1
MEMORANDUM
DATE:
June
20,
2005
SUBJECT:
Population
and
Projection
of
Stationary
Engines
FROM:
Tanya
Parise,
Alpha­
Gamma
Technologies,
Inc.

TO:
Sims
Roy,
EPA
OAQPS
ESD
Combustion
Group
The
purpose
of
this
memorandum
is
to
present
the
population
of
existing
and
projected
number
of
new
stationary
compression
ignition
(
CI)
and
spark
ignition
(
SI)
internal
combustion
engines.
This
memorandum
will
describe
how
the
existing
population
and
projected
number
of
new
CI
and
SI
stationary
engines
were
derived.

Introduction
The
population
of
existing
and
projection
of
new
engines
were
based
on
information
from
the
Power
Systems
Research
(
PSR)
1998
North
American
Engine
Partslink
Data
Base.
This
database
was
provided
by
the
U.
S.
EPA's
Office
of
Transportation
and
Air
Quality
(
OTAQ)
who
has
purchased
the
rights
to
use
this
database.
The
database
contains
population
estimates
for
road
and
nonroad
engines.
The
EPA
determined
that
it
would
be
appropriate
to
use
the
data
developed
by
PSR
in
estimating
the
number
of
existing
and
new
CI
and
SI
stationary
engines
for
several
reasons.
These
reasons
were
presented
in
the
memorandum
entitled
"
Decision
on
Using
Estimates
from
PSR
or
the
RICE
NESHAP
Population
Database
for
Engines
above
500
Horsepower,"
which
is
included
as
Attachment
A.

Existing
Population
of
Engines
The
existing
population
of
engines
in
the
U.
S.
was
based
on
information
from
PSR's
database.
The
most
recent
year
for
which
population
information
was
available
from
PSR's
database
was
1998.
The
information
in
PSR's
database
was
separated
by
engine
ignition
type
(
CI
and
SI),
size
range,
and
by
application
type.
The
information
from
PSR
includes
both
mobile
and
stationary
engines
so
it
was
necessary
to
determine
which
applications
can
be
used
for
stationary
purposes
and
distinguish
the
actual
fraction
of
engines
used
in
stationary
applications.
The
following
are
used
in
stationary
applications:
2
°
Generator
Sets,
°
Air
Compressors,
°
Pumps,
°
Welders,
°
Irrigation
Sets,
°
Hydro
Power
Units,
and
°
Gas
Compressors.

For
the
development
of
OTAQ's
nonroad
rule,
a
set
of
fractional
splits
for
mobile
engines
was
established
for
the
PSR
engine
population
information.
The
stationary
fraction
of
engines
was
determined
by
subtracting
the
mobile
fraction
from
1.
The
stationary
fraction
applies
to
the
total
engine
population
as
a
function
of
horsepower
(
HP)
rating.
Table
1
shows
the
stationary
fractions
by
HP
range
used
to
determine
the
number
of
stationary
engines
in
PSR's
database.
The
fractions
shown
in
Table
1
were
also
assumed
to
apply
to
the
projected
number
of
new
engines,
discussed
in
the
next
section
of
this
memorandum.

Table
1.
Stationary
Fraction
of
Engines
HP
Range
Stationary
Fraction
of
Total
Engines
50­
100
0.30
100­
175
0.80
175­
300
0.85
300­
600
0.90
600­
750
1.0
>
750
1.0
An
assumption
was
made
that
there
are
no
stationary
engines
smaller
than
50
HP.
This
is
consistent
with
other
information
EPA
has
reviewed
including
information
gathered
by
the
California
Air
Resources
Board
for
the
proposed
Airborne
Toxic
Control
Measure
for
stationary
CI
engines,
which
showed
that
no
stationary
engines
below
25
HP
were
identified.
In
addition,
other
state
rulemakings
have
used
50
HP
as
a
cutoff
for
stationary
engines.
Therefore,
only
the
stationary
fractions
for
engines
greater
than
50
HP
are
shown
in
Table
1.
The
population
of
stationary
CI
and
SI
engines
broken
down
by
HP
range
is
presented
in
Table
2.
3
Table
2.
Population
of
CI
and
SI
Stationary
Engines
(
as
of
1998)

HP
Range
Population
CI
SI
50­
100
124,064
44,872
100­
175
197,076
53,983
175­
300
130,744
13,583
300­
600
140,058
6,502
600­
750
24,359
1,027
>
750
47,478
5,814
Total
663,780
125,780
Projection
of
New
Engines
Engine
sales
information
from
PSR
and
information
from
the
Engine
Manufacturers
Association
(
EMA)
was
used
to
project
the
number
of
new
CI
and
SI
engines
from
the
years
2006
through
2017.
The
information
from
PSR
did
not
contain
any
data
on
the
population
of
CI
engines
above
3,000
HP,
therefore
EPA
asked
EMA
to
provide
an
estimate
of
the
number
of
new
engines
above
3,000
HP.
Joe
Suchecki
of
EMA
provided
an
estimate
to
Sims
Roy
of
EPA
in
an
email
on
April
28,
2005;
this
email
can
be
found
in
the
docket
for
the
New
Source
Performance
Standards
for
stationary
CI
engines
(
Docket
Number
OAR­
2005­
0029).
The
EMA
estimated
that
approximately
160
stationary
CI
engines
larger
than
3,000
HP
are
sold
each
year.
This
estimate
is
included
in
the
number
of
engines
shown
in
Table
3.

As
with
the
existing
population
of
engines
from
PSR,
the
engine
sales
information
from
PSR
also
included
both
mobile
and
stationary
engines
and
all
applications.
To
determine
the
number
of
stationary
engines
sold,
the
stationary
fractions
listed
in
Table
1
were
applied
to
total
engine
sales
of
applications
used
for
stationary
purposes.
It
was
also
assumed
that
there
would
not
be
many
(
if
any)
stationary
engines
less
than
50
HP
sold.
The
number
of
projected
new
CI
engines
sold
was
estimated
based
on
PSR
engine
sales
from
the
years
1990
through
2000.
The
difference
in
engine
sales
from
one
year
to
the
next
was
calculated
for
each
HP
range
to
obtain
an
average
increase
in
engine
sales
over
the
10
year
period
per
HP
range.
The
average
increase
in
sales
over
the
10
year
period
was
added
to
the
2000
engine
sales
to
project
engine
sales
in
2001.
The
average
increase
in
sales
over
the
10
year
period
was
added
to
the
projected
number
of
engines
sold
in
2001
to
obtain
estimates
of
engine
sales
in
2002,
and
so
on
to
reach
the
2006
through
2017
timeframe.
4
A
similar
methodology
was
followed
to
estimate
the
number
of
new
SI
engines,
although
less
recent
sales
data
were
available
for
SI
engines.
The
most
recent
year
of
sales
available
for
SI
engines
from
PSR's
database
was
1998.
The
projected
number
of
new
CI
engines
and
SI
engines
is
presented
in
Tables
3
and
4,
respectively.

Table
3:
Projected
Number
of
New
CI
Engines
by
HP
Range
from
2006
through
2017
HP
Range
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
50­
100
17,669
18,432
19,194
19,956
20,719
21,481
22,243
23,006
23,768
24,530
25,293
26,055
100­
175
16,292
16,846
17,400
17,954
18,507
19,061
19,615
20,168
20,722
21,276
21,829
22,383
175­
300
15,825
16,506
17,187
17,868
18,549
19,230
19,911
20,592
21,273
21,954
22,635
23,316
300­
600
6,181
6,345
6,510
6,674
6,839
7,004
7,168
7,333
7,498
7,662
7,827
7,991
600­
750
877
894
910
927
943
960
976
993
1,009
1,026
1,042
1,059
>
750
4,150
4,276
4,402
4,529
4,655
4,781
4,908
5,034
5,160
5,287
5,413
5,539
Total
60,994
63,298
65,602
67,907
70,212
72,516
74,820
77,125
79,429
81,734
84,038
86,343
Table
4:
Projected
Number
of
New
SI
Engines
by
HP
Range
from
2006
through
2017
HP
Range
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
50­
100
5,940
6,134
6,327
6,520
6,713
6,906
7,099
7,293
7,486
7,679
7,872
8,065
100­
175
7,499
7,555
7,611
7,667
7,723
7,779
7,836
7,892
7,948
8,004
8,060
8,116
175­
300
1,774
1,844
1,915
1,986
2,056
2,127
2,198
2,269
2,339
2,410
2,481
2,552
300­
600
797
824
851
879
906
933
960
988
1,015
1,042
1,070
1,097
5
600­
750
55
51
48
45
41
38
35
32
28
2
>
750
1,955
2,051
2,146
2,242
2,337
2,433
2,528
2,624
2,719
2,8
Total
18,019
18,459
18,898
19,338
19,777
20,217
20,656
21,096
21,535
21,9
The
engine
sales
information
from
PSR
used
to
project
the
number
of
new
SI
engines
from
2006
through
2017
was
not
broken
down
into
engine
type.
The
EPA
consulted
information
developed
for
the
national
emission
standards
for
hazardous
air
pollutants
(
NESHAP)
for
stationary
reciprocating
internal
combustion
engines
(
RICE).
The
EPA
used
the
breakdown
of
SI
engines
in
the
projected
number
of
new
SI
RICE
developed
for
the
RICE
NESHAP.
That
information
contained
a
breakdown
of
SI
engine
types,
i.
e.,
2
stroke
lean
burn
(
2SLB),
4
stroke
lean
burn
(
4SLB),
and
4
stroke
rich
burn
(
4SRB),
for
engines
greater
than
500
HP.
It
was
assumed
that
the
breakdown
by
engine
type
for
engines
in
the
50
to
500
HP
range
obtained
from
PSR
is
the
same
as
the
breakdown
by
engine
type
for
engines
between
500
and
1,000
HP
obtained
from
information
developed
for
the
RICE
NESHAP.
The
percent
each
SI
engine
type
represents
out
of
the
total
number
of
SI
engines
is
shown
in
Table
5.

Table
5:
Percentage
Each
Engine
Type
Represents
out
of
Total
SI
Engines
HP
Range
Percent
of
Total
SI
Engines
2SLB
4SLB
4SRB
50­
1,000
11%
47%
42%

1,000­
5,000
None
59%
41%

5,000­
10,000
None
50%
50%

The
projected
number
of
new
SI
engines
broken
down
by
engine
type
is
presented
in
Tables
6
through
8.
6
Table
6:
Projected
Number
of
New
2SLB
Engines
by
HP
Range
from
2006
through
2017
HP
Range
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
50­
100
653
675
696
717
738
760
781
802
823
845
866
887
100­
175
825
831
837
843
850
856
862
868
874
880
887
893
175­
300
195
203
211
218
226
234
242
250
257
265
273
281
300­
600
88
91
94
97
100
103
106
109
112
115
118
121
600­
750
6
6
5
5
5
4
4
3
3
3
2
2
>
750
44
46
48
50
52
54
56
58
60
62
64
66
Total
1,811
1,851
1,891
1,931
1,971
2,010
2,050
2,090
2,130
2,170
2,209
2,249
Table
7:
Projected
Number
of
New
4SLB
Engines
by
HP
Range
from
2006
through
2017
HP
Range
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
50­
100
2,792
2,883
2,974
3,064
3,155
3,246
3,337
3,427
3,518
3,609
3,700
3,791
100­
175
3,525
3,551
3,577
3,604
3,630
3,656
3,683
3,709
3,735
3,762
3,788
3,814
175­
300
834
867
900
933
967
1,000
1,033
1,066
1,100
1,133
1,166
1,199
300­
600
375
387
400
413
426
439
451
464
477
490
503
516
600­
750
26
24
23
21
19
18
16
15
13
12
10
9
>
750
1,105
1,159
1,213
1,268
1,322
1,376
1,430
1,485
1,539
1,593
1,647
1,701
7
Total
8,655
8,871
9,087
9,303
9,519
9,735
9,951
10,166
10,382
10,598
10,814
11,030
Table
8:
Projected
Number
of
New
4SRB
Engines
by
HP
Range
from
2006
through
2017
HP
Range
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
50­
100
2,495
2,576
2,657
2,738
2,819
2,901
2,982
3,063
3,144
3,225
3,306
3,387
100­
175
3,150
3,173
3,197
3,220
3,244
3,267
3,291
3,315
3,338
3,362
3,385
3,409
175­
300
745
775
804
834
864
893
923
953
982
1,012
1,042
1,072
300­
600
335
346
358
369
380
392
403
415
426
438
449
461
600­
750
23
22
20
19
17
16
15
13
12
10
9
8
>
750
806
845
884
924
963
1,002
1,042
1,081
1,120
1,160
1,199
1,238
Total
7,553
7,736
7,920
8,104
8,288
8,472
8,655
8,839
9,023
9,207
9,391
9,574
8
ATTACHMENT
A
9
MEMORANDUM
DATE:
June
22,
2004
SUBJECT:
Decision
on
Using
Estimates
from
PSR
or
the
RICE
NESHAP
Population
Database
for
Engines
above
500
Horsepower
FROM:
Melanie
Taylor
and
Chuck
Zukor,
Alpha­
Gamma
Technologies,
Inc.

TO:
Sims
Roy
and
Jaime
Pagán,
EPA
OAQPS
ESD
Combustion
Group
The
purpose
of
this
memorandum
is
to
provide
a
recommendation
of
which
data
source
to
use
for
estimating
the
population
of
existing
and
new
stationary
RICE
above
500
hp
for
both
the
NSPS
and
112(
k)
rule
development.
The
options
are
to
use
the
data
from
the
PSR
database
provided
by
OTAQ,
or
the
estimates
of
existing
and
new
engines
developed
for
the
RICE
NESHAP.
Alpha­
Gamma
has
analyzed
the
1998
PSR
data
provided
by
OTAQ
for
spark
ignition
(
SI)
population
and
sales
and
compression
ignition
population,
and
the
2000
PSR
data
for
CI
sales.

A
comparison
of
the
PSR
data
with
the
RICE
NESHAP
estimates
shows
that
the
estimates
for
CI
RICE
vary
significantly.
The
PSR
data
indicate
approximately
eight
times
more
existing
CI
RICE
above
500
hp
than
accounted
for
in
the
RICE
NESHAP
database.
Projections
of
new
CI
RICE
above
500
hp
based
on
the
PSR
sales
data
are
approximately
3.5
times
those
used
for
the
RICE
NESHAP.
For
SI
RICE,
the
PSR
data
indicate
fewer
existing
engines
above
500
hp
than
the
RICE
NESHAP
database
(~
9,000
for
PSR
and
~
13,000
for
RICE
NESHAP
database).
The
projections
of
new
SI
RICE
above
500
hp
are
fairly
similar
for
both
PSR
and
the
RICE
NESHAP.

Alpha­
Gamma
recommends
using
the
PSR
data
instead
of
the
RICE
NESHAP
estimates
for
the
following
reasons:

°
Alpha­
Gamma's
rationale
for
using
the
RICE
NESHAP
data
above
500
hp
for
the
area
source
population
was
based
primarily
on
using
consistent
data/
methods
in
the
development/
integration
of
the
112(
k)
and
NESHAP
rules.
However,
the
PSR
data
were
not
available
in
time
to
affect
the
RICE
NESHAP
(
which
is
now
final).
Thus,
EPA
may
consider
any
new
and
enhanced
data
sources
to
develop
the
112(
k)
and
NSPS
rules.
10
°
The
inventory
numbers
that
CARB
has
show
more
than
5,000
CI
engines
that
are
greater
than
500
hp,
of
which
4,270
are
above
750
hp
and
are
therefore
based
on
actual
permit
inventory.
Using
the
numbers
for
the
NESHAP,
we
would
approximate
about
14,000
existing
CI
engines
above
500
hp
nationwide.
Comparing
that
to
the
CARB
number
makes
Alpha­
Gamma
believe
the
NESHAP
estimate
is
low,
since
it
is
doubtful
California
has
a
third
of
the
CI
engines
in
the
U.
S.

­
Note
that
CARB
concluded
that
the
PSR
database
underestimated
the
number
of
CI
engines
greater
than
750
hp.
They
said
that
their
district
permit
data
were
assumed
to
be
more
representative
of
the
numbers
of
large
engines
(
greater
than
750
hp)
and
the
PSR
data
were
used
to
allocate
smaller
engines
(
less
than
750
hp).

°
The
NESCAUM
report
has
a
table
of
permitted
CI
engines
in
the
northeast
states
(
actual
engines,
not
PSR
estimates),
and
it
shows
more
than
5,000
engines
above
500
hp.
Adding
that
to
the
CARB
inventory
totals
to
more
than
10,000
existing
CI
engines
just
in
California
and
the
northeast.
Comparing
that
to
the
total
NESHAP
estimate
of
14,000
would
appear
to
justify
the
argument
that
the
NESHAP
estimate
is
not
accurate
and
underestimates
the
population.

One
potential
explanation
for
the
large
difference
in
CI
RICE
population
counts
is
that
the
RICE
NESHAP
Population
Database
is
only
representative
of
permitted
engines.
For
example,
it
is
possible
that
most
of
the
"
unaccounted"
CI
RICE
are
in
emergency
use
applications,
thus
they
do
not
appear
in
permits.
