Nonroad Engine Population Growth
Estimates in MOVES2014b
United States
Environmental Protection
tl Agency
-------
Nonroad Engine Population Growth
This technical report does not necessarily represent final EPA decisions or
positions. It is intended to present technical analysis of issues using data
that are currently available. The purpose in the release of such reports is to
facilitate the exchange of technical information and to inform the public of
technical developments.
Estimates in MOVES2014b
Assessment and Standards Division
Office of Transportation and Air Quality
U.S. Environmental Protection Agency
NOTICE
&EPA
United States
Environmental Protection
Agency
EPA-420-R-18-010
July 2018
-------
Table of Contents
Table of Contents 1
1 Introduction 2
2 Background 2
3 Methodology 3
3.1 Surrogate Data for Proj ecting Future Nonroad Engine Populations 4
3.2 Surrogate Data for Constructing Historical Nonroad Engine Populations 5
4 Results 8
5 Peer Review of Draft Report 15
5.1 Overview of the Peer-Review 15
5.2 General Comments from Robert F. Sawyer 17
5.3 General Comments from Phil Lewis 17
6 References 19
1
-------
1 Introduction
Computing accurate nonroad air pollution emissions inventories depends on estimations of
emission factors and engine activity levels. Changes in nonroad emission factors are driven
primarily by regulations and will not be discussed here (detailed discussions of nonroad emission
factors used in MOVES2014b can be found in Technical Reports NR-009d1 and NR-010f2).
Changes in nonroad engine activity levels over years are the result of complex interactions
between human population growth, changes in national and local economic factors, and changes
in the markets for nonroad engines and products they produce. Because trends in nonroad engine
activity levels are rarely directly measured, MOVES instead starts with base year engine
populations and estimates growth in the populations of nonroad engines while applying constant
annual activity values for every engine type (see Technical Report NR-005d3 for a technical
discussion of annual activity values). This report focuses on the methodology for estimating
growth in nonroad engine populations in MOVES2014b.
2 Background
Previous versions of EPA's nonroad emissions inventory model - including NONROAD2008,
which was added to the MOVES model in 2014 - based engine population growth projections on
a time series analysis of historical (1989 through 1996) nonroad engine populations taken from
the Power Systems Research (PSR) Parts Link database4'5. This database contains detailed
information about each engine family sold in the United States and was used to segregate
nonroad engines by market sector and fuel type. Total engine populations, segregated by fuel
type, were calculated for each year from 1989 through 1996 for the following broad market
sectors: Construction, Agriculture, Industrial, Lawn & Garden, Commercial, Logging, Railroad
Support, Recreational, Recreational Marine, and Airport Service. Future engine populations were
projected by extrapolating from a simple linear regression of historical populations.
Some adjustments were made to the PSR-based approach, including:
Engine population growth in the Oil Field sector was based on the Department of
Commerce's Bureau of Economic Analysis (BEA) estimates of gross state product from
domestic oil production. This data source was preferred over the PSR database for this
sector because the 1989-to-1996 decline in oil field equipment reflected in the PSR
database would have resulted in all oil field equipment disappearing by 2006.
Recreational equipment population growth rates derived from the PSR database were
applied to the Recreational Marine equipment category, but pre-1996 back-casted
populations of personal watercraft were modified to force a zero population in 1970, as it
was assumed that personal watercraft were not available in significant numbers prior to
1970.
Within the Recreational equipment sector, population growth rates for all-terrain vehicles
and off-highway motorcycles were revised based on historical sales information and sales
growth projections supplied by the Motorcycle Industry Council. Additionally, growth
projections of snowmobiles were based on information provided by the International
Snowmobile Manufacturers Association.
2
-------
In addition to forming the basis of nonroad engine population growth rates in MOVES2014,
PSR's engine population database also supplied the model's default base year national
populations, which were used as a starting point to estimate future and past year engine
populations. Population base years varied by engine type and are either 1996, 1998, 1999, or
2000 (see Technical Report NR-006e6 for a technical discussion of the derivation of base year
equipment population estimates). Geographic allocation factors derived from surrogate
information sources (e.g., business activity data, human population data, geographic data) were
used to allocate national engine populations to the county level (see Technical Report NR-014d7
for a detailed discussion of geographic allocation factors).
MOVES2014b will continue to use the base-year engine populations and geographic allocation
factors used in NONROAD2008,MOVES2014, and MOVES2014a. However, to address
concerns that long-term national growth rates derived from seven years of engine population data
limits the model's ability to accurately portray nonroad engine/emissions growth at the regional
or state levels, EPA has developed a set of annual, state-level growth indices for projecting
nonroad engine populations from the population base years.
3 Methodology
EPA considered four primary methods for projecting nonroad engine population growth trends
(Table 3.1)8:
Table 3.1 Methods for projecting nonroad engine activity and population growth trends
Projection Methodology
Description
Equipment activity projections
Tend to focus on a narrow or specialized equipment sectors (e.g., the Federal
Aviation Administration's Terminal Area Forecasts tool for projecting
commercial aviation operations could be used to project airport service
equipment activity)
Census population projections
Considered reliable because population demographics are well understood.
Often used in nonpoint source emission inventory projections where activity
is closely correlated with human population size and could be used in certain
nonroad equipment sectors (e.g., recreational vehicles, residential lawn and
garden).
Economic projections
Sometimes used as surrogates to approximate changes in emissions-
generating activity. Examples of economic data include employment
statistics, gross domestic product, and volumes of product output.
Energy use (fuel consumption)
projections
Fuel consumption data and fuel consumption projections from specific
economic sectors such as construction, agriculture, and mining can be
adopted as surrogates for projecting future nonroad engine
populations/emissions trends in those sectors. This methodology is
appropriate for projecting nonroad engine activity and emissions because
fuel consumed by nonroad engines is proportional to their use levels.
Because the model assigns constant hours-per-year activity rates to each piece of nonroad
equipment, changes in emissions-generating activity levels are instead approximated by
estimating changes in nonroad engine populations. Projections such as those described in Table
3.1 can be adopted as surrogates for projecting future nonroad engine populations.
3
-------
EPA identified sets of projections to serve as surrogates for constructing annual growth indices
from 2014 to 2040 for each nonroad equipment sector (see Section 3.1). The projections are
independent of fuel type. The growth indices function as annual multipliers that are applied to
base year nonroad engine populations in order to estimate the engine population for a given year.
For example, a growth index of 2.0 for a particular year indicates that the engine population in
that year is double that of the base year population. The model then linearly extrapolates, based
on projected engine population estimates for 2039 and 2040, to project populations further into
the future.
EPA also identified historical datasets to serve as surrogates for constructing annual, equipment
sector-specific growth indices from the population base years (1996, 1998, 1999, or 2000) to
2014 (see Section 3.2). The selected historical datasets closely resemble the 2014-2040
projections with which they're mapped (e.g., fuel sales data are matched with fuel consumption
projections; human population data are matched with human population projections).
EPA's methodology for applying annual, sector-specific growth indices to estimate nonroad
engine populations beyond the population base years is summarized in Table 3.2.
Table 3.2 Methodology for applying nonroad engine population growth indices in MOVES2Q14b
Calendar Years
Method
Population base year (1996, 1998, 1999, 2000)
through 2014
Apply annual historical growth indices to base year
populations
2014 through 2040
Apply annual projection growth indices to estimated
populations for 2014
2040 through 2060
Linearly extrapolate from the 2039 and 2040 population
estimates
3.1 Surrogate Data for Projecting I
-------
National projections of revenue ton miles and recreational marine fuel consumption from the
Energy Information Administration (EIA)'s Annual Energy Outlook (AEO) 2016 are used for the
Rail Maintenance12 and Recreational Marine13 sectors, respectively.
Finally, census region (Table 3.4) projections of energy consumption from the EIA's AEO2016
are applied to the Construction14, Agriculture15, Logging16, Oil Field17, and Underground
Mining18'19 equipment sectors. EIA furnishes unpublished energy consumption projections for
specific economic sectors; these sectors are mapped to corresponding nonroad equipment
sectors, as noted in Table 3.3.
Table 3.3 Surrogate data for projecting future (2014-2040) growth of nonroad engine populations20
Equipment Sector
Surrogate Data
Source
Surrogate Data for Future Projections
Industrial
Moody's Analytics
GDP from warehousing sector
Commercial
Moody's Analytics
Economy-wide GDP
Lawn and Garden
(residential and commercial)
U.S. Census Bureau
Human population
Recreational
U.S. Census Bureau
Human population
Airport Service
FAA TAF Model
Number of commercial aviation operations
Rail Maintenance
EIA AEO
Revenue ton miles
Recreational Marine
EIA AEO
Fuel consumption (recreational marine)
Construction
EIA AEO
Energy consumption (construction sector)
Agriculture
EIA AEO
Energy consumption (agriculture sector)
Logging
EIA AEO
Energy consumption (other agriculture sector)
Oil Field
EIA AEO
Energy consumption (oil and gas mining sector)
Underground Mining
EIA AEO
Energy consumption (sum of the coal sector and
metallic & non-metallic mining sector)
Table 3.4 Census regions of the United States21
Census Region
States
West
Alaska, Arizona, California, Colorado, Hawaii, Idaho, Montana, Nevada, New Mexico,
Oregon, Utah, Washington, Wyoming
Midwest
Illinois, Indiana, Iowa, Kansas, Michigan, Minnesota, Missouri, Nebraska, North
Dakota, Ohio, South Dakota, Wisconsin
South
Alabama, Arkansas, District of Columbia, Florida3, Georgia, Kentucky, Louisiana,
Maryland, Mississippi, North Carolina, Oklahoma, South Carolina, Tennessee, Texas,
Virginia, West Virginia
Northeast
Connecticut, Delaware, Maine, Massachusetts, New Hampshire, New Jersey, New York,
Pennsylvania, Rhode Island, Vermont
Note:
a Growth indices developed for Florida are used to estimate growth in nonroad engine populations in Puerto Rico
and the U.S. Virgin Islands.
3.2 Surrogate Data for Constructing Historical Nonroad Engine Populations
A second set of annual growth indices is required in order to project nonroad engine populations
from the population base years (1996, 1998, 1999, or 2000) to 2014. EPA identified publicly
5
-------
available datasets (Table 3.5) to serve as surrogates to estimate historical engine populations,
seeking to match the projection methodologies used for constructing historical populations with
those used to project growth in engine populations beyond 2014. With the exception of the
national-scale revenue ton miles dataset (Oak Ridge National Laboratory (ORNL)'s
Transportation Energy Data Book22) used as growth surrogates for Rail Maintenance equipment,
state-level data are used as historical growth surrogates for all equipment sectors.
Table 3.5 Surrogate data for projecting nonroad engine populations from the base years to 2014
Equipment Sector
Surrogate Data Source
Surrogate Data for Historical
Projections
Industrial
Bureau of Economic Analysis
GDP from multiple economic sectors3
Commercial
Bureau of Economic Analysis
GDP from multiple economic sectors3
Lawn and Garden
(residential)
U.S. Census Bureau
Human population
Lawn and Garden
(commercial)
U.S. Census Bureau
Number of landscaping services
establishments
Recreational
U.S. Census Bureau
Human population
Airport Service
FAA Terminal Area Forecasts
Number of commercial aviation operations
Rail Maintenance
ORLN Transportation Energy Data Book
Revenue ton miles
Recreational Marine
National Marine Manufacturers Association
Boat registrations
Construction
EIA Fuel Oil and Kerosene Sales
Fuel delivered to off-highway
(construction) consumers
Agriculture
EIA Fuel Oil and Kerosene Sales
Fuel delivered to farm consumers
Logging
EIA Fuel Oil and Kerosene Sales
Fuel delivered to off-highway (non-
construction) consumers
Oil Field
EIA Fuel Oil and Kerosene Sales
Fuel delivered to oil company consumers
Underground Mining
EIA Fuel Oil and Kerosene Sales
Fuel delivered to industrial consumers
Note:
a See Table 3.6 for a list of economic sectors used in the analysis.
For the Industrial and Commercial equipment sectors, GDP by state from several industries
(Table 3.6) thought to utilize Industrial and Commercial equipment were obtained from the U.S.
Department of Commerce Bureau of Economic Analysis23. GDP by state from the selected
industries were summed to create the time series of state-level GDP which serve as surrogates
for projecting Industrial and Commercial engine populations from the base years to 2014.
6
-------
Table 3.6 North American Industry Classification System (NAICS)24 economic sectors from which
state-level GDP serve as surrogates for projecting Industrial and Commercial engine populations
NAICS
Sector
NAICS Sector Description
Industrial
Equipment
Commercial
Equipment
11
Agriculture, Forestry, Fishing, and Hunting
X
X
21
Mining, Quarrying, and Oil and Gas Extraction
X
X
22
Utilities
X
X
23
Construction
X
X
31-33
Manufacturing
X
X
42
Wholesale Trade
X
X
44-45
Retail Trade
X
48-49
Transportation and Warehousing
X
X
51
Information
X
53
Real Estate and Rental and Leasing
X
56
Administrative and Support &
Waste Management and Remediation Services
X
X
To maintain consistency with the surrogate data selected to project future Lawn and Garden
(residential) and Recreational Vehicles equipment populations, state-level annual human
population data sourced from the U.S. Census Bureau25 are used as historical growth surrogates
for these equipment categories. Similarly, the number of commercial aviation operations in each
state from FAA's Terminal Area Forecasts model are used as growth surrogates for Airport
Service equipment. The number of landscaping establishments for each state, as reported in the
U.S. Census Bureau County Business Patterns Database (NAICS code 561730)26, serve as
historical growth surrogates for Commercial Lawn & Garden equipment.
State recreational boat registrations for the period 2004-2013, compiled by the National Marine
Manufacturers Association27, are used as historical growth surrogates for the Recreational
Marine equipment category. A simple linear regression of 2004-2013 registrations is used to
construct historical growth indices for the missing periods of 1996-2004 and 2013-2014.
Finally, EPA selected data from EIA's Fuel Oil and Kerosene Sales (FOKS) report28 to serve as
surrogates for projecting engine populations from the base years to 2014 for the Construction,
Agriculture, Logging, Oil Field and Underground Mining equipment sectors. The Adjusted
Distillate Fuel Oil and Kerosene Sales by End Use survey reports different grades of diesel and
distillate fuels for various end use sectors in major economic sectors. EPA mapped FOKS end
use sectors to nonroad equipment sectors and compiled state-level sales data for the reported fuel
grades provided in Table 3.7.
7
-------
Table 3.7 Fractions of reported fuel grades used to construct state-level historical trends,
by FOKS end use sector and nonroad equipment sector
Distillate Fuel Grade (Fraction)
FOKS End Use
Sector
Nonroad
Distillate
Fuel Oil
No. 2 Distillate -
No. 2 Distillate -
No. 1
Distillate
Equipment
Sector
High Sulfur
Diesel
Low Sulfur
Diesel
Off-Highway:
Construction
Construction
1.0
Off-Highway:
Non-Construction
Logging
1.0
Industrial
Underground
Mining
1.0
1.0
0.4
Oil Company
Oil Field
1.0
Farm
Agriculture
1.0
Due to the volatile behavior of the raw sales series, we elected not to use a value for a single year
as the basis from which to forecast future fuels sales. Consequently, we performed aggregation
of the raw series, using a technique common in econometric analysis.
As a first step in constructing historical sales trends for each state, the 5-point weighted-centered
moving average (WCMA5) of the raw sales trend was calculated, with fuel sales expressed as
thousands of gallons (1,000 gal). Five time points (years) were included, i.e. t-2, t-\, t, t+1 and
t+2, with each weighted as 2, 3, 5, 3 and 2, respectively. To obtain the average, the weighted
sales sum was divided by 15, or the total sum of the weights.
In the final year of the series (2014), at which point five time points are not available, the
average was calculated as a 3-point trailing moving average that included points t-2,1-1 and t
(weighted, respectively, as 2, 3 and 5), and with a reduced sum of weights of 10.
After calculating the WCMA5, the average sales were indexed to a population base year of 2000.
However, in some cases, values in specific years were modified to avoid nonsensical or
unreasonable results in some states in some sectors. Specifically, if the WCMA5 in any year and
state in any of the target sectors was between zero and 1,000 gallons, the values were reset to
1,000 gallons.
4 Results
Annual, state-level growth indices for each equipment sector are applied to every equipment type
within the sector, from their respective population base year to 2040. The resulting
nrgrowthindex database contains over 55,000 entries, so in the interest of brevity, a sample of
results are presented here. Figures 4.1- 4.13 show time series of annual growth indices
corresponding to one state from each U.S. Census Region (Table 3.4). Although the model
linearly extrapolates engine populations from 2039 and 2040 for years beyond 2040, for
illustrative purposes, Figures 4.1- 4.13 include linearly extrapolated growth indices. Example
states vary by sector to better illustrate the sector growth. The national average index across all
states is also plotted.
8
-------
Figure 4.1 shows the results from the Industrial equipment sector (1998 base year). Surrogate
data for historical indices is GDP from multiple economic sectors; surrogate data for future
projections is GDP from the warehousing sector (NAICS sector 48-49). Growth indices for
Illinois (Midwest Census Region), Massachusetts (Northeast Census Region), Texas (South
Census Region), and Arizona (West Census Region) are shown.
1998 2003 2008 2013 2018 2023 2028 2033 2038 2043 2048 2053 2058
IL *MA ©AZ TX National Average
Figure 4.1 Annual growth indices from 1998 base year to 2060 for Industrial equipment. The vertical
line at 2014 indicates the transition between historical and projection indices.
Results from the Commercial equipment sector (1998 base year) for Illinois, Massachusetts,
Arizona, and Texas are shown in Figure 4.2. Surrogate data for historical indices is state4evel
GDP from multiple economic sectors; surrogate data for future projections is the economy-wide
state GDP.
1998 2003 2008 2013 2018 2023 2028 2033 2038 2043 2048 2053 2058
IL JKMA 0AZ TX National Average
Figure 4.2 Annual growth indices from 1998 base year to 2060 for Commercial equipment. The
vertical line at 2014 indicates the transition between historical and projection indices.
9
-------
Example growth indices corresponding to the Lawn and Garden equipment sector are shown in
Figure 4.3 (residential lawn and garden equipment) and Figure 4.4 (commercial lawn and garden
equipment). Both plots assume a 1998 population base year and results from Illinois,
Massachusetts, California, and Texas are highlighted. Human population projections are used as
surrogates to project future populations of residential and commercial lawn and garden
equipment. Historical growth indices for residential lawn and garden equipment populations are
based on human population data; the number of landscaping services establishments is the basis
for the historical indices used for commercial lawn and garden equipment.
3.0
nj
a)
>-
CO
ai
<31
o
s
- * "
- -
- -
---
- ** " orr^1
«»- M i*> AiOOP
2.5
2.0
1.5
1.0
0.5
0.0
1998 2003 2008 2013 2018 2023 2028 2033 2038 2043 2048 2053 2058
IL SKMA 0CA TX National Average
Figure 4.3 Annual growth indices from 1998 base year to 2060 for Residential Lawn and Garden
equipment. The vertical line at 2014 indicates the transition between historical and projection indices.
n:
rj
>-
-Q
CO
<31
<31
-o
c
o
U
3.0
2.5
2.0
1.5
- 1.0
0.5
0.0
i
\
\
V
i
** i ..
^
^ ^ - ... |^aaAAhOvv^
1998 2003 2008 2013 2018 2023 2028 2033 2038 2043 2048 2053 2058
IL *MA 0CA TX National Average
Figure 4.4 Annual growth indices from 1998 base year to 2060 for Commercial Lawn and Garden
equipment. The vertical line at 2014 indicates the transition between historical and projection indices.
10
-------
Equipment population growth indices for the Recreational equipment sector (1998 base year) for
Minnesota, Massachusetts, Wyoming, and Texas are shown in Figure 4.5. Human population
data and projections serve as surrogates for both historical and projected equipment populations.
TO
-Q
CO
CT1
Ol
01
¦u
3.0
2.5
2.0
- 1.0
0.5
0.0
-+-
-+-
-+-
-+-
-+-
1998 2003 2008 2013 2018 2023 2028 2033 2038 2043 2048 2053 2058
MN *MA eWY IX National Average
Figure 4.5 Annual growth indices from 1998 base year to 2060 for Recreational equipment. The
vertical line at 2014 indicates the transition between historical and projection indices.
State-level equipment population growth indices for the Airport Service equipment sector (1998
base year) for Illinois, New York, Arizona, and Texas are shown in Figure 4.6. Both the
historical and future projections are based on the number of commercial aviation operations
reported by FAA's TAF model.
fO
¦
_Q
CO
cn
"O
c
1998 2003 2008 2013 2018 2023 2028 2033 2038 2043 2048 2053 2058
¦NY
¦ AZ TX
National Average
Figure 4.6 Annual growth indices from 1998 base year to 2060 for Airport Service equipment. The
vertical line at 2014 indicates the transition between historical and projection indices.
11
-------
The national-scale growth indices for the Railway Maintenance equipment sector (1998 and
2000 base years) are shown in Figure 4.7. National revenue ton miles are the basis for both
historical and future projections.
X
a;
"O
5
o
u
1998 2003 2008 2013 2018 2023 2028 2033 2038 2043 2048 2053 2058
1998 base year O2000base year
Figure 4.7 Annual growth indices from 1998 and 2000 base years to 2060 for Railway Maintenance
equipment. The vertical line at 2014 indicates the transition between historical and projection indices.
Figure 4.8 shows the growth indices for the Recreational Marine equipment sector (1998 base
year) for Minnesota, New York, Washington, and Florida. The pre-2014 historical indices are
based on state-level boat registration data; human population projections are used to project
future populations of Recreational Marine equipment.
0!
i)
>-
-Q
00
01
CT1
T3
C
0
1
'J
2.0
1.8
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
---
f ^ imiuuuinmi nwuir'^- |-| |
1 1 1-
1 1 1 1 1 1 1 1 1
1998 2003 2008 2013 2018 2023 2028 2033 2038 2043 2048 2053 2058
MN 3KNY 9WA FL NationaI Average
Figure 4.8 Annual growth indices from 1998 base year to 2060 for Recreational Marine equipment.
The vertical line at 2014 indicates the transition between historical and projection indices.
Equipment population growth indices for the Construction equipment sector (2000 base year) for
Illinois, New York, Arizona, and Texas are shown in Figure 4.9. Historical indices are based on
12
-------
sector-specific Fuel Oil and Kerosene Sales data, while the future growth indices are based on
projected fuel consumption in the construction sector.
CO
a;
>-
CD
_Q
O
o
o
"O
c
3.0
2.5
2.0
1.5
- 1.0
o
i
0.5
0.0
a
Mr . »
^ y
- y * «
2000 2005 2010 2015 2020 2025 2030 2035 2040 2045 2050 2055 2060
IL NY 0AZ IX National Average
Figure 4.9 Annual growth indices from 2000 base year to 2060 for Construction equipment. The
vertical line at 2014 indicates the transition between historical and projection indices.
Similar to the Construction equipment sector, the Agricultural equipment sector uses sector-
specific fuel sales data and energy consumption projections are surrogates for historical and
future growth indices. Results from Nebraska, Vermont, California, and North Carolina (2000
base year) are highlighted in Figure 4.10.
-------
delivered to off-highway consumers, while the future projections are based on projected energy
consumption in the "other agriculture" sector.
TO
0)
>
ro
-Q
o
o
o
T3
C
2.5
2.0
1.5
1.0
o 0.5
o.o
l" 'li 'f 1' I'»T'lri'iimf*"*"*1
N ^ ~
2000 2005 2010 2015 2020 2025 2030 2035 2040 2045 2050 2055 2060
MN *VT $OR GA National Average
Figure 4.11 Annual growth indices from 2000 base year to 2060 for Logging equipment. The vertical
line at 2014 indicates the transition between historical and projection indices.
Equipment population growth indices for the Oil Field equipment sector (2000 base year) for
South Dakota, Pennsylvania, Alaska, and Texas are shown in Figure 4.12. Projections of energy
consumption in the oil and gas mining sector serve as surrogates for future growth indices, and
fuel sales to oil company consumers are used to construct historical growth indices.
60.0
5 50.0
>¦
(TJ
-Q 40.0
o
o
o
CNj
30.0
- 20.0
o
i
10.0
o.o
I
AK
/J
^^^^^ot,^r,oooooooaoooooo00000000000i>
2000 2005 2010 2015 2020 2025 2030 2035 2040 2045 2050 2055 2060
SD s*PA $AK TX National Average
Figure 4.12 Annual growth indices from 2000 base year to 2060 for Oil Field equipment. The vertical
line at 2014 indicates the transition between historical and projection indices.
Figure 4.13 shows the growth indices for the Underground Mining equipment sector (2000 base
year) for Illinois, Pennsylvania, Wyoming, and West Virginia. Historical indices are based on
14
-------
sales of fuel delivered to industrial consumers, while the future projections are based on
projected energy consumption in the coal sector and metallic & non-metallic mining sector.
03
a;
>
a>
to
03
-Q
o
o
o
-------
of the presentation. For this review, no independent data analysis is required. Rather, we ask
that you assess whether the information provided is representative of the state of current
understanding, and whether incorporating the information into EPA 's MOVES model will result
in appropriate predictions and conclusions.
We request you provide us comments on substantive content sequentially. These will be listed as
an appendix to the final published report, along with EPA 's responses. Comments on
organization, formatting, and other minor issues are welcome, but should be provided
separately.
Below are questions to define the scope of the review; we are not expecting individual responses
to the questions, but would like them to help guide your response.
General Questions to Consider:
1. Does the presentation describe the selected data sources sufficiently to allow the reader
to form a general view of the quantity, quality and representativeness of data used in the
analysis? Are you able to recommend alternate data sources that might better allow the
model to estimate national or regional default values?
2. Is the description of analytic methods and procedures clear and detailed enough to allow
the reader to develop an adequate understanding of the steps taken and assumptions
made by EPA while developing the model inputs? Are examples selectedfor tables and
figures well-chosen and effective in improving the reader's understanding of approaches
and methods?
3. Are the methods and procedures employed technically appropriate and reasonable, with
respect to the relevant disciplines, including physics, chemistry, engineering,
mathematics and statistics? Are you able to suggest or recommend alternate approaches
that might better achieve the goal of developing accurate and representative model
inputs? In making recommendations, please distinguish between instances involving
reasonable disagreement in adoption of methods as opposed to instances where you
conclude that current methods involve specific technical errors.
4. Where EPA has concluded that applicable data is meager or unavailable, and
consequently has made assumptions to frame approaches and arrive at solutions, do you
agree that the assumptions are appropriate and reasonable? If not, and you are able to
do so, please suggest alternative assumptions that might lead to more reasonable or
accurate model inputs.
5. Are the resulting model inputs appropriate and, to the best of your knowledge and
experience, reasonably consistent with physical and chemical processes involved in
mobile source emissions, formation and control? Are the resulting model inputs
empirically consistent with the body of data and literature with which you are familiar?
Specific Questions:
16
-------
In addition to the general review, we request specific responses to the following questions:
1. This report describes a method for aggregating Fuel Oil and Kerosene Sales (FOKS)
data to generate growth indices for various nonroad equipment sectors. Are there better
or alternative methods for aggregating time series data such as the FOKS dataset?
2. Are there data sources EPA selected to serve as surrogates for historical andfuture
growth indices appropriate for the nonroad equipment sectors to which they are
assigned?
5.2 General Comments from Robert F. Sawyer
The methodology outlined to establish the baseline population is a reasonable, but
unfortunate, compromise based on a lack of contemporary data. The database, now 22 years old,
is valuable in that it contains a reasonably detailed breakdown nonroad engines by sector and
fuel. It would appear that there is no provision for electricity as a fuel in any of the sectors.
(Perhaps this will be accounted for in the emission factor part of the model.) In Table 3.1, the
fourth projection methodology is unclear.
RESPONSE: MOVES does not currently account for the relatively small fraction of
nonroad equipment powered by electricity. EPA recognizes that electricity as a fuel for
nonroad mobile sources is growing in market share, and that this penetration will need to
be addressed in future versions of MOVES.
The description for adopting fuel consumption data and projections as surrogates for
estimating growth in nonroad engine populations has been editedfor clarity.
The surrogate data for estimating nonroad engine population growth methodology
outlined in Table 3.3 and Table 3.5 is reasonable. The use of fuel sales to develop population is
an important check on projections based on historical trends, but appears, incorrectly, to assume
that engine efficiencies are constant over time. Statistical methods for smoothing data are
reasonable. Projections to 2060 are, of course, highly uncertain but this is not critical since the
model and base data will certainly continue to be improved periodically. The methodology for
developing state level nonroad engine populations is reasonable.
RESPONSE: We take the reviewer's point, and agree that the projections described here
are not sophisticated enough to explicitly account for projected trends in fuel efficiency.
However, developing this sophistication would require investment of significant effort to
predict an additional uncertain parameter. Given the major uncertainties involved in
long-term projections, it is not clear that these efforts would substantially improve the
utility of the model..
5.3 General Comments from Phil Lewis
General Comments:
17
-------
In my opinion, the reports include an adequate presentation of the methods used to update the
Nonroad model. As a user of previous versions of Nonroad, I frequently referred to the
documents and reports to gain better insight to the application of the model. I believe these
reports provide a sufficient explanation of the assumptions and data used in the updated Nonroad
model. As a potential user of Nonroad in the future, I am satisfied with the explanations given
for the assumptions and data sources provided in the reports. Furthermore, I have no
recommendations for alternate data sources.
Description of the analytical methods are thorough enough for the reader to gain an
understanding of the general approaches used to update the Nonroad model; however, it would
be somewhat difficult to duplicate the methodologies in their entirety. In order to duplicate the
methodologies, I believe more detail is needed, perhaps in the form of sample calculations. I do
not believe, however, that duplication is the primary objective of the reports; therefore, I do not
recommend adding unnecessary details. I also found the examples provided for tables and figures
to be appropriate for their intended use.
I believe the methods and procedures are technically and scientifically sound. This is a
reasonable approach to predict the future of nonroad equipment on a grand scale. I have no
suggestions or recommendations for improving the approach.
When trying to predict the future, no one has a High Definition crystal ball; therefore, we are left
with assumptions. I believe the assumptions chosen to fill in the gaps in data are sufficient
enough to provide reasonable model outputs.
Unfortunately, I have limited expertise with the physical and chemical processes associated with
the formation and speciation of emissions. I believe the model inputs are empirically consistent
and adequate based on my limited knowledge.
Specific Comments:
I believe the FOKS methodology is the most direct approach to estimating growth indices for
some nonroad equipment sectors. It may be possible to refine the approach in an attempt to gain
a higher resolution estimate, although I am not sure the added benefits would justify the extra
effort. Considering the Construction sector as an example, it may be possible to use population
and GDP data (particularly in specific regions) to estimate the growth in market for construction
which would result in the need for more nonroad equipment. Presumably, the FOKS
methodology takes this into account but it is not apparent in the report. I believe for the intended
use of Nonroad (and ultimately MOVES), the FOKS approach as presented in sufficient.
RESPONSE: As tailpipe emissions are created during fuel combustion, we considered
fuel sales and projected consumption to be perhaps the most appropriate surrogate
available to forecast equipment populations. Nonetheless, for future versions of MOVES,
EPA is interested in exploring projection methodologies that utilize multiple data sets
that might provide a more holistic view of nonroad engine activity and population
growth.
18
-------
To the best of my knowledge, the surrogates are appropriate to the sectors they were assigned.
As a general comment, I feel that the approach is rather clever and certainly provides some
insight in an area where it is difficult to gain any at all.
Additional Overall Comments:
I have a minor qualm with the following phrase:
"Because trends in nonroad engine activity levels are never directly measured, ..."
In my opinion, this phrase makes me think that the data has never been available nor will it ever
be available. I agree that heretofore it has been very difficult to acquire real-world activity data
from nonroad equipment; however, with the advent of telematics and portable activity
measurement systems (PAMS), collecting large amounts of activity data is becoming more of a
reality. I realize that future data collection cannot have an impact on this version of Nonroad but
EPA should consider sponsoring large scale data collection, storage, management, and analysis
efforts to gain this missing component. Real-world activity data will help fill the gaps and refine
the assumptions that are inherent in this version of Nonroad.
RESPONSE: We agree that real-world activity data measured by portable activity and
telematics measurement systems could serve to enhance the assumptions and refine the
underlying data in MOVES. Whilst real-world activity data collection efforts in the
nonroad sector are not yet as widespread as in the onroad sector, and require major and
sustained research investments, EPA supports these efforts and intends to incorporate
such datasets into future versions of the model. The opening paragraph has been
modified to indicate that trends in nonroad engine activity are rarely directed measured.
6 References
1 U.S. Environmental Protection Agency (2018). Exhaust and Crankcase Emission Factors for Nonroad
Compression-Ignition Engines. EPA-420-R-18-009. Assessment and Standards Division, Office of
Transportation and Air Quality, U.S. Environmental Protection Agency, Ann Arbor, MI. luly, 2018.
https://www.epa.gov/moves/moves-technical-reports
2 U.S. Environmental Protection Agency (2010). Exhaust Emission Factors for Nonroad Engine
Modeling - Spark-Ignition. EPA-420-R-10-019. Assessment and Standards Division, Office of
Transportation and Air Quality, U.S. Environmental Protection Agency, Ann Arbor, MI. luly, 2010.
htips://tiepis.epa.gov/Exe/ZvPDF.egi?Dockev=P.1.00S \\ !' ,>df
3 U.S. Environmental Protection Agency (2010). Median Life, Annual Activity, and Load Factor Values
for Nonroad Emissions Modeling. EPA-420-R-10-016. Assessment and Standards Division, Office of
Transportation and Air Quality, U.S. Environmental Protection Agency, Ann Arbor, MI. luly, 2010.
h tips: //tie p is. epa. gov/Exe/ZvPDF. eg i ?Docke v=P .1.0081. RY. pdf
4 Power Systems Research (1998). US Parts Link Edition 6.2. St. Paul, MN.
5 U.S. Environmental Protection Agency (2004). Nonroad Engine Growth Estimates. EPA-420-P-04-008.
Assessment and Standards Division, Office of Transportation and Air Quality, U.S. Environmental
19
-------
Protection Agency, Ann Arbor, MI. April, 2004.
https://ti.epis.epa.gov/Exe/ZvPDF.cgi?Dockev=P .1.000 lW2.pdf
6 U.S. Environmental Protection Agency (2010). NonroadEngine Population Estimates. EPA-420-R-10-
017. Assessment and Standards Division, Office of Transportation and Air Quality, U.S. Environmental
Protection Agency, Ann Arbor, MI. July, 2010.
https://nepis.epa.gov/Exe/ZvPDF.cgi7Dockev-P10081T6.pdf
7 U.S. Environmental Protection Agency (2005). Geographic Allocation of Nonroad Engine Population
Data to the State and County Level. EPA-420-R-05-021. Assessment and Standards Division, Office of
Transportation and Air Quality, U.S. Environmental Protection Agency, Ann Arbor, MI. December,
2005. https://nepis.epa.gov/Exe/ZvPDF.cgi?Dockev=P 1004LDX.pdf
8 Wolf, M. and Baker, R. (2016). Summary of Proposed Approaches for Generating Projections of
Nonroad Population Growth Trends: Technical Memorandum. EPA Contract No. EP-C-12-017, Work
Assignment No. 4-14, Eastern Research Group (ERG), May 23, 2016.
9 Moody's Analytics (2016). Gross Domestic Product Forecasts.
10 University of Virginia Weldon Cooper Center for Public Service (2016). Observed and Total
Population for the U.S. and the States, 2010-2014; May 2016 Update.
http://demographics.coopercenter.org/national-population-proiections/
11 Federal Aviation Administration (2015). Terminal Area Forecast (TAF) Model.
12 U.S. Energy Information Administration (2016). Annual Energy Outlook 2016 - Table 7
(Transportation Sector Key Indicators and Delivered Energy Consumption). September 15, 2016.
https://www.eia.gov/outlooks/archive/aeol6/
13 U.S. Energy Information Administration (2016). Annual Energy Outlook 2016 - Table 37
(Transportation Sector Energy Use by Fuel Type Within a Mode). September 15, 2016.
https://www.eia.gov/outlooks/archive/aeol6/
14 U.S. Energy Information Administration (2016). Annual Energy Outlook 2016 - unpublished data
(INDUSA File - Table 6 [Energy and Macroeconomic Profile for Non-Manufacturing Industry -
Construction]). Files provided to ERG on July 5, 2016 for EPA Contract No. EP-C-12-017, Work
Assignment No. 4-14.
15 U.S. Energy Information Administration (2016). Annual Energy Outlook 2016 - unpublished data
(INDUSA File - Table 1 [Energy and Macroeconomic Profile for Non-Manufacturing Industry - Crop
Agriculture]). Files provided to ERG on July 5, 2016 for EPA Contract No. EP-C-12-017, Work
Assignment No. 4-14.
16 U.S. Energy Information Administration (2016). Annual Energy Outlook 2016 - unpublished data
(INDUSA File - Table 2 [Energy and Macroeconomic Profile for Non-Manufacturing Industry - Other
Agriculture]). Files provided to ERG on July 5, 2016 for EPA Contract No. EP-C-12-017, Work
Assignment No. 4-14.
17 U.S. Energy Information Administration (2016). Annual Energy Outlook 2016 - unpublished data
(INDUSA File - Table 4 [Energy and Macroeconomic Profile for Non-Manufacturing Industry - Oil &
20
-------
Gas Mining]). Files provided to ERG on July 5, 2016 for EPA Contract No. EP-C-12-017, Work
Assignment No. 4-14.
18 U.S. Energy Information Administration (2016). Annual Energy Outlook 2016 - unpublished data
(INDUSA File - Table 5 [Energy and Macroeconomic Profile for Non-Manufacturing Industry - Metallic
& Non-metallic Mining]). Files provided to ERG on July 5, 2016 for EPA Contract No. EP-C-12-017,
Work Assignment No. 4-14.
19 U.S. Energy Information Administration (2016). Annual Energy Outlook 2016 - unpublished data
(INDUSA File - Table 3 [Energy and Macroeconomic Profile for Non-Manufacturing Industry - Coal
Mining]). Files provided to ERG on July 5, 2016 for EPA Contract No. EP-C-12-017, Work Assignment
No. 4-14.
20 Wolf, M. and Baker, R. (2016). Develop Final Population Growth Rates and Supporting
Documentation: Technical Memorandum. EPA Contract No. EP-C-12-017, Work Assignment No. 4-14,
Eastern Research Group (ERG), November 15, 2016.
21 U.S. Census Bureau (2010). Geographic Terms and Concepts - Census Divisions and Census Regions.
https://www.census.gov/geo/reference/gtc/gtc census divreg.html
22 Davis, S., Williams, S., and Boundy, R., Oak Ridge National Laboratory Center for Transportation
Analysis (2016). Transportation Energy Data Book: Edition 35. Oak Ridge, TN, October 31, 2016.
http://cta.oml.gov/data/index.shtml
23 U.S. Department of Commerce Bureau of Economic Analysis (2016). Interactive Tables: Gross
Domestic Product (GDP), https://www.bea.gov/regional/index.htm
24 Office of Management and Budget (2017). North American Industry Classification System.
Washington, D.C. https://www.censiis.gov/eos/www/naics/2017:NAICS/ JAICS Manual.pdf
25 U.S. Census Bureau (2016). Population and Housing Unit Estimates Datasets.
https://www.census.gov/programs-survevs/popest/data/data-sets.html
26 U.S. Census Bureau (2016). County Business Patterns Database, https://www.census.gov/programs-
sh rvevs/cbp/data/datasets .htm 1
27 National Marine Manufacturers Association (2016). https://www.nmma.org/
28 U.S. Energy Information Administration (2016). Fuel Oil and Kerosene Sales.
https://www.eia.gov/petroleum/flieloilkerosene/
29 U.S. Environmental Protection Agency (2018). Science Inventory. Peer Review: Nonroad (NR)
Updates to Population Growth, Compression Ignition (CI) Criteria, Toxic Emission Factors and
Speciation Profiles, https://cfpub.epa.gov/si/si public record report.cfm?dirEntrvId=339678
21
------- |