Locomotive Emission Standards
Regulatory Support Document
&EPA
United States
Environmental Protection
Agency
-------
Locomotive Emission Standards
Regulatory Support Document
Office of Mobile Sources
Office of Air and Radiation
U.S. Environmental Protection Agency
United States
Environmental Protection
Agency
EPA-420-R-98-101
April 1998
-------
NOTE TO READERS
This document is a revised version of the December 1997 Regulatory
Support Document that was placed in the public docket at the time of
signature. This version includes editorial revisions, corrections, and some
additional information. The technological feasibility, environmental, and
economic analyses are unchanged in content. The electronic version of
this document also contains tables and figures that were included in the
hard copy of the December 1997 document, but were not included in the
electronic version.
-------
Table of Contents
Chapter/Section Page
1.0 Industry Background 1
1.1 Locomotive Manufacturing 1
1.2 Railroads 1
1.3 Locomotive Remanufacturers and Suppliers 4
1.4 Leasing 5
1.5 Locomotive Safety Regulations 7
1.6 Commercial Role of Railroads 7
1.7 Environmental Impacts of a Modal Shift 10
2.0 Locomotive Background 11
2.1 Energy Supply Sources for Locomotives 11
2.2 Hotel Power 11
2.3 Types of Locomotives and Locomotive Design Features 12
2.4 Maintenance 14
2.5 Replacement Rates 15
2.6 Remanufacturing 17
3.0 Emission Reduction Technology 20
3.1 Locomotive Pollutants 20
3.2 Locomotive Operating Characteristics 21
3.3 Emission Reduction Technologies 22
3.4 Expected Availability of Technologies 43
3.5 Adjustments for Ambient Conditions 47
3.6 Reliability and Durability 48
4.0 Emission Standards and Supporting Analyses 49
4.1 Duty-Cycles 49
4.2 Useful Life 55
4.3 Baseline Emission Rates 59
4.4 Emission Standards 62
4.5 Feasibility and Compliance with Standards 66
-------
5.0 Test Procedures 73
5.1 Locomotive Testing 73
5.2 Test Sequence 75
5.3 Sampling and Analysis 80
5.4 Fuel Quality 85
5.6 Atmospheric Conditions During Testing 86
5.7 Other Issues 90
6.0 Emission Benefits 94
6.1 Methodology 94
6.2 Class I Railroad Analysis 96
6.3 Class II and III Railroad Analysis 100
6.4 Passenger Railroad Analysis 101
6.5 Summary of Environmental Benefits 102
7.0 Costs and Cost-effectiveness 105
7.1 Initial Cost Increase for Locomotives 106
7.2. Incremental Operating Cost Increases 115
7.3. Total Program Costs and Cost-effectiveness 118
APPENDICES 123
APPENDIX A Locomotive Remanufacture Mileage Data from AAR
APPENDIX B Locomotive Emission Data by Throttle Notch
APPENDIX C Calculation of Baseline Locomotive Emission Rates
APPENDIX D Locomotive Smoke Emissions
APPENDIX E Compliance Margins for On-Highway HDDEs
APPENDIX F Supplementary Cost-Effectiveness Data
APPENDIX G Results of SwRI Testing for EPA
APPENDIX H Graphical HC and CO Data
APPENDIX I Environmental Analysis
APPENDIX J Terms and Abbreviations Used in the Rulemaking
APPENDIX K Calculation of Weighting Factors for ABT Credits
APPENDIX L Exclusion of Pre-1973 Locomotives
APPENDIX M NOx Concentrations as a Function of Test Sequence
APPENDIX N 1995 Emission Inventory Data
APPENDIX O Corrections to Environmental Analysis
APPENDIX P Conversion of Emission Rates to g/kW-hr
-------
1.0 Industry Background
The regulations for locomotives and locomotive engines are expected to directly
impact four industries. These industries are: 1) manufacturers of locomotives and loco-
motive engines (original equipment manufacturers (OEMs)); (2) owners and operators
of locomotives (railroads); 3) remanufacturers of locomotives and locomotive engines
(OEMs, railroads, independent remanufacturers); and 4) suppliers of locomotive parts
(OEMs, railroads, independent remanufacturers, parts manufacturers). A brief
overview of these directly impacted industries follows, along with descriptions of the
general economic impact of railroads on the nation and current regulation of railroads.
1.1 Locomotive Manufacturing
Locomotives used in the United States are primarily produced by two
manufacturers: the Electromotive Division of General Motors (EMD) and General
Electric Transportation Systems (GETS). EMD manufactures its locomotives primarily
in London, Ontario and their engines in La Grange, Illinois. The GETS locomotive
manufacturing facilities are located in Erie, Pennsylvania, while their engine
manufacturing facilities are located in Grove City, Pennsylvania. These
manufacturers produce both the locomotive chassis and propulsion engines. They also
remanufacture engines. MotivePower Industries (formerly MK Rail Corporation) has
produced some locomotives using engines manufactured by Caterpillar, Inc. The
Peoria Locomotive Works also manufactures locomotives using Caterpillar Engines,
and Republic Locomotives manufactures locomotives powered by Detroit Diesel
Corporation engines. The Cummins Engine Company, Inc. also produces engines
which may be used in locomotives.
1.2 Railroads
In the United States, freight railroads are subdivided into three classes based
on annual revenue by the Federal government's Surface Transportation Board (STB).
The STB is an adjudicatory body that was formed in 1966 to settle disputes and
regulate the various modes of surface transportation within the U.S. Organizationally
a part of the Department of Transportation, the STB deals with railway rate and
service issues, railway restructuring and various other issues, including classification
of railroads. (STB regulations for the classification of railroads are contained in 49
CFR Chapter X.) In 1994 the STB classified a railroad as a Class I railroad if its
annual revenue was $255.9 million or greater, as a Class II railroad if its annual
revenue is between $20.5 and 255.8 million, and as a Class III railroad if its annual
revenue is less than $20.5 million. The Class I railroads are the nationwide, long-
distance, line-haul railroads which carry the bulk of the railroad commerce. In 1994,
there were 12 Class I railroads operating in the U.S. Due to recent mergers, there are
currently 9 Class I freight railroads operating in the country. Class I railroads
presently operate approximately 21,000 locomotives in the U.S. Class I railroads
-------
account for over 90 percent of the ton-miles of freight hauled annually and consumed
3.60 billion gallons of diesel fuel in 1996, which is about 91 percent of all locomotive
fuel used in the U.S. Of these, the five largest Class I railroads, BNSF, CSX, Conrail,
Norfolk Southern, and Union Pacific, account for the vast majority of the Class I
locomotives currently in service in the U.S. (These five will soon be four, as Conrail is
absorbed by CSX and Norfolk Southern.)
Statistics compiled by the Association of American Railroads (AAR) and the
American Short Line Railroad Association (ASLRA) show that there are approximately
530 Class II and III railroads (not including commuter and insular railroads). A more
detailed breakdown of these can be found in Table 1-1. They consist primarily of
regional and local line-haul and switching railroads1, which operate in a much more
confined environment than do the Class I railroads. For example, the average length
of haul for a regional railroad is 167 miles, while the average length of haul for a local
line-haul railroad is 41 miles (as opposed to an average of about 800 miles for a Class
I). Class II and III railroads operate approximately 4,200 locomotives.2 All but about
400 of these are owned by the operating railroad. Over 85 percent of the locomotives
owned by the Class II and III railroads were originally manufactured prior to 1973.
Class II and III railroads use about 215 million gallons of fuel annually, about 6
percent of the total used by locomotives. These last two facts indicate that Class II and
III railroads are responsible for less than one percent of all emissions from post-1972
locomotives.
Some of the smaller railroads are owned and operated by Class I railroads, many
of which are operated as formal subsidiaries for financial purposes, but are run as
standalone entities. In 1995, there were 222 local line-haul railroads and 253
switching and terminal railroads, including subsidiaries (regional and local railroads
may also have subsidiaries). A few of these are publicly held railroads and some are
shipper-owned. Insular in-plant railroads are not included in this total. ASLRA
estimated that there are probably about 1,000 insular railroads in the U.S. These
railroads are not common carriers, but rather are dedicated to in-plant use. They
typically operate a single switch locomotive powered by an engine with less than 1000
hp. Such locomotives typically use a few thousand gallons of diesel fuel each year, and
thus are not a particularly significant source of emissions.
1 "Regional railroad" and "local railroad" are terms used by AAR that are similar, but not
identical, to "Class II" and "Class III", respectively.
2 "Locomotive Data for Small Railroads," Memorandum, Charles Moulis, U.S. EPA, to
public docket A-94-31, December 5, 1997.
-------
Table 1-1
Profile of Railroad Industry - 1995 3
Type of
Railroad
Class I Freight Railroad
Class I Subsidiaries
National Passenger Railroads
Commuter Railroads4
Regional Railroads
Local/Line-Haul Railroads
Switching & Terminal Railroads
Shipper-Owned Railroads
Government-Owned Railroads
Other/Mixed Ownership5
TOTAL
Number of
Railroads
9
25
1
15
30
222
130
79
36
8
555
Year-end
Employees
185,762
3,570
23,800
22,526
9,115
5,060
1,805
2,369
1,092
905
256,004
3 "Railroad Ten Year Trends", Association of American Railroads, 1995.
4 Does not include all-electric railroads.
5 Does not include insular railroads.
-------
Amtrak is the sole large-scale provider of inter-city passenger transport. In
their comments on the NPRM, Amtrak indicated that their fleet includes 315 diesel-
electric locomotives, consuming 72 million gallons of diesel fuel annually, plus a
number of all-electric locomotives operating in the Northeast corridor.6 They offer
service to 44 states, extending over 23,000 miles of track. With a few exceptions, most
of the trackage upon which Amtrak trains operate is owned by the freight railroads.
Based on gross revenue, Amtrak is classified as a Class I railroad by the STB.
However, unlike the Class I freight railroads, Amtrak's current operating expenses
exceed its gross revenue. There are also 15 independent commuter rail systems
operating in 12 U.S. cities, consuming 61 million gallons of diesel fuel annually.7
Finally, there are a handful of very small passenger railroads that are primarily
operated for tours. These tourist railroads are included within the Class II and III
railroads.
1.3 Locomotive Remanufacturers and Suppliers
While the original manufacturers provide much of the remanufacturing services
to their customers, there are several smaller entities that also provide remanufacturing
services for locomotive engines. Moreover, some of the Class I and II railroads
remanufacture locomotive engines for their own units and on a contract basis for other
railroads. EPA has been able to identify nine independent remanufacturers that are
small business entities. Due to the necessarily limited demand, most of them find it
advantageous to diversify their operations. Many of them remanufacture marine or
other large diesel engines in addition to locomotive engines. A few apparently
remanufacture locomotives primarily for resale or lease, while six of them
remanufacture engines for operating railroads or industrial customers. Many of these
six remanufacturers also manufacture or remanufacture locomotive engine
components. A few also offer contract maintenance. This may be tied to a locomotive
lease, or may be offered separately to owners of locomotives.
In addition to the above original manufacturers and remanufacturers, EPA has
been able to identify fourteen independent suppliers and remanufacturers of locomotive
engine components for use by railroads or by the independent remanufacturers. All
but two are small business entities who produce and/or remanufacture locomotive
engine components. Again, due to the limited size of the market, many of them
produce and/or remanufacture other locomotive components (in addition to engine
components) or components for marine or large industrial engines). Some of them also
serve as locomotive wreckers and deal in used components.
6 Docket item #A-94-31-IV-D-28.
7 "1996 Transit Fact Book", American Public Transportation Association.
-------
It should be noted that railroad equipment is a broad general industrial category
that includes such components as wheels and axles, traction motors, main
generators/alternators, brakes, air compressors, governors and electronic controls and
a host of other products and services related to railroad cars, as well as locomotive
engines and components. It also includes services such as repainting and conducting
required inspections. Many of the suppliers identified also deal in some of these other
components and services as well as in locomotive engine components. None of these
suppliers list railroad equipment as their primary or secondary business activity. The
primary business codes that they do list range from "plating and polishing" to "turbines
and turbine generators". Most list "industrial machinery and equipment", "internal
combustion engine service" or "necessary repair services" as their primary or secondary
business code.
1.4 Leasing
Locomotives are available for lease from OEMs, re manufacturers, and a small
number of specialized leasing companies formed for that purpose. Leasing practices
appear to be fairly standardized throughout the industry. Although lease contracts can
be tailored on an individual basis, most leases seem to incorporate standard boilerplate
language, terms and conditions. Under a typical lease, the lessee takes on the
responsibility for safety certification and maintenance (parts and scheduled service)
of the locomotive (including the engine), although these could be made a part of the
lease package if desired. The lease duration ranged between 30 days and 5 years, with
the average being 3 years.
As can be seen from the Table 1-2, the use of leasing has increased greatly in
recent years among Class I railroads, with almost two-thirds of the locomotives placed
in service in 1994 being leased. Leasing among Class II and III railroads is not nearly
as widespread, with only about 5 percent of the total number of locomotives being
leased.
-------
Table 1-2
Purchase and Leasing of New Locomotives8
Class I Railroads Only
Year
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
Total New
Locomotives
522
280
131
356
609
534
472
323
504
821
Number
Purchased
430
173
110
320
523
417
273
213
379
288
Number
Leased
92
107
21
36
86
117
199
110
125
533
Percent
Leased
18%
38%
16%
10%
14%
22%
42%
34%
25%
65%
"Railroad Ten Year Trends", Association of American Railroads, 1995.
-------
1.5 Locomotive Safety Regulations
Achieving and maintaining the safe operation of commercial (common carrier)
railroads in the U.S. falls under the jurisdiction of the Federal Railroad Administration
(FRA), which is a part of the Department of Transportation. The FRA was created in
1966 to perform a number of disparate functions, including rehabilitating Northeast
Corridor rail passenger service, supporting research and development for rail
transportation, and promoting and enforcing safety regulations throughout the railway
system.
FRA safety regulations apply to railroads on a nationwide basis. These
regulations require a safety inspection of each locomotive used in commercial
operations every 92 days. The inspections are usually performed by the railroad which
owns or leases the locomotive. FRA personnel review the findings of these inspections
and any corrective actions identified and taken. Since each locomotive is required to
be out of revenue service for inspection every 92 days, railroads commonly schedule
their performance of preventive maintenance at these times. It appears likely that
each locomotive is out of service for 12 to 24 hours during each FRA safety inspection
and preventative maintenance period.9 To limit the time that locomotives are out of
service for these safety inspections and preventive maintenance, railroads maintain
suitable facilities distributed across the nation. Thus, it appears that the railroads
have had a long history of compliance with federal regulations, and have developed
strategies to live within the regulations and to minimize any adverse business impacts
that may have resulted.
1.6 Commercial Role of Railroads
Current railroad networks (rail lines) are geographically widespread across the
United States, serving every major city in the country. Approximately one-third of the
freight hauled in the United States is hauled by train. There are few industries or
citizens in the country who are not ultimate consumers of services provided by
American railroad companies. According to statistics compiled by AAR, rail revenue
accounted for 0.5 percent of Gross National Product in 1994. Thus, efficient train
transportation is a vital factor in the strength of the U.S. economy.
9 Values are an approximate estimate by FRA personnel.
7
-------
In order for Class I railroads to operate nationally, they need unhindered rail
access across all state boundaries. If different states regulated locomotives differently,
a railroad could conceivably be forced to change locomotives at state boundaries, and/or
have state-specific locomotive fleets. Currently, facilities for such changes do not exist,
and even if switching areas were available at state boundaries, it would be a costly and
time consuming disruption of interstate commerce. A disruption in the efficient
interstate movement of trains throughout the U.S. could have an impact on the health
and well-being of not only the rail industry, but the entire U.S. economy as well.
Rail transport currently holds about 41% of the market for inter-city freight ton-
miles. The remaining segments of this market are accommodated through other modes
of transport, including trucking (27%), river canal/barge (11.5%) and pipeline (18%) of
total ton-miles. Rail is a primary means of transport for many bulk commodities, such
as chemicals (26%), autos (65%), coal (55%) and grain (26%).10 Being a primary
source/mode of transporting these items, the railroad industry normally sets the
industry standard price ($/ton-mile). Rail transport is typically more fuel efficient and
less expensive than other land-based sources of transport. In terms of BTUs of energy
expended per ton-mile of freight hauled, Department of Energy statistics indicate that
rail transport can be as much as seven times more efficient than truck transport. The
AAR has asserted that one double-stack train can carry the equivalent of 280
truckloads of freight.11
Figure 1-1 and Table 1-3 show the long term growth trends for the amount of
freight carried by Class I railroads and the amount of fuel consumed in carrying that
freight.12 As can be seen from these data, the ton miles of freight carried have more
than doubled, while total fuel consumption has remained relatively constant or
decreased slightly. The reason for this is that locomotive manufacturers have made
continual progress in reducing the fuel efficiency of their engines and the electrical
efficiency of their alternators and motors. It is reasonable to project that the growth
in the amount of freight hauled will continue in the future. It is less certain, however,
whether fuel consumption will increase significantly in the near future. (Note: the
analysis of the environmental and economic impacts of this rulemaking assume no
growth in fuel consumption.)
10
Association of American Railroad.
11 Quoted from May 15, 1997 testimony by Bruce Wilson representing AAR. Docket item
#A-94-31-IV-D-7.
12 "Railroad Facts", Association of American Railroads, 1996 edition.
8
-------
Table 1-3
Annual Fuel Consumption and Revenue Freight
For Class I Railroads
Year
1955
1960
1965
1970
1975
1980
1985
1990
1995
Revenue Freight
(Million Ton-Miles)
623,615
572,309
697,878
764,809
754,252
918,958
876,984
1,033,969
1.305.688
Fuel Consumption
(Million Gallons)
3,384
3,463
3,592
3,545
3,657
3,904
3,110
3,115
3.480
es
Revenue Ton
Figure 1-1 Fuel Consumption and
Revenue Ton-Miles for Class I Railroads
1500
1000
500
0
10
8
6
4
2
0
1955 1960 1965 1970 1975 1980 1985 1990 1995
-B- Ton-Miles -e- Fuel
«
o
S
uel Consumed
Gallons of
9
-------
1.7 Environmental Impacts of a Modal Shift
Another important point which required consideration in the regulation of
locomotives is the potential for a modal shift. A modal shift is a change from one form
of transportation, such as trains, to another form, such as trucks. Modal shift can have
negative or positive effects on national and local emissions inventories. Negative
modal shift occurs when there is a shift to a more polluting form of transportation.
Thus, when negative modal shift occurs, the environment suffers.
Information currently available to EPA shows that truck-based freight
movement generates more pollutants per ton-mile of freight hauled than current,
unregulated rail-based forms of freight movement. Estimates quantifying the
difference indicate that locomotives are on the order of three times cleaner than trucks
on an emissions per ton-mile basis.13 Other studies conducted in the U.S. and Canada
have also concluded that freight transported by rail is three times cleaner than that
transported by truck. For example, Department of Energy data show that HD trucks
produce almost 2.5 times the quantity of NOx emissions as do railroads, but only
account for 75 percent as many ton-miles of freight hauled.14 Thus, any freight
normally carried by rail that is instead hauled by trucks would increase the overall
mass of emissions, even at current emissions rates.
Regulations that were overly stringent could raise equipment and/or operating
costs to the point that it might be a wiser economic choice to move current rail freight
by truck. A disruption in interstate commerce resulting from delays caused by
changing locomotives at state boundaries, due to separate state locomotive regulations,
could be costly to railroad companies. These increased costs would be reflected in the
price of hauling freight by rail, and could even eliminate some rail carriers from the
market. In both of these cases, customers would likely switch to the higher-polluting
trucks for the movement of their freight. Overly stringent regulation of the rail
industry or a disruption in interstate rail movement could therefore result in increases
in the emissions inventory, through a negative modal shift. (Note: as is described in
Chapter 7, EPA estimates that this rule will cost railroads about $80 million annually,
which is about one-quarter of one percent of annual freight revenue for U.S. railroads.)
13 American Society of Mechanical Engineers, Statement on Surface Transportation of
Intercity Freight, May 18,1992.
14 U.S. DOE, Transportation Energy Book, Edition 16 (1994), 1996.
10
-------
2.0 Locomotive Background
2.1 Energy Supply Sources for Locomotives
Locomotives may be subdivided into two general groups on the basis of the
source of energy for powering the locomotive: 1) "all-electric" and 2) "engine-powered".
In the "all-electric" group, externally generated electrical energy is supplied to the
locomotive by means of a catenary or some other electrical transfer device. An example
of this type of locomotive is commonly seen moving commuter trains. Power to operate
the locomotive is not generated by an onboard engine. Emission control requirements
for all-electric locomotives would be achieved at the point of electrical power
generation, and thus are not included in this rulemaking.
In the "engine-powered" group of locomotives, fuel (usually diesel in the U.S.)
is carried on the locomotive. The energy contained in the fuel is converted to power by
burning the fuel in the locomotive engine. A small portion of the engine output power
is normally used directly to drive an air compressor to provide brakes for the
locomotive and train. However, the vast majority of the output power from the engine
is converted to electrical energy in an alternator or generator which is directly
connected to the engine. This electrical energy is transmitted to electric motors
(traction motors) connected directly to the drive wheels of the locomotive for
propulsion, as well as to motors which drive the cooling fans, pumps, etc., necessary
for operation of the engine and the locomotive.15 In the case of passenger locomotives,
electrical energy is also supplied to the coaches of the train for heating, air
conditioning, lighting, etc. (i.e., "hotel power"). In some passenger trains, electrical
energy required for the operation of the passenger coaches is supplied by a locomotive-
mounted auxiliary engine.
2.2 Hotel Power
As noted above, the design of locomotives for use in passenger train service
provides for the locomotive to be operated in either of two distinct modes. In one mode,
the locomotive engine provides only propulsion power for the train. In this mode, the
engine speed changes with changes in power output, resulting in operation similar to
freight locomotives. In the second mode, the locomotive engine supplies electrical
"hotel power" to the passenger cars, in addition to providing propulsion power for the
train. Hotel power provided to the passenger cars can amount to as much as 800 kW
(1070 hp). In contrast to operation in the non-hotel power mode, the engine speed
remains constant with changes occurring in power output when operating in hotel
power mode. Thus, the two modes of operation utilize different speed and load points
15 Essentially all "engine powered" locomotives used in the U.S. employ a diesel engine
and the electrical drive system described. The term "diesel-electric" has therefore become
the most common terminology for these locomotives.
11
-------
to generate similar propulsion power. These differences in speed and load points mean
that locomotive engines will have different emissions characteristics when operating
in hotel power mode than when operating in non-hotel power mode.
2.3 Types of Locomotives and Locomotive Design Features
Locomotives generally fall into three broad categories based on their intended
use. Switch locomotives, typically 2000 hp or less, are the least powerful locomotives,
and are used in freight yards to assemble and disassemble trains, or for short hauls of
small trains. Some larger road switchers can be rated as high as 2300 hp. Passenger
locomotives are powered by engines of approximately 3000 hp, and may be equipped
with an auxiliary engine to provide hotel power for the train, although they may also
generate hotel power with the main engine, as discussed above. Freight or line-haul
locomotives are the most powerful locomotives and are used to power freight train
operations over long distances. Older line-haul locomotives are typically powered by
engines of approximately 2000-3000 hp, while newer line-haul locomotives are powered
by engines of approximately 3500-5000 hp. In some cases, older line-haul locomotives
(especially lower powered ones) are used in switch applications. The industry expects
that the next generation of line-haul locomotives will be powered by 6000 hp engines.
Locomotives also vary with respect to size, similar to the variation in
horsepower. Switch locomotives tend to be about 40 to 55 feet long, while line-haul
locomotives are typically 60 to 70 feet long. Locomotive length is roughly correlated
with engine size, and thus the difference in length has become more significant as
locomotive engines have become larger and more powerful. Locomotive length is also
related to the number of axles that a locomotive has. In the past, the typical
locomotive had four axles (two trucks with two axles each). While, there still are a
large number of four-axle locomotives in service, nearly all newly manufactured line-
haul locomotives have six axles (two trucks with three axles each). There are two
primary advantages of having more axles on the locomotive. First, with the additional
axles, locomotives can be heavier, without increasing the load on each individual axle
(and thus the load on the rail). Second, six-axle locomotives typically have greater
tractive power at low speeds, which can be critical when climbing steep grades. Four-
axle locomotives, however, are somewhat better for higher speed service. Nevertheless,
it appears that the major railroads are attempting to standardize their fleets by
purchasing six-axle locomotives almost exclusively. This is likely to lead to the future
discontinuation of the practice of converting old line-haul locomotives into switch
locomotives, since these larger six-axle locomotives are probably too long to be practical
in most switch applications.
12
-------
One unique feature of locomotives that makes them different than other,
currently regulated mobile sources is the way that power is transferred from the engine
to the wheels. Most mobile sources utilize mechanical means (i.e., a transmission) to
transfer energy from the engine to the wheels (or other point where the power is
applied). Because there is a mechanical connection between the road vehicle engine
and the wheels, the relationship between engine rotational speed and vehicle speed is
mechanically dictated by the gear ratios in the transmission and final drive (e.g., the
differential and rear axle). This results in engine operation which is very transient in
nature, with respect to changes in both speed and load. In contrast, locomotive engines
are typically connected to an electrical alternator or generator to convert the
mechanical energy to electricity. As noted above, this electricity is then used to power
traction motors which turn the wheels. The effect of this arrangement is that a
locomotive engine can be operated at a desired power output and corresponding engine
speed without being constrained by vehicle speed. The range of possible combinations
of locomotive speed and engine power vary from a locomotive speed approaching zero
with the engine at rated power and speed, to the locomotive at maximum speed and the
engine at idle speed producing no propulsion power. This lack of a direct, mechanical
connection between the engine and the wheels allows the engine to operate in an
essentially steady-state mode, in a number of discrete power settings, or notches, which
are described below.
Another design feature unique to railroad locomotive engines is the design and
operation of the throttle. Power settings for railroad engines (throttle position)
generally involve eight discrete positions, or notches, on the throttle gate, in addition
to idle and the dynamic brake function (which will be considered later). Each throttle
notch position is numerically identified, with notch position one being the lowest power
setting, other than idle, and position eight being maximum power. Because of this
design, each notch on the throttle corresponds to a discrete setting on the fuel delivery
system of the engine. These are the only engine power settings at which the locomotive
can operate. The net effect of this method of control is that the engines can operate at
only eight distinct power levels for propulsion, and at idle and dynamic brake.
Railroads and engine manufacturers have, however, indicated the possibility of either
changing the number of throttle notch positions provided or eliminating throttle
notches entirely in the future.
Dynamic braking is another unique feature of locomotives setting them apart
from other mobile sources. In dynamic braking the traction motors act as generators,
with the generated power being dissipated as heat through an electric resistance grid.
While the engine is not generating motive power (i.e., power to propel the locomotive,
also known as tractive power) in the dynamic brake mode, it is generating power to
operate resistance grid cooling fans. As such, the engine is operating in a power mode
that is different than the power notches or idle settings discussed above. While most
diesel-electric locomotives have a dynamic braking mode, some do not (generally switch
locomotives).
13
-------
Another unique design feature of locomotives is the design of the engine cooling
system and procedures used to control engine coolant temperature. Normal practice
in locomotive design has been to mount the radiator on the top of the locomotive and
not to use a thermostat. Control of coolant temperature is achieved by controlling the
heat rejection rate at the radiator. The rate of heat rejection at the radiator can be
controlled by means such as turning fans on and off or employing a variable speed fan
drive, or by controlling the amount of coolant flow to the radiator (using non-
thermostat controls). A related point of difference between road vehicle and locomotive
engine cooling systems is that antifreeze is not generally used in locomotives. Many
factors contribute to this design approach. The very large size of a locomotive engine
causes both a very high degree of difficulty in starting the engine at low temperatures
when the engine has cooled to ambient temperature, and a high probability that
coolant leakage will occur during both warm up and cool down cycles. Because of
problems inherent in starting these engines when allowed to cool to relatively low
ambient temperatures, and the potential for engine damage due to leakage of coolant,
locomotive engines tend not to be shut down for long periods of time, especially during
cold weather. When the engines are shut down, restarts are generally performed
before significant cooling of the engine occurs, to avoid or minimize restart problems
and coolant leakage problems.
The final unique design feature noted here is the manner in which new designs
and design changes are developed. The initial design of any new models/modifications
and production of prototype models are done in much the same manner as is the case
with other mobile sources. Locomotive manufacturers indicated that this process can
be expected to require from 12 to 24 months for significant changes such as those
required to comply with the new Tier 0 standards. Prototype locomotives are typically
sold or leased to the railroads for extended field reliability testing, normally of one to
two years duration. Only after this testing is completed can the new design/design
change be certified and placed into normal production.
2.4 Maintenance
Locomotive maintenance practices also present some unique features. As is the
case with other mobile sources, locomotive maintenance activities can be broken down
into a number of subcategories. Routine servicing consists of providing the fuel, oil,
water, sand (which is applied to the rails for added traction), and other expendables
necessary for day-to-day operation. Scheduled maintenance can be classified as light
(e.g., inspection and cleaning of fuel injectors) or heavy, which can range from repair
or replacement of major engine components (such as power assemblies) to a complete
engine re manufacture. Wherever possible, scheduled maintenance, particularly the
lighter maintenance, is timed to coincide with periodic federally-required safety
inspections, which normally occur at 92-day intervals. Breakdown maintenance, which
may be required to be done in the field, consists of the actions necessary to get a
locomotive back into service. Because of the cost of a breakdown in terms of lost
14
-------
revenue that could result from a stalled train or blocked track, every effort is made to
minimize the need for this type of maintenance. In general, railroads strive to
maintain a high degree of reliability, which results in more rigorous maintenance
practices than would be expected for most other mobile sources. However, the
competitive nature of the business also results in close scrutiny of costs to achieve the
most cost-effective approach to achieving the necessary reliability. This has resulted
in a variety of approaches to providing maintenance.
Maintenance functions were initially the purview of the individual railroads.
Some major railroads with extensive facilities have turned to providing this service for
other railroads, and a few of the smaller railroads also have done the same, in
particular for other small railroads. However, the tendency in recent years has been
toward a diversification of maintenance providers; a number of independent companies
have come into existence to provide many of the necessary, often specialized services
involved (e.g., turbocharger repair or remanufacture). The trend toward outside
maintenance has also been accelerated by the policies of some of the larger railroads
to divest themselves of not only maintenance activities, but ownership of locomotives
as well. The logical culmination of this trend is the "power by the mile" concept,
whereby a railroad can lease a locomotive with all the necessary attendant services for
an agreed-upon rate.
2.5 Replacement Rates
Due to the long total life span of locomotives and their engines, annual
replacement rates of existing locomotives with freshly-manufactured units are very
low. EPA estimated a replacement rate for locomotives and locomotive engines based
on historical data supplied by AAR, and from conversations with manufacturers and
railroad operators. Table 2-1 illustrates the historical replacement rates for
locomotives in the Class I railroad industry. From 1955 through 1995, annual sales
of freshly-manufactured locomotives fluctuated significantly, but have averaged
approximately 500 units over the last ten years. This replacement rate indicates a
fleet turnover time of about 40 years for Class I railroads. Fleet turnover is the time
required for the locomotive fleet to be entirely composed of locomotives that were not
in service as of the base year. Although the percentage of new units in the fleet has
increased significantly since 1994, it is too soon to determine whether this is a general
trend or merely cyclical variation.
15
-------
Table 2-1
Historical Replacement Rates for Locomotives18
Year
1955
1960
1965
1970
1975
1980
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
Total Locomotives in
Service
31,395
29,031
27,780
27,077
27,846
28,094
25,448
24,117
22,548
20,790
19,647
19,364
19,015
18,835
18,344
18,004
18,161
18,505
18,512
5 year average
10 year average
Freshly Manufactured
Locomotives Purchased17
1,097
389
1,387
1,029
772
1,480
200
436
522
280
131
356
609
530
472
323
504
821
928
610
495
Percent of Fleet which were
Freshly Manufactured Units
3.5%
1.3%
5.0%
3.8%
2.8%
5.3%
0.8%
1.8%
2.3%
1.3%
0.7%
1.8%
3.2%
2.8%
2.6%
1.8%
2.8%
4.4%
5.0%
3.33%
2.62%
16 Data are for Class I railroads only.
17 Includes leased vehicles.
16
-------
It should be noted that the above data apply to Class I railroads only. Similar
data for Class II and III railroads were not available. Purchasing practices have
historically been for Class I railroads to buy virtually all of the freshly-manufactured
locomotives sold. As the Class I railroads replace their equipment with freshly-
manufactured units, the older units are either sold by the Class I railroads to smaller
railroads, are scrapped, or are purchased for remanufacture and ultimate resale (or
leasing) by companies specializing in this work. The industry-wide replacement rate
for locomotives would therefore actually be lower than those indicated for the Class I
railroads only. This would mean that the time required for the total locomotive fleet
to turn over would be longer.
Additionally, independent of cyclic changes in the industry, future locomotive
replacement rates could actually decrease. Locomotive manufacturers are now
producing locomotives that have significantly more horsepower than older locomotives.
Railroads have requested this change so that fewer locomotives are needed to pull a
train. Placing more horsepower on a locomotive chassis increases overall train fuel
efficiency. For example, it would be more fuel-efficient to use two 6000 hp locomotives,
rather than three 4000 hp locomotives, to pull the same weight train, because the
weight of an entire locomotive can be eliminated. Thus, whereas three old locomotives
may be scrapped, only two new locomotives may need to be bought as replacements.
On the other hand, the business outlook for the railroad industry has been
improving in the last few years. As railroads have become increasingly cost-
competitive, they are attracting more business. This in turn increases demand for
locomotive power to move the additional freight. Thus, while purchases of new
locomotives may increase in the next few years, these locomotives will likely
supplement, rather than replace, existing locomotives. Moreover, if freight demands
continue to increase, it may become cost-effective to operate locomotives for longer
periods than are estimated here.
2.6 Remanufacturing
Since most locomotive engines are designed to be remanufactured a number of
times, they generally have extremely durable engine blocks and internal parts. Parts
or systems that experience inherently high wear rates (irrespective of design and
materials used) are designed to be easily replaced so as to limit the time that the unit
is out of service for repair or remanufacture. The prime example of parts that are
designed to be readily replaceable on locomotive engines are the power assemblies( i.e.,
the pistons, piston rings, cylinder liners, fuel injectors and controls, fuel injection
pump(s) and controls, and valves). Within the power assemblies, parts such as the
cylinder head in general do not experience high wear rates, and may be reused after
being inspected and requalified. The power assemblies can be remanufactured to bring
them back to as-new condition or they can be upgraded to incorporate the latest design
configuration for that engine. In addition to the power assemblies there are numerous
17
-------
other parts or systems that may also be replaced.18 Engine remanufactures may be
performed either by the railroad that owns the locomotive or by the original
manufacturer of the locomotive. Remanufactures are also performed by companies
that specialize in performing this work.
During its forty-plus year total life span, a locomotive engine could be
remanufactured as many as ten times (although this would not be considered the
norm). Locomotive engine re manufacturing events are thus routine, and are usually
part of the scheduled maintenance. It is standard practice for the Class I railroads
within the railroad industry to remanufacture a line-haul locomotive engine every four
to eight years. Typically newer locomotives, which have very high usage rates, are
remanufactured every four years. Older locomotives usually are remanufactured less
frequently because they are used less within each year. Such remanufacturing is
necessary to insure the continued proper functioning of the engine. Remanufacturing
is performed to correct losses in power or fuel economy, and to prevent catastrophic
failures, which may cause a railroad line to be blocked by an immobile train. The
trend toward higher power locomotives is naturally resulting in a trend of fewer
locomotives per train, thereby increasing the likelihood that a train would become
immobilized by the failure of a single locomotive. Road failures are very costly to the
railroads because the importance of timeliness to their customers, and the difficulty
in getting replacement locomotives to the location of the failure.
When a locomotive engine is remanufactured, it receives replacement parts
which are either freshly-manufactured or remanufactured to as-new condition (in
terms of their operation and durability).19 This includes the emission-related parts
which, if not part of the basic engine design, are also generally designed to be
periodically replaced. The replacement parts are also often updated designs, which are
designed to either restore or improve the original performance of the engine in terms
of durability, fuel economy and emissions. Because of a locomotive engine's long life,
a significant overall improvement in the original design of the parts, and therefore of
the engine, is possible over the total life of the unit. Since these improvements in
design usually occur in the power assemblies (i.e., the components where fuel is burned
and where emissions originate), remanufacturing of the engine essentially also makes
the locomotive or locomotive engine a new system in terms of emission performance.
A remanufactured locomotive would therefore be like-new in terms of emissions
generation and control.
18 Bottom end components, such as crankshafts and bearings, are often remanufactured
only during every other remanufacture event. Remanufacture events that do not include
these bottom end components are sometimes referred to as "partial remanufactures"
19 In some cases, some components are remanufactured by welding in new metal and
remachining the component to the original specifications.
18
-------
While Class I locomotives are remanufactured on a relatively frequent and
scheduled basis of 4 to 8 years, Class II and III locomotives may be remanufactured on
a longer schedule or may not be remanufactured at all. The typical service life of a
locomotive (40 years) is often exceeded by small railroads that continue to use older
locomotives. It is important to note that there is no inherent limit on how many times
a locomotive can be remanufactured, or how long it can last. Rather, the service life
of a locomotive or locomotive engine is limited by economics. For example, in cases,
where it is economical to cut out damaged sections of a frame, and weld in new metal,
an old locomotive may be salvaged instead of being scrapped. Remanufacturers can
also replace other major components such as the trucks or traction motors, to allow an
older locomotive to stay in service. However, at some point, most railroads decide that
the improved efficiency of newer technologies justifies the additional cost, and thus
scrap the entire locomotive. Nevertheless, many smaller railroads, especially
switching and terminal railroads, are still using locomotives that were originally
manufactured in the 1940s.
19
-------
3.0 Emission Reduction Technology
This chapter provides an overview of emission reduction technology which may
be used on locomotives. The first two sections are brief background discussions of
locomotive pollutants and locomotive operating characteristics. The third section is an
overview of technologies which EPA considered with respect to their potential to reduce
emissions from locomotives. The last section is a discussion of those technologies that
EPA believes will be both available and cost-effective by the time of the locomotive
emissions standards are scheduled to take effect.
3.1 Locomotive Pollutants
The emission constituents of greatest concern from locomotive diesel engines are
oxides of nitrogen (NOx), particulate matter (PM) and smoke. NOx is formed at high
temperatures and pressures associated with combustion of fuel in the engine, when
nitrogen in the air combines with available oxygen in the combustion chamber. PM
generally results from incomplete evaporation and burning of the fuel droplets (and
lubricating oil) in the combustion chamber. Thus, PM emissions are generally
associated with low combustion temperatures, inadequate combustion air in the
vicinity of fuel droplets, and fuel impurities. Since NOx formation is associated with
high combustion temperatures and PM is associated with low combustion
temperatures, many technologies and emission compliance strategies which reduce one
tend to increase the other. Unburned hydrocarbons (HC) and carbon monoxide (CO)
are generally emitted at low levels from properly functioning diesel engines due to the
presence of excess oxygen which allows nearly complete combustion of these
intermediate combustion products. Smoke emissions are typically caused by an
inadequate supply of air for combustion during times of engine acceleration, or by low
combustion temperatures. In general, technologies and compliance strategies that
reduce PM emissions tend to reduce HC, CO and smoke emissions as well, although
to differing degrees.
Ambient conditions also affect emission rates from diesel engines. Temperature
and humidity can both affect NOx emission rates as a result of their effect on
combustion temperatures. NOx emissions are higher at lower ambient humidity
levels, and can be higher at higher ambient temperatures. While not as strong a
relationship, the opposite effect is observed for HC and PM emissions which can be
higher with lower combustion temperatures. Low barometric pressures, which occur
at higher altitudes, can also cause higher smoke and PM emissions with some
reduction in NOx.20 At high altitudes, where the air is less dense, the engine draws a
smaller mass of air, which can result in a lower air/fuel ratio. At such conditions, less
oxygen is available for the combustion process, which results in less complete
20 Chaffin, C., Ullman, T., "Effects of Increased Altitude on Heavy-Duty-Diesel
Emissions." SAE Paper 940669, 1994.
20
-------
combustion of hydrocarbons, as well as increased PM and smoke emissions. This loss
in combustion efficiency also results in lower temperatures which generally means
lower NOx. This potential effect is most prevalent in non-compensated naturally-
aspirated engines and less prevalent for turbocharger engines where fuel flow is
adjusted for pressure changes. For naturally aspirated engines, higher altitudes can
actually result in higher peak temperatures due to an increased ignition delay. This
delay is due to a lower density in the cylinder, which leads to less frequent molecular
collisions (between oxygen and fuel), and thus a slower reaction rate. However, the
ignition delay is not observed in engines with sufficient intake air charge.21
Evaporative emissions from diesel engines are insignificant due to the low
volatility of diesel fuel. Diesel engines can have significant crankcase emissions. Since
an engine's piston rings cannot provide a perfect seal with the cylinder walls, a small
fraction of the products of combustion leak past the piston rings into the crankcase.
This flow of material, known as blowby gases, mixes with the mist of lubricating oil
present in the crankcase and must be vented from the crankcase.
3.2 Locomotive Operating Characteristics
Currently, almost all locomotives used in the U.S. are powered by petroleum-
fueled diesel engines. As noted in Chapter 2, power produced by the engine is
converted to electricity in an alternator and is subsequently used to move the
locomotive by means of electric motors at the wheels. This mechanical decoupling of
the engine from the drive wheels is significant, in that it allows locomotives to be
designed such that their engines operate in generally steady-state mode in several
discrete load and speed points. It also allows the engine speed for a given power output
to be optimized for fuel economy (and/or emissions) at that power output level. This
is in contrast to most other vehicles, in which the engine and drive wheels are
mechanically connected through a transmission, allowing highly transient operation.
The crankshaft rotational speed of engines (engine speed) used in highway vehicles is
related to the speed of the vehicle, while engine power output at any engine speed can
vary from zero to the maximum possible at that engine speed.
From the perspective of NOx formation in locomotives at a given power output,
it would be desirable to operate the engine at higher rather than lower speed, since
higher speeds are associated with lower combustion temperatures. This operational
approach is undesirable, however, from the perspective of both fuel efficiency and the
formation of PM. As previously described, locomotive engines are designed to operate
only at specific power output levels and engine speeds, and these engine operating
conditions are decoupled from locomotive speed. With the exception of idle and
dynamic brake, locomotive engines used in freight operations tend to be operated at or
21 Lizhong, S., et al., "Combustion Process of Diesel Engines at Regions With Different
Altitudes," SAE Paper 950857, 1995.
21
-------
close to the maximum power levels achievable at the corresponding engine speed. This
method of operation, while optimizing fuel efficiency and producing relatively low PM
emissions, causes high combustion temperatures and corresponding high NOx
emissions. Changing the basic operating pattern by operating the engine at higher
speeds for given power output levels would tend to reduce combustion temperatures
and thereby NOx formation. The consequences of this approach could, however, be a
deterioration in fuel efficiency, an increase in PM formation, and the need for larger
engines to achieve the same power output, which would tend to further reduce fuel
efficiency.
Finally, because locomotive engines are operated at discrete steady-state
operational points, the design of low emissions control strategies is expected to be
significantly easier than for a engine undergoing more transient operation, such as a
highway truck engine. A locomotive engine will only need to be optimized at about ten
discrete points for low emissions rather than at an infinite set of conditions over the
entire engine torque map. Additionally, at each point the engine operates in an
essentially steady-state condition so that changes in speed and load do not have to be
considered over the entire engine map. Therefore, calibration of parameters such as
aftercooling temperature, fuel injection timing, EGR rate, and others can be better
optimized. Furthermore, a less sophisticated electronic control system is required than
would be necessary for the same emissions control in highway applications.
3.3 Emission Reduction Technologies
This section provides a summary of emission reduction technologies that EPA
considered with respect to their potential applicability to locomotives. The
technologies discussed are broken into the following categories; engine technologies,
exhaust aftertreatment technologies, and changes in fuel (including the use of
alternative fuels) and in lubricating engine oils. Many of these emission reduction
technologies are based on industry experience used to reduce emissions from similar
but smaller engines,22 used in highway trucks since the 1970's. While many of the
emission-control technologies for highway trucks are conceptually applicable to
locomotives and locomotive engines, the design and operation of locomotives and
locomotive engines may reduce the effectiveness of some of these technologies in
locomotive applications. Among the major differences which may potentially impact
the application of these controls to locomotives are the size and design of the engine
(for example, a substantial portion of the current locomotive fleet consists of two-
stroke engines, whereas most heavy-duty truck engines are four-stroke), the engine
operating speeds (locomotive engines tend to have top speeds of just over 1000 rpm,
whereas highway engine top speeds tend to be well over 2000 rpm), and size
constraints of the vehicle and the infrastructure in which it operates (locomotives
22 In many cases the swept volume of a single cylinder of a locomotive engine is as large as
the total swept volume of all the cylinders of a highway diesel truck engine.
22
-------
must often operate in tunnels and on side-by-side tracks). This last difference is
important when considering available space for charge air cooling and aftertreatment
devices.
It is also important to note that some of the technologies discussed in this
section are still in the research and development phase, and may not be available for
locomotives for many years, if at all. Section 3.4 summarizes which of these
technologies EPA believes will be available in the time frames being considered here.
One final note with respect to the technologies discussed in this section is that, while
each is discussed in isolation, the effective use of some of these technologies can be
optimized through the use of other technologies, and adverse effects of some
technologies can be limited or eliminated through the application of other technologies.
Thus, many of the technologies presented here are more appropriately viewed as
components of larger emission reduction systems or strategies.
3.3.1 Engine Technologies
Fuel Delivery System
Fuel injection improvement is an area that has high potential for emission
reductions from locomotive diesel engines. Emissions can be improved by modifying
fuel injection pressure, fuel spray pattern, injection rate and timing, and by the use of
electronics to control injection rate and timing. While each of these changes, taken
separately, can provide emission reductions, these modifications can be used together
to optimize individual pollutant reductions and minimize fuel economy penalties.
Electronic controls have been instrumental in coordinating these individual
modifications for diesel trucks and have played a significant part in allowing diesel
truck engines to meet very stringent EPA emission standards for NOx, PM, HC, and
smoke, while maintaining fuel economy.
The design of the fuel injector nozzle and the pressure applied to the fuel
determines the fuel spray pattern. The spray pattern from the nozzle needs to be
optimized in conjunction with the configuration of the combustion chamber and
induction swirl to achieve emission reductions. The objectives of this optimization
would be to reduce fuel dropout on the surfaces of the combustion chamber, reduce sac
volume (the volume at the tip of the injector that retains fuel after the injection) to
limit end of injection "dribble", improve fuel atomization and achieve more thorough
mixing with the intake air. These potential changes would reduce HC, PM and smoke
emissions by promoting full combustion and reduce NOx emissions by reducing local
"hot spots" in the combustion chamber.
Modified injection timing is expected to be one of the primary strategies used for
meeting the Tier 0 and Tier 1 emission standards, since optimizing injection timing
and duration can achieve significant NOx emissions reductions at minimal cost.
23
-------
Injection timing modifications can reduce NOx emissions by lowering peak combustion
pressures and temperatures. Emissions data generated for AAR indicate that
retarding fuel injection by 4 degrees from factory specifications reduces NOx emissions
by over 30 percent on some locomotives.23 24 The greatest reductions (31 percent) were
seen on a two-stroke cycle switch locomotive. All of the line-haul locomotives tested
showed reductions around 25 percent, while the two passenger locomotives tested
showed reductions of 13 to 18 percent. However, due to the lower combustion
temperatures, an accompanying loss in fuel economy was found. The measured
increases in fuel consumption varied from 0.8 to 2.5 percent, with most falling in the
1.0 to 1.3 percent range. EPA expects that some or all of this fuel economy penalty
could be made up through the use of other technologies discussed in this chapter.
However, given the limited lead time available to comply with the Tier 0 emission
standards, it is expected that the primary efforts of the manufacturers and
re manufacturers will initially go toward emissions compliance, and that efforts to
minimize fuel consumption will only take priority after emissions compliance has been
achieved. Thus, it may take a significant amount of time after the effective dates of the
standards for the fuel economy penalty to be erased.
In addition to the fuel economy impact just discussed, increased particulate and
smoke emissions can result from the use of a significant degree of retarded injection
timing, due to the reduced opportunity for the particles to burn. The actual sensitivity
depends strongly on the shape of the NOx vs. PM tradeoff curve for each engine
configuration. This in turn depends on several design variables which can be modified
to minimize the effect. Lower fueling rates is a simple approach to reducing smoke
emissions with retarded injection timing, although it would result in reduced power
output. Such an approach will likely be used only for those locomotive applications
that would not be significantly impacted by a moderate power reduction.
Increasing the overall injection rate can be used to shorten the duration of the
fuel injection event. This shortened duration allows a delay in the initiation of fuel
injection, similar to the effect of retarded injection timing, causing lower peak
combustion temperatures and reduced NOx formation. Increasing the injection rate
tends to reduce the PM and fuel economy penalties of retarded injection timing,
because the termination of fuel injection is not delayed. However, increased injection
rates mean increased injection pressure, and thus increased loading on the components
which power the injector. A redesign of these parts to maintain durability would
probably be necessary.
23 Locomotive Exhaust Emission Field Tests - Phase I, Association of American Railroads
Report No. R-877, October 1994.
24 Locomotive Exhaust Emission Field Tests - Phase II, Association of American Railroads
Report No. R-885, March 1995.
24
-------
Injection rate shaping is a strategy in which the rate of fuel injection is varied
in a controlled manner over the duration of the injection period to reduce emission
formation. In one approach, a small, early burst of fuel (known as pilot injection)
initially enters the combustion chamber, and injection of the majority of the necessary
fuel is slightly delayed until the fuel in the combustion chamber ignites. Injection rate
shaping can reduce NOx formation, because delaying injection of most of the fuel
lowers peak combustion temperatures, thus reducing NOx formation. One study on
a heavy-duty diesel truck engine showed a 25 percent reduction in NOx at 1000 rpm
by using a pilot injection while still improving fuel consumption and reducing PM and
smoke emissions.25 In another study, a strategy using multiple injections was shown
to be capable of offsetting the negative impacts of timing retard on PM emissions by
maximizing the burning of PM late in the combustion event, while maintaining flame
temperatures low enough to avoid NOx formation.26
Increasing injection pressure improves the atomization of the fuel and increases
the mixing of the fuel with the intake air in the combustion chamber. This
combination of reduced droplet size and improved mixing leads to more complete
combustion and decreased formation of PM. Tests on a high-speed diesel engine
showed PM emissions to be reduced by half, with NOx held constant through timing
retard, when injection pressures were raised from 90 to 160 MPa.27 One drawback of
higher injection pressures is the need to strengthen the fuel-injection system and other
components of the engine to deal with higher injection pressures. Without
strengthening of components, a loss in durability could result. The ability to increase
injection pressure, either on existing engines or on future engines derived from present
designs is highly subject to the existing design. On some engines, such as those which
use unit injectors, the magnitude of the changes necessary should not be extreme. On
other engines, those using separate injection pump, fuel distribution lines and
injectors, the necessary redesign would be more substantive.
Another desirable goal for fuel injectors is to limit sac volume. Sac volume is the
small volume of fuel remaining in the tip of the injector at the end of injection. This
fuel may dribble out of the injector at the end of injection and at low loads, causing
increased HC and PM emissions. At high loads this effect is limited during the
25 "Reduction of Diesel Engine NOx Using Pilot Injection," T. Minami, et al, SAE 950611,
1995.
26 "Reducing Particulate and NOx Using Multiple Injections and EGR in a D.I. Diesel," D.
Pierpont, et al, SAE 950217, 1995.
27 "Effects of Injection Pressure and Nozzle Geometry on D.I. Diesel Emissions and
Performance," D. Pierpont, R. Reitz, SAE 950604, 1995.
25
-------
combustion stroke, because the much higher pressures in the combustion chamber tend
to hold the fuel in the nozzle. However, during the low-pressure exhaust period the
fuel may dribble from the injector and be emitted as HC. Some fuel injection nozzles
have been designed to close the injector spray orifices quickly and completely at the
end of the injection stroke.
At steady-state, smoke emitted by properly operating diesel locomotive engines
is minimal. However, it can occur during periods of engine acceleration, and some
engine designs are more prone to generate smoke at this time than others. Most
typically, the puff of smoke produced during acceleration is caused by "turbo lag,"
which is time required for the turbocharger to reach its appropriate operating speed
after an increase in exhaust flow occurs. As described below, this results in insufficient
air output from the turbocharger for the amount of fuel supplied during accelerations.
Due to pressures from public reaction, locomotive manufacturers have been reducing
visible smoke from the exhaust of properly operating diesel locomotive engines with a
component called a "puff limiter", which limits the rate at which additional fuel is
supplied to the engine during accelerations, and thereby compensates for turbo lag.
The primary disadvantage of a puff limiter is that it causes slower or delayed
acceleration of the engine. This would not be as great a problem for locomotives as for
highway trucks, since locomotive operation is less transient. Costs associated with the
use of a mechanical puff limiter are relatively low. Electronically controlled puff
limiters are more expensive, but should be more effective and less susceptible to
tampering than mechanical designs. As an option, a turbocharger capable of increased
air flows can be used to ensure that enough air is available to avoid reaching the smoke
limit during accelerations. If such a turbocharger were used, a wastegate could be
used during steady-state operation to prevent overcharging of the cylinder. The
wastegate would essentially route some of the exhaust away from the turbine.
Charge Air Compression and Cooling
The typical goal in engine design is to achieve the desired power output and
reliability, while constraining weight and size, at the lowest possible cost. A common
method to achieve an increase in power output without an increase in engine size
(displacement) and weight, is the use of charge air compression (e.g., turbocharging).
A turbocharger uses the energy of the hot exhaust gases to turn a turbine, which in
turn is attached to an air compressor in the intake air stream. This approach increases
the mass of air entering the engine's combustion chambers, which allows more fuel to
be used, which increases power output. However, charge air compression also heats
the intake air, which increases the formation of NOx. Durability also becomes a
significant concern due to elevated temperatures and loads. These problems can be
resolved through the use of charge air aftercooling, which will be discussed shortly.
Charge air compression can be accomplished in various ways, but the dominant
method used for diesel engines is turbocharging. The charge air compression process
26
-------
must be closely linked to the engine fuel control system to ensure that the two work
together properly, thus ensuring that fuel consumption and emissions formation are
minimized, including preventing smoke generation due to turbo lag. The magnitude
of this problem varies between engines produced by different locomotive
manufacturers. As noted above, one method commonly used to address this problem
is to slowly increase the fueling rate following a change in throttle position. Other
methods to address this problem include the use of a variable geometry turbocharger
(VGT), multiple turbochargers, electronic matching of the turbocharger and fuel
injection, or mechanical drive of the compressor (i.e., the use of a supercharger rather
than a turbocharger). A variable geometry turbocharger can significantly reduce the
problems of turbocharger lag. By changing the geometry of the gas flow passages in
the turbine, the response time of the turbocharger can be improved. For instance, one
design uses a sliding mechanism to block part of the exhaust flow passage in the
turbine. Because the same amount of exhaust has to pass through a smaller flow
passage, the exhaust gasses must travel faster which, in turn, causes the turbine to
spin faster. VGTs require slightly more space, are more costly than a conventional
turbocharger. Over a section of the on-highway transient Federal Test Procedure,
particulate reductions of up to 34 percent have been achieved on a heavy-duty diesel
truck engine through the use of a VGT, without increasing NOx emissions.28
The heating that is caused by compression of the charge air increases NOx
emissions. This can be minimized by cooling the charge air after compression. Charge
air cooling also increases the mass of air available for combustion, and may reduce
durability problems associated with high combustion temperatures. While the lower
temperatures can cause some increase in PM emissions, if the decrease in combustion
temperatures is small enough, PM emissions may be unaffected or may even decrease
(due to the additional oxygen available for combustion). One study showed that charge
air cooling resulted in decreased smoke, HC, NOx, and PM emissions and fuel
consumption, especially at high loads where most NOx is created.29
There are two basic types of charge air coolers (or aftercoolers): air-to-liquid
units and air-to-air units. Historically, in air-to-liquid systems, engine coolant has
been used as the cooling medium both for highway trucks and locomotives. The
amount of cooling with an engine coolant system is limited by the relatively high
temperature of the coolant. Air-to-air aftercoolers use a stream of outside air flowing
through the device to cool the charge air. By using ambient air, an air-to-air
28 "Optimization of Heavy-Duty Diesel Engine Transient Emissions by Advanced Control
of a Variable Geometry Turbocharger," A. Pilley, et al, SAE 890395, 1989.
29 "Performance and Emissions Trade-Offs for a HSDI Diesel Engine - An Optimization
Study," Z. Bazari, B. French, SAE 930592, 1993.
27
-------
aftercooler can cool the compressed intake air to a temperature approaching that of the
ambient air, and thus can be more effective in lowering the temperature of the
compressed charge air and thus reducing NOx levels.
Air-to-air aftercoolers are widely used at present in diesel engines for highway
truck operations, and the potential exists for their use on locomotives as well.
However, air-to-air aftercoolers generally rely heavily on ram air30 to get cooling air
into the aftercooler system. While this is not a problem for trucks operating at
highway speeds and which have the engine, the aftercooler and the engine radiator
mounted in the front of the vehicle, it could be more problematic for locomotives. For
multi-locomotive trains, there exists no comparable direct frontal area for introduction
of ram air into the aftercooler systems of locomotive engines. Air scoops or air dams
could be used to divert air into an aftercooler system. Any air scoops or air dams
installed on a locomotive would need to be bidirectional, facing the "front" and the
"rear" of the locomotive. This is because locomotives can and do operate in both
directions. The size and shape of any air scoop system is also limited by the space
constraints on the locomotive itself, and space constraints due to the infrastructure in
which the locomotive must operate. In addition, locomotives often operate at
maximum engine power and low locomotive speed for extended periods of time, which
means that little air would be introduced to an air-to-air aftercooler system due to ram
air. Blower fans could be incorporated to insure that a sufficient quantity of air is
introduced into the aftercooler system. It is possible that the cooling fans already used
on locomotives for engine cooling or the fans used to dissipate the heat generated
during dynamic brake could be made to serve a dual purpose.
Air-to-liquid aftercoolers (e.g., radiators) are the type currently in general use
for locomotive applications, in one of two configurations. As previously noted, one type
uses engine coolant to lower engine intake air temperature to a level near the
operating temperature of the engine coolant. This limits the amount of cooling that
can be accomplished due to relatively high engine coolant temperatures. However, as
a result, the temperature of the engine intake air, and thus the level of emission
control, is somewhat self-regulating and remains relatively constant (near the engine
coolant temperature) over a wide range of ambient temperatures. The overall
effectiveness of this approach might be improved by increasing the overall size of the
heat exchanger area, the air-to-surface ratio in the system, or by using a cooling fluid
with a higher heat capacity.
The second type of air-to-liquid charge air cooling system is an aftercooler using
a coolant system separate from the engine coolant system. Such a system, also known
as split cooling, has been used on some recent production locomotives. Split cooling can
cool engine intake air temperatures almost as effectively as an air-to-air aftercooler.
30 The term ram air is used to denote the air forced through an aftercooler and/or a
radiator by the motion of the vehicle.
28
-------
However, coolant temperatures are tied to ambient temperatures, necessitating
appropriate controls to reduce seasonal variations in intake air temperature. The use
of antifreeze would probably be necessary to prevent freezing of the aftercooler system
under low ambient conditions (which represents a departure from current railroad
practice of not using antifreeze). Introducing a separate liquid system for aftercooling
could also be more complex than either of the other above-mentioned systems. Space
constraints on a locomotive would also need to be considered.
A third type of potential air-to-liquid system is a combination of the two previous
types with charge air being routed only through the separate coolant system
components during high ambient temperature operations, and through both the
separate coolant and engine coolant system components during low ambient
temperatures to prevent overcooling. This more complex approach to control of
overcooling, while theoretically possible, is expected to be a more costly approach than
could be achieved by other means.
Combustion Chamber Modifications
Diesel truck engine manufacturers have achieved significant emission
reductions through changes to the engine's combustion chamber. Combustion chamber
design modifications can also reasonably be expected to provide emission reductions
for locomotive engines. Redesign of the shape of the combustion chamber and the
location of the fuel injector can optimize the motion of the air and the injected fuel with
respect to emission control. Reductions in both NOx and PM are expected to be
achieved with some combustion chamber configurations currently under development.
Compression ratio is another engine design parameter that impacts emission
control. In general, lower compression ratios cause a reduction in NOx emissions and
decreased fuel economy, but also cause an increase in PM emissions, while higher
compression ratios tend to have the opposite effects. Lower compression ratios also
play a part in lowering the loads imposed on the components of the engine and, as a
result, may contribute to the durability of the units. Compression ratios employed in
locomotive engines are typically somewhat lower than those used in truck engines,
although some recent designs have gone to higher compression ratios, possibly for
lower PM emissions, but more likely for fuel economy reasons.
Increasing turbulence in the combustion chamber either as a result of
turbulence in the intake air, by chamber design, or a combination of both (i.e., inducing
"swirl") can reduce PM emissions from diesel engines by improving the mixing of air
and fuel in the combustion chamber. Historically, swirl has been induced in truck
engines by intake air routing and by chamber design. Truck manufacturers are,
however, increasingly using "reentrant" piston designs, in which a lip is formed at the
top of the piston bowl into which the air is compressed, thereby causing controlled
swirl. Manufacturers are investigating the applicability of this chamber design to
29
-------
locomotive engines. A change in materials or a change in the shape of the chamber
may be helpful in adapting this strategy to locomotive use, and the technology holds
promise for application on locomotives.
Combustion chamber design can be further optimized to decrease PM emissions.
The location of the top piston ring relative to the top of the piston has undergone
significant investigation in truck engines. The location of piston rings has been
modified to reduce the crevice volume, i.e, the space between the top ring and the top
of the piston, while retaining the durability and structural integrity of the piston and
piston ring assembly. These changes result in reduced HC and PM emissions. Raising
the top piston ring requires modified routing of the engine coolant around the cylinder
to prevent overheating of the raised ring. Applicability of this approach to locomotive
engines is under investigation.
Some designers are investigating the possibility of adding ceramic materials to
the surfaces of the combustion chamber. Ceramic coatings may provide effective
insulation, allowing more energy to be retained in the products of combustion and
thereby increasing fuel efficiency. Retaining more energy in the combustion chamber
increases peak combustion temperatures, resulting in decreased PM emissions and
increased NOx emissions. When combined with other modifications such as modified
injection timing, rate shaping and reduced fueling rate, the use of ceramics may result
in a reduction in PM without a corresponding increase in NOx emissions or fuel
consumption.
Electronic Controls
Various electronic control systems have been developed and are already in use
on some newer locomotives. EPA expects that the electronic control systems currently
in use will continue to be improved. Use of electronic controls enables designers to
implement much more precise control of the fuel injection system, such as injection
rate shaping and variable injection timing, than is possible with a mechanical system.
This allows designers to achieve improved emission control with little or no fuel
consumption penalty. The contractor cost report estimates that electronic controls
would result in a two percent fuel consumption savings in locomotives, which could
offset the fuel consumption penalty associated with retarded injection timing.31 Also,
electronic controls are necessary for implementation of advanced concepts such as rate
shaping.
31 "Cost Estimates for Meeting the Proposed Locomotive Emission Standards," Engine,
Fuel, and Emissions Engineering, Inc., September, 1997
30
-------
Reduced Oil Consumption
Manufacturers have evaluated, and continue to evaluate means for reducing the
consumption of lubricating ("lube") oil, which would result in a lower lube oil
contribution to PM emissions. The trade-off which is made in reducing oil consumption
is with engine durability. Since durability is a very high priority in locomotive
engines, EPA expects that manufacturers will continue to strive to reduce oil
consumption, but not at the expense of durability. Any significant benefits in PM
control which may be attributable to reduced oil consumption will probably be more of
a longer-term undertaking. This effort may also lead to the eventual disappearance
of two-stroke cycle engines from locomotive use, since four-stroke engines generally are
able to achieve better oil control, resulting in much less oil as a PM constituent.
Intake Air Dilution
Displacing some of an engine's intake air with inert materials is another NOx
reduction strategy. The inert material lowers combustion temperatures by diluting the
mixture in the cylinder and absorbing heat from the burning fuel. The reduced
temperatures caused by intake air dilution, with or without aftercooling of the intake
air, result in lower levels of NOx emissions. Two identifiable methods are exhaust gas
recirculation (EGR) and water injection.
Exhaust gas recirculation uses gases from the exhaust stream to dilute the
combustion mixture. The recirculated exhaust gases absorb a portion of the energy
released during combustion of the fuel, decreasing the peak combustion temperature
and reducing NOx formation and engine power. EGR in gasoline-fueled engines is
most often accomplished by routing a portion of the exhaust stream from the exhaust
system into the intake air. As an alternative, manufacturers may use "internal" EGR
by coordinating the timing of the intake and exhaust valve events so that a portion of
the exhaust gas from the previous combustion event is retained in, or drawn back into,
the combustion chamber. This approach is less costly in terms of hardware
requirements, but also less effective than external EGR. One study on a single
cylinder engine showed that when combined with turbocharging and charge air cooling,
20 percent EGR can reduce NOx by approximately 50 percent without penalties in
smoke or HC emissions.32
While EGR has been successfully applied in gasoline engines, there are some
potential drawbacks in the application of EGR to the diesel engines used in
locomotives. The abrasiveness of the particulate matter in the exhaust stream may
cause accelerated wear in the engine and turbocharger, and PM can also find its way
into the engine lubricating oil. Also, the particulate matter can form deposits on
32 "Combined Effects of EGR and Supercharging on Diesel Combustion and Emissions,"
N. Uchida, et al, SAE 930601, 1993.
31
-------
components of the engine intake system, decreasing the heat transfer capability of the
aftercooler, and potentially decreasing the effectiveness of the turbocharger if
introduced upstream of this unit. In addition, by reducing combustion temperatures
and decreasing the amount of air available for combustion, EGR may cause incomplete
combustion, resulting in increased HC, CO, PM and smoke emissions.
There are several methods of controlling the PM emissions attributed to EGR.
One method is to cool the exhaust gas recirculated to the intake manifold. With EGR
cooling, a much higher amount of exhaust gas can be added to the intake charge. A
small NOx penalty due to increased ignition delay was observed on a truck engine at
light loads, but at high loads, some additional NOx reduction resulted from EGR
cooling.33 Another method to offset the negative impacts of EGR on PM is through the
use of higher intake air boost pressures. By turbocharging the intake air, exhaust gas
can be added to the charge without reducing the supply of fresh air into the cylinder.34
Because locomotive engines generally operate at a discrete number of steady-state
conditions in use, it should be much simpler to optimize the use of EGR than for a
highway application which is typically characterized by highly transient operation.
Particularly through the use of electronics, EGR rates can be optimized with the air
and fueling strategy independently for each notch. Concerns associated with transient
operation, such as those related to the problem excess EGR during decelerations, are
minimal.
One study considered a technology package designed to solve the problems of
minimizing the amount of intake charge displaced by exhaust gas and of fouling the
turbocharger and intercooler.35 This technology package uses a variable geometry
turbocharger (VGT), an EGR control valve, and a venturi mixer to introduce the
recirculated gas into the inlet stream after the intake air is compressed and cooled. In
addition to compressing the intake air, the VGT is used to build up pressure in the
exhaust stream. Once the pressure is high enough, the EGR control valve is opened
and the recirculated gas is mixed in to the high pressure inlet stream. Although the
recirculated gas is cooled, this cooling is kept minimal to prevent both fouling in the
cooler (due to condensation) and a large pressure drop across the cooler.
33
"NOx Reduction Strategies for DI Diesel Engines," Herzog, P., et al, SAE 920470, 1992.
34 "Combined Effects of EGR and Supercharging on Diesel Combustion and Emissions,"
Uchida, N., et al, SAE 930601, 1993.
35 "New EGR Technology Retains HD Diesel Economy with 21st Century Emissions,"
Baert, R., et al, SAE 960848, 1996.
32
-------
Several bypass oil filtration designs exist for diesel truck engines which will
filter the smaller particles (those missed by the primary filtration system) out of engine
oil.36 With bypass filtration, a portion of the oil is run through a secondary unit which
results in better filtration of the oil. This type of filtration system could be used to
minimize negative effects of PM in the oil that is associated with high levels of EGR.
At least one of design claims efficiencies of up to 99% in capturing 1-micron particles.
Another design is capable of removing water, as well as particles less than 1 micron
in size. To accelerate vaporization of impurities and to maintain oil viscosity, a heated
diffuser plate is used in a third design. A low-voltage soot removal device that reduces
the PM in the recirculated gas by 50 to 84 percent has been developed. Engine wear
was shown to be greatly reduced on a test truck engine as a result of this device.
Testing was performed at 30 percent EGR.37 Another strategy for reducing particles
in EGR is to recirculate the exhaust gas after it has passed through a particulate trap.
Traps typically can remove more than 90 percent of particulate matter, whereas some
designs have achieved a 99 percent particle collection efficiency.38 39
A hybrid EGR system is also being studied as a potential solution to durability
problems associated with recirculated diesel exhaust.40 In this system, a small gasoline
engine is used to drive the supercharger for a larger diesel engine. A portion or all of
the gasoline engine exhaust can then be fed into the intake stream of the diesel engine.
Because of the lack of sulfuric acid and the very low carbon content in the gasoline
filtration, a portion of the oil is run through a secondary unit, engine exhaust, the
problems of wear and erosion of parts in the diesel engine associated with EGR are
alleviated. Another bonus of this system is that the boost pressure is independent of
the load and speed of the diesel engine. Therefore, there is more flexibility in
optimizing the emissions and fuel consumption of the diesel engine. The study
referenced above showed that the hybrid EGR system had about the same fuel
consumption as a conventional EGR engine, but with a larger NOx decrease. However,
such systems are far from being ready for practical use.
36 Fleet Owner, "Hardware Report: What's new in... Bypass Filtration," magazine article,
January 1997.
37 "The EGR System for Diesel Engine Using a Low Voltage Soot Removal Device,"
Yoshikawa, H., et al, SAE 930369, 1993.
38 "Reducing Diesel Particulate and NOx Emissions via Filtration and Particle-Free
Exhaust Gas Recirculation," Khalil, N., et al, SAE 950736, 1995.
39 "An Optimization Study on the Control of NOx and Particulate Emissions from Diesel
Engines," Larsen, C., et al, SAE 960473, 1996.
40 "An Elegant Solution for Vehicular Diesel's Emission and Economy - Hybrid EGR
System," Akiyama, M., et al, SAE 960842, 1996.
33
-------
Water injection is a second form of intake air dilution. Water injection works
like EGR, absorbing heat to vaporize the water and heat the resulting steam, which
lowers peak combustion temperatures and decreases NOx formation. Since the water
would be introduced as a liquid, only a very small portion of the charge air volume
would be displaced, and the effects on engine power would not be large. Testing on a
diesel engine has shown a 40 percent reduction in NOx with a water-fuel ratio of 0.5
with only a slight increase in smoke emissions.41 However, water injection imposes
significant technical and regulatory problems. First, it can lead to increases in PM and
smoke emissions. Second, the water can cause corrosion of engine components, and the
use of water with dissolved impurities (such as calcium carbonate) could lead to
deposits in the water injection system and in the engine. Purified water would be
needed to avoid such deposits. Also, insulation in the water tank and injection system
or additives to the water would be necessary to prevent the water from freezing in
winter. Finally, water injection would require the engine operator to refill the water
reservoir periodically.
Emission control systems that require an operator to physically perform
alterations or additions to a system may not be effective in the field in achieving
emission benefits, especially if not performing those acts would not seriously decrease
engine performance. Because the engine could potentially function without the water
as well as it would with it, there may be no incentive for the operator to comply with
the requirement. Because of the large amounts of water that would be needed, water
injection for a locomotive application could possibly require the addition of a "water
car" behind the locomotive, adding additional weight to the train. Additionally,
railroad companies may have to develop infrastructure in the form of purified water
storage tanks in train yards and along track, similar in concept to the water towers
used for old steam powered locomotives.
Turbo-Compounding
Turbo-compounding is the addition of a second power recovery turbine in series
with the turbine of the turbocharger. This second turbine captures and converts some
of the remaining energy in the exhaust gases to useful work. This useful work is
transmitted to the crankshaft by a gear train, thereby increasing overall engine
efficiency. By reducing fuel consumption, addition of such a waste heat turbine would
lead to a corresponding reduction in overall exhaust emissions on a g/bhp-hr basis. A
waste heat turbine may require the application of ceramic engine designs. The
durability of such systems has yet to be addressed for locomotive applications, although
they were successfully used in piston-driven aircraft applications for a number of years.
41 "Reduction of Smoke and NOx by Strong Turbulence Generated During the
Combustion Process in D.I. Diesel Engines," SAE 920467, 1992.
34
-------
Also, the addition of the extra turbine could pose packaging problems for current
locomotives. Costs associated with this approach are expected to be high for the
amount of emission reductions achieved. If utilized, EPA expects that it would only
occur in the long term.
Closed Crankcase
Preventing the discharge to the atmosphere of the mixture of blowby gases and
lubricating oil mist in the crankcase is referred to as "closing the crankcase" and in the
case of gasoline-fueled engines has been achieved by routing the blowby gases to the
engine air intake. EPA regulations applicable to highway diesel engines require the
use of closed crankcase systems only on naturally aspirated diesel-fueled engines. EPA
exempted diesel-fueled truck engines with charge air compression because of the
possibility of blowby gases decreasing the effectiveness of turbochargers and
aftercoolers. For highway diesel engines equipped with charge air compression, closing
the crankcase would depend on the development of designs that protect turbochargers
and aftercoolers from damage. On some turbocharged locomotive engines, crankcase
closure is effectively achieved by routing the blowby gases into the exhaust stream
after the turbocharger. Since exhaust HC emission measurements for such systems
include HC emissions in the blowby gases, this approach can be considered as meeting
the intent of a closed crankcase. There appear to be no negative effects on fuel
efficiency or engine performance.
3.3.2 Exhaust Aftertreatment Technologies
In order to meet EPA's NOx and PM standards for on-highway heavy-duty diesel
engines, manufacturers have investigated exhaust aftertreatment as a supplement to
engine-based emission control technologies. This technology is theoretically
transferrable to diesel engines used in locomotives, giving consideration to space
availability on the locomotive and the durability of the technology. In general,
incorporating exhaust aftertreatment is more expensive than modifying engine
designs, but aftertreatment can result in additional emission reductions beyond those
achievable through changes in engine design. Aftertreatment may also lessen the
trade-offs involved in controlling both NOx and PM emissions while retaining fuel
efficiency. For example, the use of aftertreatment devices to control PM emissions
would provide engine designers more flexibility to focus on reducing NOx formation in
the engine. In general, development efforts for aftertreatment devices for locomotive
applications are behind those efforts associated with improving engine-out
performance.
Work is being done to develop reduction catalysts that would specifically reduce
NOx emissions. These devices, known as selective catalytic reduction (SCR) systems
require an adequate concentration of substances called reductants or reducing agents
that react readily with NOx. Reduction catalyst technology is currently in the
35
-------
development stage for mobile diesel applications, although such catalysts have been
used successfully in industrial applications. The goal of these catalysts is to lower NOx
emissions in the presence of the oxygen-rich exhaust gases characteristic of diesel
engines. These types of catalysts are therefore sometimes called lean-NOx catalysts.
While finding space on the locomotive to install an aftertreatment device is a
concern, the primary challenge in the use of catalyst technology to control NOx from
locomotive diesel engines is the need to provide adequate supply of a reducing agent
in the exhaust stream. Two means of achieving this are the supplemental injection of
reducing agents and a diesel exhaust-NOx catalyst which utilizes the exhaust
hydrocarbons already present as a reducing agent. Both types of catalyst systems
typically make use of zeolite molecular sieves, selectively trapping the molecules of
reactant materials. Zeolite sieves are substances with a crystalline structure capable
of trapping molecules of certain sizes while allowing others to pass through.
The reducing agents most typically considered for injection into the exhaust
stream are urea, ammonia and diesel fuel. Urea and ammonia are effective reducing
agents in industrial applications, achieving high NOx reduction efficiencies in steady-
state operation.42 Developing such catalysts for highway diesel applications has been
difficult due to the transient operating characteristics of highway vehicles, which
require these systems to be effective under frequently varying load conditions. This
would be less of a consideration in designing for locomotive operation, which is more
like a series of steady-state modes. One SCR supplier has claimed that recent
advances in the predictive modeling of emissions at various operating modes and the
development of corresponding reductant metering strategies has resulted in NOx
reductions of 90%, with concurrent PM reductions of 50% and HC reductions of 85%.43
Relying on urea or ammonia for effective catalyst operation also raises
regulatory concerns. These chemicals are consumed to achieve NOx reduction, so
vehicle operators would need to maintain an adequate supply on their vehicles.
Operators may have no practical incentive for maintaining an adequate supply of the
required reducing agents unless the control systems were designed such that running
out of the reducing agent would cause vehicle performance to be seriously degraded.
Without such a safeguard there would be little incentive to keep an adequate supply
of reducing agent in a locomotive, since the lack of reductant would otherwise be
transparent to the operator. One must not only consider the initial cost of the hardware
and the ongoing cost of the reducing agent, but also the cost of the reducing agent
refilling infrastructure that must be developed to support locomotives utilizing such
technology.
42 "Catalytic NOx Reduction in Net Oxidizing Exhaust Gas," W. Held, SAE 900496, 1990.
43 Docket items A-94-31-IV-D-8 and A-94-31-IV-E-3.
36
-------
The use of ammonia and urea as reductants on locomotives poses another
potential problem because exhaust gas temperatures from locomotive engines in most
throttle notches may be lower than the minimum required for proper operation of the
catalyst. Usage could be limited to high loads where NOx is formed, or another
suggested approach to solving this problem involves the injection of additional diesel
fuel into the exhaust gases for combustion, thereby increasing exhaust gas
temperatures to levels sufficient for the reduction reactions to occur. In such a design,
a small amount of diesel fuel is sprayed into the exhaust stream upstream of the
catalytic converter in metered amounts, corresponding to engine NOx output levels.
NOx reduction efficiencies of 30 to 80 percent have been reported in steady-state
experimental systems.4445 The reported NOx reduction efficiency of 80 percent
corresponded with a 5 percent loss in fuel economy. Fuel used for this purpose would
obviously have negative effects on fuel efficiency, increasing operating costs and
possibly raising HC, CO, PM and smoke emissions. (Fuel efficiency decreases, because
fuel that is injected for NOx reduction does not produce power output from the engine.)
Using unburned diesel fuel as the injected reducing agent would, however, tend to
resolve the concern over operator participation because the fuel would always be
available on a locomotive.
Chemical aftertreatment with cyanuric acid is also being investigated as a
method to reduce NOx emissions. As exhaust gas passes over cyanuric acid pellets, the
pellets give off cyanic acid gas (HCNO), which reacts with NOx to form nitrogen,
carbon dioxide and water. California has identified this as a "best available
technology" for stationary sources which have largely steady-state operation with
relatively gradual changes in output power levels. This type of operation is similar in
some ways to that for locomotive engines.
Aftertreatment devices such as oxidation catalysts may be considered for use on
locomotives to reduce PM emissions. PM emissions from diesel engines are composed
of carbonaceous particles, a soluble organic fraction, sulfates and adsorbed water.
Oxidation catalysts tend to reduce the soluble organic fraction and have little effect on
the carbonaceous portion of PM in diesel exhaust, since soluble organic fractions arise
from unburned fuel and lubricating oil. This limits the reduction in PM emissions that
an oxidation catalyst can achieve. Furthermore, oxidation catalysts convert a portion
of the sulfur dioxide present in the exhaust stream to sulfate PM. Because the
increased sulfate PM can offset the reduction in the soluble organic fraction, an
oxidation catalyst may not be as effective as expected in reducing total PM emissions,
44 "Catalytic Reduction of NOx and Diesel Exhaust," S. Sumiya, et al, SAE 920853, 1992.
45 "Catalytic Reduction of NOx in Actual Diesel Engine Exhaust," M. Konno, et al, SAE
920091, 1992.
37
-------
especially when high sulfur fuel is used. A major goal in the development of oxidation
catalysts for diesel engines is therefore to increase oxidation of the soluble organic
fraction while having little if any effect on the oxides of sulfur present. Optimized
designs (addressing operating temperature, exhaust gas flow path, selection of catalyst
materials, and other parameters) appear capable of reducing total PM emissions from
diesel engines using low sulfur fuel by 20 to 30 percent.4647. Actual reductions are very
dependent on the magnitude of the soluble organic fraction (compared to the
carbonaceous portion) from engine-out exhaust. The mandatory use of low sulfur diesel
fuel in locomotives would have to be considered if catalyst use became widespread.
The durability of diesel engine catalytic converters is unproven for large diesel
engines such as are used to power locomotives. However, experience gained from light
and medium heavy-duty truck and urban bus applications supports the expectation
that oxidation catalysts with sufficient durability can eventually be developed for
larger, more durable engines used in locomotives.
Another type of aftertreatment is the particulate trap oxidizer (trap) which
filters particulate matter from the exhaust stream with subsequent oxidation of the
filtered particulate. The basic element of a trap system is a structural shell containing
the filter material. Filters currently under development for truck and bus engines are
either ceramic wall-flow monolith filters or filter tubes covered with multiple layers of
a yarn-like ceramic material. The filter material contains many small holes that allow
the exhaust gases to pass through while collecting the particulate from the raw
exhaust.
The particulate matter collected by the filter eventually needs to be removed.
It is generally burned off in a process called regeneration. Two general approaches for
regeneration of the trap have been investigated. One approach employed is the use of
catalytic material on the filter which causes regeneration once a predetermined trap
loading is reached. The other approach includes a system for heating the filter to
oxidize the particulate, a microprocessor for controlling filter regeneration and, in some
systems, a supplemental air supply system.
46 "Effects of Sulfate Adsorption on Performance of Diesel Oxidation Catalysts,"
N. Harayama, et al, SAE 920852, 1992.
47 "Technical Feasibility of Reducing NOX and Particulate Emissions From Heavy-Duty
Engines," Acurex Environmental Corporation, April 30, 1993, p. 3-38.
38
-------
3.3.3 Changes in Fuel and Lubricating Oil
Diesel fuel
There are some changes that could be made to the composition of diesel fuel to
reduce emissions, at the expense of requiring the railroads to use new and more costly
fuels. Also, the possible changes to diesel fuel tend to be interdependent. For example,
an increase in a fuel's cetane number is usually associated with a decrease in aromatic
content and an increase in volatility. It is therefore difficult in some cases to determine
separately the effects of changing individual parameters.
The cetane rating is a measure of the tendency of a fuel to autoignite. The effect
of raising the cetane rating of diesel fuel is that it increases autoignition of the fuel in
the combustion chamber, generally improving combustion. A fuel's cetane rating can
be increased either with a fuel additive that enhances autoignition, or through
modified processing of diesel fuel at the refinery. Currently, the nationwide average
cetane number is approximately 44 for highway diesel fuels48. A typical gallon of diesel
fuel also consists of 20 to 45 percent aromatic hydrocarbons by volume. Decreasing the
aromatic content of diesel fuel, which is closely correlated with increased cetane rating,
seems to have a greater potential for decreasing both PM and NOx emissions than
changing fuel composition in other ways. In one study reducing the aromatic content
of diesel fuel used in truck engines from 40 to 20 percent, and increasing the cetane
number from 44 to 53 resulted in a 4 percent reduction in NOx and a 7 percent
reduction in PM49. In another study, a 10 percent reduction in PM emissions was
reported as a result of reducing the aromatic content from 40 to 10 percent.50
Sulfur occurs naturally in crude oil and, unless removed, also occurs in refined
diesel fuel. Two basic emission problems are associated with sulfur in diesel fuel.
First, the sulfur in the fuel reacts to form oxides of sulfur (SOx), including about 3
percent of the fuel sulfur that is directly emitted as particulate in the form of sulfuric
acid.51 SOx can also react in the atmosphere to form sulfate PM. Second, for vehicles
equipped with particulate traps or catalysts, fuel sulfur can cause deterioration in the
substrate materials, decreasing the effectiveness and durability of the trap or catalyst.
48 "National Fuel Survey, Diesel Fuel, Summer 1992," and "National Fuel Survey, Diesel
Fuel, Winter 1992," Motor Vehicle Manufacturers Association, 1992.
49 "Diesel Fuel Property Effects on Exhaust Emissions from a Heavy-Duty Diesel Engine
That Meets 1994 Emission Requirements", McCarthy, et al, SAE 922267, 1992.
50 "Effects of Fuel Aromatics, Cetane Number and Cetane Improver on Emissions from a
1991 Prototype Heavy-Duty Diesel Engine," T.L. Ullman, et al, SAE 902171, 1990.
51 "Cost-Effectiveness of Diesel Fuel Modifications for Particulate Control," M.C. Ingham
and R.B. Warden, Chevron, SAE 870556, 1987.
39
-------
Modifying diesel fuel to increase its volatility (i.e., tendency to evaporate) would
decrease the time required for each fuel droplet to evaporate in the combustion
chamber. More rapid and complete evaporation would result in more complete
combustion of the fuel and thus reduced PM emissions. However, the resultant
increased combustion pressures would probably require stronger engine components
to maintain durability at current levels. The effect of increased volatility on NOx is
also not well understood. It should be noted that the small increases in the volatility
of the diesel fuel referred to here would not lead to significant evaporative emissions.
Derivatives of vegetable oils or animal fats can be mixed with diesel fuel for
combustion in a diesel engine. Such fuel mixtures, known as biodiesel, have received
attention in Europe as a potential source of renewable fuel. Also, Congress identified
biodiesel as an alternative fuel in the National Energy Policy Act of 1992. The
biodiesel fuel mixture includes oxygen and, as with oxygenate additives, would likely
lead to a decrease in PM emissions while risking an increase in NOx emissions. The
costs and emission-reducing potential of these changes are not well understood or
quantified at the present time.
Additional modifications to diesel fuel can also reduce diesel exhaust emissions.
First, increasing the kinematic viscosity of diesel fuel has been correlated with reduced
PM emissions.52 Second, addition of detergents or other chemicals may reduce the
formation of deposits that impair precise control of fuel flow and can lead to increased
emissions.
Alternative Fuels
The use of fuels other than diesel fuel has received much interest in the context
of on-highway vehicles. The alternative fuels that have received the most attention are
natural gas (stored in both compressed gas and liquefied states), alcohols (methanol
and ethanol), and liquefied petroleum gas. Currently, only natural gas has received
any serious consideration as a potential fuel for locomotives. Thus, although other
fuels could someday be available for use in locomotives, this discussion is limited to
natural gas.
Natural gas is made up primarily of methane, but also tends to contain some
ethane, propane, butane and trace amounts of inert ingredients. It is a gas at standard
conditions, which requires that storage onboard a vehicle be approached differently for
natural gas than for liquid fuels. The current approach most utilized for on-highway
vehicles is to store it as a compressed gas in high pressure cylinders. However, the
high fuel usage rate of locomotives is problematic, due to compressed natural gas'
extremely low energy density (i.e., amount of energy per unit volume of fuel), and thus
52 "Description of Diesel Emissions by Individual Fuel Properties," Noboru Miyamoto,
Hokkaido University, et al, SAE 922221, 1992.
40
-------
the space needed to store sufficient fuel. The second approach to storing natural gas
is to liquefy it and store it cryogenically in insulated containers. While much more
practical for locomotives, even this approach results in a much lower energy density
than diesel fuel, and a separate tender car for the fuel would be required for a liquefied
natural gas-powered locomotive which is traveling any great distance.
In contrast to diesel fuel's ability to self-ignite in a compression ignition engine,
natural gas has a very high resistance to autoignition. In order to use natural gas in
a locomotive engine an ignition source must be provided. The two general approaches
to providing the required ignition source are diesel fuel pilot injection and spark
ignition.
The GasRail USA late-cycle, high-injection-pressure project is an example of the
former approach, and has shown NOx emission reductions of up to 75 percent with no
increases in HC and CO emissions.53 Under the pilot injection approach, a premixed
charge of natural gas and air is introduced into the combustion chamber. A small
amount of diesel fuel is then injected into the combustion chamber. The diesel fuel
autoignites, which in turn ignites the natural gas. Diesel fuel pilot injection systems
can be designed to use different ratios of natural gas to diesel fuel, but are generally
designed to maximize the use of natural gas. As a result, such systems typically use
natural gas for well over 90 percent of the energy needs of the locomotive. However,
they usually operate on pure diesel fuel at idle and low load operating points where
fuel consumption is low.
The second approach to igniting natural gas is to use a spark ignition system
similar to that used on gasoline engines. Since the spark ignition approach is a
dedicated natural gas system, it has the advantage of allowing an engine to be
optimized for natural gas. However, it is a much more significant departure from a
standard diesel engine than the pilot injection system. The pilot injection system is
essentially the addition of a natural gas fueling system to a diesel engine. As such, the
pilot injection system is much more suited to the retrofit of existing diesels than is the
spark ignition system.
53 Alternative Fuels Today, July 15, 1997.
41
-------
Data show that natural gas strategies can be used to achieve significant
reductions in NOx and PM. Over the line-haul duty cycle, a dual-fuel locomotive
engine achieved NOx and PM levels of 4.2 and 0.33 g/bhp respectively.54 A dedicated
spark-ignition natural gas-fueled locomotive engine was shown to be capable of levels
as low as 2.0 g/bhp-hr NOx and 0.09 g/bhp-hr PM over a three mode cycle.55 However,
the HC and CO emissions did not meet the Tier 2 levels for either of these engines.
The long term durability of natural gas engines in locomotive use remains unknown.
Also, the use of natural gas would require that new refueling facilities be installed at
any rail yard which services natural gas-fueled locomotives. One potential area of
application would therefore be in switching and terminal operations.
Lubricants
Reductions in PM emissions can be obtained both by changes in the composition
of lubricating oil and by a reduction of the oil consumption rate. Refiners are
conducting research to formulate such lubricating oils that form less PM for truck
engines. Some of this research may or may not be applicable to locomotive engines,
since the additives used in lubricants for locomotive engines are different than those
used in truck engines. One possibility is to replace the metal additives commonly used
in lubricating oil with nonmetallic compounds in order to reduce the noncombustible
(ash) portion of the oil.
The use of synthetic oils or partial synthetic oils may reduce formation of PM
emissions. Conventional lubricating oil formulations evaporate over a wide range of
temperatures. The portion that evaporates at lower temperatures may diffuse into the
combustion chamber, increasing PM emissions. Synthetic oils can be formulated to
evaporate over a narrow, high-temperature range. Using such synthetic oils would
reduce the oil contribution to PM emissions. Partial synthetic oils, made by displacing
most of the more volatile portion of a conventional oil with synthetic material, may
yield equivalent results. Evaporated oil components would also be present to some
extent in the blowby gases introduced into the locomotive exhaust. Since HC exhaust
emissions from locomotive engines are already very low, and the measured values may
also include blowby gases, EPA does not believe that contributions from present oil
formulations to HC exhaust emissions are a problem.
54 Fritz, S., "Exhaust Emissions from a Dual Fuel-Locomotive; Final Report," Southwest
Research Institute, March 1992.
55 Comments from W.C. Passie at Caterpillar to Docket A-94-31, "Emission Standards for
Locomotives and Locomotive Engines," June 10, 1997.
42
-------
3.4 Expected Availability of Technologies
The preceding sections were intended to provide a brief overview of the possible
technologies that manufacturers of locomotives and locomotive engines could employ
in complying with the requirements of this rule making. It should be emphasized,
however, that these are not mandated technologies and that manufacturers and
remanufacturers will develop their own optimized emission control strategies. EPA
has also determined that many of the technologies summarized above are not likely to
be feasible for use in the near term, at least for the Tier 0 and Tier 1 standards. This
section discusses the emission control strategies EPA expects to be available and cost-
effective at the time the locomotive emissions standards take effect. Additional
information on locomotive technology and costs is included in the docket.5657
Tier 0 Locomotives
Tier 0 locomotives are those originally manufactured from 1973 to 2001.
Historically, two designs have dominated the locomotive engine market. The first is
a two-stroke engine design using uniflow-scavenging and unit injection. The second
is a four-stroke engine design using unit pump injection. Both designs are
turbocharged and aftercooled for line-haul applications. There are also a few thousand
Roots-blown58 two-stroke engines and naturally-aspirated four-stroke engines used in
switch locomotives.
Locomotives currently equipped with turbocharged engines will be able to
employ modified/improved fuel injectors, enhanced charge air cooling, injection timing
retard, and in some cases improved turbochargers, to reduce NOx emissions.
Moreover, it will be practical and cost-effective to equip some of these locomotives with
electronic controls as a means of avoiding the penalty in fuel efficiency often associated
with injection timing retard. Within the category of improved fuel injectors,
modifications are expected to include, injection rate changes, modifications to the
injector spray patterns, and reduced sac volumes. Remanufacturers could also use
limited modifications to the piston design, and enhanced smoke controls. Some of these
technologies are already available. In 1994, two-stroke engines using electronically
controlled unit fuel injectors were introduced into service in passenger locomotives
56 Acurex Environmental, "Locomotive Technologies to Meet SOP (sic) Emission
Standards," Prepared for the U.S. Environmental Protection Agency, Contract No. 68-C5-
0010, August 13, 1997.
57 Engine, Fuel, and Emissions Engineering Inc., "Cost Estimates for Meeting the
Proposed Locomotive Emission Standards," Prepared for the U.S. Environmental Protection
Agency, September 12, 1997.
58 A Roots-blower is a positive displacement pump driven by the crankshaft which is used
to force air into the combustion chamber.
43
-------
used in California. The electronics package is similar to a design used in heavy-duty
on-highway diesel engines. Also in 1994, electronic fuel injection was offered as an
option in four-stroke locomotive engines. Although there were some reliability
problems with the initial system, these problems have been resolved.
Enhanced charge air cooling will also be available, most likely taking the form
of improvements in existing air-to-liquid aftercooling systems. EPA expects that
engine coolant will continue to be employed as the cooling medium in most cases,
rather than a separate cooling system. Although this charge-air cooling should not be
too difficult to implement, some additional aftercooling hardware may be necessary
such as an additional pump and radiator and whatever lines and fittings may be
necessary to reroute the engine coolant. In the late 1980's a four-pass aftercooler was
introduced on some two-stroke designs, which proved to be more effective than the
older two-pass design. EPA anticipates that it will be cost-effective to replace nearly
all remaining two-pass aftercoolers with four-pass aftercoolers during the
remanufacturing process.
Overall, it appears likely that re manufacturers will achieve as much cooling of
charge air as possible, improve fuel systems and combustion chambers, while using
timing retard to the least extent possible so as to minimize negative effects on fuel
consumption. Some four-stroke engines may require improved turbochargers to
overcome problems with smoke during acceleration.
In the case of naturally-aspirated and Roots-blown engines, the tools available
to manufacturers for reducing emissions are modifications to the fuel system,
modifications to the combustion chamber and injection timing. In theory, these
engines could be retrofitted with turbochargers and charge air coolers, which if fueling
rates were not changed, would not increase power ratings and would lower peak
combustion temperatures and thereby NOx formation. Most of these locomotives are
employed in switching and terminal applications. It is probable that many of these
will be replaced with lower horsepower line-haul units as the latter are replaced with
newer, higher power units, per the current railroad practice.
Tier 1 Standards
Tier 1 locomotives are those that will be manufactured in 2002 through 2004.
Tier 1 locomotives will be able to incorporate the technologies outlined above for the
Tier 0 locomotives, but these technologies will likely be more effective in the Tier 1
locomotives because more optimization will be possible when they are included in the
original design than is possible with retrofit technology. There are additional
approaches that should also be available; these are discussed below.
44
-------
Several recent locomotive models have already shown relatively low emissions
of HC, CO, and PM, and will be able to achieve significant NOx reductions through
minor incremental changes. These changes include a comprehensive emission
management system consisting of optimized fuel injection strategies through the use
of electronic controls and incorporation of separate circuit aftercooling. Separate
cooling systems, which have been introduced in recent years, have resulted in
enhanced effectiveness of the aftercooler. In these systems, separate circulation pumps
and radiators are used for the aftercooler rather than for the engine's power
assemblies. Therefore, the coolant to the aftercooler is not subject to the heat added
by the engine.
Other technologies can be applied to many locomotive engines in addition to
electronic controls and enhanced aftercooling. Through the use of electronics and
enhanced aftercooling, further timing retard can be used to reduce NOx without a
negative impact on PM. Additional technologies that will be available for some models
include in-cylinder and turbocharger modifications. Changes in the configuration of
the combustion chamber and piston ring location may begin to appear in engines
complying with the Tier 1 standards. Increased compression ratios could be used to
reduce PM emissions and ignition delay. In addition, upgraded turbocharger designs
would help reduce smoke emissions by providing an improved response to transience.
In the case of switch locomotives, two approaches appear to be available to
manufacturers. One approach would be the continued use of large displacement
naturally aspirated engines employing electronic control of the fuel system, improved
fuel injection and improved combustion chambers. Another approach would be to use
turbocharging and other technologies used on line-haul locomotives, but with a
reduction in engine size to achieve the desired lower power rating. A reduction in
engine size could be achieved either through the use of fewer power assemblies of the
same configuration as those used on line-haul locomotives or by the use of a different
engine design than that used in line-haul applications. Locomotive manufacturers
could also use large nonroad engines (1000-2000 hp) that were originally designed for
use in non-locomotive applications. As with all other design choices, manufacturers
will base final decisions on costs (both initial and fuel usage), on durability, and on
serviceability and maintainability with respect to the availability of parts common to
both switch and line-haul locomotives. There is also the possibility, given the ability
of railroads to use older line-haul locomotives as switchers, that manufacturers could
choose to not offer new switch locomotives during the Tier 1 period.
Tier 2 Standards
The Tier 2 NOx standards will require HC and PM control as well as additional
NOx control. These standards will apply to locomotive engines originally
manufactured in 2005 and later.
45
-------
Locomotive engine manufacturers have been in the process of developing new
engine designs with a focus on fuel economy and significantly increased power output,
to take advantage of developments in traction motor technology. In the past,
locomotive engines have been limited to about 4000 hp. This limitation was a direct
result of space constraints associated with the axle-hung DC motors. However, the
advent of high-performance AC motors has resulted in the availability of traction
motors that have a much higher power density than the old designs. As a result of this
change from DC to AC traction motors, new locomotive engines which are capable of
up to 6000 hp are now under development by the two major locomotive manufacturers.
Because they are in the process of developing these new engine models and have
been for several years, locomotive manufacturers are in a good position to meet the
Tier 2 standards. Both of the high-power engines currently under development are
four-stroke engines. (It should be noted that particulate emissions from two-stroke
engines have been observed to be largely made up of entrained lubricating oil.) Four-
stroke engines, however, have shown lower PM emissions because they generally
achieve much better oil control. Furthermore, the best approach to designing engines
inherently lower in emissions and most effectively implementing such strategies is to
build them in from the beginning rather than adding them later in the process of a
retrofit. This is true for the fuel management, combustion chamber, charge air cooling,
electronic control, and other strategies discussed below and in Chapter 4.
A table listing potential Tier 2 technologies and resultant emission reductions
is presented in Chapter 4. In general, EPA expects that additional NOx and PM
emission reductions will be possible in these locomotive engines through continued
refinements in charge air cooling, fuel management, and combustion chamber
configuration, in conjunction with further improvements in electronic control systems.
Improved fuel management would include strategies such as increased injection
pressure, optimized nozzle hole configuration, and rate-shaping. Potential combustion
chamber redesigns include the use of reentrant piston bowls and increased compression
ratio. It may also be possible to reduce oil consumption by optimizing the ring pack
design and bore honing technique. However, reduced oil consumption could require the
use of more advanced oils to achieve comparable lubrication. Also, while EPA
anticipates that locomotive engine manufacturers will strive to avoid using EGR to the
greatest extent possible, the Agency believes that moderate rates of EGR may be used
in some instances. When the engine is operating at the lighter load steady-state
notches, EGR becomes a more attractive strategy. At these conditions, more excess air
is available, so engine performance and PM emissions are less sensitive to EGR.
Because the operation is steady-state, it makes optimization of the EGR and the fuel
and air much easier that for an engine operating under transient conditions.
Many of the more advanced technologies mentioned in section 3.3 do not appear
in the list of technologies that EPA expects will be used for compliance with the Tier
2 standards. This is because EPA cannot confidently project that the technical and/or
46
-------
practical obstacles to these technologies will be overcome. However, it should be
remembered that the Tier 2 standards will not take effect for another seven years.
This represents a substantial amount of lead time, which will provide an excellent
opportunity for additional technological advances. Two emission control strategies that
have shown some promise are selective catalytic reduction and alternative-fueled
engines. These technologies and others in even more preliminary stages of
consideration may ultimately prove to be more attractive emission control alternatives
with sufficient development. Nevertheless, EPA cannot conclude that these
technologies will be available in the time frame being considered.
3.5 Adjustments for Ambient Conditions
Manufacturers and remanufacturers can incorporate various technologies to
account for changes in ambient conditions. The effects of changes in ambient
temperature can be minimized by controlling the rate of heat rejection in the
aftercoolers and engine cooling system. Current locomotives are already designed to
have a fairly constant engine coolant temperature over a broad range of temperatures.
This is achieved primarily by varying the flow of air past the radiator. Greater control
at cooler temperatures is possible using a thermostat type system to bypass or
partially bypass the heat rejection components of the cooling system (e.g., radiators)
unless the coolant reaches the appropriate temperature. The potentially adverse
effects of higher ambient temperatures can be reduced by increasing the cooling
capacity of the engine, where possible.
The effects of low barometric pressures can be minimized either by reducing the
amount of fuel injected or increasing the amount of air forced into the cylinder so that
the appropriate excess air ratio can be maintained. Although reducing the amount of
fuel injected at lower barometric pressures can result in maintaining emissions
compliance, a loss in engine power usually results. Current diesel engine designs have
been reported to operate with 24 percent less power and 5 percent higher specific fuel
consumption at elevations approaching 7,000 feet.59 An intermittent supercharger has
been developed for truck engines that assists the turbocharger at high altitudes to
compensate for lower air density effects.60
59 Lizhong, S., et al., "Combustion Process of Diesel Engines at Regions With Different
Altitudes," SAE Paper 950857, 1995.
60 Kapich, D., "Very High Speed Hydraulic Driven Supercharging System," SAE Paper
951822,1995.
47
-------
3.6 Reliability and Durability
EPA understands that the reliability and durability of locomotive engines are
important to the railroad industry. Any time that a locomotive breaks down on a track,
it can result in severely disrupted traffic flows, especially where the locomotive must
be repaired in place. The technology projected to be applied to meet the locomotive
standards has been applied to trucks and other applications. Experience to date
suggests that there is no reason that any of these technologies should inherently cause
any sort of reliability or durability problem for locomotives.
Although the emission control technologies described above are themselves
likely to be inherently reliable and durable, there may be a learning curve associated
with confidence in the application of these technologies to locomotives. With the seven
years of lead time given in this rule, potential learning curve difficulties should not be
a limiting factor. There will be adequate time for prototypes to be durability tested,
for demonstration fleets to prove out the reliability of these technologies, and for
upgrades and optimization to be implemented. Because of the importance of engine
reliability and durability to the railroad industry, EPA believes that locomotive
manufacturers will make sufficient efforts to prove out their new engine designs.
48
-------
4.0 Emission Standards and Supporting Analyses
This chapter describes the development of EPA's emission standards for
locomotives. As will be described, the regulations require that cycle-weighted brake-
specific HC, CO, NOx, and PM emissions from each new locomotive be: 1) below the
line-haul standards when weighted using line-haul duty-cycle; and 2) below the switch
standards when weighted using switch duty-cycle61. The regulations also require that
smoke emissions be below the specified smoke standards. Compliance with these
standards is required throughout the full useful life of the locomotive. This chapter
also describes the development of the duty-cycles, useful life periods, and baseline
emission rates on which the numeric standards are based. Finally, this chapter also
analyzes feasibility of these standards.
62
4.1 Duty-Cycles
EPA believes that the most cost-effective means of achieving national locomotive
emission reductions is to set standards for emissions weighted by typical in-use duty-
cycles. Unlike other vehicles which have a continuously variable throttle, locomotives
are limited to a predetermined number of throttle notches. Individual standards for
each throttle notch might be able to achieve similar emission reductions, but with less
flexibility, and thus, probably at a higher cost. This section describes the development
of the two duty-cycles used in this regulation, as well as the development of a
passenger locomotive duty-cycle that is presented here for informational purposes.
Industry Freight Locomotive Duty-Cycles
Industry has historically used two distinct types of duty-cycles for freight
locomotives: line-haul and switching. The term line-haul refers to the movement of
freight between cities or other widely separated points. Several previously-used line-
haul duty-cycles are shown in Table 4-1. Switching refers to the process of assembling
and disassembling trains in a relatively small area (a switchyard); thus switching
operations are also often referred to as yard operations. Two historical duty-cycles for
switch locomotives are shown in Table 4-2. Dynamic braking is not included in these
cycles because switch locomotives are usually not equipped with this function.
61 Exception: existing switch locomotives will not be required to comply with the line-
haul duty-cycle standards.
62 For locomotives, a duty-cycle is a usage pattern expressed as the percent of time in
use in each of the throttle notches.
49
-------
Table 4-1
Historical Information — Line-Haul Duty-Cycles
(Percent Time in Notch)
Throttle
Position
Idle
Dynamic
Brake
1
2
3
4
5
6
7
8
GE
Min.
59.0
1.5
6.5
6.5
6.5
6.5
2.9
2.9
2.5
5.2
GE
Max.
40.0
7.0
2.5
2.5
2.5
2.1
1.8
1.8
1.8
38.0
GE
Average 1
54.0
4.0
5.0
2.5
2.0
5.0
2.0
2.0
2.5
21.0
GE
Average 2
53.0
5.5
5.1
3.9
3.4
3.3
2.8
3.4
2.6
17.0
EMD
Heavy
41.0
8.0
3.0
3.0
3.0
3.0
3.0
3.0
3.0
30.0
EMD
Medium
46.0
9.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
17.0
AAR
43.0
8.0
3.0
3.0
3.0
3.0
3.0
3.0
3.0
28.0
Table 4-2
Historical Information - Switch Duty-Cycles
(Percent Time in Notch)
Throttle Position
Idle
Dynamic Brake
1
2
3
4
5
6
7
8
ATSF
77.0
0.0
10.0
5.0
4.0
2.0
1.0
1.0
0.0
0.0
EMD
77.0
0.0
7.0
7.0
4.0
2.0
1.0
0.5
0.5
1.0
50
-------
EPA Duty-Cycles
In response to a request by EPA, several freight railroads collected time-in-notch
data for both line-haul and switch operations. Amtrak also collected data for passenger
locomotive operations. The data were presented to EPA in different formats, but each
allowed the calculation of average time in each notch.63 It is important to note that not
all of the idle time reported corresponds to time during which the locomotive was not
moving. As much as one-third of idle time can be spent while a locomotive is coasting
(i.e., moving forward because of either gravity or momentum without any power being
applied to the traction motors).
In the case of line-haul operations, the data came from 63 trains64 operated by
five Class I railroads. Train operations were spread over many regions of the nation
and represented approximately 2,475 hours of freight train operations. Data on switch
operations came from two railroads and represented approximately 333 hours of switch
locomotive operations. Amtrak provided data from 20 locomotives covering
approximately 57,500 hours of operation.
These data were reviewed to identify duty-cycles applicable to line-haul, switch
and passenger operations. Results from the data collected by the railroads are
summarized in Tables 4-3 through 4-5 below. In addition to the average duty-cycles
developed from the current database, the highest and lowest individual percentages
of the time in notch from all line-haul and switch locomotive data were shown in
Tables 4-3 and 4-4. This information shows the presence of very wide variations
around the averages for the two types of operations.
63 EPA assumed equal time is notches 1 and 2 because some of the line-haul data did not
separate those notches from one another.
64 The term train is used for line-haul, instead of locomotive, because many of the trains
for which data were reported included more than one locomotive in the consist. For
example, a given train could require 12,000 total horsepower for propulsion, which would be
accomplished with three 4,000 horsepower locomotives in series.
51
-------
Table 4-3
Current Locomotive Operations - Line-Haul Duty-Cycles
(Percent Time in Notch)
Data Source: 63 Trains
Throttle
Position
Idle
Dynamic Brake
1
2
3
4
5
6
7
8
Average
38.0
12.5
6.5
6.5
5.2
4.4
3.8
3.9
3.0
16.2
Highest
77
41
23
23
13
11
12
11
18
39
Lowest
1
0
0
0
2
1
0
0
0
0
Table 4-4
Current Locomotive Operations - Switch Duty-Cycles
(Percent Time in Notch)
Data Source: 8 Locomotives
Throttle
Position
Idle
Dynamic Brake
1
2
3
4
5
6
7
8
Average
59.8
NA
12.4
12.3
5.8
3.6
3.6
1.5
0.2
0.8
Highest
82
NA
18
18
20
17
15
10
1
4
Lowest
23
NA
7
7
1
1
0
0
0
0
52
-------
Table 4-5
Current Locomotive Operations - Passenger Duty-Cycles
(Percent Time in Notch)
Data Source: 20 Locomotives
Throttle
Position
Idle
Dynamic Brake
1
2
3
4
5
6
7
8
Average
47.4
6.2
7.0
5.1
5.7
4.7
4.0
2.9
1.4
15.6
Highest
60
8
11
6
7
6
5
6
2
19
Lowest
40
2
5
4
4
4
2
1
1
8
The average line-haul duty-cycle from current operations is generally consistent
with the historical cycles. Current operations, however, show higher usage of dynamic
brake than in any of the historical cycles. In the case of switching operations, the
current data show a lower percentage of time at idle and higher percentages of time in
notches 1 and 2 than are shown in the historical cycles. The historical cycles bracket
the results from current data for the other throttle notches.
EPA believes that it is more appropriate to use the results of the recent
operational data for this rulemaking since their source is known (while the exact
source of the information used to develop the historical cycles is not fully known).
Moreover, the data are known to represent operations in widely dispersed areas of the
nation. The three average duty-cycles (line-haul, passenger65 and switch) are
summarized in Table 4-6 below.
65 The passenger duty-cycle is not used in the regulations, and is shown here only for
informational purposes.
53
-------
Table 4-6
EPA Estimated Duty- Cycles for Current In-Use Locomotive
(Percent Time in Notch)
Throttle Notch
Idle
Dynamic Brake
1
2
3
4
5
6
7
8
Line-haul
38.0
12.5
6.5
6.5
5.2
4.4
3.8
3.9
3.0
16.2
Passenger
47.4
6.2
7.0
5.1
5.7
4.7
4.0
2.9
1.4
15.6
Switch
59.8
0.0
12.4
12.3
5.8
3.6
3.6
1.5
0.2
0.8
Emissions Impacts of Duty-Cycles
EPA believes that requiring locomotives to comply with emission standards
using both a high-power duty-cycle (line-haul) and low-power duty-cycle (switch) will
not only achieve cost-effective national emission reductions, but will also minimize the
geographic variation in the reductions. The line-haul cycle is fairly representative of
a national average duty-cycle. However, because it weights high-power emissions so
heavily, it would be possible for a locomotive to comply with line-haul emission
standards with little or no emission reduction at idle. This is not an unreasonable
scenario to consider, especially for current locomotives, since the emission controls
expected to be employed are generally most effective at higher power levels. If such
a compliance strategy were allowed, then urban areas where locomotives are operated
more frequently at idle and in low power notches could potentially see little or no
emission reductions. This is why EPA is requiring that all locomotives be required to
comply with emission standards for both the line-haul and switch duty-cycles.
It should be noted that while the idle weighting factors for the two duty-cycles
are not that different from one another (0.380 and 0.598), the relative importance of
idle emissions in the two cycle-weighted emission calculations is very different. This
is because the calculation weights mass emission rates (g/hr), which can be nearly 100
54
-------
times greater at full power (notch 8) than at idle. While idle emissions are not very
significant for the line-haul cycle, they can actually be more important than notch 8
emissions for the switch cycle. For example, consider a locomotive that has mass
emission rates of 700 g/hr at idle and 50,000 g/hr at full power. Using the line-haul
weightings, the weighted emission rates for idle and full power would be 266 (700 x
0.380) and 8,100 (50,000 x 0.162) g/hr, respectively. In this case, the contribution of
notch 8 emissions would be 30 times that of the idle emissions. On the other hand,
using the switch weightings, they would be 419 (700 x 0.598) and 400 (50,000 x 0.008)
g/hr, respectively. In fact, using the switch weightings, weighted emissions for all
notches will generally be of similar magnitude. This means that manufacturers and
remanufacturers would have a strong incentive to achieve significant emission
reductions for each notch.
While the available data indicate that duty-cycles for passenger locomotives are
different from those of freight locomotives, EPA believes that it is not necessary to use
a passenger-specific duty-cycle to achieve the desired emissions reductions from
passenger locomotives. The average passenger locomotive's duty-cycle is similar to the
average line-haul cycle, except that it includes significantly more idling time. Thus,
requiring all locomotives, including passenger locomotives, to comply with emissions
standards for the line-haul and switch cycles should achieve essentially the same
emission reductions as would be achieved by using the passenger cycle. Moreover, as
noted previously, manufacturers and remanufacturers are likely to reduce emissions
significantly in all notches, which would make duty-cycle concerns less significant.
4.2 Useful Life
Definition of Useful Life
Useful life is the term EPA uses to designate the period during which a vehicle
is required to comply with emissions standards. For highway vehicles, the period of
compliance is usually expressed in terms of years and mileage, with the requirement
for compliance ending when either parameter is exceeded. Useful life periods currently
applicable to highway vehicles were originally intended to approximately correspond
to the median time and mileage at which a unit is scrapped or remanufactured. The
range of current highway vehicle useful life values are from 10 years and 100,000 miles
for light-duty vehicles (passenger cars) to 10 years and 435,000 miles (beginning in
2004) for heavy heavy-duty diesel engines (i.e., engines used in largest trucks).
For locomotives, EPA is explicitly defining useful life to be the period during
which a locomotive is designed to remain properly functional with respect to power
output and fuel economy. This definition recognizes that different designs of
locomotives can have different useful lives. EPA's approach is to specify default useful
life values, but require manufacturers and remanufacturers to specify longer useful
lives for locomotives that are designed to last significantly longer than the default
55
-------
period. In some limited cases, EPA will also allow re manufacturers to specify shorter
useful lives for locomotives that are designed to last significantly less than the default
period, provided that the manufacturer can provide adequate technical justification.
This will only be allowed for engines that were designed for use in non-locomotive
applications.
This design period is very similar to the average period between
remanufactures, and takes into account the intense maintenance which most
locomotives undergo in use and the strong desire on the part of railroads for reliability.
Thus, EPA based its default useful life values on current information about
remanufacturing intervals. However, EPA did not set the default values to be equal
to current average remanufacture intervals. Rather, it projected future intervals,
based on current average intervals for the more recent locomotive models, and on
comments from industry regarding future technology.
Current Class I Remanufacturing Practices
Figure 4-1 shows remanufacturing interval data provided by AAR.66 Although
these data were collected from a single railroad (ATSF), they are fairly representative
of Class I remanufacturing practices for the current fleet. These data show a median
remanufacture interval of about 20,000 MW-hr (about 5.7 MW-hr per hp), and that
about 95 percent of locomotives are remanufactured before they reach 28,000 MW-hr
(about 8.0 MW-hr per hp). Similar data for mileage intervals, which are shown in
Appendix A, indicate a median remanufacture interval of about 700,000 miles.67
The figure also shows that there is an increase in the MW-hr remanufacturing
intervals for newer model locomotives. To some extent, this is caused by an increase
in engine horsepower. However, it also results from improved durability.
Manufacturers have made numerous improvements over the years to significantly
increase engine life. Railroads have also increased engine life by constantly improving
maintenance practices. EPA expects that these trends will continue, resulting in
marginal increases in median engine life each year.
66 Data are for ATSF remanufacturing between 1989 and 1994 of locomotives
manufactured 1973-1990. Essentially all of the locomotives in this data set were between
3000 and 4000 hp, with an average of approximately 3500 hp.
67 EPA received other similar data from AAR regarding remanufacturing intervals.
These data have been placed in public docket A-94-31.
56
-------
100%
Figure 4-1
Remanufacture Megawatt Hour Distributions
1973 and Later Locomotives
1,000
15,000 20,000 25,000
Remanufacture Megawatt Hours
30,000
57
-------
Default Useful Life
EPA believes that the most appropriate way to determine useful life is set
default values, and then specify variations from that default on a case-by-case basis.
EPA believes that the default values in MW-hr should be equal to 7.5 times the rated
horsepower of the engine. For a typical 3500-hp locomotive, like those shown in Figure
4-1, this would mean a useful life period of 26,250 MW-hr.
In selecting this default value (i.e., 7.5 times horsepower), EPA sought a value
that would generally be feasible for the current fleet, even if there are no future
improvements in engine life, while ensuring the desired in-use control of emissions
from future locomotives under the most likely engine life scenario. While the selected
default value is somewhat greater than the median remanufacture interval for the
current fleet, EPA is confident that re manufacturers will be able to comply with the
standards during this period. EPA also believes that this value will be reasonably
close to the median remanufacturing interval that will be observed for Class I railroads
after these standards go into effect.
EPA recognizes that some Tier 0 locomotives will not be equipped with
megawatt-hour meters. For these locomotives, EPA has set the default useful life at
750,000 miles, or ten years, whichever occurs first. EPA is including the year
specification to account for switch locomotives, or other low-use locomotives. In
practice, EPA expects that most Tier 0 line-haul locomotives will reach the 750,000
mile point before ten years, while most Tier 0 switch locomotives will not. Moreover,
EPA is not confident that mileage accumulation values would be meaningful for switch
locomotives operating within a switchyard, where miles have little relevance.
Variations From Default Values
EPA expects that some future locomotives will be designed to be operated (and
actually will be operated in use) significantly beyond the default useful life values
defined here. In such cases, EPA will require that manufacturers and remanufacturers
specify a useful life that is longer than the default values. Generally, EPA would
require that the useful life value be at least as long as the median remanufacturing
interval of those locomotives in use. However, the Agency does recognize that there
could be cases in which the median remanufacturing interval would not be appropriate
for the useful life because the railroads were actually using the locomotives beyond
their legitimate design life. Such special cases would be indicated by very significant
increases in fuel consumption and/or decreases in reliability or power output, or
excessive maintenance costs before the locomotives were remanufactured.
Nevertheless, EPA believes these would be rare cases.
58
-------
EPA will allow manufacturers of nonroad engines used to repower locomotives
to petition for a shorter useful life. EPA recognizes that many of these engines will
have been designed for other applications where they may not expected to last as long
as a locomotive engine. In these cases, EPA will allow a significantly shorter useful
life, but will require substantial supporting information from the manufacturer.
4.3 Baseline Emission Rates
Locomotive manufacturers (EMD and GE)68 provided EPA with information on
locomotive emissions for HC, CO, NOx and PM. Most of the information was provided
at the time that EPA was developing revisions to the calculations for locomotive
emissions in "Compilation of Air Pollutant Emission Factors, Volume II: Mobile
Sources". The revisions to that document applicable to locomotives were published in
1992.69 EPA weighted these data by both the line-haul and switch duty-cycles to
estimate baseline emission rates. These cycle-weighted emission rates are shown in
Appendix B with the individual notch emission rates. Shown in the Appendix G are
the results of testing by Southwest Research Institute (SwRI) of late model
locomotives. The NOx and PM emission rates are also shown in Figures 4-2 and 4-3,
and summarized in Table 4-7. The HC and CO emission rates are shown graphically
in Appendix H.
Table 4-7
Range of NOx and PM Emission Rates by Engine Type (g/bhp-hr)
EMD 645
EMD 710
GE
Line-Haul Cycle
NOx
11.5-18.2
10.6-14.2
10.3-15.0
PM
0.25-0.31
0.23-0.35
0.22-0.41
Switch Cycle
NOx
14.1-33.1
14.2-17.3
9.2-15.8
PM
0.28-0.44
0.28-0.39
0.22-0.86
68 Note: some additional information was subsequently provided by Caterpillar, Inc., on
a diesel engine appropriate for line-haul locomotives.
69 Procedures for Emission Inventory Preparation, Volume IV: Mobile Sources; U. S.
EPA, Emission Planning and Strategies Division, Office of Mobile Sources and Technical
Support Division, Office of Air Quality Planning and Standards; EPA-450/4-81-026d
(Revised).
59
-------
In addition to the preceding information provided by manufacturers, information
was made available to EPA by AAR70. The data supplied by AAR were collected by
Southwest Research Institute (SwRI) under contract to AAR. Only a portion of the
more recently provided information from AAR was utilized here since much of it was
collected at only three power levels.
It is important to recognize that there is significant uncertainty associated with
estimating baseline locomotive emissions, especially given the inherent variability of
uncontrolled emission rates (e.g., line-haul NOx emission rates vary from 10.55 to
15.54 g/bhp-hr) and the limited amount of data that are available. The following
analysis represents the Agency's best estimate of baseline emission rates.
Line-Haul Baseline Emission Rates
Line-haul emission rates were estimated for the 1990 fleet by cycle-weighting
the emission rates of each engine type by their respective in use populations and
horsepower. Similarly, weighted emission rates for later model locomotives were also
calculated. However, the data used for these later model years were less complete.
These analyses are summarized in Appendix C. EPA applied a deterioration factor to
the weighted average HC and PM emission rates to account for the expected difference
between average in-use emissions rates and emission rates calculated from test
engines or relatively new locomotives; the weighted average HC and PM emission rates
were multiplied by 1.15. This deterioration factor was estimated by EPA from its
experience with other diesel engines, and from confidential business information
provided by the locomotive engine manufacturers.
It should be noted that no attempt was made to account for potential differences
in in-use usage rates (i.e., whether newer models are used more frequently than older
models) for this baseline analysis. As can be seen from the data in the Appendices B
and C, there no clear trend of emissions of older locomotives having very different
emissions from the later model locomotives. The data from EPA's testing of several
later model locomotives showed some evidence of higher NOx emissions and lower HC,
CO, and PM emissions than were seen from the older locomotive models, but this trend
was not conclusive. Nevertheless, in some cases, the NOx emissions from these newer
locomotives were as high as 15 g/bhp-hr. Thus it is possible that, without EPA
regulation, the average in-use NOx emission rate would be much higher in the future
than is estimated for the current fleet.
70 Data have been provided by AAR on more than one occasion. The first set of data
provided was collected in the early 1980s and was used in EPA's Report to Congress on
emissions from locomotives. That set of data was not employed in development of this
proposal because of its age.
60
-------
Figure 4-2 Line-Haul Cycle Emissions Data
NOxandPM (g/bhp-hr)
°n
£. \j
15
x
0 10
5
n
O
8
0 0 (ft
o
o
_°P oo o
"
-
1 1 1 1
0 0.1 0.2 0.3 0.4 0.5
PM
40
30
0 20
10
0
C
Figure 4-3 Switch Cycle Emissions Data
NOxandPM (g/bhp-hr)
o
0 °
0
ODvO) fe
V&ffo0
O ^5 O
0.2 0.4 0.6 0.8 1
PM
61
-------
Switch Baseline Emission Rates
Baseline emission rates for switch locomotives were estimated from cycle-
weighted average emission rates for the three switch engine models that are used in
the vast majority of existing switch locomotives. The same deterioration factor of 1.15
as was used in the line-haul analysis was also applied to the weighted average HC and
PM emissions.
The following table presents estimated baseline emission rates for both line-haul
and switch locomotives.
Table 4-8
Estimated Baseline In-Use Emission Rates (g/bhp-hr)
Line-Haul*
Switch**
HC
0.48
1.01
CO
1.28
1.83
NOx
13.0
17.4
PM
0.32
0.44
* Line-haul locomotives over the line-haul duty-cycle
** Switch locomotives over the switch duty-cycle
Baseline Smoke Emissions
The available data for smoke emissions are summarized in Appendix D. These
data are mainly for steady-state smoke levels. They show steady-state smoke levels
ranging from 0 to 35 percent opacity for current locomotives (with standard injection
timing) when measured with an optical path length of approximately one meter. Data
for newer, well-maintained locomotives show steady-state smoke levels that are
typically less than 10 percent.
4.4 Emission Standards
Gaseous and Particulate Emission Standards
The final emission standards for both the line-haul and switch duty-cycles are
shown in Table 4-9. Also shown in the table are the expected design targets and
average percent reductions from baseline emission rates. Based on the information
currently available, EPA has concluded that these standards are the most stringent
standards that can be achieved at a reasonable cost within the time period being
considered. EPA is not necessarily projecting that all existing locomotive
configurations will be able to achieve the Tier 0 NOx emission level at a reasonable
cost, but rather that average NOx emission levels will be below the Tier 0 standard.
62
-------
Full compliance will probably require significant use of the emission averaging,
trading, and banking (ABT) provisions, and may require that some specific high-
emitting engine configurations be converted to other configurations for which emission
controls are more available. The Tier 0 standards for HC, CO, and PM will be
achievable without averaging. For the Tier 1 and Tier 2 standards, ABT will be
necessary for both NOx and PM since manufacturers may need this flexibility to
ensure compliance for all freshly manufactured locomotives during the first few years.
Absent ABT, less stringent standards or more lead time may have been necessary.
Given the emission control technologies that are expected to be available, EPA
is establishing standards that will achieve very significant reductions in NOx
emissions from the beginning of the program. However, the standards will not begin
to achieve significant reductions in the emissions of other pollutants such as HC and
PM until 2005. The Tier 0 and Tier 1 standards were set at levels that should allow
the use of retarded injection timing, which may cause emissions of HC and/or PM to
increase. This is appropriate because NOx is the only pollutant for which locomotive
emissions contribute more than one percent of the estimated national inventories
(locomotives contribute less than one-quarter of a percent for HC, CO, and PM). EPA
did consider more stringent Tier 0 and Tier 1 emission standards for HC, CO, and PM,
but concluded that the Tier 0 and Tier 1 emission standards for NOx might not be
achievable if significant reductions in HC, CO, and PM were also required.
Nevertheless, EPA does believe that average emissions for these pollutants will not
increase, even though the standards will allow the emissions to increase for some
specific locomotives. For these cases (which are denoted with asterisks in Table 4-9),
EPA is projecting that average emissions will remain unchanged from baseline levels.
In considering what emission standards would be achievable, EPA also
considered the need for compliance margins.71 Based on data contained in Appendix
E, it was assumed that manufacturers and remanufacturers would incorporate
different compliance margins into the locomotive designs for different pollutants. The
biggest compliance margin used was 20 percent for PM. The reasons for this are
testing variability and the potential for in-use deterioration. For HC and CO, 15
percent compliance margins were used because of HC and CO measurement variability
is significantly less than for PM. For NOx emissions, a 10 percent compliance margin
was used because both deterioration and test variability are expected to be less of a
concern for NOx than they are for other pollutants. This 10 percent margin is higher
than the average compliance margin for 1993-1995 on-highway diesel engines that is
shown in the Appendix E (8 percent) because EPA believes that manufacturers and
remanufacturers will incorporate slightly larger compliance margins to account for the
additional risk associated with the extensive in-use testing program.
71 A compliance margin is a difference between the emission standard and the design
emission level. Manufacturers incorporate compliance margins to account for production
and testing variability.
63
-------
It should be emphasized that the estimated percent reductions are the
reductions from the estimated 1990 baseline levels. As was noted earlier, there is some
reason to believe that NOx emission rates would increase significantly (and HC and
PM could decrease) in the future without these emission standards. Thus, the true
reduction of the Tier 2 NOx standard, relative to some of the locomotives currently
being produced, could be as high as 67 percent. The HC and PM reductions, on the
other hand, could be somewhat lower.
Table 4-9
Emission Standards and Expected Emission Reductions
Line-
Haul
Cycle
Switch
Cycle
NOx
PM
HC
CO
NOx
PM
HC
CO
Standard
(g/bhp-hr)
TierO
9.5
0.60
1.0
5.0
14.0
0.72
2.1
8.0
Tier 1
7.4
0.45
0.55
2.2
11
0.54
1.2
2.5
Tier 2
5.5
0.20
0.30
1.5
8.1
0.24
0.60
2.4
In-Use Emissions
(g/bhp-hr)
TierO
8.6
0.32*
0.48*
1.3*
12.6
0.44*
1.0*
1.8*
Tier 1
6.7
0.32*
0.47
1.3*
9.9
0.43
1.0*
1.8*
Tier 2
5.0
0.16
0.26
1.3*
7.3
0.19
0.51
1.8*
Percent Reduction
TierO
34%
0%
0%
0%
28%
0%
0%
0%
Tier 1
49%
0%
3%
0%
43%
2%
0%
0%
Tier 2
62%
50%
47%
0%
58%
56%
50%
0%
* Baseline emission level
Smoke Standards
EPA is setting both "steady-state" and "peak" smoke standards for locomotives.
These standards are shown in Table 4-10. The steady-state standards will apply to
locomotive exhaust after the transition period during which the fueling rate is slowly
increased (typically within 15 seconds) after a notch change. The peak standards apply
to all times, but are intended to limit smoke during the transition period. The 3-second
and 30-second standards apply to the maximum smoke level observed during any
continuous 3-second and 30 second period, respectively. The same peak standards
apply to all tiers of standards, while the steady-state standards become more stringent
with each tier. There are several reasons for this. First, the Agency believes that
steady-state smoke levels are more environmentally significant than peak levels for
locomotives because of the largely steady-state manner in which locomotives are
operated. Thus, it is more critical that the steady-state standards be sufficiently
64
-------
stringent. Second, EPA is more confident that Tier 1 and Tier 2 locomotives will be
able to comply with more stringent smoke standards at steady-state than they will
during the transition period after notch changes. Finally, EPA is concerned that overly
stringent peak standards might adversely affect locomotive performance, and that such
an effect might be unwarranted, given the marginal value of lower peak smoke
standards.
Table 4- 10
Smoke Standards for Locomotives
(Percent Opacity for One-Meter Path Length)
TierO
Tier 1
Tier 2
Steady-state
30
25
20
30-sec peak
40
40
40
3-sec peak
50
50
50
Alternate Emission Standards
The general emission standards discussed previously are based on an analysis
of the potential for emission reductions from diesel-powered locomotives. However,
EPA recognizes that locomotives powered by alternatively-fueled engines could
potentially have even lower emissions for some pollutants, while having difficulty
complying with the standards for other pollutants. Therefore, EPA is establishing an
optional alternate set of standards that could be applied to locomotives which operate
on alternative fuels. The alternate standards allow higher CO emissions, but also
require lower particulate emissions. Although these alternate standards are primarily
intended to address issues associated with alternative fuels, EPA intends that they be
available for application to any locomotive. Manufacturers and remanufacturers would
be allowed to certify to the alternate CO standards instead of the normal standards,
as long as they also comply with the more stringent alternate PM standards, as well
as the normal NOx and hydrocarbon standards. They will not be allowed to mix the
alternate CO standards with the primary particulate standards for a single engine
family.
In developing these alternate standards, EPA focused on the emission
characteristics of current natural gas-fueled locomotives. Such locomotives generally
have higher (and more difficult to control) CO and hydrocarbon (i.e., total hydrocarbon)
emissions than diesel-fueled locomotives, but lower PM emissions. The status of
attainment for CO is much better nationwide than that for PM. Thus, it would be
inappropriate to effectively defer the introduction of low-PM technologies like natural
65
-------
gas engines from market because they would have difficulty in complying with the CO
standards. EPA believes that the environmental benefit of the additional PM
reductions from these standards for this limited number of locomotives is greater than
the environmental cost associated with allowing somewhat higher CO emissions.
4.5 Feasibility and Compliance with Standards
Manufacturers and remanufacturers will be required to comply with these
standards during certification, production and in-use. This section discusses the ability
of manufacturers and remanufacturers to comply with these requirements. Additional
information regarding technological feasibility can be found in Chapter 3.
Tier 0 Certification
Table 4-11 outlines technologies that EPA believes will be available for
retrofitting Tier 0 locomotives in 2000. The projected emission reductions listed in the
table represent the reductions that would be expected for a typical Tier 0 locomotive.
However, the existing fleet is diverse, and the effectiveness of these technologies will
vary from model to model. Based on these projections, EPA has determined that the
Tier 0 standards being adopted are the most stringent standards that can be adopted
for the existing fleet. Both the manufacturers and the railroads have agreed with EPA
that these standards are feasible, but that they will require extensive use of averaging,
and may lead to a few locomotive models being removed from service. Compliance with
the Tier 0 standards can be considered more precisely by dividing the subject
locomotives into three groups: 1) older and low-volume locomotives; 2) recent and high-
volume locomotives with relatively low PM and smoke emissions; and 3) locomotives
requiring significant smoke and particulate control.
Remanufacturers of the first group of locomotives are expected to be able to
comply largely by significant retarding of the injection timing. Such an approach,
however, will result in a one to two percent increase in fuel consumption. The degree
of timing retard needed can be lessened by using enhanced charge air cooling such as
changing from a two-pass to a four-pass aftercooler. In some cases, remanufacturers
of such locomotives are expected to use credits generated by other engine families with
more advanced emission controls. For especially high-emitting locomotives,
remanufacturers might also reconfigure the engine to more closely resemble other
lower emitting configurations. Many improvements have been made over the last
decade such as advances in fuel injection systems and turbocharger designs which
could be applied to older engines. Also, it is expected that some of the old, inefficient
locomotives will be removed from service instead of being rebuilt.
Manufacturers have indicated that, for most engine designs, the NOx emission
standard is technologically feasible using technologies that can be incorporated
during re manufacture. One locomotive engine design has shown unusually high
66
-------
emissions and is not expected to be able to meet the standard without more
extensive retrofitting. However, the averaging and banking provisions of this
rulemaking are expected to allow these locomotives to comply with the standard,
especially since the number of these locomotives which are still in service is not very
large.
Table 4- 11
Tier 0 Locomotive Technology and Effectiveness
(Expressed as Percent Reduction from Baseline; Increases Shown in Parentheses)
Technology
Timing retard
Electronic control
Fuel injection*
Improved turbocharging
4-pass aftercooling
Engine modification
Puff limiter
COMBINED
REDUCTIONS
HC
(5-15%)
0-10%
5-15%
0-10%
-
0-5%
-
10-20%
CO
(5-10%)
-
5-10%
0-5%
-
0-5%
-
0-10%
NOx
15-25%
0-10%
-
(0-5%)
5-10%
-
-
30-45%
PM
(10-20%)
0-5%
10-20%
0-5%
0-5%
0-5%
-
0-15%
Smoke
worse
better
better
better
even
even
better
better
BSFC
(1-2%)
0-1%
0-1%
0-0.5%
(0-1%)
-
-
(2)-l%
* improvements in nozzle geometry, sac volume, and fuel injection control for mechanical and electronic
injectors
For the second group, where the number of locomotives within a family is high, or
where the remaining service life is expected to be long, railroads are likely to desire
higher technology remanufacturing kits that will minimize possible fuel consumption
effects. To provide such kits, remanufacturers will likely use optimized electronic fuel
injection systems that provide some of the NOx benefits of timing retard without the
fuel consumption penalty. In some cases, remanufacturers will likely make some
improvements to the aftercooling systems such as using four-pass heat exchangers.
There are also opportunities for lower emissions and fuel consumption through the use
of improved fuel injectors. Potential modifications to the fuel injection system would
include injection rate, nozzle spray geometry, and reduced sac volumes. Since such
approaches will likely result in NOx emissions well below the Tier 0 standards, many
of these engines are expected to generate emission credits. Although many of the new
locomotive technology improvements offer the potential for lower fuel consumption,
some sacrifices may need to be made to fuel economy improvements to achieve lower
emissions. Ultimately, each remanufacturer will make choices between first cost and
fuel costs so as to minimize total cost of compliance over the remaining portion of the
locomotive's total life.
67
-------
Application of injection timing retard to the last group might be limited because
of the PM and smoke increases that would result. However, turbocharger designs have
improved in the past decade, and improved turbocharging allows for reductions in
smoke and PM. In addition to improved turbochargers, such locomotives are expected
to require significant improvements to aftercooling systems (e.g., more efficient heat
exchangers) and electronically controlled fuel injection systems. They might also
require combustion chamber redesign to reduce PM and smoke emissions by achieving
more complete combustion. Some of these locomotives are expected to use NOx
emission credits to the extent that they are available.
It is important to note that the ability of remanufacturers to fully comply with the
Tier 0 emission standards is limited to some extent by their ability to devote adequate
resources to developing emission controls for multiple engine families. Therefore, EPA
is phasing-in the Tier 0 requirements to allow remanufacturers to focus their resources
sequentially, beginning with the latest models. The initial focus on later model
locomotives is appropriate because these locomotives are more heavily used than older
locomotives. This approach has the additional benefit of minimizing market
disruptions for small businesses that participate significantly in the remanufacturing
of the older locomotives.
Tier 1 Certification
It is likely that both major locomotive manufacturers will be able to comply with
the Tier 1 emission standards by further modification of their current production
locomotives beyond the expected changes necessary to comply with the Tier 0 emission
standards. The technologies that will be available are listed in Table 4-12. In
addition, both major manufacturers are planning to introduce new low-emission
locomotive designs, and have indicated that they will be able to comply with the Tier
1 emission standards by January 1, 2002. In fact, once optimized, these new
locomotive models will be the models that the manufacturers use to comply with the
Tier 2 standards. Thus, the manufacturers may be able to use these new models to
generate emission credits, which could be used to ease compliance for their current
models. However, even though EPA believes that some locomotive models would be
able to comply with standards more stringent than the Tier 1 standards before
January 1, 2005 (which is when the applicability of the Tier 1 standards ends for
freshly manufactured production), EPA believes that the Tier 1 standards are the most
stringent standards that will be feasible on average for all freshly manufactured
production during the 2002-2004 time frame.
68
-------
Table 4- 12
Tier 1 Locomotive Technology and Effectiveness
(Expressed as Percent Reduction from Baseline; Increases Shown in Parentheses)
Technology
Timing retard
Electronic control
Fuel injection*
Improved turbo-
charging
Separate circuit
aftercooling
Combustion chamber
Engine modification
Puff limiter
COMBINED
REDUCTIONS
HC
(10-25%)
10-15%
10-20%
5-10%
-
0-5%
0-5%
-
10-20%
CO
(5-15%)
0-5%
5-10%
0-5%
-
0-5%
0-5%
-
0-10%
NOx
20-30%
10-15%
-
(0-5%)
10-15%
-
-
-
45-55%
PM
(15-25%)
5-10%
10-20%
0-5%
0-5%
0-5%
0-5%
-
0-20%
smoke
worse
better
better
better
even
even
even
better
better
BSFC
(1-2%)
0-2%
0-1%
0-0.5%
(0-1%)
0-0.5%
-
-
(2)-2%
* improvements in nozzle geometry, injection pressure, sac volume, and fuel injection control for
electronic injectors
Tier 2 Certification
Manufacturers will need to perform significant work to apply available on-
highway emission control technology to locomotives to meet the Tier 2 standards.
However, because the Tier 2 standards do not go into place until 2005, there will be a
relatively long lead time thus sufficient opportunity for technological development,
prove out, and certification.
On-highway diesel truck engines have shown large reductions in HC, NOx, PM,
and smoke over the past 25 years while still increasing power and fuel economy. Many
of the technologies developed for highway applications are expected to be able to be
applied to locomotives. Therefore, much of the experience gained by the truck engine
manufacturers can be applied to locomotive engine designs. Table 4-13 presents
emission reductions achieved from heavy-duty on-highway diesel engines and those
anticipated for locomotive engines.
Chapter 3 describes the technologies that will likely be used on locomotive engines
to meet the Tier 2 standards. It should be noted that these are generally similar to
systems being used on truck engines today. EPA believes that manufacturers would
make use of some of these strategies regardless of emissions regulations in order to
satisfy customer demands for higher power and better fuel economy. Given the long
69
-------
lead time, EPA believes that the manufacturers will be able to make use of these
technologies to meet the Tier 2 standards. Table 4-14 includes the technologies that
EPA expects will be used to meet the Tier 2 standards as well as the approximate
effectiveness of each strategy.
Table 4- 13
Emission Reductions for Trucks and Locomotives
HC
CO
NOx
PM
On-Highway Heavy-Duty Diesel Engine History (HD-FTP transient cycle)
1988 Truck Emission Standards (g/bhp-hr)
1998 Truck Emission Standards (g/bhp-hr)
2004 Truck Emission Standards (g/bhp-hr)
Percent Reduction Achieved in Past 10 Years
Percent Reduction Anticipated in 2004
1.3
1.3
0.5*
0%
62%
15.5
15.5
15.5
0%
0%
10.7
4.0
2.0*
63%
81%
0.6
0.1
0.1
83%
83%
Locomotive Engine Projections (based on line-haul cycle)
Baseline Locomotive Emission Levels (g/bhp-hr)
Tier 2 Emission Standards (g/bhp-hr)
Design Target w/ Compliance Margin (g/bhp-hr)
Percent Reduction Required in 2005
0.50
0.30
0.26
47%
1.3
1.5
1.3
0%
13
5.5
5.0
62%
0.32
0.20
0.16
50%
* Nominal values, based on NMHC+NOx standard of 2.4 g/bhp-hr or 2.5 g/bhp-hr provided that NMHC
does not exceed 0.5 g/bhp-hr.
EGR will likely be one of the primary technologies used by truck engines to meet
future emission standards. However, EPA agrees with locomotive manufacturers that
using EGR on locomotive engines at this time would require overcoming significant
technical challenges. Moreover, EGR may not provide the same degree of emission
reduction for locomotives as for trucks since it would not likely be used during high
power operation which is more prevalent in the locomotive duty-cycle than the Federal
Test Procedure transient cycle for truck engines. Nevertheless, it is possible that
moderate rates of EGR may still be used in some instances.
Other technologies that may be on the horizon but are not expected to be available
for widespread use in 2005 are SCR and natural-gas fueled engines. At this time,
there are durability concerns associated with the vibrational and thermal stresses that
would be seen by catalysts used in locomotive applications, and more investigation is
needed on the durability and effectiveness of SCR. More investigation is also needed
on both the operation of, and the fuel infrastructure needed for, natural-gas fueled
locomotive engines. Nevertheless, EPA hopes that the potential to generate emission
70
-------
credits will provide an incentive for manufacturers and/or railroads to participate in
demonstration programs that could ultimately lead to the viability of one or both of
these technologies.
In summary, EPA believes that the NOx and PM standards established for Tier
2 locomotives are the most stringent standards that will be feasible in 2005. While
EPA's analysis suggests that slightly more stringent standards could potentially be
feasible if the effectiveness of each of the technologies listed in Table 4-14 were at the
maximum level within the projected range, EPA believes that this is unlikely to be the
case. In addition, there is some uncertainty in EPA's estimate of current baseline
emissions. Thus, even if a 70 percent reduction in NOx emissions is ultimately
possible, that does not guarantee that a 3.9 g/bhp-hr NOx emission rate (i.e., 30
percent of 13.0 g/bhp-hr) would be achievable. EPA believes that the NOx and PM
standards being established appropriately balance the need for maximum stringency
with the uncertainty inherent in the analysis.
Table 4- 14
Tier 2 Locomotive Technology and Effectiveness
(Expressed as Percent Reduction from Baseline; Increases Shown in Parentheses)
Technology
Four-stroke cycle
Timing retard
Electronic control
Fuel injection*
Rate shaping
Improved turbocharging
Separate circuit
aftercooling
Improved oil control
Combustion chamber
Puff limiter
COMBINED
REDUCTIONS
HC
0-10%
(10-25%)
10-15%
15-25%
0-5%
10-15%
-
5-10%
5-10%
-
50-65%
CO
0-10%
(5-15%)
0-5%
5-10%
0-5%
5-10%
-
-
0-5%
-
10-25%
NOx
-
20-30%
10-15%
-
15-25%
(0-5%)
10-15%
-
-
-
60-75%
PM
0-20%
(15-25%)
5-10%**
20-30%
5-15%
5-10%**
0-5%
5-15%
0-5%
-
45-60%
smoke
better
worse
better
better
even
better
even
better
even
better
better
BSFC
-
(1-2%)
0-2%
0-1%
-
0-0.5%
(0-1%)
-
0-0.5%
-
(2)-2%
* improvements in nozzle geometry, injection pressure, and sac volume
** greater reductions would be expected during transient operation.
It is worth noting that EPA's analysis also projects reductions in HC and CO
emissions that are larger than required by the Tier 2 standards. However, EPA
believes that the projected emission reductions will probably be obtained even without
71
-------
tighter standards for HC and CO, since they are projected to result from the
technologies required to comply with the PM standards. Given that the primary focus
of this rulemaking has been to reduce NOx emissions, EPA believes that it would be
inappropriate to risk compromising the feasibility of the NOx standards by setting very
stringent HC and CO that could limit the ability of manufacturers to use injection
timing retard to lower NOx emissions.
Production Line and In-Use Compliance
Compliance with the new emission standards during production will require that
manufacturers and remanufacturers continue (or institute) proper quality assurance
programs to limit deviation of production locomotives from the design specifications.
Nevertheless, small but significant production variability is expected. As mentioned
previously, EPA expects that manufacturers and remanufacturers will address this
variability by incorporating compliance margins into their designs.
Because of the generally high expectations that railroads currently have for in-use
locomotive performance in terms of power output, fuel consumption, and reliability,
compliance with these standards in use during the full useful life is not expected to be
significantly more difficult for manufacturers and remanufacturers than compliance
at certification. In some cases, however, it may require somewhat more rigorous
maintenance than some railroads are currently practicing. This will be especially true
for Tier 1 and Tier 2 locomotives. Railroads may need to inspect, repair, and replace
(as needed) fuel injectors or power assemblies more frequently than in current practice.
They may also need to more carefully inspect the condition of the turbocharger and
aftercooler systems and repair or replace these components more frequently.
High Altitude and High Ambient Temperature Compliance
Compliance with the emission standards under different ambient operating
conditions may require that manufacturers and remanufacturers design their
locomotives to be relatively insensitive to changes in ambient conditions or to adjust
to such changes to the extent possible. Manufacturers and remanufacturers will need
to continue their practice of designing control features into their locomotive aftercooler
systems that minimize the effect of ambient temperature on the charge air and
combustion temperatures. For barometric pressure (or altitude), manufacturers and
remanufacturers will need to design control features into their locomotive fuel and/or
charge air systems that will minimize the effect of ambient pressure on the engines
air/fuel ratio. This will likely require either charge air designs that compensate for
changes in ambient pressure, or deration of the horsepower out by decreasing the
fueling rate. This technology is employed on many locomotives today.
72
-------
5.0 Test Procedures
This section describes the Federal Test Procedure (FTP) for locomotives. Since the
FTP is, in essence, a part of the emission standards, it is necessary for it to be very
specific in order to minimize testing variability. Thus, the FTP should be considered
the standard test procedure. However, as with EPA's test procedures for other vehicles
and engines, the locomotive test regulations also allow alternate test procedures to be
used, provided that they have been demonstrated to yield results equivalent or
superior to those obtained from the FTP.
The FTP for locomotives is a nominally steady-state test procedure that measures
gaseous (HC, CO, CO2, and NOx), particulate, and smoke emissions from locomotives;
that is, a procedure wherein measurements of emissions are performed with the engine
at a series of steady-state speed and load conditions. Measurement of emissions would
actually be performed during both steady-state operations and during the limited
periods of engine accelerations between notches. The reason for this is that in-use
locomotive operation is not truly steady-state. Rather, locomotive operation is a
combination of short periods of largely steady-state operation at individual notches,
and short transient periods between notch changes. In developing the final test
procedure, EPA sought to ensure that all measured emission rates are representative
of actual in-use emissions.
The test procedures, other than the test sequence, are based largely on the test
procedures previously established for on-highway heavy-duty diesel engines in 40 CFR
86 Subparts D and N. Specifically, the raw sampling procedures and many of the
instrument calibration procedures are based on Subpart D, and the dilute particulate
sampling procedures and general test procedures are based on Subpart N. The most
significant aspects of the test procedures are described below.
5.1 Locomotive Testing
In previous regulation of mobile sources EPA has based its emission standards on
either chassis testing (e.g., on-highway light-duty vehicles) or engine testing (e.g., on-
highway heavy-duty engines). In general, chassis testing is preferred because it more
closely represents the operation of the vehicle in actual use. However, EPA recognizes
that it can be impractical to require chassis testing for some sources. For example,
EPA does not currently require chassis testing for on-highway heavy-duty engines
because it would require an extremely large chassis dynamometer, and because a given
engine model is often used in very different vehicle configurations. For this rule, EPA
determined that it is appropriate to base the emission standards on chassis testing (or
locomotive testing), as is described below.
73
-------
Chassis Testing
Chassis testing for locomotives is reasonably practical because of the inclusion of
electrical alternators as part of standard locomotive design. The power output of the
engine, which is converted to electrical power by the alternator, can easily be
dissipated as heat during an emission test. Thus, EPA believes that engine testing is
not necessary, except for the few specific cases described in the "Engine Testing"
section. Moreover, EPA believes that chassis testing offers significant potential
advantages over engine testing; that is, there are problems that can occur with engine
testing, but that would not be expected with chassis testing. First, the performance
of engine coolant and intake air systems used during engine testing often deviate
significantly from the performance of the actual locomotive systems in use. Second,
engine test facilities are often designed with engine control systems that are different
than the actual in-use systems. This is particularly important for smoke
measurements, since one of the most common means of controlling smoke is to reduce
the fueling rate during transients. Thus, if the fueling controls do not match the in-use
fueling controls, then neither will the smoke measurements. Third, it can be difficult
to simulate the actual in-use load on the engine during engine testing. Each of these
three potential problems could significantly affect testing accuracy of engine testing.
However, the most important problem with engine testing is that engine testing would
be extremely impractical in use, because it would require pulling the engine out of the
locomotive.
Engine Testing
EPA sees two cases in which it is somewhat reasonable to not require chassis
testing of locomotives. First, it is acceptable to allow certification data to be generated
from engine testing of a development engine. The concerns about accuracy are
lessened to some degree by the fact that engine family would eventually be required
to be chassis tested as part of the in-use testing programs. Similarly, EPA is willing
to allow production line testing be conducted on the engine.
The Agency believes that when engine testing is conducted, it is critical that the
testing be as representative of actual locomotive operation as can practically be
achieved. Thus, EPA is requiring that important operating conditions such as engine
speed, engine load, and the temperature of the charge air entering the cylinder be
essentially the same as in a locomotive in use.
74
-------
5.2 Test Sequence
Background
Locomotives differ significantly from road vehicles, in that the power from the
engine is transmitted to the drive wheels by electrical and mechanical components,
instead of only by mechanical components. With road vehicles, the relationship
between engine speed (rpm) and vehicle speed (mph) is mechanically dictated by the
gear ratios in the transmission and in the final drive. Locomotives, on the other hand,
are powered by an engine through an electric alternator to electric motors that are
connected to the drive wheels of the locomotive. The effect of this is that a locomotive
engine is operated at a desired power output and corresponding engine speed without
being constrained by locomotive speed. With the electrical coupling between the
engine and the drive wheels of a locomotive, engine lugging is not possible.
Another design feature unique to railroad locomotive engines is the design and
operation of the throttle. Power settings for railroad engines (throttle position)
generally include eight discrete positions or notches on the throttle gate in addition to
idle and the dynamic brake function. Each throttle notch position is numerically
identified, with notch position one being the lowest power setting (other than idle) and
position eight being maximum power. Because of this design, each notch on the
throttle corresponds to a discrete setting on the fuel delivery system of the engine.
These are the only engine power settings at which the locomotive can operate. The net
effect of this method of control is that the engines can operate at only eight distinct
combinations of fueling rate, power output and engine speed (in addition to idle and
dynamic brake).
As described in Chapter 4, data collected by the freight railroads were recorded
on throttle clocks with a capability for recording throttle position on a second-by-second
basis. This allowed the analysis of the lengths of time that the throttle was
continuously in each throttle notch. The data were reviewed to determine the
continuous periods of time that locomotive engines typically remain at discrete power
settings. These results are summarized in Table 5-1 for times continuously in notch.
Inspection of the data shows that the time that locomotives are continuously operated
at a given power level is typically relatively short. Only for idle, dynamic brake and
notch 8 does the average time in notch exceed one minute. There are two reasons for
this. First, it is often necessary for locomotives to slow down for safety purposes as
they cross intersections in urban areas. Second , even when a constant locomotive
speed is allowed, it often requires that the engineer constantly adjust the throttle to
account for subtle changes in grade.
75
-------
Table 5-1
Average Time Continuously in Throttle Notch (Minutes)
Throttle Notch
Idle
Dynamic Brake
1
2
3
4
5
6
7
8
Line-Haul
2.8
5.6
0.5
0.5
0.5
0.5
0.5
0.6
0.5
4.9
Switching
1.7
NA
0.5
0.5
0.5
0.3
0.9
0.4
0.3
0.9
All Operations
2.5
5.6
0.5
0.5
0.5
0.5
0.5
0.6
0.5
1.2
AAR provided continuous traces of NOx emissions from locomotives manufactured
by both GE and EMD, under the following two conditions. First, starting with the
engine at idle, NOx concentration in the exhaust was measured continuously as engine
power was increased to full power and returned to idle. The time in each throttle notch
was one minute during both increases and decreases in power. In the second test, the
same sequence of power settings was employed but the time in each notch was
increased to three minutes. Results from the testing performed are shown in Appendix
M. The results of this testing showed: 1) that NOx concentrations in the exhaust
during decreases in power repeated the values measured during increases in power,
2) that for current locomotive calibrations, equivalent NOx results would be obtained
from the one minute in notch test and the three minutes in notch test, and 3) that the
values obtained from the short times in notch tests equaled results obtained after
prolonged periods of equilibration.
Sequence for Testing
The test sequence for locomotives and locomotive engines calls for the locomotive
or engine to be operated at idle, dynamic brake and at the eight throttle notch
positions. The test sequence begins with the engine at idle and at operating
temperature and proceeds through dynamic brake and each power level to rated power.
To begin the sequence, the engine is started, if not already running, and warmed up
to normal operating temperature in accordance with warm-up procedures for in-service
locomotives as specified by the manufacturer. For locomotive testing, the engine
76
-------
remains in the locomotive chassis, and the power output would be dissipated as heat
from resistive load banks (internal or external). The engine is considered to be warmed
up, and ready for emissions testing when coolant and lubricant temperatures are
approximately at the normal in-service operating temperatures for these materials as
specified by the manufacturer. After the engine has reached normal operating
temperature, the engine is operated at full power (i.e., highest power notch) for 5
minutes, then returned to idle, or low idle if so equipped. Testing must be commenced
within 15 minutes after the end of the 5 minutes of full power to prevent the engine
from cooling off significantly. The 5-minute period at full power is intended to ensure
that the engine is at a realistic operating temperature, and to improve test
repeatability.
EPA is not including engine starting as part of the test procedure because the
starting of locomotive engines which are fully cooled to ambient conditions seldom
occurs in use. Locomotive engines tend not to be shut down, especially in cold weather,
because it can be difficult to start these engines if they are allowed to cool to relatively
low ambient temperatures and because of the potential for engine damage due to
leakage of coolant. When the engines are shut down, restarts are generally performed
before significant cooling of the engine occurs to avoid or minimize restart problems
and coolant leakage problems.
Measurement of exhaust emissions, fuel consumption, power output, etc. begins
with the locomotive idling, after the engine is warmed up. These measurements
continue as the locomotive is operated in each notch. The minimum duration of each
test point is 6 minutes, except for the maximum power point (notch 8), where the
minimum duration of operation is 15 minutes. EPA concluded that longer periods of
notch operation are not required to develop adequate test data. The most time-
consuming requirement of emissions measurement is the time required to sample
particulate matter. When the highest feasible particulate filter face velocity is
employed, minimum sample time required to collect accurate particulate material
samples from current locomotives is about four minutes per engine power level. This
four minute value was increased to six minutes to provide sufficient time for accurate
sample collection, especially under the Tier 2 standards for particulate emissions. In
cases where a manufacturer is concerned about measurement variability, it will be
allowed to perform replicate tests and to use the average of the replicate
measurements. However, a manufacturer would not be allowed to replicate only tests
that are high. If a manufacturer chooses to replicate tests, then it must replicate all
tests.
It should also be noted that in-use locomotives are rarely operated continuously
in a single notch for longer than these minimum sampling periods, especially for the
intermediate power notches (1 through 7). Significant continuous operation for longer
periods only occurs for idle, dynamic brake, and notch 8, but still accounts for less than
a third of all operation. Such operation should not have any adverse environmental
77
-------
impact because the regulations prohibit any changes in the engine calibration after the
sampling period that would increase emissions. It should also be noted that total mass
emission rates (g/hr) for idle and dynamic brake are relatively low, and that long
periods of continuous operation in dynamic brake and notch 8 are likely to be in rural
areas. Both of these factors further minimize any potential for adverse environmental
impacts from such operation.
In the event of test equipment failure during data acquisition, testing may be
resumed by repeating the voided test mode, provided the engine is at normal operating
temperature. EPA is allowing three approaches for ensuring that the engine is at
normal operating temperatures. In the first approach, when a test mode is voided, the
engine is returned to the next lower notch position (e.g., notch 3, if the notch 4 test
mode is voided), and kept running in that notch until the problem is corrected. In the
second approach, the engine is returned to idle while the problem is corrected, then the
two test modes prior to the voided mode are repeated. With either approach, the
sequence is then continued, starting with the voided test mode. The third approach is
simply to start the test sequence over.
In order to minimize the testing burden, EPA is requiring that locomotives
equipped with dynamic brake (DB) only be tested in one DB position. The DB position
to be tested would be the notch generating nearest to 75 percent of the maximum
braking capacity. This is the DB notch that EPA expects will be most environmentally
important. While other DB positions would not have to be tested in the same manner,
they would have to incorporate the same emission controls (see "Defeat Devices"
section). Given that periods of dynamic brake use are limited, especially in urban
areas, the potential for adverse environmental impacts from this approach is fairly
minimal.
Similarly, for passenger locomotives that generate hotel power from the main
propulsion engine, EPA is not requiring that testing for compliance be conducted with
the engine producing hotel power. However, manufacturers and remanufacturers
would be required to incorporate similar emission controls for hotel mode, especially
in the area of injection timing (see "Defeat Devices" section). The Agency also retains
the authority to require emission data collection from the locomotive or engine when
it is generating hotel power.
78
-------
Alternate Locomotive Configurations
The Agency recognizes that the potential exists for future locomotives to include
additional power notches, or even continuously variable throttles, and is creating
alternate testing requirements for such locomotives. Using the standard FTP sequence
for such locomotives would result in an emissions measurement that does not
accurately reflect their in-use emissions performance. Instead locomotives having
additional notches would be tested at each notch, and the mass emission rates for the
additional notches would be averaged with the nearest "standard" notch.
Locomotives having continuously variable throttles would be tested at idle,
dynamic brake, and 15 power levels assigned by the Administrator (including full
power), with average emission rates for two power levels (excluding full power)
assigned to the nearest "standard" notch. The 15 power levels assigned represent one
level for full power and two, to be averaged, for each of the seven intermediate power
levels used on current locomotives. The Administrator retains the authority to
prescribe other procedures for alternate throttle/power configurations.
Information provided to EPA by manufacturers on engine power levels in the
various notch positions showed that the power in propulsion notches 1 through 8
followed a similar pattern when expressed as a percentage of the rated power of the
engine. These power levels, which are shown in Table 5-2 below, would be used to
adapt the normal duty-cycle weightings. As an example, for a locomotive equipped
with 9 power notches (4, 11, 23, 35, 48, 64, 80, 90, and 100 percent of rated power),
emissions measured from notches 1 through 6 would be weighted in the usual manner,
notches 7 and 8 (80 and 90 percent of rated power) would be averaged and weighted
by the normal notch 7 weighting factor, and notch 9 (100 percent of rated power) would
be weighted by the normal notch 8 weighting factor.
Table 5-2
Typical Power Distribution by Notch
Throttle Notch
Percent of
Rated Power
1
4.5
2
11.5
3
23.5
4
35.0
5
48.5
6
64.0
7
85.0
8
100
79
-------
5.3 Sampling and Analysis
Gaseous Sampling and Analysis
Gaseous exhaust pollutants (HC, CO2, CO, and NOx) are measured by drawing
samples of the raw exhaust directly to chemical analyzers (with appropriate filtering).
The sample is generally drawn through these analyzers using pumps that are located
downstream of the analyzer. The sampling procedures are based on EPA's previous
experience sampling diesel exhaust from on-highway heavy-duty engines (40 CFR 86).
The required analyzers are: a chemiluminescence analyzer for NOx, a heated flame
ionization detector (HFID) for HC, and a nondispersive infrared (NDIR) detector for
CO and CO2.
Testing of current locomotives has shown that exhaust concentrations generally
reach their steady-state levels shortly after a notch change. Thus, EPA is requiring
that a single steady-state concentration be used for gaseous emissions calculations, for
each mode and each pollutant. EPA has created special provisions for locomotives that
have long stabilization periods, which could require integration of the concentration
in some cases. These provisions are criteria that specify when a steady-state emission
measurement is considered to be representative of the typical period in notch. The
criteria were developed so that a single steady-state value would be allowed for each
mode for locomotives that reach their steady-state emission levels as soon after a notch
change as most current locomotives do.
In order to ensure good reliability of test results, EPA is also establishing
calibration and verification requirements similar to those applicable to on-highway
heavy-duty engines. However, in these regulations, unlike the regulations for highway
engines, the Agency is not establishing dilute sampling procedures for the total
exhaust stream for gaseous emissions because it is not necessary to dilute the total
exhaust stream prior to sampling with steady-state operation. In addition, the
equipment that is required for dilute sampling would be very large and expensive.
Nevertheless, not including such provisions does not preclude the use of dilute
sampling as an alternative procedure.
Particulate Sampling and Analysis
Particulates are measured by drawing a sample of the exhaust through a filter
and weighing the mass of particulate collected. Particulate sampling requires dilution
of the sample to lower the temperature below 125°F, which minimizes evaporation of
volatile particulate matter from the filters. The particulate sampling procedures are
also based on the procedures found in 40 CFR 86. However, since the locomotive test
cycle is essentially steady-state (and thus the volumetric flow rate of the exhaust is
essentially constant for each test mode), the dilution and sampling systems can be
simpler than are required for on-highway engine testing. The regulations do not
80
-------
require any critical flow venturi, or heat exchangers, since these are only used to
maintain proportional sampling when the exhaust volumetric flow rate is changing
continuously.
The regulations include specific requirements for the weighing of particulate
filters. These requirements are intended to minimize contamination of the filters by
dust or water vapor, which could otherwise lead to significant inaccuracy in the
measurement. Also included in the regulations are recommended target loadings to
ensure that sufficient particulate mass is collected to allow a meaningful
measurement. Finally, the regulations include limits on the range of acceptable filter
face velocities (i.e., the volumetric sample flow rate divided by the cross-sectional area
of the filter) based on testing of on-highway engines72 and industry comments.
Exhaust Ducting and Sample Probes
During locomotive testing, the exhaust emissions are routed through a metal duct,
or exhaust stack extension, which is located directly over the exhaust outlet of the
locomotive. It is important that the duct design is sufficiently open so that it does not
cause any significant increase in the exhaust back pressure in the engine. The purpose
of this duct is to prevent disturbance or dilution of the exhaust plume during sampling,
and to provide for standardization of the sample probe location. The regulations call
for the sample probes to be located in the duct between 2 and 5 feet downstream of the
exhaust outlet during locomotive testing, or the nearest practical equivalent during
engine testing. Although sample probe is not expected to have a large impact on
repeatability, this specification is appropriate since it does not represent a significant
burden. The allowance for different sample probe locations during engine testing is a
recognition that issues such as safety or accessibility of the probes in engine test
facilities may require that the probes be located more than 20 feet downstream of the
exhaust outlet.
When testing locomotive engines with more than one exhaust outlet, the
regulations require that all exhaust outlets be ducted together prior to sampling.
However, for locomotive testing, it is not necessary to duct the exhaust outlets
together, provided that a proportional sample is collected from each outlet, and a check
is performed to ensure that the exhaust flows from each outlet are similar. More
specifically, the regulations require that CO2 concentrations be measured from each
outlet, and that they be within 5 percent of each other. EPA believes that similar CO2
concentrations are a reasonable indication that the exhaust flows are similar.
Assuming that an engine was designed to have similar exhaust flows for each outlet,
then the types of malfunction that would cause dissimilar flows should also cause
72 Guerrieri, D., V. Rao, and P. Caffrey, "An investigation of the Effect of Differing
Filter Face Velocities on Particulate Mass Weight from Heavy-duty Diesel Engines," SAE
Paper No. 960253, February 1996.
81
-------
dissimilar CO2 concentrations. For example, malfunctions that caused a loss of power
in a given cylinder will typically lead to an increase in fuel consumption, as the engine
controls increase the fueling rate in an attempt to maintain the power output. Such
malfunctions would thus change the air/fuel ratio and should result in increased CO2
concentrations for the exhaust outlet of the malfunctioning cylinder.
The three sample probe designs specified in the regulations are intended to collect
a representative sample during testing. The basic designs have been successfully used
previously for testing other diesel engines. The gaseous sample probe design is the
same as is specified for general nonroad engine testing in 40 CFR 89. The dilute
particulate probe design is the same as the design used for on-highway engine testing
(40 CFR 86). The raw particulate probe design was used previously for diesel engine
testing at SwRI. All three specific designs were tested for EPA at SwRI, and shown to
work very well for locomotive emission testing.
Smoke Measurement
Smoke is measured with a smoke opacity meter mounted on top of the exhaust
stack extension. The instrument specifications are the same as those used for on-
highway heavy-duty engines. The meter measures opacity of the exhaust across the
longest (nondiagonal) dimension of the extension. Emission measurements are
normalized, using the following equation, to be equivalent to measurements made with
an instrument that had a one-meter long (3.281 feet) path length. This equation is
derived from the Beer-Lambert law that states that the logarithm of the transmissivity
of a gas (i.e., one minus the opacity) is proportional to the path length.
1-
1-
N
m
100
\IL
Where Nn is the normalized percent opacity, Nm is the measured percent opacity, and
L is actual path length of the measurement system in meters (only that portion of the
optical path that actually passes through the plume).
For simplicity, EPA has also included, as an appendix to the regulations, a table
which allows an approximate comparison of measured smoke levels directly to the
standard without normalization. This table shows comparable smoke standards for
various path lengths in 10-centimeter increments. The comparable smoke standards
were calculated from the low end of each range (e.g., 70.0 cm for the 70.0-79.9cm
range), so that use of these standards for non-normalized measurements would not be
less stringent than the official standard for normalized measurements.
82
-------
Power and Fuel Measurement
Brake horsepower is the sum of the horsepower supplied to the main alternator
and the mechanical horsepower required to operate accessories such as a secondary
alternator. Power required to operate oil and fuel pumps, or to circulate coolant for the
engine are not included in brake horsepower. The horsepower supplied to the main
alternator is calculated as the product of the output voltage and current, divided by the
alternator efficiency. The output voltage and current are measured directly using a
voltmeter and current shunt. The alternator output may also be read from an onboard
measurement system, provided that it meets the accuracy and precision requirements
specified in the regulations. The alternator efficiency and accessory loads do not need
to be measured during testing, provided that this information is supplied by the
manufacturer or remanufacturer. The alternator efficiency must be expressed as a
function of power input. The effect of temperature on the alternator efficiency must
also be taken into account, where possible.
Fuel flow rates are determined by measuring the change in mass of an external
fuel supply tank. For locomotive testing, the fuel intake and recycle lines are
disconnected from the locomotive fuel tank and connected to the external tank. For
many systems, a heat exchanger will be necessary to dissipate the heat from the
recycled fuel. It is important to maintain a relatively constant fuel temperature, in
order to obtain a stable measurement of the mass of fuel in the tank. Also, given
currently available technology, a one-minute averaging period is recommended in order
to obtain an accurate fuel flow rate measurement for all notches except idle, where a
three-minute averaging period is recommended.
Natural Gas-Fueled and Alcohol-Fueled Engines
EPA is currently aware of only a few natural gas-fueled locomotives, and is not
aware of any alcohol-fueled locomotives. Nevertheless, EPA believes that it is
appropriate to include test procedures for such locomotives in these regulations. For
this reason, the NMHC, alcohol and aldehyde measurement procedures that are
currently applicable to on-highway natural gas- and methanol-fueled engines (40 CFR
part 86) are being used for natural gas- and alcohol-fueled locomotives. EPA
recognizes, however, the possibility of unforeseen problems that could result during the
use of such procedures with locomotive engines, especially with alcohol-fueled
locomotives (which currently do not exist). There is a lack of information on whether
the specifications for dilute alcohol and aldehyde sample temperatures and flow rates
are appropriate for locomotives, as well as a complete lack of such specifications for raw
exhaust. Previous testing of highway vehicles has shown the potential for losses of
alcohols and aldehydes through condensation if the sample line temperature is too low,
or chemical reactions if it is too high. Nevertheless, at this time, EPA believes that it
is appropriate to specify the on-highway procedures, but may reconsider alcohol and
83
-------
aldehyde sampling issues on a case-by-case basis, should alcohol-fueled locomotives
come into use. Other aspects of the test procedure are the same as for diesel
locomotives.
Calculations
The calculations specified for raw exhaust analysis are based on those in Subpart
D of part 86. The algebraic form of some of the equations was changed slightly to
clarify their meaning. Also, equations for both wet and dry analysis are included for
each regulated species. The greatest change from the Subpart D calculations is that
the calculation of the wet-to-dry conversion factor (Kw) is done in several steps, with
iteration, instead of in a single equation. The calculation of DH2O is an algebraic
solution of a hydrogen balance around of the engine, excluding hydrogen bound to
organic matter in the exhaust. The sources of hydrogen are water entering in the
intake air (2Y»DVolair) and hydrogen released from the fuel during combustion
(2a(DCO2+DCO)DVol, where DCO2 and DCO are volume fractions);73 the small
amount of hydrogen released from incomplete combustion reactions which do not
produce CO2 or CO is assumed to be negligible. The hydrogen in the exhaust is
assumed to be in two forms: water and elemental hydrogen (H2). The elemental
hydrogen is assumed to be in equilibrium with respect to the following reaction:
H2 + CO2 = H2O + CO (equilibrium constant = K)
The calculation of DVolair is an algebraic solution of a balance of the number of dry
moles (excluding water) in the gas phase; the difference in the number of dry moles
before and after combustion, neglecting H2 formation and volatilization of the fuel, and
assuming that all NOx produced is in the form of NO, is the amount of O2 consumed
minus the amount of CO2 and CO produced. The calculation of DVol is an algebraic
solution of a carbon balance of the moles of carbon coming into the engine (Wf/CMWf)
and the moles of carbon in the exhaust (DVol-Vm» (DHC+DCO+DCO2), where DHC is
the volume fraction of organic carbon (ppmC/106) and DCO and DCO2 are volume
fractions (ppm/106)). The regulations also include an approximate non-iterative option
for calculating Kw. This option uses an approximate calculation for DH2O that was
recommended by the Engine Manufacturers Association in its comments on the
NPRM.74
The calculations specified for dilute exhaust analysis are identical to those in
Subpart N of part 86, except that the equations for calculating the dilution factor and
the fraction of exhaust that is diluted (if partial dilution is used) are new. The dilution
73 Y = the water vapor concentration of the intake air expressed as a volume fraction; and
a = the atomic hydrogen/carbon ratio of the fuel.
74 Docket item #A-94-31-IV-D-36.
84
-------
factor (DF) equation comes from a CO2 balance around the dilution tunnel, and the
fraction diluted (Vf) equation comes from a mass balance of the carbon in the diluted
exhaust with the carbon coming into the engine in the fuel. The dilution factor is the
ratio of the volume of dilution air to the volume of exhaust sample that is diluted. (The
ratio of the total volume of dilution air plus exhaust sample to the volume of exhaust
sample that is diluted is equal to one plus DF.)
5.4 Fuel Quality
Effects of Fuel Quality on Emissions
Changes in diesel fuel quality can have a significant effect on exhaust emissions.
Perhaps the most important fuel parameter, with respect to emissions, is the weight
fraction of sulfur in the fuel. Sulfur is a normal contaminant in crude oil and diesel
fuel that is converted primarily to sulfur dioxide during combustion. A small amount
of the sulfur, however, is converted to sulfate particulate. Thus, lowering the sulfur
content of diesel fuel can lead to lower particulate emissions. A study by SwRI showed
that lowering the sulfur content of diesel fuel from 0.315 weight percent (a typical in-
use level for nonroad engine fuel) to 0.033 weight percent resulted in particulate
reductions of 0.05 to 0.08 g/bhp-hr from uncontrolled locomotives.75 Other studies have
been less conclusive. It is important to note that, while the effect of sulfur on
particulate levels is expected to vary from engine to engine because of different
combustion properties, it should be relatively independent of total particulate levels.
This is because sulfate particulate results from the oxidation of the fuel sulfur, rather
than from incomplete combustion, which is the cause of organic particulates. It is
known that railroads do occasionally purchase low-sulfur diesel for use in their
locomotives, however, EPA is not aware of any reliable data regarding the relative
amounts of high-sulfur and low-sulfur fuel used by the railroad industry. It is probable
that the nationwide fraction of total fuel use that is low-sulfur is on the order of ten
percent.
There are few data available on the effects of other fuel parameters on emissions
from locomotives. However, data collected from on-highway heavy-duty emission
testing have shown that increasing cetane number and decreasing aromatics content
contribute to somewhat lower NOx and PM emissions. A combined reduction in
aromatics content as well as an increase in cetane number was seen to give additional
benefits in the reduction of NOx and PM.
75 Emission Measurements, Locomotives, Southwest Research Institute Report for EPA,
August, 1995.
85
-------
Test Fuel Specifications
The Agency is establishing test fuel specifications for compliance testing
(certification, PLT and manufacturer/remanufacturer in-use testing) which are
intended to be representative of in-use nonroad diesel fuel. More specifically, they are
consistent with test fuel specifications for on-highway heavy-duty engine certification
testing, with the exception of the sulfur specification. EPA is establishing a lower
sulfur limit of 0.2 weight percent. The cetane and aromatics requirements are a cetane
number of 40 to 48 and an aromatics content of at least 27 percent. These
specifications are intended to approximate reasonable, but somewhat worst case in-use
conditions. Since EPA is not regulating in-use fuel quality for locomotives, there is no
reason to believe that in-use locomotives will use only low sulfur on-highway fuel,
especially given the potential price differences between low and high sulfur diesel
fuels, and potential availability problems in some areas of the country. Should EPA
regulate in-use locomotive fuels in the future, it would also adjust the test fuel
specifications as appropriate. It should be noted that any improvements in in-use fuel
quality would also necessitate a reconsideration of the appropriate levels of the
standards.
5.6 Atmospheric Conditions During Testing
As is described in Chapter 3, ambient conditions such as temperature, humidity,
and barometric pressure are known to affect emissions from diesel engines. These test
conditions need to be limited to some extent to keep the compliance burden for
manufacturers and remanufacturers reasonable. However, in order to ensure that
these regulations achieve the expected environmental benefits during normal railroad
operations, and to allow for outdoor testing, EPA is specifying a fairly wide range of
temperatures and pressures for testing, with no restrictions on humidity. The specified
range of test conditions is an attempt to balance the environmental needs with those
of the manufacturers and remanufacturers. An important factor which supports the
wide range is that the large size of locomotives makes outdoor testing desirable. While
indoor testing of a locomotive engine under controlled temperature conditions could
reasonably be expected to be practical, testing of a complete locomotive indoors would
be more costly and may prove to be very difficult to implement in many cases.
86
-------
Specifically, EPA is establishing a range for the test temperature of 45-105°F.
The upper limit temperature of 105°F is specified to address summer high
temperatures throughout the southern and western areas of the country, and to
facilitate testing of locomotives under most conditions which can occur over much of
the nation during the summer. By choosing this upper limit for the test temperature,
EPA expects to also address any potential loss in control of NOx emissions, and even
PM and smoke, if charge air density decreases sufficiently at very high ambient
temperatures. There is a potential for NOx emissions to increase at very high
temperatures if the effectiveness of charge air cooling decreases at these high
temperatures.
The lower limit for the test temperature (45°F) was based on: 1) the prevailing
temperatures in the areas of the nation where locomotives are manufactured and
remanufactured (e.g., Pennsylvania, Illinois and Idaho), 2) the fact that ozone pollution
is essentially a warm weather problem, and 3) the potential test variability that could
occur at lower temperatures. Under low ambient temperatures, the effectiveness of
charge air cooling as a NOx control strategy can be expected to increase, especially if
engine coolant is not used as the cooling medium. At the same time, some increase in
PM and smoke emissions can be expected if excessive cooling of the charge air is
allowed to occur. This lower limit should allow outdoor testing during much of the
year, without unreasonable test variability problems. Acquisition of test data when
prevailing ambient temperatures at manufacturers' or remanufacturers' facilities are
below the lower test limit can be achieved by providing indoor facilities. EPA could
also allow testing at lower temperatures, but only in cases where a manufacturer or
remanufacturer can demonstrate that it would not result in any compliance advantage.
EPA also sees the need to include compliance with the standards up to 7000 feet
above sea level to achieve control of locomotive emissions at high altitude. This is
necessary because some emission control technologies suitable for the control of NOx
emissions, e.g., delaying the start of fuel injection, will tend to exacerbate other
emissions, e.g., PM and smoke, especially at high altitudes. The inclusion of
equipment that would assure the same absolute pressure of the charge air in the
intake manifold at both low and high altitudes would reduce or control this concern.
However, there is a significant burden associated with testing over a very wide range
of barometric pressures to demonstrate compliance, which EPA believes should be
limited. Therefore, EPA is establishing a lower limit on ambient pressure of 26 inches
of mercury for testing. This corresponds to an elevation of about 4000 feet above sea
level. EPA is requiring compliance at higher elevations, but is only requiring an
engineering analysis to demonstrate the performance of the locomotive emission
controls at elevations between 4000 and 7000 feet (about 23 inches of mercury).
87
-------
In the proposal, the Agency recognized the need to correct NOx emissions for the
effects of humidity and temperature during testing, but it did not propose specific
correction factors. EPA considered using the NOx-humidity correction factor that is
currently being used for on-highway and general nonroad diesel engines (40 CFR 86
and 89), but concluded that the data upon which that correction factor was based are
not adequate for this rulemaking. In particular, EPA had concerns about the
applicability of data from older pre-control on-highway engines to current and future
locomotives that incorporate NOx-reduction technologies. More importantly, however,
the data are inappropriate as a basis for such correction factors for locomotives because
the range of test conditions for locomotives being established is much broader than was
used in the collection of that data. Therefore, EPA contracted with SwRI to provide an
analysis that would support the development of locomotive specific correction factors.
The contractor provided a partial analysis, which EPA used for this rulemaking, but
was not able to complete the analysis before the rule was finalized. This partial
analysis can be found in the docket.76
That analysis recommended the following correction factors for the effects of
ambient humidity and temperature:
C +C e(^°-0143)(10-714)
^rr—
n
C1+C2^0-0143)(1000//)
Where:
C1 = -8.7+164.5e-°-0218(A/F)
C2= 130.7 +3941e-°-0248(A/F)
H = The specific humidity on a dry basis of the intake air (grams of water per kilogram of dry
air).
(A/F) = Mass of moist air intake divided by mass of fuel intake.
T30 = The measured intake manifold air temperature in the locomotive when operated at 30°C.
TA = The measured intake manifold air temperature in the locomotive as tested.
It is important to note that the correction factor for temperature actually uses the
intake air temperature rather than the ambient temperature. This is because it is the
effect of ambient temperature on the intake air that matters most, but this effect is
highly variable from locomotive to locomotive. EPA is not allowing correction for the
76 Docket item #A-94-31-IV-A-2.
88
-------
effects of ambient temperatures above 86°F. For most current locomotives, ambient
temperature does not significantly affect intake air temperature over a broad range of
ambient temperatures. This may not continue to be true for future locomotives, and
EPA does want to provide an adjustment that would "correct" for a poorly designed
aftercooler system. Finally, EPA recognizes that changes in ambient temperature can
have other effects on measured emission rates (other than the effect of intake air
temperature), but does not have any information on the effects at this time.
Since the effects of humidity and temperature on NOx emissions from locomotives
are not fully understood at this time, EPA has decided to include conservative default
correction factors in the final rule (i.e., factors that are more likely to overestimate
emissions rather than underestimate emissions), but to allow manufacturers and
remanufacturers to use their own correction factors where they are appropriate for
their specific locomotives. The correction factor being used in the final regulations
(KNOx) is the product of the temperature and humidity correction factors developed by
the contractor (KT • KH) multiplied by an adjustment factor (F) to address uncertainty
in the estimated correction. The adjustment is defined as:
F = [l^0.25(\og(KTKH-)2-)}
Thus, the final correction factor is calculated as:
KNOx=
The effect of this uncertainty factor is shown in Table 5-3. As can be seen from the
table, the uncertainty factor is always greater than one, and becomes smaller as the
correction factor becomes closer to 1.000. Thus, this factor appropriately accounts for
the uncertainty in the correction, which is least for those test conditions which are
closest to 86°F and 75 grains (or 10.71 g of moisture per kg of dry air).
89
-------
Table 5-3
Adjustment to Correction Factors
(KT)(KH)
0.800
0.900
0.950
0.985
1.000
1.030
1.050
1.100
1.200
Uncertainty Factor
1.048
1.023
1.011
1.003
1.000
1.006
1.011
1.021
1.040
•"^•NOx
0.839
0.921
0.961
0.988
1.000
1.037
1.061
1.123
1.248
The Agency recognizes that the correction factors being established in these
regulations may not be appropriate for the long term, but believes that they are
appropriate at this time. During the first several years of this program, EPA expects
that nearly all manufacturers and remanufacturers will perform engine testing rather
than locomotive testing, and will therefore be able to perform all testing under
controlled conditions where the effect of the correction factors will be small (i.e. near
86°F and 75 grains). Moreover, where the manufacturer or remanufacturer believes
that the default correction factors penalize them, they will be able to develop and use
their own correction factors. Nevertheless, EPA expects to refine these correction
factors in the future when better information becomes available.
5.7 Other Issues
Defeat Devices
A defeat device is a device or element of design that reduces the effectiveness of
emissions control during actual operation, but that does not significantly affect
emissions control under test conditions, or affects it to a significantly lesser degree.
Such devices are prohibited by the Clean Air Act. The procedures and penalties for a
finding of a violation of the defeat device prohibition are detailed in the regulations.
Penalties for a violation could be as much as $25,000 per locomotive per day. EPA
recognizes a significant potential for the use of defeat devices in locomotives, especially
those equipped with electronic engine controls. Some examples are described below.
The simplest example of a defeat device would be an electronically controlled
engine that had two injection timing calibrations: a low emission calibration (e.g.,
retarded timing) that is used during testing, and a low fuel consumption calibration
90
-------
(e.g., advanced timing) that is used all other times. Such a device could easily
recognize that a locomotive was being tested because there would be no power going
to the traction motors.
A second potential defeat device would be a calibration that had high emissions
just after a notch change, but had much lower emissions near the end of the test mode
(e.g., five minutes after the notch change). This would allow the locomotive to be
calibrated for low fuel consumption during the first few minutes, but still have a
"steady-state" emission level that is reasonably low. This is significant because the six-
minute sampling period required by this regulation is significantly longer than the
typical time in the intermediate power notches in actual use. Thus, a locomotive could
be operated in a high-emission/low-fuel consumption mode the majority of the time in
use, and spend relatively little time in the low-emission mode. This issue is addressed
to some extent by the steady-state stability criteria included in the test procedure
regulations. Similarly, a calibration that advanced timing after the end of the
sampling period would also be considered to be a defeat device.
It should be noted that EPA does recognize that, to a limited extent,
manufacturers can have legitimate reasons for calibrations that change with time.
Most notably, it is often necessary to increase the fueling rate slowly after a notch
change in order to prevent smoke emissions. In these cases, the variability of the
calibration generally lasts only a few seconds, and should not adversely affect any
emissions. EPA would not consider such a calibration to be a defeat device, provided
that the manufacturer could provide adequate technical justification at the time of
certification.
A more subtle form of defeat device would be a calibration that varied significantly
by notch to the extent that it would influence the operation of the locomotive,
discouraging the use of certain notches. EPA believes that if a locomotive was
calibrated with severely retarded timing in one notch, operators may rarely use that
notch because of concerns about fuel consumption or power output. As with the
previous example, EPA would consider such a calibration to be a defeat device, unless
the manufacturer provided adequate technical justification for the calibration.
Passenger locomotives designed to provide hotel power from the traction engine
create the potential for a defeat device because the engine can be operated in an
infinitely variable mode, depending on the hotel power load being placed on the engine.
Such an engine could be calibrated to have low emissions when generating no hotel
power (i.e., the test condition), but have higher emissions whenever it generates hotel
power. Such calibration would clearly be a defeat device.
If a locomotive engine was never subject to testing in the locomotive chassis, it is
possible that the engine could be tested at slightly different conditions (e.g., fueling
rate or load) than those at which it would be operated in use, and thus could have a
91
-------
completely different calibration and emission performance. This is one of the reasons
why EPA believes that it is necessary that every engine model be subject to in-use
chassis testing.
Finally, unnecessary loading of the engine at idle and during dynamic brake
operations by means such as operation of cooling fans, etc., when not required for the
proper operation of the engine or locomotive, would also be considered to be a defeat
device. EPA expects that any loads imposed during testing will also generally be
imposed during actual locomotive operations. By requiring that any such load be
imposed in use, it is expected that concerns about maximizing fuel efficiency will
curtail any unnecessary loading of the engine during testing.
Idle Shutdown
The Agency is finalizing a regulatory incentive for the development of an
automatic shutdown mechanism that could shut off an engine automatically after some
extended period of idling. (Current locomotive engines tend not to be shut down for
long periods of time, especially in cold weather, because it can be difficult to start these
engines if they are allowed to cool to relatively low ambient temperatures and because
of the potential for engine damage due to leakage of coolant.) The approach would be
to reduce the weighting factor for the idle emission rate for engines equipped with
automatic shutdown mechanisms, but use the higher power weighting factor that is
specified in the regulations. This approach would account for the emissions benefits
of a shutdown mechanism whereas the standard calculations would not.
At this time, EPA is not establishing specific alternate weighting factors for these
locomotives because it does not have adequate information to accurately predict how
much idling time would be reduced by such a feature. The Agency expects that an
automatic shutdown would not have much of an impact on idling time for newer line-
haul locomotives because they rarely are left at idle for long periods of time. However,
an automatic shutdown mechanism may have a very significant impact on idling time
for switch locomotives. Nevertheless, even for switch locomotives, the impact is hard
to predict at this time. The impact would be greater for locomotives that shutdown
after a shorter period of time, or for applications in which long idling periods are more
common. Thus, EPA will require that any manufacturer or re manufacturer desiring
credit for an automatic shutdown mechanism demonstrate the average percent
reduction of idling time that it would achieve.
This option is expected to be especially useful for compliance with the switch-cycle
standards because the high idle weighting factor in the switch cycle (0.598) will put
significant pressure on manufacturers and remanufacturers to reduce idling emissions.
With this option, a manufacturer or remanufacturer would get the same credit for
reducing the idle emission rate (g/hr) by half as it would for reducing idling time by
half. For example, consider two locomotives: 1) a locomotive with an idle emission rate
92
-------
of 350 g/hr; and 2) a locomotive with an idle emission rate of 700 g/hr, but with an
automatic shutdown mechanism that was shown to reduce idling time by half.
Assuming that power output rates at all notches are the same, and that emission rates
for all non-idle notches are the same, then the two locomotives would have the same
calculated cycle-weighted emission rates. This is because the contribution of idle
emissions to the cycle-weighted total would be 209 g/hr in both cases; (350)X(0.598) in
the first case, and (700)X(0.598)X(0.5) in the second case. (Note: the weighting factor
for the power output would be 0.598 in both cases.) The two cases would also be
equivalent from an environmental perspective.
Correlation Criteria for Alternate Measurement Systems
EPA has included as an appendix to the regulations a set of recommended criteria
for demonstrating equivalence of alternate measurement systems. The criteria specify
the appropriate test procedures, number of replicate tests, and degree of correlation.
These criteria are based on recommendations made by the Engine Manufacturers
Association (EMA) in their comments on the NPRM.77 The final criteria differ from
EMA's criteria in that they allow less variation for systems that measure low (while
EMA allowed 5 percent variation for both high and low measurements), and they
include specific requirements for replicating outliers. It is important to note that these
criteria are merely recommendations. Manufacturers are allowed to submit for EPA
approval other types of data which they believe demonstrate equivalence. These
recommendations, however, provide the manufacturers more certainty regarding the
type of data that EPA will normally expect.
77 Docket item #A-94-31-IV-D-36.
93
-------
6.0 Emission Benefits
This section discusses the emission benefits expected from this rulemaking.78
Emission benefits are presented both in terms of percent reductions and tons of
pollutants reduced annually. This analysis separately addresses the impacts of:
1) Class I (freight-only) line-haul locomotives
2) Class I (freight-only) switch locomotives
3) Class II and III locomotives
4) Passenger locomotives
Emission benefits were calculated in the following manner. First, as described in
Chapter 4, controlled and uncontrolled emission rates for in-use locomotives were
estimated. Average in-use emission factors were estimated by assuming purchase,
scrappage, and remanufacturing behavior for each type of locomotive service. National
emission inventories (tons per year) were calculated from the emission factors and fuel
consumption data. Reductions were calculated from a projected 1999 baseline. It is
important to note that locomotive use, purchase, scrappage, and remanufacturing
schedules vary from railroad to railroad, and that they are all relatively sensitive to
economic conditions. Therefore, this analysis, which is based on typical historical
behavior, merely represents the Agency's best estimate of impacts given the
information available at the time of the analysis.
6.1 Methodology
National emission inventories were calculated for each type of locomotive service
by first multiplying the fuel consumption rates (gal/yr) by a conversion factor of 20.8
bhp-hr/gal79 to obtain total fleet bhp-hr/yr values. These fleet bhp-hr/yr numbers were
then multiplied by the applicable fleet average emission rates to calculate emissions
inventories (tons/yr). The fleet average emission rates for each year were calculated
based on the number of each type of locomotive projected to be in the fleet at the end
of the respective year. The total reductions expected for each future year were
78 The results described in this chapter are identical to those found in the December 1997
version of this document, including several minor errors. Results of corrected analyses are
contained in Appendix O.
79 This conversion factor was calculated from data in the SwRI report. It thus represents
the conversion factor for locomotives manufactured in the mid-1990s. Older locomotives
would be expected to produce less useful work from each gallon of fuel (i.e., have a smaller
conversion factor), while future locomotives are likely to produce more work from each
gallon of fuel (i.e., have a larger conversion factor).
94
-------
calculated by subtracting the expected controlled inventory from the estimated 1999
baseline inventory. Locomotive emissions and their contribution to the national
inventories are shown in Appendix N. Based on the fuel consumption data shown in
Chapter 1, EPA is assuming no growth in locomotive emissions after 1996.
Fleet average emission rates were calculated as weighted averages of
uncontrolled, Tier 0, Tier 1, and Tier 2 emission rates. These emission rates were
weighted by estimated relative classwide fuel consumption rates (e.g., the percent of
total fuel consumed by Tier 1 locomotives in a given year). The relative fuel
consumption rates for each class were calculated assuming that they were proportional
to the product of number of locomotives(Nloc), average horsepower (HPavg), and a
relative use rate factor (FRU) based on average locomotive age, as shown below.
RelativeFuelComumption=_2£A—^A RUI
This part of the analysis was simplified for all locomotives other than Class I line-haul
freight locomotives (i.e., switch, Class II/III, and passenger locomotives) by neglecting
differences in average horsepower and relative use rates. This was done due to a lack
of specific information for these classes. Emissions factors were weighted by only
numbers of locomotives. This simplification does not significantly affect the overall
analysis because the differences in locomotive horsepower and usage rates for these
classes, as a function of the tier of applicable standards, are less significant than for
Class I freight locomotives. Moreover, these locomotives are relatively minor
contributors to the total national emission inventories.
The average baseline emission rates described in Chapter 4 were used for all
future uncontrolled emission rates. EPA estimated average in-use emission rates for
future controlled locomotives by subtracting a compliance margin from the level of each
of the standards. However, because the standards for HC, CO, and PM emissions from
Tier 0 and Tier 1 locomotives were set to prevent significant increases in these
pollutants, rather than to require reductions, EPA used the baseline emission rates in
the cases where the standards would not force emission reductions.
95
-------
TART.F fi-1
Estimated Emission Rates for Class I Locomotives (g/bhp-hr)
HC
CO
NOx
PM
Uncontrolled
TierO
Tier I
Tier II
Uncontrolled
TierO
Tier I
Tier II
Uncontrolled
TierO
Tier I
Tier II
Uncontrolled
TierO
Tier I
Tier II
Line-Haul
Locomotives
0.48
0.48
0.47
0.26
1.28
1.28
1.28
1.28
13.0
8.6
6.7
5.0
0.32
0.32
0.32
0.16
Switch
Locomotives
1.01
1.01
1.01
0.51
1.83
1.83
1.83
1.83
17.4
12.6
9.9
7.3
0.44
0.44
0.43
0.19
6.2 Class I Railroad Analysis
Assumptions
There are currently about 21,000 Class I locomotives being operated in the United
States.80 (AAR believes that this estimate is high; they estimate that there are less
than 20,000 locomotives actively being operated currently.) About 17,500 of these
locomotives were originally manufactured after 1972, and are thus subject to these
80 Official Locomotive Rosters and News, 1997 special edition-Class I railroads, James
W. Kerr, July 31, 1997.
96
-------
regulations. These locomotives are used primarily in line-haul service. Most of the
roughly 3,500 older locomotives are used as switchers. Based on these numbers, EPA
is projecting that, in 1999, Class I railroads will have about 15,200 post-1972
locomotives, and 1,300 older locomotives in line-haul service. EPA is also projecting
that they will have about 2,500 post-1972 locomotives, and 2,000 older locomotives in
switch service. It is assumed that by 2008 (i.e., within six years of the time that this
rule takes full effect), nearly all 1973 through 1999 line-haul locomotives will have
been remanufactured to meet EPA's standards for these locomotives and locomotive
engines.
More specifically, EPA is assuming that 13,200 of the post-1972 line-haul
locomotives will be brought into compliance with the Tier 0 standards by 2008, and
that 2,000 post-1972 locomotives and all 1,000 older line-haul locomotives will have
been removed from Class I line-haul service by 2010. EPA is also assuming that 3,000
Class I switch locomotives will be brought into compliance by 2017, including many
older locomotives that will be repowered or upgraded, and that the remainder of the
existing Class I switch fleet will have been removed from Class I service by 2024.
EPA is assuming that fuel consumption will remain constant at the 1996 level of
3.601 billion gallons per year.81 EPA recognizes that there is a short-term trend of
increasing fuel consumption, but is not confident that it will continue. The long-term
trend is for fuel consumption to remain fairly constant. This is the result of continual
improvements in locomotive fuel economy, which have offset the significant increase
in ton-miles of freight hauled. EPA is also assuming that 7.5 percent of fuel
consumption by Class I railroads is for switching.82
With respect to new production of freshly manufactured line-haul locomotives,
EPA is assuming 400 new units for years 2000-2004, 600 new units for years 2005-
2010, and 300 new units for all subsequent years. The higher number assumed for
years 2005-2010, is because the two largest western railroads are expected to purchase
large numbers of Tier 2 locomotives during this period in order to accelerate their
introduction into Southern California. After this period, sales are expected to slow
somewhat in terms of number of units due to the higher horsepower output of each
locomotive. EPA is also projecting that switcher sales will range from 50 to 100 per
year.
81 "Railroad Facts", Association of American Railroads, 1997 edition.
82 "Railroad Ten Year Trends," Association of American Railroads, 1995.
97
-------
Results
The results of the Class I emission analysis are summarized in Tables 6-2 (line-
haul) and 6-3 (switch). More detailed results are shown in Appendix I. These results
project a 44 percent reduction in NOx emissions from line-haul locomotives by 2010.
Figure 6-1 shows those NOx emission reductions expected in the first ten years from
each tier of locomotive. As can be seen in the figure, the majority of emission
reductions during the first ten years of the program come from Tier 0 locomotives.
Emissions from switch locomotives are not expected to decrease as rapidly as emissions
from line-haul locomotives because of a slower fleet turnover rate to Tier 2 technology.
(Although not shown in the figure, emission reductions from Class I switch locomotives
are distributed similarly with respect to locomotive tier.) Significant reductions in HC
and PM emissions are not expected until Tier 2 locomotives are phased into the fleet
in large numbers.
Table 6-2
Results of Class I Line-Haul Emission Analysis83
Year
1999
2000
2005
2010
2015
2020
2025
2030
2035
2040
Emission
(Metric Tons/Year)
HC
33,300
33,300
31,900
27,300
25,200
23,300
21,500
19,900
18,600
17,900
NOx
901,000
889,000
614,000
485,000
453,000
423,000
395,000
370,000
350,000
345,000
PM
22,200
22,200
21,300
18,000
16,500
15,100
13,800
12,700
11,700
11,300
Reductions
(Metric Tons/Year)
HC
0
0
1,400
5,900
8,000
10,000
11,800
13,400
14,700
15,300
NOx
0
11,000
287,000
415,000
448,000
478,000
506,000
531,000
551,000
556,000
PM
0
0
900
4,100
5,600
7,000
8,300
9,500
10,400
10,900
Percent Reduction
HC
0%
0%
4%
18%
24%
30%
35%
40%
44%
46%
NOx
0%
1%
32%
46%
50%
53%
56%
59%
61%
62%
PM
0%
0%
4%
19%
25%
32%
38%
43%
47%
49%
83 The results shown in this table are the same as those found in the December 1997
version of this document. Results of corrected analyses are contained in Appendix O.
98
-------
Table 6-3
Results of Class I Switch Emission Analysis
Year
1999
2000
2005
2010
2015
2020
2025
2030
2035
2040
Emission
(Metric Tons/Year)
HC
5,670
5,670
5,640
5,470
5,260
5,030
476
4,440
4,040
3,580
NOx
98,700
98,700
90,900
81,400
71,200
65,200
60,500
57,200
52,800
47,900
PM
2,470
2,470
2,460
2,370
2,270
2,150
2,020
1,860
1,660
1,430
Reductions
(Metric Tons/Year)
HC
0
0
30
210
410
650
920
1,230
1,640
2,100
NOx
0
0
6,800
16,400
26,500
32,500
37,200
40,500
44,900
49,800
PM
0
0
20
100
210
320
460
610
810
1,040
Percent Reduction
HC
0%
0%
1%
4%
7%
11%
16%
22%
29%
37%
NOx
0%
0%
7%
17%
27%
33%
38%
41%
46%
51%
PM
0%
0%
1%
4%
8%
13%
18%
25%
33%
42%
Figure 6-1
Class I Line-Haul NOx Reductions by Tier of Locomotive
500
o 400
o -
3 1
Q) §
OJ
300
200
100
0
Tier 0
Tier 1
Tier 2
2000
2005
2010
99
-------
6.3 Class II and III Railroad Analysis
Assumptions
Information provided to EPA by the American Short Line Railroad Association,
which represents most of the Class II and Class III railroads, shows that there were
approximately 4,200 locomotives in service with Class II and III railroads in 1994, and
that they consumed about 215 million gallons of diesel that year.84 However, only
about 15 percent of these locomotives were originally manufactured after 1972. Based
on these numbers, EPA is projecting that there will be about 600 post-1972 locomotives
and 3600 older locomotives in the 1999 Class II and III fleet. Due to a lack of specific
information, average Class II and III emission rates are assumed to be the same as the
average emission rates for Class I line-haul locomotives. It is possible that actual
emission rates could be somewhat higher since smaller railroads typically have lower
power duty-cycles (i.e., more time at idle and low power notches, and less at notch 8),
especially those railroads performing primarily switch and terminal services.
EPA is assuming that, during the first 10 years of the program (assuming that the
remanufacture cycle for small railroads is about 10 years), Class II and III railroads
will bring about 50 locomotives per year into compliance with the Tier 0 standards in
order to have them covered as "new locomotives" by the CAA preemption provisions.
EPA is also assuming that in 2012 these railroads will begin to purchase about 150
complying Tier 0 locomotives per year from Class I railroads. It is unlikely that they
would be able to purchase any complying locomotives prior to this time. In cases where
a Class I railroad makes the investment to bring an existing locomotive into
compliance, it will almost certainly retain it for a period equivalent to two full useful
life periods. Thus, they are not likely to sell any locomotives that they have brought
into compliance until at least 12 years after they were originally brought into
compliance.
Results
The results of the Class II and III emission analysis are summarized in Table 6-4.
More detailed results are shown in Appendix I. These results show that NOx
emissions from small railroads represent only about 6 percent of all locomotive
emissions. These emissions will decrease slowly because of a fairly slow fleet turnover
rate to Tier 0 technology. The analysis projects no significant reductions in HC and
PM emissions because Tier 2 locomotives are not expected to be phased into Class II
and III fleets in large numbers in the foreseeable future.
84 "Locomotive Data for Small Railroads," Memorandum, Charles Moulis, U.S. EPA, to
public docket A-94-31, December 5, 1997.
100
-------
Table 6-4
Results of Class II and III Emission Analysis
Year
1999
2000
2005
2010
2015
2020
2025
2030
2035
2040
Emission
(Metric Tons/Year)
HC
2,150
2,150
2,150
2,150
2,150
2,150
2,150
2,150
2,150
2,150
NOx
58,100
58,100
57,200
56,000
52,900
49,400
45,800
42,300
38,700
38,200
PM
1,430
1,430
1,430
1,430
1,430
1,430
1,430
1,430
1,430
1,430
Reductions
(Metric Tons/Year)
HC
0
0
0
0
0
0
0
0
0
0
NOx
0
0
900
2,100
5,200
8,800
12,300
15,900
19,400
19,900
PM
0
0
0
0
0
0
0
0
0
0
Percent Reduction
HC
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
NOx
0%
0%
2%
4%
9%
15%
21%
27%
33%
34%
PM
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
6.4 Passenger Railroad Analysis
Assumptions
According to the American Public Transit Association (APTA) there were
approximately 463 diesel locomotives in commuter rail service in 1995, with 397 of
these locomotives originally manufactured after 1972. These 463 locomotives
consumed about 61 million gallons of diesel fuel that year.85 In addition, Amtrak
currently has 315 diesel locomotives in service, consuming about 72 million gallons of
diesel fuel per year. EPA is projecting that 100 locomotives will be brought into
compliance during each of the first five years of the program, and that all uncontrolled
locomotives will be removed from passenger service by 2011. Sales of freshly
manufactured passenger locomotives are assumed to be 30 new units per year.
Average passenger locomotive emission rates are assumed to be the same as the
average emission rates for Class I line-haul locomotives.
85 "1996 Transit Vehicle Fact Book", American Public Transit Association.
101
-------
Results
The results of the passenger locomotive emission analysis are summarized in
Table 6-5. More detailed results are shown in Appendix I. These results show that a
39 percent reduction in NOx emissions is expected by 2010. Significant reductions in
HC and PM emissions are not expected until Tier 2 locomotives are phased into the
fleet in large numbers. The rate of phase-in is expected to be roughly similar to the
rate for Class I freight service.
Table 6-5
Results of Passenger Emission Analysis88
Year
1999
2000
2005
2010
2015
2020
2025
2030
2035
2040
Emission
(Metric Tons/Year)
HC
1,330
1,330
1,300
1,180
1,060
940
820
710
710
710
NOx
36,000
36,000
26,700
20,900
18,800
16,900
15,000
13,700
13,700
13,700
PM
890
890
870
780
700
610
530
440
440
440
Reductions
(Metric Tons/Year)
HC
0
0
30
150
270
390
510
620
620
620
NOx
0
0
9,200
15,100
17,100
19,100
21,000
22,300
22,300
22,300
PM
0
0
20
100
190
270
360
440
440
440
Percent Reduction
HC
0%
0%
2%
11%
20%
29%
38%
47%
47%
47%
NOx
0%
0%
26%
42%
48%
53%
58%
62%
62%
62%
PM
0%
0%
2%
12%
21%
31%
40%
50%
50%
50%
6.5 Summary of Environmental Benefits
Estimated emissions and emission reductions for the total U.S. locomotive fleet
are shown in Table 6-6 and Figure 6-2. The totals are dominated by the emissions and
reductions from Class I freight locomotives, which represent about 90 percent of all
current emissions from locomotives. The marginal benefits of each tier of standards,
which are used in Chapter 7 to calculate marginal cost-effectiveness, are shown in
Appendix I.
86 The results shown in this table are the same as those found in the December 1997
version of this document. Results of corrected analyses are contained in Appendix O.
102
-------
Figure 6-2 Projected Emissions
For All Freight and Passenger Locomotives
(0
0)
c
o
50
40
30
¥ s 20
10
1200
1000
800
ro
0)
g
•g 'E
• u
600
-------
Table 6-6
Results of Emission Analysis for All Locomotives
Year
1999
2000
2005
2010
2015
2020
2025
2030
2035
2040
Emission
(Metric Tons/Year)
HC
42,400
42,400
41,000
36,100
33,700
31,400
29,200
27,100
25,400
24,300
NOx
1,093,000
1,081,000
789,000
644,000
596,000
554,000
516,000
483,000
456,000
445,000
PM
27,000
27,000
26,000
22,600
20,900
19,300
17,800
16,400
15,300
14,600
Total Reductions (2000-2040)
Reductions
(Metric Tons/Year)
HC
0
0
1,400
6,300
8,700
11,000
13,200
15,300
17,000
18,100
417,000
NOx
0
11,000
304,000
449,000
496,000
538,000
576,000
610,000
637,000
648,000
20,100,000
PM
0
0
900
4,400
6,000
7,600
9,100
10,500
11,700
12,400
288,000
Percent Reduction
HC
0%
0%
3%
15%
21%
26%
31%
36%
40%
43%
NOx
0%
1%
28%
41%
45%
49%
53%
56%
58%
59%
PM
0%
0%
3%
16%
22%
28%
34%
39%
43%
46%
104
-------
7.0 Costs and Cost-effectiveness
Background
EPA estimated the costs of compliance with the proposed standards in the Draft
RSD, based in part on materials supplied by locomotive manufacturers and the
railroad industry. After the proposal, EPA contracted with ICE, Incorporated, with its
subcontractors, Acurex Environmental Corporation and Engine, Fuel, and Emissions
Engineering, Incorporated (EF&EE), to update the economic analysis, based on the
estimated costs for the most likely compliance technologies.88 In most instances, the
results of the cost study tended to confirm the EPA estimates, including the cost-
effectiveness estimates. The current estimated compliance costs are based largely on
the contractor studies, although some differences exist as a result of public comments
received or other information available to EPA. Such differences from the Draft RSD
and the contractor cost study will be noted where present. As a result of such
differences, EPA is presenting a range of costs in several areas. In general, the cost
estimates presented here tend to be somewhat conservative; that is, for those costs
with significant uncertainty, EPA used the higher end of the estimated range.
Incremental compliance costs are presented for Tier 0, Tier 1 and Tier 2
locomotives on a total and per-locomotive basis. This incremental approach is
appropriate because EPA believes that the technology will be applied sequentially, as
the standards become more stringent. Locomotive cost components consist of initial
equipment costs, and operating costs. The equipment costs consist of fixed costs and
variable costs for the necessary hardware, which, when adjusted to include the
manufacturer's markup for overhead and profit, comprise the initial cost increase to
the operator. Fixed costs include engineering costs for compliance technology
development; testing costs for development, certification, production line testing (PLT)
and in-use testing; tooling costs to enable production of the hardware for compliance;
and technical support costs for training and technical support publications to be used
by personnel operating and maintaining the locomotives. Operating costs include
incremental fuel costs, (incremental to those that would be incurred without the
compliance technology) and incremental maintenance costs, including re manufacturing
costs.
A breakdown of the testing costs used in calculating fixed costs is presented in
Table 7-1, while the total fixed costs, including testing costs, are shown in the per-
locomotive costs calculated in Tables 7-2A (base case) and 7-2B (high range). This
latter table is included as a sensitivity analysis to show the effects of modifying base
case assumptions regarding development costs, testing costs and operating costs (which
88 "Cost Estimates for Meeting the Proposed Locomotive Emission Standards", Engine,
Fuel, and Emission Engineering, Inc., September, 1997; and "Locomotive Technologies to
Meet SOP (sic) Emission Standards", Acurex Environmental Corp., August, 1997.
105
-------
are detailed below). Tables 7-2 A and 7-2B also include the estimated locomotive
populations used in allocating the fixed costs. The model categories listed in these
tables, (e.g., "Tier 0 - A" or "Tier 2 - B") represent different locomotive model types,
consistent with the descriptions of Chapters 3 and 4. Where applicable, costs are
presented in actual and discounted format.89
7.1 Initial Cost Increase for Locomotives
The initial cost increase for locomotives (sometimes known as the first price
increase (FPI)) consists of the fixed costs to the manufacturer for such things as
investment capital, research and development, etc. and the variable costs, which are
the costs for the necessary emission control hardware on each individual locomotive.
Variable costs vary with production, while fixed costs are allocated to whatever
production level is realized.
7.1.1 Fixed Costs
Fixed costs are those that represent the initial investments that must be made by
the manufacturer before the beginning of production. These will vary in magnitude
with the emission standard (Tier 0, Tier 1, Tier 2) for which compliance is required,
and with the amount of emission control required by individual manufacturers for their
product lines, but will tend to fall into the same general categories (e.g., engineering
costs or technical support costs). It is important to note that the fixed costs necessary
to develop Tier 0 certification kits apply independently to each remanufacturer
developing a kit. Thus, if three different remanufacturers independently develop a kit
for the same model of locomotive, then the fixed cost component for that locomotive
model would be three times what it be if only one remanufacturer developed a kit. The
numbers of suppliers assumed for each model category of Tier 0 locomotive in this
analysis are shown in Tables 7-2A and 7-2B. These numbers are EPA's projection that
are based on the current numbers of independent part suppliers and remanufacturers
for the various locomotive models. EPA assumed that many current suppliers and
remanufacturers will seek to avoid the need to certify by making business
arrangements with other businesses that are certifying. Thus, the number of suppliers
assumed for each kit is less than the total number of suppliers.
Because the fixed costs are for goods and services that are useful for more than
one year of production, it is appropriate to amortize or allocate these costs over more
than one year of production. In its rulemakings, EPA normally assumes that the
manufacturers would recover their development costs within the first five years of
production. For Tier 2, this becomes important because the standards are effective for
a considerable length of time. The Tier 2 development costs were therefore assumed
to be recovered by 2010, and a separate reduced fixed cost component was used in the
89 Discounted costs represent the net present value of costs.
106
-------
lifetime benefit calculation after that year. After 2010 fixed costs were limited to
testing costs for PLT and in-use testing. For Tier 1, the effective period for allocation
of development work is shorter, more on the order of three years. The period of
compliance for Tier 0 is longer, due to the long useful life and staggered remanufacture
schedule of most Tier 0 locomotives, so a period of five years can be used. It is not
necessary to calculate separate compliance costs reflecting fully-recovered fixed costs
for Tiers 0 and 1 since the initial hardware costs occur only at original manufacture
(for Tier 1) or the first remanufacture (for Tier 0), and thus are applicable only during
the first few years of the program. This does not materially affect the cost-effectiveness
of the standards, and any minor error involved would be on the side of conservatism
in cost determination. Fixed costs are normally also allocated on a per-locomotive
basis according to the number of units produced. Such costs include developmental
engineering and testing costs (the latter including facility, equipment and operating
costs), as well as tooling and technical support costs as described above.
TABLE 7-1
TESTING COSTS
Capital
Costs3
Test Systems
Auxiliary Equipment
Equipment Subtotal
Facility Costs
TOTAL CAPITAL
Number
6
4
4
Unit Cost
$340,000
$185,000
$2,000,000
Total Cost
$2,040,000
$740,000
$2,780,000
$8,000,000
$10,780,000
Annual Cost
$395,809
$1,139,020
$1,534,829
Operating
Costs
Technicians
Engineers
Maintenance
Subtotal
Overhead
TOTAL OPERATING
12
4
6
$70,000
$100,000
$30,000
$840,000
$400,000
$180,000
$1,420,000
$213,000
$1,633,000
TOTAL COSTS
$3.167.829
Number of Tests/Year
Allocated Cost/Test
Consumables
Lost Service11
Cost/Test
2000-2010
200
$15,839
$5,000
$300
$21.139
After 20 10a
50
$32,660
$5,000
$1,200
$38.860
a Capital costs assumed to be recovered by 2010.
b Average cost for lost service time varies because of different mix of development, certification,
production line and in-use testing.
107
-------
a) Q
H
Q E
Si
108
-------
P
M
S <
P
I CO
P
5 <
P
oJ o
P
S o
P
I CD
P
P
5 LU
P
I o
P
S o
P
I CO
P
S <
P
HIGH CASE
Standards
Model Cateqory
n
g
1
g
S
1
g
g
m
3
1
1
g
INUMBER OF LOCOMOTIVES I
XIII II
g g g g S
« ^J €B ^ ^
XIII II
XIII II
1 1 1 1 1
XIII II
1 pss
X III i1 1
1 IP1
X III II
1 II 1 i 1
X III
g 1 111 £
X III
X III III
1 1 1 1
X 1 II III
X II 1 1 1 1 1
S S g
XX 1 1 1 1 1 1 1
S
X X 1 1 1 1 1 1 1 1 1
•til!, f
S ! 5 s £' 11
1
S
S
1
S
S
**
S
£
S
£
S
cn
eo
g
eo
«*
ITOTAL VARIABLE COSTS
, i , ,
^ S
^;
"i
' 1 i 1 1
s 1 1 1 §
"5"
S 1 1 1 1
^ S S **
1 1 5 1 1
LT)' S 2 ^
S § 51 § 0
J S ° »
g g -T- g g
WWW
S S '
= 11 1
' 1 1 1
s g ?: g
J3 « | £
11,1
FIXED COSTS:
Number of tests
Engineering costs
Testing costs
Tooling
Technical Support
1
^
^
£
cH
g
I
5:
"
^
S
^
^
cH
^
5
S
E
1
s
S
1
1
O
TS
CO
Lg LT, ^
5 S
**
CO LD T—
CO Lfl
a *i
1
eo in
i "
ir
R **
»
; ? ""
a w
S
" 5
S 5
S
s.|. ^
S
CO i—
cn
Lrt
cn en
cn
in eo T
cn eo'
eo
S
co in ^
" **
»
r- i—
r- r-
S
TOTAL FIXED COSTS
Cost per Locomotive
Number ot suppliers
cn
cn
cri
*»
en
en
S
S
CO
eo1
CO
CO
CO
**
S
cn
S
eo
1
s
Co
^
«*
s
TOTAL MFG COSTS:
(Includes 20% Mir. markup)
1 ? z
S X g g
5 ^ s|
I'll
I'll
I ~ 1 1
§ S § 1
1 ~ 1 1
lel|
g *r CO r^
1 S 1 1
§ S | |
|£i|
IT 53
OPERATING COSTS:
FUEL COST
Average Fuel Consumption
FE Penalty
Gallons offuel^yeor
Cost peryear® (0 7D/GO.I)
=. »
= go
3 **
**
= S'
I l?
**
cn
ta
"
5 = ^
ID **
" § ?
**
§ ^
I P
", S
CO
I 1"
*i
g
TOTAL FUEL COST
INCREMENTAL WAINT
Cost peryear
Service life
cri
=
en1
to
cri
1
eri
**
1
5
|
S
J
s
1
**
ITOTAL MAINT. COST
i 1
O CD
2 S
T S*
" ™L
™ :S
*»
1 1
#i
g s
ey **
e3 ^
§ **
i i
**
? K
1 i
CO S
o ^
CO ^
** ***
to ^
TOTAL OPERATING COSTS:
PRESENT VALUE
£
g
£
i
S
cn i^
= ^
e^j 5-
£ 1
i ^
1 1
**
CD ^
1 s
rn S
en' ^'
*» ***
1 §
TOTAL COST TO RR
NPV COST TO RR
109
-------
Testing Costs
Testing costs include developmental testing, as well as certification testing,
production line testing and in-use testing. Developmental testing, which represents
the largest segment of the projected testing costs, is necessary for developing and
evaluating the necessary compliance strategies. The amount of developmental testing
that is necessary is determined by each individual manufacturer or remanufacturer.
Certification and PLT provide evidence of the manufacturer's initial compliance with
the applicable emission standards, and in-use testing provides evidence of the
continued durability of emission control systems. The amount of certification, PLT and
in-use testing are specified by EPA.
Testing costs include the costs of any necessary additional facilities and equipment
for emissions testing, plus engineering, operating and maintenance costs for the testing
facility. A detailed breakdown of these costs may be found in the EF&EE contractor
study, and they are summarized in Table 7-1. The contractor assumed that the major
manufacturers and a limited number of aftermarket remanufacturers would need test
equipment sufficient to equip six test cells, at a cost of $340,000 per cell, plus auxiliary
facility equipment, such as calibration gas handling equipment, particulate weighing
equipment, and a spur track for bringing locomotives to the test site, at an estimated
cost of $ 185,000 per facility. EF&EE assumed one such auxiliary system could service
two test cells, and that four such systems would be required, since the two major
manufacturers would already have one test site and would develop one more each. The
contractor allocated these costs over a ten year period.
The contractor assumed that although the two major manufacturers have existing
test cells, each would require an additional test cell for Tier 0 testing at an estimated
cost of $2 million per cell, but included these costs in the Tier 0 costs. EPA believes
that these facilities will be used for Tier 1 and Tier 2 testing as well as for Tier 0.
Therefore, the Agency has included the costs for two additional test cells for the major
manufacturers and two for the independent remanufacturers in the calculation of the
testing costs applicable for all tiers. EPA also received a number of comments
regarding the lack of estimated test facility costs in the Draft RSD. The comments
contained estimates ranging from $ 1 to 3 million. EPA believes that the cost study
estimate of $2 million each for additional test facilities is reasonable and represents
the mid-point of the range. Allocated facility costs are thus included along with
allocated equipment costs in the cost per test calculation.
In addition to allocated equipment and facility costs, testing costs also include an
operating cost component. In the cost study, EF&EE assumed that two technicians
and a supervisory engineer would be required for each test cell (or group of two test
cells in the case of the major manufacturers). The contractor also assumed costs for
test system calibration and maintenance and for overhead at a rate of 15 percent of
direct costs. EPA is assuming $5,000 per test for consumables (test fuel, reference
110
-------
gases, electrical power, etc.), which is intended to be conservative. Since a fraction of
the tests will be for in-use testing, EPA also assumed a cost for removing the
locomotive from service for testing. Although the Agency received comments placing
this cost at up to 17 days lost time at a cost of $ 14,000, EPA will use the AAR estimate
of 8 days lost time at a cost of $6,000, due to the considerable AAR member experience
with ERA 92-day safety inspections. EPA estimates that in-use testing will represent
about five percent of the total test load during the early years of the program when a
great deal of development testing will be conducted, but will increase to approximately
20 percent of the total test load after 2010, when the developmental testing should be
completed and the total testing load diminishes. These figures are also reflected in
Table 7-1.
As can be seen from Table 7-1, these costs when allocated over the estimated
testing requirement, amount to about $21,000 per test prior to 2010 and about $39,000
per test after 2010 when the developmental testing is completed. This differs from
both EPA's original estimate of $10,000 per test and the estimate of $10,237 per test
provided in the EF&EE cost study. However, as noted above, EPA has added allocated
facility costs and allowances for consumables and lost locomotive service to the
estimated costs, and has adjusted the cost-per-test calculation accordingly.
EPA also received a number of testing cost estimates in the public comments; the
near-term estimates for the type of testing that will most likely be required tended to
fall into the $20,000 to $30,000 range. The above-mentioned adjusted cost per test falls
into the lower end of that range, so as a sensitivity analysis EPA assumed a high range
cost per test of 1.5 times both of the costs listed in Table 7-1; the resulting testing costs
will be used in calculating the high range cost and cost/benefit figures that are
presented in Table 7-2B.
Engineering Costs
The engineering costs category represents the estimated average cost for the
number of engineering work years projected to be required to develop the calibrations
and hardware necessary for meeting the emission standards. This also includes the
effort for any ancillary changes that must be made to the locomotives to accommodate
the required new hardware. The engineering costs used in the current analysis were
taken from the EF&EE cost study and are shown in Tables 7-2A (base case) and 7-2B
(high range). In Table 7-2B the engineering costs were increased by a factor of 1.5 as
a part of the sensitivity analysis.
Tooling Costs
Tooling costs are also shown in the fixed costs portion of Tables 7-2 A and 7-2B.
These include costs for any additional or modified tooling necessary to produce the
emission control hardware, as well as for any required setup changes. Because Tier
111
-------
0 compliance is estimated to be achieved through calibration changes or through
hardware obtained from suppliers (particularly in the case of aftermarket
remanufacturers), no specific tooling costs were estimated for Tier 0. Tooling costs for
Tier 0 parts were assumed to be included in the component costs to the
remanufacturer. Tooling costs for the other tiers were taken from the cost study.
Technical Support
Technical support consists of any changes that would be required in the technical
support that manufacturers provide to users. This would include any necessary
operator or maintenance training and changes to technical publications that provide
operating and maintenance guidance. These costs were not included in the Draft RSD,
but were estimated in the EF&EE cost study, and are included in the current analysis.
These costs are also included in the fixed costs portion of Tables 7-2A and 7-2B.
7.1.2. Variable Costs
Hardware
Hardware requirements for meeting the applicable emission standards will vary
with the stringency of the standards, i.e., Tier 0, Tier 1, or Tier 2, and by
manufacturer, according to the hardware changes required to meet the standards.
These variations are outlined below and are discussed more fully in the Acurex
technology study and the EF&EE cost study. In general, however, manufacturers will
need to use some combination of fuel injection timing calibration; fuel injector,
turbocharger, and/or charge air cooling improvements; and possibly combustion
chamber or piston modifications. Particularly for Tier 0, different combinations of
these strategies will be used for different locomotives. A listing of these strategies,
along with their projected costs and estimated usage, is presented in Table 7-3. For
example, it is projected that about 50 percent of Tier 0 locomotives will have timing
retarded by 2 degrees, and the other 50 percent will have timing retarded by 4 degrees.
Similarly, it is projected that 60 percent of Tier 0 locomotives will add 4-pass
aftercoolers. Representative model combinations are shown in Tables 7-2A and 7-2B.
No separate model cases are shown for passenger locomotives, since these are generally
variations of similar line-haul locomotives and should thus incur similar costs (fixed
as well as variable). However passenger locomotives are included in the total cost
calculations since the costs should be similar to those for freight locomotives.
112
-------
Assembly Costs
Assembly costs include the labor and overhead costs for retrofitting (in the case
of Tier 0) or for initial installation of the new or improved hardware. These will also
vary with the characteristics of individual locomotives and the type of hardware
necessary for compliance with the applicable emission standards. These costs are
included in the summary figures in Tables 7-2A and 7-2B, and are discussed more fully
in the contractor cost study.
Table 7-3
Incremental Hardware Cost And Usage
Expected
Technology
2 deg timing retard
4 deg timing retard
4 pass aftercooler
Improved mechanical
injectors
Add electronic fuel injection
Improved electronic injectors
Engine Modifications
Improved turbocharger
Split cooling
High pressure injection
Combustion chamber design
Cost per
Locomotive
—
—
$5,000
$800
$35,000
$2,000
$800
$25,000
$25,000
$2,000
$800
Percent of Locomotives Using
Technology
TierO
50%
50%
60%
30%
13%
37%
20%
30%
Tier 1
100%
100%
50%
25%
75%
100%
100%
Tier 2
100%
100%
100%
100%
100%
7.1.3. Total Locomotive Cost Increase
The fixed and variable costs, together with a manufacturer markup for overhead
and profit, comprise the total manufacturing costs which represent the initial cost
increase, or FPI, to the operator. A conservative manufacturer markup of 20 percent
was assumed, based on the amount normally used as a target figure in the automotive
industry.90 Although the actual markup for automobiles and trucks tends to vary with
economic conditions and other factors, this target figure was used for locomotives for
the sake of conservatism and in the absence of any data on the actual markups
90 "Cost Estimations for Emission Control Related Component/ System and Cost
Methodology Description", LeRoy H. Lindgren, March, 1978 (EPA-460/3-78-002).
113
-------
achieved in the locomotive industry. Given the competitive nature of the industry,
EPA believes the actual markups are likely to be lower. The total manufacturing costs
or FPI are shown in Tables 7-2A and 7-2B.
Tier 0 Locomotives
In calculating the estimated FPI for Tier 0 locomotives, EPA assumed that the
locomotives could be grouped into 5 categories (or engine families): older and newer
line-haul locomotives from the two major manufacturers, and switch locomotives from
one manufacturer. This latter group was not included in the EF&EE cost study
because the contractor did not feel it would be cost-effective to upgrade these
locomotives. However, EPA believes a significant number will be certified, due to the
relatively low cost involved. The EF&EE cost study estimated that it may be more
cost-effective to improve the existing unit injectors of older locomotives, since they
typically do not accumulate as many miles each year as the newer locomotives.
However, based on comments received, EPA does not necessarily concur that this will
be true for all such locomotives, and has included electronic injection systems for some
of the older locomotives as well. The other major difference in costs results from the
need for an improved turbocharger for locomotives with higher smoke emissions, which
would not be required for locomotives which did not have this problem. Changes in
injection timing calibrations would likely be used for all these scenarios.
Tier 1 Locomotives
The estimated initial cost increase for Tier 1 locomotives are shown in Tables 7-2A
and 7-2B. The EF&EE cost study made the distinction between mechanical and
electronic injection designs for Tier 1 locomotives. Fuel injection cost differences would
thus arise from the fact that some current locomotives are already equipped with
electronic injector systems, which would merely need to be upgraded, rather than
replaced. EPA believes that all Tier 1 locomotives will have injection systems that are
upgraded versions of the current electronic injectors. Therefore, the costs used in all
calculations are based on the upgraded electronic injector scenario, rather than
conversion from mechanical to electronic injection. Improved charge air cooling and
limited in-cylinder modifications, e.g., for reduced oil consumption or higher
compression ratio, as well as changes in injection timing calibrations, were also
estimated to be required. EPA also believes that early versions of the new engine
designs that will be used to meet the Tier 2 standards will make their appearance
during the Tier 1 period. Thus, the tables show two Tier 1 models for each
manufacturer.
114
-------
Tier 2 Locomotives
The estimated initial cost increases for Tier 2 locomotives are shown in Tables 7-
2A and 7-2B. Costs for Tier 2 compliance are based on new and improved engine
designs already under development by the two major manufacturers. The existence
of these designs, plus the 7 years of lead time available for development should result
in cleaner engines that are able to meet the Tier 2 standards with minimum
incremental costs over the developmental costs that would otherwise have been
incurred for development of the uncontrolled engines. Again, the basic strategies likely
to be used are injection timing calibrations, low-temperature charge air coolers and
fuel injection improvements, largely the addition of rate shaping capability in the
latter case. The Acurex report projected a possible need for EGR for some Tier 2
locomotives, however, EF&EE and EPA did not include EGR in the projected
compliance strategies. EPA believes that if EGR were to be used at all, it would only
be used on the lower power notches. It is also important to note that since the
manufacturers are expected to introduce early versions of these new models during the
Tier 1 time period, some of the emission control development cost is assigned to the
Tier 1 locomotives. Compliance with the Tier 2 standards will actually be a
continuation of the Tier 1 development process.
7.2. Incremental Operating Cost Increases
The incremental operating costs include any increase or decrease in the amount
of fuel consumed as a result of the new standards, plus any incremental maintenance
costs. This latter category would include maintenance on any new hardware required
or additional maintenance on existing hardware. Any incremental subsequent
remanufacture costs would also be included in the total. These costs are shown in the
operating cost sections of Tables 7-2A and 7-2B.
7.2.1. Incremental Fuel Cost
EPA estimated the fuel economy ramifications of various emission control
technologies in the Draft RSD. Public comments on the subject were also received, and
incremental fuel costs (or savings) were also presented for the strategies considered in
the contractor cost study. These are expressed in terms of percent increases or
decreases from current levels. However, estimates for current fuel consumption levels
show considerable variation.
In the Draft RSD, EPA calculated an annual per-locomotive cost of a one
percent decrease/increase in fuel economy by multiplying the cost of fuel used by
Class I railroads in 1992, $1,913,000,000, by a one percent fuel economy
penalty/benefit and divided by the number of locomotives in use during 1992,
rounded to the nearest thousand (i.e. approximately 18,000 locomotives). In terms
of gallons of fuel consumed (at $0.63 per gallon), this represents roughly 167,900
115
-------
gallons. Total fuel economy penalties/benefits of one percent from pre-regulated
baseline levels would thus result in a calculated annual per-locomotive cost or
benefit of $1,062.
On the other hand, in the contractor cost study, EF&EE assumed total per
locomotive fuel consumption rates of 104,000 and 297,000 gallons per year, depending
on application (i.e., light or heavy usage). EF&EE assumed a fuel consumption rate
of 445,000 gallons per year for the advanced technology engines projected to be used
for Tier 2 compliance. However, in their comments on the NPRM, EMD projected a
fuel consumption rate of 350,000 gallons per year for their new 6,000hp locomotives.
Since the OEM has developmental data for their products and should thus be in a
better position to make fuel consumption estimates for their new technology, EPA will
use this estimate for Tier 2, along with the Tier 0 and Tier 1 estimates from the cost
study. The fuel consumption rates assumed are shown in the operating costs section
of Tables 7-2A and 7-2B. These fuel rates, however, are only used for these per
locomotive calculations; total costs were calculated from total fuel consumption
estimates provided by the railroads.
As stated above, EPA received comments concerning projected fuel economy
penalties. Some of these comments projected penalties as high as 5-10 percent,
however this latter estimate appears to assume use of EGR on all models for meeting
the Tier 2 standards. As stated above, EPA does not believe this will be required, and
has projected fuel economy penalties of two percent as its best estimate of the Tier 2
fuel economy effects of the current rulemaking. Based on past developments in the
industry, EPA further believes that manufacturers will make every effort to eliminate
any initial fuel consumption penalties, and will have largely succeeded by 2010.
However, projecting the necessary developmental costs involved and resulting fuel
consumption levels is difficult at this time. Thus, for the sake of conservatism, the two
percent projected fuel economy penalty will be retained in the analysis for the full 41
years covered by the analysis.
EPA also projected fuel economy penalties of 1-2 percent for Tier 0 locomotives,
depending on other modifications that could serve to mitigate any fuel economy
decrease involved. However, some Tier 0 locomotive models converting from
mechanical to electronic fuel injection may actually see a decrease in fuel consumption.
For Tier 1, EPA has assumed a one percent penalty, since the additional lead time
afforded for meeting Tier 1 standards should allow additional development for the
manufacturers to address any more negative fuel consumption effects. Nevertheless,
as a sensitivity analysis, the Agency has also included a scenario which doubles the
base-case incremental fuel consumption estimates. These are in the high range cost
estimates presented in Table 7-2B and the high-range cost-effectiveness analysis
shown in Table 7-5.
116
-------
7.2.2. Incremental Maintenance
Routine Maintenance
EPA did not estimate any incremental maintenance costs in the Draft RSD.
However, comments submitted by EMD stated that additional cost would be incurred
for periodic replacement of electronic fuel injectors (which are routinely replaced at
periodic intervals between remanufactures), and electronic injection wiring harnesses
(which would also require replacement outside the normal remanufacture cycle due to
embrittlement of the insulation from the heat generated by the engine). EF&EE
agreed in its cost study that these would be legitimate costs, and provided estimated
incremental costs for their replacement. EPA is including the cost of these
replacements in the estimated maintenance costs for Tier 0 and Tier 1 locomotives, and
the cost of improved injector replacement for Tier 2 locomotives. In addition, the
contractor projected a small additional cost for routine periodic replacement of
improved unit injectors, based on the difference in cost between the standard and
improved versions. EPA has included the incremental costs for Tier 0 mechanical
injector replacement in the maintenance costs. All of these periodic costs are converted
to an average annual component and are included in the incremental maintenance
costs shown in the operating costs section of Tables 7-2A and 7-2B. For purposes of
this analysis, Tier 0 locomotives were assumed to have an average remaining service
life at the time of remanufacture of 15 and 21 years, respectively, for older and newer
locomotives, as shown in Tables 7-2A and 7-2B.
Subsequent Remanufactures
In addition to the initial equipment costs at time of original manufacture, and
incremental maintenance costs for components that were upgraded at the time of
manufacture/initial remanufacture, there will be some increase in costs at the time of
each subsequent remanufacture of the locomotives. These costs are the incremental
price increases in equipment that is routinely replaced at time of remanufacture.
These costs are increased due to the improved parts necessary to meet the applicable
emission standards. These costs are also included in the incremental maintenance
costs presented in Table 7-3, and are discussed in greater detail in the EF&EE cost
study.
7.2.3. Total Cost Increase
The estimated increased fuel costs and maintenance costs are included in the
incremental operating costs for Tier 0, Tier 1 and Tier 2 locomotives shown in Tables
7-2A and 7-2B. These operating costs are presented as annual incremental costs and
lifetime incremental costs for the estimated service life of the locomotive, in both actual
and net present value (NPV) form. These operating costs are added to the first price
increase to yield the lifetime costs per locomotive presented in the table. As discussed
117
-------
above, EPA has based its analysis on public comments received and on engineering
judgement, as well as on the contractor study. Again, since there was considerable
variation in the estimated fuel consumption figures contained in the comments and
other sources, the incremental fuel consumption estimates from the base-case analysis
were doubled to form a high-range estimate, which is used as a sensitivity analysis.
7.3. Total Program Costs and Cost-effectiveness
Tables 7-4 and 7-5 summarize the lifetime costs and emission benefits of the final
locomotive rulemaking. While the costs presented are applicable for both NOx and PM
reductions, the calculated benefits shown are for NOx only. However, it should be
remembered that there are also significant emission reductions in PM (275,000 metric
tons) and HC (400,000 metric tons). The costs are presented in an undiscounted
(actual), and discounted (7% NPV) format. Costs and benefits were computed over a
forty-one year program run to ensure complete fleet turnover, due to the extremely
long service life of the typical locomotive. For the sake of consistency, total fuel
consumption was calculated using the classwide fuel consumption rates used for
calculation of the benefits in Chapter 6. Additional tables showing the year-by-year
costs and emission benefits in both undiscounted and discounted form for the entire 41-
year program can be found in Appendix F. Table F-l shows undiscounted benefits and
Table F-2 shows the benefits discounted at a rate of 7 percent.
The Draft RSD estimated the NOx cost-effectiveness for the program as a whole
at $175 per metric ton. Based on the costs presented above and recomputed benefits,
EPA has recalculated the cost-effectiveness for the total program at $163 per metric
ton of NOx reduction for the base case (shown in Table 7-4) and $253 per metric ton for
the high range case (shown in Table 7-5). As stated earlier, the high range case
includes developmental engineering and testing costs at 150 percent of the base case
and fuel consumption costs that are double the base case. Both of these sets of figures
compare quite favorably to other NOx control strategies that have been adopted in
recent years. For example, the cost-effectiveness for the large nonroad engines
rulemaking was $160 to $360 per ton of NOx. The rule becomes even more cost-
effective when the above-mentioned HC and PM benefits are considered.
EPA also calculated the marginal cost-effectiveness of the three tiers. The results
of this calculation are shown in Appendix F, and are summarized in Table 7-6. To
determine the marginal costs and benefits, EPA considered two additional scenarios:
one in which the Tier 2 standards were eliminated so that the Tier 1 standards
continued to apply after 2005; and one in which both the Tier 1 and Tier 2 standards
were eliminated so that the Tier 0 standards continued to apply for the duration of the
program. The total costs, NOx benefits and cost-effectiveness for these two scenarios
are shown in the appendix. From these scenarios, EPA calculated the marginal cost-
effectiveness for each of the three tiers of standards by dividing the marginal costs by
the marginal benefits, as shown Table 7-6. The Tier 0 figures show the cost, benefits,
118
-------
and cost-effectiveness of continuing the Tier 0 standards for the full period 2000-2040.
The Tier 1 figures show the marginal costs, benefits and cost-effectiveness associated
with the Tier 1 standards, assuming that they were continued after 2005. The Tier 2
figures show the marginal costs, benefits and cost-effectiveness of the Tier 2 standards,
which is calculated from the difference in costs and benefits between the total costs
and benefits for all three tiers of standards (as shown in Table 7-4) and the previous
scenario where the Tier 1 costs and benefits were continued from 2005-2040. As can
be seen from the table, the costs for Tier 1 and Tier 2 are significantly higher than
those for Tier 0. The differential would likely have been even greater if the costs were
estimated over a shorter period of time. For example, when the cost-effectiveness is
calculated for a 20 year period, rather than 41 years, the cost per ton for the total
program increases by more than 40 percent, which could mean costs of upwards of $600
per ton at the high end of the cost range.
119
-------
Table 7-4
Cost-Effectiveness Analysis: Base Case
Category
Total Costs
TIERO
Average number of Tier 0 locomotives produced per year (2000-2013)
Average number of Tier 0 locomotives in the fleet (2000-2040)
INCREMENTAL COSTS:
Initial Manufacture
Fuel consumption
Maintenance
TOTAL
NPV
1,214
10,803
$470,446,480
$435,742,226
$217,159,792
$1,123,348,498
$584,926,672
TIER1
Average number of Tierl locomotives produced per year (2002-2004)
Average number of Tier 1 locomotives in the fleet (2002-2040)
INCREMENTAL COSTS:
Initial Manufacture
Fuel consumption
Maintenance
TOTAL
NPV
480
1,324
$102,890,062
$79,754,324
$32,013,080
$214,657,466
$132,572,277
TIER 2
Average number of Tier 2 locomotives produced per year (2005-2040)
Average number of Tier 2 locomotives in the fleet (2005-2040)
INCREMENTAL COSTS:
Initial Manufacture
Fuel consumption
Maintenance
TOTAL
NPV
$669
$1,186
$78
$1,935
$613
462
9,078
994,839
615,407
433,920
044,166
541,238
TOTAL COSTS
NPV
TOTAL NOx BENEFIT (Tons-M)
COST EFFECTIVENESS ($/Ton)
NPV
$3,273,
$1,331,
20.
050,130
040,187
052,552
$163
$66
120
-------
Table 7-5
Cost-Effectiveness Analysis: High Range
Category
Total Costs
TIERO
Average number of Tier 0 locomotives produced per year (2000-2013)
Average number of Tier 0 locomotives in the fleet (2000-2040)
INCREMENTAL COSTS:
Initial Manufacture
Fuel consumption
Maintenance
TOTAL
NPV
1,214
10,803
$502,544,778
$871,484,452
$217,159,792
$1,591,189,022
$782,324,482
TIER1
Average number of Tierl locomotives produced per year (2002-2004)
Average number of Tier 1 locomotives in the fleet (2002-2040)
INCREMENTAL COSTS:
Initial Manufacture
Fuel consumption
Maintenance
TOTAL
NPV
480
1,324
$134,317,119
$159,508,649
$32,013,080
$325,838,848
$167,500,335
TIER 2
Average number of Tier 2 locomotives produced per year (2005-2040)
Average number of Tier 2 locomotives in the fleet (2005-2040)
INCREMENTAL COSTS:
Initial Manufacture
Fuel consumption
Maintenance
TOTAL
NPV
$706
$2,373
$78
$3,158
$951
462
9,078
878,325
230,814
433,920
543,059
097,583
TOTAL COSTS
NPV
TOTAL NOx BENEFIT (Tons-M)
COST EFFECTIVENESS ($/Ton)
NPV
$5,075,
$1,900,
20.
570,929
922,399
052,552
$253
$95
121
-------
Table 7-6
Marginal Cost-Effectiveness of Each Tier of Standards
Locomotive Standards
TIERO
Total Costs
Total NOx Benefits
Cost-effectiveness ($/Ton)
TIER 1
Marginal Costs
Marginal NOx Benefits
Cost-effectiveness ($/Ton)
TIER 2
Marginal Costs
Marginal NOx Benefits
Cost-effectiveness ($/Ton)
Base Case
$2,422,421,411
12,809,089
$189
$850,628,719
4,172,037
$204
$618,662,877
3,071,426
$201
High Range
$3,572,022,108
12,809,089
$279
$1,503,548,820
4,172,037
$360
$1,229,466,813
3,071,426
$400
122
-------
APPENDICES
-------
APPENDIX A
Locomotive Remanufacture Mileage Data From AAR, June 1996
(Data for Santa Fe Railway)
Remanufacture Miles for 1973 and Later Locomotives
T3
£
i
•i
V
OL
8
£
Q.
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
400
000 500,000 600,000 700,000 800,000 900,000 1,000,000
Remanufacture Mileage
L All data
11/1/83 and earlier
4/1/85 and later
A-l
-------
APPENDIX B
Locomotive Emission Data
by Throttle Notch
B-l
-------
Notch
DB
1
1
2
3
4
5
6
7
8
DB
1
1
2
3
4
5
6
7
a
DB
1
1
2
3
4
5
6
7
8
DB
I
1
2
3
4
5
6
7
8
Model
EMD 1 S-645E3
EMD 16-64SE3
EMO 16-645E3
EMD 16-645E3
EMD 16-645E3
EMO 16-64SE3
EMD 16-M5E3
EMO 16-645E3
EMO 16-645E3
EMD 16-645E3
EMD 16-645E3
EMD 16-64SE3
EMD 20-645E3
EMD 20-645E3
EMO 20-645E3
EMD 20-645E3
EMO 20-64SE3
EMD 20-645E3
EMO 20-645E3
EMD 20-64SE3
EMD 20-645E3
EMO 20-645E3
EMD 20-645E3
EMD 20-645E3
EMD 12-645E3B
EMD 12-645E3B
EMD 12-64SE3B
EMD 12-64SE3B
EMD 12-645E3B
EMD 12-64SE38
EMD 12-645E3B
EMD 12-64SE3B
EMO 12-645E38
EMD 12-64SE3B
EMD 12-64SE3B
EMD 12-64SE3B
EMD 16-64SE3B-
EMD 16-645E3H
EMD 16-645E38
EMD 16-645E3B
EMD 16-645E3B
EMD 16-645E3B
EMD 16-645E3B
EMD 16-64SE3B
EMD 16-645E3B
EMD 16-645E3B
EMD 16-64SE3B
EMD 16-645E3B
Ratea
Power, bhp
3000
3000
3000
3000
3000
3000
3000
3000
3000
3000
Line-haul
Switch
3600
3600
3800
3600
3800
. 3800
3800
3800
3600
3800
Line-haul
Switch
2500
2500
2500
2500
2500
2500
2500
2500
2500
2500
Una-haul
Switch
3000
3000
3000
3000
3000
3000
3000
3000
3000
3000
Line-haul
Switch
Power 1/1
Fuel Rale HC
Notcn, bhp Itvhf
69
17
10S
395
686
1034
1461
1971
2661
3159
853
262
95
17
111
435
781
1219
1741
2299
3344
3819
1023
301
36
11
111
417
594
878
1105
1S17
2103
2451
674
224
138
17
105
363
721
1030
1438
1621
2492
3070
635
255
114
40
94
167
275
404
556
740
994
1177
157
47
68
187
310
468
665
365
1227
1432
76
28
54
159
223
324
404
549
749
872
12fl
279
296
361
432
528
657
827
1066
1186
* HC, CO. and PM data shown ham an from EMO 645E3B passenger locomotive .
g/btip-nr
4,26
10.88
1.48
0.51
0.36
0.31
0.29
0.31
0.33
0.37
0.48
0.33
3.45
14.24
1.48
0.53
0.39
0.31
0.30
0.32
035
0,36
0.48
0.39
1.23
9.36
0.67
0.4S
0.25
0.22
0.23
0.22
0.21
0.20
0.29
0.58
1.70
35.35
1.32
0.70
0.48
0.31
0.20
0.12
0.10
0.10
0.47
1.78
CO
grt)hp-hf
9.56
33,18
2.54
0.74
0.43
0,42
0.52
0.97
1.89
1.87
1.85
2.12
fl.96
S9.B2
2.85
1.04
0.58
0.43
0.64
1.02
0.65
0.68
1.18
2.82
225
31.73
0.86
0.54
0.43
0.65
1.18
2.53
3.80
2.24
2.27
2.00
3.75
90.12
3.27
1.72
1.18
0.81
0.81
0.46
0.55
0.52
1.40
4.64
NOx
g/bhp-hr
S9.91
96.18
26.74
15.29
14.84
14.90
14.30
12.97
11.72
11.69
13.64
18.00
46.63
90.65
24.77
13.23
13.85
13.52
11.96
11.18
13.60
12.34
13.48
15.93
16.42
31.00
13.03
11.29
11.26
11.90
12.60
11.89
13.15
14.05
13.76
14.13
NA-
MA*
NA'
NA*
NA-
NA*
NA-
NA*
NA"
NA*
NA-
NA*
PM
gft>hp-hr
1.16
2.02
0.34
0.34
0.33
0.25
023
0.26
0.24
0.26
0.29
0.36
1.16
3.24
0.34
0.34
0.33
0.25
0.23
0.28
0.24
O.Z7
0.30
0.38
1.47
3.00
0,27
021
0.31
024
0.22
0.27
0.23
0.25
0.29
0.33
0.64
2.59
0.33
0.34
0.33
0.25
0.23
0.28
024
026
029
0.38
NQx data used in tw analysis came from EMD 645E3B freight
locomotive - inatviduaJ notch data not avaHaBJe,
DB
I
1
2
3
4
5
6
7
8
EMD 16-64SF3
EUO16-645F3
EMO 18-64SF3
EUD16-645F3
EMD 1S-645F3
EMD16-645F3
EWD16-645F3
EMD16-645F3
EMD 16-645F3
EMD 16-645F3
EMD 16-645F3
EMD 16-645F3
3500
3500
3500
3500
3500
3500
3500
3500
3500
3500
Line-haul
Switch
74
15
250
444
725
1131
1635
2212
3182
3661
966
305
94
66
137
203
305
445
615
818
1150
1345
3,67
15.20
3.67
0.23
0.53
0.36
029
026
027
0.31
0.49
1.12
7.25
41.13
1.21
0.69
0.49
0.36
0.36
0.75
1.11
1.24
1.33
1.87
62.00
201.60
23.35
18.76
16.41
16.86
16.11
14.63
13.86
12.64
15.54
22.39
0.89
5.50
0.30
0.36
0.35
0.25
0.23
0.27
0.24
0.26
0.30
0.44
B-2
-------
Notcn
Model
Hated Power in Fuel Rate
Power, bhp Notch. Ofip Ib/hr
DB
1
1
2
3
4
5
6
7
a
DB
I
1
2
3
4
5
6
7
3
DB
1
1
2
3
4
5
6
7
a
OB
)
1
2
3
4
5
6
7
a
DB
1
1
2
3
4
5
6
7
8
EMD 12-645F3B
EMC 12-64SF3B
EMD 12-645F3B
EMD 12-645F3B
EMD 12-645F3B
EMD 12-645F3B
EMD 12-64SF3B
EMD 12-64SF3B
EMD 12-645F3B
EMD 12-645F3B
EMD 12-645F3B
EMD 12-645F38
EMD 16-64SF3B
EMD 16-64SF3B
EMD 16-645F3B
EMD 16-645F3B
EMO 18-645F3B
EMD 1 6-645 F3B
EMD 16-645F38
EMO 16-645F38
EMD 16-645F3B
EMD16-64SF3B
EMD 16-645F3B
EMD 16-645F38
EMD 12-71033
EMD 12-710G3
EMD 12-710G3
EMD 12-710G3
EMO 1 2-71 OG3
EMD 12-71033
EMD 12-710O3
EMD 12-710G3
EMD 12-710G3
EMD12-710G3
EMD12-710G3
EMD 12-710G3
EMD 16-71 OG3
EMD 16-710G3
EMD 16-710(33
EMD 18-71QG3
EMD 1 6-71 OG3
EMD 16-710G3
EMD 16-710G3
EMD 1 6-71 OG3
EMD 16-710G3
EM018-710G3
EMD 16-710G3
EMD 16-710G3
EMD 1 2-71 OG3A
EMD 12-71063A
EMD 12-710O3A
EMD 1 2-71 OG3A
EMD 12-71 OG3A
EMD 12-710G3A
EMD12-710G3A
EMO 12-710G3A
EMD 12-710G3A
EMD 12-710G3A
EMD 12-710G3A
EMD 12-710G3A
2350
2350
2850
2850
2850
2850
2350
2350
2350
2850
Line-haul
Switch
3600
3600
3600
3600
3600
3600
3600
3600
3600
3600
Line-haul
Swttcn
3200
3200
3200
3200
3200
3200
3200
3200
3200
3200
Line-haui
Switch
3600
3600
3800
3600
3600
3600
3000
3800
3600
3600
Line- haul
Switch
3200
3200
3200
3200
3200
3200
3200
3200
3200
3200
Una-haul
Switch
15
8
222
339
735
993
1322
1704
2389
2823
769
253
38
9
20S
475
1005
1353
1878
2766
3454
3666
1073
343
22
6
197
356
664
1025
1353
1719
2568
3023
807
252
11
9
203
439
981
1361
1787
2306
3541
4094
1084
329
94
a
209
372
717
1053
1402
1696
2534
3196
846
262
33
18
95
134
274
364
484
615
841
991
91
22
92
179
363
480
652
919
1136
1281
110
35
90
140
256
372
491
617
363
1035
107
26
91
171
354
479
623
799
1190
1363
142
19
91
141
256
372
491
587
648
1077
HC
g/bhp-hr
6.20
6.63
O.G2
0.47
0.38
0.39
0.36
0.34
0.32
0.27
0.35
0.52
7.75
8.44
0.19
0.44
0.31
0.26
0.26
0.24
0.27
0.28
0.33
0-42
16.05
13.00
0.37
0.26
0.21
0.22
023
0.25
0.32
0-28
0.36
0.43
14.40
12.80
0.57
0.40
0.31
0.30
02&
0.28
0.30
0.33
0.38
0.53
1.09
6.88
0.40
0.22
0.17
0.13
0.12
0,11
0.09
0.11
0.15
0.30
CO
-------
Notch
DB
I
1
2
3
4
5
6
7
8
DB
I
I
2
3
4
5
6
7
8
DB
I
1
, 2
3
4
5
6
7
8
OB
1
1
2
3
4
5
6
7
8
Model
EMD 16-710G3A
EMD 16-710G3A
EMD 18-710G3A
EMD 16-710G3A
EMD 16-71QG3A
EMD 16-710G3A
EMD 16-710G3A
EMD 16-7IOG3A
EMO 16-710G3A
EMD 16-71QG3A
EMD 16-710G3A
EMD 16-710G3A
GE12
GE12
GE12
GE12
GE12
GE12
GE12
GE12
GE12
GE12
GE 12
GE12
GE12
GE12
GE12
GE12
GE12
GE12
GE12
GE12
GE12
6E12
GE12
GE12
GE12
GE12
GE12
GE12
GE12
GE12
GE12
GE12
GE12
GE12
Power, brip
3600
3600
3600
3600
3600
3600
3600
3600
3600
3600
Line-haul
Switch
2500
2500
2500
2500
2500
2500
2500
2500
2500
2500
Line-haul
Switch
3000
3000
3000
3000
3000
3000
3000
3000
3000
3000
Una-haul
Switch
3300
3300
3300
3300
3300
3300
3300
3300
3300
3300
Notcti, ohp
23
S
198
430
975
1351
1817
2637
3496
4035
1086
330
96
15
114
266
535
857
1256
1863
2004
2500
686
212
96
15
137
319
642
1024
1507
1998
2405
3000
819
252
62
12
100
200
700
1000
1500
2000
2600
3300
Ibitir
134
23
88
167
351
478
635
838
1147
1328
98
18
52
105
205
320
455
590
697
862
98
18
SO
121
238
368
523
679
802
991
NA*
17
50
86
273
368
532
660
858
1082
Fueling rate data not available.
DB
I
1
2
3
4
5
6
7
8
GE12
GE12
GEIfl
GE16
GEIfl
GE18
GE18
GE 18
GE 16
6E16
GE16
GE16
GE16
GE16
Une-hau
SwHdl
3000
3000
3000
3000
3000
3000
3000
3000
3000
3000
Una-haul
Switch
860
236
117
25
150
300
700
1050
1550
2050
2600
3000
639
265
130
24
66
133
259
405
578
748
882
1090
HC
grt>Hp-hr
9.58
7.41
0.41
0.26
0.18
o.ta
0.14
0.13
0.17
0.18
0.21
0.26
4.40
13.80
1.09
0.47
0.39
0,29
0.28
023
0.26
0.29
0.48
0.95
4.40
17.27
0.91
0.31
0.43
0.28
0.23
025
0-29
025
0.45
0.94
2.90
21.58
1.08
0.60
O.S1
0.33
0-22
0.15
0.14
0.14
0.32
1.00
11.97
19.12
1.18
1.16
0.48
0.36
021
0.28
0.27
0.27
0.73
1.56
CO
gft*iD-hr
6.50
9.88
0.46
024
0.24
0.41
0.90
2.07
2.85
2.80
2.30
1.02
5.53
18,87
1.74
1.59
2.11
2.94
2.85
3.65
2.32
1.37
2.12
3.11
5.53
23.60
1.45
1.06
2.32
2.88
3.06
228
1.37
1.00
1.73
3.04
8.57
2950
1.98
1.69
2.13
2.96
3.68
227
0.99
1.00
1.68
325
15.80
19.66
1.85
1.85
1.16
1,26
1.98
2.00
2.12
2.12
2.44
2.74
NOx
g/bfip-fir
131.41
171.76
20.76
18.60
14.55
14.13
12,48
10.81
9.71
9.08
11.04
1553
S.14
16.33
7.28
6.65
7.60
9.70
11.21
10.80
11.75
10.46
10.32
9.60
5.14
20.40
6.08
4.44
8.36
9.51
10.83
11.51
10.44
11.00
10.56
9.39
7.95
25,50
6.30
7.08
7.67
9.78
10.88
11.49
10,48
11.00
10,75
10.18
11.41
12.36
6.66
6.66
7.SO
7,65
10.01
12.30
12.13
12.13
11.35
9.22
PM
g/X>hp-hr
4.09
3.18
0,25
0.31
0.30
0.23
0.21
0.25
021
0.23
025
0.28
1.38
3.47
0.95
0.56
0.50
0.33
0.24
0.22
0.15
0.14
0.26
0.51
1.88
4.33
0.79
0.38
0.58
0.32
022
0.15
0.14
0.14
0.24
0.48
2.90
5.42
1.08
0.60
0.51
0.33
0.22
0.15
0.14
0.14
024
0.51
5.32
9.12
0.67
0.67
0.35
0.45
0,24
0,18
0.18
0.18
0.41
0.86
B-4
-------
Notch Model
DB GE16
I GE16
1 GE16
2 GE16
3 GE16
4 GE16
5 QE1E
6 GE16
7 GE16
8 GE16
GE16
GI16
OB GE 16
f GE16
1 GE16
2 GE16
3 GE16
4 GE16
S GE16
6 GE16
7 GE16
3 GE16
GE16
GE16
1 Fueling rate data not available.
Rated Power in Fuel Hals
Power, bhp Notch, bhp Ibrtif
3600
3600
3600
3600
3600
3600
3600
3600
3600
3600
Line-haul
Switch
4100
4100
4100
4100
4100
4100
«100
4100
4100
4100
Line-haul
117
25
175
375
850
1250
1650
2450
3100
3600
1001
315
117
25
195
400
950
1400
2050
2770
3440
4100
1127
349
150
28
76
153
298
466
662
B53
1014
1254
NA*
32
78
172
384
526
756
967
1180
1415
HC
gftlhp-hf
11.97
19.12
1.16
Q.4fl
0.38
0.21
0.29
0.27
0.25
024
0.62
1.27
11.97
19.12
1.16
0.48
0.36
0.21
0.29
0.27
0.24
0-24
0.58
1.18
CO
g/bhp-hr
15.60
19.66
1 85
1.16
126
1.96
1.90
2.12
1.50
0.96
1.67
2.47
15.80
19.68
1 as
1.16
1.26
1.98
1.90
2.12
0.96
0.74
1.44
2,38
NOx
gtohp-hr
11.41
12.36
6.68
7.50
7.65
10.01
12.48
12.13
11.75
11.56
11.29
10.10
11.41
12.36
6.66
7.50
7.65
10.01
12.48
12.13
11.56
11.47
11.23
10.09
PM
g/bhp-hr
5.32
B.I 2
0.67
0.35
0,45
0.24
0.17
0.18
0,16
0.17
0.38
0,71
5,32
9.12
0.67
0.35
0.4S
0.24
0.17
0.16
0.17
0.17
0.34
0,67
B-5
-------
Switch Locomotives
Notch
Model
OB
1
1
2
3
4
5
6
7
8
EMD 16-567C
EMD 16-567C
EMD 16-567C
EMD 16-567C
EMD 16-S67C
EMD 16-567C
EMD 16-567C
EMD 16-567C
EMD 16-567C
EMD 16-567C
EMD 16-567C
EMD 16-567C
Fueling rate data not available.
OB
1
1
2
3
4
5
6
7
8
OB
I
1
2
3
4
5
6
7
8
EMD 12-645G
EMD 12-645E
EMD 12-64SE
EMD 12-645E
EMD 1Z-645E
EMD 12-645E
EMD12-645E
EMD 12-64SE
EMD 12-645E
EMD12-645E
EMD 12-645E
EMD 12-645E
EMD 16-64SE
EMO 16-645E
EMD 16-645E
EMD 16-645E
EMD 16-64SE
EMD 16-fidSE
EMD 16-645E
EMD 16-645E
EMD 16-645E
EMD 16-645E
EMD 16-645E
EMD 16-645E
Rated Power in Fuel Hate
Power. bhp Notch, bhp Ib/rtr
NA*
NA"
NA'
NA'
NA-
NA'
NA*
NA'
NA'
NA'
1750
1750
1750
1750
1750
1750
1750
1750
1750
1750
Line-haul
Switch
1500
1500
1500
1500
1500
1500
1SOO
1500
1500
1500
Una-haul
Switch
2000
2000
2000
2000
2000
2000
2000
2000
2000
2000
Line-ham
Switch
50
12
80
181
431
668
960
1231
1528
1B20
SOS
139
70
15
72
233
440
669
385
1109
1372
1586
462
160
82
15
98
333
589
871
1161
1465
1810
2124
613
212
80
26
41
95
167
249
332
419
529
630
103
32
55
137
226
331
442
567
710
854
HC
yohp-hr
6.58
3.04
3.10
3.78
542
6.74
9.78
13.28
1758
22.20
0.79
1.25
2.07
6.60
1.30
0.50
0.33
0.29
0.31
0.34
0.36
0.42
0.51
0.76
3.28
8.27
1.24
0.45
0.32
0.30
0.32
0.32
0.36
0.38
0.49
0.74
CO
gfthp-hr
11.96
468
4.28
7.20
8.60
6.22
11.30
18.76
38.84
125.58
2.72
2.13
S.OO
12.07
2.54
1.28
077
0.53
0.47
0.61
1.52
3.60
2.61
1.71
8.52
18.B7
2.46
1.29
0.73
0.55
0.52
0.63
0.96
1.87
1.69
1.72
NOx
gfljhp-hr
37.62
20.20
22.50
12.94
107.40
t85.48
262.52
382.66
513.26
616.26
16.97
17.27
48.79
65.80
17.22
11.91
12.99
14.64
15.97
16.23
15.96
15.15
16.42
17.39
34.18
83.13
18.82
13.02
13.81
14.25
14.82
15.86
16.36
16.36
16.75
17.55
PM
g/tjhp-fir
1.48
0.52
0.38
1.60
3.30
3.S4
4.84
736
7.62
10.46
0.32
0.40
0.80
2.07
0.32
0.33
0.31
0.24
0.23
0.28
0.2S
0.28
0.31
0.38
0.88
2.53
0.32
0.33
0.32
0.24
0.23
028
0,26
029
0.31
0.38
B-6
-------
APPENDIX C
Calculation of Baseline
Locomotive Emission Rates
This appendix contains a numerical description of how EPA developed its
baseline emission factors. The cycle-weighted emission factors listed here were
calculated from the notch data in Appendix B using the respective line-haul or switch
duty-cycle. Cycle-weighted horsepower (i.e., the average power used over the cycle)
was also calculated. The emission factors were weighted by the product of number of
locomotives in the fleet and the average power:
AVG
These weighted average emission factors were multiplied by deterioration factors to
be more representative of in-use emissions. The line-haul emissions were determined
for two locomotive categories (pre-1991 and 1991-1995) because of the limits of the data
set. The emission factors for these two categories were weighted together in the same
manner as the individual emission factors.
C-l
-------
Weighted Average Line-Haul Emissions
Model
EMD 16-645E3
EMD 20-645E3
EMD 16-645E3B
EMD 16-645F3
EMD 12-645F3B
EMD 16-645F3B
EMD 12-710G3
EMD 16-710G3
EMD 12-710G3A
EMD 16-710G3A
GE 12 - 2500
GE 12 - 3000
GE 12 -3300
GE 16 - 3000
GE 16- 3600
GE 16-4100
Number in
1990 Fleet
1562
723
2693
232
6
400
2
537
17
250
843
145
0
801
451
1029
Cycle-Wtd
Power (hp)
853
1023
835
988
769
1073
807
1084
846
1086
686
819
860
839
1001
1127
Weighted Average
Deterioration Factors
1990 and Earlier In-Use
HC
g/bhp-hr
0.48
0.49
0.47
0.49
0.35
0.33
0.36
0.38
0.15
0.21
0.48
0.45
0.32
0.73
0.62
0.58
0.49
1.15
0.57
CO
g/bhp-hr
1.85
1.18
1.40
1.33
1.17
0.63
0.90
0.52
1.09
2.30
2.12
1.73
1.68
2.44
1.67
1.44
1.53
1.00
1.53
NOx
g/bhp-hr
13.64
13.46
13.12
15.54
11.52
15.23
10.55
11.55
10.75
11.04
10.32
10.56
10.75
11.35
11.29
11.23
12.53
1.00
12.53
PM
g/bhp-hr
0.29
0.30
0.29
0.30
0.25
0.25
0.25
0.26
0.25
0.25
0.26
0.24
0.24
0.41
0.36
0.34
0.30
1.15
0.35
GE
EMD
3000
1500
1200
1200
Weighted Average
Deterioration Factors
1991-1995 In-Use
0.30
0.30
0.30
1.15
0.35
0.90
0.80
0.87
1.00
0.87
13.50
14.00
13.67
1.00
13.70
0.20
0.30
0.23
1.15
0.27
Estimated In-Use Emission Factors for Entire 1995 Line-Haul Fleet:
HC
CO
NOx
PM
0.48
1.28
13.0
0.32
C-2
-------
Weighted Average Switch Emissions
Model
EMD 16-567C
EMD 12-645E
EMD 16-645E
Number in
1990 Fleet
1279
1216
1763
Cycle-Wtd
Power (hp)
159
160
212
Weighted Average
Deterioration Factors
In-Use Baseline
HC
g/bhp-hr
1.25
0.76
0.74
0.88
1.15
1.01
CO
g/bhp-hr
2.13
1.71
1.72
1.83
1.00
1.83
NOx
g/bhp-hr
17.27
17.39
17.55
17.44
1.00
17.44
PM
g/bhp-hr
0.40
0.38
0.38
0.38
1.15
0.44
Estimated In-Use En
lission Factors
for Entire 1995 Line-Haul Fleet:
HC
CO
NOx
PM
1.01
1.83
17.4
0.44
c-:
-------
APPENDIX D
Locomotive Smoke Emissions
Table 25 (page D-2): SwRI Report for EPA, Emission Measurements - Locomotives,
SwRI 5374-024
Table 12 (page D-3) Association of American Railroads, Locomotive Exhaust Emission
Field Tests, AAR Report R-885
Table 28 (page D-4) Association of American Railroads, Locomotive Exhaust Emission
Field Tests, AAR Report R-877
D-l
-------
TABLE 25. SMOKE OPACITY TEST RESULTS
Unit
AT&SF601
BN 9457
SP502
AT&SF202
NS 8842
Smokemeter
Date Fuel Position
2/9/95 EM-1880-F Perpendicular
2/10/95 Diagonal
4/10/95 EM-1902-F Perpendicular
4/11/95 Diagonal
4/12/95 EM-1880-F Diagonal
4/13/95 Perpendicular
5/3/95 EM-1880-F Perpendicular
5/4/95 Diagonal
5/5/95 EM-1902-F Perpendicular
5/6/95 Diagonal
5/23/95 EM-1902-F Perpendicular
5/24/95 Diagonal
5/24/95 EM-1880-F Diagonal
5/25/95 Perpendicular
5/29/95 EM-1902-F Perpendicular
5/30/95 Diagonal
5/31/95 EM-1880-F Diagonal
6/1/95 Perpendicular
Steady-State Smoke, % Opacity
LI
5
6
10
8
9
10
DB Idle
4 5
4 5
2 2
0 0
0 0
1 1
11 14
9 12
11 13
9 12
2 2
2 2
1 2
3 3
5 6
4 4
4 5
4 5
N1
5
4
2
0
0
2
9
5
7
5
2
3
2
4
5
4
4
4
N2
5
4
3
0
1
2
9
7
8
4
3
4
3
4
4
4
3
4
N3
5
4
2
0
0
2
17
15
17
14
4
3
3
4
5
6
2
3
N4
6
5
2
0
0
2
14
12
13
11
3
2
2
3
7
7
2
4
N5
5
4
2
1
0
2
12
8
10
8
4
2
3
3
7
6
2
4
Nti
5
4
1
0
0
2
11
5
7
4
4
1
3
3
9
7
1
3
N7
5
5
0
0
*
1
10
3
4
1
3
1
2
3
__*
8
0
3
N8
5
9
0
1
*
2
9
2
3
0
3
1
2
3
__*
10
1
3
Transient Smoke
Unit
AT&SF601
BN 9457
SP502
AT&SF202
NS 8842
"bulb failure
"heavy rain
Smokemeter
Date Fuel Position
2/9/95 EM-1880-F Perpendicular
2/10/95 Diagonal
4/10/95 EM-1902-F Perpendicular
4/11/95 Diagonal
4/12/95 EM-1880-F Diagonal
4/13/95 Perpendicular
5/3/95 EM-1880-F Perpendicular
5/4/95 Diagonal
5/5/95 EM-1902-F Perpendicular
5/6/95 Diagonal
5/23/95 EM-1902-F Perpendicular
5/24/95 Diagonal
5/24/95 EM-1880-F Diagonal
5/25/95 Perpendicular
5/29/95 EM-1902-F Perpendicular
5/30/95 Diagonal
5/31/95 EM-1880-F Diagonal
6/1/95 Perpendicular
LI
12
13
18
14
17
16
Maximum %
DB Idle
12 15
13 11
7 3
1 0
3 1
5 2
60 87
56 86
26 71
43 76
2 7
3 7
2 7
3 7
5 14
8 6
9 13
10 10
Opacity During Throttle Notch
N1
22
32
5
1
2
3
55
62
56
61
4
5
4
5
21
21
22
23
N2
36
40
6
1
3
4
55
58
60
56
5
5
5
6
36
61*
21
33
N3
13
13
9
0
3
4
59
58
58
60
5
4
5
5
12
11
8
13
N4
19
16
8
3
3
4
44
41
40
41
4
3
4
4
14
12
7
11
N5
17
17
8
2
1
4
17
13
16
14
6
5
6
5
15
13
6
10
Change
N6
10
10
11
2
8
8
39
24
19
14
8
6
6
6
12
10
4
7
N7
9
10
4
1
__*
4
21
5
7
4
4
3
4
5
..*
11
3
5
N8
8
12
2
2
__*
2
24
5
6
3
6
3
4
5
..*
17
7
6
D-2
-------
TABLE 12. STEADY-STATE SMOKE TEST RESULTS
Locomotive Model
Standard Timing
4° Retarded Timing
Maximum Steady State Smoke3
(% Opacity)
Increase in
Maximum Opacity
(% Opacity)
EMD SD40 2 (E3B)
UP 3808
UP 3953
UP 3938
UP 3959
UP 3228
30
9
12
8
5
3
15
21
12
6
O
6
9
4
1
GE C40 8
UP 9108
UP 9113
UP 9126
UP 9133
17
25
35
28
34
33
55
51
17
8
20
23
EMD SD40 2 (E3)
SP 8270
SP7315
SP 8303
SP 7323
9
21
12
<5
11
40
17
<5
2
19
5
0
Note: a - Maximum steady- state smoke opacity in any notch during "UP"
smoke test.
D-3
-------
TABLE 28. STEADY-STATE SMOKE TEST SUMMARY
Locomotive Unit
Number
SP 2706b
SP 2754b
SP 2742b
SP 2720b
SP 2739b
Amtrak514c
Amtrak 229C
Amtrak 806C
CSX 8704
SP 6344
UP20
CSX 8709
AT&SF 601
SP 6344d
Model
EMD MP15AC
EMD MP15AC
EMD MP15AC
EMD MP15AC
EMD MP15AC
GE DASH8-B32
EMD F40PH
GE AMD- 103
EMD SD60
EMD GP35
Republic RD20
EMD SD60
GE DASH9-
44CW
EMD GP35
Standard Timing
4° Retarded
Timing
Maximum Steady-State Smoke (%
Opacity)"1
7
6
8
11
10
10
11
9
10
35
<5
8
5
34
8
8
16
15
23
10
11
14
12
35
30
16
11
49
Increase in
Maximum Opacity
(% Opacity)
1
2
8
4
13
0
0
5
2
0
30
8
6
15
Note: a - Maximum steady-state smoke opacity in any notch during "UP" smoke cycle.
b - Highest reading from either of the two exhaust stacks.
c - Non-HEP
d - Repeat testing of SP 6344.
D-4
-------
APPENDIX E
Compliance Margins for On-Highway
Heavy-Duty Diesel Engines
In complying with engine emission standards, engine manufacturers typically
strive to produce engines that emit at levels below the applicable standards. This
compliance margin (i.e., the difference between the level of the standard and the
engine's certification level or FEL) is used by the manufacturers to assure that factors
such as testing, production variability or unexpected deterioration will not result in
noncompliance. Since engines typically operate at emission levels below the standards,
it is important to consider the engine's actual emissions, rather than the levels of the
standards, when quantifying the emission benefits of regulations.
An analysis of actual compliance margins from on-highway heavy-duty diesel
engine (HDDE) certification results was performed in order to estimate the actual
locomotive emission levels that will result from the federal locomotive emission
standards. HDDEs were chosen because they are similar in many ways to locomotive
engines, and it is expected that locomotive engine manufacturers will use compliance
strategies similar to those currently used for on-highway heavy-duty diesel engines.
The 1993 through 1995 model years were chosen for the analysis of compliance
margins. The 1993 to 1994 transition represents a change in the particulate standard
for HDDEs (from 0.25 to 0.10 g/bhp-hr) and could be seen as indicating what might
happen in the case of previously unregulated locomotives coming under new emission
standards. For the 1995 model year, the analysis was limited to those engine families
which had been certified as of January 30, 1995. Although a few more 1995 HDDEs
were certified, those included in this analysis represent the majority of 1995 HDDEs.
When determining what compliance margin to use for a given engine family,
HDDE manufacturers take many factors into account. One of the most important
factors is the expected potential for deterioration of critical emission control
components. Until recently, HDDE manufacturers have met the applicable standards
with modifications to the engine itself. However, due to the increasing stringency of
the HDDE particulate standard there has been increasing use of oxidation catalysts
on HDDEs. Given that exhaust aftertreatment technologies typically have higher
rates of deterioration compared to engine technologies, it is likely that manufacturers
would utilize different compliance strategies for engines with catalysts compared to
engines without catalysts. For this reason only HDDEs which did not utilize exhaust
aftertreatment were considered in the analysis of HDDE compliance margins.
In addition to excluding HDDEs with exhaust aftertreatment from the
compliance margin analysis, engines certified to the urban bus standards were also
excluded. Given that urban bus engines are required to meet more stringent
E-l
-------
particulate standards than other HDDEs, they cannot be directly compared to other
HDDEs and must be analyzed separately. Additionally, only two manufacturers
currently certify urban bus engine families, and the total number of urban bus engine
families is very small in relation to all HDDE families. In the emissions averaging,
banking and trading program (ABT) urban bus engines constitute a separate class and
cannot be averaged with other HDDEs. The situation of few engine families and an
exclusive ABT class leads to manufacturers using unique compliance strategies for this
class of HDDEs such as using a trap oxidizer on one family for particulate emissions
well below the standard and using these credits to offset production of another family
which emits well above the standard. These strategies lead to very few urban bus
engine families with widely varying emissions which are not representative of HDDEs
as a whole. Thus, urban bus engines are also excluded from this analysis.
The results of the analysis are summarized in Table E-1. NOx and particulate
matter (PM) compliance margins were analyzed separately. Additionally, for each
model year the engines were sorted according to HDDE subclass (light, medium and
heavy) and according to whether the engine was certified to the applicable standards
for NOx and PM or whether they were certified to a family emission limit (FEL) above
or below the standard under the ABT program. The purpose of this sorting was to
determine whether there are trends which should lead to the exclusion of particular
groups of engines, as was done with aftertreatment-equipped and urban bus engines.
As can be seen from Table E-l, neither the change in PM standards from the 1993 to
the 1994 model year, nor the sorting according HDDE subclass and FEL yields any
strong trends which would lead to excluding any of these groups from the analysis. In
addition to analysis by model year, the average of all three model years is included in
Table E-l.
E-2
-------
Table E-l
NOx and Particulate Compliance Margins of Heavy-Duty Diesel Engines1 2
NOx FELstd
PM FELstd
Class=L
Class=M
Class=H
Total
1993
NOx
7.7
7.7
9.8
8.1
7.5
7.3
11.9
6.8
7.6
7.8
PM
19.7
19.3
11.9
19.0
19.1
17.7
19.2
16.1
20.9
18.9
1994
NOx
9.0
8.2
8.9
NA
6.1
9.4
12.6
9.1
6.9
8.3
PM
29.3
17.1
18.0
NA
18.5
18.3
22.0
12.0
21.0
18.4
1995
NOx
5.8
8.5
8.2
NA
8.8
7.4
9.4
10.1
7.0
8.1
PM
35.9
20.1
14.6
NA
23.5
20.9
21.6
16.1
24.9
22.0
Average
NOx
7.5
8.1
9.0
NA
7.5
8.0
11.3
8.7
7.2
8.1
PM
28.3
18.8
14.8
NA
20.4
19.0
20.9
14.7
22.3
19.8
Avg NOx+PM
17.9
13.1
11.9
NA
14.0
13.5
16.1
11.7
14.8
14.0
1. Compliance margins expressed as a percentage of applicable family emission limit. Urban bus
engines and engines utilizing exhaust aftertreatment excluded.
2. The NOx standard is 5.0 g/bhp-hr for all three model years. The PM standard is 0.25 g/bhp-hr for
the 1993 model year and 0.10 g/bhp-hr for the 1994 and 1995 model years.
-------
APPENDIX F
Supplementary Cost-Effectiveness Data
Year-by-Year Costs and Benefits
Tables F-l and F-2 contain year-by-year costs and benefits for the locomotive
standards rulemaking described in Section 7.3. (Note: Benefits are consistent with
the uncorrected benefits of Chapter 6.) Table F-l shows the costs and benefits in
undiscounted form, while Table F-2 shows the year-by-year results discounted at an
interest rate of 7 percent. (Note: Discounted benefits were calculated upon request
by the Office of Management and Budget; they were not used by EPA during this
rulemaking.) Summing the annual discounted costs yields a 41-year fleet wide cost
of $1.2 billion and emission reductions of 5.3 million tons of NOx, 84,000 tons of HC,
and 58,000 tons of PM. The resulting 41-year annualized fleetwide costs and
emission reductions are $89 million per year and 390,000 tons of NOx, 6,000 tons of
HC, and 4,000 tons of PM, respectively. A copy of the spreadsheet prepared for this
41-year cost and benefit analysis has been placed in the Public Docket for this
rulemaking.
Marginal Cost Effectiveness
Tables F-3A through F-4B show the marginal cost effectiveness calculations
described in Section 7.3. Tables F-3A and F-4A are based on the costs contained in
the base case cost scenario shown in Table 7-4, while Tables F-3B and F-4B
represent the same calculations made using the high range costs contained in Table
7-5. Tables F-3A through F-4B are shown in a format similar to the format of
Tables 7-4 and 7-5. The format is similar in that economic information is shown
separately for the same three sets of locomotives: 1) those originally manufactured
1973-2001; 2) those originally manufactured 2001-2004; and 3) those originally
manufactured 2005-2040. Tables F-3A and F-3B show the how the cost
effectiveness would have changed if the Tier 2 standards had not been finalized, so
that newly manufactured locomotives would have continued to be certified to the
Tier 1 after 2004. Tables F-4A and F-4B show the how the cost effectiveness would
have changed if neither the Tier 1 nor the Tier 2 standards had been finalized, so
that newly manufactured locomotives would have continued to be certified to the
Tier 0 after 2001; the total costs from these tables are the marginal Tier 0 costs.
Marginal costs for Tier 1 are calculated by subtracting the total costs of Table F-4A
(F-4B) from the total costs of Table F-3A (F-3B). Marginal costs for Tier 2 are
calculated by subtracting the total costs of Table F-3A (F-3B) from the total costs of
Table 7-4 (7-5).
F-l
-------
TABLE F-l
UNDISCOUNTED COSTS AND BENEFITS OF RULEMAKING
YEAR
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
COST
$14,752,491
$48,065,475
$109,864,672
$116,983,557
$124,777,069
$68,250,344
$76,457,368
$81,976,166
$86,155,991
$90,346,096
$88,538,250
$80,012,843
$81,482,532
$82,962,502
$84,452,752
$85,953,282
$87,464,092
$88,535,431
$89,617,050
$90,708,949
$91,811,127
$92,923,586
$94,046,325
$95,179,344
$96,255,180
$97,206,371
$98,163,345
$99,124,829
$74,766,683
$77,040,784
$79,240,704
$80,258,167
$82,314,167
$84,370,167
$86,426,167
$88,482,167
$90,538,167
$92,594,167
$94,650,167
$96,706,167
$98,762,167
BENEFITS - Metric Tons
HC
0
0
44
87
127
1433
2614
3687
4611
5440
6277
6663
7185
7703
8216
8696
9171
9640
10104
10562
11016
11463
11906
12343
12774
13200
13621
14036
14444
14847
15257
15641
16019
16393
16762
16967
17092
17346
17593
17833
18065
NOx
11211
45880
110000
171261
231215
303684
359785
393564
411517
427815
448821
456429
466628
476765
486832
496414
505915
514137
522277
530336
538314
546211
554026
561759
569231
576263
583200
590030
596578
603033
609532
615772
621924
627988
633965
637232
638943
641226
643455
645631
647756
PM
0
0
0
1
1
928
1765
2525
3178
3764
4354
4622
4986
5346
5703
6036
6365
6690
7011
7327
7640
7948
8252
8553
8849
9141
9428
9712
9992
10267
10544
10801
11054
11303
11549
11678
11750
11918
12080
12236
12387
F-2
-------
TABLE F-2
DISCOUNTED COSTS AND BENEFITS OF RULEMAKING
YEAR
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
COST
$14,752,491
$44,921,005
$95,960,060
$95,493,429
$95,191,828
$48,661,552
$50,946,773
$51,050,636
$50,143,571
$49,142,290
$45,008,357
$38,013,525
$36,179,219
$34,426,489
$32,752,233
$31,153,425
$29,627,114
$28,028,050
$26,514,451
$25,081,780
$23,725,740
$22,442,262
$21,227,494
$20,077,786
$18,976,383
$17,910,194
$16,903,286
$15,952,195
$11,245,075
$10,829,070
$10,409,623
$9,853,536
$9,444,820
$9,047,409
$8,661,574
$8,287,500
$7,925,300
$7,575,021
$7,236,654
$6,910,140
$6,595,375
BENEFITS - Metric Tons
HC
0
0
39
71
97
1021
1742
2296
2683
2959
3191
3166
3190
3196
3186
3152
3106
3052
2989
2921
2847
2769
2687
2604
2518
2432
2345
2259
2172
2087
2004
1920
1838
1758
1680
1589
1496
1419
1345
1274
1206
NOx
11211
42878
96078
139800
176393
216522
239740
245092
239506
232703
228158
216846
207188
197840
188802
179923
171371
162763
154523
146642
139111
131917
125051
118501
112222
106176
100424
94954
89727
84764
80072
75600
71360
67342
63536
59685
55930
52458
49197
46134
43257
PM
0
0
0
1
1
662
1176
1572
1850
2048
2213
2196
2214
2218
2212
2188
2156
2118
2074
2026
1974
1920
1863
1804
1744
1684
1624
1563
1503
1443
1385
1326
1268
1212
1157
1094
1028
975
924
874
827
F-2
-------
TABLE F-3A - MARGINAL COST-EFFECTIVENESS ANALYSIS
(BASE CASE ANALYSIS WITHOUT TIER 2 STANDARDS)
CATEGORY
TOTAL COSTS
TIER 0 LOCOMOTIVES (Model Years 1973-2001)
INCREMENTAL COSTS:
Initial Manufacture
Fuel consumption
Maintenance
TOTAL
NPV
$470,446,480
$435,742,226
$217,159,792
$1,123,348,498
$584,926,672
TIER 1 LOCOMOTIVES (Model Years 2002-2004)
INCREMENTAL COSTS:
Initial Manufacture
Fuel consumption
Maintenance
TOTAL
NPV
$102,890,062
$79,754,324
$32,013,080
$214,657,466
$132,572,277
TIER 1 LOCOMOTIVES (Model Years 2005-2040)
INCREMENTAL COSTS:
Initial Manufacture
Fuel consumption
Maintenance
TOTAL
NPV
$644,639,666
$593,307,704
$78,433,920
$1,316,381,289
$435,028,487
TOTAL COSTS
NPV
TOTAL NOx BENEFIT (Tons-M)
COST EFFECTIVENESS ($/Ton)
NPV
$2,654,387,253
$1,152,527,435
16,981,126
$156
$68
F-4
-------
TABLE F-4A - MARGINAL COST-EFFECTIVENESS ANALYSIS
(BASE CASE ANALYSIS WITHOUT TIER 1 OR TIER 2 STANDARDS)
CATEGORY
TOTAL COSTS
TIER 0 LOCOMOTIVES (Model Years 1973-2001)
INCREMENTAL COSTS:
Initial Manufacture
Fuel consumption
Maintenance
TOTAL
NPV
$470,446,480
$435,742,226
$217,159,792
$1,123,348,498
$584,926,672
TIER 0 LOCOMOTIVES (Model Years 2002-2004)
INCREMENTAL COSTS:
Initial Manufacture
Fuel consumption
Maintenance
TOTAL
NPV
$42,648,927
$79,754,324
$12,392,160
$134,795,412
$73,219,144
TIER 0 LOCOMOTIVES (Model Years 2005-2040)
INCREMENTAL COSTS:
Initial Manufacture
Fuel consumption
Maintenance
TOTAL
NPV
$492,
$593,
$78,
$1,164,
$374,
535,877
307,704
433,920
277,501
170,736
TOTAL COSTS
NPV
TOTAL NOx BENEFIT (Tons-M)
COST EFFECTIVENESS ($/Ton)
NPV
$2,422
$1,032
12
421,411
316,552
809,089
$189
$81
F-5
-------
TABLE F-3B - MARGINAL COST-EFFECTIVENESS ANALYSIS
(HIGH RANGE ANALYSIS WITHOUT TIER 2 STANDARDS)
CATEGORY
TOTAL COSTS
TIER 0 LOCOMOTIVES (Model Years 1973-2001)
INCREMENTAL COSTS:
Initial Manufacture
Fuel consumption
Maintenance
TOTAL
NPV
$502,544,778
$871,484,452
$217,159,792
$1,591,189,022
$782,324,482
TIER 1 LOCOMOTIVES (Model Years 2002-2004)
INCREMENTAL COSTS:
Initial Manufacture
Fuel consumption
Maintenance
TOTAL
NPV
$134,317,119
$159,508,649
$32,013,080
$325,838,848
$191,771,133
TIER 1 LOCOMOTIVES (Model Years 2005-2040)
INCREMENTAL COSTS:
Initial Manufacture
Fuel consumption
Maintenance
TOTAL
NPV
$664,026,920
$1,186,615,407
$78,433,920
$1,929,076,247
$600,514,176
TOTAL COSTS
NPV
TOTAL NOx BENEFIT (Tons-M)
COST EFFECTIVENESS ($/Ton)
NPV
$3,846,104,116
$1,574,609,791
16,981,126
$226
$93
F-6
-------
TABLE F-4B - MARGINAL COST-EFFECTIVENESS ANALYSIS
(HIGH RANGE ANALYSIS WITHOUT TIER 1 OR TIER 2 STANDARDS)
CATEGORY
TOTAL COSTS
TIER 0 LOCOMOTIVES (Model Years 1973-2001)
INCREMENTAL COSTS:
Initial Manufacture
Fuel consumption
Maintenance
TOTAL
NPV
$470,446,480
$871,484,452
$217,159,792
$1,559,090,724
$759,309,513
TIER 0 LOCOMOTIVES (Model Years 2002-2004)
INCREMENTAL COSTS:
Initial Manufacture
Fuel consumption
Maintenance
TOTAL
NPV
$45,900,000
$159,508,649
$12,392,160
$217,800,809
$107,770,438
TIER 0 LOCOMOTIVES (Model Years 2005-2040)
INCREMENTAL COSTS:
Initial Manufacture
Fuel consumption
Maintenance
TOTAL
NPV
$530
$1,186
$78
$1,795
$546
081,249
615,407
433,920
130,576
921,611
TOTAL COSTS
NPV
TOTAL NOx BENEFIT (Tons-M)
COST EFFECTIVENESS ($/Ton)
NPV
$3,572
$1,414
12
022,108
001,562
809,089
$279
$110
F-7
-------
APPENDIX G
Results of SwRI Testing for EPA
From "Emission Measurements - Locomotives"
SwRI 5374-024, August 1995
G-l
-------
TABLE 23. AVERAGE DUTY-CYCLE WEIGHTED EXHAUST EMISSIONS
USING LOW-SULFUR CERTIFICATION DIESEL FUEL
Locomotive
AT&SF No. 601
GE DASH 9-44CW
BN No. 9457
EMD SD70MAC
SP No. 502
MK5000C
AT&SF No. 202
EMD SD75M
NS No. 8842
GE DASH 9-40C
Duty Cycle
EPA Freight
EPA Switcher
AAR 3-Mode
EPA Freight
EPA Switcher
AAR 3-Mode
EPA Freight
EPA Switcher
AAR 3-Mode
EPA Freight
EPA Switcher
AAR 3-Mode
EPA Freight
EPA Switcher
AAR 3-Mode
g/hp-hra
HC
0.21
0.40
0.18
0.31
0.44
0.28
0.51
0.81
0.43
0.28
0.43
0.28
0.29
0.52
0.27
CO
1.10
1.63
1.28
0.60
0.68
0.57
1.07
1.62
0.99
0.94
0.88
0.87
0.85
1.33
0.87
NOx
10.84
12.23
10.80
13.28
13.62
13.01
15.46
20.88
14.59
13.23
15.19
13.05
14.07
14.88
14.38
PM#1a
NA
NA
0.05
NA
NA
0.23
NA
NA
0.11
NA
NA
0.23
NA
NA
0.11
PM#2b
0.07
0.13
0.06
0.22
0.22
0.21
0.16
0.39
0.11
0.24
0.25
0.24
0.11
0.21
0.11
PM#3C
0.08
0.15
0.07
0.24
0.24
0.23
0.19
0.45
0.13
0.30
0.31
0.30
0.15
0.28
0.14
Notes: a - PM sampling system #1 - 30 cm/s face velocity and 30-minute maximum sampling time.
b - PM sampling system #2 - 70 cm/s face velocity and 15-minute maximum sampling time.
c - PM sampling system #3 - 70 cm/s face velocity and 5-minute maximum sampling time.
G-2
-------
TABLE 24. AVERAGE DUTY-CYCLE WEIGHTED EXHAUST EMISSIONS
USING HIGH-SULFUR DIESEL FUEL
Locomotive
AT&SF No. 601
GE DASH 9-44CW
BN No. 9457
EMD SD70MAC
SP No. 502
MK5000C
AT&SF No. 202
EMD SD75M
NS No. 8842
GE DASH 9-40C
Duty Cycle
EPA Freight
EPA Switcher
AAR 3-Mode
EPA Freight
EPA Switcher
AAR 3-Mode
EPA Freight
EPA Switcher
AAR 3-Mode
EPA Freight
EPA Switcher
AAR 3-Mode
EPA Freight
EPA Switcher
AAR 3-Mode
g/hp-hra
HC
***
***
***
0.28
0.38
0.25
0.54
0.90
0.45
0.33
0.52
0.31
0.34
0.58
0.32
CO
***
***
***
0.64
0.78
0.63
1.09
1.66
1.00
0.88
0.86
0.79
0.89
1.38
0.92
NOx
***
***
***
14.19
14.80
13.72
15.86
21.33
14.96
13.79
15.75
13.48
15.01
15.80
15.32
PM#1a
***
***
***
NA
NA
0.28
NA
NA
0.18
NA
NA
0.28
NA
NA
0.18
PM#2b
***
***
***
0.29
0.29
0.28
0.24
0.46
0.19
0.29
0.32
0.29
0.19
0.27
0.19
PM#3C
***
***
***
0.30
0.30
0.30
0.27
0.55
0.22
0.35
0.39
0.34
0.22
0.33
0.21
Notes: a - PM sampling system #1 - 30 cm/s face velocity and 30-minute maximum sampling time.
b - PM sampling system #2 - 70 cm/s face velocity and 15-minute maximum sampling time.
c - PM sampling system #3 - 70 cm/s face velocity and 5-minute maximum sampling time.
*** - No tests were conducted on AT&SF No. 601 using high-sulfur fuel.
G-3
-------
APPENDIX H
Graphical HC and CO Data:
Numerical Data are in Appendices B and C
NOx and PM Data are in Chapter 4
(Note: Not all of the data presented here or in Figures 4-2 and 4-3 were used in the
baseline emission analysis; and thus, not all the data in this appendix are shown in
Appendices B and C.)
H-l
-------
Line-Haul Cycle Emissions Data
HC and CO (g/bhp-hr)
1
0.8
0.6
o
1C
0.4
0.2
D
DD
DDD ° o o D D
D
D
1
CO
Switch Cycle Emissions Data
2
1.5
O
I 1
0.5
0
c
HC and CO (g/bhp-hr)
D
D
D
° ° D
-
D D
BD o
DDDD ° °
r^D
tr
D D
I I I I
) 1 2345
CO
H-2
-------
APPENDIX I
Environmental Analysis
This appendix contains the original uncorrected environmental analysis. See Chapter
6 for details, and Appendix O for corrected tables.
1-1
-------
The figure below shows the relationship between relative usage rate and
average fleet age assumed by EPA for Class I line-haul locomotives. It is based on the
following assumptions:
(1) That each individual newly manufactured locomotive is used to the
maximum extent possible for the first 10 years of its service life. This means
that this maximum usage rate (hours of use per year) applies to new Tier 1 or
Tier 2 fleets for the first 10 years of production, which corresponds to an
average fleet age of up to 5.5 years.
(2) That after this point, the relative usage rate decreases linearly with average
fleet age such that the usage rate for a fleet with an average age of 40 years
would be 50 percent of the usage rate of a newly manufactured locomotive.
1.2
1
0.8
I
0.6
0.4
0.2
0
Locomotive Usage Rates as a Function of Age
Relative Use of each Tier of Locomotive by Class I Railroads
10 20 30
Average Age of Fleet (Years)
40
50
1-2
-------
Year
Relative Usage Factors
Base Tier 0 Tier I
Tier II
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
0.790 0.790
0.783 0.783
0.775 0.775
0.768 0.768
0.761 0.761
0.754 0.754
0.746 0.746
0.739 0.739
0.732 0.732
0.725 0.725
0.717
0.710
0.703
0.696
0.688
0.681
0.674
0.667
. 0.659
0.652
0.645
0.638
0.630
0.623
0.616
0.609
0.601
0.594
0.587
0.580
0.572
0.565
0.558
0.551
0.543
0.536
0.529
0.522
0.514
0.507
0.500
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
0.983
0.967
0.950
0.933
0.917
0.900
0.883
0.867
0.850
0.833
0.817
0.800
0.783
0.767
0.750
0.733
0.717
0.700
0.683
0.667
0.650
0.633
0.617
0.600
0,583
0.567
0.550
0.533
0.517
1.000
1.000
1.000
.000
.000
.000
.000
.000
1.000
1.000
0.993
0.986
0.978
0.971
0.964
0.957
0.949
0.942
0.935
0.928
0.920
0.913
0.906
0.899
0.891
0.884
0.877
0.870
0.862
0.855
0.848
0.841
0.833
0.826
0.819
0.812
1-3
-------
Class 1 Line-Hail Locomotives
1
1 Year
1990
2000
2001
2002
2003
2004
2005
2006
2007
2009
2008
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
202 5
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
£038
2039
2040
Number ol Locomotives
Base Tier 0 Tier 1 Tier II
16500 000
15900 600 0 0
14300 2500 0 0
11700 4700 400 0
9100 6900 BOO 0
6SOO 9400 1200 0
3600 11700 1200 600
1700 13400 1200 1200
700 14000 1200 1800
500 14000 1200 2400
500 14000 1200 3000
0 14000 1200 3600
0 14000 1200 3900
0 13400 1200 4200
0 12BOO 1200 4500
0 12200 1200 4800
0 I1BOO 1200 51OO
0 11000 1200 5400
0 10400 1200 5700
0 9800 1200 6000
0 9200 1200 6300
0 8600 1200 6600
0 8000 1200 6900
0 7400 1200 7200
0 6600 1200 7500
Q 6200 1200 7800
0 5600 1200 8tOO
0 5000 1200 8400
0 4400 1200 8700
0 3800 1200 9000
0 3200 1200 9300
0 2600 1200 9600
0 2000 1200 9900
0 1400 1200 10200
0 600 1200 10500
0 200 1200 10600
0 0 1200 11100
0 0 1200 11400
0 0 1000 11700
0 0 BOO 12000
0 0 600 12300
0 0 400 12600
Fuel Usage 1
Base Tier 0 Tin 1 Tier II
100% 0% 0% 0%
96% i 4% 0% 0%
85% 15% 0% D%
68% 27% 5% 0%
51% 39% 10% 0%
35% 51% 14% 0%
19% 50% 13% 8%
8% 64% 13% 1S%
3% 63% 12% 22%
3% 59% 11% 27%
2% 55% 11% 32%
0% 52% 10% 37%
0% 50% 10% 40%
0% 46% 10% 42%
0% 45% 10% 45%
0% 42% 10% 49%
9% 40% 8% 51%
0% 37% 9% 64%
0% 35% 9% 56%
0% 33% 9% 59%
0% 30% 9% 61%
0% 28% 9% 64%
0% 26% 8% 66%
0% 23% 8% 68%
0% 21% 8% 71%
0% 19% 8% 73%
0% 17% 8% 75%
0% 1S% 8% 77%
0% 13% 7% 79%
0% 11% 7% B2%
0% 8% 7% 04%
D% 9% 7% 86%
0% 6% 7% 88%
0% 4% 7% 89%
0% 2% 6% 91%
0% 1% 9% 93%
0% 0% 6% 94%.
0% 0% 6% 94%
0% 9% 5% 95%
0% 0% 4% 98%
0% 0% 3% 97%
.0% .0%. 3& 98?.
Emission Factors (gtohp-tiO
HC CO NOx PM
048 1.2B 13.00 0.32
0,48 1.26 12.84 0.32
0.48 1.28 12.34 032
0.48 1.28 11.47 0.32
0.48 1.28 1065 0.32
0.48 1.28 9.84 0.32
0.46 1.28 8,86 0.31
0.44 178 612 0.30
0.43 1.28 7.67 0.29
0.42 1.26 7.46 0.2B
0.41 1.28 7.27 0.27
0.39 1,26 7.01 0.26
0.39 1,28 6.94 0.26
0.38 1.28 6.83 0.25
0.38 1.28 6.73 0,25
0.37 1.28 6.63 0.24
0.36 1.28 6.54 0.24
0,36 1.28 6.45 0.23
0.35 1.28 6.36 0,23
0.35 1.28 627 0.23
0.34 1.28 6.19 0.22
0.34 1.26 6.10 0.22
0.33 1.28 8,02 0.21
0.33 1.26 5.94 0.21
0.32 .28 5,96 0.21
0.32 .28 5.78 0.20
0.31 .28 5.70 0.20
0.31 .28 5.62 0.20
0.30 .28 S.SS 0.19
0.30 1.26 546 0,19
0.29 1.28 5.41 0.19
0.29 128 5.34 0.18
0.28 1.26 527 0.18
0.28 1.28 5.20 0.16
0.27 1.28 5-14 0.17
0.27 1.28 5.08 0.17
0.27 1.28 5.05 0.17
0,27 1.28 5.05 0.17
0.26 1.28 5.03 0.17
0.26 128 5.01 0.17
0.26 1.28 4.99 0.16
O.ae. 1*8 4.98 0.16
PerceM Reduction
HC CO NOx PM
0% 0% 0% 0%
0% 0% 1% 0%
0% 0% 5% 0%
0% 0% 12% 0%
0% 0% 18% 0%
0% 0% 24% 0%
4% 0% 32% 4%
8% 0% 38% 8%
11% 0% 41% 11%
13% 0% 43% (4%
15% 0% 44% 16%
18% 0% 46% 19%
19% 0% 47% 20%
20% 0% 47% 21%
22% 0% 48% 23%
23% 0% 49% 24%
24% 0% 50% 25%
25% 0% 50% 27%
27% 0% 51% 28%
28% 0% S2% 29%
29% 0% 52% 31%
30% 0% 53% 32%
31% 0% 54% 33%
32% 0% 54% 34%
33% 0% 55% 35%
34% 0% 56% 36%
35% 0% 56% 38%
36% 0% 57% 39%
37% 0% 57% 40%
38% 0% 56% 41%
39% 0% 58% 42%
40% 0% 59% 43%
41% 0% 58% 44%
42% 0% 60% 45%
43% 0% 60% W%
44% 0% 61% 47%
44% 0% 61% 47%
44% 0% 61% 47%
45% 0% 61% 48%
45% 0% 61% 48%
46% 0% 62% 46%
4J% 0% 62% 4|£
Emissions (Metnc Tons/Vear)
HC CO NOx PM
33256 8B683 900662 22171
332S6 88G83 889471 22171
33256 BBBB3 854(103 22171
33213 88683 794315 22171
33171 88683 737745 22171
33133 68683 681953 22171
31883 86683 613992 21377
30758 86683 562423 20473
29742 88683 S3 1620 19747
28878 68603 516666 19126
28109 88693 503396 18577
27333 88683 485441 18023
27008 88683 460561 17791
26532 66663 473464 17464
26099 88683 466453 17141
25653 88683 4S9S37 16823
25240 68683 453131 16528
24634 68683 446632 16239
24435 68683 440639 15954
24043 88683 434553 15674
23657 88683 428573 15398
23277 88683 422700 15128
22905 68683 416933 14862
22539 88683 411273 14600
22160 88683 405719 14344
21627 68683 400273 14092
21481 66683 394933 1384S
21142 86683 389700 13603
20810 86663 304574 13365
20484 86683 37955$ 13133
20 165 86683 374644 12905
19652 68603 3S9840 12661
19548 68683 365143 12463
19247 88663 360553 12249
IBaSS 63663 356071 12040
18669 88663 351696 11836
185SO 88683 350053 11750
18512 88683 349750 11721
18347 88683 348425 11597
18192 88683 347178 11461
18047 88683 3460GB 11371
17911 88683 344914 11269
Reductions (Metric Tons/Year)
HC CO NO* PM I
0 0 1t2t1 0
0 0 45860 0
43 0 105838 0
8S 0 162937 C
123 0 218729 0
1373 0 286690 894
2498 0 338259 169B
3514 0 369062 2424
4376 0 384014 3043
5147 0 397286 35M
5923 0 415241 4148
6247 0 420121 4380
6704 0 427218 4706
7157 0 434229 5029
7603 0 441145 5348
8016 0 447551 5642
8422 0 453850 5932
8821 0 460043 6217
9213 0 466129 6497
9599 0 472109 6772
9978 0 477982 7043
103S1 0 483749 7309
10717 0 489409 7S70
11076 0 494963 7327
11429 0 500410 6079
11775 0 505749 8326
12114 0 510902 6568
12446 0 516108 B605
12772 0 52! 127 9038
13091 0 S26038 9266
13404 0 530843 94B9
13710 0 535540 9708
14009 O 540129 9921
14301 0 544812 10130
14587 0 548986 10334
14706 0 550629 10421
14744 0 5S0932 10449
14909 0 552257 10573
15064 0 553504 10690
15209 0 S54674 10799
15345 0 555768 10902
-------
Class I Switch Locomotives
Year
1909
2000
2001
2002
2003
2004
2005
2006
2007
£008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
202S
2026
2027
2026
2029
2030
2031
2032
2033
2034
2036
2036
2037
2036
2039
2040
Number of Switchers
Base Tart) Tied Tier II
4500 000
4500 000
4500 0 0,0
4250 200 50 ' 0
4000 400 100 0
3750 600 150 0
3500 800 150 50
3248 1000 ISO 102
2994 1200 150 156
2738 1400 150 212
2480 1600 ISO 270
2220 1800 150 330
1958 2000 150 392
1694 2200 150 456
1*28 2400 150 522
1160 2600 150 590
BOO 2600 ISO 660
616 3000 150 732
544 3000 150 806
468 3000 150 862
390 3000 150 960
310 3000 ISO 1040
226 3000 150 1122
144 3000 150 1206
58 3000 ISO 1292
0 2970 150 1380
0 2880 150 1470
0 2788 ISO 1562
0 2694 15G 1656
0 259S 150 1752
0 2500 150 1850
0 2350 150 1950
0 2200 ISO 2050
0 2050 ISO 2150
0 1900 150 2250
0 1750 150 2350
0 1600 ISO 2450
0 1450 160 2550
0 1300 150 2850
0 1150 150 2750
0 1000 150 2B50
0 850 1 SO 2950
Emission Factors (
-------
Class II and ill Locomotives
Year
1999
2000
2001
2002
2003
2004
2005
2006
2007
2006
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
Number of Locomotives
Base T»rO Tier! Tier (1
4200 000
4200 000
4200 0 00
4150 SO 0 0
4100 100 0 0
4050 150 0 0
4000 200 0 0
3950 250 0 0
3900 300 0 0
3650 350 0 0
3800 400 0 0
3750 450 0 0
3700 500 0 0
3550 650 0 0
3400 900 0 0
3250 950 0 0
3100 1100 0 0
2950 1250 0 0
2800 1400 0 0
2650 1550 0 0
2500 1700 0 0
2350 1850 0 0
2200 2000 0 0
2050 2160 0 0
1900 2300 0 0
1750 2450 0 0
1600 2600 0 0
1450 27SO 0 0
1300 2900 0 0
1150 3050 0 0
1000 3200 0 0
850 3350 0 0
700 3500 0 0
650 3650 0 0
400 3800 0 0
250 3950 0 0
100 4100 0 0
0 4200 0 0
0 4200 0 0
0 4200 0 0
0 4200 0 0
0 4200 0 0
Emission Factors (fj/bhp-hr)
HC CO NOx PM
0.48 1.28 13.00 0.32
0.48 128 13,00 0.32
0.40 128 13.00 0.32
046 128 12.95 0.32
0.48 1.28 1289 032
0.48 1.26 12.84 0.32
0.48 128 12.79 0.32
0.48 128 12.74 0.32
0.48 1.28 12.68 0.32
0.48 1.26 12.63 0.32
0.48 1.2B 1258 0.32
0.48 1.28 1252 0.32
048 1.2B 12.47 0.32
0.48 1.26 12.31 032
0.48 1.28 12.15 0.32
0.48 128 1199 0.32
0.48 1.28 11.83 0.32
0.40 1.28 1168 0.32
0.48 1.28 11.52 ' 0.32
0.48 1.28 11.36 0.32
0.48 1.2B 11.20 0.32
0.46 128 11.04 0.32
0.48 128 10.88 0.32
0.48 1.28 10.72 0.32
0.48 128 10.56 0.32
0.48 126 10.40 0.32
0.48 1.26 10.25 0.32
0.48 1.28 10.09 0.32
0.48 1.28 9.93 0.32
0.48 1.28 9.77 0.32
0.48 1.28 9.61 0.32
0.48 1-20 9.45 0.32
0.48 .28 329 0.32
0.48 .26 9,13 0.32
0.48 28 8.97 0.32
0.48 .28 8H! 0.32
0.48 .28 6.6B 0.32
0.48 28 6.55 0.32
0,48 28 8.55 0.32
048 .28 8.55 0.32
0.48 .26 8.55 0.32
0.48 28 8.SS 0.32
Parcanl Reduction
HC CO NOx PM
0% 0% 0% 0%
0% 0% 0% 0%
0% 0% 0% 0%
0% 0% 0% 0%
0% 0% 1% 0%
0% 0% 1% 0%
0% 0% 2% 0%
0% 0% 2% 0%
0% 0% 2% 0%
0% 0% 3% 0%
0% 0% 3% 0%
0% 0% 4% 0%
0% 0% 4% 0%
0% 0% 5% 0%
0% 0% 7% 0%
0% 0% 8% 0%
0% 0% 9% 0%
0% 0% 10% 0%
0% 0% 11% 0%
0% 0% 13% 0%
0% 0% 14% 0%
0% 0% 15% 0%
0% 0% 16% 0%
0% 0% 18% 0%
0% 0% 19% 0%
0% 0% 20% 0%
0% 0% 21% 0%
0% 0% 22% 0%
0% 0% 24% 0%
0% 0% 25% 0%
0% 0% 26% 0%
0% 0% 27% 0%
0% 0% 29% 0%
0% 0% 30% 0%
0% 0% 31% 0%
0% 0% 32% 0%
0% 0% 33% 0%
0% 0% 34% 0%
0% 0% 34% 0%
0% 0% 34% 0%
0% 0% 34% 0%
0% 0% 34% 0%
Emissions (Metric Tons/Year)
HC CO NOx PM
2147 S724 58136 1431
2147 5724 58138 1431
2147 5724 58136 1431
2147 5724 57899 1431
2147 6724 57662 1431
2147 5724 57425 1431
2147 5724 57188 1431
2147 5724 56951 1431
2147 5724 56715 1431
2147 5724 56478 1431
2147 5724 56241 1431
2147 5724 56004 1431
2147 5724 55767 1431
2147 5724 55066 1431
2147 5724 54345 1431
2147 5724 53635 1431
2147 5724 52924 1431
2147 5724 52213 1431
2147 5724 51503 1431
2147 5724 50792 1431
2147 5724 50081 1431
2147 5724 49370 1431
2147 5724 48660 1431
2147 5724 47049 1431
2147 5724 47238 1431
2147 5724 46527 1431
2147 5724 45817 1431
2147 5724 45100 1431
2147 5724 44395 1431
2147 5724 43685 1431
2147 5724 42974 1431
2147 5724 42263 1431
2147 5724 41552 1431
2147 5724 40842 1431
2147 5724 40131 1431
2147 5724 39420 1431
2147 5724 36709 1431
2147 5724 38238 1431
2147 5724 38236 1431
2147 5724 38236 1431
2147 5724 38236 1431
2147 5724 36236 1431
Reductions (Metric Tons/Year)
HC CO NOx PM
0000
0000
0 0 237 0
0 0 474 0
0 0 71 t 0
0 0 948 0
0 0 1185 0
0 0 1421 0
0 0 1658 0
0 0 1895 0
0 0 2132 0
0 0 2369 0
0 0 3060 0
0 0 3791 0
0 0 4501 0
0 0 5212 0
0 0 59?3 0
0 0 tuM 0
0 0 7344 0
0 0 8055 0
0 0 6766 0
0 0 9476 0
0 0 10187 0
0 0 10898 0
0 0 11609 0
0 0 12319 0
0 0 13030 0
0 0 13741 0
0 0 14451 0
0 0 15162 0
0 0 15673 0
0 0 16584 0
0 0 17294 0
0 0 18005 0
0 0 18716 0
0 0 19427 0
0 0 19900 0
0 0 19900 0
0 0 19900 0
0 0 19900 0
0 0 19900 0
-------
Passenger Locomotives
Year
1B99
2000
2001
2002
2003
2004
2005
2006
2007
2006
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2 038
2039
2040
Number of Locomotives
Base Tiar 0 Tier I Tier II
778 0 0 0
778 0 0 0
778 0 0 ,0
648 100 30 '0
518 200 80 0
368 300 90 0
258 400 90 30
128 500 90 60
98 500 90 90
63 500 SO 120
36 500 90 150
6 500 90 160
0 478 90 210
0 448 90 240
0 418 90 270
0 388 90 300
0 358 90 330
0 328 90 360
0 298 90 390
0 268 90 420
0 230 90 430
0 206 90 480
0 176 90 510
0 148 90 540
0 118 00 570
0 86 90 600
0 56 90 630
0 28 90 660
0 0 88 690
0 0 56 720
0 0 28 750
0 0 0 778
0 0 0 778
0 0 0 778
0 0 0 778
0 0 0 778
0 0 0 778
0 0 0 778
0 0 0 778
0 0 0 776
0 0 0 778
0 0 0 778
Emission Factors (g/Wip hf)
HC CO NOx PM
048 1.28 13.00 0.32
0.48 1.28 13.00 0.32
0.48 128 13.00 0.32
0.48 128 12.18 0.32
0.48 1.28 11.37 0.32
0.4B 1.28 10.55 0.32
0,47 t.28 9.67 0.31
0.48 .28 8.79 0.31
0.45 28 848 0.30
0.44 .28 6.17 0.30
0.44 .28 7.85 029
0.43 .28 7.54 0.28
0,42 .28 7.36 0.28
0.41 ,28 7,22 0.27
0.40 .28 7.08 026
0.38 .28 6.94 0.26
0,38 .28 6.80 0.25
0.37 ,28 8.07 0.2S
0.37 1.28 853 0.24
0.36 1.2B 6.39 0.23
0.35 1.28 8.25 073
034 1.28 8.11 0.22
0.33 1.28 5.97 0.22
0.32 1.28 5.03 0.21
0.31 128 5.09 0.20
0.31 1.28 5.56 0.20
0.30 1.28 5.42 0.19
0.29 1.28 5,28 018
0.26 1.28 S.14 0.18
0.27 1.2B 5.08 0,17
0,28 1.28 5.01 0.17
0.2G 28 4.95 0.16
0.26 .28 4.95 0.18
0.26 .28 4.95 0.16
0.26 .28 4.95 0.16
0.28 .28 4.95 0.18
0.26 .28 4.95 0.16
0.26 .28 4.95 0.16
0.26 .28 4.95 0.16
0.26 .28 4.95 0.16
0.28 .28 4.95 0.16
026 .28 4.95 0.16
Percent Reduction
HC CO NO* PM
0% 0% 0% 0%
0% 0% 0% 0%
0% 0% 0% 0%
0% 0% 6% 0%
0% 0% 13% 0%
0% 0% 19% 0%
2% 0% 26% 2%
4% 0% 32% 4%
6% 0% 35% 6%
8% 0% 37% 8%
9% 0% 40% 10%
11% 0% 42% 12%
13% 0% 43% 13%
15% 0% 44% 15%
17% 0% 48% 17%
18% 0% 47% 19%
20% 0% 48% 21%
22% 0% 49% 23%
24% 0% 50% 25%
26% 0% 51% 27%
27% 0% 52% 29%
29% 0% 53% 31%
31% 0% 54% 33%
33% 0% 55% 35%
35% 0% 56% 37%
36% 0% 57% 39%
38% 0% 58% 40%
40% 0% 59% 42%
42% 0% 60% 44%
44% 0% 61% 46%
45% 0% 61% 48%
47% 0% 62% 50%
47% 0% 62% 50%
47% 0% 62% 50%
47% 0% 62% 50%
47% 0% 62% 50%
47% 0% 52% 50%
47% 0% 52% 50%
47% 0% 62% 50%
47% 0% 62% 50%
47% 0% 62% 50%
47% 0% 62% 60%
Emissions (Metric Tons/Year)
HC CO NOx PM
1328 3541 35963 885
1328 3541 35963 BBS
1328 3541 35963 865
1327 3541 33705 865
1325 3541 31446 885
1324 3541 29187 885
1300 3541 26746 86B
1278 3541 24305 851
1252 3541 23446 834
1228 3541 22568 617
1204 3541 21729 800
1180 3541 20870 783
11 56 3541 20360 766
1132 3541 19976 749
1106 3541 19592 732
1084 3541 19208 715
1060 3541 18824 690
1030 3541 18440 600
1012 3541 18056 663
986 3541 17872 646
964 3541 17288 629
940 3541 16903 612
916 3541 16519 595
692 3541 16135 578
668 3541 1S751 561
844 3541 15367 544
620 3541 14963 527
796 3541 14599 510
772 3541 14229 493
749 3541 14048 476
727 3641 13884 459
705 3641 13694 443
705 3541 13694 443
705 3541 13694 443
705 3641 13694 443
705 3541 13694 443
705 3541 13694 443
705 3541 13694 443
705 3541 13694 443
705 3541 13694 443
705 3541 13694 443
705 3541 13694 443
Reductions (Metric Ton a/Year)
HC CO NOx PM
0
0000
0000
1 0 2259 0
3 0 4517 0
4 0 6776 0
28 0 9217 17
52 0 11656 34
76 0 12517 51
100 0 13375 68
124 0 14234 85
148 0 15093 102
172 0 15603 119
196 0 15988 137
220 0 16372 154
244 0 16756 171
268 0 17140 188
292 0 17524 205
316 0 17903 222
340 0 18292 239
364 0 18676 256
388 0 19O60 273
412 0 19444 290
436 0 19828 307
460 0 20212 324
404 0 20596 341
500 0 20960 356
532 0 21364 375
568 0 21734 393
579 0 21917 410
601 0 22099 427
622 0 22270 443
322 0 22270 443
622 0 22270 443
622 0 22270 443
622 0 22270 443
622 0 22270 443
622 0 22270 443
622 0 22270 443
622 0 22270 443
622 0 22270 443
622 0 22270 443
-------
All Locomotives
Year
1999
2000
2001
2002
2003
2004
2005
2006
2007
2006
2009
2010
2011
2012
2013
2014
2015
2018
2017
201 B
2019
2020
2021
2022
2023
2024
202S
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2036
2039
2040
Emissions (Metric Tons/Year)
HC CO NOx PM
42404 108228 1092527 26959
42fKM 106228 1081316 269S9
42404 106228 1046647 26959
42360 100228 982527 26958
42317 108228 921266 28958
42277 106228 861312 26957
40971 108228 788843 26031
39790 108228 732742 25194
38717 108220 698963 24434
37793 108228 681010 23780
36984 108228 664712 23194
38128 106228 643706 22605
35741 108228 636096 22337
35219 108228 625899 21973
34702 108228 615762 21613
34188 108228 605695 21256
33708 108228 596113 20923
33234 108228 586612 20S94
32764 106228 578390 20269
32300 108228 670250 19948
31642 108228 562191 19631
31388 106228 554213 19319
30941 108228 5483)6 19011
30498 106228 5385O1 16706
30061 108228 530768 16408
29630 108228 $23295 18110
29204 108228 516264 17818
28783 108228 509327 17530
28368 108228 502497 17247
27960 108228 495049 16967
27557 108228 489494 16692
27147 108228 482995 16415
26783 108228 476755 16158
26385 108228 470603 15906
26011 108228 464S39 15655
25642 106228 458562 15410
25437 108228 455295 15281
25312 108228 453584 15209
25058 108228 451301 15041
24811 108228 449072 14878
24571 108228 446696 14722
24339 108228 444771 14S72
Reductions {Metric Tons/Year)
HC CO NOx PM
0000
0 0 11211 0
0 0 4 56 BO 0
44 0 110000 0
87 0 171261 1
127 0 231215 1
1433 0 303684 92B
26)4 0 359785 1765
3687 0 393564 2S2S
4611 0 411517 3178
5440 0 427315 3764
6277 0 448821 4354
6663 0 456423 4622
7185 0 466626 4986
7703 0 476765 5346
8216 0 486832 5703
8696 0 496414 6036
9171 0 &0&9IS 6365
9640 0 514137 6690
10104 0 522277 7011
10562 0 530336 7327
11016 0 538314 7640
11463 0 546211 7948
11906 0 554026 8252
12343 0 S61759 8553
12774 0 569231 8849
13200 0 576263 914 t
13621 0 560200 9428
14036 0 590030 9712
14444 0 S96578 9992
14847 0 603033 10267
15257 Q 609532 10544
15641 0 615772 10801
16019 0 621924 11054
18393 0 627988 11303
16762 0 633965 11549
16967 0 637232 11678
17092 0 638943 11750
17346 0 641226 U918
17593 0 643455 12080
17B33 0 645631 12236
180J5 0 647756 12387
Percent Reductions (Weuic Tons/Year)
HC CO NO* PM
0.0% 00% 1.0% 0.0%
0,0% 00% 4.2% 0.0%
0.1% 0.0% 10.1% 0.0%
0.2% 00% 15.7% 00%
0.3% 0.0% 21.2% 0.0%
3.4% 0.0% 27.8% 3,4%
6.2% 0.0% 32.9% 6.S%
8.7% 0.0% 36.0% 9.4%
10.9% 0.0% 37.7% 11.8%
12.8% 0.0% 39.2% 14.0%
14.8% 0.0% 41.1% 16.1%
15.7% 0.0% 41.8% 17.1%
16.9% 00% 42.7% 18.5%
18.2% 0.0% 43,6% 19.8%
19.4% 0.0% 44.6% 21.2%
20.5% 0.0% 45.4% 224%
21.8% 00% 463% 23.6%
22.7% 0.0% 47.1% 24,8%
23,8% 0.0% 47.8% 26.0%
24.9% 0.0% 48.5% 27.2%
28.0% 0.0% 493% 28.3%
27.0% 0.0% 50.0% 29,5%
28.1% 0.0% 50.7% 30.6%
29.1% 0.0% 51.4% 31.7%
30.1% 0.0% 521% 32,8%
31.1% 0.0% 52.7% 33.9%
32.1% 0.0% 53.4% 35.0%
33.1% 0.0% 54.0% 36.0%
34.1% 0.0% 54.6% 37.1%
35.0% 0.0% 55.2% 38.1%
36.0% 0.0% 55.8% 39.1%
36.9% 0.0% 56.4% 40.1%
37.8% 0.0% 569% 41.0%
38.7% 0.0% 57,5% 41,9%
39.5% 00% 58.0% 42.8%
40.0% 0.0% 58.3% 43.3%
40.3% 0.0% 58,5% 43,6%
409% 0,0% 58,7% 44.2%
41.5% 0.0% 58.9% 44.8%
421% 0.0% 591% 45.4%
42.6% 00% 593% 45.9%
-------
Fleet Average Emission Factors
For AJI Locomotives
Year
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
HC
0.52
0.52
0.52
0.52
0.52
0.51
0.50
0.48
0.47
1 0.46
0.45
0.44
0.44
: 0.43
0.42
i 0.42
: 0.41
: 0.40
'! 0.40
' 0.39
•'•. 0.39
0.38
' 0.38
; 0.37
0.37
0.36
' 0.36
: 0.35
•i 0.35
0.34
0.34
0.33
0.33
0.32
0.32
; 0.31
: xr.31
i 0.31
' 0.31
0.30
0.30
0.30
(g/bhp-
CO
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1,32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
' 32
hr)
NOx
13.30
13.16
12.74
11.96
11.22
10.49
9.60
8.92
8.51
8.29
8.09
7.84
7.74
7.62
7. SO
7.37
7.26
7.14
7.04
6.94
6.84
6.75
6.65
6.56
6.48
6.37
6.29
6.20
6.12
6.04
5.96
5.88
5.80
5.73
5.66
5.58
5.54
5.52
5.49
5.47
5.44
5.41
PM
0.33
0.33
0.33
0.33
0.33
0.33
0.32
0.31
0.30
0.29
0.28
0.28
0.27
0.27
0.26
0.26
0.25
0.25
0.25
0.24
0.24
0.24
0.23
0.23
0.22
0.22
0.22
0.21
0.21
0.21
0.20
0.20
0.20
0.19
0.19
0.19
0.19
0.19
0.18
0.18
0.18
0.18
HC
10.7
10.7
10.7
10.7
10.7
10.7
10.4
10.1
9.8
9.6
9.4
9.1
9.1
8.9
8.8
8.7
8.5
8.4
8.3
8.2
8.1
7.9
7.8
7.7
7.6
7.5
7.4
7.3
7.2
7.1
7.0
6.9
6.8
6.7
6.8
6.5
6.4
6.4
6.3
6.3
6.2
6.2
CO
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
(9/gai)
NOx
276.7
273.8
265.0
248.8
233.3
218.1
199.8
185.6
177.0
172.5
168.3
163.0
161.1
158.5
155.9
153.4
151.0
148.5
146.5
144.4
142.4
140.3
138.3
136.4
134.4
132.5
130.7
129.0
127.2
125.6
124.0
122.3
120.7
119.2
117.6
118.1
115.3
114.9
114.3
113.7
113.2
112.6
PM
6.8
6.8
6.8
6.8
6.8
6.8
6.6
6.4
6.2
6.0
5.9
5.7
5.7
5.6
5.5
5.4
5.3
5.2 |
5,1 !
5.1 ',
5.0
4.9
4.8
4.7 :
4.7 !
4.6 !
4.5 I
4.4 !
4.4 ;
4.3
4.2 '
4.2 i
4.1 •
4.0 !
4.0 i
3.9 |
3.9 i
3.9 !
3.8
3.8
3.7
3.7 ;
1-9
-------
This table shows the projected marginal NOx benefits of the Tier 0, Tier 1, and
Tier 2 standards in the columns labeled "Tier 0", "Tier 1", and "Tier 2", respectively.
The column labeled "All Tiers" shows the total projected NOx benefits of the entire
program (as finalized) consistent with Chapter 6. The column labeled "Tiers O&l"
shows the total projected NOx benefits of the program without the Tier 2 standards
(i.e., assuming that the Tier 1 standards continue to apply to newly manufactured
locomotives after 2004). The marginal benefits of the Tier 1 and Tier 2 standards are
calculated from the first three columns.
vlai-glnal NOx :E
benefit Analysis
(Metric Tons)
Year
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
Total
All Tiers
0
11211
45880
110000
171261
231215
303684
359785
393564
411517
427815
448821
456429
466628
476765
486832
496414
505915
514137
522277
530336
538314
546211
554026
561759
569231
576263
583200
590030
596578
603033
609532
615772
621924
627988
633965
637232
638943
641226
643455
645631
647756
200525S2
Tiers 0 41
0
11211
45880
110000
171261
231215
293785
340943
366605
377574
387614
402324
407066
41 3384
41 9673
425928
431 954
437943
442697
447413
452092
456734
461 338
465905
470435
474748
478663
482530
486333
489899
493415
496961
500460
503912
S07318
51 0678
51 2575
51 3527
514018
51 4520
51 5035
51 5563
16981126
TierO
0
10908
44631
100363
153538
205959
257207
294403
311455
315551
319427
327526
32941 7
331666
333928
336200
338465
340781
341923
343077
344242
345419
346608
347809
349021
350069
350780
351491
352202
35291 2
353623
3S4334
355044
355755
356466
3571 77
357887
358361
358361
358361
358361
358361
12809089
Tierl
0
305
1249
9637
17723
25256
36578
46540
55149
62022
68187
74798
77650
81718
85745
89728
93469
97162
100773
104336
107849
111314
114730
118097
121414
124678
127883
131039
134131
138986
139792
142627
145415
148157
150852
153502
154688
155166
155656
156159
156674
157202
4172037
Tier 2
0
0
0
0
0
0
9899
18842
26959
33943
40201
46497
49363
53244
57092
60904
64460
67972
71440
74864
78244
81580
84872
88120
91324
94484
97599
1 00670
1 03697
1 06680
109618
1 1 2571
115312
118012
1 20670
1 23287
1 24657
125416
1 27208
1 28935
1 30596
132193
3071426
1-10
-------
APPENDIX J
Terms and Abbreviations Used in the Rulemaking
J-l
-------
Terms Used in the Rulemaking
Aftercooling - (see "charge air cooling")
Baseline - relating to uncontrolled locomotives.
Charge air cooling - an engine technology (for turbocharged or supercharged engines) that
reduces NOx emissions by lowering combustion temperatures. Can also improve power output.
Class I railroads - the largest railroads in the U.S., based on annual revenue, as defined by the
U.S. Surface Transportation Board in 49 CFR Chapter X. Class I railroads comprise about 90
percent of the railroad industry.
Class II and III railroads - small railroads, as defined by the U.S. Surface Transportation Board
in 49 CFR Chapter X.
Commuter railroad - a passenger railroad that operates within a single metropolitan area,
using diesel-powered locomotives.
Compliance margin - the degree to which a locomotive's emissions are below the applicable
standard. Compliance margins are expressed as percent of the standard, so that a ten percent
compliance margin for the Tier 0 NOx standard (9.5 g/bhp-hr) would be 0.95 g/bhp-hr.
Consist - a series of two or more locomotives pulling the same train. Railroads use more than
one locomotive in a consist when they need more power to pull the train than can be supplied
by a single locomotive. The total amount of power need is determined by the total weight of the
train and the steepest grade that it must climb.
Diesel-electric locomotive - the standard type of locomotive in the U.S.,it is a locomotive that
uses a diesel engine to power electrical traction motors connected to the wheels.
Duty-cycle - a description of the amount of time a typical locomotive spends in each throttle
notch, expressed as percent of total time in use.
Dynamic brake - a means of slowing a train by using the traction motors as generators,
effectively converting the momentum of the train to electrical energy which is dissipated as
heat.
Emission inventory - an emission total for a given pollutant and a given class of source, usually
expressed as tons of emissions per year.
Four-stroke - relating to a type of engine that uses four piston strokes per combustion event.
Freshly manufactured - newly manufactured (and not yet remanufactured).
J-2
-------
Grams per brake horsepower hour - the ratio of the mass of emissions from an engine to the
amount of power produced by the engine at the same time. Use of the word "brake" in this
context indicates that the power includes power supplied to accessories, in addition to the power
available for propulsion. One gram per brake horsepower-hour is equal 0.7457 grams per
kilowatt-hour.
Hotel power - electrical power supplied by the engine for use in passenger cars.
Injection timing - the time at which fuel is injected into the engine for combustion. Retarding
the timing of a diesel engine (i.e., delaying the point at which the fuel is injected into the
cylinder) reduces NOx emissions, but can increase PM emissions, smoke, and fuel consumption.
Advancing the timing (i.e, injecting the fuel earlier) can have the opposite effects.
Insular railroad - an industrial facility that uses locomotives to move rail cars short distances
on its own property. Locomotives used in this way usually have very low power ratings, and
use very little fuel each year.
Line-haul - relating to the movement of trains across reasonably long distances. Most railroad
operations are line-haul operations.
Local railroad - a railroad that operates within a very limited geographic range. Local railroad
is roughly equivalent to Class III railroad.
Lugging - reducing the speed of an engine by increasing the engine load.
Megawatt-hour - unit of work equivalent to the total work perform in one hour at a constant
power rate of one megawatt. One megawatt-hour is equal to 1341 horsepower-hours.
Net present value - an economic term used to account for the time value of money.
Notch - (see "throttle notch")
Particulates or particulate matter - very small solid particles emitted by engines and other
sources. Particulates are formed by incomplete combustion, and are often related to visible
smoke. Particulates from diesel engines are typically a few microns in diameter, or smaller.
Power assembly - a cylinder of an engine and related components that are detachable from the
engine as a single system.
Rated horsepower - the maximum power out of a locomotive engine.
Regional railroad - a railroad that operates within a moderate geographic range. Regional
railroad is roughly equivalent to Class II railroad.
Remanufacture - to thoroughly overhaul (or remanufacture) an engine, such that it is
functionally equivalent to its original condition (or better).
Roots-blower - a mechanical blower that is used to force air into an engine for combustion. A
roots-blower operates at much lower pressures than turbochargers and superchargers.
J-3
-------
Service life - the entire period during which a locomotive is in service, from the time it is
manufactured, until it is scrapped. The service life of a locomotive is typically about 40 years.
Switch - relating to movement of railroad cars over short distances, usually within a switch
yard. Most switch locomotives have rated power of less than 2000 horsepower.
Throttle notch - a discrete power setting of a locomotive throttle. Most locomotives have eight
throttle notches for propulsion, plus notches for dynamic brake and idle.
Train - a series of rail cars and locomotives. Individual locomotives are not trains.
Truck - the part of a locomotive that contains the traction motors, axles and wheels. (Note: the
term truck is also used in this document to refer to large highway vehicles.)
Turbocharger - a turbine device that uses energy from exhaust gases to compress intake air.
Two-stroke - relating to a type of engine that uses two piston strokes per combustion event.
Upgrade - the process of converting an uncontrolled locomotive that was built before 1973 (and
therefore not subject to these regulations) into a locomotive that complies with the Tier 0
standards. Upgrading is optional.
Useful life - the period (expressed as MW-hrs of work performed by the engine) during which
a locomotive is designed to be properly functioning with respect to power out, reliability and
fuel consumption. These regulations require that locomotives also comply with emission
standards during this period. A typical useful life period is about six years.
Yard - (see "switch")
J-4
-------
Abbreviations Used in the Rulemaking
AAR - Association of American Railroads
ABT - averaging, banking and trading
ASLRA - American Short Line Railroad Association
ATSF - Atchison, Topeka, and Santa Fe Railway
BN - Burlington Northern Railroad
BNSF - Burlington Northern Santa Fe Railway (formerly BN and ATSF)
CFR - Code of Federal Regulations
CSX - CSX Transportation
DOE - U.S. Department of Energy
EF&EE - Engine, Fuel, and Emission Engineering, Inc.
EGR - exhaust gas recirculation
EMA - Engine Manufacturers Association
EMD (or EMD-GM) - Electro-Motive Division of General Motors
EPA - Environmental Protection Agency
FPI - first price increase
FRA - Federal Railroad Administration
FTP - Federal Test Procedure
GE or GETS - General Electric Transportation Systems
g/bhp-hr - grams per brake horsepower hour
HC - hydrocarbons
HFID - heated flame ionization detector
HP - horsepower
kW - kilowatt
MW-hr - megawatt-hour
OEM - original equipment manufacturer
NDIR - nondispersive infrared detector
NOx - oxides of nitrogen
NPV - net present value
PLT - production line testing
PM - particulate matter
PM-10 - particulate matter in the size range of 0 to 10 microns
J-5
-------
RSD - Regulatory Support Document
SAE - Society of Automotive Engineers
SCR - selective catalytic reduction
SF - (see ATSF)
SOx - oxides of sulfur
STB - Surface Transportation Board
SwRI - Southwest Research Institute
UP - Union Pacific Railroad
VGT - variable geometry turbocharger
J-6
-------
APPENDIX K
Calculation of Weighting Factors for ABT Credits
The ABT program requires that locomotives certified to an FEL other than the
applicable standard be recertified to that FEL at all subsequent remanufactures. The
result of this requirement is that credit calculations are based on the total emissions
of that locomotive for its remaining service life. Thus, in order to allow ABT credits
generated by remanufactured locomotives to be used by freshly manufactured
locomotives (and vice versa), it is necessary to prorate emission credits. For simplicity,
the prorating factors are assumed to be a function of locomotive age. These factors,
which are shown in the table, are the estimated fraction of the service life that is
remaining for a locomotive.
These factors were calculated assuming that a typical locomotive remains in
service for 40 years, and is remanufactured 6 times. This means that a typical
locomotive will experience 7 useful lives during its service life. Due to the fact that a
locomotive's usage rate (MW-hrs per year) typically declines with age, the
remanufacturing interval in terms of years is expected to change with locomotive age.
For this analysis, developed an assumed remanufacture schedule, based on a typical
locomotive. During the first 12 years, a locomotive is assumed to be remanufactured
every 4 years (after 4, 8, and 12 years). During the next 12 years, a locomotive is
assumed to be remanufactured every 6 years (at 18 and 24 years). A locomotive is
assumed to be remanufactured once more at 32 years, and scrapped at 40 years. Each
one of these points represents 1/7 of the locomotive's service life. For example, at 12
years, which is assumed to be the point at which the locomotive needs to be
remanufactured for the third time, a locomotive is assumed to have expended 3/7 of its
service life and have 4/7 (0.571) of its service life remaining. Any locomotive that is 32
or more years old is assumed to be in its final useful life, and is therefore assumed to
have 1/7 (0.143) of its service life remaining at the point of its remanufacture.
K-l
-------
Determination of Fractional Service Life Remaining (F) for ABT
Calculations
Age
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
F
0.964
0.929
0.893
0.857
0.821
0.786
0.750
0.714
0.679
0.643
0.607
0.571
0.548
0.524
0.500
0.476
Age
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
F
0.452
0.429
0.405
0.381
0.357
0.333
0.310
0.286
0.268
0.250
0.232
0.214
0.196
0.179
0.161
0.143
K-2
-------
APPENDIX L
Exclusion of Pre-1973 Locomotives
Locomotives originally manufactured prior to 1973 are excluded from the
regulations, unless these units undergo certain modifications resulting in post-1972
configurations. EPA is not including pre-1973 locomotives in the regulations for the
following reasons: First, the number of locomotives in these subgroups is small and
utilization of individual locomotives is generally low. As a result, contributions to the
national emission inventory are low. Second, technologies for reducing emissions from
these "old" units either do not exist in a form that could be applied to the locomotives,
or the cost of applying available technology would greatly exceed the value of the
locomotive.
In developing a description for "old" locomotives which would not be included
in the final rule, EPA sought to identify points either in engine design or in the
sourcing of components which could lead to substantial technical problems or high
costs in achieving compliance. However, to achieve the greatest benefit from the
regulations, EPA sought to identify a date which would allow for the inclusion of a
significant portion of the national locomotive fleet at the time that standards became
effective. To avoid the potential for establishing unequal burdens between railroads
as a function of the original manufacturer of the locomotives in their fleets, EPA
decided to use a single dividing line for all locomotives. As a result of reviews of dates
at which either design changes or the sourcing for major parts occurred, EPA identified
the following: EMD introduced its 645 series of engines in 1966, and continued this
series into the 1980s. GE changed design and sourcing for turbochargers, fuel injection
pumps, and fuel injectors at the start of 1973. In mid-1983 GE again changed sourcing
for fuel injection pumps, and in mid 1986 again changed sourcing for fuel injectors.
The results of the review lead EPA to select January 1, 1973 as the date of original
manufacture for separation between included and excluded locomotives. Since this
date precedes the effective date of the standards by 27 years, a significant fraction of
the existing fleet of locomotives will be included in the regulation. Pre-1973 engines
still in operation by the year 2000 will be almost exclusively used by Class II and III
railroads or for switching operations. They will therefore have relatively low usage
and emissions rates when expressed as grams per year. Since benefits from emission
control could be low, expressed as an annual mass of emissions, and the costs of control
high, exclusion of these locomotives from the regulation appears to be appropriate.
L-l
-------
APPENDIX M
NOx Concentrations as a Function of Test Sequence
This appendix contains continuous traces of NOx emissions provided by AAR,
for locomotives manufactured by both GE and EMD, measured under different test
sequences. Starting with the engine at idle, NOx concentration in the exhaust was
measured continuously as engine power was increased to full power and returned to
idle.
M-l
-------
Figure M-l
1000
900'
ja.
o
1
I
o
Q
SP Unit 2706 'Transient' NOx
EMD 1S-84S Roafc itowi Enaln* 1SOO HP
SP 8706 - EMD MP13 Sw«Ch«r LacnmoHw
100 '200
Sv»fti
3CX3 400 EGO 600 TOO
Test Time (seconds)
eoo
900
1000
M-2
-------
2500
Figure M-2
Amtrak Unit 514 "Transient1 NOx
300
1000 1500 2000
Test Time (Sec)
2500
3000
Swflt
M-2
-------
Figure M-3
Amtrak Unit 229 Transient NOx
1BDO
1*00
1BJQ- —
1000- •—
o
x<
BUD 1fr*«£38 3.300 HP »nfl*r»
soo
1000 1500 2000
Time (sec)
JSOQ
3000
M-4
-------
Figure M-4
Amtrak Unit 806 Transienf NOx
2500
2000
1500'
1000
500
Frtigh*Mo
-------
APPENDIX N
1995 Emissions Inventory Data
TABLE N-l
Pollutant
HC
CO
NOx
PM-10
Emissions*
(Thousand Metric Tons
Total
20786
83726
19799
38760
Mobile Source
7596
67496
9637
634
per Year)
Locomotives
42
108
1093
27
Percent of
Total
0.20%
0.13%
5.52%
0.07%
Percent of
Mobile Source
0.56%
0.16%
11.34%
4.25%
* Total and Mobile Source emissions from 1996 "National Air Pollutant
Emission Trends, 1900-1995" EPA-454/R-96-007; Locomotive emissions
from this document.
N-l
-------
APPENDIX O
Corrections to Environmental Analysis
This appendix contains corrections to the environmental impacts analysis. The
corrections are described below, followed by corrected versions of the "Paisenger
Locomotives", "Class I Lane-Haul Locomotives", "All Locomotives", and "Fleet Average
Emission Factors" tables of Appendix I. These corrections have a minimal effect on the
projection of total benefits, and are presented here only in* the purpose of
completeness. The tables found in Chapter 6 and Appendix I are unchanged from the
December 1997 version of this document.
Correction #1 - Delay of Tier 0 standards for passenger locomotives
The original passenger table showed the Tier 0 standards taking effect in 2002,
instead of 2007 as was specified in the regulations. This correction has an effect on
emission projections only for years 2002-2010.
Correction #2 - Line-haul ^manufacturing and retirement schedule
The original Class I line-haul table was not consistent with the description of
the assumptions made in the text regarding the remanufacturing and retirement of
locomotives. The table on the next page shows the correct schedule'in greater detail
than was presented in the December 1997 version of this document.
O-l
-------
Summary of Corrected Remanufacturing and Retirement Schedule During the Period
2000-2010
Year
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
Uncontrolled
Locomotives
RefYiani ifincta IT
edintolferO
Configuration
500
1600
2250
2250
2550
2350
1750
650
Uncontrolled
Locomotives
txmt are
Retired from
Service
100
100
400
400
400
400
400
400
300
100
Newly
Manufactured
Uncontrolled
Locomotives
300
100
Newly
Manufactured
TSerO
Locomotives
100
300
,
Total Number
of Uncontrolled
Locomotives
16500
16200
14600
11950
9300
6350
3600
1450
400
100
0
0
Total Number
of Tier 0
Locomotives
0
600
2500
4750
7000
9550
11900
13650
14300
14300
14300
14300
Notes: This table accounts for the projection of 300 pre-1973 locomotives being
remanufactured into Tier 0 configurations by adding 50 locomotives to the
second column for years 2002-2007.
This table properly accounts for the new production of uncontrolled locomotives
in 2000 and 2001, and their first remanufacture into complying configurations
in 2004 and 2005.
This table accounts for all 3000 existing locomotives that are projected to be
retired from service: 100 pre-1973 locomotives per year during 2000-2009 (as
part of a normal retirement schedule); 300 later locomotives per year during
2002-2007; and 200 later locomotives in 2008.
0-2
-------
Passenger Locomotives - Corrected
Year
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2018
2017
2016
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2036
2039
2040
Number of Locomotives
Base Tier 0 Tier I Tier II
778 0 0 0
778 0 0 0
770 0 0 ; 0
746 0 30 0
718 0 60 0
fiBB 0 90 0
658 0 90 30
628 0 90 60
498 100 90 90
366 200 90 120
238 300 90 150
106 400 90 160
0 478 90 210
0 448 90 240
0 418 90 270
0 388 90 300
0 358 90 330
0 328 90 360
0 298 90 390
0 268 90 420
0 238 90 450
0 206 90 480
0 178 90 510
0 148 90 540
0 118 90 570
0 88 90 600
0 58 90 630
0 28 90 660
0 0 88 890
0 0 58 720
0 0 28 750
0 0 0 778
0 0 0 776
0 0 0 778
0 0 0 778
0 0 0 778
0 0 0 778
0 0 0 778
0 0 0 778
0 0 0 778
0 0 0 778
.0 00 .778
Emission Factors (g/bhp-hr)
HC CO NOx PM
0.48 1.28 13.00 0.32
0.48 1.2B 13.00 0.32
0.40 1.28 13.00 0.32
0.43 1.28 12.76 0,32
0.46 1.28 12.51 0,32
0.48 1.28 12.27 0.32
0.47 1.28 11.96 0.31
0.46 1.28 11.65 0.31
0.45 1.28 10.76 0.30
0.44 1.26 9.68 0.30
0.44 1,28 9.00 0.29
0.43 1.28 8.12 0.28
0.42 1.28 7.36 0.28
0.41 1.26 7.22 0.27
0.40 1.26 7.08 0.26
0.39 1.28 6.94 0.26
0.38 1.28 6.80 0.25
0,37 1,28 6.67 0.25
0,37 1.28 6,53 0.24
0.36 1.28 6.39 0.23
035 1.28 6.2S 0.23
0.34 1.28 6.11 0.22
0.33 1.28 5.97 0.22
0.32 1.28 5.83 0.21
0.31 1.28 5.69 0.20
0,31 1.28 5.56 0.20
0.30 1.28 '5.42 0.19
0.29 1.28 5.28 0.18
0.28 1.26 5.14 0.18
0.27 1.28 5.08 0.17
0.26 1.28 5.01 0.17
0.28 1.28 4.95 0.16
0.26 1.26 4.95 0.16
0.26 1.28 4.96 0.16
0.26 1.28 4.95 0,16
0.26 1.28 4.95 0.16
0.26 1,28 4.95 0.16
0.26 1.28 4.95 0.16
0.26 1.28 4.95 0.16
0.26 1.28 4.95 0.16
0.26 1.28 4.95 0.16
0.26 1.26 4.95 0.16
Percent Reduction
HC CO NOx PM
0% 0% 0% 0%
0% 0% 0% 0%
0% 0% 0% 0%
0% 0% 2% 0%
0% 0% 4% 0%
0% 0% 6% 0%
2% 0% 8% 2%
4% 0% 10% 4%
6% 0% 17% 6%
6% 0% 24% 8%
9% 0% 31% 10%
11% 0% 38% 12%
13% 0% 43% 13%
15% 0% 44% 15%
17% 0% 46% 17%
18% 0% 47% 19%
20% 0% 48% 21%
22% 0% 49% 23%
24% 0% 50% 25%
26% 0% 51% 27%
27% 0% 52% 29%
28% 0% 53% 31%
31% 0% 54% 33%
33% 0% 55% 35%
35% 0% S6% 37%
36% 0% 57% 39%
36% 0% 58% 40%
40% 0% 59% 42%
42% 0% 60% 44%
44% 0% 61% 46%
45% 0% 61% 48%
47% 0% 62% 50%
47% 0% 62% 50%
47% 0% 62% 50%
47% 0% 62% 50%
47% 0% 62% 50%
47% 0% 62% 50%
47% 0% 62% 50%
47% 0% 62% 50%
47% 0% 62% 50%
47% 0% 62% 50%
47% 0% 62% 50%
Emissions (Metric Tons/Year)
HC CO NOx PM
1328 3541 35963 885
1328 3541 35963 885
1328 3541 35963 885
1327 3541 35287 885
1325 3541 34611 885
1324 3541 33934 685
1300 3541 33076 668
1276 3541 32217 851
1252 3541 29776 834
1228 3541 27335 817
1204 3541 24894 800
1180 3541 22453 783
1156 3541 20360 766
1132 3541 19976 749
1108 3541 19592 732
1084 3541 1920S 715
1060 3541 18824 698
1036 3541 18440 680
1012 3541 18056 663
988 3541 17672 646
964 3541 17288 629
940 3541 16903 612
916 3541 16519 595
892 3541 16135 578
868 3541 15751 561
844 3541 15367 544
820 3541 14983 527
796 3541 14599 510
772 3541 14229 493
749 3541 14046 476
727 3541 13864 459
705 3541 136S4 443
705 3541 13694 443
705 3541 13394 443
705 3541 13694 443
705 3541 13694 443
705 3541 13694 443
705 3541 13694 443
Reductions (Metric TonsA ear) i|
HC CO NOx PM [
000 0 ,
000 0
1 0 676 0
3 0 1353 0
4 0 2029 0
28 0 2888 17
52 0 3746 34
76 0 6187 51
100 0 8628 68
124 0 11070 65
148 0 13511 102
172 0 15603 119
196 0 15988 137
220 0 16372 154
244 0 16756 171
268 0 17140 188
292 0 17524 205
316 0 17908 222
340 0 18292 239
364 0 18676 256
388 0 19060 273
412 0 19444 290
436 0 19828 307
460 0 20212 324
484 0 20596 341
508 0 20980 358
532 0 21364 375
556 0 21734 393
579 0 21917 410
601 0 22099 427
622 0 22270 443
622 0 22270 443
622 0 22270 443
622 0 22270 443
622 0 22270 443
622 0 22270 443
622 0 22270 443
705 3541 13694 443 j 622 0 22270 443
705 3541 13694 443: 622 0 22270 443
705 3541 13694 443
705 3541 13694 443
622 0 22270 443
622 0 22270 443 |
-------
Class I Une-HaJ Locomotives - Cdirected
!
4 Year
1998
2000
2001
2002
,2003
.2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2018
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2028
2000
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
NiBflbet ol Locomotives
Base Tier 0 Tier I Tier N
16500 000
16200 600 0 0
14600 £500 0 0
11950 4750 400 0
9300 7000 800 0
6350 9550 1200 0
3600 11900 1200 600
1450 13650 1200 1200
400 14300 1200 1800
100 14300 1200 2400
0 14300 1200 3000
0 14300 1200 3600
0 14300 1200 3900
0 13700 1200 4200
0 13100 1200 4500
0 12500 1200 4800
0 11900 1200 6100
0 11300 1200 5400
0 10700 1200 5700
0 10100 1200 6000
0 9500 1200 6300
0 8900 1200 6600
0 6300 1200 6900
0 7700 1200 7200
0 7100 1200 7SOO
0 8500 1200 7800
0 5900 1200 8100
0 5300 1200 8400
0 4700 1200 8700
0 4100 1200 8000
0 3500 1200 8300
0 2900 1200 9600
0 2300 1200 9900
0 1700 1200 10200
0 1100 1200 10500
0 SOO 1200 10800
0 0 1200 11100
0 0 1200 11400
0 0 1000 11700
0 0 BOO 12000
0 0 600 12300
0 0 400 12600
Fuel Usage
Biusa TierO Tier I Tier II
100% Off, 0% 0%
96% 4% 0% 0%
65% 15% 0% 0%
68% 27% 5% 0%
52% 39% 10% 0%
34% 52% 14% 0%
18% 60% 13% 8%
7% 65% 13% 15%
2% 64% 12% 22%
0% 80% 12% 28%
0% 56% 11% 33%
0% 53% 10% 37%
0% 51% 10% 38%
0% 48% 10% 42%
0% 45% 10% 46%
0% 43% 9% 48%
0% 40% 0% 50%
0% 38% 9% 53%
0% 36% 9% 56%
0% 11% 8% 56%
D% ;„ 9% 80%
0% fj% 8% 83%
0% 26% 6% 65%
0% 24% 8% 88%
0% 22% 8% 70%
0% 20% 8% 73%
0% 18% 8% 74%
0% 18% 7% 77%
0% 14% 7% 79%
0% 12% 7% 81%
0% 10% 7% 83%
0% 8% 7% 85%
0% 7% 7% 87%
0% 5% 7% 89%
0% 3% 6% 91%
0% 1% 6% 92%
0% 0% 6% 94%
0% 0% 6% 94%
0% 0% 5% 95%
0% 0% 4% 96%
0% 0% 3% 97%
0% 0% 2% 22%.
Emission Factors (gft>hp-hf)
HC CO NOx PM
0.48 1.28 13.00 0.32
0.48 1.28 12.B4 0.32
0.48 1.28 12.35 0.32
0.48 1.28 11,49 0.32
0.48 1.28 10.66 0.32
0.48 1.28 9.81 0.32
0.46 128 8.02 0.31
0.44 1.28 8.08 0.30
0.43 .28 7.61 0.29
0.42 .28 7.35 0.28
0.41 .28 7,17 0,27
0.40 .28 7.02 0.26
0.39 .28 6.95 0.26
0.38 .28 6.85 0,25
0.36 .28 6.75 0.25
0-37 .28 6.65 0.24
0.37 .28 6.56 0.24
0.36 .28 6.47 0.24
0.35 .28 6.36 0.23
0.35 .28 6.29 0.23
0.34 .28 6,21 0.22
0.34 .28 6.12 0.22
0.33 .28 6.04 0.22
0.33 .28 5.96 0,21
0.32 .28 5.88 0.21
0.32 1.28 5.BO 0.20
031 1.28 5.73 0.20
0.31 1.28 5.65 0.20
0.30 1.28 5.58 0.19
0.30 1.28 5.51 019
0.28 1.28 5.43 0.19
0.29 1.28 5,37 0.18
0.28 1.28 5,30 0.18
0.28 1.28 5,23 0.18
0.28 128 517 0.17
0.27 1.28 5.10 0.17
0.27 1.28 5.05 0.17
0.27 1.28 5.05 0.17
0,26 1.26 5.03 0.17
0.26 128 5.01 017
026 1.28 499 016
026 128 498. 016
Percent Reduction
HC CO NOx PM
0% 0% 0% 0%
0% 0% 1% 0%
0% 0% 5% 0%
0% 0% 12% 0%
0% 0% 18% 0%
0% 0% 25% 0%
4% 0% 32% 4%
8% 0% 38% 8%
11% 0% 41% 11%
13% 0% 43% 14%
16% 0% 45% 16%
18% 0% 46% 19%
19% 0% 47% 20%
20% 0% 47% 21%
21% 0% 48% 22%
23% 0% 49% 24%
24% 0% 50% 25%
25% 0% 50% 26%
26% 0% 51% 28%
27% 0% 52% 29%
29% 0% . 52% 30%
30% 0% 53% 31%
31% 0% 54% 33%
32% 0% 54% 34%
33% 0% 55% 35%
34% 0% 55% 36%
35% 0% 58% 37%
36% 0% 57% 36%
37% 0% 57% 39%
36% 0% 58% 40%
39% 0% 58% 41%
40% 0% 59% 42%
41% 0% 59% 43%
42% 0% 60% 44%
43% 0% 60% 45%
44% 0% 61% 46%
44% 0% 81% 47%
44% 0% 61% 47%
45% 0% 61% 48%
45% 0% 61% 48%
46% 0% 62% 49%
46% 0% 62% 49%
Emissions (Metric Tons/Year)
HC CO NOx PM
33256 88683 B006B2 22171
33256 88663 689671 22171
33256 88683 855607 22171
33214 88683 795775 22171
33173 88683 736746 22171
33133 86683 679457 22171
31883 88683 610869 21277
30758 66663 558752 20473
29742 88663 527466 19747
28341 88683 509522 19102
28068 88683 498610 18548
27399 88683 486624 1BO69
27076 88683 481755 17838
26822 88683 474718 17514
26174 66683 467764 17194
25731 88663 460901 16877
25322 B8683 454547 16566
24919 88683 448297 16296
24523 88683 442150 16016
24133 88683 436108 15738
23750 88683 430170 15464
23374 88683 424336 15195
23004 88683 418606 14931
22640 88683 412361 14672
22283 88683 40/460 14417
21933 88683 402044 14167
21589 88683 396733 13921
21251 88683 391526 13680
20920 88683 366424 13444
20596 88683 381428 13212
20276 88683 376537 12985
19967 88683 371750 12763
19683 88683 367070 12545
19365 88683 362494 12332
19073 88683 358024 12124
16788 68663 353660 11921
18550 88683 350053 11750
18512 08683 349750 11721
1B347 88683 348425 11597
18192 88683 347178 11461
tB047 88683 346008 11371
17911 88683 344914 11269
Reduaions (Melnc Tons/Year)
HC CO NOx PM
0 0 11011 0
0 0 45075 0
42 0 104907 0
83 0 161936 0
123 0 221225 0 |l
1373 0 269813 H94 j|
2498 0 341930 1698 j|
3514 0 373214 2424
4415 0 391160 3069
5188 0 404072 3622
5857 0 414058 4102
6180 0 418827 4333
6634 0 425964 4657
7082 0 432918 4977
7525 0 439781 5293
7934 0 446135 5585 I
B 337 0 452385 5872
8733 0 458532 6155
9123 0 464574 6433
9506 0 470512 6706
9882 0 476346 6975 .
10252 0 482076 7239
10616 0 487701 7499
10973 0 493222 7754
11323 0 498638 8004
11 667 0 503950 8250
12005 0 5091 56 8491
J2336 0 514258 8727
12660 0 519254 895S
12978 0 52414B 9185
13289 0 523932 9406
13593 0 533613 9625
13891 0 538186 9838
14183 0 542658 10046
14468 0 547022 10250
14706 0 550829 10421
14744 0 550932 10449
14909 0 552257 10573
15064 0 553504 10690
15209 0 554674 10799
15345 0 555768 10902
-------
All Locomotives - Corrected
Emissions (Melric Tons/Year)
Year ', HC CO NOx PM
1i99 42404 106228 1092527 26959
£000 1 42404 108228 1081516 26959
2001 42404 108228 1047452 26959
2002 42360 108228 965040 26956
2003 !' 42318 108228 925431 26958
2004 i 42277 108226 663563 26957
2005 '! 40971 108226 792050 26031
2006 ' 39790 106226 7369B2 25194
2007 38717 108228 701140 24434
2008 37756 108223 676611 23754
2009 l! 36924 106226 661091 2316*1
2010 :[ 36193 100228 646471 22651
. 2011 35807 108228 637291 22363
2012 I 35290 108228 627153 22023
2013
2014
2015 •
2016 j
. 2017
34776 10822B 617073 2166S
34267 108228 6070S9 21311
33790 108226 5Q7528 20960
33318 108228 568076 20653
32852 108228 579901 20331
1 2018 1 32391 108228 571805 20012
2019 J 3193S 106228 563787 19698
2020 i' 31485 108228 555849 19387
2021 ' 31039 108228 547989 19060
£022 i 30599 108288 540209 18778
2023
2024 j
2025
2026 1
2027 1
2028
2029
30165 108228 532508 18479
29735 108228 525067 18185
29311 106228 518064 17894
28892 100228 511153 17606
28479 108228 504347 1732S
28072 108228 497831 17047
27670 108228 491387 16772
2030 fl 27262 108228 464906 16496
2031
2002
2033
2034
2035
2036
2037
2038
2039
2040
26600 108228 476682 16240
26502 108228 472545 15983
26129 1Q822B 466493 15739
25761 108226 460526 15494
25437 108228 455295 15281
2S312 100228 453584 15209
25058 108228 451301 15041
2481 1 108228 449072 14876
24571 106228 446896 1472Z
24339 108228 444771 14S72
Reductions {Metric Tons/Year)
HC CO NO* PM
0000
0 0 11011 0
0 0 45075 0
44 0 107487 0
86 0 167096 1
127 0 226964 1
1433 0 300477 928
2614 0 3S5545 1765
36B7 0 391386 2525
4646 0 413916 3204
5481 0 431436 3793
£211 0 446056 4308
6597 0 455236 4575
7114 0 465374 4936
7626 0 475454 5294
8138 0 485468 5648
Bfl14 0 494999 5979
9066 0 504451 6305
9552 0 512626 6628
10013 0 520722 6947
10469 0 528740 7261
10919 0 S36678 7572
11365 0 544537 7B78
11805 0 552318 8181
12239 0 560018 8480
12669 0 56746O 8774
13093 0 574463 9065
13S12 0 561374 9351
13925 0 588180 9634
14332 0 594706 9912
14734 0 601140 10186
15142 0 607621 10462
15524 0 613845 10719
15902 0 619932 10971
16275 0 628034 11220
16643 0 632001 11465
16967 0 637232 1167B
17092 0 638943 11750
17346 0 641226 1191B
17593 0 643455 12080
17033 0 645631 12236
18065 0 647756 12387
"ercent Reduclions {Metric Tons/Yea')
HC CO NOx PM
0.0% 0.0% 1 0% 0.0%
0.0% 00% 4.1% 00%
0,1% 0.0% 98% 0.0%
0.2% 0,0% 15.3% 00%
0.3% 00% 210% 0.0%
34% 0.0% 276% 3.4%
62% 00% 325% 65%
8.7% 00% 35H% 9.4%
110% 0.0% 37.9% 11.9%
129% 00% 39.5% 14.1%
14.6% 00% 408% 16d%
156% 0.0% 41.7% 17.0%
16.8% 00% 42.6% 183%
18,0% 0.0% 43.5% 196%
19.2% 00% 444% 21.0%
20.3% 0.0% 45.3% 22.2%
21.4% 0.0% 46.2% 23,4%
22.5% 0.0% 469% 24.6%
23.6% 0.0% 47.7% 25.8%
24.7% 0.0% 484% 269%
25.B% 0.0% 491% 28.1%
26.8% 0.0% 49.8% 29.2%
27.8% 0.0% 506% 30.3%
28.9% 0.0% 51.3% 31.5%
29.9% 0.0% 51.9% 32.5%
30.9% 0.0% 52.6% 33.6%
31.9% 0,0% 53.2% 347%
32.8% 0.0% 538% 35.7%
33.8% 0.0% 54.4% 36.8%
34,7% 0.0% 550% 37.8%
35.7% 0.0% 556% 38.8%
366% 0.0% 56.2% 39.8%
37.5% 0.0% 56.7% 407%
38.4% 0.0% 57.3% 41.6%
39.2% 0.0% 57.8% 42.5%
40.0% 0.0% 58.3% 43.3%
40.3% 0.0% 58.5% 43.6%
40.9% 0.0% 58.7% 44.2%
41.5% 0.0% 58.9% 446%
42.1% 0.0% 59.1% 45.4%
42.6% 00% 593% 45.9%
-------
Fleet Average Emission Factors - Corrected
(g/bhp-hr)
Year
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013,
2014
2015
2016
2017
2018
' 2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
HC
0.52
0.52
0.52
0.52
0.52
0.51
0.50
0.48
0.47
0.46
0,45
0.44
0.44
0.43
0.42
0.42
0.41
0.41
0.40
0.39
0.39
0.38
0.38
0.37
0.37
0.36
0.36
0.35
0.35
0.34
0.34
0.33
0.33
0.32
0.32
0.31
0.31 -
0.31
0.31
0.30
0.30
0.30
CO
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1,32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
1.32
NOx
13.30
13.17
12.75
11.99
11.27
10.51
9.64
8.97
8.54
8.26
8.05
7.87
7.76
7.64
7.51
7.39
7.27
7.16
7.06
6.96
6.86
6.77
6.67
6.58
6.48
6.39
6.31
6.22
6.14
6.06
5.98
5.90
5.83
5.75
5.68
5.61
5.54
5.52
5.49
5.47
5.44
5.41
(g/gai) ;
PM HC
0.33 10.7
0.33 : 10.7
0.33 10.7
0.33 i 10.7
0.33 i 10.7
0.33 10.7
0.32
0.31
0.30
0.29
0.28
0.28
0.27
0.27
0.26
0.26
0.26
0.25
0.25
0.24
0.24
0.24
0.23
0.23
0.22
0.22
0.22
0.21
0.21
0.21
0.20
0.20
10.4
10.1
9.8
9.6
9.4
9.2
9.1
8.9
8.8
6.7
8.6
8.4
8.3
8.2
8.1
8.0
7.9
7.7
7.6
7.5
7.4
7.3
7.2
7.1 '
7.0
6.9
0.20 j 6.8
0.19 6.7
0.19 | 6.6
0.19
0.19
0.19
0.18
0.18
0.18
0.18 j
6.5
6.4
6.4
6.3
6.3
6.2
6.2
CO
27.4
27.4
27,4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27,4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
27.4
NOx
276.7
273.9
265.2
249.4
234.3
218.7
200.6
186.6
177,5
171.8
167.4
163.7
161.4
158.8
156.3
153.7
151.3
148.9
146.8
144.8
142.8
140.8
138.8
138.8
134.8
133,0
1*31.2
129.4
127.7
126.1
124.4
122.8
121.2
119.7
118.1
116.6
115.3
114.9
114.3
113.7
113.2
112.6
PM
6.8
6.8
6.8
6.8
6,8
6.8
6.6
6.4
6.2
6.0
5.9 ''
5.7 i
5.7
5.6
5.5
5.4
5.3
5,2
5.1
5.1
5.0 !
4.9
4.8
4.8
4.7
4.6
4.5 ;
4.5
4.4 :
4.3
4.2
4.2
4,1
4.0 :
4.0
3.9
3.9 i
3.9
3.8
3.8
3.7
3.7
O-6
-------
APPENDIX P
Conversion of Emission Rates to g/kW-hr
The emission standards and rates presented in this document, which are
expressed in g/bhp-hr, can be converted to g/kW-hr by multiplying them by 1.341. The
converted baseline emission rates and emission standards for the line-haul and switch
cycles are shown below.
Baseline Emission Factors and
Standards Expressed as g/kW-hr
Line-Haul
Baseline
TierO
Tier 1
Tier 2
HC
0.644
1.341
0.738
0.402
CO
1.716
6.705
2.950
2.012
NOx
17.433
12.740
9.923
7.376
PM
0.429
0.805
0.603
0.268
Switch
Baseline
TierO
Tier 1
Tier 2
HC
1.354
2.816
1.609
0.805
CO
2.454
10.728
3.353
3.218
NOx
23.333
18.774
14.751
10.862
PM
0.590
0.966
0.724
0.322
p-1
------- |