions
on
a
case­
by­
case
basis
from
the
presumptions
that
are
established
pursuant
to
the
BART
guidelines.

Response:
The
requirement
to
achieve
greater
reasonable
progress
than
BART
is
a
separate
issue
from
how
emission
reductions
and
visibility
improvements
achievable
by
BART
are
estimated.
The
latter
issue
is
where
the
role
of
presumptive
control
requirements
enter
into
the
trading
context.
This
issue
will
be
addressed
in
the
forthcoming
rulemaking
on
BART
trading
programs.

Comment:

Three
commenters
(
A200028­
266,
A200028­
278,
A200028­
312)
asked
what
the
criteria
are
for
"
greater
reasonable
progress"
on
the
20%
worst
visibility
days
and
on
the
20%
best
visibility
days?

Response:
As
proposed
in
the
BART
guidelines
and
finalized
as
40
CFR
51.308(
e)(
3),
the
criteria
for
greater
reasonable
progress
than
BART
are
(
1)
no
degradation
at
any
Class
I
area,
and
(
2)
greater
overall
improvement
based
on
average
improvement
at
all
affected
Class
I
areas
As
we
explained
in
the
final
BART
preamble,
we
believe
that
each
prong
of
the
test
should
ideally
be
based
on
an
examination
of
both
the
worst
20%
and
best
20%
of
days.

Comment:
Seven
commenters
(
A200028­
239,
A200028­
247,
A200028­
249,
A200028­
266,
A200028­
289,
A200028­
352,
A200028­
362)
asserted
that
the
visibility
results
from
any
cap­
andtrade
program
should
be
better
than
under
BART.
It
is
not
enough
to
simply
reduce
tons
of
emissions
nationwide,
as
the
locations
of
the
emissions
reductions
are
critical
to
improving
visibility
in
specific
parks
and
wilderness
areas.
Seven
commenters
said
the
rule
must
ensure
that
the
required
reductions
will
result
in
improved
visibility
in
all
affected
national
parks
and
wilderness
areas
(
0147,
0148,
0159,
0162,
0163,
0164,
0292).

Other
commenters
said
that
reasonable
progress
is
a
goal
and
that
natural
conditions
preclude
achieving
that
goal
(
0207).
"
Better
than
BART"
should
be
evaluated
in
the
context
of
reasonable
progress.
Commenter
0217
said
the
proposed
approach
to
demonstrate
reasonable
progress
would
undermine
the
fundamental
role
of
the
states
in
determining
BART
or
what
constitutes
greater
reasonable
progress
than
BART.
249
Commenter
0259
said
the
proposed
guidelines
should
not
limit
state
and
tribal
ability
to
consider
all
of
the
reasonable
progress
factors
in
the
Act
in
demonstrating
greater
reasonable
progress.
The
guidelines
should
not
limit
a
demonstration
of
greater
reasonable
progress
to
only
visibility
impacts.
Other
reasonable
factors
include
greater
environmental
certainty
(
through
the
use
of
a
cap),
reduced
overall
compliance
cost
(
through
the
use
of
an
emissions
trading
mechanism),
and
consensus
effort
on
the
part
of
stakeholders.

Response:
The
EPA
agrees
that
visibility
improvement
is
obviously
a
critical
factor
in
deciding
whether
a
program
makes
greater
reasonable
progress
towards
the
national
visibility
goal
than
would
BART.
If
an
alternative
program
results
in
a
different
geographic
distribution
of
emission
reductions
than
would
BART,
it
may
not
be
possible
to
predict
greater
reasonable
progress
based
solely
on
the
fact
that
the
alternative
produces
greater
emissions
reductions.
Therefore,
in
such
circumstances,
the
"
better
than
BART"
test
requires
a
demonstration
based
on
visibility
modeling.
In
the
forthcoming
rule
proposal
on
BART
trading
programs,
we
will
seek
comment
on
how
other
factors
might
be
considered.

Comment:
Three
commenters
asked
under
what
conditions
will
a
dispersion
modeling
analysis
be
required
(
i.
e.,
how
different
must
the
spatial
distribution
in
emission
reductions
be)?

Response:
We
believe
it
is
not
necessary
for
EPA
to
define
in
the
rule
the
degree
of
difference
in
the
geographic
distribution
of
emissions
necessary
to
trigger
application
of
the
two
prong
test.
For
our
CAIR
analysis,
we
explained
in
the
SNPR
that
the
fact
that
CAIR
would
produce
greater
emissions
reductions
than
BART
in
most
States,
but
less
reductions
than
BART
in
a
few
States,
was
sufficient
reason
to
employ
the
two
pronged
visibility
test.
69
FR
32704.
For
other
programs
developed
by
States,
a
State
would
have
the
ability
to
make
a
reasonable
decision
as
to
whether
there
was
a
sufficient
basis
to
make
the
demonstration
that
an
alternative
program
would
be
better
than
BART
based
on
modeling
of
the
emissions
distributions
alone,
or
whether
the
State
should
proceed
with
the
two­
pronged
visibility
test.
The
State's
discretion
is
subject
as
always
to
the
condition
that
it
must
be
reasonably
exercised,
and
must
be
supported
by
adequate
documentation
of
the
analyses.

Comment:
Four
commenters
said
the
guidance
should
only
require
that
the
trading
program
achieve
"
equivalent"
progress,
rather
than
greater
progress,
compared
to
BART.
(
A200028­
272,
A200028­
275,
A200028­
333,
A200028­
7001).
They
say
requiring
greater
progress
would
be
by
definition
overstepping
the
Congressional
mandate
and
that
there
is
no
justification
for
such
a
position.
Other
commenters
said
the
requirement
to
achieve
greater
emissions
reductions
provides
a
strong
disincentive
to
the
implementation
of
a
trading
program,
and,
therefore,
is
not
appropriate.

Response:
250
The
EPA's
interpretation
of
CAA
169A(
b)(
2)
as
allowing
for
an
alternative
to
BART
where
the
alternative
makes
greater
reasonable
progress
was
upheld
by
the
Ninth
Circuit
in
Central
Arizona
Water
Conservation
District
v.
EPA,
990
F.
2d
1531
(
9th
Cir.
,
Mar.
25
1993).
In
that
case
the
court
held
that:
"
Congress's
use
of
the
term
`
including'
in
§
7491(
b)(
2)
prior
to
its
listing
BART
as
a
method
of
attaining
`
reasonable
progress'
supports
EPA's
position
that
it
has
the
discretion
to
adopt
implementation
plan
provisions
other
than
those
provided
by
BART
analyses
in
situations
where
the
agency
reasonably
concludes
that
more
`
reasonable
progress'
will
thereby
be
attained."
Id.
at
1543
(
emphasis
supplied).

This
interpretation
was
more
recently
upheld
by
the
D.
C.
Circuit
in
CEED
v.
EPA,
where
the
court
stated
that
the
petitioner
had
"
shown
neither
that
Congress's
language
precluded
non­
BART
alternatives
where
BART
wasn't
`
necessary
to
make
reasonable
progress,'
nor
that
EPA's
reading
is
otherwise
unreasonable.
CEED
v.
EPA,
Slip.
Op.
at
13
(
D.
C.
Cir.,
Feb.
18
2005).
As
in
Central
Arizona
Water,
in
CEED
our
determination
that
BART
was
not
necessary
to
make
reasonable
progress
was
premised
on
the
determination
that
the
alternative
program
made
greater
reasonable
progress
than
BART.
In
this
case
the
program
at
issue
 
a
backstop
emission
trading
program
for
certain
western
States
and
Tribes
in
the
Western
Regional
Air
Partnership
 
was
invalidated
on
other
grounds.
(
Specifically,
because
the
program
was
constrained
by
an
invalid
methodology
for
determining
the
emissions
reductions
achievable
from
BART.
Id.
at
14).
Nothing
in
the
court's
decision
in
CEED
suggests
that
a
"
greater
reasonable
progress"
test
as
a
criteria
for
an
alternative
to
BART
program
would
be
"
overstepping
the
Congressional
mandate."
In
fact,
the
court
noted
that
"
Congress's
addition
of
§
169B
.
.
.
clarified
that
the
focus
of
the
Clean
Air
Act
was
to
achieve
`
actual
progress
and
improvement
in
visibility'
.
.
.
not
to
anoint
BART
the
mandatory
vehicle
of
choice."
Id
at
13,
citations
omitted.
Comment:
Two
commenters
discussed
visibility
impairment
under
existing
conditions.
Commenter
0255
described
differences
between
the
Acid
Rain
program
and
the
BART
guidance.
The
BART
guidance
focuses
on
emissions
from
individual
combustion
sources
that
are
significantly
contributing
to
local
visibility
problems
in
national
parks
and
wilderness
areas.
Because
EPA
has
identified
50
km
as
a
sensitive
corridor
around
these
Class
I
areas,
sources
within
this
50
km
area
should
not
have
the
option
to
opt
out
if
they
are
found
to
be
BART
eligible.

Response:
Any
localized
impacts
from
BART­
eligible
sources
such
as
the
commenter
raises
can
be
addressed
by
BART
for
reasonably
attributable
visibility
impairment
if
necessary.

10.2.2
Averages
as
a
Measure
of
Improvement
Comment:
One
commenter
(
A200028­
258)
stated
that
to
ensure
that
trading
budgets
result
in
greater
reasonable
progress,
dispersion
modeling
must
be
required.
Further,
it
is
not
acceptable
to
allow
averaging
of
the
difference
between
improvements
due
to
BART
and
due
to
the
emissions
trading
program
among
Class
I
areas
to
show
an
overall
improvement.
251
Five
commenters
opposed
using
averages
as
a
measure
of
improvement
(
0179,
0221,
0227,
0267,
0305).
Commenter
0179
said
it
is
inappropriate
to
justify
the
use
of
any
trading
program,
or
other
BART
provision,
based
on
"
averaging"
of
improvements
in
visibility
over
Class
1
areas.
To
satisfy
the
legal
requirements
of
the
BART,
trading
programs
should
be
allowed
only
if
it
can
be
shown
that
they
will
reduce
pollution
and
improve
visibility
more
than
a
source­
bysource
clean
up
would
and
visibility
is
shown
to
improve
at
each
and
every
Class
1
area.
Commenters
(
0221,
0227,
0292)
said
averaging
is
contrary
to
the
Act
because
section
169A
and
the
RHR
are
designed
to
reduce
and
eventually
eliminate
visibility
impairment
in
each
Class
I
area.
Commenter
0221
added
that
visibility
improvement
is
not
a
commodity
that
can
be
"
traded"
among
states
or
Class
I
areas
because
they
each
have
a
different
required
rate
of
visibility
progress
and
different
emission
reduction
requirements
to
meet
a
specific
visibility
progress
rate.
Commenter
0267
was
concerned
that
trading
would
allow
sources
near
Class
I
areas
to
purchase
allowances
rather
than
install
emission
controls,
and
that
visibility
improvement
at
all
Class
I
areas
would
not
be
achieved.

Commenter
0267
said
the
CAIR
should
satisfy
the
requirements
laid
out
at
40
CFR
51.308(
e)(
2)
to
demonstrate
it
is
better
than
BART.
Showing
that
visibility
is
at
worst,
unchanged
under
CAIR
does
not
represent
reasonable
progress.
The
commenter
asked
how
EPA
can
use
an
average
visibility
improvement
calculation
based
on
44
sites
to
conclude
that
CAIR
is
"
better
than
BART",
because
the
RHR
required
improvement
at
all
Class
I
areas.

Response:
As
we
explained
in
the
CAIR
NFR
preamble,
we
disagree
with
comments
that
CAA
§
169A(
b)(
2)'
s
requirement
of
BART
for
sources
reasonably
anticipated
to
contribute
to
impairment
at
any
Class
I
area
means
that
an
alternative
to
the
BART
program
must
be
shown
to
create
improvement
at
each
and
every
Class
I
Area.
Even
if
a
BART
alternative
is
deemed
to
satisfy
BART
for
regional
haze
purposes,
based
on
average
overall
improvement
as
opposed
to
improvement
at
each
and
every
Class
I
Area,
CAA
§
169A(
b)(
2)'
s
trigger
for
BART
based
on
impairment
at
any
Class
I
area
remains
in
effect,
because
a
source
may
become
subject
to
BART
based
on
"
reasonably
attributable
visibility
impairment"
at
any
area.
See
40
CFR
§
51.302.

In
addition,
within
a
regional
haze
context,
not
every
measure
taken
is
required
to
achieve
a
visibility
improvement
at
every
class
I
area.
BART
is
one
component
of
long
term
strategies
to
make
reasonable
progress,
but
it
is
not
the
only
component.
The
requirement
that
the
alternative
achieves
greater
progress
based
on
the
average
improvement
at
all
Class
I
areas
assures
that,
by
definition,
the
alternative
will
achieve
greater
progress
overall.
Though
there
may
be
cases
where
BART
could
produce
greater
improvement
at
one
or
more
class
I
areas,
the
no­
degradation
prong
assures
that
the
alternative
will
not
result
in
worsened
conditions
anywhere
than
would
otherwise
exist,
and
the
possibility
of
BART
for
reasonably
attributable
visibility
protects
against
any
potential
"
hot
spots."
Taken
together,
the
EPA
believes
these
factors
make
a
compelling
case
that
the
proposed
test
properly
defines
"
greater
reasonable
progress."
The
EPA
anticipates
that
regional
haze
implementation
plans
will
also
contain
measures
addressing
other
sources
as
necessary
to
make
progress
at
every
mandatory
federal
Class
I
area.
252
With
respect
to
the
number
of
Class
I
areas
used
in
the
demonstration,
in
updated
analysis
in
the
CAIR
NFR,
we
analyzed
the
results
for
81
IMPROVE
sites
representing
116
Class
I
areas.

10.2.3
Visibility
Impairment
Hot
Spots
&
Reasonably
Attributable
Visibility
Impairment
Comment:
Three
commenters
provided
general
comments
on
hot
spots
(
0201,
0304,
0323).
Commenters
(
0201,
0304)
said
the
need
to
include
geographic
enhancements
within
any
national
trading
program
should
be
strongly
emphasized
to
ensure
that
no
visibility
"
hot
spots"
are
created
as
a
product
of
trading.
Commenter
0323
said
the
guidance
should
cite
the
WRAP
approach
as
an
example
of
implementing
the
"
geographic
enhancement"
provisions
of
the
regional
haze
rule.

Response:
Regional
haze
rule
§
308(
e)(
4)
as
finalized
does
provide
States
with
the
option
of
utilizing
geographic
enhancements.

Comment:
Two
commenters
discussed
how
FLMs
should
certify
impairment
(
0201,
0246).
Commenter
0201
said
for
sources
subject
to
the
IAQR/
CAIR,
the
RPOs
should
work
with
the
FLMs
to
establish
criteria
for
future
certifications
which
recognize
that
the
IAQR/
CAIR's
overall
regional
emissions
reductions
will
achieve
more
visibility
improvement
across
a
broad
geographic
area
than
implementation
of
BART
alone.
Such
criteria
would
address
pollutant
trends
measured
at
Class
I
areas,
limits
on
distance
from
a
given
class
I
area
for
certification
of
BART­
eligible
sources,
and
recognition
of
existing,
or
commitments
to
future,
emissions
controls.
Should
the
IAQR/
CAIR
be
finalized,
the
RPOs
should
encourage
the
FLMs
and
affected
sources
to
communicate
regularly
during
implementation
of
the
rule
regarding
Class
I
area
monitoring
data
so
that
planned
emissions
reductions
would
ensure
reasonable
geographic
distribution
of
the
benefits
and
thereby
avoid
the
likelihood
of
any
future
source­
specific
certifications
under
40
CFR
51.302.
The
FLMs
and
states
should
retain
the
authority
to
"
certify"
impairment
at
specific
class
I
areas
which
may
be
"
reasonably
attributable"
to
sources
subject
to
the
IAQR/
CAIR.

Commenter
0246
said
there
is
a
need
for
guidance
about
the
specific
kinds
of
circumstances
that
would
cause
an
FLM
to
identify
a
"
hot
spot"
leading
to
certification
of
impairment.
The
current
criteria
(
Class
I
areas
in
which
sulfates
are
not
decreasing
may
result
in
scrutiny
of
sources
within
200
kilometers
controlled
at
less
than
85%
capture
rate)
may
not
provide
adequate
guidance.
Another
alternative
that
is
the
continuing
discussions
between
stakeholders
and
the
FLMs
under
the
WRAP
would
determine
how
new
data
and
information
can
be
used
to
inform
regulatory
intervention
in
the
section
309
backstop
market
program.

Response:
Today's
action
does
not
affect
the
ability
of
FLMs
or
States
to
certify
reasonably
attributable
visibility
impairment.
We
believe
regulatory
certainty
may
be
improved
by
agreements
with
FLMs
regarding
the
circumstances
under
which
RAVI
may
be
certified,
coupled
if
necessary
with
"
geographic
enhancements"
to
the
trading
program
to
accommodate
sources
subject
to
RAVI
253
BART.
We
therefore
encourage
cooperation
and
communication
between
States,
Tribes,
and
RPOs
on
these
issues.
However,
providing
additional
guidance
on
these
issues
is
beyond
the
scope
of
this
rule.

Comment:
Two
commenters
discussed
the
role
of
BART
for
reasonably
attributable
visibility
impairment
(
0217,
0246).
Commenter
0217
had
concerns
about
the
role
of
BART
for
reasonably
attributable
visibility
impairment
(
RAVI)
until
the
final
milestone
is
achieved
under
a
trading
program
in
2018.
Even
if
the
role
of
this
regulatory
tool
is
likely
to
be
limited,
the
regulated
community
thinks
it
creates
significant
uncertainty.
The
commenter
urged
the
states,
FLMs
and
stakeholders
to
further
define
the
limited
circumstances
(
e.
g.,
to
address
"
hot
spot"
concerns)
when
this
regulatory
mechanism
could
come
into
play,
and
to
be
mindful
of
the
central
role
of
the
states
in
making
attribution
determinations
and
in
fashioning
BART
remedies.
Commenter
0246
urged
the
states,
tribes,
FLMs
and
other
stakeholders
to
continue
to
explore
ways
in
which
uncertainty
from
the
potential
use
of
this
program
can
be
minimized
and
creative
remedies
might
be
fashioned
under
the
"
geographic
enhancement"
program.

Response:

See
previous
response.
EPA
encourages
cooperation
and
communication
between
States,
Tribes,
and
RPOs
on
defining
RAVI;
However,
providing
additional
guidance
on
these
issues
is
beyond
the
scope
of
this
rule.

Comment:
Five
commenters
disagreed
with
the
creation
of
"
reasonably
attributable
BART"
as
exceeding
EPA's
authority
(
28­
357,
28­
305,
28­
36,
28­
364,
28­
395).
There
is
only
one
CAA
BART
process
and
EPA's
efforts
to
create
two
are
arbitrary
and
unfair
(
28­
395).
Commenter
28­
357
added
that,
in
event
EPA
persists
with
the
concept
of
two
BARTs,
the
Guidelines
should
permit
states
to
integrate
them
in
a
manner
that
clearly
distinguishes
the
applicability
of
both
programs
and
eliminates
the
potential
"
double
jeopardy"
and
resulting
uncertainty
facing
regulated
sources.
In
effect,
where
a
determination
is
made
that
a
cap
and
trade
program
like
309
would
achieve
that
greater
reasonable
progress,
the
BART
issue
for
all
eligible
participating
sources
of
SO
x
would
be
resolved.
A
second
alternative
is
to
clearly
and
specifically
reconcile
and
limit
the
role
of
BART
for
RAVI
until
all
BART
tolls
for
affected
sources
in
the
applicable
region.

Commenter
28­
395
said
the
original
BART
guidelines
and
EPA's
proposed
new
guidelines
should
be
integrated.
From
a
statutory
perspective,
the
visibility
protection
program
of
the
Clean
Air
Act
does
not
refer
to
either
type
of
visibility
impairment.
The
statute
merely
refers
to
"
visibility
impairment".
Similarly,
the
Act
does
not
refer
to
one
BART
for
reasonably
attributable
visibility
impairment
and
another
BART
for
contributors
to
regional
haze.
The
visibility
protection
program
merely
refers
to
"
BART".
While
the
focus
of
the
original
1980
visibility
protection
rules
on
plume
blight
was
a
useful
device
at
the
time,
the
need
for
any
distinction
between
types
of
visibility
impairments
has
been
eliminated
now
that
EPA
is
implementing
the
entire
visibility
protection
program.
EPA
should
revise
the
BART
guidelines
for
254
large
fossil­
fuel
fired
stream
electric
generating
units
so
that
there
is
one
BART
for
emissions
that
impair
visibility
in
MCIF
Areas.

Commenter
0252
said
greater
"
reasonable
progress"
ensures
that
there
will
be
visibility
improvements
in
all
Class
I
areas;
therefore
there
is
no
reason
to
retain
reasonably
attributable
BART.
EPA
should
provide
greater
regulatory
certainty
by
clarifying
that
once
a
source
has
complied
with
the
regional
haze/
BART
rule,
either
through
source­
specific
BART
or
through
a
cap­
and­
trade
program
such
as
CAIR,
it
has
complied
with
reasonably
attributable
BART.

Commenter
0323
agreed
with
EPA
that
any
"
market"
approach
to
address
the
regional
haze
BART
requirement
does
not
necessarily
address
all
"
reasonably
attributable"
impairment.
Commenter
0304
said,
until
they
are
in
compliance
with
the
final
emissions
budgets
(
through
reductions
or
the
purchase
of
allowances,)
sources
subject
to
the
IAQR
should
be
subject
to
visibility
protection
regulations
regarding
"
reasonably
attributable"
visibility
impairment.

Response:
The
EPA
established
in
the
1999
regional
haze
rule
that
the
requirement
to
install
and
operate
BART
is
applicable
in
the
regional
haze
context
under
CAA
169B
as
well
as
under
the
1980
"
reasonably
attributable"
regulations
at
40
CFR
51.302.
Comments
or
challenges
to
this
interpretation
were
ripe
at
that
time.
As
noted
in
the
CAIR
NFR
preamble,
the
EPA
continues
to
maintain
that
BART
based
on
reasonably
attributable
visibility
impairment
(
RAVI)
must
be
retained
in
order
to
protect
against
the
possibility
of
localized
degradation,
even
though
current
modeling
does
not
predict
such
"
hot
spots."
Integration
of
"
RAVI"
BART
into
a
trading
program
may
be
accomplished
through
geographic
enhancements.

Comment:
Four
commenters
discussed
the
process
for
determining
whether
reasonably
attributable
visibility
impairment
(
0224,
,
0246,
0252,
0323).
Commenter
0246
said
it
is
the
primary
responsibility
and
role
of
the
states
and
tribes
to
determine
reasonable
attribution
and
BART
under
the
RAVI
program.
Commenter
0224
said
guidance
(
e.
g.
how
to
weight
BART
factors,
such
as
cost
effectiveness),
or
presumptive
numerical
BART
emission
limits
that
apply
for
regional
haze
should
apply
in
the
same
manner
to
"
reasonably
attributable"
visibility
impairment.
Applying
technology
(
or
performance)
presumptions
to
regional
haze
without
applying
the
same
presumptions
to
"
reasonably
attributable"
visibility
impairment
would
be
applying
statutory
factors
inconsistently
and
therefore
would
be
contrary
to
the
statute.

Response:
The
1980
regulations
lay
out
the
process
for
addressing
reasonably
attributable
BART.
See
40
CFR
51.302(
c).
Other
than
making
a
minor
change
to
the
BART
determination
process
for
sources
found
to
have
a
reasonably
attributable
contribution
to
visibility
impairment,
we
are
not
revising
these
regulations.

10.3
Determination
that
CAIR
is
"
Better
than
BART"
255
Comment
Commenter
0267
was
concerned
with
emissions
"
leakage"
issues
for
CAIR
border
areas
and
how
that
will
affect
nearby
Class
I
areas.
A
demonstration
of
the
distribution
of
emissions
needs
to
be
made
to
ensure
it
is
not
different
than
the
source­
by­
source
BART
approach
to
ensure
that
the
trading
program
achieves
better
visibility
improvement.
If
the
distribution
is
different
under
the
two
approaches,
a
modeling
study
for
each
Class
I
area
is
necessary
for
both
the
worst
and
best
20%
days.
Also,
reductions
seen
in
emissions
under
Title
IV
have
been
in
proportion
to
the
size
of
the
emission
sources,
so
Class
I
areas
that
are
remote
from
individual
"
large"
sources,
and/
or
at
the
edges
of
the
CAIR
trading
region,
may
not
make
enough
reasonable
progress
without
enhanced
control
of
nearby
sources.
Since
the
reasonable
progress
goals
have
not
been
determined
for
the
Class
I
areas
it
is
not
known
how
the
projected
improvement
in
visibility
under
CAIR
compares
to
what
will
be
necessary.
States
may
need
to
get
additional
reductions
from
some
EGUs
above
and
beyond
CAIR
on
a
case­
by­
case
basis
to
achieve
reasonable
progress
at
some
Class
I
areas.

Response:
The
EPA
has
determined
that
the
CAIR
will
achieve
greater
reasonable
progress
than
BART
for
EGUs
and
therefore
may
be
treated
by
States
as
a
substitute
for
BART
for
affected
EGUs.
This
does
constitute
a
finding
that
CAIR
fully
satisfies
the
reasonable
progress
requirement
for
each
Class
I
area.
The
EPA
agrees
that
States
may
need
to
get
more
emission
reductions
than
provided
by
CAIR
in
order
to
meet
reasonable
progress
goals
at
some
Class
I
areas.
States
have
the
discretion
to
determine
from
what
sources
to
obtain
those
reductions,
and
are
not
required
to
treat
CAIR
as
having
satisfied
BART
for
EGUs.

Comment:
Commenter
0221
said
EPA's
proposal
to
categorically
exclude
sources
from
BART,
without
opportunity
for
the
states
to
evaluate
the
adequacy
of
the
reductions
in
light
of
the
BART
requirement,
to
consider
their
overall
visibility
protection
needs,
or
to
effectuate
the
core
requirements
and
purposes
of
the
visibility
protection
program,
tramples
over
section
169A,
the
relative
roles
and
responsibilities
of
federal
and
state
governments
embodied
in
section
110
of
the
Act,
and
the
court's
affirmance
of
state
prerogatives
in
American
Corn
Growers.

Response:
EPA's
determination
that
CAIR
may
substitute
for
BART
for
affected
EGUs
does
not
infringe
on
State
prerogatives
because
States
are
not
compelled
to
accept
this
determination.

10.3.1.
Applicable
Sources
Comment:
Seventeen
commenters
discussed
how
sources
subject
to
CAIR
and
BART
should
be
treated
(
0169,
0189,
0201,
0207,
0208,
0223,
0230,
0252,
0253,
0271,
0272,
0277,
0279,
0299,
0304,
0308,
0310,
0317).
Some
commenters
opposed
exempting
sources
subject
to
the
CAIR
from
the
BART
rule
(
0169,
0310,
0317).
Commenter
0275
said
this
exemption
would
be
unfair
to
sources
that
have
already
taken
steps
to
reduce
regional
haze,
such
as
through
the
production
of
clean
fuels.
Commenter
0310
said
an
overlapping
regulatory
scheme
would
impose
256
unnecessary
and
contradictory
requirements
on
this
industry
that
is
already
heavily
regulated,
that
has
already
achieved
significant
emissions
reductions,
and
that
provides
a
necessary
and
significant
contribution
to
the
Texas
and
national
economies.
Accordingly,
the
commenter
concurs
with
EPA's
recommendation
that
BART
requirements
must
not
be
made
applicable
to
entities
subject
to
the
CAIR.

Response:

We
agree
that
the
overall
advantages
provided
by
integrating
BART
requirements
with
CAIR
and
thereby
allowing
BART
for
EGUs
to
be
satisfied
in
the
most
cost
effective
way
possible
outweighs
any
inequity
(
if
it
exists)
towards
sources
that
have
already
reduced
emissions.
Such
sources
will
continue
to
enjoy
the
benefits
of
lower
emissions,
such
as
lower
costs
to
comply
with
emission
allowance
requirements.

Comment:
Commenter
0189
said
visibility
improvement
requirements
following
2018
should
be
determined
on
the
basis
of
visibility
impact
and
cost­
effectiveness
as
the
primary
factors,
with
other
secondary
factors
such
as
state
of
development
of
control
technology
and
administrative
costs
also
considered.
Sources
should
not
be
exempted
from
control
or
review
simply
because
they
were
controlled
under
CAIR
or
BART.
However,
commenter
0253
said
affected
states
should
be
required
to
look
initially
to
non­
EGU
sources
for
any
additional
emissions
reductions
that
might
be
required
by
statute
or
regulation.

Response:
The
regional
haze
rule
requires
a
comprehensive
SIP
revision
in
2018
to
establish
a
new
long
term
strategy.
Although
all
BART
requirements
will
of
necessity
have
been
satisfied
at
that
time,
we
cannot
prejudge
what
requirements
may
be
deemed
appropriate
for
the
next
10
year
planning
period.

Comments:
Several
commenters
addressed
the
role
that
CAIR
opt­
in
requirements
should
play
relative
to
BART.
Commenters
(
0189,
0201,
0207,
0223,
0308)
said
EPA
must
provide
that
CAIR
compliance
equals
BART
compliance
for
non­
EGU
sources
that
"
opt
in"
to
CAIR
coverage.
This
rule
of
per
se
compliance
should
apply
just
as
much
to
opt­
in
sources
that
comply
through
emissions
allowance
trading
as
to
opt­
in
sources
that
install
controls.
Other
commenters
(
0207,
0208)
said
that
EPA
should
take
necessary
steps
to
assure
power
plant
owners
that
the
requirements
of
the
CAIR
meet
the
Acts's
BART
obligations,
as
well
as
the
reasonable
further
progress
criteria
of
the
RHR.
Commenter
0230
said
appropriate
"
opt­
in"
provisions
are
needed.
Alternatively,
commenter
0272
said
an
effective
control
of
SO
2
and
NO
x
emissions
from
non­
EGU
sources
could
be
achieved
by
introducing
an
approach
similar
to
BART
in
CAIR
for
non­
EGU
sources.
"
BART
for
CAIR"
would
only
be
applied
on
major
non­
EGU
sources
and
the
limiting
criteria
for
the
stringency
of
control
would
be
cost
effectiveness,
existing
controls
and
probably
other
criteria
determined
by
EPA
in
a
corresponding
guideline.

Response:
257
Our
determination
that
CAIR
is
"
better
than
BART"
was
based
on
emission
projections
and
air
quality
modeling
for
EGUs
only.
For
an
opt­
in
source
to
use
CAIR
to
satisfy
BART,
a
supplementary
demonstration
would
be
required.

Comment:
Commenter
0208
provided
additional
comments
on
the
need
to
properly
design
the
"
western
CAIR"
and
why
this
is
the
best
way
to
implement
the
BART
trading
program
in
the
west.
However,
commenters
(
0223,
0230,
0304)
urged
EPA
to
encourage
states
to
participate
in
national
trading
program,
rather
than
regional
trading
programs.
Commenter
0252
urged
EPA
to
make
CAIR
a
nationwide
program.
A
nationwide
program
would
reduce
unnecessary
burden
on
states
to
develop
an
interstate
trading
program
in
coordination
with
EPA.
Furthermore,
it
would
provide
a
more
robust
market
for
all
states,
especially
those
in
the
west.

Response:
As
explained
in
the
CAIR
RTC,
The
EPA
agrees
there
could
be
some
advantages
to
an
expanded
geographic
scope
for
CAIR,
but
such
an
expansion
would
also
raise
numerous
technical
and
legal
challenges
which
could
delay
promulgation
and
implementation
of
the
rule.
Therefore
the
EPA
determined
that
it
was
appropriate
to
focus
in
the
CAIR
on
obtaining
the
health
and
welfare
benefits
related
to
implementation
of
PM2.5
and
Ozone
standards.
This
does
not
foreclose
the
possibility
of
expanding
a
CAIR
like
approach
to
the
West
at
a
later
date.

10.3.2
Specific
Comments
Regarding
Whether
CAIR
Satisfies
BART
Comment:
Thirty­
eight
commenters
said
that
the
CAIR
can
or
should
satisfy
BART
requirements
(
0181,
0185,
0187,
0188,
0189,
0195,
0198,
0201,
0206,
0207,
0210,
0213,
0214,
0215,
0216,
0223,
0224,
0228,
0230,
0236,
0238,
0242,
0245,
0250,
0252,
0253,
0254,
0256,
0265,
0276,
0294,
0300,
0303,
0304,
0308,
0310,
0319,
0323)
for
several
reasons,
including
the
following:
°
CAIR
reductions
should
satisfy
a
state's
obligation
to
satisfy
BART
under
the
RHR,
including
the
reasonable
progress
requirements
(
0181,
0185,
0189,
0224,
0245,
0250,
0252,
0253,
0254,
0256,
0308,
0300,
0310).
°
The
emission
reductions
required
under
the
CAIR
will
result
in
cost­
effective
emission
reductions
(
0181,
0210,
0213,
0228,
0308,
0319,
0294).
°
The
CAIR
emissions
reductions
are
scheduled
to
take
effect
in
the
2010­
2015
time
frame,
which
coincides
with
the
schedule
for
BART
implementation
(
0181).
°
Allowing
the
CAIR
to
replace
BART
avoids
subjecting
sources
to
overlapping
regulatory
requirements
and
provides
regulatory
certainty
(
0187,
0216,
0236,
0250,
0252,
0253,
0300,
0310,
0319).
°
Since
the
purpose
of
BART
is
visibility
improvement,
there
is
no
reason
to
impose
BART
in
states
where
visibility
improvement
goals
can
be
met
without
it
(
0189,
0253).
°
Greater
participation
in
the
trading
program
will
result
in
greater
opportunities
for
trading
and
cost
savings
(
0250,
0319).
258
Commenter
0201
said
that
until
sources
are
in
compliance
with
the
final
emissions
budgets
(
through
reductions
or
the
purchase
of
allowances)
the
sources
subject
to
the
IAQR/
CAIR
should
be
subject
to
visibility
protection
regulations
regarding
"
reasonably
attributable"
visibility
impairment.

Response:
As
explained
in
the
preamble,
EPA
has
determined
that
CAIR
makes
greater
reasonable
progress
than
BART
and
may
substitute
for
BART
for
EGUs
in
CAIR
affected
States.
Except
for
the
intimation
that
it
can
be
determined
at
this
time
whether
CAIR
satisfies
all
reasonable
progress
goals,
the
EPA
agrees
with
the
commenters.

Comment:
Fourteen
commenters
said
CAIR
cannot
or
should
not
satisfy
BART
requirements
for
several
reasons,
including
the
following
(
0147,
0150,
0151,
0165,
0203,
0204,
0221,
0227,
0235,
0255,
0267,
0280,
0305,
0317):
°
IAQR
does
not
protect
the
all
156
Class
I
areas
defined
in
the
1977
CAA
amendments
and
will
not
produce
mandated
reductions
in
the
first
10
years
(
0147,
0150,
0151,
0203,
0204,
0221,
0227,
0317).
°
CAIR
needs
significant
strengthening
to
adequately
protect
polluted
communities
(
0203,
0204)
°
The
goals
of
the
programs
are
different,
resulting
in
a
different
spectrum
of
controls
under
CAIR
than
would
occur
under
BART
(
0267,
0280)
°
Adoption
of
CAIR
appears
to
preclude
application
of
source­
by­
source
BART
to
EGUs,
and
all
Class
I
areas
may
not
experience
visibility
benefits
from
a
trading
program
equal
to,
or
greater
than,
the
benefits
realized
from
source­
by­
source
BART
(
0267).
°
The
proposal
is
inconsistent
with
the
Act
(
0203,
0204,
0221,
0227).
°
Due
to
the
availability
of
control
technologies
to
cost
effectively
reduce
SO
2
and
NO
x
emissions
from
coal­
fired
boilers,
BART­
eligible
sources
should
not
be
allowed
to
opt
out
of
the
program
without
meeting
the
mandated
emissions
limits
(
0255).

Commenter
0221
added
that
several
arguments
describing
why
the
"
better
than
BART"
provision
violates
the
Act.
For
example,
EPA
has
not
found
superior
visibility
improvement
resulting
from
application
of
CAIR
compared
to
source­
by­
source
BART
controls
in
each
and
every
Class
I
area
that
may
be
impacted
by
BART­
eligible
sources
in
the
CAIR
region.
EPA's
attempt
to
use
average
emissions
over
selected
areas
is
not
sufficient.
Commenters
(
0221,
0235)
added
that
instead
of
implementing
the
Act,
EPA
is
attempting
to
amend
it.
Commenter
0221
also
described
serious
policy
flaws
in
the
proposal
such
ignoring
the
need
to
virtually
eliminate
SO
2
emissions
in
order
to
reach
the
national
visibility
goal.

Response:
The
policy
of
allowing
another
program
to
substitute
for
BART
is
not
an
illegal
substitution
of
the
Agency's
judgment
for
that
of
Congress,
but
rather
is
based
on
a
reasonable
259
interpretation
of
the
ambiguous
language
of
CAA
§
169A(
b)(
2).
This
section
provides
that
EPA
must
require
SIPs
to
contain
certain
measures
"
as
may
be
necessary
to
make
reasonable
progress
toward
meeting
the
national
goal
.
.
.
including
 
"
among
other
things,
a
requirement
that
BARTeligible
sources
install
BART.
See
42
USC
7491(
b)(
2).
The
EPA
has
interpreted
this
as
meaning
that
BART
is
required
only
"
as
may
be
necessary
to
make
reasonable
progress."
A
logical
corollary
is
that
if
an
alternative
strategy
covering
a
given
source
or
source
sector
makes
greater
progress
than
BART,
then
BART
is
not
necessary
for
that
source
or
sources.

In
Central
Arizona
Water
Conservation
District
v.
EPA,
990
F.
2d
1531,
(
1993),
the
Ninth
Circuit
applied
the
two­
step
Chevron
test
(
Chevron,
U.
S.
A.,
Inc.
v.
NRDC,
467
U.
S.
837,
842­
45
(
1984)
and
found
that
the
statutory
language
was
ambiguous
and
that
EPA's
interpretation
was
reasonable.
990
F.
2d
at
1543.
The
court
therefore
agreed
that
the
statute
provided
EPA
with
"
discretion
to
adopt
implementation
plan
provisions
other
than
those
provided
by
BART
analyses
in
situations
where
the
agency
reasonably
concludes
that
more
`
reasonable
progress'
will
thereby
be
attained."
Id.
In
that
case,
the
BART­
alternative
in
question
applied
to
a
single
facility.

In,
Center
for
Energy
and
Economic
Development
v.
EPA,
No.
03­
1222,
(
D.
C.
Cir.
Feb.
18,
2005)
("
CEED"),
the
D.
C.
Circuit
similarly
applied
the
two­
step
Chevron
test
and
also
found
that
the
statutory
language
was
ambiguous
and
that
EPA's
interpretation
was
reasonable.
CEED,
slip
op.
at
12­
13.
In
CEED
,
the
BART­
alternative
at
issue
(
the
WRAP
Annex)
applied
not
just
to
one
facility
but
to
all
BART­
eligible
sources
(
as
well
as
other
sources)
in
participating
States.
Therefore,
the
CEED
decision
clarified
that
EPA's
interpretation
of
§
169A(
b)(
2)
allowing
a
BART­
alternative
is
reasonable
not
only
in
a
source­
specific
context
but
also
in
the
context
of
larger
programmatic
BART­
alternatives.
(
The
court
vacated
the
Annex
on
other
grounds,
holding
that
EPA
did
not
rationally
exercise
its
discretion
with
respect
to
the
particular
methodology
required
of
States
to
demonstrate
that
the
Annex
would
result
in
greater
progress
than
BART.
CEED,
slip
op.
at
14).

The
CAIR
cap
and
trade
program
for
EGUs
represents
an
intermediate
case
between
the
single­
source
BART
alternative
in
Central
Arizona
Water
and
the
all­
BART­
sources
alternative
in
CEED
.
A
BART­
alternative
program
can
only
achieve
emission
reductions
at
sources
included
in
the
program
and
so
cannot
achieve
greater
reasonable
progress
than
BART
with
respect
to
BART­
eligible
sources
which
are
not
covered
by
the
program.
Therefore,
the
relevant
question
for
CAIR
is
whether
it
achieves
greater
reasonable
progress
from
EGUs
than
would
BART
for
EGUs.
If
it
can
be
shown
that
CAIR
does
so,
then
under
EPA's
interpretation
of
§
169A(
b)(
2),
upheld
in
both
these
cases,
BART
for
EGUs
is
not
"
necessary
to
make
reasonable
progress."
Because
this
finding
would
be
applicable
only
to
EGUs,
it
would
not
frustrate
Congress's
intent
in
identifying
26
source
categories
for
BART
applicability.
(
Note
that
as
explained
in
the
NFR,
a
State
which
chooses
to
meet
its
CAIR
emission
reduction
requirements
partially
or
wholly
from
non­
EGU
sources
could
consider
BART
satisfied
for
any
BART­
eligible
non­
EGUs
affected,
but
the
State
would
have
to
conduct
a
better­
than­
BART
analysis
covering
this
alternative
means
of
meeting
its
CAIR
requirements).
260
The
EPA
believes
the
analysis
presented
in
the
CAIR
preamble
and
TSD
convincingly
demonstrate
that
CAIR
achieves
greater
reasonable
progress,
with
respect
to
SO
2
and
NO
x
emissions
from
EGUs,
than
would
source­
specific
BART
for
EGUs
in
the
CAIR
region.

Moreover,
the
EPA
does
not
believe
that
it
is
a
prerequisite
of
BART­
alternative
programs
that
the
emission
reductions
be
developed
explicitly
for
visibility
purposes
in
the
first
place.
As
long
as
the
end
result
is
greater
visibility
improvement
than
BART,
it
is
of
no
consequence
that
the
emission
reductions
were
initially
developed
to
serve
other
air
quality
purposes.
As
the
D.
C.
Circuit
stated
in
CEED,
"
Congress's
addition
of
§
169B
.
.
.
clarified
that
the
focus
of
the
Clean
Air
Act
was
to
achieve
`
actual
progress
and
improvement
in
visibility,'
.
.
.
not
to
anoint
BART
the
mandatory
vehicle
of
choice."
CEED,
slip.
op.
at
13.
As
long
as
visibility
is
actually
improved
more
than
it
would
be
under
BART,
it
should
not
matter
whether
the
emission
reductions
were
developed
explicitly
for
visibility,
as
under
BART,
or
for
some
other
purpose,
as
under
CAIR.

Comment:
Nine
commenters
provided
additional
comments
regarding
whether
CAIR
satisfies
BART
(
0186,
0231,
0234,
0237,
0267,
0275,
0280,
0285,
0304).
Some
commenters
pointed
out
instances
where
complying
with
CAIR
may
not
satisfy
BART.
Commenters
(
0186,
0237,
0275)
said
there
can
be
scenarios
where
the
"
highly
cost­
effective"
metric
for
CAIR
controls
may
not
rise
to
the
level
of
BART,
which
must
also
consider
a
number
of
other
factors,
including
improvements
in
visibility.
Commenter
0231
said
if
trading
is
allowed,
how
can
a
state
be
sure
that
the
impacts
on
the
Class
I
area
by
any
source
are
addressed?
A
source
close
to
and
significantly
impacting
a
Class
I
area
could
elect
to
buy
allowances
from
another
source
hundreds
of
miles
away
that
has
no
impact
whatsoever.
The
Class
I
area
impacts
therefore
would
not
be
addressed.

Response:
While
the
highly
cost
effective
criterion
used
in
CAIR
might
not
result
in
BARTequivalent
controls
on
individual
units,
in
the
aggregate
the
CAIR
results
in
millions
of
tons
more
emissions
reductions
than
BART
and
has
been
shown
to
make
greater
progress
based
on
visibility
projections.

Comment:
Commenter
0267
said
CAIR,
as
proposed,
is
not
better
than
BART
but
it
could
be
made
better
than
BART
by
lowering
the
emission
caps
for
SO
2
and
NO
x.
Commenter
0280
said
unless
the
stringency
of
CAIR
is
increased,
BART
should
apply
to
all
affected
sources,
including
EGUs.
Visibility
improvement
should
be
a
certainty
and
not
subject
to
a
market
decision
on
where
emission
reduction
investments
are
placed.
However,
there
is
a
link
between
CAIR
and
visibility
improvement
and
emission
reductions
as
a
result
of
installing
BART
control
should
count
toward
meeting
the
emissions
budget
established
in
CAIR.
Commenter
0267was
concerned
that
unless
the
CAIR
emission
caps
are
lowered
to
levels
consistent
with
moderate
costs
per
ton,
that
they
will
be
better
than
BART.
261
Response:
Our
analysis
has
demonstrated
that
the
emission
reductions
required
by
CAIR
are
in
fact
"
better
than
BART"
according
to
the
criteria
we
previously
proposed
and
have
now
finalized
at
regional
haze
rule
308(
e)(
3).

Comment:
Commenter
0285
asked
EPA
to
clarify
whether
CAIR
would
satisfy
the
BART
requirements
for
SO
2
and
NO
x
only
or
for
all
four
pollutants?
Also,
at
plants
where
CAIR
applies,
would
a
BART
evaluation
be
required
of
activities
other
than
the
boilers
themselves?

Response:

Our
determination
that
CAIR
is
better
than
BART
applies
only
to
SO2
and
Nox.
If
an
EGU
covered
by
CAIR
for
SO2
and
NOx
does
not
qualify
as
a
BART­
eligible
source
based
on
potential
to
emit
250
tons
of
any
other
visibility
impairing
pollutant,
then
it
would
not
be
subject
to
BART
and
would
not
have
go
through
a
BART
determination.
For
units
in
all
other
BARTcategories
the
BART
guidelines
should
be
followed
to
determine
which
units
are
subject
to
BART.

Comment:
Four
commenters
provided
other
comments
on
the
CAIR
(
0214,
0228,
0242,
0267).
Commenter
0214
said
if
EPA
chooses
not
to
determine
that
CAIR
satisfies
BART,
the
agency
should
specify
in
the
final
guidelines
that,
in
their
individual
source
assessments,
Eastern
sources
should
be
able
to
model
for
visibility
impairment
assuming
that
the
CAIR
is
in
effect.

Response:
EPA
has
made
the
determination
that
CAIR
satisfies
BART
so
the
comment
is
not
relevant.

Comment:
Commenter
0228
asked
EPA
to
clarify
the
final
guidelines
by
specifying
that
surplus
reductions
resulting
from
measures
adopted
to
meet
requirements
of
the
Act
as
of
the
baseline
date
of
a
state's
regional
haze
SIP
means
the
date
of
the
emission
inventories
upon
which
the
regional
haze
SIPs
rely,
which
the
BART
Guidelines
specify
is
the
2002
base
year
emission
inventory
(
which
would
not
include
the
additional
emission
reductions
to
be
obtained
through
the
CAIR
rule)
.
Also,
EPA
should
clarify
that
states
which
fully
implement
the
CAIR
trading
program
and
for
which
EPA
has
approved
a
SIP
revision
implementing
the
CAIR
reductions,
be
relieved
from
the
technical
and
analytical
requirements
to
demonstrate
"
greater
reasonable
progress"
or
greater
visibility
improvement
than
would
be
achieved
through
the
installation
and
operation
of
sourcespecific
BART
as
it
applies
to
EGUs.
Commenter
0242
said
emission
reductions
that
will
be
finalized
in
CAIR
are
"
surplus
to
those
reductions
resulting
from
measures
adopted
to
meet
the
requirements
of
the
CAA
as
of
the
baseline
of
the
SIP"
because
the
baseline
of
the
regional
haze
SIP
is
the
year
of
the
base
emissions
inventory
(
2002).
Therefore,
since
CAIR
will
be
finalized
after
that
date,
CAIR
emission
reductions
were
not
required
prior
to
the
regional
haze
baseline
date.
262
Response:
EPA
has
stated
at
several
places
within
the
CAIR
rulemaking
that
2002
is
the
base
year
for
regional
haze
emissions
inventory
and
that
CAIR
reductions
are
therefore
surplus
to
base
year
requirements.
We
also
agree
that
States
that
opt
to
utilize
CAIR
as
a
BART
substitute
may
rely
on
our
technical
determination
rather
than
being
required
to
submit
their
own
demonstration.
The
State
option
to
utilize
CAIR
is
provided
by
regional
haze
rule
section
308(
e)(
4).
