Table
of
Contents
Overview
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
1
The
Charge
to
EPA
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
2
Background
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
2
Impact
on
Investment
in
New
and
Existing
Utility
and
Refinery
Generation
Capacity
and
Energy
Efficiency
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4
New
Sources
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5
Existing
Sources
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
8
Impact
on
Industries
Other
than
Electric
Utilities
and
Petroleum
Refineries
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
22
NSR
Applicability
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
23
Energy
Efficiency
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
25
Impact
on
Environmental
Protection
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
27
Conclusion
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
31
1
New
Source
Review:
Report
to
the
President
Overview
The
New
Source
Review
(
NSR)
program
is
one
of
many
programs
created
by
the
Clean
Air
Act
to
reduce
emissions
of
air
pollutants
 
particularly
"
criteria
pollutants"
that
are
emitted
from
a
wide
variety
of
sources
and
have
an
adverse
impact
on
human
health
and
the
environment.
Other
key
programs
include
the
Title
IV
Acid
Rain
Program,
"
MACT"
standards
and
other
air
toxics
standards,
New
Source
Performance
Standards,
the
22­
state
NO
x
"
SIP
Call,"
the
Regional
Haze
Program,
numerous
mobile
source
programs,
and
other
state
and
local
SIP­
based
emissions
standards.
Government
officials
from
both
major
political
parties
and
industry
groups
have
expressed
the
belief
that
the
NSR
program
is
unnecessarily
complicated
and
often
serves
as
an
unnecessary
obstacle
to
environmentally
beneficial
projects
in
the
energy
sector,
such
as
those
that
improve
energy
reliability
and
efficiency
and
promote
the
use
of
renewable
resources.

The
President's
National
Energy
Policy
Development
Group
asked
EPA
to
investigate
whether
the
NSR
program
does,
in
fact,
have
such
impacts.
The
Agency's
review
of
the
NSR
program
was
broad­
based.
EPA
held
four
public
hearings,
had
individual
meetings
with
over
100
groups
representing
the
public,
industry
and
State
and
local
agencies,
and
reviewed
over
130,000
comments
from
private
citizens,
environmental
groups,
state
officials
and
industry
representatives.

With
regard
to
the
energy
sector,
EPA
finds
that
the
NSR
program
has
not
significantly
impeded
investment
in
new
power
plants
or
refineries.
For
the
utility
industry,
this
is
evidenced
by
significant
recent
and
future
planned
investment
in
new
power
plants.
Lack
of
construction
of
new
greenfield
refineries
is
generally
attributed
to
economic
reasons
and
environmental
restrictions
unrelated
to
NSR.

As
applied
to
existing
power
plants
and
refineries,
EPA
concludes
that
the
NSR
program
has
impeded
or
resulted
in
the
cancellation
of
projects
which
would
maintain
and
improve
reliability,
efficiency
and
safety
of
existing
energy
capacity.
Such
discouragement
results
in
lost
capacity,
as
well
as
lost
opportunities
to
improve
energy
efficiency
and
reduce
air
pollution.

For
the
refining
and
other
industries,
EPA
concludes
that
NSR
as
applied
to
existing
plants
discourages
projects
that
would
have
provided
needed
capacity
or
efficiency
improvements
and
would
not
have
increased
air
pollution
 
in
fact
in
some
cases
air
pollution
may
have
decreased.
EPA
believes
this
can
result
in
lost
capacity
or
foregone
opportunities
to
increase
capacity
without
increasing
emissions.

Finally,
with
regard
to
environmental
protection,
EPA
concludes
that
preventing
emissions
of
pollutants
covered
by
NSR
does
result
in
significant
environmental
and
public
health
benefits.
Specifically
quantifying
the
NSR
program's
contribution
to
these
benefits
is
very
difficult
because
of
the
variety
of
Clean
Air
Act
programs
that
address
these
pollutants
and
because
there
is
no
tracking
by
any
government
agency
of
the
reductions
in
emissions
that
sources
make
due
to
the
program.
Moreover,
EPA
recognizes
that
the
Agency
does
not
currently
have
other
information
that
would
be
necessary
to
quantify
risk
reduction
benefits
associated
with
the
program.
1
See,
Resolution
Number
01­
12,
Environmental
Council
of
States
on
Reform
of
the
New
Source
Review
Regulations
dated
August
28,
2001,
National
Governors
Association
Policy
Position,
NR­
18
Comprehensive
National
Energy
Policy;
Section
18.6.

2
However,
EPA
believes
that
the
inability
to
make
exact
estimates
does
not
mean
that
the
benefits
of
the
NSR
program
are
insignificant.
EPA
also
believes,
however,
that
for
particular
industry
sectors
the
benefits
currently
attributed
to
NSR
could
be
achieved
much
more
efficiently
and
at
much
lower
cost
through
the
implementation
of
a
multi­
pollutant
national
cap
and
trade
program.
In
particular
the
President's
Clear
Skies
initiative
is
a
much
more
certain
and
effective
way
of
achieving
emissions
reductions
from
the
power
generation
sector.

For
virtually
the
entire
history
of
the
NSR
program,
representatives
of
industry,
state
and
local
agencies,
and
environmental
groups
have
worked
with
EPA
on
developing
improvements
to
the
NSR
program.
These
efforts
came
to
a
head
in
1996,
when
EPA
proposed
a
rule
to
"
reform"
the
NSR
program.
Even
after
the
proposal,
stakeholders
have
invested
countless
hours
in
trying
to
find
ways
to
make
the
program
better.
Based
on
the
conclusions
of
this
study
and
the
recommendations
from
the
State
Governors
and
Environmental
Commissioners1
and
other
stakeholders,
EPA
now
plans
to
finish
the
task
of
improving
and
reforming
the
NSR
program.

I.
The
Charge
to
EPA
In
its
May,
2001
National
Energy
Policy
Report,
the
National
Energy
Policy
Development
(
NEPD)
Group
recommended
that
the
Administrator
of
the
Environmental
Protection
Agency
(
EPA),
in
consultation
with
the
Secretary
of
Energy
and
other
Federal
agencies,
"
review
New
Source
Review
regulations,
including
administrative
interpretations
and
implementation,
and
report
to
the
President
within
90
days
on
the
impact
of
the
regulations
on
investment
in
new
utility
and
refinery
generation
capacity,
energy
efficiency,
and
environmental
protection."
Consistent
with
this
recommendation,
EPA
conducted
its
examination
and
is
now
issuing
this
report.
This
report
describes
EPA's
conclusions
about
the
impacts
of
NSR
on
these
three
issues
based
on
its
review
of
the
available
information
and
comments.

II.
Background
EPA
assembled
an
interagency
team
for
this
project,
including
representatives
from
the
Department
of
Energy
(
DOE),
Department
of
the
Interior
(
DOI),
Office
of
Management
and
Budget
(
OMB),
White
House
Council
on
Environmental
Quality
(
CEQ),
and
the
National
Economic
Council
(
NEC).
In
consultation
with
this
group,
EPA
prepared
a
background
paper,
which
was
released
on
June
22,
2001
(
EPA
Background
Paper).
This
paper
described
available
data
relevant
to
the
three
issues
EPA
was
charged
with
reporting
on:
investment
in
utility
and
refinery
capacity,
energy
efficiency,
and
environmental
protection.
The
background
paper
included
EPA's
own
data,
as
well
as
data
provided
in
a
supporting
report
by
ICF
Consulting
Inc.
(
ICF
Report),
which
summarized
ICF's
survey
of
the
available
literature
and
public
statements
on
NSR
issues.
The
background
paper
presented
the
data
to
facilitate
public
comment,
and
to
provide
the
opportunity
for
external
reviewers
to
provide
additional
relevant
data.
The
background
paper
did
not
draw
conclusions
or
make
recommendations.
2
Note
that
many
parties
submitted
comments
concerning
issues
unrelated
to
the
NEPD's
recommendation
for
EPA
to
review
on
the
impact
of
the
regulations
on
investment
in
new
utility
and
refinery
generation
capacity,
energy
efficiency,
and
environmental
protection.
For
example,
numerous
parties
offered
comments
as
to
the
merits
of
pending
NSR
enforcement
cases.
This
report
does
not
summarize
issues
unrelated
to
the
NEPD's
charge.

3
Following
the
background
paper's
release,
EPA
initiated
an
intensive
public
outreach
effort,
consisting
of
three
components:
(
1)
a
30­
day
public
comment
period;
(
2)
a
series
of
four
public
hearings
held
in
locations
across
the
country;
and
(
3)
a
series
of
meetings
with
more
than
100
stakeholder
groups,
including
environmental
organizations,
industry
representatives,
and
State
and
local
governments.
During
this
public
outreach
period,
EPA
received
written
comments
from
over
130,000
individuals
and
organizations.
A
total
of
255
people
testified
at
the
four
hearings.
All
of
the
materials
received
during
the
public
outreach
period,
including
written
comments,
transcripts
of
the
hearings,
and
attendance
lists
and
written
materials
in
connection
with
the
stakeholder
meetings,
are
available
in
public
docket
number
A­
2001­
19
at
the
EPA's
Office
of
Air
and
Radiation
Docket
and
Information
Center.

This
report
discusses
the
statutory
and
regulatory
provisions
of
the
New
Source
Review
(
NSR)
pre­
construction
permitting
program.
While
the
report
explains
the
views
of
many
parties
regarding
the
requirements
of
the
NSR
program,
it
is
not
intended
to
affect
the
NSR
program
or
actions
that
EPA
has
taken
to
implement
or
enforce
the
NSR
program2.
This
report
does
not
substitute
for
statutory
provisions
or
regulations,
nor
is
it
a
guidance
document
reflecting
EPA's
interpretation
of
statutory
or
regulatory
provisions.
Its
purpose
is
to
summarize
information
that
EPA
has
received
relating
to
the
NSR
program
and
to
report
on
EPA's
findings
concerning
whether
the
NSR
program
has
affected
investment
in
new
utility
and
refinery
generation
capacity,
energy
efficiency,
and
environmental
protection.

New
Source
Review
EPA
is
strongly
supportive
of
the
goals
of
the
NSR
permitting
program,
whose
basic
requirements
are
established
in
parts
C
and
D
of
Title
I
of
the
Clean
Air
Act
(
CAA).
The
purpose
of
the
NSR
program
is
to
protect
public
health
and
welfare,
as
well
as
national
parks
and
wilderness
areas,
as
new
sources
of
air
pollution
are
built
and
when
existing
sources
are
modified
in
a
way
that
significantly
increases
air
pollutant
emissions.
Specifically,
NSR's
purpose
is
to
ensure
that
when
new
sources
are
built
or
existing
sources
undergo
major
modifications:
(
1)
air
quality
improves
if
the
change
occurs
where
the
air
currently
does
not
meet
federal
air
quality
standards;
and
(
2)
air
quality
is
not
significantly
degraded
where
the
air
currently
meets
federal
standards.
The
fundamental
philosophy
underlying
the
NSR
program
is
that
a
source
should
install
modern
pollution
control
equipment
when
it
is
built
(
for
new
sources)
or
when
it
makes
a
major
modification
(
for
existing
sources).
Congress
believed
that
incorporating
pollution
controls
into
the
design
and
construction
when
new
units
are
built,
or
when
major
modifications
occur,
is
generally
more
efficient
than
adding
on
controls
after
construction.

The
NSR
program
is
by
no
means
the
primary
regulatory
tool
to
address
air
pollution
from
existing
sources.
The
Clean
Air
Act
provides
for
several
other
public
health­
driven
and
visibilityrelated
control
efforts:
for
example,
the
National
Ambient
Air
Quality
Standards
Program
3
The
term
NSR
usually
refers
to
the
overall
program,
but
is
sometimes
also
used
as
shorthand
to
refer
to
nonattainment
NSR,
which
may
be
a
source
of
confusion.
In
this
document,
we
will
use
NSR
to
refer
to
the
general
program
(
both
nonattainment
NSR
and
PSD),
and
will
use
nonattainment
NSR
when
referring
specifically
to
NSR
for
nonattainment
areas.
4
Connecticut,
Delaware,
Maine,
Maryland,
Massachusetts,
New
Hampshire,
New
Jersey,
New
York,
Pennsylvania,
Rhode
Island,
Vermont,
and
Washington,
D.
C.

4
implemented
through
enforceable
State
Implementation
Plans,
the
NO
X
SIP
Call,
the
Acid
Rain
Program,
the
Regional
Haze
Program,
etc.
Thus,
while
NSR
was
designed
by
Congress
to
focus
particularly
on
sources
that
are
newly
constructed
or
that
make
major
modifications,
Congress
provided
numerous
other
tools
for
assuring
that
emissions
from
existing
sources
are
adequately
controlled.
For
example,
the
national
cap
on
S0
2
emissions
established
under
the
Acid
Rain
Program
applies
to
all
existing
electricity
generating
units,
without
regard
to
the
date
of
construction
or
whether
a
given
source
has
been
modified.

NSR
operates
by
requiring
a
source
to
obtain
a
permit
prior
to
construction
or
major
modification.
The
permit
establishes
various
actions
that
the
source
must
undertake
to
control
its
emissions
of
air
pollution.
However,
NSR
only
applies
if
the
construction
project
will
emit
air
pollution
that
exceeds
threshold
levels
established
in
the
NSR
regulations.
For
a
new
source,
NSR
is
triggered
only
if
the
potential
emissions
qualify
as
major.
For
an
existing
major
source
making
a
modification,
NSR
is
only
triggered
if
the
modification
will
result
in
a
significant
net
increase
in
emissions.

The
major
NSR
program
comprises
two
separate
parts:
Nonattainment
NSR
and
Prevention
of
Significant
Deterioration
(
PSD).
3
These
two
programs
have
separate
requirements
to
address
the
differing
air
quality
planning
needs
in
the
areas
where
they
apply.
Nonattainment
NSR
applies
in
areas
where
air
is
unhealthy
to
breathe
­
i.
e.
where
the
established
national
ambient
air
quality
standards
(
NAAQS)
for
a
CAA
criteria
pollutant
are
not
being
met.
These
areas
are
called
nonattainment
areas.
Nonattainment
NSR
for
major
sources
of
certain
pollutants
also
applies
in
the
federally
designated
ozone
transport
region
(
OTR),
which
consists
of
eleven
northeastern
States
and
Washington,
D.
C.
4
PSD
applies
to
major
sources
located
in
areas
where
air
quality
is
currently
acceptable
­
i.
e.,
where
the
NAAQS
for
CAA
criteria
pollutants
are
being
met.
These
are
called
attainment
areas.
Because
nonattainment
areas
have
poorer
air
quality,
nonattainment
NSR
requirements
are
generally
more
stringent
than
PSD
requirements.

III.
Impact
on
Investment
in
New
and
Existing
Utility
and
Refinery
Generation
Capacity
and
Energy
Efficiency
The
EPA
begins
by
examining
the
question
of
whether
the
NSR
program
has
an
impact
on
investment
in
projects
that
would
increase
or
preserve
utility
and
refinery
generation
capacity
or
that
would
improve
energy
efficiency.
We
received
extensive
comments
on
this
issue,
reflecting
widely
varying
views
on
whether
there
is
an
impact
and,
if
so,
on
its
nature
and
extent.

In
general,
comments
made
by
both
the
electric
utility
industry
and
the
petroleum
refining
industry
consistently
assert
that
the
NSR
program
has
a
significant
and
adverse
impact
on
5
These
comments
were
consistently
raised
by
companies
representing
virtually
all
types
(
e.
g.,
coal­
fired;
oil­
fired
or
gas­
fired)
and
sizes
of
electric
generating
facilities.
See,
e.
g.,
Comments
of
the
Clean
Energy
Group
(
CEG)
[
II­
D­
291];
Comments
of
the
Utility
Air
Regulatory
Group
(
UARG)
[
II­
D­
303];
Comments
of
Class
of
`
85
Regulatory
Response
Group
(
Class
of
`
85
Group)
[
II­
D­
268];
Comments
of
National
Rural
Electric
Cooperative
Associations
(
NRECA)
[
II­
D­
322].
The
members
of
these
groups,
as
well
as
individual
utilities
that
filed
comments
expressing
the
same
conclusion,
span
the
entire
United
States.
See,
e.
g.,
Comments
of
Northeast
Utilities
Service
Company
(
NUSCO)
[
II­
D­
331];
Comments
of
Cinergy
[
II­
D­
270];
Comments
of
Sunflower
Electric
Power
Corporation
[
II­
D­
292];
Comments
of
Tri­
State
Generation
and
Transmission
Association
[
II­
D­
335];
Comments
of
West
Associates
[
II­
D­
216];
Comments
of
Salt
River
Project
(
SRP)
[
II­
D­
320].
Even
waste­
to­
energy
facilities
agreed
with
this
conclusion.
See
e.
g.,
Comments
of
American
Ref­
Fuel
[
II­
D­
214].
The
refining
industry
offered
similar
comments.
See
NPRA
Letter
to
Stephanie
Daigle,
EPA,
7/
23/
2001.
6
See
comments
by
Michigan
Department
of
Environmental
Quality,
representing
a
workgroup
including
Alabama,
Michigan,
North
Carolina,
South
Carolina,
Virginia
and
West
Virginia
permitting
staff.
[
II­
E­
09].
7
For
other
State
comments,
see
STAPPA/
ALAPCO,
[
II­
D­
313],
CARB
[
II­
D­
468],
RAPCA
[
II­
D­
302],
Wisconsin,
Missouri,
et.
al.
For
environmental
groups,
see,
Clean
Air
Task
Force
[
II­
D­
236],
NRDC,
Sierra
Club
[
II­
D­
437],
et.
al.
8
See
Natural
Resources
Defense
Council
(
NRDC)
comments
[
II­
D­
267]
at
1.

5
investment
in
expanding
and
preserving
capacity,
as
well
as
on
energy
efficiency.
5
These
commenters
assert
that
the
program
is
in
need
of
fundamental
reform.
Other
industries
(
as
discussed
in
Section
IV
below)
made
similar
assertions,
as
did
some
State
permitting
authorities.
These
commenters
said
that
investment
is
hindered
by
(
1)
regulatory
uncertainty
and
lack
of
flexibility
resulting
from
alleged
recent
policy
"
re­
interpretations"
related
to
the
applicability
of
the
program's
requirements;
and
(
2)
the
added
costs
and
delays
imposed
by
the
NSR
process.
6
Other
commenters,
including
environmental
groups
and
some
State
and
local
permitting
authorities,
expressed
the
opposite
view.
They
assert
that
NSR
does
not
appear
to
be
significantly
hindering
such
investment,
adding
that
NSR
has
resulted
in
large
benefits
to
the
environment
while
allowing
for
increased
energy
and/
or
fuel
supplies.
7
One
environmental
commenter
does
not
believe
that
there
is
sufficient
information
to
conclude
that
NSR
is
a
primary
factor
driving
decisions
to
invest
or
not
to
invest
in
capacity.
8
This
section
discusses
our
conclusions
based
on
a
review
of
the
available
data
and
comments
received
regarding
investment
in
new
capacity
and
energy
efficiency.
Because
the
issues
associated
with
new
and
modified
source
permitting
differ,
this
paper
will
discuss
separately
the
impact
on
new
sources
and
the
impact
on
existing
sources
undergoing
changes.

A.
New
Sources
Focusing
first
on
the
impacts
of
NSR
on
investment
in
new
capacity,
the
EPA
finds
that
NSR
does
not
appear
to
have
a
significant
impact
on
investment
in
new
utility
or
refinery
plants.
The
discussion
below
indicates
that,
for
utilities,
significant
new
capacity
has
been
permitted
in
recent
years
and
substantial
additional
greenfield
capacity
is
planned.
For
refiners,
decisions
about
whether
to
construct
new
greenfield
refineries
are
primarily
driven
by
economic
and
9
This
214
GW
increase
would
represent
a
30
percent
increase
over
the
current
installed
capacity
level,
and
would
restore
national
reserve
margins
to
about
25
percent,
from
a
low
of
8
percent
in
1999.
10
See
New
Jersey
DEP
comments
[
II­
D­
310].
11
The
State
of
Kentucky,
in
fact,
put
a
hold
on
any
new
permit
applications
for
electrical
generation
sources
until
it
can
analyze
the
environmental
impacts
of
the
large
volume
of
pending
permit
applications.
12
See,
e.
g.,
California
Air
Resources
Board
(
CARB)
[
II­
D­
468],
Georgia
Department
of
Natural
Resources
(
DNR)
[
II­
D­
341],
Wisconsin
DNR
[
II­
G­
71],
STAPPA/
ALAPCO
[
II­
D­
303],
Clean
Air
Task
Force
[
II­
D­
236],
NRDC
[
II­
D­
267]
and
other
similar
comments.

6
environmental
considerations.
It
does
not
appear
that
NSR
has
a
significant
impact
on
these
considerations.

1.
Utilities
For
electric
utilities,
significant
new
sources
were
permitted
in
recent
years
(
dominated
by
natural
gas­
fired
systems)
and
more
are
planned.
The
background
paper
noted
current
plans
of
certain
companies
to
bring
into
service
units
producing
more
than
120
Gigawatts
(
GW)
in
the
coming
years.
An
analysis
by
the
NorthBridge
group,
prepared
for
the
Clean
Air
Task
Force,
uses
RDI's
NewGen
database
to
estimate
that
it
is
likely
that
214
GW
­
and
possibly
as
much
as
400
GW
­
of
new
generating
capacity
will
come
online
before
2005,
based
on
a
survey
of
data
on
plants
at
various
stages
of
development.
9
Several
State
commenters
presented
similar
data.
For
example,
New
Jersey
stated
that
it
had
permitted
over
2500
MW
of
new
electric
generation
since
July
1999,
and
had
proposed
to
approve
another
1700
MW
in
July
of
200110.
Another
5800
MW
of
applications
were
under
review,
and
another
2000
MW
of
projects
were
in
the
pre­
application
meeting
stage.
These
projects
cover
22
facilities
and
49
units.
This
12,000
MW
will
result
in
a
60%
increase
over
the
18,000
MW
of
existing
generating
capacity
in
New
Jersey.
11
Other
States
and
environmental
group
commenters
presented
similar
data.
12
Although
most
of
these
projects
will
be
subject
to
NSR,
the
program
does
not
appear
to
be
hindering
their
development.

In
general,
the
DOE's
experience
is
that
far
more
capacity
is
planned
than
is
ever
actually
realized.
As
it
related
to
the
analysis
by
the
NorthBridge
group,
the
DOE
projects
in
its
2001
Annual
Energy
Outlook
that
only
a
small
fraction
of
the
capacity
estimates
by
NorthBridge
will
actually
come
on
line
by
2005.
For
the
period
of
1999
to
2005,
DOE
estimates
the
following:

°
Overall
generation
will
increase
from
3386
billion
kilowatt­
hours
(
BKWH)
to
3810
BKWH.

°
Overall
capacity
will
increase
by
74
GW
(
from
745
gigawatts
(
GW)
to
819
GW).

°
For
coal­
fired
power
plants,
capacity
will
decrease
slightly
(
from
306
GW
to
301
GW),
while
generation
increases
from
1833
BKWH
to
2085
BKWH,
as
existing
units
increase
their
hours
of
operation.
13
STAPPA/
ALAPCO
comments
[
II­
D­
313]
at
6.
14
The
primary
air
pollution
control
requirement
commonly
imposed
on
natural
gas
combustion
is
selective
catalytic
reduction,
which
adds
about
$
30
per
kilowatt
to
the
cost
of
a
combined
cycle
generation
system.
New
pulverized
coal
systems
require
electrostatic
precipitators
or
fabric
filters
for
particulate
matter
control,
scrubbers
for
sulfur
dioxide
control,
selective
catalytic
reduction
for
nitrogen
oxide
control,
and
perhaps
additional
control
technology
for
air
toxics.
Cumulatively,
the
systems
needed
for
coal­
based
generation
cost
over
$
200
per
7
°
For
gas­
fired
plants,
combined­
cycle
units
will
increase
in
capacity
from
20
GW
to
50
GW,
while
generation
increases
from
371
BKWH
to
584
BKWH.

While
these
data
indicate
continued
expansion
in
new
generating
capacity,
some
industry
commenters
assert
that
NSR
can
nevertheless
introduce
costs
and
delays
to
the
process
of
bringing
new
generating
units
online,
as
well
as
have
an
impact
on
fuel
supply
flexibility.
Utilities
cited
implementation
of
the
requirements
for
preconstruction
monitoring,
modeling,
and
consultation
with
Federal
Land
Managers,
saying
that
the
processing
time
by
Federal,
State
and
local
governments
and
potential
permit
appeals
can
result
in
significant
costs
and
delays
in
obtaining
a
permit.
In
particular,
industry
commenters,
as
well
as
some
State
permitting
authorities,
attribute
a
significant
portion
of
the
delay
in
obtaining
NSR
permits
to
the
large
body
of
NSR
guidance
that
has
been
issued
over
the
course
of
many
years,
by
both
EPA
and
State
agencies
administering
delegated
programs.
This
guidance
frequently
is
case­
specific
in
nature.
Many
commenters
consider
the
guidance
to
be
ambiguous
and,
in
some
cases,
inconsistent.

Among
the
various
aspects
of
the
NSR
program
that
industry
commenters
more
specifically
identified
as
concerns
for
new
sources
included
the
following:

°
How
to
determine
which
emissions
control
technologies
qualify
as
best
available
control
technology
("
BACT")
or
lowest
achievable
emissions
rate
("
LAER")
technology
using
EPA's
"
top
down"
policy
and
the
Agency's
BACT/
LAER
clearinghouse.

°
Procedural
concerns
about
guidance
issued
by
Federal
Land
Managers
related
to
permitting
near
Class
I
areas.

°
The
limitation
on
construction
activities
prior
to
issuance
of
a
permit,
which
is
of
particular
concern
when
(
1)
the
permit
undergoes
lengthy
appeals
processes,
or
(
2)
the
climate
is
cold
and
the
construction
season
is
thus
shorter.

°
The
cost
and
availability
of
offsets
in
nonattainment
areas.
Commenters,
particularly
in
California
and
New
York,
noted
that
shortages
in
available
offsets
have
the
potential
to
significantly
increase
the
cost
of
NSR
permitting
in
certain
limited
areas.
Permitting
authority
commenters
noted
that
offsets
represent
from
1­
6
%
of
the
cost
of
a
new
power
plant.
13
Commenters
further
stated
that
NSR
control
requirements
affect
fuel
supply
choices
for
new
installations.
They
point
out
that
the
cost
of
air
pollution
control
represents
a
much
greater
proportion
of
the
cost
of
construction
at
coal­
fired
facilities
than
at
gas­
fired
plants.
14
Operation
kilowatt,
and
add
about
20%
to
the
cost
of
a
new
coal­
fired
system.
For
a
1000
MW
unit,
these
translate
into
a
cost
of
$
200
million.
15
See,
e.
g.,
STAPPA
[
II­
D­
313]
at
3,
New
Jersey
DEP
[
II­
D­
310]
at
2.
16
See
CARB
[
II­
D­
468]
at
4.
17
Wisconsin
DNR
comments
[
II­
G­
71]
at
1.
18
See,
Testimony
of
the
National
Petrochemical
and
Refiners
Association
(
NPRA)
before
the
Senate
Subcommittee
on
Clean
Air,
Wetlands,
Private
Property
and
Nuclear
Safety
on
Apr.
5,
2001.

8
and
maintenance
costs
are
also
higher.
They
believe
this
discourages
investment
in
new
coal­
fired
plants.

Other
stakeholders
offered
a
different
view.
Several
State
and
local
permitting
authorities
noted
that
the
NSR
process
can
generally
be
accomplished
in
a
reasonable
time,
and
within
the
same
time
frame
as
the
other
elements
involved
in
planning
of
a
typical
electric
generator
project.
15
Some
States
reported
acceleration
of
permitting
times
for
new
utility
sources
consistent
with
that
reported
in
the
EPA
Background
Paper.
16
One
State
commenter
suggested
that
the
perception
that
NSR
is
lengthy,
cost­
intensive,
and
uncertain
is
really
not
the
norm,
though
it
can
be
true
in
exceptional
cases.
17
In
EPA's
experience,
NSR
has,
in
some
individual
cases,
impeded
new
power
projects.
However,
as
a
general
matter,
available
information
indicates
that
NSR
typically
does
not
represent
a
significant
barrier
to
the
construction
of
new
electricity
plants.
As
for
the
impact
of
NSR
on
fuel
choices
for
new
facilities,
EPA
notes
that
NSR
typically
does
not
require
significantly
greater
levels
of
control
at
new
coal­
fired
plants
than
the
recently
updated
NSPS
for
large
electric
generating
units.
Thus,
NSR
itself
is
not
the
only
driver
with
regard
to
air
pollution
control
costs
at
new
coal­
fired
units
and
does
not
appear
to
significantly
influence
fuel
choices
at
new
facilities.

2.
Refineries
As
noted
earlier,
the
construction
of
new
"
greenfield"
petroleum
refineries
in
the
near
future
seems
unlikely
for
various
economic
and
regulatory
reasons,
primarily
unattractive
profit
margins.
Industry
has
reported
that
the
rates
of
return
for
refineries
have
averaged
about
5
percent
in
the
last
decade,
roughly
equivalent
to
the
return
from
a
passbook
savings
account,
but
with
much
greater
risk.
As
a
result,
building
new
plants
at
new
sites
is
highly
unlikely.
18
The
EPA
agrees
with
this
assessment.
Moreover,
while
any
new
refinery
would
be
required
to
obtain
an
NSR
permit,
the
available
information
does
not
indicate
that
NSR
permitting
is
among
the
most
significant
impediments
to
the
construction
of
new
refineries.
Refinery
commenters
indicate
that
any
additional
U.
S.
refinery
capacity
must
come
from
either
efficiency
improvements
or
expansion
at
existing
refineries
(
discussed
below).

B.
Existing
Sources
9
The
vast
majority
of
concerns
about
NSR
raised
during
the
review
pertained
to
existing
sources.
As
discussed
below,
the
EPA
believes
that
commenters
have
identified
areas
where
NSR
can
discourage
investment
in
both
preserving
and
maintaining
utility
and
refinery
generating
capacity
as
well
as
in
improving
energy
efficiency
and
expanding
capacity.

1.
Utilities
With
respect
to
existing
sources,
comments
from
across
the
spectrum
of
the
utility
industry
consistently
asserted
that
the
NSR
program
imposes
significant
burdens
on
the
utility
practices
necessary
to
maintain
the
safety,
availability,
efficiency
and
reliability
of
the
electricity
supply
at
existing
sources.
They
further
assert
it
can
have
a
highly
negative
impact
on
the
nation's
power
supply.
The
result,
they
conclude,
is
that
the
program
hinders
investment
in
projects
intended
to
expand
and
preserve
generating
capacity
at
existing
electric
generation
units.
In
addition,
as
discussed
below,
many
utility
commenters
believe
that
the
current
NSR
program
has
actively
discouraged
efficiency
improvement
projects,
which
they
believe
not
only
can
have
net
environmental
benefits,
but
also
can
provide
an
effective
short­
term
response
to
tight
reserve
margins
at
many
locations
in
the
United
States.
On
the
other
hand,
environmental
groups
do
not
believe
that
there
is
sufficient
information
to
conclude
that
NSR
is
the
primary
factor
driving
decisions
to
invest
in
new
capacity
at
existing
sources
or
that,
absent
NSR,
significant
investments
would
have
been
made
that
are
presently
not
being
made
in
recapturing
lost
existing
capacity
due
to
deterioration
of
equipment.
This
section
examines
more
closely
the
capacity
issues
at
electric
utilities,
followed
by
the
energy
efficiency
issues.

a.
Impact
on
Utility
Projects
to
Maintain
the
Availability,
Reliability,
and
Safety
of
the
Electric
Power
Supply
(
i)
NSR
Applicability
The
utility
industry
comments
predominantly
focused
on
the
exclusion
from
major
NSR
permitting
requirements
for
activities
that
represent
"
routine
maintenance,
repair
and
replacement."
They
asserted
that,
in
recent
years,
EPA
has
narrowed
its
interpretation
of
this
exclusion
to
the
point
where
NSR
potentially
applies
to
repair
and
replacement
activities
that
are
customarily
undertaken
within
the
industry
to
assure
the
availability,
reliability,
and
safety
of
power
plant
operations.
Commenters
believe
that
under
such
an
interpretation
NSR
would
be
required
whenever
the
work
involved:
(
1)
a
component
that
is
replaced
infrequently
in
the
life
of
an
industrial
facility;
(
2)
a
component
that
is
large
and
expensive
(
in
absolute
terms);
or
(
3)
a
replacement
component
that
is
better
designed
and
will
improve
the
availability
or
efficiency
of
the
facility.

Thus,
according
to
the
utility
commenters,
because
electricity
generation
units
are
inherently
large,
complex,
and
expensive
(
in
absolute
terms),
most
power
plant
repair
and
replacement
activity
would
not
be
covered
by
the
exclusion.
Because
of
the
costs
and
potential
delays
associated
with
NSR,
they
believe
that
this
has
discouraged
activities
intended
to
maintain
the
reliability,
availability,
and
safety
of
existing
power
plants;
and/
or
has
required
generators
to
limit
the
output
of
their
power
plants
to
avoid
triggering
NSR,
regardless
of
their
capacity,
in
order
to
maintain
the
units
during
their
normal
useful
lives.
NSR
costs
and
delays
are
of
particular
concern
to
commenters
for
such
changes
at
existing
units
because
(
1)
while
certain
19
See
Clean
Air
Task
Force
Comments
[
II­
D­
236],
Appendix
D.
20
UARG
Comments
[
II­
D­
303]
at
29­
32.
21
UARG
Comments
[
II­
D­
303]
Attachment
C.
22
See
Jerry
Golden,
TVA,
Routine
Maintenance
of
Electric
Generating
Stations
(
February
2000)
("
TVA
2000
Report"),
described
in
UARG
Comments
[
II­
D­
303]
at
29­
31.
23
TVA
2000
Report
at
25.

10
projects
might
be
relatively
inexpensive
absent
NSR,
they
believe
the
cost
of
controls
resulting
from
NSR
can
make
them
cost­
prohibitive
to
undertake,
which,
in
turn,
can
adversely
affect
the
availability
and
reliability
of
plant
operations
and
discourage
such
projects,
and
(
2)
they
believe
that
units
may
need
to
be
offline
until
permitting
can
occur,
so
delays
in
permitting
can
have
significant
impacts
on
energy
supply
through
lost
generation
during
this
time.

Although
utilities
stated
that
NSR­
required
controls
are
expensive
relative
to
the
gains
associated
with
projects
that
might
trigger
NSR,
other
commenters
noted
that
these
costs
are
small
compared
to
the
company's
revenue.
The
Clean
Air
Task
Force
submitted
a
study
by
MSB
Energy
Associates
performed
on
a
sample
of
51
existing
coal­
fired
utility
units.
The
study
concludes
that
if
these
units
triggered
NSR
and
had
to
install
BACT­
level
controls,
the
cost
would
be
modest
relative
to
the
size
and
revenue
level
of
the
companies.
19
In
the
commenters'
view,
this
impact
is
exchanged
for
significant
environmental
benefits,
estimated
at
2.8
million
tons
per
year
of
sulfur
dioxide
(
SO
2
)
(
22%
of
all
power
plant
SO
2
emissions
in
the
U.
S.)
and
1.0
million
tons
per
year
of
NO
X
(
19%
of
all
power
plant
NO
X
emissions
in
the
U.
S.).

According
to
industry,
thousands
of
repair
and
replacement
projects
are
undertaken
by
facilities
each
year
and
that,
as
a
result,
NSR
permitting
is
potentially
triggered
early
in
the
life
of
virtually
every
electric
utility
plant,
and
then
repeatedly
thereafter.
20
The
industry
commenters
submitted
information
about
the
types
of
projects
they
stated
that
they
typically
undertake,
which
they
maintain
are
required
to
ensure
reliability,
availability,
or
safety
of
their
facilities,
but
which
they
believe
EPA
would
classify
as
non­
routine
and
therefore
would
potentially
be
subject
to
NSR
if
they
resulted
in
a
significant
net
emissions
increase.
21
For
example,
a
survey
undertaken
by
the
Tennessee
Valley
Authority
(
TVA)
reported
the
frequency
with
which
particular
repair
and
replacement
projects
are
undertaken
within
the
electric
utility
industry.
22
The
TVA
survey
covered
approximately
20%
of
the
electric
utility
industry
­­
219
units
totaling
about
80,000
MW
­­
and
included
a
review
of
case
studies
and
statistics
regarding
cyclone
replacement,
balanced­
draft
conversion,
reheater
replacement,
and
economizer
replacement.
For
example,
their
survey
states
that,
at
the
190
units
in
the
survey
that
had
reheaters,
there
were
213
reheater
replacement
projects
(
some
reheaters
were
replaced
more
than
once).
At
the
202
units
in
the
survey
that
had
economizers,
there
were
98
economizer
replacement
projects.
For
both
components,
replacements
occurred
as
early
as
5
years
after
initiation
of
a
unit's
commercial
operation,
or
as
late
as
40
to
50
years.
Similarly,
at
151
boilers
originally
constructed
as
forced
draft
systems,
utilities
replaced
79
systems
with
balanced
draft
systems,
primarily
to
address
"
equipment
degradation,
maintenance
problems,
health
and
safety
concerns,
and
pollution
control
requirements."
23
Finally,
the
TVA
survey
reported
that,
since
1979,
300
cyclones
out
of
701
had
been
replaced
at
the
96
electricity­
generating
stations
in
the
United
States
powered
by
cyclone
boilers.
UARG
similarly
reported
a
more
complete,
recent
24
UARG
Comments
[
II­
D­
303]
at
31­
32.
25
NRECA
Comments
[
II­
D­
322]
at
14­
15;
see
also
Class
of
`
85
Group
Comments
[
II­
D­
268]
at
9
("
Electric
generating
plant
personnel
have
been
placed
in
the
untenable
position
of
not
being
able
to
correct
and
improve
the
reliability
and
efficiency
of
their
plants,
resulting
in
compromised
safety
to
plant
employees
and
the
general
public,
without
risking
an
enforcement
action.");
Dairyland
Comments
(
II­
D­
324)
at
4
(
EPA's
current
"
interpretation
may
compromise
the
reliability
and
efficiency
of
existing
plants
and
could
undermine
the
preservation
of
a
diverse
energy
supply.").
26
See,
e.
g.,
RAPCA
[
II­
D­
302],
Adirondack
Council
[
II­
D­
136],
Public
Citizen
[
II­
D­
327].
27
Public
Citizen
[
II­
D­
327].

11
census
of
the
entire
coal­
fired
steam
electric
generating
industry.
24
This
census
sought
industrywide
information
regarding
the
frequency
of
maintenance,
repair
and
replacement
activities
that
they
believe
EPA
considers
non­
routine.
The
census
results
are
reported
to
show:

°
The
industry
has
undertaken
tens
of
thousands
of
such
maintenance,
repair
or
replacement
activities;

°
Every
unit
in
the
industry
has
undertaken
such
activities;

°
Approximately
50%
of
the
units
in
the
industry
will
have
undertaken
such
activity
within
five
years
of
the
unit's
in­
service
date;

°
Each
unit
in
the
industry
undertakes
on
average
annually
at
least
one
such
activity.

In
short,
in
the
view
of
many
industry
commenters,
an
inappropriately
narrow
routine
maintenance
exclusion
would
not
exclude
many
common
maintenance
projects.
According
to
these
commenters,
this
would
leave
nearly
every
coal­
fired
generating
unit
in
a
constant
state
of
obligation
to
evaluate
whether
each
of
these
numerous
projects
would
trigger
NSR,
and
if
so,
whether
the
costs
associated
with
NSR
(
including,
if
applicable,
the
costs
of
add­
on
controls
and
potential
downtime)
would
render
such
projects
cost­
prohibitive.
As
discussed
below,
if
such
projects
are
found
to
be
cost
prohibitive,
commenters
predict
steady
deterioration
of
existing
capacity,
and
limited
investment
in
the
recovery
of
such
capacity
at
existing
sources.
Many
industry
commenters
echoed
this
conclusion
and
asserted
that
the
situation
is
unacceptable
and
must
be
corrected
to
reflect
the
real
environment
surrounding
routine
maintenance
within
the
electrical
utility
industry.
25
On
the
other
hand,
environmental
group
commenters
and
some
permitting
authorities
felt
that
the
routine
maintenance
exclusion
is
appropriate.
They
believed
that
a
less
narrow
exclusion
would
allow
the
exception
to
swallow
the
rule.
In
this
vein,
commenters
expressed
concerns
that
large­
scale
capital
projects,
such
as
major
life
extension
projects,
should
not
qualify
as
routine.
26
One
of
these
commenters
expressed
concern
that
a
facility
could
be
virtually
rebuilt
without
triggering
NSR
under
industry's
preferred
interpretations
of
the
routine
maintenance
exemption27.
28
See,
e.
g.,
Class
of
`
85
Group
Comments
[
II­
D­
268]
at
9­
10.
29
Commenters
also
complained
of
delays
in
the
minor
NSR
permitting
process
(
an
average
of
3­
8
months
in
one
utility's
service
area.)
See
Jerry
L.
Golden
&
Donald
P.
Houston,
TVA,
Impacts
of
EPA's
Reinterpretation
of
New
Source
Review
Requirements
­­
Potential
Loss
of
Generating
Capability
on
the
TVA
System,
at
8
(
July
19,
2001)
("
TVA
2001
Report")
(
Attachment
E
to
UARG
Comments
[
II­
D­
303]).
30
See
UARG
Comments
at
39­
42;
see
also
EPA
Background
Paper
at
7.

12
After
reviewing
the
comments,
the
EPA
notes
that
there
are
differing
opinions
amongst
the
commenters
about
the
appropriate
scope
of
the
routine
maintenance
exemption
and
the
resulting
NSR
impacts.
In
determining
whether
an
activity
is
"
routine"
for
purposes
of
being
excluded
from
NSR,
EPA
consistently
has
taken
a
case­
by­
case
approach,
weighing
the
nature,
extent,
purpose,
frequency
and
cost
of
the
work,
as
well
as
other
relevant
factors.
Nevertheless,
the
Agency
recognizes
that
many
industry
commenters
expressed
uncertainty
about
the
scope
of
the
routine
exclusion
and
argued
that
this
uncertainty
will
cause
them
to
delay
or
forego
projects
critical
to
maintaining
the
availability,
reliability
and
safety
of
their
facilities.
In
light
of
the
volume
of
anecdotal
evidence
presented,
the
EPA
concludes
that
concern
about
the
scope
of
the
routine
maintenance
exclusion
is
having
an
adverse
impact
on
projects
that
affect
availability,
reliability,
efficiency,
and
safety.
Changes
to
the
NSR
program
that
add
to
the
clarity
and
certainty
of
the
scope
of
the
routine
maintenance
exclusion
will
improve
the
process
by
reducing
the
unintended
consequences
of
discouraging
worthwhile
projects
that
are
in
fact
outside
the
scope
of
NSR.

(
ii)
Energy
Impacts
According
to
utility
commenters,
the
energy
impact
of
an
inappropriately
narrow
NSR
routine
maintenance
exclusion
would
be
adverse
and
potentially
quite
significant.
In
addition,
the
industry
commenters
stated
that
an
inappropriately
narrow
exclusion
would
leave
many
activities
potentially
subject
to
NSR.
This
circumstance,
they
believe,
would
result
in
limited
alternatives
for
utility
managers.
They
describe
three
alternatives.

First,
utilities
could
go
through
the
NSR
pre­
construction
permitting
process.
The
principal
complaints
against
this
alternative
were
protracted
processing
delays
and
the
attendant
costs,
including
the
costs
of
pollution
control
retrofits.
28
In
addition,
commenters
feared
that,
if
the
interpretation
of
routine
were
to
be
narrowed,
thousands
of
projects
would
trigger
NSR
per
year,
and
would
result
in
even
more
substantial
delays
by
flooding
the
permit
process
with
more
permit
applications
than
it
has
the
capacity
to
process
quickly.

Second,
a
company
could
accept
enforceable
emissions
limits
(
through
a
"
minor"
NSR
permit)
in
the
form
of
a
cap
on
emissions
from
the
affected
units.
29
Commenters
stated,
however,
that
acceptance
of
such
a
cap
would
require
a
utility
to
limit
the
affected
unit's
hours
of
operation
and
production
rates
to
representative
emission
levels
just
prior
to
the
change,
which
could
restrict
the
electricity
supply
in
a
particular
area.
30
Commenters
also
could
limit
emissions
by
adding
pollution
control
technology,
but
commenters
felt
this
was
also
not
a
workable
NSR
avoidance
strategy
because
it
also
could
be
infeasible,
cost­
prohibitive,
and
would
only
be
a
31
See
UARG
comments
at
39­
42.
32
See,
e.
g.,
Class
of
`
85
Group
Comments
[
II­
D­
268]
at
7,
TVA
2001
report
at
7
(
Attachment
E
to
UARG
Comments
[
II­
D­
303]).
33
See,
e.
g.,
id.
34
TVA
2001
report
at
12­
14.
35
Southern
Company,
The
Dismantling
of
Energy
Supply
Capacity
Through
New
Source
Review
(
Attachment
D
to
UARG
Comments
[
II­
D­
303]).
36
First
Energy
Comments
[
II­
D­
261]
at
1.
37
West
Associates
Comments
[
II­
D­
216]
at
7.
38
NRECA
Comments
[
II­
D­
322]
at
7.
Other
commenters
that
submitted
similar
analyses
include:
Minnesota
Power
Comments
[
II­
D­
165]
(
25%
lost
production);
Dairyland
Comments
[
II­
D­
324]
at
7
(
41%
lost
generating
capacity);
SRP
Comments
[
II­
D­
320]
at
6
(
18.5%
loss).

13
temporary
solution.
31
Moreover,
commenters
stated
that
the
delays
associated
with
the
minor
NSR
process
required
to
create
the
limit
still
severely
impact
a
unit's
ability
to
replace
components
necessary
to
get
back
online
quickly
after
a
forced
outage.
32
For
example,
when
a
turbine
rotor
shaft
cracks
or
slag
falls
and
destroys
a
boiler
floor,
the
utility
must
repair
the
component
as
quickly
as
possible
and
restore
the
unit
to
service.
Commenters
claim
that,
if
the
necessary
repairs
were
not
considered
routine
maintenance,
repair
and
replacement,
the
repair
could
not
be
made
until
the
source
obtained
an
NSR
permit.
In
the
meantime,
the
commenters
believe
that
the
utility
could
lose
the
entire
capacity
of
the
unit,
which
could
endanger
the
stability
of
the
electrical
grid
and
create
a
risk
of
regional
blackouts.
33
Commenters
also
argued
that
avoiding
NSR
by
accepting
caps
on
emissions
through
operational
limits
would
constrain
electrical
system
operators'
flexibility
to
deliver
necessary
electricity
at
the
least
cost.
In
this
regard,
several
utilities
analyzed
their
systems
to
estimate
the
restrictions
on
their
ability
to
produce
electricity,
had
what
they
consider
to
be
a
narrow
interpretation
of
the
routine
exclusion
been
applied
over
the
last
twenty
years
and
had
the
utilities
elected
to
obtain
minor
NSR
permits
limiting
generation
to
recent
levels
in
every
instance
they
undertook
certain
replacement
projects.

For
example,
TVA
(
serving
approximately
2.3
million
homes
in
the
Tennessee
River
Valley),
34
reported
that,
over
the
last
twenty
years,
it
would
have
lost
32%
of
its
coal
system's
energy
capability,
or
34
million
megawatt­
hours
(
MW­
hr)
annually.
In
a
similar
analysis,
the
Southern
Company
found
that,
by
the
year
2000,
it
would
have
had
an
energy
shortfall
of
57.5
million
MW­
hr,
and
that
it
would
not
have
been
able
to
meet
38%
of
its
customer
demand.
35
Similarly,
First
Energy
estimated
that
it
would
have
lost
39%
of
its
coal­
fired
generating
capacity
between
1981
and
2000.36
West
Associates
(
a
western
utility
with
a
younger
fleet
of
generating
units)
estimated
a
loss
of
27%
of
generating
capacity
of
one
of
its
plants
just
in
the
next
six
years.
West
Associates
also
estimated
that,
after
10
years
of
operation
under
this
"
cap
system,"
the
Western
System
Coordinating
Council
(
WSCC)
would
have
lost
65
million
MW­
hr
of
generating
capacity,
or
the
equivalent
of
32
power
plants
with
a
net
capacity
of
250
MW
each.
37
The
National
Rural
Electric
Cooperative
Association
(
NRECA)
estimated
that,
in
one
maintenance
cycle,
the
loss
of
capability
for
the
approximately
21,000
MW
of
cooperative­
owned
plants
would
be
12%
to
24%.
38
Nationally,
using
this
analysis
method,
one
commenter
stated
that
it
would
take
200
new
500
megawatt
power
plants
just
to
make
up
the
lost
capacity,
that
is,
to
stay
at
the
39
See
UARG
Comments
[
II­
D­
303]
at
39.
40
See
Ralph
L.
Roberson
&
Richard
D.
McRanie,
Thoughts
on
Power
Plant
Efficiency,
at
7
(
Attachment
F
to
UARG
Comments
[
II­
D­
303])
(
RMB
Report);
see
also
Class
of
`
85
Group
Comments
[
II­
D­
268]
at
5­
6
(
noting
that
utilization
of
base­
loaded
units
displaces
less
efficient,
more
polluting
plants).
41
First
Energy
Comments
[
II­
D­
261]
at
1.

14
current
levels
of
available
supply.
39
Maximizing
the
utilization
of
existing
generation
capacity
can
be
critical
to
ensuring
the
ability
of
utilities
to
meet
consumer
demand
in
peak
periods.

Third,
according
to
industry
commenters,
a
company
could
simply
choose
not
to
undertake
the
needed
maintenance,
repair
and
replacement
projects
in
question,
so
as
to
avoid
triggering
NSR.
They
believe
this
would
result
in
a
loss
of
electricity
generating
capacity,
because
delayed
and
foregone
maintenance
leads
to
a
decrease
in
availability
and
reliability.

In
addition,
commenters
suggest
that
such
a
decrease
also
could
have
a
negative
impact
on
the
energy
efficiency
of
the
unit
and
the
overall
efficiency
of
a
utility
system.
This
is
because,
if
a
larger
utility
unit
becomes
unavailable
during
a
period
when
it
would
have
been
utilized
to
meet
consumer
demand,
then
multiple
smaller,
less
efficient
units
often
must
be
utilized
in
its
place.
40
One
utility
commented
that
only
through
maintenance
of
highly
efficient
low­
cost
baseline
generation
is
the
retirement
of
more
inefficient
units
possible.
41
The
commenter
asserted
that
less
efficient
units
are
more
costly
to
operate
and
generally
produce
more
pollution
per
unit
of
electric
output.

EPA
notes
that
the
possible
energy
impacts
predicted
by
industry
commenters
appear
to
flow
from
the
industry's
reported
uncertainty
regarding
the
scope
of
the
routine
maintenance
exclusion.
Consistent
with
our
conclusion
in
the
previous
section
of
this
report,
we
conclude
that
concern
about
the
scope
of
the
routine
maintenance
exclusion
is
having
an
adverse
impact
on
projects
that
would
improve
the
reliability
and
availability
of
existing
electric
generating
facilities.
We
also
note
that,
when
catastrophic
forced
outages
have
occurred
in
the
past,
the
Agency
has
consistently
worked
with
industry
and
State
and
local
permitting
authorities
to
allow
the
facility
to
get
the
unit
back
and
running
quickly.

B.
Impact
on
Efficiency
Improvement
Projects
(
i)
NSR
Applicability
With
respect
to
the
issue
of
energy
efficiency,
a
significant
number
of
industry
commenters
stated
that
an
inappropriately
narrow
routine
maintenance,
repair
and
replacement
exclusion
would
prevent
electricity
generators
from
taking
advantage
of
opportunities
to
improve
their
generating
efficiency.
One
measure
of
such
efficiency
is
"
heat
rate,"
or
the
amount
of
fuel­
bound
energy
required
to
produce
a
unit
of
electrical
power
(
typically
expressed
in
million
BTU
per
kWhr
Improving
an
electric
unit's
efficiency
 
e.
g.,
its
heat
rate
­­
means
that
less
fuel
is
required
to
produce
the
same
amount
of
electrical
power,
reducing
pollution
per
unit
of
production
output.
Alternatively,
improved
efficiency
may
allow
a
unit
to
produce
more
electricity
for
the
same
amount
of
fuel
burned
(
i.
e.,
with
no
greater
amount
of
emissions).
New
electric
generation
42
EPA
Background
Paper
at
28.
43
RMB
Report
at
6
(
Attachment
F
to
UARG
Comments
[
II­
D­
303]).
44
Industry
commenters
state
that
most
energy
efficiency
improvements
can
be
linked
with
tangible
benefits
to
the
environment
and
that
unless
the
power
source
is
in
close
proximity
to
the
process
in
which
energy
efficiency
is
improved,
the
emissions
benefits
are
not
necessarily
local.
If
the
power
source
is
a
grid,
it
may
not
be
possible
to
predict
where
all
the
benefits
will
occur,
nor
what
their
magnitude
would
be.
Nevertheless,
commenters
believe
that
energy
efficiency
should
be
an
important
aspect
of
meeting
national
air
pollution
goals
because
the
energy
saved
is
energy
that
would
have
otherwise
been
generated.
45
EPA
Background
Paper
at
28
(
citing
Detroit
Edison
Applicability
Determination,
May
23,
2000).
46
Under
EPA's
"
WEPCO
rule,"
NSR
is
not
triggered
for
existing
utility
sources
unless
there
is
a
significant
net
increase
in
actual
emissions
using
an
actual
to
predicted
future
actual
methodology.

15
technologies
often
lead
to
energy
efficiency
improvements,
but
industry
raised
concerns
that
applying
these
new
technologies
(
i.
e.,
replacing
boiler
or
turbine
components
with
components
of
better
design
and
materials)
often
could
trigger
NSR
 
in
some
cases
even
if
the
unit's
emissions
rate
does
not
increase
 
because
the
source
uses
the
more
efficient
unit
more
than
it
used
the
old
one.

These
commenters
stated
that
the
turbine
blade
project
that
was
the
subject
of
the
Detroit
Edison
applicability
determination
is
a
good
example
of
such
a
project.
42
Industry
reports
that,
under
a
voluntary
self­
reporting
program
initiated
by
the
Energy
Information
Administration
(
EIA),
utilities
have
reported
numerous
projects
that
are
expected
to
increase
efficiency.
43
Commenters
cited
as
examples
projects
ranging
from
load
optimization
programs
and
improved
boiler
controls
to
replacing
turbine
blades
and
rotors,
to
upgrades
or
replacements
of
components
like
superheaters
and
condensers.
44
Industry
commenters
noted
that
EPA
views
such
energy
efficiency
projects
as
the
Detroit
Edison
turbine
blade
upgrade
as
"
markedly
different
from
the
frequent,
inexpensive,
necessary,
and
incremental
maintenance
and
replacement"
of
deteriorated
components
and,
therefore,
not
within
the
scope
of
the
routine
maintenance
exclusion.
45
Industry
commenters
expressed
concern
that
this
could
result
in
the
discouragement
of
energy
efficiency
improvements
because
they
could
be
subject
to
NSR.
For
utilities,
this
is
a
particular
concern
in
any
jurisdiction
that
has
not
incorporated
the
WEPCO
rule
emission
increase
methodology
because
the
"
actual­
to­
potential"
test
applies
in
these
jurisdictions.
46
In
non­
WEPCO
jurisdictions,
and
in
all
jurisdictions
for
nonutility
activities,
industry
commenters
said
that
NSR
could
apply
to
any
project
that
both
corrects
availability/
reliability
problems
and
improves
efficiency
(
because
of
the
belief
that
any
project
that
corrects
availability/
reliability
problems
could
result
in
an
emissions
increase
under
the
actual­
topotential
test),
and
to
any
efficiency
improvement
project
at
a
unit
that
is
not
at
the
very
top
of
a
system's
loading
order.
Even
for
units
that
are
at
the
top
of
the
loading
order
of
a
particular
system,
like
Detroit
Edison's
Monroe
units,
industry
commenters
expressed
concern
about
whether
any
efficiency
improvement
could
be
shown
not
to
increase
emissions,
because
an
efficiency
improvement
almost
always
makes
the
improved
unit
more
attractive
to
run.
47
See,
e.
g.,
Class
of
`
85
Group
Comments
[
II­
D­
268]
at
5;
UARG
Comments
[
II­
D­
303]
at
45.
48
See
Comments
of
Xcel
Energy
[
II­
D­
213]
at
6­
7.
49
EPA
Background
Paper
at
28.
50
See
Class
of
`
85
Group
Comments
[
II­
D­
268]
at
5­
6;
see
also
FirstEnergy
Comments
[
II­
D­
261]
at
1­
2.

16
Utility
commenters
stated
that
the
Detroit
Edison
applicability
determination
discourages
utilities
from
undertaking
efficiency
improvement
projects.
47
They
suggested
that
utilities
are
likely
to
forego
efficiency
improvements
in
order
to
avoid
the
uncertainty,
delays
and
potential
costs
associated
with
NSR
applicability.
One
commenter
sought
to
illustrate
this
point
in
responding
to
the
EPA
Background
Paper's
inquiry
regarding
whether
NSR
applicability
alters
the
economics
of
efficiency
improvement
projects
by
evaluating
a
typical
turbine
efficiency
improvement
project.
This
evaluation
showed
that
such
a
project
would
cost
approximately
$
937,000
for
a
250
MW
unit,
and
would
be
expected
to
yield
additional
revenues
of
$
21.5
million
(
present
value).
For
such
a
unit,
however,
the
commenter
determined
that
NSR
applicability
would
result
in
expensive
retrofits,
with
a
capital
cost
(
i.
e.,
excluding
operation
and
maintenance
of
the
retrofits)
approximating
$
68.4
million.
48
Industry
commenters
said
that
discouraging
efficiency
improvement
projects
also
results
in
more
emissions
than
if
the
projects
could
go
forward
without
NSR.
They
argue
that,
on
a
megawatt
basis,
efficiency
improvements
reduce
pollution,
49
and
that,
even
if
utilization
increases
at
the
unit
with
improved
efficiency,
the
dynamics
of
economic
dispatch
of
electric
generating
units
mean
that
the
increased
utilization
at
that
unit
necessarily
displaces
less
efficient,
and
therefore
more­
polluting,
plants.
50
Thus,
the
industry
concludes
that
discouraging
efficiency
improvements
almost
always
results
in
higher
emissions
than
if
these
improvements
had
been
made.
As
an
example,
the
Detroit
Edison
case
was
again
cited,
where
the
use
of
the
more
efficient
blades
would
have
permitted
each
generating
unit
to
produce
the
same
amount
of
electricity
as
it
had
in
1994
while
burning
112,635
fewer
tons
of
coal.
The
result,
according
to
commenters,
would
have
been
a
reduction
of
1,826
tons
per
year
(
tpy)
in
SO
2
emissions,
1,402
tpy
in
NO
X
emissions,
and
259,111
tpy
in
carbon
dioxide
(
CO
2
)
emissions,
assuming
that
input
design
parameters
(
maximum
heat
input
and
fuel
consumption
specifications)
remained
the
same.
Detroit
Edison
estimated
that
more
than
1,000
other
electric
utility
units
in
the
United
States
have
the
capability
to
achieve
similar
reductions
through
similar
turbine
blade
replacements
and
other
projects;
thus,
extrapolating
based
upon
these
estimates,
they
predict
that
by
encouraging
the
adoption
of
blading
efficiency
improvements,
CO
2
emissions
would
be
reduced
by
81
million
tons
per
year
or
more,
provided
input
design
parameters
(
maximum
heat
input
and
fuel
consumption
specifications)
remained
the
same.
They
predict
that
SO
2
and
NO
X
emissions
would
also
be
reduced
significantly.

In
contrast,
commenters
from
environmental
groups
believe
that
NSR
treats
energy
efficiency
improvement
projects
appropriately.
They
stated
that
NSR
only
applies
when
a
project
results
in
an
emissions
increase
and
that
the
types
of
projects
discussed
above
where
significant
reductions
are
achieved
would
not
trigger
NSR.
However,
if
an
energy
efficiency
project
also
results
in
a
significant
emissions
increase,
these
commenters
felt
that
it
would
be
inappropriate
to
51
See,
e.
g.,
July
20
testimony
of
John
Walke,
NRDC.
52
NRDC
Comments
[
II­
D­
267].
53
This
was
the
case
in
Detroit
Edison,
where
there
was
no
expected
increase
and
therefore
the
proposed
project
did
not
trigger
NSR.
[
See
Detroit
Edison
Applicability
Determination]

17
exempt
the
increase
from
review
under
NSR.
51
One
commenter
also
questioned
whether
NSR
is
the
predominant
factor
in
influencing
a
decision
about
whether
to
proceed
with
an
efficiency
project,
noting
that
some
analysts
believe
that
the
regulation
of
utility
rates
 
and
specifically
their
treatment
of
cost
recovery
 
has
lessened
the
incentive
for
heat
rate
improvements.
52
In
reviewing
the
information
regarding
energy
efficiency
projects,
the
EPA
concludes
that
NSR
may
discourage
some
energy
efficiency
improvements.
EPA
notes
that
as
long
as
utilization
remains
constant,
energy
efficiency
improvements
can
result
in
significant
emissions
reductions.
Such
projects
would
not
trigger
NSR
if
there
were
not
a
significant
emissions
increase.
53
Because
such
projects
are
not
subject
to
the
NSR
regulations,
NSR
generally
has
a
negligible
impact
in
such
cases.
However,
as
noted
above,
energy
efficiency
improvements
are
often
associated
with
increases
in
utilization,
because
the
more
efficient
generating
units
are
dispatched
more
often.
Efficiency
improvements
can
also
result
in
an
increase
in
capacity
or
availability.
In
such
cases,
there
can
be
local
emissions
increases
that
trigger
NSR
if
the
projects
are
not
routine
maintenance.
For
example,
in
Detroit
Edison,
if
a
five
percent
increase
in
operation
were
to
result,
actual
increases
on
the
order
of
800
tons
of
NO
X
and
2000
tons
of
SO
2
would
occur.
Even
if
these
emissions
increases
occur
at
the
same
time
as
emissions
decrease
somewhere
else,
some
commenters
expressed
concerns
about
the
localized
impacts
of
potentially
large
emissions
increases,
and
felt
that
review
under
NSR
was
needed
to
address
them.

Congress
provided
that
where
physical
changes
at
a
plant
result
in
significant
increases
in
air
pollution,
these
plants
should
go
through
NSR
and
take
steps
to
control
emissions.
Even
if
a
physical
change
is
relatively
inexpensive
when
compared
to
the
cost
of
the
controls
that
are
projected
to
result
from
NSR,
the
change
could
still
result
in
emissions
increases
that
Congress
believed
should
undergo
review.
However,
as
noted
in
the
example
turbine
efficiency
improvement
project
above,
and
echoed
throughout
many
comments,
the
costs
associated
with
NSR,
particularly
the
costs
to
retrofit
pollution
controls,
can
render
these
projects
uneconomical.
Thus,
the
EPA
finds
that
NSR
discourages
some
types
of
energy
efficiency
improvements
when
the
benefit
to
the
company
of
performing
such
improvements
is
outweighed
by
the
costs
to
retrofit
pollution
controls
or
to
take
measures
necessary
to
avoid
a
significant
net
emissions
increase.
The
EPA
recognizes
the
need
to
promote
the
development
of
efficient
and
more
environmentally
friendly
designs.

On
the
other
hand,
it
is
also
clear
that
a
wide
range
of
activities
at
an
electric
utility
can
have
energy
efficiency
benefits,
from
everyday
maintenance
to
major
capital
projects.
In
general,
the
EPA
encourages
efficiency
improvements
wherever
feasible.
However,
the
scope
and
magnitude
of
some
of
the
kinds
of
changes,
their
impact
on
recovering
capacity
that
had
been
lost
to
deterioration
of
equipment,
their
impact
on
significantly
extending
the
life
of
the
boiler,
turbine,
etc.,
and
the
resulting
significant
emissions
increase,
necessitates
that
certain
projects
which
may
result
in
efficiency
improvements,
must
be
reviewed
under
NSR.
Though
projects
of
this
54
Clean
Air
Task
Force
comments
[
II­
D­
236]
at
49
and
App.
C.

18
magnitude
still
may
go
forward
once
their
air
quality
impacts
are
addressed,
the
EPA
finds
that
NSR
can
discourage
companies
from
undertaking
them.

(
ii)
Energy
Impacts
The
ICF
report
in
support
of
the
EPA
Background
Paper
referred
to
various
data,
such
as
those
of
the
National
Coal
Council
(
NCC)
May
2001
report,
which
estimate
that
repairs
and
replacements
that
improve
efficiency
at
existing
coal­
fired
facilities
could
result
in
an
increase
in
capacity
of
5%
to
10%.
Applied
across
the
entire
coal­
fired
electric
generation
capacity
of
the
United
States
(
over
300
GW)
this
would
result
in
an
additional
capacity
of
15,000­
30,000
MW.
This
is
the
equivalent
to
30­
60
new
500
MW
plants
or
enough
power
for
10­
20
million
homes.

Similarly,
as
noted
in
the
EPA
Background
Paper,
the
NCC
report
found
that
coal­
fired
units
over
20
years
of
age
had
been
substantially
derated,
and
concluded
that:
"
If
all
existing
conditions
resulting
in
a
derating
could
be
addressed,
approximately
20,000
MWs
of
increased
capacity
could
be
obtained
from
regaining
lost
capacity
due
to
unit
deratings."
Likewise,
the
NCC
reported
that
20,000
MW
of
additional
capacity
could
be
gained
by
"
increasing
heat
input
and/
or
electrical
output
from
[
existing]
generating
equipment."
Moreover,
the
NCC
found
that
this
restoration
and
increase
of
capacity
from
existing
units
could
only
be
economically
viably
pursued
by
the
facility
owners
if,
among
other
factors,
the
increased
availability
and/
or
electrical
output
would
clearly
not
trigger
NSR.
Other
industry
representatives
supported
this
estimate.

Conversely,
environmental
group
commenters
expressed
the
view
that
such
investments
are
not
as
profitable
as
investments
in
completely
new
electric
generation
capacity
and
that
this
is
why
the
industry
is
not
pursuing
them,
as
opposed
to
NSR
being
the
major
impediment.
54
They
also
estimate
that
the
emissions
reductions
from
efficiency
improvement
projects
would
be
small
compared
to
the
reductions
that
would
be
achieved
if
NSR
applied.

In
conclusion,
for
the
utility
industry,
with
respect
to
existing
sources,
and
in
contrast
to
new
sources,
the
EPA
finds
that
the
available
information
indicates
that
the
NSR
program
is
having
an
adverse
impact
on
investment
in
both
electric
generation
capacity
and
energy
efficiency.
While
there
are
only
limited
data
that
prove
that
NSR
has
resulted
in
the
cancellation
of
otherwise
economical
projects
of
either
type,
a
significant
number
of
industry
commenters
presented
a
variety
of
projects
at
existing
sources
that
could
have
increased
capacity,
improved
reliability,
or
enhanced
efficiency,
but
were
made
uneconomical
due
to
delays
and
costs
associated
with
NSR.
The
EPA
finds
many
of
these
cases
to
be
credible
and
based
on
real­
world
examples,
and
believes
that
they
demonstrate
that
NSR
has
an
adverse
impact
on
such
investment
at
existing
sources.
It
is
reasonable
to
conclude
that
the
foregone
investment
has
resulted
in
foregone
capacity
increases
through
decreased
reliability
and
availability
that
are
not
recovered,
and
through
foregone
efficiency
improvements.

2.
Refineries
55
API
Comments
[
II­
D­
134]
at
1­
2.
56
See
NPRA
Comments
[
II­
E­
27]
at
2.
57
See
BP
America
comments
[
II­
D­
307]
at
2.
58
NPRA
letter
to
Stephanie
Daigle,
EPA,
7/
23/
2001.

19
Turning
to
the
question
of
NSR
impacts
on
investment
in
capacity
at
existing
refineries,
the
EPA
finds
that
the
comments
again
highlight
areas
where
NSR
may
adversely
impact
investment
in
capacity
and
energy
efficiency
projects.
These
areas
are
examined
further
in
this
section
in
order
to
assess
their
nature
and
extent.

Refinery
commenters
observe
that
the
refining
industry
differs
considerably
from
the
electric
utility
industry
in
several
respects.
For
example,
it
is
operating
much
closer
to
full
capacity
than
the
utility
industry,
and
it
is
not
transitioning
from
an
economically
regulated
basis
to
a
market
basis.
Even
while
operating
at
very
high
utilization
rates,
commenters
noted
that
the
industry
must
be
able
to
respond
rapidly
to
changes
in
raw
material
availability,
market
demands,
and
environmental
requirements.
API
explained
that,
"[
r]
efiners
are
required
by
law
to
make
adjustments
to
fuel
specifications
from
one
season
to
another,
produce
fuels
meeting
multiple
specifications
in
various
regions
of
the
country,
and
reconfigure
to
refine
cleaner
burning
low
sulfur
diesel
and
gasoline,
all
while
being
able
to
supply
fuels
to
meet
constantly
changing
customer
demand."
55
API
suggested
that
these
requirements
necessitate
frequent
and
rapid
responses
that
may
involve
changes
to
a
refinery's
facilities
and
processes.
Moreover,
they
note
that,
to
meet
demand
for
petroleum
products
and
avoid
market
disruptions
that
can
lead
to
shortages
and
price
volatility,
the
refining
industry
must
be
able
to
maintain
the
availability,
reliability,
and
safety
of
its
facilities.
NPRA's
comments
noted,
"
Refining
operations
are
continuous
and
complex.
They
depend
on
the
simultaneous
operation
of
many
individual,
but
inter­
related,
pieces
of
equipment
("
units").
A
delay
or
inability
to
change
or
improve
operations
of
a
single
unit
can
have
a
significant
cumulative
impact
on
the
refinery's
ability
to
produce
the
fuels
that
its
customers,
and
the
national
economy,
rely
upon."
56
To
meet
increasing
demand
without
major
construction
of
new
refining
facilities,
commenters
believe
that
the
industry
must
improve
the
efficiency
of
its
existing
facilities,
and
it
must
engage
in
what
one
industry
commenter
described
as
a
"
continuous
incremental
improvement
in
production
capacity."
57
Finally,
as
noted
in
the
Background
Paper,
and
above,
with
no
new
refineries
likely
to
be
built
in
the
near
future,
assessing
the
impact
of
NSR
on
existing
sources
is
particularly
critical.

As
with
utilities,
refineries
maintain
that
the
exclusion
for
"
routine
maintenance
repair
and
replacement"
has
been
narrowed
by
EPA
in
recent
years
and
undercuts
their
ability
to
respond
quickly
to
market
changes
and
raw
material
availability.
In
addition,
refinery
industry
commenters
expressed
concern
about
the
test
used
to
determine
whether
a
change
results
in
an
emissions
increase
at
non­
utility
source
categories
(
i.
e.,
the
"
actual
to
potential"
test).
In
the
view
of
many
refinery
commenters,
the
NSR
program
has
the
effect
of
constraining
the
industry's
ability
to
(
1)
expand
domestic
refining
capacity,
(
2)
increase
the
supply
of
cleaner
burning
fuels,
and
(
3)
enhance
energy
efficiency.
58
The
commenters
said
that
under
the
NSR
program,
numerous
common
activities
at
a
refinery
 
whether
required
to
respond
to
demand
changes,
to
repair
or
replace
a
broken
piece
of
equipment,
to
improve
efficiency,
to
expand
refining
capacity,
or
even
59
See
API
Comments
[
II­
D­
134]
at
2;
ExxonMobil
Comments
[
II­
D­
418]
at
2;
NPRA
Comments
[
II­
E­
27]
at
3.
60
See
Marathon
Ashland
Petroleum
LLC
(
MAP)
Comments
[
II­
D­
253]
at
2.
61
See,
e.
g.,
ExxonMobil
Comments
[
II­
D­
418]
at
11(
commenting
that
actual­
to­
potential
test
"
fabricate[
s]
emission
increases"
where
no
increases
actually
occur).
62
See
ExxonMobil
Comments
[
II­
D­
418]
at
12;
BP
America
Comments
[
II­
D­
307]
at
2;
MAP
Comments
[
II­
D­
253]
at
2.
63
See
BP
America
Comments
[
II­
D­
307]
at
2.
64
NPRA
Comments
[
II­
E­
27]
at
Attachment
1,
No.
1.
65
See,
e.
g.,
NPRA
comments
[
II­
D­
400]
and
API
comments
[
II­
D­
134].
66
NPRA
Comments
[
II­
E­
27]
at
2.

20
to
respond
to
environmental
requirements
 
are
potentially
subject
to
NSR
permitting.
59
One
industry
commenter
states
that
hundreds
of
such
activities
are
undertaken
each
year
at
existing
U.
S.
refineries.
60
According
to
commenters,
the
lengthy,
costly,
and
uncertain
nature
of
the
current
NSR
permitting
process
discourages
those
activities
to
which
it
potentially
applies,
or
at
least
introduces
significant
delays
in
and
constraints
on
the
ability
of
the
operator
to
make
the
required
changes
in
an
efficient
and
timely
manner.

Refining
industry
commenters
also
noted
that,
in
their
opinion,
the
NSR
emissions
increase
test
for
non­
utilities
(
the
"
actual­
to­
potential"
comparison)
presumes
that
virtually
any
activity
at
a
refinery
increases
emissions
within
the
meaning
of
NSR,
even
if
the
activity
were,
in
fact,
to
result
in
decreased
actual
emissions.
61
Thus,
these
commenters
stated
that,
of
the
activities
undertaken
at
a
given
refinery,
only
those
activities
ultimately
deemed
to
constitute
"
routine
maintenance,
repair
or
replacement"
might
avoid
NSR.
However,
according
to
industry
commenters,
few
activities
beyond
the
most
mundane
maintenance
activities
that
may
be
undertaken
each
year
at
a
given
facility
would
be
deemed
"
routine"
under
the
NSR
regulations.
62
One
commenter
maintained
that
the
NSR
program
would
apply
NSR
to
any
change
that:
(
a)
results
in
an
increase
in
capacity
or
capacity
utilization
of
an
existing
process
unit;
or
(
b)
increases
the
efficiency
or
lowers
the
unit
operating
costs;
or
(
c)
extends
the
useful
life
of
that
unit
..."[
or
(
d)]
increase[
s]
unit
reliability."
63
According
to
industry,
these
are
precisely
the
types
of
activities
that
U.
S.
refineries
must
constantly
undertake
to
meet
demand
and
minimize
fuel
supply
disruptions
and
price
volatility.
Moreover,
commenters
suggest
that
the
use
of
an
actualto
potential
test
encourages
industry
to
maximize
current
actual
emissions
within
permit
limits,
rather
than
providing
incentives
for
emissions
reductions.
64
Industry
commenters
provided
a
list
of
activities
that
they
reportedly
undertake
to
maintain
reliability,
improve
efficiency,
and
expand
capacity
that,
in
their
view,
are
typically
undertaken
in
the
industry
but,
nevertheless,
are
potentially
subject
to
NSR
under
the
current
program.
65
According
to
industry,
the
potential
applicability
of
NSR,
which
they
believe
could
encompass
virtually
any
given
project,
tends
to
discourage
operators
from
undertaking
particular
projects
because
NSR
would
add
significant
delays
and
costs.
66
Industry
commenters
observed
that
the
EPA
Background
Paper's
estimate
for
the
length
of
time
typically
necessary
to
obtain
an
NSR
permit
did
not
include
the
time
spent
prior
to
submittal
of
a
complete
application.
If
such
time
is
included,
the
length
of
the
NSR
permitting
process
in
the
experience
of
refinery
67
See
API
comments
[
II­
D­
134]
at
8.
68
See
ExxonMobil
Comments
[
II­
D­
418]
at
16.
69
See
BP
America
Comments
[
II­
D­
307]
at
2.
70
See
id.;
see
also
ExxonMobil
Comments
[
II­
D­
418]
at
18
(
noting
both
the
cost
and
scheduling
impacts
of
NSR
on
project
economics).
71
See
BP
America
Comments
[
II­
D­
307]
at
2­
3.
72
See
API
Comments
[
II­
D­
134]
at
2.
73
NRDC
comments
[
II­
D­
267]
at
5.

21
commenters
is
at
least
7
to
22
months,
excluding
any
post­
issuance
appeals
and
challenges.
67
An
industry
commenter
further
predicted
that,
if
the
listed
activities
are
viewed
as
non­
routine,
the
refining
industry,
as
well
as
other
U.
S.
industries,
would
experience
much
longer
lead
times
in
obtaining
NSR
permits
than
already
occur.
68
Like
utilities,
refiners
also
raised
the
concern
that
there
would
be
limited
options
for
projects
that
are
potentially
subject
to
NSR.
69
They
described
three
options.
First,
the
operator
could
seek
to
obtain
an
NSR
permit,
accepting
the
delays,
uncertainties,
and
potentially
significant
costs
that
commenters
say
are
associated
with
such
permits.
70
Alternatively,
an
operator
could
seek
to
"
avoid"
NSR
by
limiting
emissions
to
past,
actual
levels
through
a
minor
NSR
permit
(
a
permit
which,
according
to
industry,
can
take
3­
12
months
to
obtain),
thus
giving
up
refinery
capacity
and
"
deprive[
ing]
the
source
of
the
`
headspace'
between
actual
and
allowable
emissions
that
is
crucial
to
long­
term
operating
flexibility
and
the
ability
to
respond
quickly
to
changes
in
demand."
71
A
third
option
would
be
to
simply
cancel
the
project,
and
forego
the
projected
benefit
that
was
the
reason
for
the
project
in
the
first
place.

Overall,
the
comments
submitted
by
refinery
and
other
commenters
during
this
review
process
emphasize
their
belief
that
by
imposing
significant
costs
and
delays,
the
NSR
program
discourages
investment
in
projects
that
are
necessary
to
maintain
the
reliability
of
existing
refineries,
improve
their
efficiency,
expand
capacity,
and
respond
flexibly
to
rapidly
changing
consumer
demand
for
petroleum
products.
According
to
one
commenter,
what
the
industry
most
needs
is
certainty
and
flexibility
in
its
efforts
to
meet
both
the
energy
needs
of
the
Nation
and
environmental
requirements.
72
In
contrast,
NRDC's
comments
suggest
that
poor
return
on
investment
is
more
important
than
environmental
considerations
(
of
which
NSR
is
only
a
small
part,
and
is
not
specifically
named
by
sources
examined
in
the
EPA
Background
Paper)
in
any
decisions
not
to
invest
in
new
capacity.
73
They
point
to
information
presented
in
the
Background
Paper
showing
that,
in
recent
years,
there
has
been
significant
investment
in
refinery
capacity
at
existing
sources.

As
discussed
above
for
utilities,
the
EPA
notes
that
for
refineries
there
are
also
differences
of
opinion
amongst
the
commenters
about
the
scope
of
the
routine
maintenance
exclusion
and
the
resulting
impacts.
In
determining
whether
an
activity
is
"
routine"
for
purposes
of
being
excluded
from
NSR,
EPA
consistently
has
taken
a
case­
by­
case
approach,
weighing
the
nature,
extent,
purpose,
frequency
and
cost
of
the
work,
as
well
as
other
relevant
factors.
However,
EPA
acknowledges,
as
it
did
for
utilities,
that
the
comments
report
significant
uncertainty
about
the
scope
of
the
"
routine"
exemption.
Such
uncertainty
can
result
in
the
delay
or
cancellation
of
22
projects.
Changes
to
the
NSR
program
that
add
to
the
clarity
and
certainty
of
the
scope
of
the
routine
maintenance
exclusion
will
improve
the
process
by
reducing
the
unintended
consequences
of
discouraging
worthwhile
projects
that
are
in
fact
outside
the
scope
of
NSR.

A
key
difference
between
utilities
and
refineries
is
the
fact
that
refineries
use
the
"
actualto
potential
test"
for
determining
NSR
applicability,
while
utilities
generally
do
not.
The
EPA
has
reviewed
a
number
of
examples
where
projects
could
have
provided
capacity
increases
or
energy
efficiency
improvements,
and
likely
could
have
done
so
without
increasing
actual
emissions,
and
in
some
cases
the
projects
appear
likely
to
decrease
actual
emissions.
Such
projects,
if
they
occur
at
units
operating
below
capacity,
could
trigger
NSR
or,
at
least,
trigger
a
need
to
cap
the
units
below
capacity
or
install
pollution
controls
to
avoid
NSR.
Again,
the
determination
of
whether
a
change
results
in
an
emissions
increase
is
a
case­
by­
case
determination,
but
the
EPA
believes
that
the
commenters'
examples
make
a
credible
case
that
some
capacity
or
efficiency
projects
that
do
not
increase
actual
emissions
are
not
undertaken
because
they
trigger
NSR
under
the
actual­
topotential
test.
Although
the
information
is
mostly
anecdotal
in
nature,
the
EPA
believes
that
the
information
presented
is
based
on
real
world
experience,
and
makes
a
credible
case
that
some
projects
are
not
going
forward
in
part
because
of
NSR.
The
EPA
believes
that
this
results
in
lost
refining
capacity,
or
foregone
opportunities
to
increase
capacity
without
increasing
emissions.

IV.
Impact
on
Industries
Other
than
Electric
Utilities
and
Petroleum
Refineries
In
addition
to
the
information
supplied
to
EPA
by
utility
and
refinery
commenters,
the
Agency
received
numerous
comments
from
other
industries
regarding
the
NSR
program's
impact
on
energy
use,
efficiency,
and
capacity.
These
comments
came
principally
from
a
variety
of
industry
associations
and
coalitions
of
manufacturers
representing
the
automobile,
aerospace,
chemical,
electronics,
food,
aluminum
and
steel,
packaging,
paper,
printing,
pharmaceutical,
and
other
manufacturing
sectors.
Like
the
utility
and
refining
industries,
these
commenters
were
primarily
concerned
with
the
current
application
of
the
NSR
program
to
existing
sources.
They
noted
many
anecdotal
instances
where
projects
would
have
reduced
energy
demand
and/
or
increased
energy
efficiency,
but
were
abandoned
because
of
NSR
permitting
delays
and/
or
costs
associated
with
the
retrofit
of
existing
equipment
with
the
BACT
or
LAER
emissions
controls
mandated
by
NSR
rules.
Other
commenters
presented
similar
examples
of
pollution
control
and
pollution
prevention
projects
abandoned
because
of
potential
NSR
applicability.
According
to
the
commenters,
the
cancellation
of
projects
that
would
have
improved
energy
efficiency
or
decreased
pollution
means
that
NSR
is
having
an
adverse
impact
on
investment
in
both
energy
efficiency
and
environmental
protection.

Among
the
general
concerns
voiced
by
commenters
in
addition
to
pollution
control
costs
were
claims
that
(
1)
the
NSR
program
is
complex
and
gives
rise
to
uncertainty
and
associated
delays,
(
2)
it
hinders
flexibility
for
industry
to
quickly
make
needed
changes,
and
(
3)
that
it
results
in
the
loss
of
production
capacity
where
NSR
is
triggered
based
on
the
application
of
the
actualto
potential
test,
even
if
emissions
will
not
actually
increase.
Furthermore,
commenters
argued
that
if
a
source
wants
to
avoid
NSR,
it
faces
the
undesirable
outcome
of
accepting
new
emissions
74
See,
e.
g.,
Comments
of
NEDA/
CARP
[
II­
D­
272]
at
9­
10.
75
See,
e.
g.,
FPA
Comments
[
II­
D­
271]
at
2­
3.
76
NEDA/
CARP
Comments
[
II­
D­
272]
Attachment
A,
Example
#
1.
77
NEDA/
CARP
Comments
[
II­
D­
272]
Attachment
A,
Example
#
4.
According
to
this
example,
only
2
tons
per
year
of
regulated
emissions
would
have
resulted
from
the
change,
but
potential
emissions
could
have
increased
over
100
TPY
of
VOC
because
operation
of
an
incinerator
with
a
98%
control
efficiency
voluntarily
installed
by
the
company
is
not
considered
to
be
"
federally
enforceable."

23
limits
in
the
NSR
permit
that,
according
to
commenters,
effectively
reduce
a
plant
or
unit's
productive
capacity.
74
A.
NSR
Applicability
1.
Routine
Maintenance,
Repair
&
Replacement
As
with
utilities
and
refineries,
many
commenters
from
other
industry
sectors
focused
on
the
NSR
"
routine
maintenance,
repair
and
replacement"
exclusion.
Like
the
industries
discussed
above,
they
believe
that
EPA
has
narrowed
the
exclusion
in
recent
years.
Thus,
they
stated
this
was
the
day­
to­
day
largest
problem
in
maintaining
the
availability,
reliability,
and
safety
of
production
equipment.
75
In
particular,
commenters
asserted
that
projects
involving
repair
or
replacement
components
incorporating
"
state­
of­
the­
art"
improvements
in
materials
or
design
may
be
subject
to
NSR
since
they
may
not
qualify
as
routine
maintenance,
or
may
result
in
more
efficient
utilization
of
fuel
and/
or
raw
materials
that
may
potentially
increase
a
facility's
emissions.
For
instance,
at
one
plant,
a
company
states
that
it
elected
not
to
replace
spray
nozzles
in
a
process
dryer,
even
though
it
determined
that
significant
energy
savings
could
result,
because
it
concluded
that
the
new
Teflon
coated
nozzles
would
not
be
equivalent
parts
and,
therefore,
the
project
would
not
be
exempt
from
NSR
as
routine.
According
to
the
commenter,
the
new
nozzles
would
have
resolved
the
repeated
need
to
replace
the
existing
equipment,
and
may
have
provided
a
safer
and
more
reliable
operating
environment.
76
Similarly,
commenters
complained
that
NSR
application
discouraged
engineering
design
innovations
that
provide
better
quality
and
control
assurances
during
sometimes­
dangerous
production
processes.
One
example,
provided
by
the
chemical
industry,
was
the
installation
of
a
temperature
regulating
system
on
a
thermal
jacket
around
a
dryer
that
is
equipped
with
a
heated
jacket
that
uses
a
temperature
control
system
in
the
jacket.
The
temperature
control
system
works
by
regulating
the
flow
of
steam
or
hot
liquids
similar
to
radiator
fluids
in
the
jacket
that
surrounds
the
dryer.
The
current
system
uses
an
older
design
and
is
relatively
ineffective
because
of
the
system's
wide
temperature
variation,
which
causes
risks
of
explosion
and
lengthens
the
drying
process
time.
Both
problems
could
be
eliminated
with
the
installation
of
a
temperature
regulating
system,
which
would
also
reduce
energy
demands
on
the
process
by
20%.
Although
work
is
often
performed
on
the
jacket
regulating
system,
the
company
suggested
that
it
did
not
go
forward
with
the
change
because
work
on
the
temperature
regulating
system,
utilizing
a
unique
new
system,
would
not
be
considered
"
routine."
77
78
AFPA
Comments
[
II­
E­
15],
Tab
3,
Case
in
Point
#
4.
79
Comments
of
American
Chemistry
Council
[
II­
D­
416]
example
1.

24
It
was
also
suggested
that
application
of
the
NSR
program
impeded
the
ability
of
companies
to
undertake
projects
to
ensure
the
reliability
of
their
equipment
that
might
also
result
in
significant
energy
efficiency
gains.
Commenters
presented
a
number
of
examples
of
such
projects,
including
examples
from
the
chemical,
packaging,
aluminum
and
general
manufacturing
sectors.
One
illustration
from
the
American
Forest
and
Paper
Association
described
replacement
of
outdated
analog
controllers
at
a
series
of
six
batch
digesters.
The
original
controllers
were
no
longer
manufactured,
although
new
digital
controllers,
costing
approximately
$
50,000,
are
capable
of
receiving
inputs
from
the
digester
vessel
temperature,
pressure
and
chemical/
steam
flow.
The
new
controllers
would
have
more
precisely
filled
and
pressurized
digesters
with
chips,
chemicals
and
steam
(
whereas
the
old
controllers
added
materials
in
timed
sequence),
thus
bringing
a
batch
digester
on
line
faster.
However,
the
source
determined
that
under
the
NSR
program
this
project
would
not
be
considered
to
be
routine
because,
although
repairs
to
the
analog
system
might
have
been
frequent
at
the
company
involved,
replacement
of
the
system
with
a
digitalized,
computerized
system
would
not
qualify
as
"
routine."
78
As
with
utilities
and
refineries,
EPA
notes
that
there
are
widely
differing
views
on
the
scope
of
the
routine
maintenance
exclusion
on
other
industries.
As
before,
we
therefore
conclude
that
concern
about
the
scope
of
the
routine
maintenance
exclusion
is
having
an
adverse
impact
for
industries
outside
the
energy
sector.
It
also
is
credible
to
conclude
that
projects
have
been
discouraged
that
might
have
been
economically
and/
or
environmentally
beneficial
without
increasing
actual
emissions.
Changes
to
the
NSR
program
that
add
to
the
clarity
and
certainty
of
the
scope
of
the
routine
maintenance
exclusion
will
improve
the
process
by
reducing
the
unintended
consequences
of
discouraging
worthwhile
projects
that
are
in
fact
outside
the
scope
of
NSR.

2.
Pollution
Prevention
Projects
Another
series
of
examples
provided
by
commenters
from
the
manufacturing
sector
involved
pollution
prevention
projects,
many
with
significant
energy
savings
potential.
Pollution
prevention
projects
at
manufacturing
facilities
may
qualify
for
exemption
under
the
NSR
program.
This
determination
is
made
on
a
case­
by­
case
basis
under
EPA's
1994
guidance
which
addresses
pollution
control
projects
and
NSR
applicability.
Although
this
guidance
was
intended
to
create
incentives
for
industry
to
undertake
such
projects,
some
comments
suggested
that
it
might
actually
discourage
such
projects.
One
example
comes
from
the
chemical
industry.
In
that
case,
a
chemical
facility
considered
installation
of
a
new,
more
efficient
CFC
refrigeration
system.
Completion
of
this
project,
according
to
the
commenter,
would
have
resulted
in
decreased
CFC
emissions
and
less
electricity
demand,
reducing
overall
emissions
from
the
facility's
power
generating
plant.
However,
this
project
would
not
have
qualified
for
the
pollution
control
project
exclusion
because
the
primary
purpose
of
the
project
was
not
to
reduce
emissions.
Therefore,
because
the
project
otherwise
would
have
triggered
NSR,
the
company
elected
not
to
undertake
it.
79
80
NEDA/
CARP
Comments
[
II­
D­
272]
Attachment
A,
Example
#
14.
81
NEDA/
CARP
Comments
[
II­
D­
272]
Attachment
A,
Example
#
15.

25
In
a
second
example,
an
aerospace
company
suggested
that
it
was
unable
to
avoid
NSR,
using
EPA's
1994
pollution
control
project
policy,
because
the
purpose
of
a
particular
project
was
to
improve
energy
efficiency,
although
significant
pollution
control
benefits
would
also
have
resulted.
The
company
had
proposed
to
speed
up
its
manufacturing
process
(
for
parts
and
subassemblies)
by
using
a
new
adhesive
that
would
dry
(
or
cure)
faster.
The
company
stated
that
the
project
would
have
resulted
in
pollution
prevention
both
because
the
new
adhesive
had
a
lower
volatile
organic
compound
(
VOC)
content
than
the
one
in
use
and
because
more
parts
could
be
processed
in
less
time,
consuming
less
energy
overall.
However,
this
project
could
not
qualify
for
the
pollution
control
project
exclusion
because
its
purpose
was
to
improve
efficiency,
rather
than
to
abate
pollution
and
because
the
new
adhesive
system
would
have
increased
the
utilization
of
production
equipment
at
the
plant.
Because
the
project
otherwise
would
have
triggered
NSR
applicability,
the
company
declined
to
make
the
change.
80
EPA
believes
that
these
examples
indicate
that
NSR
is
having
an
adverse
impact
on
some
pollution
control
and
prevention
projects.

B.
Energy
Efficiency
The
Agency
also
received
a
number
of
industry
comments
explaining
the
NSR
program's
effect
on
energy
efficiency
and
demand.
These
comments
suggest
that
the
delays
and
costs
associated
with
NSR
have
discouraged
the
adoption
or
implementation
of
various
energy
conservation
and
efficiency
measures.
Examples
provided
by
commenters
included
efforts
to
conserve
fuel
and
programs
that
will
result
in
energy
demand
reductions
at
major
industrial
plants.
The
commenters
allege
that,
in
many
cases,
the
projects
would
ultimately
reduce
actual
emissions,
but
nonetheless
trigger
NSR
under
the
actual­
to­
potential
test.

For
instance,
NSR
was
cited
as
a
principal
reason
for
not
undertaking
energy
efficiency
projects
for
the
installation
of
heat
exchangers
and
overfire
air
by
various
manufacturing
sectors
including
the
electronics
and
appliance
industries,
plastics,
and
paper
industries.
Heat
exchangers
recover
heat
from
boiler
flue
gas
streams
to
heat
water
used
in
the
system's
deaerator
units.
By
preheating
the
water
used
in
the
deaerator
units,
the
heat
exchanger
reduces
the
steam
needed
to
run
the
deaerators.
This
increases
the
overall
efficiency
of
the
boiler
house
and
reduces
fuel
usage.
It
also
reduces
annual
boiler
emissions.
At
a
plastics
plant,
a
commenter
pointed
out
that
installation
of
a
heat
exchanger
would
be
expected
to
reduce
natural
gas
consumption
by
7.5
percent,
NO
X
emissions
by
7.5
percent,
SO
2
emissions
by
5.8
percent
and
carbon
monoxide
(
CO)
emissions
by
7.6
percent,
particulate
matter
(
PM)
emissions
by
9
percent,
and
VOC
emissions
by
9.3
percent.
The
project
achieves
these
benefits
through
pollution
prevention
rather
than
add­
on
controls.
81
In
this
case,
the
industry
applicant
sought
exclusion
from
NSR
applicability
under
the
pollution
control
project
exclusion.
However,
this
project
did
not
qualify
as
a
pollution
control
project
because
its
primary
purpose
was
not
pollution
control
or
prevention.
Moreover,
because
the
boilers
required
back­
up
firing
with
oil
during
the
winter
to
ensure
operation,
the
"
actual
to
potential"
emission
test
would
have
caused
the
project
to
trigger
NSR.
To
avoid
the
installation
of
new
controls
that
would
be
mandated
as
the
result
of
NSR
applicability,
the
source
states
that
82
AFPA
Comments
[
II­
E­
15],
Tab
3,
Case
in
Point
#
1.
83
FPA
Comments
[
II­
D­
271]
at
6­
7.

26
it
is
considering
burning
more
fuel
oil
over
the
next
two
years
to
increase
base
level
of
emissions
(
actual
emissions).

Another
example
from
a
boiler
at
a
pulp
and
paper
mill
illustrates
a
similar
problem.
According
to
the
comment,
the
mill's
industrial
boiler
currently
experiences
extensive,
internal
erosion
as
a
result
of
the
carryover
of
solids
such
as
sand
and
wire
from
the
burning
of
tire­
derived
fuel,
and
burned
bark
particles,
which
have
led
to
decreased
boiler
efficiency.
As
a
result,
the
mill
proposed
to
install
a
new
overfire
air
system
to
allow
for
more
complete
combustion
of
the
bark
fuel.
By
getting
more
heating
value
from
the
same
amount
of
bark
burned,
less
natural
gas
would
be
required
to
provide
supplemental
heat
at
an
annual
natural
gas
savings
of
about
$
1
million
(
in
July,
2001
dollars).
According
to
the
comment,
future
actual
emissions
of
NO
X
,
CO
and
VOCs
would
decrease
after
completion
of
this
project.
However,
because
the
boiler
is
currently
operating
below
its
rated
capacity,
the
potential
emissions
after
completion
of
the
project
would
increase
over
past
actual
emissions,
triggering
NSR.
The
commenter
estimates
that
the
cost
of
NSR
controls
would
be
$
17
million.
82
At
the
time
this
project
was
under
consideration,
the
relevant
company
estimated
that
the
annual
savings
in
natural
gas
usage
equated
to
roughly
200
million
cubic
feet
of
natural
gas.
This
amount
of
gas
has
a
heating
value
of
approximately
0.2
trillion
Btu.

The
Department
of
Energy
has
estimated
that
overfire
systems
could
be
installed
on
20
percent
of
the
200
coal
fired
boilers
in
the
industry,
resulting
in
680,000
MW­
hr
in
energy
savings
annually.
Additional
energy
savings
reportedly
are
possible
if
overfire
air
provides
similar
benefits
in
wood­
fired
systems.
Potential
reductions
in
NO
X
,
S0
2
,
CO,
PM,
VOCs
and
other
pollutants
such
as
mercury
would
accompany
such
energy
savings.

Commenters
also
expressed
a
need
for
operational
flexibility,
and
asserted
that
NSR
delays
can
limit
such
flexibility,
with
the
result
that
if
changes
are
projected
to
trigger
NSR,
even
changes
that
improve
energy
efficiency,
they
are
no
longer
economically
viable.
Because
some
industries
must
make
rapid
changes
in
their
product
lines
it
is
very
difficult
for
them
to
manage
NSR
compliance.
One
such
example
was
provided
by
the
flexible
packaging
industry.
In
that
case,
the
industry
has
been
moving
steadily
toward
the
replacement
of
solvent­
based
inks
and
coatings
with
water­
based
inks
and
coatings
in
the
production
of
packaging
for
foods,
drugs,
cosmetics,
and
other
household
goods.
However,
certain
product
orders
reportedly
require,
from
time­
to­
time,
solvent­
based
inks
or
coatings,
and
these
operations
are
required
to
operate
large
thermal
oxidizers
by
their
permits.
In
addition
many
of
the
low
VOC
coatings
contain
materials
that
can
poison
a
thermal
oxidizer's
catalyst.
Therefore,
the
plant
asked
its
permitting
agency
to
change
its
permit
to
run
the
oxidizer
only
when
it
runs
VOC­
based
coatings.
83
In
this
instance,
the
operator
calculated
that
the
change
could
save
approximately
15,000
cubic
feet
of
gas
and
650
kWh
of
electricity
each
day.
However,
the
commenter
felt
that
the
change
would
probably
be
a
change
in
the
plant's
method
of
operation,
triggering
NSR,
even
though
actual
emissions
were
expected
to
be
reduced
by
the
change.
Because
of
the
nature
of
its
operations,
involving
product
batches
sometimes
constituting
only
hours
of
a
day's
run,
the
84
Id.
85
NEDA/
CARP
Comments
[
II­
D­
272]
Attachment
A,
Example
#
12.

27
company
did
not
feel
it
could
accept
limits
on
its
hours
of
operation.
Therefore,
the
project,
which
according
to
the
commenter
was
conceived
as
a
way
to
create
large
energy
savings,
did
not
go
forward.
84
A
number
of
commenters
claimed
to
have
abandoned
energy
conservation
projects
because
they
determined
that
NSR
would
apply
and
make
the
project
cost­
prohibitive.
For
instance,
at
one
commenter's
automobile
assembly
plant,
the
company
wanted
to
eliminate
one
shift
of
a
two­
shift
operation
due
to
downward
market
fluctuations.
This
would
have
resulted
in
a
reduction
of
roughly
30%
(
0.4
billion
cubic
feet)
of
annual
natural
gas
usage
in
the
plant's
boilers,
ovens,
thermal
oxidizers
and
other
fuel
combustion
equipment
at
a
cost
savings
of
greater
than
$
2
million
dollars
annually.
In
addition,
electrical
power
consumption
would
have
been
reduced
by
roughly
10%,
at
a
cost
savings
of
greater
than
$
700,000
annually.
In
order
to
accommodate
this
change,
however,
the
facility
needed
to
install
certain
pieces
of
equipment,
consisting
mostly
of
assembly
motors
to
increase
the
production
capability
of
a
single
shift
by
two
automobiles
per
hour.
According
to
the
comment,
because
of
the
actual­
to­
potential
test,
and
the
source's
reluctance
to
take
a
cap
limiting
it
to
one­
shift
operation,
the
project
would
have
triggered
NSR
and
the
project
would
no
longer
have
been
economically
viable.
85
Overall,
the
comments
received
from
industries
other
than
utilities
and
refineries
also
provide
additional
evidence
suggesting
that
the
current
NSR
program
is
having
an
adverse
impact
on
energy
efficiency
by
discouraging
projects
that
may
improve
energy
efficiency,
or
may
increase
capacity
and
reliability
without
actually
increasing
pollutant
emissions.
In
some
cases
it
may
even
be
discouraging
projects
that
decrease
emissions,
because
of
the
"
actual­
to­
potential"
test
used
for
these
industries.

V.
Impact
on
Environmental
Protection
Overall,
EPA
believes
that
preventing
emissions
of
pollutants
covered
by
NSR
does
result
in
significant
environmental
and
public
health
benefits.
Attempting
to
specifically
quantify
the
NSR
program's
contribution
to
these
benefits
is
very
difficult
because
of
the
variety
of
Clean
Air
Act
programs
that
address
these
pollutants
and
because
there
is
no
tracking
by
any
government
agency
of
the
reductions
in
emissions
that
sources
make
due
to
the
NSR
program.
Moreover,
EPA
recognizes
that
measuring
risk
reduction
benefits
associated
with
any
given
reduction
in
emissions
requires
complex
risk
assessments
that
would,
in
turn,
require
more
specific
information
than
has
been
gathered
in
the
context
of
this
review.

We
note
that
NSR
is
implemented
in
the
context
of
several
other
significant
Clean
Air
Act
programs.
Available
information
indicates
that
these
other
programs
result
in
substantial
emissions
reductions.
For
example,
the
Title
IV
Acid
Rain
Program
has
reduced
SO
2
emissions
from
the
electric
utility
industry
by
more
than
7
million
tons
per
year.
The
Tier
2
motor
vehicle
emissions
standards
and
gasoline
sulfur
control
requirements
will
ultimately
achieve
NO
x
reductions
of
2.8
million
tons
per
year.
Standards
for
highway
heavy­
duty
vehicles
and
engines
will
reduce
NO
x
emissions
by
2.6
million
tons
per
year.
Standards
for
non­
road
diesel
engines
are
86In
response
to
public
requests
for
more
such
information,
the
Agency
has
added
to
the
docket
some
general
benefits
information
about
reductions
in
emissions
of
pollutants
likely
to
be
impacted
by
the
NSR
regulations.
(
A)
U.
S.
Environmental
Protection
Agency,
Office
of
Air
Quality
Planning
and
Standards.
Regulatory
Impact
Analysis
for
the
NO
X
SIP
call,
FIP,
and
Section
126
Petitions:
Volume
1,
Cost
and
Economic
Impacts.
September,
1998.
Located
on
the
Internet
at
www.
epa.
gov/
ttn/
oarpg/
otag/
sipriav1.
zip;
(
B)
U.
S.
Environmental
Protection
Agency,
Office
of
Air
Quality
Planning
and
Standards.
Regulatory
Impact
Analysis
for
the
NO
X
SIP
call,
FIP,
and
Section
126
Petitions:
Volume
2,
Health
and
Welfare
Benefits.
December,
1998.
Located
on
the
Internet
at
www.
epa.
gov/
ttn/
oarpg/
otag/
sipriav2.
zip;
(
C)
U.
S.
Environmental
Protection
Agency,
Office
of
Air
Quality
Planning
and
Standards.
Regulatory
Impact
Analysis
for
the
Final
Regional
Haze
Rule.
April,
1999.
Located
on
the
Internet
at
28
anticipated
to
reduce
NO
x
emissions
by
about
1.5
million
tons
per
year.
The
NO
x
"
SIP
Call"
will
reduce
NO
x
emissions
by
over
1
million
tons
per
year.
Altogether,
these
and
other
similar
programs
achieve
emissions
reductions
that
far
exceed
those
attributable
to
the
NSR
program.
Moreover,
most
of
these
other
programs
are
much
more
efficient,
streamlined,
and
simple
than
NSR
because
they
do
not
entail
the
same
resource­
intensive,
case­
by­
case
review
that
is
required
under
NSR.

It
would
be
very
difficult
to
estimate
or
quantify
the
benefits
of
the
NSR
program.
However,
EPA
believes
that
the
inability
to
make
exact
estimates
does
not
mean
that
these
benefits
are
insignificant
or
nonexistent.
Notably,
industry
concerns
about
NSR
focused
almost
exclusively
on
problems
associated
with
applying
the
program
to
existing
sources.
These
comments
illustrated
a
potential
dichotomy
in
that
the
benefits
of
the
NSR
program
are
largely
attributable
to
new
sources
while
the
existing
sources
reportedly
are
more
burdened
by
the
program.

Electric
utilities
and
petroleum
refineries
are
significant
sources
of
air
emissions.
The
major
regulated
air
pollutants
emitted
from
power
plants
are
SO
2
,
NO
X
,
PM,
and
mercury.
Refineries
primarily
emit
SO
2
and
NO
X
,
as
well
as
VOCs.
Based
on
2000
emissions,
the
electric
utility
industry
is
the
single
largest
source
of
SO
2
emissions
and
the
second
largest
source
of
NO
X
emissions
(
on
road
mobile
sources
are
the
largest).
In
2000,
the
electric
utility
industry
emitted
11.2
million
tons
of
SO
2
,
5.1
million
tons
of
NO
X
,
and
302,000
tons
of
PM.
In
1999,
refineries
emitted
479,000
tons
of
SO
2
,
299,000
tons
of
NO
X
and
161,200
of
volatile
organic
compounds.
Emissions
of
these
pollutants
from
all
sectors
in
1999
totaled
18.9
million
tons
SO
2
,
25.4
million
tons
NO
X
,
18.1
million
tons
VOC,
and
23.7
million
tons
PM.

There
is
a
significant
body
of
scientific
literature
linking
air
pollution
to
several
health
effects.
These
include:
premature
mortality,
chronic
asthma
and
increased
asthma
attacks,
chronic
and
acute
bronchitis,
other
chronic
respiratory
diseases
and
damage,
increased
airway
responsiveness
to
stimuli,
inflammation
in
the
lung,
respiratory
cell
damage,
premature
aging
of
the
lungs,
increased
susceptibility
to
respiratory
infection,
decreased
lung
function,
developmental
effects,
infant
mortality,
low
birth
weight,
cancer,
decreased
time
to
onset
of
angina,
other
cardiovascular
effects.
Additional
effects
include
decreased
worker
productivity;
increased
emergency
room
visits
for
respiratory
and
cardiovascular
effects,
and
more
hospital
admissions
for
respiratory
and
cardiac
diseases.
86
www.
epa.
gov/
ttn/
oarpg/
t1/
reports/
riaes.
pdf;
and
(
D)
U.
S.
Environmental
Protection
Agency,
Office
of
Air
Quality
Planning
and
Standards.
Regulatory
Impact
Analysis
for
the
Final
Section
126
Petition
Rule.
December,
1999.
Located
on
the
Internet
at
www.
epa.
gov/
ttn/
oarpg/
t1/
reports/
126fn0.
pdf.
87
First
Energy
Corporation
testimony
on
NSR,
7/
10/
2001,
stated
that
current
interpretations
of
NSR
would
have
prevented
projects
now
resulting
in
a
reduction
of
40,000
TPY
of
SO
2
and
NO
X
emissions.

29
Potential
effects
beyond
human
health
effects
include
direct
damage
to
plants
and
forests,
decreased
yields
for
crops
and
forest
products,
damage
to
ecosystem
functions,
decreased
visibility,
corrosion
and
soiling
of
buildings
and
monuments,
eutrophication
(
i.
e.,
explosive
algae
growth
leading
to
a
depletion
of
oxygen
in
the
water),
acidic
deposition
and
acidification
of
water
bodies,
and
impacts
on
recreational
demand
from
damaged
aesthetics
and
decreased
visibility.

The
EPA
Background
Paper
provided
some
preliminary
estimates
of
the
amount
of
emissions
prevented
by
the
NSR
program
for
all
industries
in
"
clean"
areas
(
e.
g.,
emissions
that
would
have
otherwise
occurred
from
construction/
modification).
The
NSR
program
in
such
clean
areas
is
known
as
the
PSD
program.
The
Paper
stated
that
for
the
period
1997
through
1999,
new
or
modified
source
compliance
with
PSD
for
all
industries
prevented
approximately
1.4
million
tons
of
air
pollution
from
being
emitted
per
year.
The
vast
majority
of
these
reductions
are
attributable
to
the
application
of
NSR
to
new
gas
fired
electric
generating
units.
The
Background
Paper
also
reported
that
this
number
underestimates
total
emission
reductions
because
it
does
not
include
estimates
of
emissions
prevented
in
nonattainment
areas
through
nonattainment
NSR
permitting
requirements
during
that
same
time
period.

Several
commenters
reiterated
this
position
and
noted
that
as
a
general
rule
these
reductions
would
be
greater
because
the
control
requirements
are
more
stringent
and
the
offset
requirements
essentially
result
in
a
net
emissions
decrease.
Although
EPA
agrees
that
there
are
additional
emission
reductions
that
result
from
compliance
with
the
offsets
requirements
of
nonattainment
NSR
program,
at
this
time
the
Agency
does
not
have
information
quantifying
those
emissions
reductions.
Finally,
other
commenters
noted
that
the
EPA
Background
Paper
failed
to
address
the
emission
reductions
of
SO
2
and
NO
X
that
occur
as
a
result
of
sources
reducing
their
emissions
so
as
to
avoid
the
applicability
of
NSR
altogether.
On
the
other
hand,
since
SO
2
emissions
from
the
utility
industry
are
capped
by
the
Title
IV
acid
rain
program,
NSR
does
not
produce
overall
net
reduction
in
SO
2
emissions
from
the
industry.
Similarly,
in
nonattainment
areas,
Title
I
effectively
caps
emissions
of
the
nonattainment
pollutant.
To
a
degree,
the
same
is
true
for
seasonally
or
geographically
limited
cap
and
trade
programs,
such
as
the
"
NO
X
SIP
call."
Furthermore,
as
noted
below,
industry
commenters
note
that
these
estimates
of
emission
reductions
attributed
to
NSR
do
not
account
for
foregone
emissions
reductions
that
they
allege
would
have
occurred
in
the
absence
of
NSR's
disincentives
to
proceed
with
projects
that
increase
efficiency.
87
A
large
number
of
commenters,
primarily
citizens
and
environmental
groups,
expressed
strong
support
for
the
benefits
that
derive
from
reducing
emissions
from
these
industrial
sectors,
either
by
installing
pollution
reduction
controls
on
new
sources
as
they
are
built,
or
on
existing
sources
as
they
are
modified.
Many
groups
argued
that
the
public
health
threat
from
the
air
30
emissions
of
power
plants
and
refineries
is
urgent
and
further
reductions
are
needed.
Noting
environmental
justice
concerns,
one
commenter
stated
that
80
percent
of
the
refineries
in
the
Texas
oil
refinery
communities
are
either
populated
by
minority
citizens
or
contain
significant
minority
representation
and
reported
that
approximately
three
million
minority
citizens
live
in
these
Texas
communities.

The
EPA
Background
Paper
also
presented
previous
estimates
of
the
health
benefits
per
ton
of
pollutant
reduced
for
SO
2
and
NO
X
emissions
based
on
a
study
of
emissions
at
utilities.
The
work
cited
in
the
EPA
Background
Paper
is
based
on
the
benefits
of
reducing
premature
mortality
associated
with
long­
term
exposure
to
PM.
However,
many
citizen
and
environmental
group
commenters
requested
a
more
detailed
discussion
of
additional
health
benefits
like
the
avoidance
of
reduced
lung
function,
asthma
attacks,
lost
work
days
and
premature
death,
which
have
been
linked
to
these
air
pollutants.
For
example,
one
commenter
representing
43
environmental
groups
cited
a
study
by
Abt
Associates
presenting
their
estimate
that
national
power
plant
emissions
accounted
for
more
than
6,000
asthma
attacks,
30,000
premature
deaths,
and
5
million
lost
work
days
per
year,
noting
that
elderly
people
with
respiratory
disease
and
children
are
at
the
greatest
risk.

Commenters
requested
that
EPA
present
information
on
the
benefits
due
to
avoided
emissions
of
other
pollutants,
including
pollutants
that
are
reduced
collaterally
when
criteria
pollutants
are
controlled
(
e.
g.,
mercury).
One
commenter
notes
that
EPA
documents
identify
coal­
fired
power
plants
as
the
largest
industrial
emitters
of
mercury,
another
pollutant
with
welldocumented
health
and
environmental
effects.
Thus,
without
addressing
the
benefits
that
derive
from
reductions
of
these
pollutants
as
well,
several
commenters
argue
that
the
EPA
Background
Paper
significantly
underestimates
public
health
and
environmental
benefits
of
NSR.

Many
commenters
also
mentioned
numerous
other
benefits
that
result
from
lower
emissions
from
power
plants
and
refineries.
They
presented
information
about
impacts
primarily
of
power
plant
emissions
on
the
environment,
particularly
in
National
Parks.
For
example,
several
groups
provided
information
regarding
the
adverse
impact
of
power
plant
emissions
in
particular
on
visibility
in
National
Parks.
Some
commenters
also
note
that
ground
level
ozone
(
smog)
not
only
impacts
vegetation
(
more
than
50
species
of
plants
and
trees
allegedly
harmed
by
ozone),
but
also
the
health
of
visitors
to
National
Parks.
Additionally,
commenters
note
the
impact
of
SO
2
and
NO
X
emissions
on
the
formation
of
acid
rain
and
its
impact
on
ecosystems
(
e.
g.,
red
spruce
decline,
fish
killed).
Finally,
many
commenters
were
also
concerned
about
CO
2
emissions
and
their
potential
to
affect
climate,
and
believed
that
NSR
plays
a
role
in
preventing
these
emissions
as
well.
Commenters
urged
EPA
to
discuss
the
benefits
generally
of
reduced
emissions
in
all
these
areas
more
explicitly,
and
quantify
them
as
they
relate
to
the
NSR
program.

In
addition,
several
commenters
noted
that
in
nonattainment
areas,
a
source's
failure
to
reduce
emissions
through
NSR
places
the
burden
on
other
sources
to
reduce
emissions.
In
other
words,
because
the
State
has
to
reduce
emissions
somewhere
in
order
to
attain
air
quality
standards,
it
will
target
other
sources
(
e.
g.,
construction
activities),
or
even
consumers
in
order
to
create
those
reductions.
Even
in
attainment
areas,
compliance
with
PSD
requirements
can
help
maintain
the
area's
ability
to
continue
to
grow.
88
See,
e.
g.
STAPPA/
ALAPCO
comments.

31
Some
state
and
local
governments
supported
the
role
NSR
plays
in
preventing
emissions
from
new
and
modified
sources.
88
They
believe,
based
on
their
experience,
that
without
NSR,
emissions
from
new
and
modified
sources
would
severely
interfere
with
their
efforts
to
attain
and
maintain
air
quality
standards.
While
there
are
several
important
programs
that
reduce
emissions
from
existing
sources,
they
felt
NSR
was
a
critical
complementary
program
because
it
minimized
emissions
from
new
sources.

Some
commenters
also
expressed
support
for
the
technology­
forcing
aspect
of
the
NSR
program,
arguing
that
it
is
the
only
CAA
program
that
automatically
mirrors
improvements
in
control
technology
over
time,
and
therefore
encourages
continued
development
of
cleaner
technology.
Commenters
urged
EPA
to
estimate
the
benefits
of
this
effect
as
well.

Industry
commenters
felt
that
the
current
NSR
program
actually
acts
as
a
barrier
to
improved
environmental
protection
in
certain
instances.
Although
NSR
is
only
triggered
when
emissions
increase,
these
commenters
argued
that
the
way
EPA
calculates
an
increase
in
emissions
can
actually
have
the
effect
of
subjecting
a
project
to
NSR
that
would
decrease
actual
emissions.
Because
of
the
delay
and
costs
associated
with
applying
NSR
to
a
project,
NSR
renders
these
environmentally
beneficial
projects
uneconomical,
and
they
may
be
rejected.
Similarly,
again
because
of
the
way
that
NSR
calculates
emissions
increases,
several
industry
commenters
noted
an
incentive
to
keep
actual
emissions
high
because
the
closer
actual
emissions
are
to
a
source's
maximum
capacity
to
emit,
the
less
likely
it
is
to
trigger
NSR.

VI.
Conclusion
Based
upon
the
information
examined
during
this
review
of
the
NSR
program,
there
appears
to
be
little
incremental
impact
of
the
program
on
the
construction
of
new
electricity
generation
and
refinery
facilities
but
a
more
dramatic
impact
on
investment
in
utility
and
refinery
generating
capacity
and
energy
efficiency
at
existing
utility
and
refinery
plants.
Looking
at
industry
as
a
whole,
there
also
is
clear
evidence
of
NSR's
benefits
for
environmental
protection.

With
respect
to
environmental
protection,
the
EPA
finds
that
NSR
is
not
designed
to
play
the
primary
role
in
reducing
emissions
from
existing
sources.
In
fact,
for
pollutants
covered
by
a
national
cap
and
trade
program
(
such
as
the
Title
IV
acid
rain
program),
the
NSR
program
does
not
necessarily
produce
any
overall
emissions
reductions.
Furthermore,
EPA
believes
that
in
particular
industry
sectors
 
especially
the
utility
sector
 
the
benefits
currently
attributed
to
NSR
could
be
achieved
much
more
efficiently
and
at
much
lower
cost
through
the
implementation
of
a
multi­
pollutant
national
cap
and
trade
program.

Nevertheless,
the
NSR
program
plays
a
role
in
attainment
and
maintenance
of
the
NAAQS,
particularly
with
regard
to
new
sources.
It
helps
ensure
that
as
industry
continues
to
grow
and
expand,
air
quality
is
managed
appropriately
(
i.
e.,
by
helping
assure
that
clean
areas
do
not
worsen
and
that
dirty
areas
get
cleaner).
It
also
helps
to
protect
sensitive
areas
like
national
parks
and
wilderness
areas,
and
promotes
new
and
more
effective
pollution
controls.
As
89
Very
few
commenters
provided
sufficiently
detailed
examples
for
EPA
to
make
definitive
judgements
as
to
whether
the
given
projects
would
have
been
considered
nonroutine
or
ultimately
triggered
NSR.
As
a
result,
EPA
cannot
quantify
the
number
of
projects
affected
or
the
corresponding
impacts
on
capacity,
reliability,
efficiency,
safety,
or
other
relevant
factors.
Based
on
the
information
presented,
it
appears
unlikely
that
many
of
the
examples
discussed
would
trigger
NSR
either
because
they
would
qualify
for
the
routine
exclusion
or
they
would
not
increase
emissions
significantly.
Nevertheless,
the
anecdotal
information
was
sufficient
to
support
our
conclusions
with
regard
to
the
overall
impact
of
the
NSR
program.

90
The
EPA
notes
that
its
conclusions
for
refiners
are
equally
valid
for
the
numerous
nonutility
non­
refinery
sources
that
commented
during
the
review.

32
described
in
this
report,
and
thoroughly
detailed
in
the
comments
and
other
references
provided,
NSR
also
provides
health
and
ecological
benefits.

With
respect
to
new
facilities,
the
NSR
program's
principal
impacts
are
in
the
form
of
delays
and
additional
costs,
but
there
is
little
evidence
that
these
delays
and
costs
are
preventing
new
source
construction
in
the
utility
industry.
Indeed
there
is
substantial
evidence
that
significant
new
generating
capacity
is
being
brought
online
within
normal
time
frames
for
planning
such
projects.

With
respect
to
the
maintenance
and
operation
of
existing
utility
generation
capacity,
there
is
more
evidence
of
adverse
impacts
from
NSR.
Credible
examples
were
presented
of
cases
in
which
uncertainty
about
the
exemption
for
routine
activities
has
resulted
in
delay
or
cancellation
of
projects
which
sources
say
are
done
for
the
purposes
of
maintaining
and
improving
the
reliability,
efficiency
and
safety
of
existing
energy
capacity.
89
Such
discouragement
results
in
lost
capacity,
as
well
as
lost
opportunities
to
improve
energy
efficiency
and
reduce
air
pollution.

There
appeared
to
be
little
impact
of
NSR
on
planning
for
new
greenfield
refineries,
because
new
refineries
are
not
being
built
for
economic
and
environmental
reasons
unrelated
to
NSR.
For
existing
refineries,
the
points
raised
above
about
the
routine
maintenance
exclusion
apply
equally
well
to
refineries
as
they
do
for
utilities
 
the
EPA
observed
that
commenters
expressed
uncertainty
about
the
application
of
the
exclusion
to
any
particular
project.
Existing
refineries,
however,
face
an
additional
issue:
the
actual­
to­
potential
emissions
test.
The
EPA
found
credible
examples
of
projects
at
existing
units
that
would
have
provided
needed
capacity
or
efficiency
improvements
and
would
likely
not
have
increased
 
and
in
some
cases
may
have
decreased
 
actual
emissions.
Due
to
the
actual­
to­
potential
test,
such
projects,
if
they
occur
at
units
operating
below
capacity,
could
trigger
NSR
unless
the
company
committed
to
continue
operating
the
units
below
capacity
or
installed
pollution
controls.
The
EPA
believes
that
this
potentially
results
in
lost
refining
capacity,
or
foregone
opportunities
to
increase
capacity
without
increasing
emissions,
which
could
contribute
to
price
volatility
and
shortages
in
fuel
supply.
90
With
respect
to
energy
efficiency,
the
EPA
recognizes
that
the
NSR
program
applies
to
certain
projects
that
have
the
effect
of
increasing
efficiency
(
e.
g.,
projects
that
increase
electricity
output
for
a
given
fuel
input).
The
ordinary
costs
and
permitting
times
associated
with
NSR
may,
in
the
EPA's
judgment,
result
in
the
delay
or
cancellation
of
certain
projects
that
could
improve
33
energy
efficiency.
EPA
encourages
energy
efficiency
improvements
wherever
feasible.
However,
the
EPA
notes
that
some
changes
that
improve
energy
efficiency
also
can
result
in
significant
emissions
increases
that
have
adverse
air
quality
impacts
that
must
be
reviewed,
even
though
the
proposed
project
could
reduce
regional
or
national
emissions.
Thus,
of
the
universe
of
possible
efficiency
improvements,
the
appropriate
focus
of
the
NSR
program
is
on
those
that
are
nonroutine
and
that
significantly
increase
emissions.
At
non­
utility
source
categories,
the
"
actual
to
potential"
emissions
test
can
discourage
efficiency
improvement
projects
even
where
there
would
not
be
an
increase
in
actual
emissions.
It
is
clear
that
some
of
these
efficiency
improvements
can
still
go
forward
(
by
going
through
NSR
or
taking
steps
to
avoid
NSR);
however,
it
also
is
clear
that
others
are
in
fact
canceled
due
to
the
costs
and
delays
associated
with
NSR.

As
noted
at
the
beginning
of
this
report,
representatives
of
industry,
state
and
local
agencies,
and
environmental
groups
have
worked
with
EPA
for
over
a
decade
on
developing
improvements
to
the
NSR
program.
Our
findings
in
this
report
ratify
a
longstanding
and
broadlyheld
belief
that
parts
of
the
NSR
program
can
and
should
be
improved.
For
example,
we
conclude
above
that
changes
to
NSR
that
add
to
the
clarity
and
certainty
of
the
scope
of
the
routine
maintenance
exclusion
will
improve
the
program
by
reducing
the
unintended
consequences
of
discouraging
worthwhile
projects
that
are
in
fact
outside
the
scope
of
NSR.
For
these
reasons,
EPA
is
recommending
a
number
of
changes
to
the
NSR
program
that
will
address
the
concerns
raised
during
this
NSR
review
as
well
as
many
other
concerns
presented
to
EPA
about
NSR
over
the
past
decade.
