1Final
Regulatory
Impact
Analysis:
Control
of
Emissions
from
Nonroad
Diesel
Engines.
Assessment
and
Standards
Division,
Office
of
Transportation
and
Air
Quality,
U.
S.
Environmental
Protection
Agency.
May
2004.
EPA/
420­
R­
04­
007.

1
MEMORANDUM
DATE:
May
31,
2005
SUBJECT:
Control
Costs
for
NOx
Adsorbers
and
CDPF
for
CI
Engines
FROM:
Tanya
Parise,
Alpha­
Gamma
Technologies,
Inc.

TO:
Sims
Roy,
EPA
OAQPS
ESD
Combustion
Group
The
purpose
of
this
memorandum
is
to
present
information
on
the
cost
of
controls
for
stationary
compression
ignition
(
CI)
internal
combustion
engines
(
ICE).
The
cost
information
presented
in
this
memorandum
will
be
used
to
estimate
cost
impacts
associated
with
the
new
source
performance
standards
(
NSPS)
for
stationary
CI
ICE.
The
control
technologies
discussed
in
this
memorandum
are
NOx
adsorbers
and
catalyzed
diesel
particulate
filters
(
CDPF),
which
are
the
technologies
that
are
the
basis
for
the
proposed
emissions
standards
for
the
control
of
NOx
and
PM,
respectively.

Introduction
The
costs
of
NOx
adsorbers
and
CDPF
presented
in
this
memorandum
were
estimated
based
on
information
obtained
from
the
final
regulatory
impact
analysis
(
RIA)
for
nonroad
diesel
engines
developed
by
the
Office
of
Transportation
and
Air
Quality
(
OTAQ)
published
in
May
2004.1
The
following
sections
describe
how
the
capital
and
annual
costs
for
these
control
technologies
were
derived
based
on
information
from
the
RIA.

Control
Costs
NOx
Adsorbers
Table
6.2­
11
of
the
RIA
presented
NOx
adsorber
system
costs
in
2002
dollars
for
2
engines
of
varying
size
and
displacement.
Several
costs
were
presented
in
Table
6.2­
11
including
the
baseline
cost
to
buyer
(
near
term
and
long
term),
the
cost
to
buyer
with
highway
learning
(
near
term
and
long
term),
and
the
cost
to
buyer
with
nonroad
learning
(
long
term).

The
RIA
indicated
that
costs
for
NOx
adsorbers
were
estimated
based
on
the
methodology
used
in
the
2007
heavy­
duty
highway
engines
rulemaking.
This
rulemaking
sets
final
emissions
standards
for
2007
and
later
engines
used
in
heavyduty
highway
vehicles.
It
was
also
indicated
that
the
control
technologies
expected
to
be
used
to
show
compliance
with
the
nonroad
standards
are
the
same
as
those
expected
for
highway
engines.
The
long
term
costs
presented
by
OTAQ
assume
that
control
system
costs
will
decrease
over
time
as
manufacturers
become
more
experienced
with
production
and
can
make
changes,
adjustments,
and
improvements
lowering
the
cost
of
production.
According
to
OTAQ,
this
is
often
described
as
the
manufacturing
learning
curve.
In
the
RIA,
it
was
indicated
that
there
would
be
a
learning
curve
associated
with
the
heavy­
duty
highway
engine
rule
as
well
as
the
nonroad
rule
for
diesel
engines.
Nonroad
diesel
engines
currently
do
not
employ
any
type
of
NOx
aftertreatment
and
CDPF
have
only
been
applied
in
limited
applications,
according
to
OTAQ.
These
are
therefore
new
technologies
for
nonroad
diesel
engines
and
will
involve
a
learning
curve
beyond
the
learning
in
response
to
the
heavy­
duty
highway
rule.
The
standards
for
nonroad
CI
engines
follow
the
implementation
of
the
heavy­
duty
highway
engines
rule
and
OTAQ
indicated
that
the
2007
heavy­
duty
highway
engines
rule
was
used
as
the
baseline
level
of
learning
for
nonroad
engines.

Stationary
CI
engines
are
similar
to
nonroad
CI
engines
and
EPA
believes
the
costs
associated
with
a
NOx
adsorber
developed
for
nonroad
CI
engines
would
be
similar
to
NOx
adsorber
costs
for
stationary
CI
engines.
Also,
the
NSPS
will
require
stationary
CI
engines
to
meet
the
nonroad
CI
engine
emissions
standards.
The
EPA
therefore
feels
it
is
appropriate
to
use
the
nonroad
control
costs
developed
by
OTAQ
for
stationary
CI
engines.
The
EPA
believes
that
it
is
appropriate
to
take
into
account
the
learning
curve
when
estimating
the
cost
of
NOx
adsorbers.
The
technology
is
currently
considered
a
new
technology
and
EPA
expects
that
as
the
technology
is
more
widely
applied,
system
costs
will
decrease
in
the
future.
The
control
costs
would
be
higher
if
the
near
term
costs
presented
by
OTAQ
were
used.
Since
the
long
term
costs
include
a
learning
curve
effect
for
portions
of
the
NOx
adsorber
system,
EPA
feels
it
is
justified
in
using
these
costs.
Finally,
since
the
technology
is
not
available
yet
but
is
expected
to
be
available
in
approximately
2011,
EPA
believes
that
for
NOx
adsorbers
it
is
appropriate
to
follow
the
timeline
for
the
nonroad
rulemaking
(
2011)
and
therefore
use
the
costs
estimated
for
the
nonroad
CI
engine
rule.

In
order
to
develop
a
relationship
between
the
NOx
adsorber
system
cost
and
engine
size
to
determine
the
capital
and
annual
costs
for
different
engine
sizes,
EPA
generated
a
plot
of
the
NOx
adsorber
system
costs
obtained
from
Table
6.2­
11
of
the
RIA
versus
the
engine
horsepower
(
HP).
Assuming
a
linear
trend,
the
following
3
functions
were
developed:

Baseline
Cost
to
Buyer
$
4.0(
x)
+
$
213
R2=
0.9926
Cost
to
Buyer
w/
Highway
Learning
$
3.3(
x)
+
$
194
R2=
0.9926
Cost
to
Buyer
w/
Nonroad
Learning
$
2.8(
x)
+
$
178
R2=
0.9927
where
x
represents
the
engine
size
in
HP.
The
linear
regression
plot
is
included
in
Attachment
A.

Based
on
the
above
functions
developed
by
EPA,
the
purchased
equipment
cost
was
calculated
for
different
engine
sizes.
The
capital
and
annual
costs
were
determined
using
the
Office
of
Air
Quality
Planning
and
Standards
(
OAQPS)
Control
Cost
Methodology
described
below:

Determine:
1
­
Total
Capital
Costs
2
­
Total
Annual
Costs
1
­
Total
Capital
Cost
Components
and
Factors:

Total
Capital
Cost
(
TCC)
=
Direct
Costs
(
DC)
+
Indirect
Costs
(
IC)

1.1
­
Direct
Costs
(
DC):
DC
=
PEC
+
DIC
1.1.1
­
Purchased
Equipment
Costs
(
PEC):

­
Control
Device
and
Auxiliary
Equipment
(
EC)
­
Instrumentation
(
10%
of
EC)
­
Sales
Tax
(
3%
of
EC)
­
Freight
(
5%
of
EC)

PEC
=
118%
EC
1.1.2
­
Direct
Installation
Costs
(
DIC)

­
Foundations
and
Supports
(
8%
of
PEC)
­
Handling
and
Erection
(
14%
of
PEC)
­
Electrical
(
4%
of
PEC)
­
Piping
(
2%
of
PEC)
­
Insulation
for
Ductwork
(
1%
of
PEC)
­
Painting
(
1%
of
PEC)
4
DIC
=
30%
PEC
DC
=
PEC
+
0.3
PEC
=
1.3
PEC
1.2
­
Indirect
Costs
(
IC):
IC
=
ICC
+
C
1.2.1
­
Indirect
Installation
Costs
(
IIC)

­
Engineering
(
10%
of
PEC)
­
Construction
and
Field
Expenses
(
5%
of
PEC)
­
Contractor
Fees
(
10%
of
PEC)
­
Startup
(
2%
of
PEC)
­
Performance
Test
(
1%
of
PEC)

IIC
=
28%
PEC
=
0.28
PEC
1.2.2
­
Contingencies
(
C)
(
3%
of
PEC)

­
Equipment
Redesign
and
Modifications
­
Cost
Escalations
­
Delays
in
Startup
C
=
3%
PEC
=
0.03
PEC
IC
=
0.28
PEC
+
0.03
PEC
=
0.31
PEC
TCC
=
1.3
PEC
+
0.31
PEC
=
1.61
PEC
=
1.61
(
1.18
EC)
=
1.9
EC
2
­
Total
Annual
Cost
Elements
and
Factors
Total
Annual
Cost
(
TAC)
=
Direct
Annual
Costs
(
DC)
+
Indirect
Annual
Costs
(
IC)

2.1
­
Direct
Annual
Costs
(
DC):

­
Utilities
­
Operating
Labor
­
Maintenance
­
Annual
Compliance
Test
­
Catalyst
Cleaning
­
Catalyst
Replacement
­
Catalyst
Disposal
2.2
­
Indirect
Annual
Costs
(
IC)
5
­
Overhead
(
60%
of
operating
labor
and
maintenance
costs)
­
Fuel
Penalty
­
Property
Tax
(
1%
of
TCC)
­
Insurance
(
1%
of
TCC)
­
Administrative
Charges
(
2%
of
TCC)
­
Capital
Recovery
=
{
I(
1+
I)
n/((
1+
I)
n­
1)*
TCC}
where
I
is
the
interest
rate,
and
n
is
the
equipment
life.

The
information
from
OTAQ's
RIA
did
not
include
any
direct
annual
costs
for
NOx
adsorbers
such
as
operating
and
maintenance
costs.
These
costs
were
therefore
assumed
to
be
zero.
Indirect
annual
costs
were
calculated
based
on
the
OAQPS
Control
Cost
Methodology
assuming
an
equipment
life
of
20
years
and
an
interest
rate
of
7
percent.
According
to
OTAQ,
the
fuel
penalty
associated
with
NOx
adsorbers
is
estimated
to
be
about
1
percent.
For
the
purposes
of
estimating
indirect
annual
costs,
EPA
assumed
the
fuel
penalty
was
negligible.
The
estimated
capital
and
annual
costs
associated
with
a
NOx
adsorber
applied
to
engines
of
varying
sizes
are
shown
in
Table
1.

Table
1:
Capital
and
Annual
Costs
Associated
with
a
NOx
Adsorber
Engine
Size
(
HP)
Purchased
Equipment
Cost
Total
Capital
Cost
Total
Annual
Cost
Total
Capital
Cost
($/
HP)
Total
Annual
Cost
($/
HP)

75
$
388
$
737
$
99
$
10
$
1
135
$
556
$
1,056
$
142
$
8
$
1
238
$
844
$
1,604
$
216
$
7
$
1
400
$
1,298
$
2,466
$
331
$
6
$
1
750
$
2,278
$
4,328
$
582
$
6
$
1
3000
$
8,578
$
16,298
$
2,190
$
5
$
1
Average
$
7
$
1
*
Costs
include
the
costs
of
an
oxidation
catalyst
and
are
based
on
the
cost
to
buyer
with
nonroad
learning.

CDPF
Table
6.2­
13
of
the
RIA
presented
CDPF
system
costs
in
2002
dollars
for
engines
of
varying
size
and
displacement.
Several
costs
were
presented
in
Table
6.2­
13
including
the
baseline
cost
to
buyer
(
near
term
and
long
term),
the
cost
to
buyer
with
highway
learning
(
near
term
and
long
term),
and
the
cost
to
buyer
with
nonroad
learning
(
long
6
term).
The
EPA
will
require
emissions
standards
for
stationary
CI
engines
that
are
based
on
the
use
of
CDPF
following
the
schedule
for
nonroad
CI
engines.
The
EPA
therefore
believes
that
for
CDPF
it
is
appropriate
to
follow
the
nonroad
rulemaking
and
use
the
cost
for
CDPF
that
incorporates
the
nonroad
engine
learning
curve.
As
with
the
NOx
adsorber
system
costs,
EPA
believes
that
it
is
also
appropriate
to
use
the
long
term
costs
associated
with
CDPF.
This
technology
is
also
relatively
new
and
EPA
expects
the
costs
of
CDPF
to
decrease
over
time
as
the
technology
is
more
frequently
applied
to
CI
engines.
The
control
costs
would
be
higher
if
near
term
costs
from
OTAQ
or
current
costs
provided
by
engine
control
vendors
for
CDPF
were
used.

In
order
to
develop
a
relationship
between
the
CDPF
system
cost
and
engine
size,
EPA
again
developed
a
linear
regression
of
the
CDPF
system
costs
obtained
from
Table
6.2­
13
of
the
RIA
versus
the
engine
HP.
Based
on
the
regression
analysis,
the
following
functions
were
developed:

Baseline
Cost
to
Buyer
$
5.8(
x)
+
$
117
R2=
0.9936
Cost
to
Buyer
w/
Highway
Learning
$
4.7(
x)
+
$
93
R2=
0.9936
Cost
to
Buyer
w/
Nonroad
Learning
$
3.7(
x)
+
$
75
R2=
0.9936
where
x
represents
the
engine
size
in
HP.
The
linear
regression
plot
is
included
in
Attachment
A.

Information
in
the
RIA
also
included
costs
for
a
CDPF
regeneration
system.
According
to
OTAQ,
some
form
of
active
regeneration
is
expected
to
be
used
as
a
backup
to
the
passive
regeneration
ability
of
the
CDPF.
It
was
further
stated
in
the
RIA
that
there
are
challenges
associated
with
implementing
CDPF
with
nonroad
applications
beyond
those
of
highway
applications.
It
is
anticipated
that
some
additional
hardware
beyond
the
filter
itself
may
be
required
in
order
to
ensure
that
regeneration
of
the
filter
occurs.
This
may
include
new
fuel
control
strategies
that
force
regeneration
or
it
may
include
an
exhaust
system
fuel
injector
to
inject
fuel
upstream
of
the
CDPF
to
provide
the
necessary
regeneration.
The
estimated
costs
of
such
a
system
were
presented
in
Table
6.2­
16
of
the
RIA
for
engines
of
varying
size
and
displacement.
Based
on
the
information
in
Table
6.2­
16,
EPA
developed
the
following
linear
relationship
between
the
CDPF
regeneration
system
cost
and
engine
size:

Cost
to
Buyer
w/
Learning
­
Regeneration
$
0.18(
x)
+
$
123
R2=
0.9706
where
x
represents
the
engine
size
in
HP.
The
linear
regression
plot
is
included
in
Attachment
A.

Note
that
the
cost
function
shown
above
for
the
regeneration
system
applies
to
engines
with
a
direct
injection
(
DI)
fuel
system.
In
a
DI
fuel
system,
fuel
is
injected
directly
into
the
main
combustion
chamber.
In
an
indirect
injection
(
IDI)
fuel
system,
fuel
is
injected
7
into
a
small
pre
chamber
above
the
main
combustion
chamber
where
combustion
begins.
The
main
combustion
chamber
is
lit
off
by
the
flame
from
the
small
chamber.
According
to
Caterpillar,
a
major
manufacturer
of
stationary
diesel
engines,
all
Caterpillar
engines
manufactured
in
the
last
15
years
have
been
DI
engines.
An
IDI
fuel
system
is
less
favorable
because
it
is
less
fuel
efficient
according
to
the
manufacturer,
but
there
are
still
a
few
manufacturers
using
the
IDI
design,
but
there
are
fewer
each
year.
The
EPA
therefore
expects
that
most
stationary
diesel
engines
would
have
a
DI
fuel
system
and
has
included
in
the
cost
of
CDPF
the
cost
of
a
regeneration
system
for
engines
with
a
DI
fuel
system.
The
regeneration
system
costs
for
engines
with
IDI
fuel
systems
would
be
twice
as
much
as
the
regeneration
system
costs
for
engines
with
DI
fuel
systems.
Based
on
the
functions
developed
by
EPA,
the
purchased
equipment
cost
was
calculated
for
different
engine
sizes
using
DI
fuel
systems.
The
capital
and
annual
costs
were
determined
using
the
OAQPS
Control
Cost
Methodology
as
previously
described.
Maintenance
costs
associated
with
a
CDPF
system
were
obtained
from
Table
6.2­
30
of
the
RIA.
The
table
indicated
that
a
maintenance
interval
of
3,000
hours
for
engines
below
175
HP
and
4,500
hours
for
engines
above
175
HP
was
appropriate.
Table
6.2­
30
further
indicated
that
the
estimated
costs
associated
with
maintenance
was
$
65
for
engines
up
to
600
HP
and
$
260
per
event
for
engines
above
600
HP.
The
EPA
used
these
costs
to
estimate
the
annual
maintenance
costs
associated
with
CDPF
for
prime
and
emergency
engines
as
the
hours
of
operation
affect
the
frequency
of
maintenance.
It
was
assumed
that
prime
engines
operate
1,000
hours
per
year
and
emergency
engines
operate
37
hours
per
year.
An
equipment
life
of
20
years
and
an
interest
rate
of
7
percent
was
used
to
estimate
the
indirect
annual
costs.
According
to
OTAQ,
the
fuel
penalty
associated
with
CDPF
is
estimated
to
be
about
1
percent.
For
the
purposes
of
estimating
indirect
annual
costs,
EPA
assumed
the
fuel
penalty
was
negligible.
This
is
consistent
with
information
received
from
a
CDPF
vendor
who
indicated
that
the
fuel
penalty
associated
with
CDPF
would
be
negligible.
The
capital
and
annual
costs
associated
with
a
CDPF
system
are
shown
in
Table
2.

Table
2:
Capital
and
Annual
Costs
Associated
with
a
CDPF
Engine
Size
(
HP)
Purchased
Equipment
Cost
Total
Capital
Cost
Total
Annual
Cost
Total
Capital
Cost
($/
HP)
Total
Annual
Cost
($/
HP)

Prime
Emergency
Prime
Emergency
75
$
489
$
929
$
160
$
126
$
12
$
2
$
2
135
$
722
$
1,371
$
219
$
186
$
10
$
2
$
1
238
$
1,121
$
2,131
$
309
$
287
$
9
$
1
$
1
400
$
1,750
$
3,325
$
470
$
448
$
8
$
1
$
1
Engine
Size
(
HP)
Purchased
Equipment
Cost
Total
Capital
Cost
Total
Annual
Cost
Total
Capital
Cost
($/
HP)
Total
Annual
Cost
($/
HP)

Prime
Emergency
Prime
Emergency
8
750
$
3,108
$
5,905
$
840
$
795
$
8
$
1
$
1
3000
$
11,838
$
22,492
$
3,115
$
3,026
$
7
$
1
$
1
Average
$
9
$
1
$
1
*
Costs
include
the
costs
of
an
oxidation
catalyst
and
are
based
on
the
cost
to
buyer
with
nonroad
learning.
**
Costs
presented
are
for
engines
with
a
DI
fuel
system.
Costs
for
engines
with
an
IDI
fuel
system
would
be
twice
the
costs
of
engines
with
a
DI
fuel
system.
9
Summary
Table
3
presents
a
summary
of
the
control
costs
associated
with
the
NSPS
for
stationary
CI
engines
that
would
be
incurred
due
to
emissions
standards
that
are
based
on
the
use
of
both
NOx
adsorbers
and
CDPF.
Note
that
in
determining
the
combined
control
costs,
the
cost
of
an
oxidation
catalyst
was
excluded
from
the
cost
of
the
NOx
adsorber
since
the
CDPF
include
an
oxidation
catalyst
element.

Table
3:
Combined
NOx
Adsorber
and
CDPF
Control
Costs
Engine
Size
(
HP)
Purchased
Equipment
Cost
Total
Capital
Cost
Total
Annual
Cost
Total
Capital
Cost
($/
HP)
Total
Annual
Cost
($/
HP)

Prime
Emergency
Prime
Emergency
75
$
809
$
1,537
$
242
$
208
$
20
$
3
$
3
135
$
1,234
$
2,345
$
350
$
317
$
17
$
3
$
2
238
$
1,963
$
3,730
$
524
$
503
$
16
$
2
$
2
400
$
3,110
$
5,909
$
817
$
796
$
15
$
2
$
2
750
$
5,588
$
10,617
$
1,473
$
1,428
$
14
$
2
$
2
3000
$
21,518
$
40,884
$
5,587
$
5,498
$
14
$
2
$
2
Average
$
16
$
2
$
2
*
Costs
presented
are
based
on
the
cost
to
buyer
with
nonroad
learning.
**
Costs
do
not
include
the
cost
of
an
oxidation
catalyst
element
for
NOx
adsorbers.
***
Costs
presented
include
the
cost
of
a
regeneration
system
for
engines
with
a
DI
fuel
system.
10
Attachment
A
­
Linear
Regression
Plots
11
NOx
Adsorber
System
Costs
vs.
Horsepower
y
=
2.7512x
+
178.13
R2
=
0.9927
y
=
3.2925x
+
193.71
R2
=
0.9926
y
=
3.97x
+
213.41
R2
=
0.9926
$
0
$
500
$
1,000
$
1,500
$
2,000
$
2,500
$
3,000
$
3,500
$
4,000
$
4,500
0
200
400
600
800
1000
1200
Horsepower
(
HP)

System
Costs
Baseline
Cost
to
Buyer
­­
Long
Term
Cost
to
Buyer
w/
Highway
Learning
­­
Long
Term
Cost
to
Buyer
w/
Nonroad
Learning
­­
Long
Term
12
CDPF
System
Costs
vs.
Horsepower
y
=
3.7416x
+
74.834
R2
=
0.9936
y
=
4.6775x
+
93.459
R2
=
0.9936
y
=
5.8462x
+
117.18
R2
=
0.9936
$
0
$
1,000
$
2,000
$
3,000
$
4,000
$
5,000
$
6,000
$
7,000
0
200
400
600
800
1000
1200
Horsepower
(
HP)

System
Costs
Baseline
Cost
to
Buyer
­­
Long
Term
Cost
to
Buyer
w/
Highway
Learning
­­
Long
Term
Cost
to
Buyer
w/
Nonroad
Learning
­­
Long
Term
13
CDPF
Regeneration
System
Costs
vs.
Horsepower
Engine
with
Direct
Injection
(
DI)
Fuel
System
y
=
0.1753x
+
122.93
R2
=
0.9706
$
0
$
50
$
100
$
150
$
200
$
250
$
300
$
350
0
200
400
600
800
1000
1200
Horsepower
(
HP)

Total
Cost
to
Buyer
­­

Long
Term
w/

Learning
