United States       Air and Radiation      EPA420-R-99-026
           Environmental Protection               November 1999
           Agency
&ERIV    Final Regulatory Impact
           Analysis: Control of
           Emissions from Marine
           Diesel Engines
                               > Printed on Recycled Paper

-------
                                                                            EPA420-R-99-026
                                                                              November 1999
                                 of
                      Engine Programs and Compliance Division
                              Office of Mobile Sources
                        U.S. Environmental Protection Agency
                                     NOTICE

    This technical report does not necessarily represent final EPA decisions or positions.
It is intended to present technical analysis of issues using data which are currently available.
         The purpose in the release of such reports is to facilitate the exchange of
      technical information and to inform the public of technical developments which
        may form the basis for a final EPA decision, position, or regulatory action.

-------
                            Table of Contents

CHAPTER 1:  INTRODUCTION	  1

CHAPTER 2:  INDUSTRY CHARACTERIZATION	  5
    I.  Introduction 	  5
    II. Variety of Marine Diesel Engines  	  5
       A. Broad Engine Categories	  5
       B. Category 1 Engine Subgroups	  6
       C. Auxiliary and Propulsion Engines  	  8
       D. Recreational and Commercial	  8
    III. Marine Engine Production	  9
    IV. Description of Engine Manufacturers  	  10
       A. Sources of Information	  10
       B. Category 1 Engines	  11
       C. Category 2 Engine Manufacturers	  17
       D. Category 3 Engine Manufacturers	  18
    V. Commercial Vessel Builders	  19
       A. Introduction  	  19
       B. Vessel Types	  19
       C. Shipbuilders	  20
       D. Boat Builders	  21

CHAPTER 3:  TECHNOLOGICAL FEASIBILITY	  25
    I.  Introduction 	  25
    II. Marine Diesel Engine Design Ratings	  25
       A. High Performance Ratings:  Recreational Marine Engines  	  25
       B. Other Ratings:  Commercial Marine Engines	  26
    III. The Marinization Process 	  27
    IV. Background on Diesel Emission Formation	  28
    V. General Description of Emission Control Strategies	  29
       A. Combustion Optimization	  29
       B. Advanced Fuel Injection  Controls  	  31
       C. Improving Charge Air Characteristics 	  32
       D. Electronic Control 	  34
       E. Exhaust Gas Recirculation	  35
       F. Exhaust Aftertreatment Devices	  36
       G. Water Emulsification	  37
    VI. Emission Measurement	  38
       A. Certification Duty Cycles  	  38
       B. Relative Stringency of Duty Cycles 	  40
       C. Not-to-Exceed  Provisions  	  43
       D. Emissions Sampling 	  52
    VII. Baseline Technology Mix	  53

-------
       A. Category 1 Marine Diesel Engines  	  53
       B. Category 2 Marine Diesel Engines  	  55
       C. Category 3 Marine Diesel Engines  	  56
   VIII. Low-Emission Category 1 Marine Engines 	  59
   IX. Anticipated Technology Mix	  63
       A. Category 1 Marine Diesel Engines  	  63
       B. Category 2 Marine Diesel Engines  	  64
       C. Category 3 Marine Diesel Engines  	  64
   X.  Test Fuel Specifications	  64
       A. Category 1 and 2 Marine Diesel Engines  	  64
       B. Category 3 Marine Diesel Engines  	  65
   XI. Impact on Noise, Energy, and Safety  	  67

CHAPTER 4:  ECONOMIC IMPACT 	  75
   I.  Methodology   	  75
   II.  Overview of Technologies	  76
   III. Technology Costs  	  78
       A. Fuel Injection Improvements	  79
       B. Engine Modifications	  81
       C. Turbocharging	  81
       D. Aftercooling	  82
       E. Rebuild Costs  	  84
       F. Certification and Compliance	  85
       G. Total Engine Costs 	  86
       H. Sensitivity	  93
   IV. Aggregate costs	  93

CHAPTER 5:  ENVIRONMENTAL IMPACTS	  95
   I.  Health and Welfare Concerns	  95
       A. Ozone 	  95
       B. Particulate Matter	  96
       C. Carbon Monoxide  	  98
       D. Smoke	  98
   II.  Emission Reductions  	  98
       A. Category 1  	  99
       C. Category 3	  107
       D. Nationwide Totals 	  110
       F. Air Toxics	  113

CHAPTER 6:  COST-EFFECTIVENESS  	  117
   I.   Engine Lifetime Cost-Effectiveness of the New Standards	  118
       A. HC+NOx 	  118
       B. PM	  121
       C. Comparison with Cost-Effectiveness of Other Control Programs  	  123
   II.  30-Year Cost-Effectiveness of the New Standards	  125

                                     ii

-------
                                                               Chapter 1: Introduction
                    CHAPTER 1:  INTRODUCTION
    EPA is adopting new standards for emissions of oxides of nitrogen (NOx), hydrocarbons (HC),
carbon monoxide (CO), and participate matter (PM) from new diesel-cycle engines with a gross
power output at or above 37 kilowatts used in marine vessels.1 This Final Regulatory Impact
Analysis (Final RIA) provides technical, economic, and environmental analyses of the new emission
standards for the affected engines. The anticipated emission reductions will support EPA's effort
to make significant, long-term improvements in air quality in many areas of the U. S.  We proj ect that
the new standards for new marine diesel engines will reduce their emissions of HC by 8 percent,
NOx by 15 percent, and PM by 11 percent in 2020. The NOx reduction is in addition to a 6 percent
reduction in 2020  from international  standards for U.S.-flagged  vessels.  Overall, the new
requirements will provide much-needed assistance to states and regions facing ozone and particulate
air quality problems that are causing a range of adverse health  effects, especially in terms of
respiratory impairment and related illnesses.

    Chapter 2 contains an overview of the manufacturers, including a brief description of the
engines or vessels, that will be affected by this rulemaking.  Chapter 3 provides a description of the
range of technologies that may be available for improving emission  control from these engines,
including detailed projections of a possible set of compliance technologies.  Chapter 4 applies cost
estimates to the  projected  technologies for  several different power categories and describes the
potential impacts on small businesses. These costs are summarized below.  Chapter 5 provides an
overview of the health and welfare issues involved and presents the calculated reduction in emission
levels resulting from the new standards.  Chapter 6 compares the costs and the emission reductions
to estimate the cost-effectiveness of the rulemaking.

    EPA has created these categories to define the three distinctive types of marine diesel engines.
Category 1  includes engines greater than  37 kW but with a per-cylinder displacement of 5
liters/cylinder or less. Category 2 includes engines with 5 to 30  liters/cylinder, and Category 3
includes the remaining, very large, engines.

    Beginning in 2000, marine diesel engines greater than or equal than 130 kW will be subject to
an international NOx emissions standard developed by International Maritime Organization. This
standard is considered to be the baseline case (or "Tier 1") for determining the benefits of the EPA
standards. The MARPOL NOx standard is presented in Table 1-1.
       1  Diesel-cycle engines, referred to simply as "diesel engines" in this analysis, may also be
referred to as compression-ignition (or CI) engines.  These engines typically operate on diesel
fuel, but other fuels may be also be used. This contrasts with Otto-cycle engines (also called
spark-ignition or SI engines), which typically operate on gasoline.

                                            1

-------
Regulatory Impact Analysis
                                        Table 1-1
                       MARPOL Marine Diesel Engine NOx Standard
Starting
Date*
2000
rated speed (n)
n >2000 rpm
130 37 kW
disp. < 0.9
0.9 < disp. < 1.2
1.2 < disp. < 2.5
2.5 < disp. < 5.0
5.0 < disp. < 15
15 < disp. < 20, and
power<3300kW
15 < disp. < 20, and
power > 3300 kW
20 < disp. < 25
25 < disp. < 30
Starting
Date
2005
2004
2004
2007
2007
2007
2007
2007
2007
NOx+HC
(g/kW-hr)
7.5
7.2
7.2
7.2
7.8
8.7
9.8
9.8
11.0
PM
(g/kW-hr)
0.40
0.30
0.20
0.20
0.27
0.50
0.50
0.50
0.50
CO
(g/kW-hr)
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0

-------
                                                               Chapter 1: Introduction
    This document presents  an  analysis of the projected regulatory impacts of the final rule.
Included in these impacts are engine and operating costs, emissions benefits, and the associated cost-
effectiveness of the rule.  Table 1-3 presents these impacts on a per-engine basis for five different
engine sizes. The long-term costs apply for engines produced starting in the sixth year of production
to show the effect of fully amortized fixed costs and learning in production. The listed engine costs
(in  1997 dollars) reflect  the  full anticipated price increment resulting from the  new  emission
standards. These costs are adjusted in the cost-effectiveness calculation to take into account the
value of non-emission benefits (such as improved performance). The operating costs presented here
are the net present value of the  lifetime operating costs using a seven percent rate.  Because the focus
of this  rulemaking is on ground-level ozone,  all of the costs are applied to HC+NOx benefits.  The
benefits are  discounted at a rate  of seven percent to the year that the engine is introduced into
commerce. All of the costs for the Tier 2 standards are incremental to the baseline  represented by
the MARPOL requirements.  The aggregate  cost-effectiveness of the final rule is $103 per ton for
HC+NOx benefits ($23 per ton using long-term costs).
              Tier
                      Table 1-3
2 Discounted Incremental Costs and Benefits by Power Rating
Power
Rating
37-225 kW
225-560 kW
560-1000 kW
1 000-2000 kW
2000+ kW
Cost
Basis
Near-term
Long-term
Near-term
Long-term
Near-term
Long-term
Near-term
Long-term
Near-term
Long-term
Engine
Cost
$1,806
$486
$3,208
$846
$25,395
$856
$22,818
$1,120
$54,192
$13,019
Operating
Costs
(NPV)
$442
$704
$206
$636
$12,430
HC+NOx
Benefits
(annual tons)
4.3
26
80
267
750
Cost-
Effectiveness
$470
$164
$137
$46
$318
$12
$87
$5
$81
$26

-------
Regulatory Impact Analysis

-------
                                                Chapter 2: Industry Characterization
      CHAPTER 2:  INDUSTRY CHARACTERIZATION
I. Introduction

    To help assess the  potential impact of this emission control  program,  it is important to
understand the nature of the affected industries. This chapter describes the marine diesel and vessel
industries along several dimensions, including the variety of engines produced for this market, the
ways in which they are produced, the types of companies that produce these engines, and the types
of companies that manufacture the vessels on which they are installed. The picture that emerges is
one of a complex industry, both in terms of products made and the markets in which they are sold.

II.  Variety of Marine Diesel Engines

A. Broad Engine Categories

    This emission control program covers all new marine diesel engines at or above 37 kW
introduced into commerce in the United States. Thus, the rule encompasses a wide range of engines,
from an auxiliary engine used on a small fishing vessel to a propulsion engine installed on an ocean-
going vessel. Because of the differences among these engines, it is not possible to design one set of
emission limits that apply to all of them.  Therefore, as discussed in the final rule preamble, EPA has
divided marine diesel engines into three subcategories. These engine groups are intended to reflect
the similarities between the marine diesel engines in each group and their land-based counterparts.

    Category 1 engines  are engines with rated power at or  above 37 kW, but with a specific
displacement of less than 5 liters per cylinder.  These engines are similar to land-based nonroad
diesel engines that are  used in applications ranging from skid-steer loaders to large earth moving
machines. Category 2  engines are engines with a specific displacement at or above 5 liters to 30
liters per cylinder. Many of these engines have counterpart locomotive models. Category 1 and
Category 2 marine diesel  engines are often derived from or use the same technologies as their land-
based counterparts.  Consequently, EPA believes that most of the technology being developed to
enable the land-based counterparts to achieve recently finalized emission control programs can be
applied to these marine diesel engines.1'2 Already, limited experience with the application of land-
based nonroad Tier 2 control technologies to marine engines, as part of low-emission demonstration
programs, shows that Category 1 marine diesel engines can achieve emission levels comparable to
the Tier 2 standards for nonroad diesel engines. It is anticipated that this will also be the case when
locomotive Tier 2 technologies are applied to Category 2 engines.  The final emission standards take
into account the fact that some Category 2  engines are larger and use different technology than
locomotive engines.  These technologies and their application to marine diesel engines are discussed
in greater detail in Chapter 3.

-------
Regulatory Impact Analysis
    Category 3 engines are very large engines, at or above 30 liters per cylinder.  These are larger
than any mobile source engines addressed by EPA. They are similar in size to land-based power
plant generators, and are used primarily for propulsion in ocean-going vessels. Because they are
currently designed for maximum fuel  efficiency and performance without consideration of the
impacts on NOx emissions, these engines can have very high NOx emissions. In general, Category 3
marine engines are maintained by crews  whose job is to optimize adjustable parameters on the
engines to reflect ambient conditions. These engines are already using advanced controls of charge
air temperature and pressure which are considered to be emission control strategies for smaller
engines. However, NOx reductions can still be achieved through fuel injection control strategies
such as rate shaping.   Therefore, the NOx  standards developed by the International Maritime
Organization are appropriate for Category 3 engines.  The marine  diesel engine categories are
summarized in Table 2-1.

                                       Table 2-1
                               Engine Category Definitions
Category
1
2
3
Displacement per Cylinder
disp. < 5 liters
(and power 37 kW)
5 disp. < 30 liters
disp. > 30 liters
Basic Engine Type
Nonroad
Locomotive
Unique, "Cathedral"
B. Category 1 Engine Subgroups

    EPA has further divided Category  1 engines into several subgroups. These subgroups are
similar to the land-based nonroad diesel engine subgroups, with one important change. EPA has
based the marine subgroups on cylinder displacement rather than engine power.  This is a more
appropriate scheme for two reasons. First, manufacturers sometimes offer different engine models
that are the same except for the number of cylinders. These engines may fall into  different power
groupings by virtue of the added power from adding cylinders.  Similarly, the common practice of
bolting two marine engines together could in many cases move the combined engine artificially into
a different regime. For example, with respect to emissions and performance, two six-cylinder 300
kW engines bolted together would operate the same as each individual engine. Yet,  by doubling the
power at the crankshaft, the engine would be  subject to less challenging requirements.  Second,
marine engines are often available in a wider  range of power than their land-based counterparts.
While it may be possible to define wider power bands for marine diesel engine subgroups, it may
not be possible to do so without creating phase-in disadvantages for particular companies, especially
in comparison to their land-based phase-in schedule. A displacement scheme should minimize these
inequities.
    In selecting the displacement values corresponding with the nonroad power ranges, EPA
examined the engine displacement and power characteristics of a wide range of existing engines.

-------
                                                 Chapter 2: Industry Characterization
Table 2-2 lists the displacement values we selected to provide the greatest degree of consistency with
the established land-based nonroad engine power groups.  The wide range in power ratings for
engines with a given per-cylinder displacement, however, led to a high degree of overlap in the
attempted correlation between displacement and power rating. As a result, some nonroad engine
models that were spread  across different power groupings are brought together under a single
displacement grouping. This has the potential to move an engine model into a group with somewhat
more or less stringent requirements, but for almost all engine models there was sufficient overlap
to avoid moving a family of engines into an entirely new grouping. The observed overlap highlights
the benefit of relying on displacement for a simplified approach.  This  should give manufacturers
opportunity to more sensibly plan an R&D effort for a family of engines that must meet a single set
of requirements with a common implementation date.

     The net effect of changing to a displacement-based grouping is hard to quantify.  Somewhat
greater emission reductions will likely result for the reasons described above, though it is difficult
to identify the relative sales volumes of engines that fall above and below the threshold under both
scenarios. The effect on costs is expected to be very small. As described above, no engines are
subject to the more stringent standards that would not have a subset of the engine line subject to
those same  standards under a power-based grouping  arrangement. As a result, there should be no
increase in R&D expenditures. Variable costs may be incurred for a greater number of engines, but
the cost analysis in Chapter 4 makes clear that fixed costs dominate  the overall cost impact of
emission  requirements.

-------
Regulatory Impact Analysis
                                        Table 2-2
                         Nonroad Power Categories Corresponding
                           to Per-Cylinder Displacement Ranges
Displacement
(liters/cylinder)
power >37 kW
displ.<0.9
0.9 displ.<1.2
1.2 displ.<2.5
2.5 displ.<5.0
Approximate Corresponding Power Band
from Land-based Nonroad Rulemaking
37 kW<75
75 kW<130
130 kW<560
kW>560
50 hp<100
100 hp<175
175 hp<750
hp>750
C. Auxiliary and Propulsion Engines

    All three categories of marine diesel engines can be further distinguished as to whether they are
used for auxiliary or propulsion purposes. As described in Chapter 3, this will have an impact on
the duty cycle used to measure emissions from the engine. Especially on larger vessels, one or more
auxiliary engines may be added to a vessel to generate electricity for navigational and crew services,
or to operate special shipboard equipment such as cargo cranes. Propulsion engines  are used to
move the vessel through the water. Large ocean-going ships may have small secondary propulsion
engines known as bow-thrusters to help the vessel maneuver in ports.

D. Recreational and Commercial

    Marine diesel engines can also be distinguished according to whether they are used on
recreational or commercial vessels.  This distinction  is important because these engines are not
identical.
    Commercial vessels are typically displacement vessels, which means the engine pushes the
vessel through the water. Optimal operation is more a function of hull characteristics, which are
designed to reduce drag, than engine size. The power ratings of the engines used in these vessels are
analogous to those used in land-based applications.  Commercial vessels are typically heavily used,
and their engines are designed to operate for as many as 4,000  to 6,000 hours a year at the higher
engine loads needed to push the vessel and its cargo through the water.  Commercial vessels are
generally  not  serially produced.   Instead, they  are designed for a specific user and  many
characteristics, including the choice of engines, are set by the purchaser.

    The operation  and  design characteristics of recreational vessels  are very different from
commercial vessels. Recreational vessels are designed for speed and therefore typically operate in
a planing  mode. To enable the vessel to be pushed onto the  surface of the water where it will
subsequently operate, recreational vessels are constructed of lighter materials and use engines with

-------
                                                  Chapter 2: Industry Characterization
high power density (power/weight). The tradeoff on the engine side is less durability, and these
engines are typically warranted for fewer hours of operation. Fortunately, this limitation typically
corresponds with actual recreational vessel use. With regard to design, these vessels are more likely
to be serially produced. They are generally made out of light-weight fiberglass.  This material,
however, minimizes the ability to incorporate purchaser preferences, not only because many features
are  designed  into the fiberglass molds, but also because these vessels are very sensitive to any
changes in their vertical or horizontal centers of gravity.   Consequently, optional features are
generally confined to details in the living quarters, and engine choice is very limited or is not offered
at all.

III.  Marine Engine Production

    Reflecting differences in their size and base engines, different marine diesel  engines are
produced in different ways. Not surprisingly, the largest of these engines, Category 3 engines, have
perhaps the most complicated production process. These engines are generally uniquely built and
designed for a particular vessel used in a particular application. Although their design may draw on
technologies used in stationary source applications, Category 3 engines are not typically derived from
a pre-developed land-based application. In addition, because they are so large, the engine often
becomes part of the structure of the vessel.

    Category 1 and 2 marine diesel engines, on the other hand, are often derived from land-based
engines.  Because of this, their production is often referred to as marinization, meaning the land-
based engine  is modified for use in the  marine environment. Marinization can be very complex or
relatively  simple.  Depending on the degree  of change to the base engine, marinization  can
significantly affect the emission characteristics of an engine.  This will often be the case if the engine
fuel system or cooling systems are modified.

    Some of the more complex changes associated with marinization are performed by large engine
manufacturers such as Caterpillar, Cummins, and Detroit Diesel. For these companies, marinization
may involve  a  significant redesign of their land-based product, including pistons, fuel systems,
cooling systems, and electronic controls. Actual production of marine engines often begins on the
same assembly line as the land-based counterparts. However,  at some stage of the production
process, the marine engine may be moved to a different assembly line or area, where production is
completed using parts and processes specifically designed for the marine derivative engine.

    Post-manufacture marinizers will often make significant changes to base engines as part of the
marinization process. These companies purchase a complete or semi-complete land-based engine
from an engine  manufacturer and finish or modify it using specially designed parts. The process is
often less  complex, although fuel and cooling systems may be extensively modified. The finished
engine is often intended for niche applications.

    A third type of marinization may be performed either by engine manufacturers or post-
manufacture marinizers. Some of these companies start with a completed engine and modify it to
make it compatible for installation on a marine  vessel,  without changing its underlying design

-------
Regulatory Impact Analysis
characteristics or engine calibrations.  These modifications  may involve changes  to  engine
mountings, electronic instrumentation and alarm systems. These "engine dressing" companies may
also add marine gears and couplings, in the case of propulsion engines, or a generator in the case of
auxiliary engines. In contrast to the other marinization processes, we do  not expect these changes
to affect the emission characteristics of the engine.

IV. Description of Engine Manufacturers

    The companies that are most immediately affected by the new emission standards are those that
manufacture marine diesel engines.  This section describes some of the main features of these
companies, with regard to their overall size and marine engine production volumes.  Manufacturers
of each category of engines will be examined separately.

A.  Sources of Information

    The goal of an industry characterization is to describe the basic features of a particular engine
market segment as a whole, focusing on those features that may be helpful for or may become
potential  obstacles to implementation of the  new emission standards.  In other diesel  engine
rulemakings, both highway and nonroad, EPA based its engine industry characterizations in large
part on information collected by Power Systems Research (PSR). This characterization of the marine
diesel engine industry also relies on PSR databases. However, for reasons described below, the PSR
databases are less helpful for describing this industrial segment. Therefore, EPA has augmented the
PSR data with information obtained through other sources, including trade journals and discussions
with various engine manufacturers.

    The two PSR databases used by EPA in this industry characterization are the OE Link database
and the Parts Link database. Both of these depend on information provided to PSR by the engine
manufacturers.  This can be an important drawback for relying on these databases when developing
a picture of the marine diesel engine market.  Specifically, because they rely on self-reporting, very
small  engine manufacturers may not be present in the databases.

    Each of the two databases have additional weaknesses that affect EPA's ability to describe this
industry with a great  deal of precision.  The OE Link database  contains data on U.S.  engine
production for the years 1980  to 1997.  While  the overall production  information is useful in
obtaining a picture of the potential impacts of the rulemaking on this industry, several features of this
database should be kept in mind. First, it includes only those companies that produce engines in the
United States. This is important because it means that engines that are imported into this country,
and that must be certified to the new standards, are not  included. Second, it includes all U.S.
production, without differentiating between the engines that will be sold in this country, and that
must meet the new emission standards, and those that will be exported.  Consequently, relying on
OE Link to paint a picture of the companies that may be affected by the new standards may result
in an overestimate or underestimate of the magnitude of the potential impact of the final rule.
                                           10

-------
                                                Chapter 2: Industry Characterization
    The Parts Link database, on the other hand, contains information about the annual population
of marine diesel engines in the United States for the years 1990 to 1997. This database is useful
because it provides the names of all engine manufacturers, domestic and foreign, who have sold
engines into the U.S. fleet.  The primary weakness of this database is that it does not yield annual
sales  into the market for a given manufacturer.   Instead, the annual information  for each
manufacturer includes the cumulative engine sales over many years, with some internal accounting
for scrappage. Consequently, it does not help EPA determine which foreign manufactures are active
in the U.S. market at this time.

B. Category 1 Engines

    1. Category 1 Engine Manufacturers

    a) Identification of Domestic Producers

    Using the PSR OE Link database, EPA assembled a list of companies that  produce marine
diesel engines in the United States.  These manufacturers, listed in  Table 2-3, are grouped into four
categories.  Domestic engine manufacturers (DEM)  are companies who make complete marine
engines. As described in the marinization discussion in the above section, these marine engines are
likely to be derived from one of the manufacturer's own land-based nonroad or highway engines.
Foreign engine manufacturers (FEM) are similar companies, but are owned by foreign companies.
Post-manufacturer marinizers (PMM) are companies that purchase a finished or semi-complete
engine from another engine manufacture and modify it for use in the marine environment. These
companies are broken down as to whether they make only an auxiliary engine or not, since those who
make only auxiliary engines are more  likely to be dressing engines and therefore not required to
recertify their product to the new emission standards.

                                       Table 2-3
                             Category 1 Engine Manufacturers
                       Source: PSR OE Link - U.S. Production Data
Domestic Engine
Manufacturers
Caterpillar
Cummins
Deere
Detroit Diesel




Foreign Engine
Manufacturers
Isuzu
Yanmar




Post-Manufacture
Marinizers — Propulsion
and Auxiliary
Alaska Diesel
Daytona Marine
Marine Corp. of Amer.
Marine Power
Onan
Outboard Marine
Peninsular Diesel
Trinity Marine
Westerbeke
Post-Manufacture
Marinizers —
Auxiliary only
Kohler
M&L Industries
R.A. Mitchell Co.
Star Power Services
Stewart & Stevenson
York


                                          11

-------
Regulatory Impact Analysis
                Figure 2-1
                                                              Figure 2-2
    U.S. Diesel Marine Engine Production
               Recreational Propulsion Engines
                                                  U.S. Diesel Marine Engine Production
                                                          Commercial Propulsion Engines
                                                              Figure 2-3
    Note that companies involved only in dressing marine engines are not specifically identified as
such in the PSR database, even though they are ostensibly producing marine diesel engines. Through
contacts with various engine manufacturers and its own  efforts, EPA identified several of these
companies.  Because of the special nature of their business, it is hard to precisely identify these
companies; there may therefore be tens if not hundreds of others.  While an exact profile of the
engine dressing industry would be interesting, it  would be  of limited  value  in this  industry
characterization since the final rule exempts them from certification requirements.

    An interesting  aspect of this industry  revealed by  the PSR data  is that, although PMMs
outnumber EMs numerically, their production of propulsion engines is much smaller. Using data
on production volumes as reported in OE Link, it
is possibleto make several observations about the
production of marine engines.   Remember that
these  reflect  total  production, and  not just
production intended for sale in the United States.
However, these figures are  nevertheless an
indicator of the basic make-up of the industry.

    As  illustrated  in Figures 2-1 and 2-2,  the
majority of propulsion engines produced in the
U. S. are made by domestic engine manufacturers,
even though the number of PMM companies
exceeds  the  number  of   large   engine
manufacturers. This is true for both commercial
and recreational propulsion engines.

    In addition to producing more propulsion
                                                  U.S. Diesel Marine Engine Production
                                                            Auxiliary Engines
                                           12

-------
                                                 Chapter 2: Industry Characterization
marine diesel engines, DEMs are also generally much larger companies than PMMs. This is to be
expected since PMM generally do not produce base engines or serve markets outside of marine
applications.  The size difference between DEMs and PMMs is confirmed by a financial analysis
performed by ICF Incorporated using Dunn and  Bradstreet information.  Of the eight large
companies examined, five were DEMs and three were PMMs.  According to this analysis, the
median large company has approximately 23,000 employees and annual sales of $3.2 billion. All
of the small companies observed were PMMs, and the median small company has approximately 60
employees and annual sales of $19 million. Conversations with various engine manufacturers also
confirm the  observation that PMMs are small companies.   Several, such as Peninsular, Alaska
Diesel, and Daytona are family-owned business that operate on a scale much smaller than the large
engine manufacturers.

    When auxiliary engines are considered, however, the relationship between output and engine
sized  is reversed.  Specifically, PMMs appear to produce  more marine auxiliary marine diesel
engines than DEMs.  In addition, as illustrated in Figure 2-3, a large number of auxiliary marine
engines seem to be produced by companies who specialize in auxiliary engines.  This,  in turn,
suggests that many of these engines are assembled by engine dressing companies.  There are at least
two explanations for this observation. It could be the case that large engine manufacturers do not
get involved in the dressing process for auxiliary engines, preferring to  leave this to  smaller
companies that can tailor their product to the special needs of marine customers. Alternatively, it
could be the  case that the data on auxiliary engine productions for DEMs and FEMs is included in
their land-based nonroad figures instead of their marine figures.

    b) Identification of Foreign Manufacturers

    As indicated above, the PSR OELink database contains information only on domestic producers
of marine diesel engines. Yet, foreign companies that sell engines in the United States may also be
affected by this rulemaking; it is therefore important to identify them and estimate their share of the
domestic market. PSR Parts Link population database contains a list of all the companies that have
engines in the current U.S. fleet.  A list of those companies is contained in Table 2-4. Note that, with
regard to domestic engine manufacturers, Table 2-4 lists only the manufacturer of the base engines.
Thus, Navistar and GM  are on  the list from the population database (Parts Link) but not on the
production database (OE Link),  since they do not produce a finished marine diesel engine. This may
be the case for some of the foreign engine manufacturers as well.  It should also be noted that the
PSRParts link database reflects  manufacturers of engines that have been sold into the U.S.  market,
and does not permit distinguishing between those that are still active in the market and those that are
not. The list also does not reflect those companies who seek to sell into the U.S. market.
                                           13

-------
Regulatory Impact Analysis
                                     Table 2-4
                           Category 1 Engine Manufacturers
                     Source: PSRParts Link - U.S. Population Data
Domestic Engine
Manufacturers
Allis-Chalmers
Caterpillar
Cummins (including
Consolidated Diesel)
Deere
Detroit Diesel
Ford
GM (Chev/Pont/Can)
Navistar
Waukesha


Foreign
Bedford/AWD
Deutz-MWM
Hino
Isotta Fraschini
Isuzu
Komatsu
Kubota
Leyland
Lister-Fetter
Mack
MAN
Mazda
Mercedes Benz
MIT Motors
Engine Manufacturers
Mitsubishi HVY
MTU
New Holland
Perkins
PSA
Renault
Rover Cars
SACM
Scania
Steyr
Toyota
VM
Volvo
Yanmar
                                        14

-------
                                                 Chapter 2: Industry Characterization
                Figure 2-4
            Figure 2-5
    U.S. Diesel Marine Engine Population
            Commercial Propulsion Engines
U.S. Diesel Marine Engine Population
          Auxiliary Engines
                Figure 2-6
            Figure 2-7
    U.S. Diesel Marine Engine Population
               Recreational Propulsion Engines
U.S. Diesel Marine Engine Population
              All Engines
    The PSRParts Link database suggests that the majority of Category 1 engines sold into the U.S.
domestic market, approximately 74 percent, were manufactured by domestic companies.  The
remaining 26  percent were manufactured by  foreign companies.   This  suggests that foreign
companies may have a considerable presence in the Category 1  engine market. When broken down
by engine category, this data suggests that the domination of DEMs is strongest in the recreational
category. These statements are illustrated in Figures 2-4 through 2-7 above.  This data supports the
above observation suggested  by the PSR OE Link database, that U.S. manufacturers  have the
majority of the domestic marine diesel market.
                                           15

-------
Regulatory Impact Analysis
    c) Other Marine Diesel Engine Manufacturers

    Finally, there are at least four other marine diesel engine manufacturers identified by EPA that
do not appear in the PSR databases.  These are American Diesel, Mercury Marine, Norpro, and
Regan Equipment. It is EPA's understanding that each of these are PMMs, and therefore their
omission will not greatly affect the above observations about the characteristics of the marine diesel
engine market.

    d) Conclusion

    The above information suggests that the U.S. diesel  engine industry is dominated  by U.S.
manufacturers and that the large engine manufacturers produce the majority of the engines.  PMMs
play a significant role in this market, although more so in the auxiliary engine market than in the
propulsion market. Foreign manufacturers are a smaller presence in the market, though they play
a larger role in commercial than in recreational applications.

    2. Category 1 Engine Applications

    With  regard to the numbers of engines produced, the  PSR OE Link database indicates total
annual production has varied over the past 17 years, from a low of 10,000 units in 1980 to a high of
nearly 20,000 units in 1988. Figure 2-8 illustrates that growth in this industry was interrupted in the
late 1980s and early 1990s.  This may have been due to the economic recession in those years or
changes in the tax code, particularly from additional taxes on luxury goods.
                                       Figure 2-8
Category 1 Engine Production
onnnn
zuuuu
1 cnnn
IOUUU
innnn
IUUUU
cnnn
ouuu
19
/ \ ^""'
^^ \ /-— - _^^


1981 1983 1985 1987 1989 1991 1993 1995 1997
80 1982 1984 1986 1988 1990 1992 1994 1996
Total Production
                                           16

-------
                                                Chapter 2: Industry Characterization
    Figure 2-9 illustrates the portion of production for each segment of the industry, commercial
propulsion engines, recreational propulsion engines, and auxiliary engines. According to this chart,
auxiliary engines have had a fairly constant share of total production, increasing from about 7 percent
at the beginning of the period to about 12 percent  at the end.  The more important feature of the
shares of production concern propulsion engines. In the early 1980s, recreational engines replaced
commercial engines as the largest segment of production. This has been the case  since then,
although production of these two types of propulsion engines was almost the same in the early 1990s.

                                      Figure 2-9
                           Category 1 Engine Production
 70%
 60%
 50%
 40%
 30%
 20%
 10%
  0%
         1981
1983
1985
1987
1989
1991
1993
1995
1997
     1980     1982     1984     1986     1988     1990    1992    1994
                                                          1996
                        Aux
                           Comm
                                      Rec
C. Category 2 Engine Manufacturers

    The greater mass and slower operating speeds of Category 2 engines generally corresponds with
better brake-specific fuel efficiency and longer total engine lifetime.  Because of the large capital
requirements and the need to supply replacement parts over many years, companies with a history
of supplying marine engines have a distinct advantage in the marketplace.  Also, because of the
relatively low sales volume of these engines, producing engines for locomotives and for the
international marine market is also very important for broadening the sales base for distributing fixed
costs. Category 2 engine sales are therefore dominated by a small number of manufacturers.

    General Motors Electromotive Division (EMD) has the longest history of supplying Category 2
marine engines and produces the majority of these engines introduced into the U.S. Caterpillar is
also actively pursuing sales of Category 2 engines.  Several other companies have engines available,
but do not focus on the marine market. General Electric sells the large majority of its engines for
locomotives, though some of these engines find their way into marine vessels. Fairbanks Morse has
                                          17

-------
Regulatory Impact Analysis
produced Category 2 marine engines for a long time, but has shifted its focus to supplying engines
into military applications, rather than the commercial applications that will be most affected by new
emission standards.  Several foreign companies produce engines that could be sold into the U.S.
market, though companies buying new vessels for domestic use have most often specified U.S.-made
engine models.  This may change somewhat in the future, especially considering the ongoing
developments in joint ventures and other business arrangements. Wartsila, MaK, MTU, Deutz, and
Yanmar are some of the leading  candidates for supplying engines into the U.S.-flagged  marine
vessels.

    Tugboats and towboats are the principal use of Category 2 marine engines. While tugboats
with total propulsive power up to about 2000 kW typically use multiple Category 1 engines, larger
models rely on one or two Category 2 engines. Similarly, the big fishing vessels, ferries,  and many
workboats use these larger engines. These high-powered engines carry greater loads and, in many
cases, have more intensive operations. In addition, multiple Category 2 engines are commonly used
for auxiliary power on an ocean-going vessel.

    With prices approaching or exceeding $1 million for a new engine, there is a strong dependence
on maintaining and remanufacturing engines in the field. A strong preventive maintenance program
is the norm, often including extensive, ongoing diagnostics for oil quality, fuel consumption, and
other engine performance parameters.  Common industry practice is to completely remanufacture
the engines every five years.  Procedures have improved to the point that engine  durability on
remanufactured engines is no different than  on new engines.  Since engine remanufacturing costs
only 20 to 30 percent as much as buying a new engine, even twenty- or thirty-year-old engines are
commonly overhauled to provide dependable power.  The quality of remanufactured engines is so
good, and the price savings so great that it is even common practice to install remanufactured marine
or locomotive engines into new marine  vessels.

D. Category 3 Engine Manufacturers

    There are currently no U.S.  manufacturers  of Category 3 marine engines for commercial
purposes.  The Agency, however, has identified 22 foreign manufacturers of these engines, a large
fraction of which are located in Germany and Japan. Due to the competitive nature of this industry,
the number of manufacturers has changed over time due to mergers between companies. In addition,
of the Category 3 engine manufacturers identified, only 12 produce engines of their own design. The
remainder of the manufacturers produce engines under licensing agreements with other companies
that control engine design.  Table 2-5 presents the Category 3 engine manufacturers identified by
EPA.  These manufacturers are divided by those that produce their own designs and those that solely
produce licensed  designs. In many cases, the manufacturer names are hyphenated which generally
is the result of a merger between two or more companies.
                                           18

-------
                                                Chapter 2: Industry Characterization
                                       Table 2-5
                    Manufacturers of Category 3 Marine Diesel Engines
                  Produce Own Designs
    Produce Engines Using
      Licensed Designs
               Akasaka
               Allen Diesel
               Daihatsu
               Deutz-MWM
               Fincantier Group/GMT
               Hanshin
               MAN-B&W
               MaK
               Matsui Iron Works
               Mitsubishi Heavy Ind. LTD
               SEMT-Pielstick
               Wartsila-NSD
Bazan
Cegielski
Coltec Industries
Dieselmotorenwerk Vulcan
GMBH
Jadranbrod ULJANK
Kawasaki Heavy Ind.
Kirloskar Oil Engines Ltd.
Koloma Plant Joint Stock
Maschinenbau Halberstadt
Niigata
V.  Commercial Vessel Builders

A. Introduction

    The industry characterization for the commercial marine vessel industry was developed by ICF
Incorporated under contract for EPA. Following is a summary of their findings. A complete copy
of this report is available from EPA Air Docket A-97-50.

    There are three main parts to this report. Parts 1 and 2 establish the location, number, size, and
relevant factors associated with shipbuilding and boat building.  Part 3 covers production practices
commonly employed by shipbuilders and boat builders.

B. Vessel Types

    The report makes a distinction between two broad groups of commercial vessels, "ships" and
"boats," based on a vessel's basic dimensions, mission and area of operation.

    •  Commercial Ships—this  category  is comprised of larger merchant vessels,
        usually exceeding 400  feet in length, that engage in waterborne trade and/or
        passenger transport. These ships tend to operate in Great Lakes, coastwise, inter-
        coastal, non-contiguous, and/or transoceanic routs.  Principle commercial ship
        types are dry cargo ships, tankers, bulk carriers and passenger ships. Passenger
        ships include cruise ships and larger ferries.
                                          19

-------
Regulatory Impact Analysis
    •   Commercial Boats—This category is comprised of smaller service and industrial
         vessels less than 400  feet in length that provide service to commercial ships,
         industrial vessels, and/or barges that perform  specialized marine functions.
         Commercial boats are found mainly in inland and coastal waters.  Principle
         commercial boat types are tugboats, towboats, offshore supply boats, fishing and
         fisheries vessels, passenger boats, and industrial boats. Passenger boats include
         crewboats, excursion boats, and smaller ferries.

C. Shipbuilders

    1. Production Information and Industry Drivers

    The vast majority of commercial ships are foreign built. Currently, the U.S. holds less than 0.5
percent of the world market share of commercial shipbuilding and repair. The late 1970s and early
1980s reflect a worldwide boom in shipbuilding, particularly in energy-related vessels such as oil
drill rigs and tankers. With falling oil  prices in the 1980's, there was a substantial decline in
commercial vessel orders throughout the world.  No commercial orders were placed  at U.S.
shipyards for 1987 through 1989. In the 1990s, however, the rate of orders has slowly increased to
the current  rate of approximately 16 ships per year.

    2. Location and Number of Builders

    There  are currently only 18 major  shipbuilding facilities in the United States. Most of the
facilities are located on the East and Gulf Coasts. The Gulf Coast is a particularly attractive location
for shipbuilders due to the relatively low labor costs for the region. New Orleans and Houston lead
the U.S. in  international trade tonnage.

    3. Size and Sales Profile

    The six largest companies in the  shipbuilding industry in terms of employment and sales
revenue are represented by The American Shipbuilding Association (ASA). Almost all commercial
ships  currently manufactured in the U. S.  constructed at an ASA yard. These six companies employ
nearly 55,000 people, representing 90 percent of new ship construction employment in the U.S.
Additionally, five of the yards are the largest employers in their states. Sales revenue for these six
companies  is estimated at approximately $5.2 billion dollars annually.

    4. Outlook

    Worldwide, the demand for ships is expected to increase during the next decade.  Chartering
companies  and class societies are tightening their requirements. Insurance will become increasingly
expensive for older tonnage. Tighter safety regulations will force single-hulled and 30-year old oil
tankers out of the market. These collective forces are expected to create a market momentum in
which new  vessels may be more attractive.
                                           20

-------
                                                 Chapter 2: Industry Characterization
    In sum, the gains due to the anticipated U.S. market growth in commercial shipbuilding are
expected to be modest. During the next decade, U.S. shipyards are projected to lose 28,000 jobs as
a result of the cut in military orders. This shift in the government market may result in U. S. Shipyard
closings or conversions to repair facilities or even non-shipbuilding projects as they have during the
lean years.  Recent reports by market analysts also suggest a tempered shipbuilding outlook due to
the financial crisis in Asia. Any increases in U.S. shipbuilding is likely to be modest.

D.  Boat Builders

    1. Production and Industry Drivers

    Because of Jones Act requirements, the vast majority of boats used in the U.S.  are made.
domestically. According to WorkBoat Magazine's annual construction survey, 569 commercial
boats were ordered in 1997, approximately 100 more than were reported in both  1995  and 1996.
While these estimates rely upon survey responses, WorkBoat Magazine's  1995 estimates closely
compare with another industry journal, Seatrade Review.

    The U. S. boatbuilding industry is currently influence by four key factors: increasing demand for
T class vessels, increasing demand for offshore supply vessels, increasing demand for oil, and the
expansion of demand for casino boats.  For further analyses of these factors, see section 3.1 of the
ICF report.

    2. Boat Types and Customers

    U.S. boatbuilders construct a wide variety of commercial boats less than 400 feet in length.
Most of these yards are  small and manufacture small and/or less complex boat designs.  A few of
the yards have the capacity to build more advanced boats, and some also have the capacity to build
small ships. U.S. Boatyards build boats primarily used on inland and coastal waterways between
U.S. ports.  Such vessels must satisfy Jones Act requirements  and, therefore, be built in  the United
States.  For this reason, the U.S. boatbuilding industry has a protected local market and does not face
intense foreign competition.

    3. Location and Number of Builders

    In contrast to the highly concentrated shipbuilding industry, there are several hundred yards that
build many different types of boats. Using various sources of information including Duns and
Bradstreet databases, industry magazines,  and  Internet searches, ICF  estimated that there are
approximately 430 firms that engage in boat building. A majority of the boat builders are located
in the Gulf Coast, the Northeast, and the West Coast. The number of boatbuilders in these three
regions  account for approximately 30 percent,  26 percent and 25  percent of the boat building
industry, respectively.
                                           21

-------
Regulatory Impact Analysis
    4. Size and Sale Profile

    Given the large number of companies in the Boatbuilder Database, it was not possible to contact
each one to gather size and sales information. Therefore, two boatbuilders were contacted from the
top, middle, and bottom third of the boatbuilder database based on the total employment for the
company, including subsidiaries. For specific information on the size and sales of boatbuilders, see
section 3.4 of the ICF report as referenced in the introduction of this section.

    5. Applications by Geographic Location

    A majority of boatbuilders are located in the Gulf Coast, the Northeast, and the West Coast.
Collectively, these three regions represent 345 boatbuilders, or 80 percent of all companies in the
Boatbuilder Database.

    The Gulf Coast is currently the leader in boat production, outperforming the East and West
Coasts due to advantages in weather, production automation, favorable labor rates and relatively
inexpensive real estate.  Approximately 30 percent of all boatbuilders are located in the Gulf Coast.
Overall, roughly 35 percent of the yards that build industrial boats, tugboats, ferries and passenger
boats are located in this region.

    The Northeast comprises about 25 percent of the firms in the database. Compared with yards
located in the Gulf Coast and the South Atlantic, a boatbuilder in the north is 20 to 25 percent more
expensive, taking into account both labor and operating costs. Few of these Northeastern yards build
industrial boats, tugboats, ferries, and passenger boats. However, nearly 70 percent of the yards that
build fishing boats are located in this region.

    The West is primarily dedicated to highly specialized boat yards. For example, boatbuilder in
this region construct vessels  such as processor catchers, tuna clippers, and small aluminum brine
shrimp boats. About 25 percent of the firms in the database are located in the West, including 30
percent of all firms that build fishing boats and nearly 42 percent of the firms that build industrial
boats, tugboats, ferries,  and passenger boats.

    ICF also identified 156 small boat builders that employ fewer than 20 workers, based  on total
company employment. These yards primarily repair and build small boats less than 55 feet in length
which are used in fishing applications such as lobster, oyster, and crab boats.

    6. Very Small Boat Builders

    ICF identified 182  small boatbuilders that employ fewer than 20 workers, based on total
company employment. These yards primarily repair and build small boats less than 55 feel in length,
which are used in numerous fishing applications. Such applications include lobster, oyster, and crab
boats.  Production of these vessels requires smaller facilities and fewer employees.  Vessel
production at these yards is cyclical, which in part explains why the primary source of revenue for
most small yards is repair services.

                                           22

-------
                                                 Chapter 2: Industry Characterization
    According to ICF, little data exists on how many of these small vessels are manufactured
annually. However, the production of small fishing vessels is extremely regional. Generally, local
demand looks to local builders for vessel production.

    Additionally, some small yards do assemble tugs,  particularly for inland use.   Tugs are
technically less difficult to develop than other types of boats and their designs have not changed
much over time.

    7. Outlook

    Decisions by leading boatbuilders to reopen facilities  and expand their labor forces are strong
indications that market growth for commercial boatbuilding is expected to be considerable. The
extent that smaller boatbuilders realize the same level of growth is uncertain.  If the commercial
boatbuilding industry continues to grow at approximately the same pace maintained over the last
decade, the overall demand for new boats may mean more business for even the smallest yards.

    Demands  for new boats in the U.S. will likely be in "hot markets" such as T class vessels,
offshore supply vessels, drillships and semisubmersible rigs, and casino boats. For this reason, the
projected increase in demand for new boats will mean more business for the U.S. boating industry,
as foreign builders are ineligible to build for this market.

    In sum, U.S. boatbuilders are expected to continue to be a major consumer of marine  diesel
engines. If market increases are realized, the number of marine diesel units used in the boatbuilding
industry may increase substantially.
                                           23

-------
Regulatory Impact Analysis
Chapter 2 References

1.  Federal Register, "Control of Emissions of Air Pollution from Nonroad Diesel Engines; Final
Rule," 63 FR 56968, October 23, 1998.

2.  Federal Register, "Emission Standards for Locomotives and Locomotive Engines; Final
Rule," 63 FR 18978, April 16,  1998.

3.   Boating Industry Magazine, " 1997 Annual Industry Reviews: The Boating Business", January
1998.

4.   "Diesel  & Gas Turbine Worldwide Catalog:  Product  Directory and Buyers Guide," 1997
Edition, Volume 62.

5.   Dunn & Bradstreet.

6.   National Marine Manufacturers Association -http://209.94.233.3/iww/customers/nmma/.

7..  National Marine Manufacturers Association - "Monthly Statistical Report of the Recreational
Marine Industry," July 1997.

8.   National Marine Manufacturers Association - "Monthly Statistical Report of the Recreational
Marine Industry, " July 1996.
                                          24

-------
                                                Chapter 3: Technological Feasibility
      CHAPTER 3:  TECHNOLOGICAL FEASIBILITY
I. Introduction

    The purpose of this chapter is to discuss the feasibility of achieving significant reductions in
exhaust emissions from marine diesel engines. Because marine diesel engines are often derived from
or use the same technologies as land-base engines, EPA believes that the marine industry will be able
to capitalize on the technological improvements being made to land-based diesel engines and achieve
similar emission reductions.

    This chapter covers a wide range of areas that affect the technological feasibility of this rule.
It begins with a rudimentary description of the classifications and applications of marine engines as
well as the marinization process. This is followed by an overview of emission formation from diesel
engines.  Then,  established diesel emission control technologies will be described, which EPA
believes can be  used on marine diesel engines to meet the new  emission standards. Next, the
emission test procedures for measuring and assessing the impacts of the new emission standards will
be discussed. This will be followed by a description of baseline  technologies currently used on
marine diesel engines and a discussion of developing emission control strategies and results achieved
from various low-emission marine engine designs.   Finally, this  chapter presents a mix of
technologies EPA believes will allow manufacturers to reach the Tier 2 emission standards.

II. Marine Diesel Engine Design Ratings

    Broadly speaking, the rating of an engine refers to the type of operating conditions the engine
is designed to handle.  For marine diesel engines, engine ratings also roughly correspond to how the
engine is intended to be used, for primarily recreational or some type of commercial application.
These are described in greater detail below.

A. High Performance Ratings: Recreational Marine Engines

    High performance engines are generally intended for light-use recreational marine purposes.
The high performance rating varies by manufacturer but generally  means that the engines are
designed to maximize the power-to-weight ratio. In other words, they are designed for maximum
performance while minimizing the space needed on the boat for the engine and minimizing the added
weight of the engine. Because the purpose of these engines is to get  high power from a small engine,
they are not built to withstand long periods of operation at rated power. For planing vessels, the
engine only needs the high power until the vessel is raised to the surface of the water, then a much
lower power is used for cruising.  Marine engines with a high performance rating are generally
designed to operate a maximum of 200 to 1000 hours per year. They are intended for variable-load
applications where the engine operates at full power no more than  15 minutes at a time and where


                                          25

-------
Regulatory Impact Analysis
the engine spends one out of every eight to twelve hours at full load.  These engines are typically
Category 1 engines.

B. Other Ratings: Commercial Marine Engines

    Light-duty commercial engines have many  of the same characteristics as the recreational
engines, except that they are designed to be more durable and thus more attractive for commercial
applications.  Greater durability is achieved by using a heavier engine design for a given power
rating. Although these engines are still only designed to be operated one hour out of every six to
eight hours at high power, they are intended to be used anywhere from 750 to 3000 hours per year,
for active seasonal to annual use.  Light-duty commercial  engines are usually used in boats with
planing hulls such as,  seasonal fishing vessels and emergency rescue boats, or to power bow thrusters
on larger vessels. While they can also be used in recreational applications, this is typically on larger
vessels that need  more power to achieve planing or are intended  to be  used more than just
occasionally.  These are typically Category 1 engines.

    Intermittent-duty commercial rated engines are intended for use in boats with either planing or
displacement hulls that are used under variable speeds and loads. These engines are designed to
operate at full load no more than half of the time  for 2000 to 4000 hours  per year.   Typical
applications for intermittent-duty engines are commercial fishing  boats, ferries, and coastal
freighters. In addition, most auxiliary marine engines fall into this category. Again, the use of these
engines is not exclusively commercial, and they can be used on large displacement hull yachts.
Intermittent-duty engines are typically Category  1  engines, but may include some Category 2
engines.

    Marine engines with a medium  continuous rating are designed to operate a large number of
hours at fairly constant speeds and loads on vessels with displacement hulls.  These engines are
recommended for applications where the engine operates from 3000 to 5000 hours per year (60-100
hours per week), but spends no more than about 80 percent of the time at full load.  The advantages
of this design are good durability and fuel consumption while still maintaining some performance
benefits.  These engines may be found in applications such  as crew and supply boats, trawlers, and
tow boats. Large auxiliary engines generally fall into this rating as well.  This rating includes mostly
Category 2 engines with some Category 1 engines.

    Marine engines with a continuous rating are designed to operate under full load up to 24 hours
per day and generally are operated more than 5000 hours per year.  These engines are designed to
maximize durability and fuel efficiency which results in low operating costs.  Typical applications
for marine engines with a continuous rating range from tugboats to ocean-going vessels.  Tugboat
applications typically use Category 2  engines, but the majority of the ocean-going vessels use
Category 3 engines.  Great Lakes vessels may use either Category 2 or Category 3 propulsion
engines.
                                           26

-------
                                                 Chapter 3: Technological Feasibility
III. The Marinization Process

    As noted in the previous chapter, marine engines are not generally built from the ground up as
marine engines.  Instead, they are often marinized land-based engines.  The following is a brief
discussion of the marinization process, as it is performed by either engine manufacturers or post-
manufacture marinizers.

    The most obvious  changes made to a land-based engine as part of the marinization process
concern the engine's cooling system.  Marine engines generally operate in closed compartments
without much air flow  for cooling.  This restriction can lead to engine performance and safety
problems.  To address engine performance problems, these engines make use of the ambient water
to draw the heat out of the engine coolant. To address safety problems, marine engines are designed
to minimize hot surfaces.  One method of ensuring this, used mostly on recreational or light-duty
commercial engines, is  to run cooling water through a jacket around the exhaust system and the
turbocharger. Larger engines generally use  a thick insulation around the exhaust pipes. Hardware
changes associated with these cooling system changes often include water jacketed turbocharger s,
water cooled exhaust manifolds, heat exchangers, sea water pumps with connections and filters, and
marine gear oil coolers. In addition, because of the greater cooling involved, it is often necessary
to change to a single-chamber turbocharger, to avoid the cracking that can result from a cool outer
wall and a hot chamber divider.

    Other important design changes are related to engine performance. Especially for planing hull
vessels used in recreational and light-duty commercial marine applications, manufacturers strive to
maximize the power-to-weight ratio of their marine engines, typically by increasing the power from
a given cylinder displacement.  The most significant tool to accomplish this is the fuel injection
system: the most direct way to increase power is to inject more fuel.  This can require changes to the
camshaft, cylinder head, and the injection timing and pressure. Currently, the design limits for
increased fuel to the cylinder are smoke and durability. Modifications made to the cooling system
also help enhance performance. By cooling the charge, more air can be forced into the cylinder. As
a result, more fuel can be injected and burned efficiently due to the increase in available oxygen.
In addition, changes are often made to the pistons, cylinder head components, and the lubrication
system. For instance, aluminum piston skirts may  be used to reduce the weight of the pistons.
Cylinder head changes include changing valve timing to optimize engine breathing characteristics.
Increased oil quantity and flow may be used to enhance the durability of the engine.

    Marinization may also involve replacing engine components with similar components made of
materials that are more  carefully adapted to the marine environment.   Material changes include
more use of chrome and brass  including changes to electronic fittings  to resist water induced
corrosion.  Zinc anodes are often used to  prevent engine components, such as raw-water heat
exchangers, from being damaged by electrolysis.

    Depending on the stage of production and the types of changes made, the marinization process
can have an impact on  the base engine's emission characteristics.  In other words, a land-based
engine that meets a particular set of emission limits may no longer meet these limits after it is

                                           27

-------
Regulatory Impact Analysis
marinized. This can be the case, for example, if the fuel system is changed to enhance engine power
or if the cooling system no longer achieves the same degree of engine cooling as that of the base
engine.  Because marine diesel engines are currently unregulated, engine manufacturers have been
able to design their marine engines to maximize performance, often obtaining power-to-weight ratios
much higher than land-based applications.  The challenge presented by the new emission standards
will be to achieve the emission limits while maintaining these favorable performance characteristics.

IV. Background on Diesel Emission Formation

    In a diesel engine, the liquid fuel is injected into the combustion chamber after the air has been
heated by compression (direct inj ection), or the fuel is inj ected into a prechamber, where combustion
initiates before spreading to the rest of the combustion  chamber (indirect injection). The fuel is
injected in the form of a mist of fine droplets or vapor that mix with the air.  Power output is
controlled by regulating the amount of fuel inj ected into the combustion chamber, without throttling
(limiting) the amount of air entering the engine. The compressed air heats the injected fuel droplets,
causing  the fuel to evaporate and mix with the available oxygen.  At several  sites where the fuel
mixes with the oxygen, the fuel auto-ignites and the multiple flame fronts spread through the
combustion chamber.

    NOx and PM are the emission components of most concern from diesel engines. Incomplete
evaporation and burning of the fine fuel  droplets or vapor result in emissions of the very small
particles of PM. Small amounts of lubricating oil that escape into the combustion chamber can also
contribute to PM. Although the air/fuel ratio in a diesel cylinder is very lean, the air and fuel are not
a homogeneous charge as in a gasoline engine.  As the fuel is injected, the combustion takes place
at the  flame-front where the air/fuel ratio is near stoichiometry.  At localized areas, or in cases where
light-ends have vaporized and burned, molecules of carbon remain when temperatures and pressures
in the cylinder become too low to sustain combustion as the  piston reaches bottom dead center.
Therefore, these heavy products of incomplete combustion are exhausted as PM.

    High temperatures and excess oxygen are  necessary for  the formation of NOx.   These
conditions are found in a diesel engine.  Therefore, the diesel combustion process can cause the
nitrogen in the air to combine with available oxygen to form NOx. High peak temperatures can be
seen in typical unregulated  diesel engine designs.  This is because the fuel is injected early to help
lead to more complete combustion, therefore, higher fuel efficiency.  If fuel is injected too early,
significantly more fuel will mix with air prior  to combustion.  Once combustion begins, the
premixed fuel will burn at once leading to a very high temperature spike.  This high temperature
spike, in turn, leads to a high rate of NOx formation. Once combustion begins, diffusion burning
occurs while the  fuel is being injected which  leads to a more constant,  lower temperature,
combustion process.

    Because of the presence of excess oxygen, hydrocarbons evaporating in the combustion chamber
tend to be completely burned and HC and CO are not emitted at high levels. Evaporative emissions
from diesel engines are insignificant due to the low evaporation rate of diesel  fuel.
                                          28

-------
                                                 Chapter 3: Technological Feasibility
    Controlling both NOx and PM emissions requires different, sometimes opposing strategies. The
key to controlling NOx emissions is reducing peak combustion temperatures since NOx forms at
high temperatures.  In contrast, the key to controlling PM is higher temperatures in the combustion
chamber or faster burning. This reduces PM by decreasing the formation of particulates and by
oxidizing those particulates that have formed. To control both NOx and PM, manufacturers need
to combine approaches using many different design variables to achieve optimum performance.
These design variables are discussed below.

V. General Description of Emission Control Strategies

    EPA believes that the new emission standards for marine diesel engines can be met using
technology that has been developed for and used on locomotive, land-based nonroad, and highway
engines. This section discusses technology used successfully on diesel engines to reduce exhaust
emissions,  including  combustion  optimization,  better fuel  control,  improved charge  air
characteristics, exhaust gas recirculation,  and aftertreatment. A more  detailed analysis of the
application of several of these technologies to marine engines is discussed later in this chapter.  The
costs associated with applying these systems to marine engines are considered in the next chapter.

A. Combustion Optimization

    Several parameters in the combustion chamber of a heavy-duty diesel engine affect its efficiency
and emissions. These engine parameters include injection timing, combustion chamber geometry,
compression ratio, valve timing, turbulence, injection pressure, fuel spray geometry and rate, peak
cylinder temperature and pressure, and charge  (or intake) air temperature and pressure.  Many
technologies that are designed to control the engine parameters listed above have been investigated.
As mentioned previously, however, a positive  influence on  one pollutant may have a  negative
influence on another.  For example, charge air cooling reduces NOx emissions, but increases PM.
Manufacturers will need to integrate all of these variables into optimized  systems to meet the new
emission standards.

    1. Timing retard

    The effect of injection timing on emissions and performance is well established.1'2'3'4 Retarded
timing will most likely be used at cruising speeds where propulsion marine engines spend most of
their operating time. NOx is reduced because the premixed burning phase is shortened and because
cylinder temperature and  pressure are lowered. Timing retard increases HC, CO, PM,  and fuel
consumption, however, because the end of injection comes later in the combustion stroke where the
time for extracting energy from fuel combustion  is shortened and the cylinder temperature and
pressure are too low for more complete oxidation of PM. One technology that can offset this trend
is higher injection pressure, which is discussed further below.
                                          29

-------
Regulatory Impact Analysis
    2. Combustion chamber geometry

    While manufacturers are already achieving emission reductions through modifications to the
combustion  chamber, EPA believes there are additional  changes  that  may  provide  further
improvements in emission control. The parameters being investigated include (1) the shape of the
chamber and the location of injection; (2) reduced crevice volumes;  and (3) compression ratio.
These parameters have been thoroughly  explored for highway engines and should  be  readily
adaptable to nonroad and marine engines.

    Efforts to redesign the shape of the combustion chamber and the location of the fuel injector for
highway and nonroad engines have been primarily focused on optimizing the relative motion of air
and injected fuel to simultaneously limit the formation of both NOx and PM. Piston crown design
must be carefully matched with injector spray pattern and pressure for optimal emission behavior.5
One strategy, reentrant piston bowl design, focuses on optimizing the radius of the combustion bowl,
the angle of the reentrant lip, and the ratio of the bowl diameter to bowl depth to optimize air motion.
An alternative is the use of higher pressure inj ection systems that decrease the need for turbulent in-
cylinder charge air motion.  While higher pressure  systems raise concerns of durability, there has
been a significant amount of progress in this area and it is expected that manufacturers will be able
to develop a durable system.6

    The second parameter being investigated is reducing crevice volumes by moving the location
of the top piston ring  relative to the top of the  piston.7  A reduced crevice volume can result in
reduced HC emissions and, to a lesser extent, reduced PM emissions. Costs associated with the
relocation of the top ring can be substantial because raising the top  of the piston ring requires
modified routing of the engine coolant through the engine block and lube oil routing under the piston
to prevent the raised ring from overheating. It is also  important to design a system that retains the
durability and structural integrity of the piston and piston ring assembly, which requires very precise
tolerances to avoid compromising engine lubrication.

    Compression ratio is another engine design parameter that impacts emission control.  In general,
higher compression ratios cause a reduction of cold start PM and improved fuel economy,  but can
also increase NOx.  Several methods can be employed to increase the compression ratio in an
existing diesel engine. Redesign of the piston crown or increasing the length of the connecting rod
or piston pin-to-crown length will raise the compression ratio by reducing the clearance volume.8
There is a limit to the benefit of higher compression ratios because of increased engine weight (for
durability) and frictional losses, which could somewhat limit fuel economy improvements.

    3. Swirl

    Increasing the turbulence of the intake air entering the combustion chamber (i.e., inducing swirl)
can reduce PM by improving the mixing of air and fuel in the combustion chamber. Historically,
swirl  was induced  by routing  the  intake air to achieve a circular motion in the cylinder.
Manufacturers are, however, increasingly using "reentrant" piston designs in which the top surface
of the piston is  cut out to allow fuel injection and air motion in a smaller cavity in the piston to

                                           30

-------
                                                 Chapter 3: Technological Feasibility
induce additional turbulence. Manufacturers are also changing to four valves per cylinder, which
reduces pumping losses and can also allow for intake air charge motion. The effect of swirl is often
engine-specific, but some general effects may be discussed.

    At low loads, increased swirl reduces HC, PM,  and smoke emissions and  lowers  fuel
consumption due to enhanced mixing of air and fuel. NOx emissions might increase slightly at low
loads as swirl increases.  At high loads, swirl  causes  slight decreases in PM emissions and  fuel
consumption, but NOx may increase because of the higher temperatures associated with enhanced
mixing and reduced wall impingement.9 A higher pressure fuel system can be used to offset some
of the negative effects of swirl, such as increased NOx, while enhancing the positive effects such as
a reduction in PM.10

B. Advanced Fuel Injection Controls

    Control of the many variables involved in fuel injection is central to any strategy to reduce
diesel engine emissions. The principal variables being investigated are injection pressure, nozzle
geometry (e.g., number of holes, hole size  and shape, and fuel spray angle), the timing of the start
of injection, and the rate of injection throughout the combustion process (e.g., rate shaping).

    Manufacturers continue to investigate new injector configurations  for nozzle geometry and
higher injection pressure (in excess of 2300 bar (34,000 psi)).11'12  Increasing injection pressure
achieves better atomization of the fuel droplets and enhances mixing of the fuel with the intake air
to achieve more complete combustion.  Though HC and PM are reduced, higher cylinder pressures
can lead to increased NOx formation.13 However, in conjunction with retarding the  start of fuel
injection,  higher fuel injection pressures can lead to reduced NOx because of lower combustion
temperatures.  HC, PM, or fuel economy penalties from this strategy can be avoided because the
termination of fuel inj ection need not be delayed. Nozzle geometry is used to optimize the fuel spray
pattern for a given combustion chamber design in order to improve mixing with the intake air and
to minimize fuel condensation on the combustion chamber surfaces.14 Minimizing the leakage of
fuel droplets is critical for reducing HC emissions.  Valve-closed orifice (VCO) tips are more
effective than sac-type nozzles, because they eliminate any droplets remaining after inj ection, which
would increase HC emissions.  Although VCO tips are subject to very high pressures, EPA believes
progress will continue in developing a durable injector tip prior to implementation of these standards.

    The most recent advances in fuel injection technology are the systems that use rate shaping or
multiple injections to vary the delivery of fuel over the course of a single injection. Igniting a small
quantity of fuel initially limits the characteristic rapid increase in pressure and temperature that leads
to high levels of NOx formation.  Injecting most of the fuel into an established flame then allows for
a steady burn that limits NOx emissions without increasing PM emissions. Rate shaping may be
done either mechanically or electronically.  Rate shaping has been shown to reduce NOx emissions
by up to 20 percent.15

    For electronically controlled engines, multiple injections may be used to shape the rate of fuel
injection into the combustion chamber. Recent advances in fuel system technology allow high-

                                           31

-------
Regulatory Impact Analysis
pressure multiple injections to be used to reduce NOx by 50 percent with no significant penalty in
PM. Two or three bursts of fuel can come from a single injector during the injection event.  The
most important variables for achieving maximum emission reductions with optimal fuel economy
using multiple inj ections are the delay preceding the final pulse and the duration of the final pulse.16
This strategy is most effective in conjunction with retarded timing, which leads to reduced NOx
emissions without the attendant increase in PM.

    A promising fuel  injection  design  is that developed by Caterpillar and  Navistar, the
Hydraulically actuated Electronically controlled Unit Inj ection (HEUI) system.17'18  The HEUI system
utilizes a common rail of pressurized oil to provide high injection pressures of 175 MPa (25,000 psi)
throughout an engine's  operating  range.  The HEUI  system provides full electronic control of
inj ection timing and duration, along with the possibility for rate shaping. The most attractive aspect
of this system is that it  operates largely independent  of engine speed so that the engine can be
optimized over a  larger range of operation. Some marine engine manufacturers are already utilizing
this system on production engines. It is expected that manufacturers will be  able to develop and
produce a full-authority electronic fuel  injection system for a reasonable cost  in time for these
standards. For larger engines (>1.5  liters/cylinder), Caterpillar has designed a Mechanically actuated
Electronically controlled Unit Injection (MEUI) system. MEUI is similar to  HEUI in that it can
control injection  pressure, timing  and rate shaping independent of engine speed. Caterpillar has
reported injection pressures as high as 200 MPa (30,000 psi) with the MEUI system.

    Common rail fuel systems may be used to gain fuel injection improvements over distributer
pump type systems.  Fuel is pressurized in the fuel  rails which allows high pressures independent
of engine speed. Cummins has recently developed an alternative technology known as the Cummins
Accumulator Pump (CAP) fuel system.19'20 The CAP system uses an accumulator between the high
pressure fuel injection pump and the distributor to achieve high pressure fuel injection of 145 MPa
(21,000 psi) and control timing independent of engine speed.  This system offers a low  cost
alternative to common rail fuel injection.

C. Improving Charge Air Characteristics

    Charge air compression (turbocharging) is primarily used to increase power output and reduce
fuel consumption from a given displacement engine. At rated power, a typical diesel engine loses
about 30 percent of its energy through the exhaust. A turbocharger uses the  waste energy in the
exhaust gas to drive a turbine linked to a centrifugal compressor, which then boosts the intake air
pressure. By forcing more air into the cylinder, more  fuel can also be added at the same air-fuel
ratio, resulting in higher power and better fuel consumption while controlling smoke and particulate
formation.

    In most recreational and light-duty commercial marine applications, the  exhaust system and
turbocharger are  generally water-jacketed to keep surface temperatures low.  In addition, raw water
is actually mixed with the exhaust  for cooling and tuning reasons. Because the exhaust is dumped
directly into the surrounding water, the more gradual cooling from mixing water with the exhaust
prevents a large pressure spike which could affect the exhaust tuning. The water j acketing causes

                                           32

-------
                                                  Chapter 3: Technological Feasibility
some of the reclaimed energy in the turbocharger to be lost to the cooling water.  This energy loss
is not large and could be reduced further by insulating the turbine from the water jacket. For larger
marine engines, insulation rather than water j acketing is generally used to keep surface temperatures
down.  For these engines, dry exhaust is directly emitted into the atmosphere.

    Aftercooling is often used to reduce NOx by reducing the temperature of the charge air after it
has been heated during compression. Reducing the charge air temperature directly reduces the peak
cylinder temperature during combustion which is  where NOx is formed. This technology was
initially developed to improve the specific power output of an engine by increasing the density of
air entering the combustion chamber. Water-to-air and air-to-air aftercooling are well established
for land-based engine applications.  For engines in marine vessels, there are two different types of
aftercooling strategies used: jacket-water, and raw-water aftercooling.  Because of the access that
marine engines have to a large cooling medium (i.e. oceans and lakes), EPA anticipates aftercooling
will play an important role in gaining significant NOx reductions from marine diesel engines.

    1. Jacket-water aftercooling

    Marine engines use jacket water systems to pull heat from the engine coolant. Raw water is
pumped into a heat exchanger which is used to cool the engine coolant. For most recreational and
light commercial applications, the  raw water is then used to cool the exhaust.  If jacket-water
aftercooling is applied, the coolant coming from the heat exchanger passes through the aftercooler
before entering the engine.  Through the use of jacket-water aftercooling, boost air may be cooled
to approximately 95-105ฐC. Another 40 to 50ฐC reduction in boost air temperature can be gained
by using a completely separate cooling circuit for the aftercooler.

    The heat exchanger is designed to resist the difficulties associated with operating in a sea-water
environment.  Water is pumped through the heat exchanger at a high velocity to help prevent
barnacle growth and to keep the passages clean.  In addition, copper is a poison to barnacles and the
heat exchangers are generally made of a copper-nickel alloy. To prevent sea water from causing the
metal in the heat exchanger to decay due to electrolysis, zinc anodes are used.

    In commercial applications where the water temperature is expected to be near freezing or very
dirty, keelcooling may be used.  In a keelcooled system, the engine coolant is passed through many
long tubes along the bottom of the boat which act as a heat exchanger.  Therefore, no ambient water
is brought into the vessel or cooling system.  This avoids problems that could be associated with
contaminants such as mud, chemicals, or ice. In displacement hull vessels keelcooling can be as
effective as a standard heat exchanger and is often used on commercial vessels. One disadvantage
of keel cooling is that it is a more complex and costly system than a standard shell and tube heat
exchanger.  Also,  keelcooling is less effective on a planing vessel,  and is not often  used for
recreational vessels.
                                           33

-------
Regulatory Impact Analysis
    2. Raw-water aftercooling

    Raw-water aftercooling means that outside water is pumped through the aftercooler rather than
engine coolant. Depending on the temperature of the water and the size of the aftercooler, raw-water
aftercooling can be used to cool the boost air to approximately 45-55ฐC. Because of the cooler
intake air temperatures associated with this technology, it can be used to achieve lower NOx levels
from a marine engine. The same strategies used to protect the jacket water heat exchanger from the
corrosive effects of sea water can also be used to protect a raw-water aftercooler. Currently, this
aftercooling strategy is only used in recreational and light commercial applications.   While
introducing raw-water aftercooling may require additional maintenance (replacing anodes), the
benefits of improved fuel efficiency, greater engine durability, and better control of NOx emissions
may lead to more widespread use in the future.
    3. Separate-circuit aftercooling

    Separate-circuit aftercooling is a compromise approach to marine engine aftercooling that shows
promise for addressing near-term emission standards. This system is much like current jacket-water
cooling systems, except that one coolant loop is devoted to engine cooling and a second loop is
devoted to removing heat from the aftercooler. The coolant loop devoted to the aftercooler provides
more effective cooling than jacket-water aftercooling, without introducing the durability concerns
of raw-water aftercooling.

    In principle, separate-circuit aftercooling is very similar to systems on highway truck engines.
The main difference between highway and marine applications is the fundamental cooling medium
(or heat sink). Highway engines use air with high relative speeds through a heat exchanger to extract
heat from the coolant running through the aftercooler. Marine engines have the same coolant loop,
but extract heat from a heat exchanger (or keel cooler) with ambient water.  Because water has a
higher convection coefficient than air (especially without high air speeds), it should generally more
efficient at  removing heat from a system.   Separate-circuit systems  should therefore have a
comparable level of cooling compared to highway engines, and  superior cooling relative to land-
based nonroad engines, which operate at slower speeds.  Separate-circuit aftercooling could even
overcool the engine intake air in some cases; a thermostat could be used to limit the cooling water
flow when the system temperature gets too low.
D. Electronic Control

    Various electronic control systems are in use or under development for nonroad, locomotive,
highway, and marine diesel engines.  Use of electronic controls enables designers to  implement
much more precise control of the fuel injection system and is necessary for advanced concepts such
as rate shaping. Through this precise control, trade-offs between various control strategies can be
minimized.  In addition, electronic controls  can be used to sense ambient conditions and engine
operation to maximize performance and minimize emissions over a wide range of conditions such

                                           34

-------
                                                 Chapter 3: Technological Feasibility
as transient operation of the engine. Electronic control is already used in limited marine applications
and has shown its ability to endure the marine environment. These same controls used today could
be used to optimize for emissions as well as performance.

E. Exhaust Gas Recirculation

    Exhaust gas recirculation (EGR) is a recent development in diesel engine control technology
for obtaining significant NOx reductions.  EGR reduces peak combustion chamber temperatures by
slowing reaction rates and absorbing some of the heat generated from combustion.  While NOx
emissions are reduced, PM and fuel consumption can be increased, especially at high loads, because
of the reduced oxygen available and longer burn times  during combustion.21'22  One method of
minimizing PM increases is to reduce the flow of recirculated gases during  high-load operation,
which would also prevent a loss in total power output from the engine. Recent experimental work
showed NOx reductions of about 50 percent, with little impact on PM emissions, using just six
percent EGR in conjunction with a strategy of multiple injections.23

    1. EGR cooling

    There are several methods of controlling any increase in PM emissions attributed to EGR. One
method is to cool the exhaust gas recirculated to the intake manifold. By cooling the recirculated
gas, it takes up less volume allowing more room for fresh intake air. With EGR cooling, a much
higher amount of exhaust gas can be added to the intake charge. At light loads, there can be a small
NOx penalty due to increased ignition delay, but at high loads, some additional NOx reduction may
result from EGR cooling.24 Another method to offset the negative impacts of EGR on PM is through
the use of high 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.25

    2. Soot removal

    Another challenge facing  manufacturers is  the potential negative effects of soot from the
recirculated  exhaust being routed into the intake stream. Soot may form deposits in the intake
system, which could cause wear on the turbocharger or decrease the efficiency of the aftercooler.
As the amount of soot in the cylinder increases, so does the amount of soot that works its way past
the piston rings into the lubricating oil, which can lead to increased engine wear.  One thing that has
been developed to reduce soot in the recirculated exhaust gas is a low-voltage soot removal device.26
Engine wear was shown to be greatly reduced as a result of this device.  Another strategy is to
recirculate the exhaust gas after it has passed through a particulate trap or filter. Demonstrations
have shown that some prototype traps can remove more than 90 percent of particulate matter.27

    One study discusses a technology package designed to solve the problems of minimizing the
amount of intake  charge displaced by  exhaust gas and of fouling of the  turbocharger and
aftercooler.28 This technology package uses a variable geometry turbocharger, an EGR control valve,
and a venturi mixer to introduce the recirculated gas into the inlet stream after the intake air is
charged and  cooled. The variable geometry turbocharger is used to build up pressure in the exhaust

                                           35

-------
Regulatory Impact Analysis
stream.  Once the pressure is high enough, the EGR control valve is opened and the recirculated gas
is mixed into the high pressure inlet stream in a venturi mixer.  Although the recirculated gas is
cooled, this cooling is minimal to prevent both fouling in the cooler (due to condensation) and a
large pressure drop across the cooler.

    Several bypass filtration designs exist to filter smaller particles out of engine oil.29 With bypass
filtration, a portion of the oil is run through a secondary unit which results in well filtered oil. This
type of filtration system could be used to minimize negative effects of soot in the oil that are
associated with high levels of EGR. At least one  design claims efficiencies of up to 99 percent 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.

F. Exhaust Aftertreatment Devices

    Researchers in industry and academia have explored various technologies for treating engine-out
exhaust emissions. In general, EPA does not expect that marine engine manufacturers will need to
utilize exhaust aftertreatment to meet the new emission standards; however, further work on these
technologies may lead to development of an approach that provides effective control at a lower cost
than today's anticipate technologies. One technical difficulty to investigate is the concern of catalyst
poisoning in a salt-water environment.

    1. Oxidation catalysts

    The flow-through oxidation catalyst provides relatively moderate PM reductions by oxidizing
both gaseous hydrocarbons and the portion of PM known as the soluble organic fraction (SOF). The
SOF consists of hydrocarbons adsorbed to the carbonaceous solid particles and may also include
hydrocarbons that have condensed into droplets of liquid.34  The carbon portion of the PM remains
largely unaffected by the catalyst.  Although recent combustion chamber modifications have reduced
SOF emissions, the SOF still comprises between 30 and 60 percent  of the total mass  of PM.
Catalyst efficiency for SOF varies with exhaust temperature, ranging from about 50 percent at 150ฐC
to more than 90 percent above 350ฐC.30

    Another challenge facing catalyst manufacturers is the formation of sulfates in the exhaust. This
is especially true for marine engines which tend to use fuels with higher sulfur levels than land-based
engines.  At higher exhaust temperatures, catalysts have a greater tendency to oxidize sulfur dioxide
to form sulfates, which contribute to total PM emissions.  For highway applications, this is helped
by the introduction of low-sulfur  fuel. In addition, catalyst manufacturers have been successful in
developing catalyst formulations that minimize sulfate formation.31 Catalyst manufacturers have also
adjusted the placement of the catalyst to a position where the needed SOF reduction is achieved, but
sulfate formation is  minimized.32 Marine fuel with sulfur concentrations higher than 0.05 weight
percent may prevent the use of more active oxidation catalysts with higher conversion efficiencies.
                                           36

-------
                                                 Chapter 3: Technological Feasibility
    2. Particulate traps

    Use of a particulate trap is a very effective way of reducing participate emissions, including the
carbon portion. Particulate traps have been extensively developed for highway applications, though
very few engines have been sold equipped with traps, primarily because of the complexity of the
systems needed to remove the collected particulate matter. Continued efforts in this area may lead
to simpler, more durable designs that  control emissions cost-effectively.  Research in this area is
focused on developing new filter materials and regeneration methods. Some designs rely on an
additive acting as a catalyst to promote spontaneous oxidation for regeneration, while other designs
aim to improve an active regeneration strategy with microwave or other burner technology.

    3. Selective Catalytic Reduction

    Selective catalytic reduction (SCR) is one of the most effective, but also most complex and
expensive, means of reducing NOx from large diesel engines. Emission reductions in excess of 90
percent can be achieved using SCR.33 In SCR systems, a reducing agent, such as ammonia, is
injected into the exhaust and both are  channeled through a catalyst where NOx emissions are
reduced. These systems are being successfully used for large stationary source applications which
operate under constant, high load conditions.

    A number of disadvantages are apparent for the use of current technology SCR systems  on
ships.  The SCR system is effective only over a narrow range of exhaust temperatures.  The
effectiveness of the system is decreased at reduced temperatures exhibited during engine operation
at partial loads. Also, excess ammonia in the exhaust can occur during transient operation, where
control of optimum ammonia injection is difficult.  However, because marine engines (especially
larger engines) operate under fairly steady-state conditions, this "ammonia slip" may be less of a
problem.

G. Water Emulsification

    Water emulsification of the  fuel is a technique which also lowers maximum combustion
temperature without an increase in fuel consumption. Water has a high heat capacity, which allows
it to absorb enough of the energy in the cylinder to reduce peak combustion temperatures. There are
at least two ways to accomplish the emulsification during combustion:  in the combustion chamber
or in the fuel tank.  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.34 Water dilution does have significant
challenges. Combining water and fuel for the first time in the chamber requires significant changes
to the cylinder head to add an injector. Using a single injector with stratified water and fuel adds
complexity to the injection system.  Combining water with fuel in the tank  may introduce
combustion problems due to unstable emulsion.  Also, this technique requires a  significantly
redesigned fuel handling system to overcome the potential risk of corrosion and to maintain power
output.  However, these problems may be overcome in the future as the strategy is refined. In any
event, extra liquid storage availability  is necessary to retain similar range.
                                           37

-------
Regulatory Impact Analysis
VI. Emission Measurement

    In any program designed to achieve emissions reductions from internal combustion engines, the
test procedures that  are used to determine emissions levels are as important as the emissions
standards  that are implemented.  Five duty cycles were investigated for use in this rule:  three
designed for propulsion marine diesel engines, one used for land-based nonroad diesel engines, and
one intended for constant-speed auxiliary diesel engines.   These cycles are designated by the
International Standards Organization (ISO) as E2, E3, E5, Cl, and D2, respectively.35 This section
discusses duty cycles for testing as well as other issues specific to testing marine diesel engines.

A.  Certification Duty Cycles

    The E3 duty cycle is designated for propulsion marine diesel engines operating on a propeller
curve.  Both the E3 and the E5 are designed to represent marine diesel engines, and both cycles focus
on operational modes which lie on a propeller curve.  However, the E3 represents heavy-duty diesel
marine engine operation, while the E5 represents diesel marine engine operation on vessels less than
24 meters  in length. In addition, the E3 operates at a higher average power than the E5.  These two
facts imply that the E3 is probably more representative for commercial while the E5 is probably more
representative for recreational marine diesel engines.  For this reason, EPA is using the E3 duty cycle
for this program. Using a single cycle to represent all propeller-curve marine diesel engines should
reduce certification burdens for marine engines that are used in both vessels under and over 24
meters in length. More detail  on this issue  may be found in the docket.36

    For auxiliary marine diesel engines, ISO specifies the C1 and D2 duty cycles. Auxiliary marine
engines include any engines used on a marine vessel that are not primarily used for propulsion. The
C1 duty cycle was developed for nonroad diesel engines and applies to variable-speed engines while
the D2 applies to constant-speed engines. The D2 duty cycle is a five-mode cycle at a constant-speed
which is intended for generator sets with an intermittent load and is currently allowed as an option
for the certification of land-based constant-speed engines.

    Many larger propulsion marine engines do not operate on a propeller curve. These engines may
run at a constant speed and use a variable-pitch propeller to control vessel speed. The E2 constant-
speed propulsion marine duty cycle applies to these engines. Figure 3-1 presents the five duty cycles
discussed  above.

    There is another class of propulsion engines that run at variable speed and use a variable pitched
propeller.  These engines are designed to operate near the power curve for the engine to maximize
fuel efficiency. In general, these engines will operate at a constant speed near peak torque, except
when maneuvering in port, where they operate along the lug curve. Because of the expense of the
system, variable-speed engines are rarely used with variable-pitche propellers. ISO does not have
a test duty cycle specifically designed for these engines.  EPA believes that the most appropriate
existing duty for these engines is the ISO Cl  duty cycle.  This is consistent with MARPOL
Annex VI.
                                           38

-------
  100%
0
g.  80%
   60%

   40%

   20%

    0%
0
g.  80%
"S  60%
N
"ro
E  40%
o
z
   20%

    0%
           Figure 3-1: CI Marine Test Duty Cycles Plotted Under the Full Load Torque Curve
               (percentages are modal weightings used to calculate a composite emissions level)
                ISO E3 Commercial Marino
             Idle
 Intermediate
Engine Speed
                                        Rated
                ISO E5 Recreational Marine
                           32%ป
           30ฐ/
             Idle
 Intermediate
Engine Speed
                                        Rated
                                              ISO C1 Land-Based CI Nonroad
                                                         100%
                                0
                                g.  80%
                                   60%

                                   40%

                                   20%

                                    0%
                                                                   15ฐ/
    Idle
 Intermediate
Engine Speed
Rated
                                 ISO E2 Variable Propeller
                                  80%
                               -a  60%
                               0
                               N
                               "ro
                               E  40%
                               o
                                                         20%
                                                          0%
   Rated
Engine Speed
                                                                                      ISO D2 Constant Speec
                                                                                    100% -
                                                                                              Rated
                                                                                            Engine Speed

-------
Regulatory Impact Analysis
B. Relative Stringency of Duty Cycles

    In general, emissions levels are sensitive to the duty cycle used to generate them. For instance,
if emissions from a given engine were lower on the marine cycle than on the land-based nonroad
duty cycle, it would suggest that lower standards would be necessary to ensure the same level of
control from marine engines as will be achieved from nonroad land-based diesel engines. However,
according to EPA's analysis in the case of Cl versus E3 and the locomotive line-haul cycle versus
E2, emissions tend to be roughly equivalent. This analysis is described below.

    1. Category 1 marine vs. land-based test procedure

    Test data on nine uncontrolled Category 1 marine diesel engines37'38'39 and two nonroad diesel
engines40 show that HC+NOx emissions are approximately the same (2% lower on the E3) when
measured on the E3 or Cl duty cycle. This suggests that emission levels determined using the E3
cycle are roughly equivalent to those determined using the Cl cycle.  PM and CO emissions were
measured to be, respectively, 30% and 46% lower on the E3 than Cl.  For controlled marine diesel
engines, EPA anticipates that it will be somewhat easier to demonstrate compliance on the E3 cycle
than the Cl cycle because of the lower measured PM and CO , the E3 only has half as many test
modes, and the E3 does not include the peak torque mode that drives many nonroad diesel  engine
emission control designs. Figure 3-2 presents these results.

              Figure 3-2:  Ratio of Emissions on the E3 vs. Cl Duty Cycles
   c/)
                 HC+NOx
PM
                                          40

-------
                                                 Chapter 3: Technological Feasibility
    For constant-speed engines, both the nonroad and marine standards, the test procedure does
not impact the stringency. Both rules use the D2 duty cycle for constant-speed engines.

    2. Category 2 marine vs. locomotive line-haul

    Category 2 emission standards are based on the line-haul standards for locomotive engines.
Although locomotive engines must certify to two duty cycles with separate standards, EPA focused
on the line-haul  standard as being more applicable to marine engines. This is because the line-haul
duty cycle has a high average power similar to the commercial marine diesel duty cycles. Another
similarity between the  locomotive and  marine duty cycles is that they are based on a path of
operation under the torque curve rather than over the whole torque curve. Locomotive engines run
at eight distinct operating points or "notches" in real world use while marine engines generally either
operate over a propeller curve or at constant speed.

    To compare the relative stringency of the marine E2 and E3 and the locomotive line-haul duty
cycles, EPA looked at baseline emission data based on these cycles. Data on thirteen locomotive
engines41 and twenty-three marine engines42'43'44'45 are presented in Figure 3-2. This data clearly
shows that combined HC and NOx baseline emissions from engines tested on the three duty cycles
are at about the  same levels.
                                           41

-------
Regulatory Impact Analysis
   Figure 3-3: Comparison of HC+NOx Emissions for Marine and Locomotive Engines
25
20
"
1 15
.*:
S
X
O
5 10
5
0
(


\
• •$** *X
• , i^ ฐ ฐ
o ^ ^^
% 1
••


	

D 1000 2000 3000 4000 5000 6000
500 1500 2500 3500 4500 5500
Rated Power [kW]
• Marine Propeller Law E3* Marine Constant Speed E^ Locomotive Line-Haul


                                     42

-------
                                                 Chapter 3: Technological Feasibility
C. Not-to-Exceed Provisions

    EPA is concerned that if a marine engine is designed for low emissions on average over a low
number of discrete test points, it may not necessarily operate with low emissions in-use. This is due
to a range of speed and load combinations that can occur on a vessel which do not necessarily lie on
the test duty cycles. For instance, the test modes for the E3 cycle lie on an average propeller curve.
However, a propulsion marine engine may never be fitted with an "average propeller."  In addition,
a light vessel with a planing hull may operate at lower torques than average while the same engine
operated on a heavy vessel with a deep displacement hull may operate at higher torques than average.
In addition, a planing hull vessel can operate at high torques at low speed prior to planing.

    It is EPA's intent that an engine operates with low emissions under all in-use speed and load
combinations that can occur on a vessel, rather than just at the discrete test modes in the selected
duty cycles. To ensure this, EPA is establishing  requirements that extend to typical in-use operation.
For propulsion marine engines certified to the  E3  duty cycle, EPA is applying a "not-to-exceed"
(NTE) zone based on the maximum power  curve  of the engine.   Under  this provision, the
manufacturers  must design their engines to comply to a not-to-exceed limit, tied to the standard, for
all of the regulated pollutants within the NTE zone.  In cases where the engine is  included in
averaging, banking, or trading of credits, the not-to-exceed limits are tied to the family emission
limits. EPA reserves the right to test an  engine in a lab or installed on a vessel to confirm
compliance to  this requirement.

    When testing the engine within the NTE zone, only steady-state operation will be considered.
It is unlikely that transient operation is necessary under the NTE concept to ensure that emissions
reductions are  achieved.  EPA designed the NTE zones to contain the operation near an assumed
propeller curve that the steady-state cycles represent. EPA believes that the vast maj ority of marine
operation in the NTE zone is steady-state.  For planing vessels, EPA  believes that the transient
operation as a vessel comes to plane generally is along the torque curve and would not be within the
NTE zone.  However, EPA does not have enough data to reliably say where under the torque curve
marine engines operate during transient operation.  Also, EPA does not believe that the NTE zone
should be extended to include areas an engine may see during transient operation but would never
see during steady-state operation.   For this reason, EPA does not believe that adding transient
operation to the NTE requirements is necessary at this time. EPA would revise its opinion in the
future if there  were evidence that in-use emissions were increased due to insufficient emission
control under transient operation.

    This NTE zone concept is consistent with the recent guidance for highway engines.46 However,
highway engines  have a much  larger NTE zone  which covers a full range  of possible engine
operation. In addition, highway engines have the additional requirement of a second cap shaped by
the emission results from the 13 mode EURO test.  For marine engines, the NTE zone is much
smaller and does not include peak torque or high-speed/low power operation.  In addition, marine
engines would not be subject to transient operation during NTE testing.
                                           43

-------
Regulatory Impact Analysis
    1.  Shape of the NTE zones

    For engines certified on the E3 cycle, the NTE zone is defined by the maximum power curve,
actual propeller curves, and speed and load limits. The E3 duty cycle itself is based on a cubic
relationship between speed and power which intersects the maximum test power point of the engine.
For the NTE zone, the upper boundary is based on a speed-squared propeller curve passing through
the 115  percent load point at maximum test speed and the lower boundary is based on a speed to the
fourth propeller curve passing through the  85 percent load point at maximum test speed.  These
propeller curves are believed to represent the full range of propeller curves seen in use (these curves
are adjusted  slightly  for Category 2  engines, as described in the Summary and Analysis of
Comments).47 To prevent imposing an unrealistic cap on a brake-specific basis, EPA has limited this
region to power at or above 25 percent of maximum test power and speeds at or above 63 percent
of maximum test speed. These limits are consistent with Mode 4 in the E3 duty cycle. The NTE
zone is illustrated in Figure 3-4.  Also,  the zone is split into two subzones with different emissions
limits. These limits are discussed below. For engines operating in applications in which the in-use
operation differs significantly from the NTE zone in Figure 3-3, manufacturers may petition for
adjustments to the NTE zone for a given engine design.

              Figure 3-4: Not-to-Exceed Zone for Category 1 E3 Engines
   o
   Q.
   CD
100%-
 90%
 80%
 70% H
 60%
   E   50%
   D)
   C
   CD
 30%
 20%
 10%
   0%
                            25% power
            50%     60%
70%      80%      90%
   engine speed
                                                            100%
                                        44

-------
                                                 Chapter 3: Technological Feasibility
    For variable-speed engines with variable pitch propellers, theNTE zone includes the entire NTE
zone for engines certified on the E3 duty cycle. In addition, this zone includes the range of operation
above 1.15 times the squared propeller curve for speeds greater than or equal to 63% of maximum
test speed. This additional portion of the NTE zone represents operation near the lug curve that EPA
understands is common for these applications. As with the NTE zone for engines certified to the E3
duty cycle, this zone is split into two subzones above and below 45% power with limits that are
discussed below.

    For constant-speed propulsion engines certified to the E2 duty cycle, EPA is using a similar
approach to  ensure that  emissions measured on the duty cycle  are representative  of emissions
generated in use.  Because these engines operate at a constant speed, the NTE zone is easier to define
than for the E3  duty cycle. The not-to-exceed limits apply to the speed where the engine is designed
to operate for all loads greater than or equal to 25 percent of maximum load at that speed. Because
a constant speed can actually operate over a small range of engine speeds in-use, the NTE zone
includes this small range of speeds.  Below 25 percent load, EPA is not adopting aNTE cap because
brake-specific emissions become large at low loads due to a small power level in the denominator.
In addition, these engines generally do not spend much time operating at low loads.  As with the
NTE zone for propeller-curve engines, this zone is split into two  subzones above and below 45%
power with limits that are discussed below. Also, any variation in speed possible for the engine in
use must  be considered in the NTE zone.

    EPA is not adopting any NTE provisions for auxiliary marine engines certified using the D2 and
Cl duty cycles. EPA does not at this time have enough information on the operation and design of
these engines.  However, EPA will investigate typical emissions from auxiliary marine engines off
of the D2 and Cl duty cycles and act appropriately.  Finally, because the impact of NTE is not fully
understood for very large propulsion engines and to be consistent with the MARPOL requirements,
EPA does not believe that it is appropriate to extend the NTE concept to Category 3 engines at the
present time.

    2.  Emissions limits for the NTE zone

    EPA is requiring emissions caps for the NTE zones which  represent a multiplier times the
weighted test result used for certification. Although ideally the engine should meet the certification
level throughout the NTE  zone, EPA understands that a  cap of 1.0 times the standard is not
reasonable, because there is inevitably  some variation in emissions over the range of engine
operation. This is consistent with the concept of a weighted modal emission test such as the steady-
state tests included in this rule.

    For propulsion engines certified to the E3,  Cl, and E2 duty cycles, EPA believes that not-to-
exceed limits of 1.2 times the emissions standard (or FEL) is appropriate for the subzone at or above
45% of maximum test power. Below 45% of maximum test power, the cap is 1.5.  EPA's decision
not to apply a tighter cap for marine engines is based on all of the emissions data in this chapter for
which individual modal results were available to EPA.  Figure 3-5 presents the ratio of the modal
emissions to the cycle weighted emissions from  more than 30 Category 1 and 2 marine diesel

                                          45

-------
Regulatory Impact Analysis
engines.13 This data shows that the 1.2 cap is easily achievable for HC+NOx at higher power, but
may be more challenging at low powers (mode 4). This is why the cap is relaxed to 1.5 below 45%
of maximum test power. Most of the engines are below these caps, and EPA believes that all marine
diesel engines can be optimized to meet these caps, particularly as the marine engine market shifts
more towards electronically controlled engines. EPA only had modal data for CO and PM on the
E3 duty cycle.  The separate caps in the NTE zone also appear to be necessary for PM. About half
of the engines already meet these limits and EPA believes that all engines could be designed for this
requirement. It should be noted that the CO emissions data presented here were well below the new
emission standards, so CO should not be a limiting factor.

    EPA also collected modal data on a mechanically controlled and an electronically controlled
commercial marine diesel engine.48  This data is presented in Figures 3-6 and 3-7 which show
operation under the whole NTE zone rather than just at four points.  The mechanically controlled
engine is very close to meeting the NTE requirement at all points, and at most points is well below
the cap. For PM, this engine is above the NTE requirement along the minimum load boundary for
the NTE zone. However this engine is not calibrated for emissions control.  We believe that if the
engine is designed to operate at these points, that it is important that manufacturers design the engine
to achieve emissions levels at these points near the certified level.

    For HC+NOx, the electronically controlled marine engine is well below the NTE caps for the
whole zone.  For PM, this engine really only has difficulty with the NTE caps at full  power.
However, this engine is not calibrated for emissions control. In addition, data on other commercial
marine engines shown in Figure 3-5 suggest that most commercial marine engines do not have this
problem at full power.
       b This data is cited and presented in more detail elsewhere in this chapter.

                                          46

-------
Figure 3-5: Ratio of Modal Emissions to Cycle Weighted Emissions
2.2
2
1.8
ฃ1.6
|1.4
ra 1.2
F 1
| 0.8
S 0.6
0.4
0.2
0
HC+NOx E3 Ratio
Ratio of Modal Emissions to the Weighted Average
1
1


1








1 L











('




r








n





100; 100 91; 75 80; 50 63; 25
% rated speed; % rated power
CO E3 Ratio
Ratio of Modal Emissions to the Weighted Average
20
.0
2C 1
.D
2 A
.4
2f\
.f.
"^- o
_c f.
^> A A
| 1.8 -
^- A c r
^) 1.6
^ -1 A
-^1.4
-C A n
^1.2
> A
=^ 1
O) n o
^^ O.O
Oc
.b
.4
0.2
0. _.






J
1 ^T
n
' T 1 rfl
_f ' rf
]
= = : ; 3



r
r







-
100; 100 91; 75 80; 50 63; 25
% rated speed; % rated power
                                                      HC+NOx E2 Ratio
                                             Ratio of Modal Emissions to the Weighted Average
                                            100;100     91;75     80;50     63; 25
                                                   % rated speed; % rated power
                                                         PM E3 Ratio
                                             Ratio of Modal Emissions to the Weighted Average
                                            100;100     91;75     80;50     63; 25
                                                   % rated speed; % rated power

-------
Regulatory Impact Analysis
       Figure 3-6: Ratio of Modal to E3 Emissions Compared to the NTE Zone

             for an Electronically Controlled Marine Diesel Engine
      Modal Variation in HC+NOx Compared to E3
                   Mechanically Controlled Marine Diesel Engine
     100% -
  s_
  
-------
                                    Chapter 3: Technological Feasibility
     Figure 3-7: Ratio of Modal to E3 Emissions Compared to the NTE Zone

            For an Electronically Controlled Marine Diesel Engine
        Modal Variation in HC+NOx Compared to E3

                  Electronically Controlled Marine Diesel Engine
   100%
,_   80%

-------
Regulatory Impact Analysis
    The use of electronic controls should give marine engine manufactures more flexibility in how
they meet the cap under all operation compared with mechanically controlled engines with fixed
injection timing.  EPA is  concerned, however, that electronic controls (or any other Auxiliary
Emission Control Devices) not be used in such a way as to result in higher emissions from marine
engines in use than would be seen during testing. EPA believes that the best way of ensuring this
is the use of the NTE provisions.  In addition, however, EPA believes that it is appropriate to require
that the manufacturers provide information such as message or parameter identification, scaling,
limit,  offset, and transfer function information required to interpret all messages and parameters
broadcast on an engine's controller area network. In addition, requiring that electronically controlled
engines broadcast their engine speeds and loads on the controller area networks would facilitate in-
use testing.

    EPA recognizes the difficulties of complying to the caps over an NTE zone with a mechanical
engine due to the restriction of a fixed injection timing. However, the split zone with different caps
should  address this  difficulty.   Nevertheless, the  requirement  for effective in-use emission
performance over the broad range  of in-use operation should not be restricted to electronically
controlled engines.  Even  with mechanical controls, EPA believes that it is feasible for marine
engines to meet the new requirements through optimization of the engine calibration.

    3. Ambient Conditions

    air temperature and humidity

    Ambient air conditions, including temperature  and  humidity, have  a significant effect on
emissions from marine engines in use. To ensure real world emissions control, the NTE zone testing
should include a wide range of ambient air conditions representative of real world conditions. EPA
believes that the appropriate ranges should be 13-30ฐC (55-86ฐF) for air temperature and 7.1-10.7
grams water per kilogram of dry  air (50-75  grains/pound of dry air) for air humidity. The air
temperature ranges were based on temperatures seen during ozone exceedences, except that the
upper end of the air temperature range has been adjusted to account for the cooling effect of a body
of water on the air above it.49  We are also aware, however,  that marine engines sometimes draw
their intake air from an engine compartment or engine room with such that intake air temperatures
are substantially higher than fresh air temperatures. We are therefore retaining 35ฐC as the upper
end of the specified range for engines that don't draw their intake air from the outdoor ambient. The
air temperature and humidity ranges are otherwise consistent with those developed for NTE testing
of highway heavy-duty diesel engines.

    For NTE testing in which the air temperature or humidity is outside of the range, the emissions
must be corrected back to the specified air temperature or humidity range.  These corrections must
be consistent with the equations  in 40 CFR Part 89, Subpart E except that these equations correct
to 25 ฐC and 10.7 grams per kilogram of dry air while corrections associated with the NTE testing
shall be to the nearest outside edge of the specified ranges.  For instance, if the outdoor ambient
temperature were higher than 30ฐC for an engine with a that drew fresh outdoor air into the intake,
                                           50

-------
                                                 Chapter 3: Technological Feasibility
a temperature correction factor may be applied to the emissions results to determine what the
emissions would be at 30ฐC.

    EPA believes that these ranges of ambient air conditions will have a small effect on emissions
within the NTE requirements. For commercial marine engines using aftercooling, the charge air
temperature is insensitive to ambient air temperature compared to  the cooling effect of the
aftercooler. SwRI tested this theory and found that when the ambient air temperature was increased
from 21.9 to 32.2ฐC, the cooling water to the aftercooler of a commercial marine engine only had
to be reduced by 0.5 ฐC to maintain a constant charge air temperature.50 At the same time, the effect
of varying temperatures on turbocharger performance and the corresponding effect on PM emissions
is not well understood.  According to the CFR correction factor, there is only a ฑ3% variation in
NOx in the NTE humidity range.

    water temperature

    Ambient water temperature also may affect emissions due to its impact on engine and charge
air cooling. For this reason, NTE testing includes a range of ambient water temperatures from 5 to
27ฐC (41 to 80ฐF). The water temperature range is based on temperatures that marine engines
experience in the U. S. in use. Although marine engines experience water temperatures near freezing,
EPA does not believe that additional emission control will be gained by lowering the minimum water
temperature below 5ฐC. At  this time, EPA  is not aware  of an established correction factor for
ambient water temperature. For this reason NTE zone testing must be within the specified ambient
water temperature range.

    EPA does not believe that the range of ambient water temperatures discussed above will have
a significant effect on the stringency of the NTE requirements. Following the normal engine test
practice recommended by SAE51 for aftercooled engines, the cooling water temperature would be
set to 25ฑ5 ฐC. This upper portion of the NTE temperature range is within the range suggested by
SAE for engine testing.  For lower temperatures, manufacturers would be able to use a thermostat
or other temperature regulating device to ensure that the charge air is not overcooled.  In addition,
the SAE practice presents data from four aftercooled diesel engines on the effects of cooling medium
temperature on emissions. For  every 5ฐC increase in temperature, HC decreases  1.8%,  NOx
increases 0.6%, and PM increases 0.1%.

    EPA is aware that many marine engine installations  are designed for operation in a given
climate.  For example, to stay within the engine manufacturer's specified operating temperatures,
a fishing vessel operating in Alaska does not need to be designed for 27ฐC water temperatures. In
this case, the boat builder could take advantage of the cooler water to build a smaller heat-exchanger
system.  This vessel operating in a warmer climate would likely not  stay within the specified ranges
for the engine. This would likely increase engine emissions, but could also drastically shorten the
engine's operating life. EPA will perform in-use testing on engines where they operate. If we find
a vessel that operates in a way that does not meet the engine manufacturer's specifications, we may
pursue an enforcement action against the boat builder or the vessel operator.
                                          51

-------
Regulatory Impact Analysis
    In addition, EPA understands that there are times when someone may need to compromise
emission control for startability or safety reasons. Manufacturers aren't responsible for the NTE
requirements under start-up conditions. In addition, manufacturers may petition to be exempt from
emission-control requirements under specified extreme conditions such as engine overheating where
emissions may increase under the engine protection strategy.

D. Emissions Sampling

    Aside from the duty cycle, the test procedures for CI marine engines are similar to those for
land-based nonroad diesel engines. However, there are a few other aspects of marine diesel engine
testing that need to be considered.  Many recreational,  auxiliary, and light-commercial  marine
engines mix cooling water into the exhaust.  This exhaust cooling is generally done to keep surface
temperatures low for  safety reasons and to tune the exhaust for performance and noise.  Since the
exhaust must be dry for dilute emission sampling, the cooling water must be routed away from the
exhaust in a test engine.

    Even though many marine engines exhaust their emissions directly into the water, EPA has
based its test procedures and associated standards on the emissions levels in the  "dry" exhaust.
Relatively little is known about water scrubbing of emissions.  EPA must therefore consider all
pollutants out of the  engine to be a risk to public health.  Additionally, EPA is not aware of a
repeatable laboratory test procedure for measuring "wet" emissions. This sort of testing is nearly
impossible from a vessel in-use.  Finally, a large share of the emissions from this  category come
from large engines which emit their exhaust directly to the atmosphere.

    The  established method for sampling emissions is through the use of full dilution sampling.
However, for larger engines  the exhaust flows  become  so large that  conventional dilute testing
requires a very large and costly dilution tunnel. One option for these engines is to use a partial dilute
sampling method in which only a portion of the exhaust is sampled. It is important that the partial
sample be representative of the total exhaust flow. The total flow of exhaust can be determined by
measuring fuel flow and balancing the carbon atoms in and out of the engine. For guidance on
shipboard testing,  the MARPOL NOx Technical  Code specifies analytical instruments, test
procedures,  and  data   reduction techniques  for  performing test-bed  and  in-use  emission
measurements.52 Partial dilution sampling methods can provide accurate steady-state measurements
and show great promise for measuring transient emissions in the near future. EPA intends to pursue
development of this method and put it in place prior to the date that the standards in this final rule
become enforceable.

      Because marine engines often become an integral part  of a vessel and  cannot easily be
removed, EPA reserves the option for in-use testing to be performed on the vessel.  There are several
portable sampling systems on the market that, if used carefully, can give fairly accurate results. For
Category 3  engines, and some Category 2 engines, it may not be feasible to test the engines on a
dynamometer. When this is the case, it may be more appropriate to test the engine aboard the vessel.
For these vessels, there should be enough space to bring a portable sampler aboard similar to those
used for stationary source testing.  Engine speed can be monitored directly, but load may have to be

                                          52

-------
                                                 Chapter 3: Technological Feasibility
determined indirectly. For constant speed engines, it should be relatively easy to set the engine to
the points specified in the duty cycles.
VII. Baseline Technology Mix

A. Category 1 Marine Diesel Engines

    EPA developed estimates of the current mix of technology for Category 1 marine diesel engines
based on  data from  the  1997 Power Systems Research (PSR) OE Link database and from
conversations with marine manufacturers. Based on this information, EPA estimates that 70 percent
of new  marine engines are turbocharged, and  70  percent of these turbocharged engines  use
aftercooling.  The majority of these engines are four-strokes, but about 20 percent of Category 1
engines  are two-strokes.  Although indirect injection (TDI) is mostly used with smaller diesels,
engines  with IDI account for about 10 percent of total CI sales for the past five years. Electronic
controls have only recently been introduced into the marine market place; however, EPA anticipates
that their use will increase as customers realize the performance benefits associated with electronic
controls and as the natural migration of technology from highway to nonroad to marine occurs.
Table 3-1 provides more detail on the baseline technology mix.

                                        Table 3-1
               Baseline Technology Mix of Category 1 Marine Diesel Engines
Technology
natural aspiration
turbocharger
no aftercooler
jacket- water aftercooler
raw- water aftercooler
direct injection
indirect injection
mechanical control
electronic control
Auxiliary
47%
53%
70%
21%
9%
78%
22%
97%
3%
Recreational
29%
71%
48%
46%
6%
90%
10%
99%
1%
Commercial
31%
69%
53%
41%
6%
96%
4%
100%
0%
Aggregate
30%
70%
50%
44%
6%
91%
9%
99%
1%
    With regard to baseline emissions, data from eight high-speed marine diesel engines are
presented in Table 3-2 and compared to the new emission standards.53'54'55'56'57'58 For the eleven of
the propulsion marine engines, the results are based on the E3 test cycle; for the other two, results
were only available on the Cl cycle. The D2 test cycle was used for the auxiliary marine engine.
                                           53

-------
Regulatory Impact Analysis
PM emissions were not sampled from all of the engines. This data shows to what extent emissions
need to be reduced from the levels of today's Category 1 marine diesel engines to meet the new
emission standards.  On average, EPA is requiring significant reductions in HC+NOx and PM;
however, the CO standards will act as a cap to prevent increased emissions..
                                         54

-------
                                                 Chapter 3: Technological Feasibility
                                       Table 3-2
             Emissions Data from Baseline Category 1 Marine Diesel Engines
Rated
Power
(kW)
20
71*
134
157
199
216
235
237
262
265
280
336
447
746
895
Technology Mix
naturally aspirated
indirect injection
naturally aspirated
turbocharged
naturally aspirated
turbocharged, aftercooled,
electronic injection
turbocharged
turbocharged
jacket- water aftercooled
turbocharged
turbocharged
raw- water aftercooled
turbocharged
jacket- water aftercooled
turbocharged
jacket- water aftercooled
turbocharged
turbocharged
jacket- water aftercooled
turbocharged, unit injection,
jacket- water aftercooled
turbocharged
jacket- water aftercooled
Emissions Data g/kW-hr
HC
0.27
3.45
0.28
—
0.16
—
—
0.03
0.37
0.35
—
0.09
0.52
0.39
0.27
NOx
6.6
8.3
8.4
10.8
10.3
7.5
8.9
9.1
7.8
9.4
7.8
7.0
9.7
16.2
9.7
CO
2.8
11.5
0.44
—
1.2
—
—
1.4
0.7
0.6
—
1.6
1.7
2.6
1.3
PM
0.99
1.17
0.16
—
0.21
—
—
—
—
—
—
0.17
0.17
—
—
New Standards
HC+NOx
—
7.5
7.2
CO
—
5.0
5.0
PM
—
0.4
0.2
Because this is an auxiliary marine engine, this data is based on the D2 test cycle.
                                          55

-------
Regulatory Impact Analysis
B. Category 2 Marine Diesel Engines

    Category 2 marine engines are essentially marinized locomotive two- and four-stroke engines.
Two-stroke engines currently make up the vast majority of the fleet; however, four-stroke engines
are expected to gain market share as locomotives go to four-stroke to meet the recently finalized
standards for locomotive engines.  Like locomotive engines,  almost all Category 2 marine diesel
engines use turbochargers to increase their power-to-weight ratios. About 90 percent of Category
2 marine diesel engines are turbocharged and jacket-water aftercooled. A small fraction of these
aftercoolers are on a separate circuit from the engine cooling system. Electronic controls have not
really made any inroads into this class of engines.

    Baseline emission data is drawn from five Category 2 marine diesel engine pairs that were
emission tested for the U.S. Coast Guard and two engines tested by Caterpillar.59'60 Also, we have
NOx data on nine locomotive engines tested over the E3 duty cycle. According to the source61 of
this locomotive data, this data should approximate marine engine data.  The E3 weighted emission
results from these engines are presented in Table 3-3.  Of the five marine engine pairs,  two have
emissions levels below the Tier 0 locomotive engine standards and a third has emission levels below
the Tier 1 locomotive engine emission standard.  These marine engines probably have somewhat
lower  baseline emissions  than  locomotive engines because locomotive engines are severely
constrained in their heat rej ection capabilities while marine engines have the advantage of being able
use to the oceans, rivers, and lakes as a large heat sink. Additional testing of Category 2 engines by
Lloyd's Register confirms that Category 2 marine engines have lower emissions than baseline
locomotive engines.62'63 The Lloyd's data is included in Figure 3-3. The Lloyd's data is particularly
interesting because it measures emissions from commercial vessels in-use.
                                           56

-------
                                                  Chapter 3: Technological Feasibility
                                        Table 3-3
       Emissions Data from Baseline Category 2 Marine Diesel and Locomotive Engines
Rated
Power
(kW)
902
1120
1183
1357
1460*
1584
1710*
1801
1828
1900*
2001
2190*
2289
2356
2380*
2745
Technology Mix
locomotive
turbocharged, unit injection,
jacket- water aftercooled
locomotive
locomotive
turbocharged
locomotive
turbocharged
locomotive
locomotive
turbocharged, aftercooled
turbocharged, unit injection,
jacket- water aftercooled
turbocharged
locomotive
locomotive
turbocharged, two-stroke
opposed piston
locomotive
Emissions g/kW-hr
HC
—
0.50
—
—
0.50
—
0.03
—
—
0.03
0.55
0.00
—
—
0.11
—
NOx
17.7
14.9
20.9
21.0
16.5
21.1
13.1
17.9
15.2
16.4
17.8
11.3
17.8
16.2
9.5
18.6
CO
—
2.0
—
—
1.1
—
1.4
—
—
4.1
2.6
4.9
—
—
0.9
—
PM
—
—
—
—
0.52
—
0.17
—
—
0.27
—
0.32
—
—
—
—
New Standards
HC+NOx
7.8
CO
5.0
PM
0.27
* The emissions results reported here are averages of engine pairs.
C. Category 3 Marine Diesel Engines

    Category 3 engines are significantly different from Category 1 and 2 engines in several ways
that can affect the technologies used and the baseline emissions. First these engines are intended for
a different sort of operation. Category 3 engines are typically used for propulsion on large ocean-
going commercial vessels whose service is characterized by principally by cruising operations.
                                            57

-------
Regulatory Impact Analysis
Typically, the engine is coupled directly to a fixed pitch propeller, and maneuvering and reverse
propulsion is accomplished by reversing the entire engine.  Category 3 engines are also used to a
lesser extent on coast-wise commercial vessels.  For maneuverability in coast-wise applications,
variable pitch propellers are used to allow the engine to operate in one direction at a constant speed.
Occasionally, smaller Category 3 engines may be used for auxiliary service on ocean-going vessels.

    Consequently, Category 3 engines have a characteristic operation profile that can be different
from Category 2 and Category 1 marine engines. Because of Category 3 engines' high power (3,000-
60,000 kW) and relatively continuous use, fuel is the most significant cost in a vessel's operation.
Therefore,  Category  3 engines are designed and operated at the lowest brake specific fuel
consumption (BSFC) of any combustion system (as low as 0.176 kg/kW-hr).64'65 To achieve this they
are operated at maximum brake mean effective pressures (BMEP) (-1500-2200 kPa) to maximize
thermal efficiency and minimum mean piston speed (7-9 m/s) to maximize mechanical and propeller
efficiency.66 Power density (rated engine power/engine weight) is sacrificed to operate at maximum
efficiency rather than maximum output. In comparison, commercial Category 2 and Category 1
marine engines are often designed and operated at maximum output rather than maximum efficiency
due to economies  gained in high average-trip speeds.

    This operation profile has an effect on total oxides of nitrogen  (NOx) and particulate matter
(PM) emissions. For diesel engines there is an exponential relationship between NOx emissions and
peak combustion temperatures, which occur at peak combustion pressures.  Keeping in mind that
BSFC is minimized if BMEP is maximized at a given speed, Category 3 engines are unfortunately
designed for and most economically operated at peak NOx formation conditions. However, these
conditions do provide for a high 95%-burn combustion temperature, which minimizes non-volatile
carbon PM emissions.67 In addition, the use of bunker fuel can result in higher NOx from Category
3 engines than for other diesel engines for two reasons:  1) nitrogen in the fuel, and 2) poor ignition
qualities  leading to increased ignition delay and higher peak temperatures.  Marine fuels are
discussed in more detail later in this chapter.

    In general, Category 3 marine diesel engines are two-stroke cycle engines using a crosshead
piston design.  This means that an additional linkage is used in the piston-crank assembly to allow
for long strokes. Charge air is generally compressed using a supercharger driven by the crankshaft
at low loads and using a turbocharger under cruising and high loads. To reduce pumping losses, the
down stroke of one piston may be used to help push intake air into the next cylinder.  This charge
air is heavily cooled. Typically, large Category 3 engines will use a two stage aftercooler where the
first stage is cooled by jacket water and the second stage is cooled by raw water.  Each cylinder
typically will have its own fuel inj ection pump so that each cylinder may be independently optimized
for peak performance.

    Lloyd's has collected baseline emission data on many marine diesel engines.68 However, from
this data, EPA was only able to determine E2  emissions from five Category 3 engines, since the
modes tested did not generally line up with the marine diesel duty cycles. This data is presented in
Table 3-4. To generate the E2 weighted results, the raw Lloyd's data was converted to mass rates
by converting  all NOx reported volume  concentrations to NO2 mass concentrations and then

                                           58

-------
                                                 Chapter 3: Technological Feasibility
multiplying by exhaust flow rates calculated by carbon balance. Power was based on reported data.
Because the emission standards are based on distillate fuel and the emission data was based on
residual fuel, an adjustment of the standard to reflect residual fuel use is provided. This adjustment
of 10 percent was based on comparing Lloyd's data on distillate and residual fuel and is consistent
with the allowance provided by MARPOL. EPA recognizes that the small ship population sampled
is a source of significant uncertainty; however, these results do indicate that an emissions benefit
exists for Category 3  engines.

                                       Table 3-4
               Emission Data from Baseline Category 3 Marine Diesel Engines
Rated Power
[kW]
4246
4780
4780
6545
7700
Rated Speed
[rpm]
570
520
512
510
510
Measured NOx
[g/kW-hr]
12.7
16.2
17.7
19.5
16.2
MARPOL NOx
Standard [g/kW-hr]
11.2
11.9
12.6
12.9
12.9
Standard Adjusted
for Bunker Fuel*
12.3
13.2
13.9
14.2
14.2
* This is a rough adjustment for comparison purposes only.
                                           59

-------
Regulatory Impact Analysis
VIII. Low-Emission Category 1 Marine Engines

    In an effort to achieve significant and timely emission reductions from marine engines in
California, many vessels have been repowered with low-emission marine engines as part of various
demonstration programs. These repower (completed and planned) programs are the result of efforts
by local agencies such as the Santa Barbara County Air Pollution Control District and the South
Coast Air Quality Management District as well as EPA's Region 9, the Port of Los Angeles, and
others.  The low-emission marine engines have been specially produced for the marine repower
programs by a number of manufacturers including Cummins, Detroit Diesel, Mercruiser, Marine
Corporation of America, and Yanmar.  At this time, more than 75 low-emission high-speed marine
engines have been  installed and are still  operating in  a wide variety  of applications, and more
replacements are planned.

    The emission control strategies used on these demonstration engines were targeted on achieving
low NOx. For the majority of these engines, timing was retarded at cruise and full load operating
conditions using either electronic or mechanical engine control. Another technology which achieved
low NOx emissions was indirect injection (TDI).  All of the engines used raw-water aftercooling.
Vessel owners reported positively on the performance and fuel efficiency of these engines.  Apart
from  a problem with the calibration of electronic controls  on one engine  type, which has
subsequently been resolved, no significant problems have arisen as a result of the low-emission
technology.0 Table 3-5 presents emissions data from several low-emission marine engines based on
the E3 marine test cycle.69'70'71'72'73

    Figure 3-8 compares the emission results from the baseline and low-emission high-speed marine
engines presented in this chapter to the new  emission standards. The emission levels  of these
demonstration engines compare favorably to the new standards.
       c The only problem that was reported was due to the calibration of the electronic controls
on one engine type. This problem has been resolved.

                                          60

-------
                                                   Chapter 3: Technological Feasibility
                                         Table 3-5
            Emissions Data from Low-Emission Category 1 Marine Diesel Engines
Rated
Power
(kW)
120
132
162
164
217
232
245*
252
298*
324
575*
Technology Mix
turbocharged, raw- water
aftercooled, electronic control
turbocharged, raw- water
aftercooled, IDI
turbocharged, raw- water
aftercooled, IDI
turbocharged, raw- water
aftercooled, electronic control
turbocharged, raw- water
aftercooled, electronic control
turbocharged, raw- water
aftercooled
turbo, raw- water aftercooled,
electronic Ctrl, 2-stroke
turbocharged, raw- water
aftercooled
turbocharged, raw- water
aftercooled, electronic control
turbocharged, jacket-water
aftercooled
turbo, raw- water aftercooled,
electronic Ctrl, 2-stroke
Emissions g/kW-hr
HC
0.09
0.05
0.07
0.31
0.14
0.45
0.4
—
0.13
0.11
0.3
NOx
5.0
4.0
3.9
4.7
6.1
6.5
6.5
5.5
8.7
6.5
6.1
CO
0.96
0.13
0.17
1.6
2.1
0.99
1.0
—
0.78
2.0
3.5
PM
—
**
**
—
—
—
0.23
—
0.09
0.78
0.40
New Standards
HC+NOx
7.2
7.2
7.2
CO
5.0
5.0
2.0
PM
0.3
0.2
0.27
* This data is based on Cl test cycle.
** No PM data was collected, but the manufacturer reported that there was no visible smoke.
                                             61

-------
Regulatory Impact Analysis
20
I15
S
X
O 10
z
+
o
I
^ 5
Z
1.4
1.2
.c
^ 0.8
S0.6
11 0.4
0.2
0
Q -_
7
. . R
'!_' D
-* 4
D)
•3
O J
0 o -
1

Figure 3-8: Diesel Marine Engine Emissions Compared to the Standards

A
*
-A.
A
0


A


T— I


* * A
* * AV
. '"W
* A
A
A

•4
;
A" ~


-

A

^




300 600
Rated Power [kW]




*
^
y






;
















^




900








0 300 600 900
Rated Power [kW]





*

A





_^_
* A *
^A M* *






*A ^
A




A








*








*



0 300 600 900
Rated Power [kW]

+ Proposed Standards
* Test Data

                                   62

-------
                                                 Chapter 3: Technological Feasibility
IX. Anticipated Technology Mix

    As discussed above, marine engines are generally derived from land-based nonroad, locomotive,
and to some extent highway engines. This allows marine engines, which generally have lower sales
volumes than other nonroad engines, to be produced more cost-effectively.  Because the marine
designs are derived from land-based  engines, EPA believes that many of the emission-control
technologies which are likely to be applied to nonroad engines to meet their Tier 2 and 3 emission
standards and to locomotives to meet their Tier 2 emissions standards will be applicable to marine
engines. EPA believes that the technologies listed below will be sufficient for meeting both the new
emission standards and the NTE requirements.

A.  Category 1 Marine Diesel  Engines

    EPA anticipates that timing retard may be used in many Category 1 marine diesel applications,
especially in certain modes of engine operation to ensure adequate control of NOx emissions across
the whole NTE zone.  The negative impacts of timing retard on HC, PM and fuel consumption can
be offset with advanced fuel injection systems with higher fuel injection pressures, optimized nozzle
geometry, and potentially through rate shaping. EPA does not expect marine engine manufacturers
to convert from direct injection to indirect injection due to these standards.

    Regardless of environmental regulations,  EPA believes that Category 1 engine manufacturers
will be making greater use of electronic engine management controls in the future to satisfy customer
demands of increased power and fuel economy.  Through the use of electronic controls, additional
reductions in HC, CO, NOx, and PM can be achieved. Electronics may be used to optimize engine
calibrations under a wider range of operation. Most of the significant research and development for
the improved fuel inj ection and engine management systems should be accomplished for land-based
nonroad diesel engines which are being designed to meet Tier 2 and Tier 3 standards. Electronically
controlled common rail engines should prove to be capable of meeting even lower emission levels
in the future, especially for smaller engines.

    EPA projects that all Category 1 marine diesel engines will be turbocharged and most will be
aftercooled to meet Tier 2 emission standards.  Aftercooling strategies will likely be a mix of j acket-
water and separate-circuit aftercooling. EPA does not expect a significant increase in the use of raw-
water charge air cooling for marine engines as a result of this rule.  Chapter 4 presents a possible
scenario of how these technologies could be used on Category 1 marine diesel engines to meet the
Tier 2 standards.

    By setting standards with implementation dates that extend well into the next decade, EPA is
providing engine manufacturers with substantial lead time for developing, testing, and implementing
emission control technologies.  This lead time and the coordination of standards with those for
nonroad engines allows for a comprehensive program to integrate the most effective emission control
approaches into the manufacturers' overall design goals related to performance, durability, reliability,
and fuel consumption.
                                           63

-------
Regulatory Impact Analysis
B. Category 2 Marine Diesel Engines

    EPA anticipates that Category 2 marine diesel engines will use the same control strategies as
Category 1 engines. As a result of the new emission standards, EPA projects that the majority of
these engines will continue to be turbocharged and aftercooled. EPA believes that separate-circuit
aftercooling will most likely be considered more attractive than jacket-water aftercooling for most
applications.  These control strategies are similar to those that are likely to be used on locomotive
engines, except that the cooling strategies on marine engines are expected to be more effective.

C. Category 3 Marine Diesel Engines

    EPA anticipates that Category 3 engines will be able to achieve the MARPOL NOx limit
without significant engine redesign. Typically, the same strategies that have been used over time to
reduce fuel consumption have generally resulted in an increase in NOx emissions. Reducing NOx
with the technology used today basically means calibrating the engines with a focus on emissions
as well as fuel consumption.  For instance, rate shaping may be used to inject a small amount of fuel
early followed by the remaining charge so that  the majority of the  fuel may undergo diffusion
burning. This reduces NOx by reducing peak cylinder temperatures associated with the burning of
fuel that is premixed with air prior to the start of combustion. Negative impacts on fuel consumption
can also be minimized through fuel injection strategies, including increasing injection pressures and
through optimizing nozzle  geometry.  Wartsila NSD, a market leader in Category 3  engine
production, has  demonstrated a combination of engine technologies that meets the MARPOL
emission standard.  Wartsila NSD has utilized a combination of late fuel injection rate shaping,
higher cylinder compression ratio, higher fuel injection pressure and an optimized combustion
chamber to achieve a 25-3 5 percent reduction in emissions.74 Furthermore, the combination of these
technologies has no negative effect on BSFC.
X.  Test Fuel Specifications

A. Category 1 and 2 Marine Diesel Engines

    EPA is applying the recently finalized test fuel specifications for nonroad diesel engines to
Category 1 and Category 2 marine diesel engines, with a change in the sulfur content upper limit
from 0.4 to 0.8 weight-percent (wt%).  EPA believes that this will simplify development  and
certification burdens for marine engines that are developed from land-based counterparts. This test
fuel has a sulfur specification range of 0.03 to 0.80 wt%, which covers the range of sulfur levels
observed for most in-use fuels.  Manufacturers will be able to test using any fuel within this range
for the purposes of certification. Thus, they will be able to harmonize their marine test fuel with
U.S. highway (<0.05 wt%) and nonroad (0.03 to 0.40 wt%), and European testing (0.1 to 0.2 wt%).

    The intent of these test fuel specifications is to ensure that engine manufacturers design their
engines for the full range of typical fuels used by Category 1  marine  engines in use. Because the
technological feasibility of the new emission standards is based on fuel with up to 0.4 wt% sulfur,

                                           64

-------
                                                  Chapter 3: Technological Feasibility
any testing done using fuel with a sulfur content above 0.4 wt% would be done with an allowance
to adjust the measured PM emissions to the level they would be if the fuel used were 0.4 wt% sulfur.
The full range of test fuel specifications are presented in Table 3-6.  Because testing conducted by
EPA is limited to the test fuel specifications, it is important that the test fuel be representative of in-
use fuels.

                                        Table 3-6
                     Category 1 and Category 2 Test Fuel Specifications
Item
Cetane
Initial Boiling Point, ฐC
10% point, ฐC
50% point, ฐC
90% point, ฐC
End Point, ฐC
Gravity, API
Total Sulfur, % mass
Aromatics, % volume
Parafms, Napthenes, Olefins
Flashpoint, ฐC
Viscosity @ 38 ฐC, centi stokes
Procedure (ASTM)
D613-86
D86-90
D86-90
D86-90
D86-90
D86-90
D287-92
D 129-21 orD2622-92
D1319-89orD5186-91
D1319-89
D93-90
D445-88
Value (Type 2-D)
40-48
171-204
204-238
243-282
293-332
321-366
32-37
0.03-0.80
10 minimum
remainder
54 minimum
2.0-3.2
B. Category 3 Marine Diesel Engines

    Category 3 engines typically burn residual fuel, which is the by-product of distilling crude oil
to produce lighter petroleum products such as gasoline, diesel fuel and kerosene. Residual fuel
possesses a high viscosity and density, and it typically has high ash, sulfur and nitrogen content in
comparison to marine distillate fuels.  Some Category 3 engines can burn straight residual fuel, but
many burn a blend of residual and distillate, which is called intermediate fuel (IF). The two most
common IF blends burned in Category 3  engines are IF  180, which contains about 88 percent
residual fuel and IF 380 which contains about 98 percent residual fuel.75 Table 3-7 summarizes
current ASTM standards for a marine distillate oil, residual fuel, and the two most common IF
blends.
                                           65

-------
Regulatory Impact Analysis
                                        Table 3-7
                         Comparison of ASTM Fuel Specifications76

ISO-F symbol
Density @ 15C, max
Viscosity @ 40C
Viscosity @ 50C
Viscosityฎ 100C
Carbon Residue, max
Ash, max
Sulfur, max
Units

kg/m3
cSt
cSt
cSt
wt%
wt%
wt%
Distillate fuel
DMA
890
1.5-6.0
—
—
0.20*
0.01
1.5
IF 180
RMF-25
991
316
180
25
20**
0.15
5.0
IF 380
RMH-35
991
-710
380
35
22**
0.20
5.0
Residual fuel
RML-55
no max
—
—
55
no max
0.20
5.0
    * Ramsbottom test
    ** Conradson test

    The use of residual fuel has two important consequences.  First, it is more difficult to handle.
Because of it's high viscosity and high impurities, the fuel must be heated and filtered before it can
be passed to the engine.   This requires additional equipment and space. Bunker fuel is kept in a
main fuel tank where it is kept heated, generally using steam coils, to just above its pour point. Prior
to use, this fuel is pumped into a settling tank, where the heavier portions settle to the bottom. Fuel
is pumped from the top of the settling tank through heaters, centrifugal separators, and filters before
entering the fuel metering pump. The centrifugal separators and filters remove water and remaining
sludge from the fuel.  The sludge is then routed to a sludge tank. In addition, a separate fuel tank is
usually necessary to store a lighter fuel which is used to start a cold engine.

    Second, residual  fuels can have detrimental effects on engine emissions.  These fuels may
contain between 0.6-2.15 percent nitrogen by weight, and fuel-bound nitrogen is almost completely
converted to NO in diesel engines.77 It is estimated that fuel-bound nitrogen contributes 0.35 g/kW-
hr per 0.1 percent nitrogen, however test results indicate that fuel  ignition  quality  also has a
detrimental effect on NOx emissions.78 For example ISO E3 test results on a Category 3 engine
indicate a 22 percent increase in ISO weighted NOx when residual fuel was substituted for distillate
fuel, but the fuel-bound nitrogen (0.4 percent) only accounted for 9 percent of this increase.  The
remainder of the NOx increase was  attributed to the fuel's poor ignition quality, which caused
excessive ignition delay at part load.  At 25 percent load on residual fuel, the engine produced 50
percent more NOx than on distillate, while at full load it produced only 25 percent more than on
distillate.79

    Residual fuel also has a detrimental effect on paniculate matter (PM) emissions. One Category
3-engine study80 indicated that on residual fuel, PM increased on average by a factor of 10 in
                                           66

-------
                                                 Chapter 3: Technological Feasibility
comparison to distillate. Furthermore, lab-to-lab comparisons showed ฑ25 percent variability in PM
results when residual fuel was tested. This unacceptable test variability has been attributed to the
effect of high sulfate concentrations  on the dilute sampling method specified by the International
Standards Organization (ISO). These results have led to a remark in ISO 8178 stating that the
method has not been validated with fuel containing more than 0.8 percent sulfur by-weight. Since
even the lowest international marine distillate fuel sulfur specification is 1.5 percent, Category 3
engines  operate on fuels for which no PM test method has been developed.

XL Impact on Noise, Energy, and Safety

    The Clean Air Act requires EPA to consider potential impacts on noise, energy, and safety when
establishing the feasibility of emission standards for marine diesel engines.  One important source
of noise in diesel combustion is the  sound associated with the combustion event itself.  When a
premixed  charge of fuel and air ignites,  the very rapid combustion leads to a sharp increase in
pressure, which is easily heard and recognized as the characteristic sound of a diesel engine. The
conditions that lead  to high noise levels also cause high levels of NOx formation.  Fuel injection
changes and  other NOx  control  strategies therefore typically reduce engine noise, sometimes
dramatically.

    The impact of the new emission standards on energy is measured by the effect on fuel
consumption from complying engines. Many of the marine engine manufacturers are expected to
retard engine timing which increases fuel consumption somewhat. Most of the technology changes
anticipated in response to the new standards, however, have the potential to reduce fuel consumption
as well as emissions.  Redesigning combustion chambers, incorporating improved fuel injection
systems, and introducing electronic controls provide the engine designer with powerful tools for
improving fuel efficiency while simultaneously controlling emission formation.  To the extent that
manufacturers add aftercooling to non aftercooled engines and shift from jacket-water aftercooling
to raw-water aftercooling, there will  be a marked improvement in fuel-efficiency.  Manufacturers
of highway diesel engines have been able to steadily improve fuel efficiency even as new emission
standards required significantly reduced emissions.

    There are no apparent safety issues associated with the new emission standards. Marine engine
manufacturers will likely use only proven technology that is currently used in other engines such as
nonroad land-based  diesel applications, locomotives, and diesel trucks.
                                           67

-------
Regulatory Impact Analysis
Chapter 3 References
1.  Herzog, P., Burgler, L., Winklhofer, E., Zelenda, P., and Carterllieri, W., "NOx Reduction
Strategies for DI Diesel Engines," SAE Paper 920470, 1992.

2.  Uyehara, O., "Factors that Affect NOx and Particulates in Diesel Engine Exhaust," SAE Paper
920695, 1992.

3.  Durnholz, M., Eifler, G., Endres, H., "Exhaust-Gas Recirculation - A Measure to Reduce
Exhaust Emission of DI Diesel Engines,"  SAE Paper 920715, 1992.

4.  Bazari, Z., French, B.,  "Performance and Emissions Trade-Offs for a HSDI Diesel Engine -
An Optimization Study," SAE Paper 930592, 1993.

5. Acurex Environmental  Corporation, "Estimated Economic Impact of New Emission Standards
for Heavy-Duty Highway Engines," prepared for U.S. EPA, March 26, 1996 (Docket A-97-50;
document IV-A-3).

6.  See reference 4—SAE 930592

7.  U.S. Environmental Protection Agency, "Emission Control Technology for Diesel Trucks:
Report to Congress," October 1993 (Docket A-97-50; document IV-A-2).

8.  See reference 5—Acurex Environmental Corporation

9.  See reference 4—SAE 930592

10. See reference 1—SAE 920470

11. See reference 4—SAE 930592

12. Pierpont, D., Reitz, R., "Effects of Injection Pressure and Nozzle Geometry on DI Diesel
Emissions and Performance," SAE Paper 950604, 1995.

13. See reference 1—SAE 920470

14. See reference 12—SAE 950604

15. Ghaffarpour, M and Baranescu, R, "NOx Reduction Using Injection Rate Shaping and
Intercooling in Diesel Engines," SAE Paper 960845, 1996

16. Piepont, D., Montgomery, D., Reitz, R., "Reducing Particulate and NOx Using Pilot
Injection," SAE Paper 950217, 1995.
                                         68

-------
                                               Chapter 3: Technological Feasibility
17.  "CAT's HEUI System: A Look at the Future?," Diesel Progress, April 1995, page 30
(Docket A-97-50; document IV-D-60).

18.  "CAT Gears Up Next Generation Fuel Systems," Diesel Progress, August 1998, page 82
(Docket A-97-50; document II-G-12).
19.  Youngblood, J., "Cummins New Midrange Fuel System," presentation at SAE Diesel
Technology for the New Millennium TOPTEC, April 21, 1998 (Docket A-97-50; document
IV-D-65).

20.  "The Year in Review," Diesel Progress, June 1998, page 34 (Docket A-97-50; document
II-G-13).

21.  See reference 1—SAE 920470

22.  See reference 16—SAE 950217

23.  Montgomery, D. and Reitz, R., "Six-Mode Cycle Evaluation of the Effect of EGR and
Multiple Injections on Particulate and NOx Emissions from a D.I. Diesel Engine," SAE Paper
960316, 1996

24.  See endnote 1—SAE 920470

25.  Uchida, N., Daisho, Y., Saito, T., Sugano, H., "Combined Effects of EGR and Supercharging
on Diesel Combustion and Emissions," SAE Paper 930601, 1993.

26.  Yoshikawa, H., Umehara, T., Kurkawa, M., Sakagami, Y., Ikeda, T., "The EGR System for
Diesel Engine Using a Low Voltage Soot Removal Device," SAE Paper 930369, 1993.

27.  Khalil, N., Levendis, Y., Abrams, R., "Reducing Diesel Particulate and NOx Emissions via
Filtration and Particle-Free Exhaust Gas Recirculation," SAE Paper 950736, 1995.

28.  Baert, R., Beckman, D., Verbeek, R., "New EGR Technology Retains HD Diesel Economy
with 21st Century Emissions," SAE Paper 960848, 1996.

29.  Fleet Owner, "Hardware Report: What's new in... Bypass Filtration," magazine article,
January 1997 (Docket A-97-50; document IV-D-59).

30.  Meeting between Manufacturers of Emission Controls Association and U.S. Environmental
Protection Agency, April 1995 (Docket A-97-50; document IV-D-66).

31.  Voss, K., Bulent, Y., Hirt, C., and Farrauto, R.,  "Performance Characteristics of a Novel
Diesel Oxidation Catalyst," SAE Paper 940239, 1994.
                                         69

-------
Regulatory Impact Analysis
32.  Johnson, J., Bagley, S., Gratz, L., Leddy, D., "A Review of Diesel Particulate Control
Technology and Emissions Effects - 1992 Horning Memorial Award Lecture," SAE Paper
940233, 1994.

33.  California Air Resources Board, Mail-Out #91-42.

34.  Konno, M., Chikahisa, T., Murayama, T., "Reduction of Smoke and NOx by Strong
Turbulence Generated During the Combustion Process in D.I. Diesel Engines," SAE Paper
920467, 1992.

35.  International  Standards Organization, 8178-4, "Reciprocating internal combustion
engines—Exhaust emission measurement—Part 4: Test cycles for different engine applications."

36.  Memorandum from Mike Samulski, U.S. EPA, to Docket A-96-40, "Selection of Duty Cycle
to Propose for High-Speed CI Marine Engines," February  19,  1997 (Docket A-97-50; document
II-B-5).

37.  Southwest Research Institute, "Emission Testing of Nonroad Compression Ignition
Engines," prepared for U.S. Environmental Protection Agency, September 1995 (Docket A-97-
50; document IV-A-1).

38.  Letter from Michael S. Brand, Cummins Engine Company, Inc. to Bill Charmley U.S.
Environmental Protection Agency, November 13, 1995 (Docket A-97-50; document IV-D-63).

39.  Southwest Research Institute, "Marine Diesel Engine  Testing," prepared for U.S.
Environmental Protection Agency, September 1999 (Docket A-97-50; document II-A-6).

40. SwRI Interim Final Report,  "Nonroad Engine Emissions Testing," prepared for U.S.
Environmental Protection Agency, September 1999 (Docket A-97-50; document IV-A-7).

41.  Acurex Environmental, "Locomotive Technologies to Meet SOP Emission Standards,"
Prepared for the U.S. Environmental Protection Agency, August 13,  1997 (Docket A-97-50;
document IV-A-6).

42.  Environmental Transportation Consultants, "Shipboard Marine Engines Emission Testing
for the United States Coast Guard," Delivery Order No. 31, 1995  (Docket A-97-50; document
IV-D-5).

43.  Lloyd's Register, "Marine Exhaust Emission Research Programme; Steady-State Operation,"
1990 (Docket A-97-50; document IV-D-68).

44.  Lloyd's Register, "Marine Exhaust Emissions Research Programme; Steady-State Operation;
Slow Speed Addendum," 1991  (Docket A-97-50; document IV-D-68).

45. Letter from Steven Meinhart, Caterpillar Inc. to Alan Stout, U.S. Environmental Protection
Agency, April 6, 1999 (Docket  A-97-50; document IV-D-64).

                                         70

-------
                                                Chapter 3: Technological Feasibility
46. "Heavy-duty Diesel Engines Controlled by Onboard Computers: Guidance on Reporting and
Evaluating Auxiliary Emission Control Devices and the Defeat Device Prohibition of the Clean
Air Act," U.S. EPA,  October 15, 1998 (Docket A-97-50; document U-B-3)
47. Wilbur, C., "Marine Diesel Engines," Butterworth & Heinemann Ltd, 1984.

48. See reference 39-Southwest Research Institute, 1999

49. Memorandum from Mark Wolcott to Charles Gray, "Ambient Temperatures Associated with
High Ozone Concentrations," U.S. Environmental Protection Agency, September 6, 1984
(Docket A-97-50; document II-B-2).

50. See reference 39-Southwest Research Institute, 1999

51. SAE J1937 (reaffirmed JAN1995), "Engine Testing with Low-Temperatue Charge Air-
Cooler Systems in a Dynamometer Test Cell," SAE Recommended Practice.

52. Annex VI of MARPOL 73/78, "Technical Code of Control of Emissions of Nitrogen Oxides
for Marine Diesel Engines," October 22, 1997 (Docket A-97-50; document IV-D-67).

53. See reference 37—Southwest Research Institute, 1995

54. See reference 38-Brand, 1995

55. See reference 40—Environmental  Transportation Consultants, 1995

56. Memorandum from Jeff Carmody, Santa Barbara County Air Pollution Control District, to
Mike Samulski, U.S. Environmental Protection Agency, "Marine Engine Replacement
Programs," December 1, 1997 (Docket A-97-50; document II-G-11).

57. See reference 45-Meinhart, 1999

58. See reference 39-Southwest Research Institute, 1999

59. See reference 42—Environmental  Transportation Consultants, 1995

60. See reference 45-Meinhart, 1999

61. Letter from William Passie, Caterpillar, to Alan Stout, U.S. EPA, August 4, 1999 (Docket
A-97-50; document IV-D-55).

62. See reference 43—Lloyd's Register, 1990

63. See reference 42—Lloyd's Register, 1991

64. Gilmer, Johnson, Introduction to Naval Architecture. U.S. Naval Institute, 1992, p. 251.

                                         71

-------
Regulatory Impact Analysis
65.  Taylor, C., The Internal Combustion Engine in Theory and Practice. MIT Press, 1990, p.
448.

66.  Heywood, J., Internal Combustion Engine Fundamentals. McGraw-Hill. New York. 1988.
p. 887.

67.  Primus, R., Hoag, K., "Fundamentals of Reciprocating Engine Performance-Class Notes,"
Prepared August 1995.

68.  See reference 44-Lloyd's Register, 1991

69.  Memorandum from Jeff Carmody, Santa Barbara County Air Pollution Control District, to
Mike Samulski, U.S. Environmental Protection Agency, "Marine Engine Replacement
Programs," July 21, 1997 (Docket A-97-50; document II-G-10).

70.  Phone conversation between Steven Moore, South Coast Air Quality Management District
and Mike Samulski, U.S. Environmental Protection Agency, July 30, 1997.

71.  See reference 57-Carmody, 1997

72.  Talwar, M., Crawford, E., Hart, D., "Low Emission Commercial Marine Engine
Development for the Santa Barbara County Air Pollution Control District," SAE Paper 950734,
1995.

73.  Facsimile from Eric Peterson, Santa Barbara County Air Pollution Control District, to Mike
Samulski, U.S. Environmental Protection Agency, "Marine Engine Replacement Programs,"
April 1, 1998 (Docket A-97-50; document II-D-14).

74.  Hellen, G., "Technologies for Diesel Exhaust Emission Reduction," NAVSEA Workshop,
August 18, 1998 (Docket A-97-50; document II-D-15).

75.  The Bunker News Daily, "Flashpoint," MRC Publications, July 17, 1997, Vol. 1 No. 137,
www.bunkernews.com (Docket A-97-50; document II-D-16).

76.  International Standards Organization 8217,  1987.

77.  See reference 66-Heywood, p. 577, 1988

78.  Bastenhof, D., "Exhaust Gas Emission Measurements: A Contribution to a Realistic
Approach," CIMAC, May 1995 (Docket A-97-50; document II-G-16).

79.  See reference 78-Bastenhof, 1995

80.  See reference 78-Bastenhof, 1995
                                         72

-------
      Chapter 3: Technological Feasibility
73

-------
Regulatory Impact Analysis
                                   74

-------
                                                         Chapter 4: Economic Impact
                CHAPTER 4:  ECONOMIC IMPACT
    EPA expects that in almost all cases, manufacturers will produce a complying marine engine
by adapting an engine that has been designed and certified to meet highway or nonroad emission
standards. This analysis considers the cost of these upgrades to the base engines as part of the impact
of new marine emission standards; variable costs are applied directly, with an additional fixed cost
added to apply the technologies to marine engines. The analysis arrives at the full cost impact by
considering changes to turbocharging and aftercooling applicable to marine engines.

    This chapter describes EPA's approach to estimating the cost of complying with the new
standards. Both engine and vessel design are considered in determining a total cost impact. The
estimated aggregate cost to society is also considered, followed by an analysis of the impact on small
businesses.

I. Methodology

    EPA has estimated a mix of current and projected technologies for complying with the new
emission standards.  The costs of individual technologies are developed in considerable detail and
then combined according to the projections of technology changes. EPA developed the costs for
individual technologies in cooperation with ICF, Incorporated  and Geraghty & Miller in  a
combination of reports related to diesel engine emission controls.1'2

    To simplify the analyses, costs were examined for five types of engines representing five power
ranges.  The five cases, shown in Table 4-1, cover four Category 1 engines and one Category 2
engine.   The selected power ratings generally correspond with the displacement values used to
differentiate among the standards.

    Costs of control include variable costs (for incremental hardware costs, assembly costs, and
associated markups) and fixed costs (for tooling, R&D, and certification).  Variable costs are marked
up at a rate of 29 percent to account for manufacturers' overhead and profit.3 For technologies sold
by a supplier to the engine manufacturers, an additional 29 percent markup is included  for the
supplier's overhead and profit.  The analysis also includes consideration of lifetime operating costs
where applicable. The result is a total estimated incremental cost for individual engines of various
sizes.  Costs are presented in 1997 dollars.

    Table 4-1 also includes information on current product offerings and sales volumes, which is
needed to calculate  amortized fixed costs for individual engines.  Estimated sales and product
offerings were compiled from the PSR database based on historical 1997 information. For the largest
engines, some sales volumes were modified slightly to take into account recent data related to vessel
populations.
                                          75

-------
Regulatory Impact Analysis
                                       Table 4-1
                Power Ranges and Nominal Power for Estimating Costs (kW)
Engine Power
Ranges
37 - 225
225 - 560
560 - 1000
1000-2000
2000 +
Nominal
Engine Power
100
400
750
1500
3000
Models
26
15
10
8
5
Annual
Sales*
3,284
1,579
142
130
68
Average Sales
per Model
126
105
14
16
14
    *Excluding recreational.
    Even though the analysis does not reflect all the possible technology variations and options that
are available to manufacturers, EPA believes that the projections presented here provide a cost
estimate representative of the different approaches manufacturers may ultimately take.

II. Overview of Technologies

    The land-based engines that often serve as the base engines for marine diesel applications will
be changing as a result of new emission standards adopted for nonroad and locomotive engines.
Most new nonroad and locomotive engines rated over 37 kW will be subject to two new tiers of
standards spanning the next ten years. These engines will be designed, manufactured, and certified
to have reduced emissions.  In most cases, the technological challenge for developing compliant
marine engines is therefore to make the necessary modifications to the land-based engines for their
use in the marine environment without significantly increasing emission levels.

    In the absence of emission standards for marine engines, manufacturers would have to decide
whether it would be more effective to marinize their low-emitting Tier 2 land-based engines, thus
maintaining a single base engine, or to continue to marinize unregulated or Tier 1  land-based
engines, thus maintaining separate production of older-technology engines for marine application.
This analysis is based on engine manufacturers choosing the latter, and accordingly assesses the
previously estimated variable costs of the land-based programs as an impact of the marine emission
standards. To the extent that manufacturers would simplify production by using their low-emitting
models anyway, the cost estimates developed in this analysis overstates the actual  impact of the
emission standards.

    As discussed in Chapter 3, manufacturers of Category 1 engines are expected to comply with
the new emission standards by conducting basic engine modifications, upgrading fuel systems, and
improving aftercooling systems.  Manufacturers of Category 2 engines are expected  to redesign
combustion chambers, improve high-pressure  fuel injection  systems,  and upgrade  or add
                                           76

-------
                                                          Chapter 4: Economic Impact
turbocharging and aftercooling.  While the final emission standards vary for different sizes of
Category 2 engines, this analysis presents only a single set of cost estimates for all these engines.
The Category 2 emission standards are based on the technological capabilities of engines, taking into
account the generally observed increase in emission levels with greater engine size. The graduated
standards therefore call for a common  set of technologies for all the different sizes of Category 2
engines. The Category 2 cost estimates therefore represent the whole range of Category 2 engines.
As described above, this is not meant to imply that manufacturers will, in practice, adopt a uniform
set of technologies to comply with emission standards.

    Except for the aftercooling changes, hardware improvements for nonroad  and  locomotive
engines should be transferrable to marine engines, in many cases with some degree of adaptation.
The  analysis  includes  a  substantial  degree  of development  work  to  make adjustments for
turbocharger matching, adjusting fuel injection parameters, and other changes that may be needed
to prepare an engine for marine applications. Also, because manufacturers will in many cases be
producing new engine designs outside the normal product development cycle, extensive development
costs are included to design a marine version of a base engine, taking into account not only direct
expenses for controlling emissions, but also considering some need for re-optimizing performance.
Finally, marine engines rely on seawater (through a heat exchanger), not the ambient air, for rej ecting
heat from the engine and  aftercooler.  The cost of adding these systems are therefore presented
separately in this chapter for different sizes of marine engines.

    The first step in estimating the incremental cost of new emission standards is to establish the
baseline technology package from which changes will be made. As described above, most of the
technologies included in the analysis of costs for the land-based rules are carried directly into the
analysis of costs for marine engines. Including these sets of technologies as a package defines both
baseline and projection scenarios. A more detailed assessment is required for turbocharging and
aftercooling. The PSR database provides information on sales and engine data for engine models
available for auxiliary and propulsion marine applications. Table 4-2 summarizes the sales-weighted
figures showing how many current engine sales include the various turbocharging and aftercooling
options. Table 3-2 provides similar information without breaking engines into different size ranges.
The predominance of naturally aspirated engines rated under 130 kW is contrasted with the nearly
universal use of turbocharging for bigger engines.  The table also  shows that most turbocharged
engines are equipped with jacket-water (or "single-circuit") aftercoolers, though a small number of
turbocharged engines in certain size ranges have either no aftercooling or the more sophisticated
separate-circuit aftercooling.
                                           77

-------
Regulatory Impact Analysis
                                        Table 4-2
                        Baseline Technologies for Marine Engines*
Engine Power
(kW)
100
400
750
1500
3000
Naturally
Aspirated
65%
0%
0%
0%
10%
Turbocharged
35%
100%
100%
100%
90%
Jacket-water
Aftercooling
15%
55%
100%
100%
80%
Separate-circuit
Aftercooling
0%
15%
0%
0%
10%
source: PSR Database.
*A11 numbers rounded to the nearest 5 percent.
    As described in the cost analysis for the land-based nonroad engine rulemaking, manufacturers
are expected to develop engine technologies not only to reduce emissions, but also to improve engine
performance.  While it is difficult to take into account the effect of ongoing technology development,
EPA is concerned that assessing the full cost of the anticipated technologies as an impact of new
emission standards would inappropriately exclude from consideration the expected benefits for
engine performance, fuel consumption, and durability.d  Short of having sufficient data to predict the
future with a reasonable degree of confidence, EPA faces the need to devise an alternate approach
to quantifying the true impact of the new emission standards. As an attempt to take this into account,
we present the full cost of the control technologies in this chapter, then apply a credit to some of
these costs for calculating the cost-effectiveness of the rulemaking, as described in Chapter 6.

III. Technology Costs

    The total estimated cost impact of new emission standards is developed by considering the cost
of each of the anticipated technologies. The following paragraphs describe these technologies and
their application to marine engines.  The analysis then combines these itemized costs into a
composite cost estimate for the range of marine engines affected by the rulemaking.4

    Cost estimates for fuel injection upgrades are derived from estimates for nonroad engines.
Variable costs are carried directly into this analysis. In some cases, cost figures represent an average
across  a broader power range or an extrapolation to a bigger engine.  Estimated R&D expenditures
are also considered.   The anticipated R&D effort should be focused primarily on transferring
established engine technologies to marine engines.  Projected fixed costs are therefore reduced from
       dWhile EPA does not anticipate widespread, marked improvements in fuel consumption,
small improvements on some engines may occur.
                                           78

-------
                                                         Chapter 4: Economic Impact
the levels anticipated for redesigning nonroad engines.  EPA's general expectation is that one-third
of the previously anticipated level of R&D will be needed to successfully implement the changes for
marine engines.

    Estimated levels of R&D in the analysis are developed for individual technologies to show the
incremental effect of combining different levels of emission control.  Engine design in practice,
however, is much more integrated. Table 4-3 shows a weighted total of the combined R&D costs
calculated for each  size engine to integrate the projected package of technologies for the new
emission standards.  To put these large capital expenditures in context, the analysis is based on each
$1 million of R&D consisting of about three engineer-years and four technician-years of effort, plus
nearly $500,000 for  testing-related expenses.

                                        Table 4-3
                                  Total Estimated R&D
                             Expenditures per Engine Family
Engine Size
100 kW
400 kW
750 kW
ISOOkW
3000 kW
R&D
$490,000
$498,000
$1,400,000
$1,400,000
$1,870,000
    As described in the preamble to the final rule, the manufacturers are responsible to comply with
emission limits at any speed and load that can occur on a vessel. EPA believes it is not appropriate
to consider additional costs for manufacturers to comply with these "off-cycle" requirements. This
is because we expect that  manufacturers can  manage engine operation to  avoid unacceptable
variation in emission levels by more effectively using the technologies that will be used to meet the
emission limits more broadly, rather than by use of additional hardware. For example, manufacturers
can adust fuel inj ection parameters to avoid excessive emissions. The split-zone approach described
in Chapter 3 is designed to accommodate normal variation in emission levels at different operating
points. This approach involves no additional variable cost. The estimated R&D expenditures reflect
the time needed to address this.  It is possible that manufacturers will need to retard inj ection timing
for some engines, which would result in a cost penalty for increased fuel consumption. This scenario
is considered as a sensitivity analysis in Section in.J. below.

A. Fuel Injection Improvements

    All engines  are expected to see significant improvements in their fuel injection systems.  We
also expect some engines to see incremental improvements in existing unit injector designs  (see
                                           79

-------
Regulatory Impact Analysis
Table 4-4). Better control of injection timing and increased injection pressure contribute to reduced
emissions.

    An additional alternative is a common rail injection system.  The principal benefit of common
rail technology is that injection pressure is not dependent on engine speed, which, in conjunction
with electronic controls, greatly increases the flexibility of tailoring the injection timing and profile
to varying modes of operation.5 Though the technology development originated with highway diesel
engines, common rail designs have been tailored for application in  much larger engines.6  Cost
estimates for common rail systems are summarized in Table 4-5.

                                         Table 4-4
                                Unit Injection Improvements

Component costs per engine
Assembly, markup, and
warranty
Hardware cost per engine
Total fixed costs
Fixed cost per engine
Composite Unit Cost (1995 $)
Composite Unit Cost (1997 $)
100 kW
$63
$32
$95
$45,000
$87
$182
$190
400 kW
$98
$46
$144
$100,000
$232
$375
$392
750 kW
—
—
—
—
—
—
—
ISOOkW
—
—
—
—
—
—
—
3000kW
$4,000
$1,882
$5,882
*
*
$5,882
$6,150
  *Fixed costs for developing unit injectors are included under Engine Modifications.
                                            80

-------
                                                         Chapter 4: Economic Impact
                                        Table 4-5
                               Common Rail Fuel Injection

Component costs per engine
Assembly, markup, and
warranty
Hardware cost per engine
Total fixed costs
Fixed cost per engine
Composite Unit Cost (1995 $)
Composite Unit Cost (1997 $)
100 kW
$80
$23
$103
$100,000
$193
$296
$310
400 kW
$116
$34
$150
$100,000
$232
$381
$398
750 kW
$205
$59
$264
$300,000
$5,153
$5,417
$5,665
ISOOkW
$630
$183
$813
$300,000
$4,503
$5,315
$5,558
3000kW
$1,500
$435
$1,935
$1,000,000
$17,933
$19,868
$20,776
B. Engine Modifications

    Manufacturers have continued to pursue improvements in basic engine technology in an effort
to improve operating performance while controlling emissions.  Such variables include routing of
the intake air, piston crown geometry, and placement and orientation of injectors and valves. Most
of these variables affect the mixing of air and fuel in the combustion chamber. Small changes in
injection timing are also considered in this set of modifications. EPA expects that the base engines
from which the marine engines are derived will have been extensively redesigned for operation at
nonroad  Tier 2 emission levels. Remaining engine modifications reflect the need to adapt these
engines for complying with marine emission limits, including requirements to maintain control of
emissions over the whole range of marine engine operation.

    Because no performance improvements beyond those already achieved for nonroad engines are
anticipated, no cost savings  for  non-emission  benefits are applied in the  cost-effectiveness
calculation. Accounting for the full cost of engine modifications as an impact of new emission
standards also reflects  the extensive effort needed  to  conduct an  overall recalibration and
reoptimization of the engine according to the schedule dictated by the implementation of new
emission standards.  On  the other hand, no fuel economy penalty for retarded injection timing is
included. EPA believes that manufacturers will either not need to retard timing, or at least will be
able to optimize other variables such as injection pressure and aftercooling to overcome any effects
of retarding timing.  The possibility  of a fuel penalty is considered in the  sensitivity analysis
described in Section V below.

    As described in the cost  analysis  for the land-based  nonroad engine rulemaking, engine
modifications are expected to require extensive R&D and retooling, but no change in hardware costs.
The hardware costs included in this analysis for Category 2 engines are based on the different
                                           81

-------
Regulatory Impact Analysis
technology projections for Tier 2 locomotive engines. Table 4-6 shows the estimated per-engine
costs for these modifications.
                                       Table 4-6
                                  Engine Modifications

Hardware cost per engine
Total fixed costs
Fixed cost per engine
Composite Unit Cost
lOOkW
—
$200,000
$386
$386
400 kW
—
$200,000
$463
$463
750 kW
—
$1,100,000
$18,893
$18,893
ISOOkW
—
$1,100,000
$16,510
$16,510
3000kW
$800
$1,500,000
$26,900
$27,700
C. Turbocharging

    EPA expects that turbocharging will be needed by all marine diesel engines rated over 37 kW.
As shown in Table 4-2, except for some engines rated below 225 kW or above 2000 kW, all marine
diesel engines are  already turbocharged.  Turbocharger costs  for the remaining engines were
developed for EPA by Arcadis Geraghty and Miller.7  Since  R&D costs were not developed
separately for turbocharging, EPA conservatively estimated that one-fourth of the R&D required for
a turbocharging and aftercooling system will be needed for adding turbocharging alone.  This
accounts for the fact that much of the capital costs for turbocharger development is borne  by the
turbocharger manufacturer, which is then reflected in the hardware cost to the engine manufacturer.
Total turbocharger cost impacts are presented in Table 4-7.

    Turbocharging can be used to increase power, which leads to higher costs for virtually  all the
powertrain components. This analysis, however, does not presume an increase in power or improved
fuel economy resulting from turbocharging.  Rather, turbocharging  in the context of emission
controls is seen primarily as a technology that enables aftercooling, which has a great potential to
control NOx emissions.   Manufacturers may choose to exploit new turbochargers to  improve
performance,  but EPA believes that such an approach is not needed to comply with emission
standards. As described at the end of Section n, these non-emission benefits support the analytical
approach of assigning only half the cost of turbocharging as  an impact of the new emission
standards.
                                          82

-------
                                                          Chapter 4: Economic Impact
                                        Table 4-7
                                      Turbocharging

Hardware cost per engine
Total fixed costs
Fixed cost per engine
Composite Unit Cost
lOOkW
$208
$100,000
$193
$401
400 kW
—
—
—
—
750 kW
—
—
—
—
ISOOkW
—
—
—
—
SOOOkW
$2,133
$300,000
$5,380
$7,513
D. Aftercooling

    Cost estimates for adding or upgrading aftercooling systems were developed for EPA by Arcadis
Geraghty and Miller.8  Separate costs were developed for adding jacket-water aftercooling and
separate-circuit aftercooling (Tables 4-8 and 4-9, respectively).  A third scenario of improving
technology can be determined by calculating the net increase in cost for converting from j acket-water
aftercooling to separate-circuit aftercooling (Table 4-10). The ample supply of cooling water and
its greater convection coefficient relative to air support our belief that optimized separate-circuit
aftercooling systems can match or exceed the cooling capability of air-to-air cooling in land-based
engines.

    Cost estimates were developed for separate-circuit aftercooling systems using tube-and-shell
heat exchangers. An alternative design sometimes used on marine vessels is keel cooling, in which
coolant is plumbed from the engine or aftercooler to the hull, where it is cooled by exposure to the
seawater.  Upgrading systems that rely on keel cooling involves somewhat higher initial costs, but
this cost premium is offset by lower operating costs for maintenance or rebuild. Costs related to keel
cooling systems are not considered separately in this analysis.
                                           83

-------
Regulatory Impact Analysis
                                      Table 4-8
                  Incremental Cost of Adding Jacket-Water Aftercooling

Component costs per engine
Assembly, markup, and
warranty
Hardware cost per engine
Total fixed costs
Fixed cost per engine
Composite Unit Cost
lOOkW
$237
$383
$620
$400,000
$826
$1,446
400kW
$899
$746
$1,645
$550,000
$3,092
$4,737
750 kW
—
—
—
—
—
—
ISOOkW
—
—
—
—
—
—
3000kW
—
—
—
—
—
—
                                      Table 4-9
                 Incremental Cost of Adding Separate-circuit Aftercooling

Component costs per engine
Assembly, markup, and
warranty
Hardware cost per engine
Total fixed costs
Fixed cost per engine
Composite Unit Cost
lOOkW
$596
$790
$1,386
$480,000
$992
$2,378
400kW
$1,800
$1,765
$3,565
$660,000
$3,711
$7,276
750 kW
—
—
—
—
—
—
ISOOkW
—
—
—
—
—
—
3000 kW
$20,492
$11,113
$31,605
$1,440,000
$27,631
$59,236
                                         84

-------
                                                          Chapter 4: Economic Impact
                                        Table 4-10
       Incremental Cost of Converting from Jacket-Water to Separate-circuit Aftercooling

Component costs per engine
Assembly, markup, and
warranty
Hardware cost per engine
Total fixed costs
Fixed cost per engine
Composite Unit Cost
lOOkW
$359
$407
$766
$80,000
$166
$932
400 kW
$901
$1,019
$1,920
$110,000
$619
$2,539
750 kW
$1,701
$2,011
$3,712
$140,000
$2,573
$6,285
ISOOkW
$3,151
$3,189
$6,340
$200,000
$3,212
$9,552
3000kW
$5,964
$5,143
$11,107
$240,000
$4,605
$15,712
E. Rebuild Costs

    Operating data related to marine engines indicates that rebuild or remanufacture typically occurs
multiple times for an engine, especially for larger engines.  To the extent that manufacturers are
adding new hardware to an engine to comply with emission requirements, there may be an increase
in the cost of repair or replacement of parts over the life of an engine. This analysis includes an
estimate of increased operating costs by projecting a schedule of parts replacement coinciding with
periodic rebuilds. To calculate a total cost for each rebuild, the analysis provides for a rebuild rate
of one-third of total long-term variable costs for each rebuild event.  For some technologies, this
represents a scenario of one-third of all engines replacing all the hardware in the system  (such as
turbochargers). For other technologies, this represents a scenario of all engines needing replacement
of one-third of the value of new or improved components. For example, a rebuilder is not expected
to replace all the elements of a fuel injection system at the point of rebuild; rather, after searching
for defective or potentially defective parts, a rebuilder is expected to replace individual components
on an as-needed  basis.  The value of replacement parts  is calculated on a long-term basis, as
described below, then marked up by a factor of three to account for the higher cost of aftermarket
parts.

    Turbocharging and aftercooling are the only technologies expected to require an increase in
labor hours to rebuild an engine.   Since these systems are virtually custom-built in commercial
vessels, the analysis uses the same assembly times as were used to develop costs for new engines.

    The rebuild schedule for Category 1 engines is based on a sixteen-year life,  with two rebuilds
occurring at the end of the fifth and tenth years.  This represents median values for the fleet of
engines, rather than incorporating scrappage rates and additional rebuilds on a subset of the fleet.
Similarly, Category 2 engine rebuilds are based on a 23-year lifetime with three rebuilds occurring
after years  six, twelve, and eighteen. Lifetime costs are discounted to a net-present-value figure at
the point of sale using a 7 percent discount rate.  Rebuild costs are shown in Table 4-11.
                                            85

-------
Regulatory Impact Analysis
                                        Table 4-11
                                Incremental Rebuild Costs

lOOkW
400 kW
750 kW
ISOOkW
3000kW
Incremental hardware costs per engine:
Common rail
Unit injection
Turbocharger
Jacket-water aftercooling
Separate-circuit aftercooling
Aftercooling upgrade
$66
$40
$133
$152
$381
$230
$96
$62
—
$575
$1,152
$577
$169
—
—
—
$2,237
$1,089
$520
—
—
—
$5,642
$2,017
$1,238
$2,560
$1,365
—
$13,115
$3,817
Incremental labor hours per engine:
Turbocharger
Jacket-water aftercooling
Separate-circuit aftercooling
Aftercooling upgrade
1
5
11
6
—
5
20
15
—
—
40
30
—
—
60
45
6
—
90
68
F. Certification and Compliance

    EPA has significantly reduced certification requirements in recent years, but manufacturers are
nevertheless responsible for generating a minimum amount of test data and other information to
demonstrate compliance with emission standards. Table 4-12 lists the expected costs for different
sizes of engines, including the amortization of those costs over five years of engine sales. Estimated
certification costs are based on two engine tests and $ 10,000 worth of engineering and clerical effort
to prepare and submit the required information.

    Until engine designs are significantly changed, engine families can be recertified each year using
carryover of the original test data.  Since these engines are currently not subject to any emission
requirements, the analysis includes a cost to recertify an upgraded engine model every five years.

    Costs for production line testing (PLT) are summarized in Table 4-13. These costs are based
on testing 1  percent of total estimated sales, then distributing costs over the fleet.  Listed costs for
engine testing presume no need to build new test facilities, since most or all manufacturers already
                                            86

-------
                                                         Chapter 4: Economic Impact
have testing facilities available.  The final rule exclude the smallest companies from requirements
to conduct production line testing.

                                       Table 4-12
                                      Certification

Total fixed costs
Fixed cost per engine
Composite Unit Cost
lOOkW
$30,000
$58
$58
400 kW
$30,000
$70
$70
750 kW
$40,000
$687
$687
ISOOkW
$40,000
$600
$600
3000kW
$50,000
$897
$897
                                       Table 4-13
                             Costs for Production Line Testing

Cost per test
Testing rate
Cost per engine
lOOkW
$10,000
1%
$100
400 kW
$10,000
1%
$100
750 kW
$15,000
1%
$150
ISOOkW
$15,000
1%
$150
3000kW
$20,000
1%
$200
G. Total Engine Costs

    These individual cost elements can be combined into a calculated total for the new emission
standards by assessing the degree to which the different technologies will be deployed. As described
above, comparing the projected need for specific technologies can be compared to the technology
baseline to determine the changes that will occur in response to the new emission standards.

    To comply with emission standards, manufacturers are expected to rely on turbocharged engines
with extensive modification to injection systems and basic engine design.  Some engines are
expected to require new or improved aftercooling systems.  In general, the projected utilization of
each technology is expected to match that of land-based counterpart engines.  One exception to that
is our expectation that manufacturers will  not need to add electronic controls to comply with the
marine diesel emission standards.  With respect to aftercooling, jacket-water cooling systems are
considered equivalent to air-to-water designs, and separate-circuit systems are considered equivalent
to air-to-air designs.

    We also expect manufacturers to take a variety of approaches to improve fuel systems. Unit
inj ection systems can generally be further optimized for higher inj ection pressures and improved rate
shaping capability.  Nonroad engines with the highest sales volumes have begun converting to
                                           87

-------
Regulatory Impact Analysis
common rail fuel systems to take advantage of the step improvement in control of inj ection variables.
EPA believes that these fuel systems will be used similarly for land-based and marine engines.

    Factoring in the degree of deployment and adding up the costs of the individual technologies
results in a total estimated cost impact for engines designed and produced to meet emission limits.
As shown in Table 4-14, estimated costs for complying with emission standards  increase with
increasing power ratings. Estimated cost impacts range from $1,800 for a 100 kW engine to $54,000
for a 3,000 kW engine. Increased rebuild costs add to the total estimated impact, though to a much
lesser degree than from the expected increase in new engine costs (Table 4-15).  The percentages
listed in the tables represent the anticipated change in the individual technologies and should not be
confused with the total anticipated deployment of these technologies.

    Long-term  costs decrease due to two  principal factors.  First, the analysis anticipates that
manufacturers recover their initial  fixed costs for tooling, and R&D certification over a five-year
period.  Fixed costs are therefore applied only to the first five model years of production.

    The second modification is related to the effects  of the manufacturing learning curve. The
learning curve literature asserts that as cumulative production increases, the input requirements may
decrease as well. The magnitude of the reduction in production requirements may be expressed as
a progress ratio, p, where each doubling of cumulative past production leads  to  a "p" percent
reduction in input requirements, hence unit cost.   Areas involving direct labor and material are
usually the source of the greatest savings. These include, but are not limited to, a reduction in the
number or complexity of component parts,  improved component production, improved assembly
speed and processes, reduced error rates, and improved manufacturing process. These all result in
higher overall production, less scrappage of materials and products, and better overall quality. As
each successive p cycle takes longer to complete, production proficiency generally reaches a
relatively stable level, beyond which increased production does not necessarily lead to markedly
decreased costs.

    Companies and industries learn differently. On  average, it appears that doubling a firm's
cumulative output may be associated  with  the decrease  in the unit cost by 20% (Alchian 1963,
Argote and Epple 1990, Benkard 1999, Dutton  and Thomas 1984).  Dutton and Thomas (1984)
report rates varying from a 45 percent savings to a 7 percent increase in unit costs (see Figure 4-1).
The effect of learning curve seems to be less in the chemical industry and the nuclear power industry
where a doubling of cumulative  output is associated with 11% decrease in cost (Lieberman 1984,
Zimmerman 1982). The effect of learning is more difficult to decipher in the computer chip industry
(Gruber 1992).

    However, there are two  qualifiers to  the general findings.  First, many of  the empirical
estimations of the effect of learning curve  implicitly assume that there is no depreciation of the
learning. In contrast, Benkard (1999) finds that a firm's  knowledge may depreciate 39% in one year.
Argote,  et al (1990) increases this estimate to 96% in one year.  This suggests that under some
circumstances the effects of learning may not result in  enduring cost savings.
                                           88

-------
                                                        Chapter 4: Economic Impact
    Second, estimating cost from learning curves is very approximate. Button and Thomas (1984)
state that without understanding the dynamics of learning, a simplistic use of learning curve to
predict future costs has proved to be unreliable. Alchian (1963) finds that the difference between
the estimated labor use using learning curves and the actual labor used is 25% on average (using
industry learning curve, the difference ranges from +73% to -45%).
                                          89

-------
Regulatory Impact Analysis
                                        Figure 4-1

                            Distribution of Progress Ratios
       15
       10
    3-
    c
    as
            Ill   I I IIIIII IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII11  I  I   I  I   III


            55  57 59  61 63 65 67 69  71 73 75 77 79 81 83 85 87 89 91  93 95 97  99 101 103 105 107


                                         Progress Ratio
       22 field studies (n = 108).
 Figure 18

 Dutton and Thomas 1984
                                         90

-------
                                                          Chapter 4: Economic Impact
    EPA applied a p value of 20 percent in this analysis.  That is, the variable costs were reduced
by 20 percent for each doubling of cumulative production.  Using one year as the base unit of
production, the first doubling occurs at the start of the third model year and the second doubling at
the start of the fifth model year.  To align with the five-year amortization of fixed-costs, EPA
incorporated the second doubling at the start of the sixth model year. Variable costs are reduced by
a total of 36 percent using two "p" cycles with "p" valued at 20 percent. For a sensitivity analysis
of this assumption, see Table 4-16.

    EPA believes the use of the learning curve is appropriate to consider in assessing the cost impact
of diesel  engine emission controls.   The  learning curve  applies  to new  technology, new
manufacturing operations,  new parts, and new assembly operations.  While all the technologies
projected in this  analysis specify either upgraded existing designs or transferred nonroad engine
developments, the changes envisioned nevertheless require manufacture of new components and
assemblies, involving new manufacturing operations.  This should be especially true with marine
engines. Because of the relatively low sales volumes, manufacturers are less likely to put in the extra
R&D effort for low-cost manufacturing. As manufacturers gain experience with these new systems,
comparable learning is expected to occur with respect to unit labor and material costs.

    Table 4-14 also lists the long-term cost estimates for these engines the sixth and subsequent
years of production after incorporating these two changes. Reductions in long-term cost estimates
up to 90 percent demonstrate the predominance of research and other fixed  costs in the total
estimated impact of the emission standards.

    Focusing on the learning curve's effect on the cost estimates shows the sensitivity of these
assumptions.  Applying one stage of learning results in a 7 percent reduction in total costs for 100,
400, and 3000 kW engines. The 750 andlSOO kW engines rely predominantly on fixed costs, so the
learning curve has a negligible effect on those engines. Applying the second stage of learning would
further decrease the cost by 6 percent to 13 percent of the total cost. Table 4-16 contrasts these two
learning curve scenarios.
                                           91

-------
                                                                     Table 4-14
                                                                   Engine Costs

Common rail
Unit injection upgrade
Engine modifications
Turbocharger
Jacket-water
aftercooling
Separate-circuit
aftercooling
Upgrade to separate-
circuit aftercooling
Certification + PLT
Total Cost per Engine
(yr. 1-5)
Total Cost per Engine
(yr. 6 and later)**
100 kW
Fraction*
0%
100%
100%
65%
30%
10%
15%
100%
—
—
Cost
—
$190
$386
$261
$434
$238
$140
$158
$1,806
$486
400 kW
Fraction*
0%
100%
100%
0%
30%
0%
30%
100%
—
—
Cost
—
$392
$463
—
$1,421
—
$762
$170
$3,208
$846
750 kW
Fraction*
100%
0%
100%
0%
0%
0%
0%
100%
—
—
Cost
$5,665
—
$18,893
—
—
—
—
$837
$25,395
$856
ISOOkW
Fraction*
100%
0%
100%
0%
0%
0%
0%
100%
—
—
Cost
$5,558
—
$16,510
—
—
—
—
$750
$22,818
$1,120
3000kW
Fraction*
0%
100%
100%
10%
0%
10%
80%
100%
—
—
Cost
—
$6,150
$27,700
$751
—
$5,924
$12,570
$1,097
$54,192
$13,019
*"Fraction" denotes the percentage of engines estimated to
**Variable costs are reduced by a total of 36 percent using
require each of the new or improved technologies.
two "p" cycles with "p" valued at 20 percent. For a sensitivity analysis of this assumption, see Table 4-16.

-------
                                                                 Table 4-15
                                                                Rebuild Costs

lOOkW
Fraction*
Cost
400 kW
Fraction*
Cost
750 kW
Fraction*
Cost
ISOOkW
Fraction*
Cost
3000kW
Fraction*
Cost
Incremental hardware costs per engine:
Common rail
Unit injection upgrade
Turbocharger
Jacket-water aftercooling
Separate-circuit aftercooling
Upgrade to separate-circuit
aftercooling
Hardware cost per engine
Incremental labor costs per
engine:
Total cost per rebuild
Total rebuild cost per engine
(npv)
0%
100%
65%
30%
10%
15%
—
—
—
—
—
$40
$87
$46
$38
$34
$245
$116
$361
$441
0%
100%
0%
30%
—
30%
—
—
—
—
—
$62
—
$173
—
$173
$408
$168
$576
$703
100%
0%
0%
0%
0%
0%
—
	
—
—
$169
—
—
—
—
—
$169
$0
$169
$207
100%
0%
0%
0%
0%
0%
—
	
—
—
$520.00
—
—
—
—
—
$520.00
$0.00
$520
$635
0%
100%
10%
0%
10%
80%
—
—
—
—
—
$2,560
$137
—
$1,311
$3,054
$7,062
$1,778
$8,840
$12,430
'"Fraction" denotes the percentage of engines estimated to require each of the new or improved technologies.

-------
Regulatory Impact Analysis
                                       Table 4-16
               Long-Term Costs—Sensitivity of Learning Curve Assumptions
Number of p
cycles
1
2
lOOkW
$606
$486
400 kW
$1,065
$846
750 kW
$904
$856
ISOOkW
$1,267
$1,120
SOOOkW
$16,109
$13,019
H.  Sensitivity

    There  has  been some concern  expressed that the technologies used to meet emission
requirements for land-based engines will be less effective at controlling emissions from marine
engines. Some of the reasons suggested for needing a more aggressive approach include the change
in duty cycle, the effects of marinizing an engine, and the need to comply with emission limits across
not-to-exceed zones. Manufacturers could rely on inj ection timing retard as a technology option for
achieving an additional measure of NOx control.  Also, manufacturers may choose, for example, to
avoid the high R&D costs of implementing a new technology for an engine family with low sales
volume by relying on timing retard as a lower-cost alternative. EPA therefore conducted a sensitivity
analysis to show the costs associated with a fuel penalty resulting from relying on retarded timing.

    Because the requirement to control emissions throughout an engine's operating range poses the
greatest challenge at low speeds and loads, EPA calculated the costs of increasing fuel consumption
by one percent at modes 2 and 3 and by three percent at mode 4 (lightest load operation). Using the
weightings  for the composite duty cycle, increased life-cycle fuel consumption from this net 1.0
percent fuel penalty can be calculated and then discounted to the present at a 7 percent rate. The
resulting estimated net-present-value cost increase ranges from $400 for a 100 kW engine to $ 19,000
for a  3000 kW engine.  This  cost results from increased fuel  consumption.   Considering the
established effectiveness  of timing retard as a strategy to control NOx emissions, this may be
considered  a viable approach, either as a substitute or a supplemental technology.

IV. Aggregate costs

    The above analysis presents unit cost estimates for each power category. With current data for
engine and vessel sales for each category and proj ections for the future,  these costs can be translated
into projected direct costs to the nation for the new emission standards in any year. Aggregate costs
are estimated at about $10 million in the first year the new standards apply, increasing to a peak of
about $29 million in 2007 as increasing numbers of engines become subject to the new standards.
The following years show a drop in aggregate costs as the per-unit cost of compliance decreases,
resulting in aggregate costs of about $5 million after 2011.
                                           94

-------
                                                       Chapter 4: Economic Impact
Chapter 4 References
1."Estimated Economic Impact of New Emission Standards for Heavy-Duty Highway Engines,"
Acurex Environmental Corporation Final Report (FR 97-103), March 1997 (Docket A-97-50;
document IV-A-3). The Acurex Environmental Corporation has since changed its name to
Arcadis Geraghty & Miller.

2."Incremental Cost Estimates for Marine Diesel Engine Technology Improvements,"
Memorandum from Louis Browning and Kassandra Genovesi, Arcadis Geraghty & Miller, to
Alan Stout, EPA, September 30, 1998 (Docket A-97-50; document II-A-2).

3."Update of EPA's Motor Vehicle Emission Control Equipment Retail Price Equivalent (RPE)
Calculation Formula," Jack Faucett Associates, Report No. JACKFAU-85-322-3, September
1985 (Docket A-97-50; document IV-A-5).

4.Tables showing the full development of cost estimates for each technology are included in a
memorandum to the docket entitled, "Final Cost Estimates for Marine Diesel Engine Emission
Control Technology," November 17,  1999  (Docket A-97-50; document IV-B-3).

5. "The Year in Review: The Beat Goes On," Diesel Progress, June 1998, pp. 42, 44, 68 (Docket
A-97-50; document II-G-13).

6."Cat Gears Up Next Generation Fuel Systems," Diesel Progress, August 1998, p. 82 (Docket
A-97-50; document II-G-12).

7. See reference 2—Arcadis Geraghty & Miller, 1998.

8. See reference 2—Arcadis Geraghty & Miller, 1998.

Alchian, Armen.  (1963) Reliability of Progress Curves in Airframe Production. Econometrica,
    31(4), 679-.
Argote, Linda, Sara L. Beckman, and Dennis Epple.  (1990)  The Persistence and Transfer of
    Learning in Industrial  Settings.  Management Science, 36(2), 140-154.
Argote, Linda, and Dennis Epple. (1990) Learning Curves in Manufacturing. Science, 247(4945),
    920-924.
Benkard, C. Lanier (1999)  Learning and Forgetting: The Dynamics of Aircraft Production. NBER
    Working Paper, #7127.
Dutton, John M., and Annie  Thomas.  (1984)   Treating Progress Functions as a Managerial
    Opportunity. Academy of Management Review, 9(2), 235-247.
Gruber, Harald.  (1992) The Learning Curve in the Production of Semiconductor Memory Chips.
    Applied Economics, 24(8), 885-.
Lieberman, Marvin B.  (1984)  The Learning Curve and Pricing in the Chemical Processing
    Industries. RAND Journal of Economics, 15(2), 213-228.

                                         95

-------
Regulatory Impact Analysis
Zimmerman, Martin B.  (1982) Learning Effects and the Commercialization of New Energy
    Technologies: the Case of Nuclear Power. Bell Journal of Economics, 13(2), 297-310).
                                        96

-------
                                                  Chapter 5: Environmental Impacts
         CHAPTER 5:  ENVIRONMENTAL IMPACTS
    This chapter describes the expected environmental impacts of the new emission standards,
which focus on reducing HC, NOx, CO, and PM emissions. Specifically, the first part of the chapter
will discuss the health and welfare impacts of these pollutants. The second part of the chapter
estimates the total nationwide emissions inventory for marine diesel engines and projects future
emissions and emission reductions.

I. Health and Welfare Concerns

    As part of the periodic review of the ozone and PM air quality standards required under the
Clean Air Act, EPA has reassessed the impacts of ozone and PM on human health and welfare,
taking into account peer-reviewed scientific information. The paragraphs below summarize some
of EPA's current concerns, as compiled in the Agency's Criteria Documents and Staff Papers for
ozone and PM. The Criteria Documents prepared by the Office of Research and Development
consist of EPA's latest summaries of scientific and technical information on each pollutant.  The
Staff Papers on ozone and PM prepared by the Office of Air Quality Planning and Standards
summarize the policy-relevant key findings regarding health and welfare effects.

A. Ozone

    Ground level ozone is formed when hydrocarbons and oxides of nitrogen react in the presence
of sunlight.  Over the past few decades, many researchers have investigated the health effects
associated with both short-term  (one-  to three-hour)  and prolonged acute (six- to eight-hour)
exposures to ozone.  In the past decade, numerous controlled-exposure studies of moderately
exercising human subjects have been conducted which collectively allow a quantification of the
relationships between prolonged acute ozone exposure and the response of people's respiratory
systems under a variety of environmental conditions. To this experimental work has been added
field and epidemiological studies which provide further evidence of associations between short-term
and prolonged acute ozone exposures and health effects ranging from respiratory symptoms and lung
function decrements to increased hospital admissions for respiratory causes. In addition to these
health effects, daily mortality studies have suggested a possible association between ambient ozone
levels and an increased risk of premature death.

    Most of the recent controlled-exposure ozone studies have shown that respiratory effects similar
to those found in the short-term exposure studies occur when human subjects are exposed to ozone
concentrations as low as 0.08 ppm while engaging in intermittent, moderate exercise for six to eight
hours.  These effects occur even though ozone concentrations and levels of exertion are lower than
in the earlier short-term exposure studies. They appear to build up over time, peaking in the six- to
eight-hour time frame.  Other effects, such as the presence of biochemical indicators of pulmonary
inflammation and increased susceptibility to infection, have also been reported  for prolonged
                                          97

-------
Regulatory Impact Analysis
exposures and, in some cases, for short-term exposures. Although the biological effects reported in
laboratory animal studies can be extrapolated to human health effects only with great uncertainty,
a large body of toxicological evidence exists which suggests that repeated exposures to ozone causes
pulmonary inflammation similar to that found in humans and over periods of months to years can
accelerate aging of the lungs and cause structural damage to the lungs.

    In addition to human health effects, ozone is known to adversely affect the environment in many
ways.  These effects include reduced yield for commodity crops, for fruits and vegetables,  and
commercial forests; ecosystem  and vegetation effects in such areas as National Parks (Class I areas);
damage to urban grass, flowers, shrubs, and trees; reduced yield in tree seedlings and noncommercial
forests; increased susceptibility of plants to pests; materials damage; and reduced visibility. Nitrogen
oxides (NOx), key precursors to ozone, also result in nitrogen deposition into sensitive nitrogen-
saturated coastal estuaries and  ecosystems, causing increased growth of algae and other plants.

B. Particulate Matter

    Particulate matter (PM) represents abroad class of chemically and physically diverse substances
that exist as discrete particles (liquid droplets or solids) over a wide range  of sizes.  Human-
generated sources of particles include a variety of stationary and mobile sources.  Particles may be
emitted directly to the atmosphere or may be formed by transformations of gaseous emissions such
as sulfur dioxide or nitrogen oxides. The maj or chemical and physical properties of PM vary greatly
with time, region, meteorology, and source category, thus complicating the assessment of health and
welfare effects as related to various indicators of particulate pollution. At elevated concentrations,
particulate matter can adversely affect human health, visibility, and materials.  Components of
particulate matter (e.g., sulfuric or nitric acid) contribute to acid deposition.

    Key EPA findings can be  summarized as follows:

1.   Health risks posed by inhaled particles are affected both by the penetration and deposition of
    particles in the various regions of the respiratory tract, and by the biological responses to these
    deposited materials.

2.   The  risks of adverse effects associated with deposition  of ambient particles in the thorax
    (tracheobronchial and alveolar regions of the respiratory tract) are markedly greater than for
    deposition  in the extrathoracic (head) region. Maximum particle penetration to the thoracic
    regions occurs during oronasal or mouth breathing.

3.   The  key health effects categories associated with PM include premature death; aggravation of
    respiratory and cardiovascular disease, as indicated by  increased hospital admissions  and
    emergency room visits, school absences, work loss days, and restricted activity days; changes
    in lung function and increased respiratory symptoms; changes to lung tissues and structure; and
    altered respiratory defense mechanisms. Most of these effects have been consistently associated
    with ambient PM concentrations, which have been used as a measure of population exposure,
    in a large number of community epidemiological studies. Additional information and insights

                                            98

-------
                                                     Chapter 5: Environmental Impacts
    on these effects are provided by studies of animal toxicology and controlled human exposures
    to various constituents of PM conducted at higher than ambient concentrations.  Although
    mechanisms by which particles cause effects are not well known, there is general agreement that
    the cardio-respiratory system is the major target of PM effects.

4.   Based on a qualitative assessment of the epidemiological evidence of effects associated with
    PM for populations that appear to be at greatest risk with respect to particular health endpoints,
    the EPA has concluded the following with respect to sensitive populations:

    a.   Individuals with respiratory disease (e.g., chronic obstructive pulmonary disease, acute
         bronchitis) and cardiovascular disease (e.g., ischemic heart disease) are at greater risk of
         premature mortality and hospitalization due to exposure to ambient PM.
    b.   Individuals with infectious respiratory  disease (e.g., pneumonia) are at greater risk of
         premature mortality and morbidity (e.g.,  hospitalization, aggravation of respiratory
         symptoms)  due to  exposure to ambient PM.   Also,  exposure to PM may increase
         individuals' susceptibility to respiratory infections.

    c.   Elderly individuals are also at greater risk of premature mortality and hospitalization for
         cardiopulmonary problems due to exposure to ambient PM.

    d.   Children are at greater risk of increased respiratory symptoms and decreased lung function
         due to exposure to ambient PM.

    e.   Asthmatic individuals are at risk of exacerbation of symptoms associated with asthma, and
         increased need for medical attention, due to exposure to PM.

5.   There are fundamental physical and chemical differences between fine and coarse fraction
    particles, and it is reasonable to expect that differences may exist between the two subclasses
    of PM-10 in both the nature of potential effects and the relative concentrations required to
    produce such effects. The specific components of PM that could be of concern to health include
    those typically within the fine fraction (e.g., acid aerosols, sulfates, nitrates, transition metals,
    diesel particles, and ultra fine particles), and those typically within the coarse fraction (e.g.,
    silica and resuspended dust).  While components of both fractions can produce health effects,
    in general, the fine fraction appears to contain more of the reactive substances potentially linked
    to the kinds of effects observed in the epidemiological studies. The fine fraction also contains
    the largest number of particles and a much larger aggregate surface area than the coarse fraction
    which enables the fine fraction to have a substantially greater potential for absorption and
    deposition in the thoracic region, as well as for dissolution or absorption of pollutant gases.

    With respect to welfare or secondary effects, fine particles have been clearly associated with the
impairment of visibility over urban areas and large multi-state regions.  Fine particles, or major
constituents thereof, also are implicated in materials damage, soiling and acid deposition. Coarse
fraction particles contribute to soiling and materials damage.
                                            99

-------
Regulatory Impact Analysis
    Particulate pollution is a problem affecting localities, both urban and nonurban, in all regions
of the United States.  Manmade emissions that contribute to airborne particulate matter result
principally from stationary point sources (fuel combustion and industrial processes), industrial
process fugitive particulate emission sources, nonindustrial fugitive sources (roadway dust from
paved and unpaved roads, wind erosion from cropland, etc.) and transportation sources. In addition
to manmade emissions, consideration must also be given to natural emissions including dust, sea
spray, volcanic emissions, biogenic emanation (e.g., pollen from plants), and emissions from wild
fires when assessing particulate pollution and devising control strategies.

C. Carbon Monoxide

    Though carbon monoxide (CO) is not the primary focus of this rule, EPA is adopting new
standards for CO.  Carbon monoxide has long been known to have substantial adverse effects on
human health and welfare, including toxic effects on blood and tissues, and effects on organ
functions, and has been linked to fetal brain damage, increased risk for people with heart disease,
and reduced visual perception, cognitive functions and aerobic capacity. Due to recent emission
standards, the number of areas in nonattainment for CO has greatly diminished in the past decade.
There are approximately 20 remaining serious or moderate CO nonattainment areas.

D. Smoke

    Smoke from diesel engines has long been associated with adverse effects on human welfare,
including considerable economic, visibility and aesthetic damage. The carbon particles that make
up diesel smoke cause reduced  visibility; soiling of urban buildings,  homes, personal property,
clothes, and skin; and are associated with increased odor, coughing, and eye irritation. In addition,
the particles causing visible smoke  are the same as those associated with the significant threats to
human health described above for particulate matter.

II. Emission Reductions

    For the purposes of this rulemaking, EPA has divided marine diesel engines into three
categories. Category  1  engines are smaller  engines used in recreational  and light commercial
applications, Category 2 engines are medium-sized engines such as those used in tugboats, and
Category 3 engines are the large engines used primarily for propulsion in ocean-going and some
Great Lakes vessels.   Because of the distinctly different characteristics between the design and
operation of the marine diesel engines in these three categories, the final rule has provisions that are
unique to each category. Also, due to distinctions between the categories, different data sources and
approaches were required to develop an inventory for the three categories.  The remainder of this
chapter will describe the inventory calculations for the categories separately then combine them to
discuss the nationwide benefits.
                                           100

-------
                                                   Chapter 5: Environmental Impacts
A. Category 1

    1. General Methodology

    In the inventory calculations, Category 1 engines were divided into recreational, commercial,
and auxiliary applications. The applications were further divided into power ranges consistent with
the way the standards apply to different sets of engines.  Each of the engine applications and power
ranges were modeled with distinct annual hours of operation, load factors, and average engine lives.
The basic equation for determining the emissions inventory, for a single year, from marine engines
is shown below:
   Emissions =
                   E
              rec, com, aux
  E
ranges
(population x power x load* annual use x emission/actor]
    This equation sums  the total  emissions for each of the power ranges  and applications.
"Population" refers to the number of marine diesel engines greater than or equal to 37 kW but with
a displacement per cylinder of five liters or less estimated to be in the U.S. in a given year. "Power"
refers to the population-weighted average rated power for a given power range. Two usage factors
are included; "load" is the ratio between the average operational power output and the rated power,
and "annual use" is the average hours of operation per year. Emission factors are applied on a brake-
specific basis  (g/kW-hr)  and represent the weighted value between levels from baseline and
controlled marine  engines operating in a given calendar year. Emission inventories from Category
1 engines were calculated for HC, NOx, CO, and PM.  Although the emission standards combine
HC and NOx,  it is useful to consider the  HC and NOx emission impacts separately.  The split
between HC and NOx for regulated engines is based on the data from Chapter 3 for technologies that
are expected to be used to meet the new emission standards.

    2. Inputs to the Inventory Calculations

    Data on the various inputs were obtained from several sources. Engine populations were taken
from the 1997 Power Systems Research (PSR) Parts Link database. This database contains marine
engine populations and technical information by engine model for 1990 through 1997. From this
data, EPA was able to focus on specific applications and power ranges.  To determine turnover rates
for the purpose of determining the introduction of controlled engines, scrappage rates are needed.
Through the combination of historical population and scrappage rates, historical sales and retirement
of engines can be estimated. EPA used the normalized scrappage rate developed by PSR and fit it
to the data using assumed average useful lives for each of the three applications. Figure 5-1 presents
the normalized scrappage curve. The average useful lives for each of the three  applications were
based on conversations with manufacturers and are estimated to be 15,13, and 17 years respectively
for recreational, commercial, and auxiliary marine diesel engines.
                                          101

-------
Regulatory Impact Analysis
                       Figure 5-1:  Normalized Scrappage Curve
  8
           0               0.5               1               1.5               2
                  Engine Age Normalized by Average Useful Life
    To project future populations, EPA applied growth rates to each of the three application types.
Based on the eight years of population data (1990-1997) supplied by the PSR Parts Link database,
growth rates of 6.3 percent for recreational, 0.6 percent for commercial, and 7.5 percent for auxiliary
were observed.  EPA was concerned that eight years of data were not enough to determine a
historical growth trend, especially since the growth rates for these years were so high for recreational
and auxiliary marine engines. To estimate growth over a larger number of years, EPA relied on
PSR's OE Link database which provides detailed information on the U.S. production of marine
diesel engines from 1980 to 1997. Based on the PSR production data, EPA calculated growth rates
of 3.5 percent for recreational, 0.9 percent for commercial, and 1.5 percent for auxiliary marine
diesel engines.  In its final calculations, EPA used  the growth rates based on U.S. production.
Although the growth rate for recreational marine diesel engines is significantly larger than for
commercial and auxiliary, EPA believes that it is reasonable due to recently expanded efforts of
marine diesel engine manufacturers to break into this potentially high profit market.

    Estimated engine populations are presented in Table 5-1. These figures show that recreational
marine engines are the largest segment of the U.S. Category 1 marine diesel engine population. As
will be shown below, however,  their inventory contribution is much less than that of the smaller
commercial population due to different usage rates.
                                         102

-------
                                                   Chapter 5: Environmental Impacts
                                        Table 5-1
                   Projected Category 1 Populations by Year [thousands]

recreational
commercial
auxiliary
1995
142
57
10
2000
180
58
13
2005
214
61
14
2010
254
64
15
2020
358
70
17
2030
505
76
20
    The remaining factors needed to estimate the category 1 emissions inventory are engine usage
and emission rates.  Recreational and commercial marine diesel engine usage characteristics are
based on a compilation of data supplied to the EPA by the Engine Manufacturers Association
(EMA). This data is based on five years of propulsion marine diesel engine sales from seven engine
manufacturers who make up the strong maj ority of the U. S. market.6 Included in the data submission
were average annual hours of use, load factors, and emission factors broken down by ranges of rated
power and rated speed.  Because the submission did not include auxiliary  marine engines,
conversations with individual manufacturers were used to develop usage inputs for these engines.
Baseline emission factors were developed using the both the EMA submission and the data for
uncontrolled marine diesel engines presented in Chapter 3. The usage inputs and baseline emission
factors, used in the inventory calculations, are presented in Tables 5-2 and 5-3.

                                        Table 5-2
                     Usage Inputs for Category 1 Inventory Calculations
Power
lange [kW]
37-75
75-130
130-225
225-450
450-560
560+
Load Factor [percent]
recreational
25
25
25
30
40
40
commercial
60
60
60
71
73
79
auxiliary
65
65
65
65
65
65
Annual Use [hours]
recreational
125
175
175
225
500
500
commercial
2320
2350
2270
3240
3770
4500
auxiliary
2500
2500
2500
2500
2500
2500
       e  Although difficult to compare directly, these sales figures seemed to agree with the PSR
data fairly well.
                                          103

-------
Regulatory Impact Analysis
                                        Table 5-3
                  Baseline Emission Factors for Category 1 Marine Engines
Power Range
[kW]
37-75
75-130
130-225
225-450
450-560
560-1000
1000+
HC
[g/kW-hr]
0.27
0.27
0.27
0.27
0.27
0.27
0.27
NOx
[g/kW-hr]
11
10
10
10
10
10
13
CO
[g/kW-hr]
2.0
1.7
1.5
1.5
1.5
1.5
2.5
PM
[g/kW-hr]
0.90
0.40
0.40
0.30
0.30
0.30
0.30
    The new emission standards, presented in Chapter 1, were used as the emission factors for the
controlled engines. Not enough information is available at this time to account for deterioration or
manufacturer compliance margins. EPA believes that these factors will not have a large effect on
the final calculations. At worst, this methodology results in conservative benefit estimates since the
standards represent levels at the useful lives of the engines.  HC and NOx are combined in a single
numerical emission limit.  To separate them for inventory analysis, EPA estimates that the HC
emissions will be reduced by 0.07 g/kW-hr reduction. This estimate is based on the emission data
from the low-emission engines presented in Chapter 3.

    3. Inventory Results

    Table 5-4 presents the estimated baseline emissions for recreational, commercial, and auxiliary
marine engines.  Although the final rule does not include standards for recreational marine diesel
engines, they are included in this analysis for completeness.
                                          104

-------
                                                  Chapter 5: Environmental Impacts
                                       Table 5-4
           Projected Category 1 Baseline Emissions by Year [thousand short tons]

HC
NOx
PM
CO
recreational
commercial
auxiliary
total
recreational
commercial
auxiliary
total
recreational
commercial
auxiliary
total
recreational
commercial
auxiliary
total
1995
0.6
9.9
0.7
11.3
21.9
382
28.3
432
0.7
12.0
1.1
13.9
3.4
60.2
4.5
68.0
2000
0.8
10.3
1.0
12.1
29.7
398
38.0
465
1.0
12.5
1.5
14.9
4.7
62.8
6.0
73.4
2005
0.9
10.8
1.1
12.8
35.3
416
40.9
492
1.1
13.1
1.6
15.8
5.5
65.6
6.4
77.6
2010
1.1
11.3
1.2
13.6
42.0
435
44.1
521
1.4
13.7
1.7
16.8
6.6
68.6
6.9
82.1
2020
1.6
12.4
1.4
15.3
59.2
476
51.1
586
1.9
14.9
2.0
18.9
9.3
75.1
8.0
92.4
2030
2.2
13.5
1.6
17.3
83.5
520
59.3
663
2.7
16.3
2.3
21.4
13.1
82.1
9.3
105
    This analysis suggests that commercial marine engines make up the majority of the Category
1 emissions (i.e. 85 percent of HC+NOx in 2000) even though they made up only 23 percent of the
population of Category 1 engines in 2000. This is mostly due to the high annual use and load factors
of these engines. By comparison, recreational engines represent 72 percent of the population in
2000, but account for only 6 percent of baseline Category 1 HC+NOx emissions. However, due to
their higher projected growth rate, recreational engines become a growing source of marine diesel
emissions in future years (10 percent of the baseline in 2020).
    4. Emissions Benefits

    Tables 5-5 and 5-6 show the expected impact of the new emission standards on HC, NOx, and
PM emissions from commercial and auxiliary Category  1 engines.  The Tier 2 standards should
reduce emissions from each of these engines by 26 percent for HC, 29 percent for NOx, and 38
                                         105

-------
Regulatory Impact Analysis
percent for PM.  These reductions vary for different power categories with the highest power
categories showing the greatest reductions.  This is reasonable since these engines are used more
intensely than the others.  It should be noted that for the cost-effectiveness calculations in Chapter
6, only the NOx reductions beyond the MARPOL standard are considered.  For Category 1, the
MARPOL standard represents a 3.5 percent reduction in NOx from new engines compared to
baseline.  The CO standard serves as a cap on the already low emissions from diesel engines;
therefore, no emissions reductions are claimed here for CO.  EPA is not including standards for
Category 1 recreational engines in this rule.  Thus, while Category 1 recreational engine emissions
will be reduced somewhat as a result of the MARPOL standards, EPA is not claiming any benefits
for  Category 1 recreational engines here.  Also, no Category  1 recreational engine emission
reductions are included in the cost-effectiveness calculations in Chapter 6.

                                        Table 5-5
                             Projected Controlled Category 1
             Commercial Marine Engine Emissions [thousand short tons per year]
Modeled Item
HC
NOx*
PM
controlled level
benefit
reduction
controlled level
benefit
reduction
controlled level
benefit
reduction
1995
9.9
0.0
0%
382
0.0
0%
12.0
0.0
0%
2000
10.3
0.0
0%
396
1.2
0%
12.5
0.0
0%
2005
10.5
0.3
3%
396
20.2
5%
12.5
0.5
4%
2010
10.0
1.3
12%
367
68.0
16%
11.4
2.3
16%
2020
9.4
3.0
24%
330
145
31%
9.8
5.1
34%
2030
10.0
3.5
26%
351
170
33%
10.3
6.0
37%
* The NOx reductions shown here include reductions from the MARPOL standards as well as those from the Tier 2
standards beyond the MARPOL standards. In 2030 the MARPOL standards account for 12 percent of the total Category
1 NOx reductions.
                                           106

-------
                                                    Chapter 5: Environmental Impacts
Table 5-
Projected Controllec
Auxiliary Marine Engine Emissions
Modeled Item
HC
NOx*
PM
controlled level
benefit
reduction
controlled level
benefit
reduction
controlled level
benefit
reduction
1995
0.7
0.0
0%
28.3
0.0
0%
1.1
0.0
0%
2000
1.0
0.0
0%
37.9
0.1
0%
1.5
0.0
0%
5
Category 1
thousand short tons per year]
2005
1.1
0.0
3%
39.4
1.5
4%
1.6
0.1
4%
2010
1.1
0.1
10%
38.8
5.2
12%
1.5
0.3
15%
2020
1.1
0.3
22%
37.7
13.4
26%
1.3
0.7
35%
2030
1.2
0.4
25%
41.3
18.0
30%
1.4
1.0
41%
* The NOx reductions shown here include reductions from the MARPOL standards as well as those from the Tier 2
standards beyond the MARPOL standards. In2030 the MARPOL standards account for 12 percent of the total Category
1 NOx reductions.
    In 2010, the new standards are expected to result in reductions of 12 percent HC, 16 percent
NOx, and 16 percent PM from Category 1  marine diesel engines.  Once the effects of the new
standards are fully phased  in  (i.e. in 2035 the new  emission  standards will lead to  estimated
reductions of 24 percent HC, 27 percent NOx, and 36 percent PM from Category 1 engines. These
figures are lower than the per-engine reductions because they do not include any potential future
emissions reductions from recreational marine diesel engines. Table 5-7 shows the actual numbers
of category 1 commercial and recreational engines subject the MARPOL and Tier 2 standards.

                                        Table 5-7
         Projected Category 1  Commercial  and Auxiliary Engines Subject to Standards

2000
2005
2010
2020
2030
Uncontrolled
67,000
50,100
32,600
6,000
600
MARPOL
3,900
17,500
17,500
8,300
1,500
Tier 2
0
6,900
28,300
72,400
93,900
                                           107

-------
Regulatory Impact Analysis
    As described in the preamble to the final rule, the manufacturers are responsible to comply with
emission limits at any speed and load that can occur on a vessel. These "not-to-exceed" provisions
are needed to keep pace with the increasing sophistication of diesel engine technology.  These
provisions help ensure that engines are in fact operating within allowable emission levels, rather than
requiring emission reductions  beyond that called  for by the emission limits.  This analysis of
emission benefits thus does not take into account  any additional benefit associated with not-to-
exceed provisions.

B. Category 2

    1. Baseline Emission Inventories

    Baseline emissions inventories for Category 2 marine diesel engines are based on an analysis
developed by Corbett and Fischbeck.1 In that analysis, two separate methods were used to develop
inventories for U.S. flagged vessels and foreign vessels used in U.S. domestic waters.  In addition,
the study investigated marine engine types and replacement rates.  Below  is an overview of the
analysis; more detail may be found in the report.

    Emissions inventories for U.S.  flagged vessels were  estimated using ship registry data on
commercial vessels greater than 100 gross registered tons.  The general methodology is based on
estimates of daily fuel consumption.  Specifically, engine rated power, operation, fuel consumption
and fuel specific emissions factors, developed by Lloyd's Register, were used to generate annual
mass of emissions from a given vessel. These emission numbers were multiplied by the number of
U. S. flagged vessels for discrete applications to generate a baseline emission inventory for domestic
ships.

    Estimated emissions inventories for foreign ships operating in U. S. navigable waters were based
on cargo movements and waterways data. Using cargo movements and geographical data, the ton-
miles of cargo moved in each region were calculated for the Great Lakes, inland waterways, and
coastal waters up to 200 miles offshore.  Emissions per ton  mile were based on the usage  and
emission factors used for U.S. vessels described above, average dead weight tonnage per ship,
assumed cargo capacity factors, and average vessel  speed for the duty cycle. The results from this
analysis are presented in Table  5-8.
                                           108

-------
                                                   Chapter 5: Environmental Impacts
                                       Table 5-8
                      Baseline Emissions from Category 2 CI Marine
               Engines Operated in U.S. Waters [thousand short tons per year]
Year
2000
2010
2020
2030
HC
11.1
12.3
13.6
15.0
NOx
267
295
325
360
PM
6.1
6.8
7.5
8.3
CO
34.1
37.7
41.7
46.0
    2. Emission Benefits

    To calculate the emissions benefits of the new emission standards, the replacement rates of old
engines with new engines must be known.  For this analysis, the average useful life of 23 years
reported in the report by Corbett and Fischbeck was used. This report also discusses a growth rate
of 2 percent based on production and turnover in recent years. However, EPA was concerned that
a few years of data was not enough to establish a growth trend for this industry.  Therefore, EPA
conservatively is using a growth rate of 1 percent which is more consistent with the Category 1
growth rates used in this analysis. By combining the 23 year average useful life with the normalized
scrappage curve presented in Figure 5-1  and the estimated growth  rate,  EPA developed the
replacement schedule used for the inventory analysis. Because of the high average age of the U.S.
fleet, EPA believes that it is likely that their will be higher turnover than usual in the near future.
This would result in greater benefits sooner. However, EPA has not made any attempts to include
this dynamic in its inventory analysis.

    Emissions reductions were based on the baseline brake-specific emission levels presented in
Chapter 3 and the emission standards presented in Chapter 1. Since EPA is only claiming NOx
reductions from the new emission standards beyond those resulting from the MARPOL standards,
the MARPOL reduction was calculated. The MARPOL standard will result in an average reduction
of 14  percent from new engines compared to baseline engines. On a per engine basis, the Tier 2
standards represent reductions of 38 percent NOx.  In the benefits modeling described here, these
reductions are only applied to engines in U.S. flagged vessels operated in U.S. waters.  No HC
benefits are claimed for Category 2 engines  due to their already low HC emissions.  Also, the PM
and CO standards for Category 2 engines serve as caps, and no reductions in those pollutants are
claimed.  Because of the slow turnover rates, the full effects of these reductions will not be seen until
after 2045. Projected benefits of the MARPOL requirements and the new standards for Category
2 engines are shown in Table 5-9. No NOx reductions are assumed for foreign flagged ships since
it is unknown at this time which countries will adopt the MARPOL requirements. Foreign flagged
ships make up about 4 percent of the Category 2 NOx inventory in U.S. waters. Table 5-10 shows
the projected number of Category 2 engine subject to the MARPOL and Tier 2 standards.

                                       Table 5-9
                                          109

-------
Regulatory Impact Analysis
                   Projected NOx Emission Reductions from Category 2
              CI Marine Engines Operated in U.S. Waters [thousand short tons]
Year
2000
2010
2020
2030
MARPOL NOx
1.4
17.0
33.8
44.5
EPA NOx
0.0
11.1
41.7
72.1
                                       Table 5-10
                     Projected Category 2 Engines Subject to Standards

2000
2005
2010
2020
2030
Uncontrolled
1,550
1,283
995
396
103
MARPOL
66
416
498
480
179
Tier 2
0
0
293
1,097
1,897
    As described for Category 1 engines above, this calculation of emission benefits for Category
 2 engines does not take into account any additional benefits for not-to-exceed provisions.

C. Category 3

    1. Baseline Emissions

    Baseline emissions were calculated for Category 3 using the same methodology that was used
for Category 2. The difference between Category 2 and 3 is that foreign vessels only make a small
fraction of the national NOx for Category 2, while they make up about half of the national NOx for
Category 3. This difference is easily explained by the fact that most Category 3 engines are used on
ocean-going vessels, and while U.S. is a large foreign trade country, most of the ocean-going vessels
that visit U.S. ports are foreign flag.  Baseline emission results  from the Category 3 inventory
calculations are presented in Table 5-11.
                                           110

-------
                                                   Chapter 5: Environmental Impacts
                                       Table 5-11
                      Baseline Emissions from Category 3 CI Marine
               Engines Operated in U.S. Waters [thousand short tons per year]
Year
2000
2010
2020
2030
HC
8.1
9.0
9.9
10.9
NOx
273
301
333
368
PM
21.2
23.4
25.8
28.6
CO
25.0
27.6
30.5
33.7
    2. Emission Benefits

    This section calculates the emissions benefits for Category 3 engines associated with adopting
the MARPOL requirements for Category 3 engines on U.S. flagged ships. Emissions benefits were
calculated using the same useful life and growth rate used for in the Category 2 analysis based on
the limited data available to EPA.  NOx reductions were determined by comparing the data in
Chapter 3 to the emission limits presented in Chapter 1.  As discussed in Chapter 3, the analysis
assumes that engines certifying on diesel oil but operating on residual fuel have 10 percent higher
NOx emissions in use.  Because the MARPOL standards are not part of this rule, and EPA is setting
no additional requirements for Category 3 engines, no benefits are claimed for Category 3 engines
in the cost-effectiveness calculations contained in Chapter 6.

    EPA estimates that adopting the MARPOL requirements will result in a 17 percent reduction
in NOx from applicable new Category 3 engines. If this  standard were only met on U.S. flagged
ships, only about a 9 percent reduction in NOx is  anticipated once all U.S. flag ships meet the
MARPOL NOx standard. However, foreign flag ships may reduce NOx as other nations adopt the
MARPOL provisions.  The U.S. flag reductions and the potential effects (in U.S. waters) of world
wide application of the MARPOL NOx standard are both presented in Table 5-12 and Figure 5-1.
                                          Ill

-------
Regulatory Impact Analysis
                                        Table 5-12
                        Projected NOx Reductions from Category 3
              CI Marine Engines Operated in U.S. Waters [thousand short tons]

2000
2010
2020
2030
U.S. Flag
1.0
11.4
22.7
29.8
Foreign Flag*
0.9
11.2
22.3
29.4
Total*
1.9
22.5
45.0
59.2
* The U.S. could only ensure that the reductions from U.S. flagged ships are achieved.
                                           112

-------
                                                  Chapter 5: Environmental Impacts
     Figure 5-1:  Projected Baseline and Controlled National NOx Emissions from
     Category 3
       500
       400
       300
       200
       100
                                                          ™ Baseline
                                                          ~ U.S. Flag Control Only
                                                          — World Fleet Control
          2000
2010
    2020
Calendar Year
2030
2040
D. Nationwide Totals

    For the purposes of this analysis, the CI marine inventory has been divided into five sections:
Category 1 recreational,  commercial, and auxiliary; Category 2; and Category 3.  Of these five
categories, Category 1 commercial, Category 2, and Category 3 engines dominate the baseline
inventory from CI marine engines.

    The reductions associated with the new emission standards come from Category 1 commercial
and auxiliary, and Category 2.  Because of the relatively less stringent standards for Category 3
engines, they become a bigger fraction (from 27 percent to 38 percent) of the emissions inventory
in 2020.

    EPA used nationwide emission estimates from the 1997 Trends Report2 to compare the relative
emissions contribution from CI marine engines to other emission sources. Because Trends does not
report estimates for 1998, figures for 2000 are used here as the baseline.  The national HC, NOx,
PM, and CO emissions inventories are summarized in Table 5-13 along with the EPA estimates for
CI marine engines. These data indicate that emissions from baseline CI marine engines account for
8.1 percent of NOx and 4.8 percent of PM from mobile sources and 4.4 percent of NOx and 1.0
                                          113

-------
Regulatory Impact Analysis
percent of PM nationwide.  CI marine engines account for less than one percent of volatile organic
sources (VOC) or CO for mobile sources.

                                       Table 5-13
                       2000 National Emissions [thousand short tons]
Emission Source
CI Marine
Other Nonroad
Highway
Total Mobile Sources
Other Sources
Total Nationwide Emissions
VOC
31
2,044
4,482
6,526
9,567
16,093
NOx
1,005
4,930
6,397
12,332
10,550
22,882
PM
42
593
238
873
3,270*
4,143
CO
133
16,286
44,244
60,663
19,199
79,862
*does not include erosion or fugitive dust.
    Table 5-14 contains the baseline annual emissions from marine diesel engines as a whole as well
as projections of the annual emissions with the MARPOL requirements and EPA's new emission
standards in place. According to this analysis, the new emission standards will yield reductions from
the baseline of 8 percent HC, 21 percent NOx, and 11 percent PM from marine diesel engines in
2020. Only NOx reductions beyond those resulting from the MARPOL standards are claimed to
result from the Tier 2 standards.  Thus,  only those reductions beyond those resulting from  the
MARPOL standards are included in the cost-effectiveness values shown in Chapter 6. In 2020 the
Tier 2 standards will result in a 15 percent NOx reduction below the levels expected from  the
MARPOL standards.

    Table 5-15 contains a similar analysis as that shown in Table 5-14 but is limited only to those
engines  covered by the Tier 2 standards (Category 1 commercial and auxiliary, and Category 2).
This analysis shows that, within the group of engines subject to the Tier  2 standards, the new
emission standards will yield reductions from the baseline of 12 percent HC, 27 percent NOx, and
24 percent PM from Category 1 commercial and auxiliary, and Category 2 marine diesel engines in
2020. In 2020  the Tier 2 standards will  result in a 23 percent NOx reduction below the levels
expected from the MARPOL standards for this group of engines.

    Nationally, the new emission standards will yield estimated reductions of 0.9 percent NOx and
0.1 percent PM, with greater percentage reductions expected in port areas.  This is especially true
for Los Angeles-South Coast, Baton Rouge, Beaumont-Port Arthur, and similar ports where marine
diesel engines account for a large fraction of the NOx emissions. It is also important to note that the
emission reductions in densely populated port areas may be of greater value.
                                          114

-------
                                                     Chapter 5: Environmental Impacts
                                        Table 5-14
              Total Emission Reductions from all Commercial CI Marine Engines

HC
103 short tons
NOx
103 short tons
PM
103 short tons
baseline
controlled
reduction
baseline
MARPOL
controlled
reduction*
baseline
controlled
reduction
2000
31.3
31.3
0%
1,005
1,001
1,001
0%
42.3
42.3
0%
2010
34.8
33.4
4%
1,117
1,074
1,004
10%
46.9
44.4
4%
2020
38.7
35.4
8%
1,244
1,167
987
21%
52.2
46.4
11%
2030
43.2
39.3
9%
1,390
1,292
1,056
24%
58.2
51.2
12%
* This reduction is from the baseline. The Tier 2 standards are expected to achieve a 15 percent reduction in 2020 from
the levels expected from the MARPOL standards.

                                        Table 5-15
                Emission Reductions from Engines Subject to Tier 2 Standards

HC
103 short tons
NOx
103 short tons
PM
103 short tons
baseline
controlled
reduction
baseline
MARPOL
controlled
reduction*
baseline
controlled
reduction
2000
22.4
22.4
0%
702.2
699.6
699.6
0%
20.1
20.1
0%
2010
24.7
23.3
6%
773.5
742.3
672.1
13%
22.2
19.7
11%
2020
27.3
24.0
12%
852.2
797.5
618.0
27%
24.4
18.6
24%
2030
30.1
26.2
13%
939.0
871.1
634.7
32%
27.0
20.0
26%
* This reductionis from the baseline. The Tier 2 standards are expected to achieve a23 percent reduction in 2020 from
the levels expected from the MARPOL standards.

E. Other Impacts of NOx Emission Reductions

    The NOx reductions described above are expected to provide reductions in the concentrations
of secondary nitrate particulates.  NOx can react with ammonia in the atmosphere to form
                                            115

-------
Regulatory Impact Analysis
ammonium nitrate particulates, especially when ambient sulfur levels are relatively low.  EPA
believes that the best estimate of atmospheric NOx to PM conversion rates available at this time is
provided by SAL3  SAI used a combination of ambient concentration data and computer modeling
that simulates atmospheric conditions to estimate the conversion of NOx to PM nitrate. Ambient
data was collected from 72 ozone, 64 NOx, and 14 NOMC monitoring sites for use in oxidation
calculations. Data was also collected from 45 nitrate/NOx monitoring sites for use in the equilibrium
calculations.  For the purpose of modeling, the 48 continental states were divided into nine regions,
and rural areas were distinguished from urban areas. The model was  designed to perform the
equilibrium calculation to estimate particulate nitrate formation for different regions, seasons, and
times of day and then was calibrated using ambient data.

    The results from the S AI report state that the fraction of NOx converted to nitrates (g/g) ranges
from 0.01 in  the northeast to 0.07 in southern California.  Using a simple average, the average
fraction of NOx converted to nitrates is approximately 0.04. The effects of the conversion fraction
of future PM  reductions are shown in Table 5-16.
                                        Table 5-16
                               Equivalent PM Emissions for
                      CI Marine Engines [thousand short tons per year]
Year
2005
2010
2015
2020
Total NOx Reductions
37
113
193
257
Equivalent PM Reductions
1.5
4.5
7.7
10.3
    The expected reductions in NOx emissions should  also positively affect  visibility, acid
deposition, and estuary eutrophication.  Both NO2 and nitrate particulates are optically active, and
in some urban areas, NO2 and nitrate particulates can be responsible for 20 to 40 percent of the
visible light extinction. The effect of this rulemaking on visibility should be small but potentially
significant, given that it is expected to reduce national NOx emissions by about 0.9 percent in 2020.
This results in an estimated 0.2 to 0.4 percent reduction in haze. For areas with active ports, this
reduction would be significantly greater.

    The NOx control described above is also expected to provide benefits with  respect to acid
deposition. The annual NOx reduction expected in 2020 from the new emission standards is about
65 percent of the 400,000 ton per year reduction expected from Phase I of the Agency's acid rain
NOx control rule (59 FR 13538, March 22, 1994), which was considered to be a significant step
toward controlling the ecological damage caused by acid deposition. It is not clear how reductions
from marine engines will compare to  reductions from stationary sources with elevated stacks;
                                           116

-------
                                                   Chapter 5: Environmental Impacts
however, some reduction in the adverse effects of acid depositions should occur as a result of this
rulemaking.

    The new emission standards should also lead to a reduction in the nitrogen loading of estuaries.
This is significant since high nitrogen loadings can lead to eutrophication of the estuary, which
causes disruption in the ecological balance.  The effect should be most significant in areas heavily
affected by atmospheric NOx emissions.  One such estuary is Chesapeake Bay, where as much as
40 percent of the nitrogen loading may be caused by atmospheric deposition. Also, marine engine
emissions by definition occur in the marine environment; therefore, reductions in exhaust emissions
from marine engines are expected to benefit the marine environment.

F. Air Toxics

    The term "hydrocarbons" includes many different molecules.  Many of these molecules are
considered toxic air emissions, including benzene, formaldehyde, acetaldehyde, and 1,3-butadiene.
Because EPA does not have speciated data on diesel marine engine hydrocarbon emissions, the
reductions in air toxins can only be estimated here.  This analysis was done using data on highway
heavy-duty diesel  engines.4'5'6'7 According to this data, hydrocarbons from an highway HDDE
include approximately 1.1 percent benzene, 7.8 percent formaldehyde, 2.9 percent acetaldehyde, and
0.6 percent 1,3-butadiene. Table 5-17 shows the estimated air toxics reductions associated with the
hydrocarbon reductions in this rule when the highway HDDE results are extended to this source.

                                       Table 5-17
                    Estimated Annual Air Toxics Reductions  [short tons]
Year
2005
2010
2015
2020
Benzene
4
15
28
36
Formaldehyde
26
109
196
255
Acetaldehyde
10
41
73
95
1,3 -Butadiene
2
8
15
20
                                          117

-------
Regulatory Impact Analysis
Chapter 5 References
1. Corbett, J., Fischbeck, P., "Commercial Marine Emissions Inventory and Analysis for United
States Continental and Inland Waterways," Order No. 8A-0516-NATX, August 21, 1998 (Docket
A-97-50; document II-A-1).

2. "National Air Pollutant Emission Trends, 1900-1996," U.S. EPA, EPA-454/R-97-011,
December 1997.

3. "Benefits of Mobile Source NOx Related Particulate Matter Reductions," Systems
Applications International, EPA Contract No. 68-C5-0010, WAN 1-8, October 1996.

4.  Springer, K., "Investigation of Diesel-Powered Vehicle Emissions VII," U.S. EPA, EPA-
460/3-76-034, 1977.

5.  Springer, K., "Characterization of Sulfates, Odor, Smoke, POM and Particulates from Light
and Heavy-Duty Engines - Part IX," U.S. EPA, EPA-460/3-79-007, 1979.

6.  Bass, E., and Newkirk, M., "Reactivity Comparison of Exhaust Emissions from Heavy-Duty
Engines Operating on Gasoline, Diesel, and Alternative Fuels," Southwest Research Institute,
SwRI9856, 1995.

7.  College of Engineering - Center for Environmental Research and Technology, "Evaluation of
Factors that Affect Diesel Exhaust Toxicity," submitted to California Air Resources Board,
Contract No. 94-312, 1998.
                                         118

-------
                                                       Chapter 6: Cost-Effectiveness
              CHAPTER 6:  COST-EFFECTIVENESS
    This chapter assesses the cost-effectiveness of the new hydrocarbon and oxides of nitrogen
emission standards. This analysis relies in part on cost information from Chapter 4 and emissions
information from Chapter 5 to estimate the cost-effectiveness of the new standards in terms of
dollars per short ton of total HC+NOx emission reductions.  This chapter also examines the cost-
effectiveness of the PM standards. Finally, the chapter compares the cost-effectiveness of this final
rule with the cost-effectiveness  of other NOx and  PM control strategies from previous EPA
rulemakings.

    The analysis presented in this chapter is performed for marine diesel engines and vessels using
the same nominal power ratings as presented in Chapter 4. An estimate of the industry-wide cost-
effectiveness of the new emission  standards, combining all of the nominal engine sizes, is also
presented.

    Benefits associated with the MARPOL provisions are not included in this analysis.  Marine
diesel engines greater than or equal to 13 0 kW will be subj ect to an international NOx standard prior
to the implementation of EPA's new emission standards. Therefore, the baseline emissions case
assumes emission control to the MARPOL standard for engines installed on vessels in 2000 and
later.   Because the  new EPA standards  apply only to Category  1  and Category 2, no cost-
effectiveness analysis is provided for Category 3.

    Two types of cost-benefit analyses are performed in this chapter. The first analysis focuses on
individual engines and examines  total costs and total emission reductions over the typical lifetime
of an average marine diesel  engine discounted to the beginning of the engine's life. The second
method looks at the net present value (NPV) of a stream of costs and benefits over a standardized
period of time (30 years).  Over  this period, the calculation includes the whole set of new
requirements.

    As described in Chapter 4, several  of the  anticipated engine technologies will  result in
improvements in engine performance that go beyond emission control. While the cost estimates
described in Chapter 4 do not take into account the observed value of performance improvements,
these non-emission benefits should be taken into account in the calculation of cost-effectiveness.

    As shown in Table 4-2, some engines are using turbocharging and aftercooling technologies
currently.  Turbocharging and  aftercooling provide important advantages  in improving fuel
consumption and power density. This shows that the turbocharging and aftercooling improvements
for these engines provide benefits independent of emission control that justify their additional cost.
Smaller engines have generally not adopted these technologies in the absence of emission standards.
We expect the same non-emission benefits would apply to these engines when they adopt the
anticipated technologies, since turbocharging and aftercooling have similar effects on engines of
varying sizes. EPA understands, however, that market forces alone are not sufficient to drive these

                                         119

-------
Regulatory Impact Analysis
changes. As a result, manufacturers that adopt these technologies to comply with emission standards
incur a cost that should be attributed to the standards.  As the larger engines demonstrate, it would
likely be appropriate to assign very close to 100 percent of the cost of these changes to non-emission
benefits.

    A similar assessment applies to fuel injection improvements.  Manufacturers upgrading to
higher-pressure inj ection with better control of inj ection parameters have the ability to improve both
emissions and engine performance generally.  The experience with highway engines shows, for
example, that manufacturers have improved fuel consumption and extended total engine lifetimes
during a period of tightening emission standards.  While it is more difficult to quantify the non-
emission benefits associated with fuel inj ection improvements, it is clear that these benefits are real.
To avoid underestimating the cost impacts of the rulemaking, EPA believes it is most appropriate
to assign equal weighting to emission and non-emission  benefits for these technologies.  The
analysis therefore assesses 50 percent of the total costs of these technologies as an impact of the
emission standards.

    For some or all of these technologies, a greater value for the non-emission benefits could likely
be justified.  This has the effect of halving the cost for those technologies in the cost-effectiveness
calculation. The cost-effectiveness values in this chapter are based on this calculation methodology.
However, all costs shown in this chapter (with the exception of the "Accounting for Non-emission
Benefits" columns in Table 6-2) are the entire costs of the technology, not just the portions attributed
to emission reduction.

I.  Engine Lifetime Cost-Effectiveness of  the New Standards

A. HC+NOx

    The cost-effectiveness of the new HC+NOx standards was calculated for the various nominal
marine engine power ratings described in Chapter 4. For this analysis, the entire cost of the program
is attributed to the control of HC and NOx emissions. As discussed in Chapter 4, the estimated cost
of complying  with the new emission  standards varies depending  on the model year under
consideration (i.e. year 1  versus year 6).  Therefore, this analysis includes the per-engine cost-
effectiveness results for the different model years during which the costs are expected to change.
This analysis focuses on costs and benefits for individual engine types; therefore, the costs presented
in this section represent the actual cost-effectiveness as it affects a given engine. All of the costs and
benefits are discounted at seven percent to the model year of the marine engine.

    1. Tier 2

    EPA calculated the costs and benefits achieved from the Tier 2 standards beyond the MARPOL
requirements. Tier 2 costs are compared to baseline technology. EPA conservatively did not make
an attempt to estimate costs for any development work or new technology that could be attributed
to the MARPOL requirements and, therefore, removed from this analysis. Table 6-1 presents the
Tier 2 discounted cost-effectiveness for marine diesel engines at the five nominal marine engine

                                          120

-------
                                                       Chapter 6: Cost-Effectiveness
power ratings.  Table 6-1 also provides a side-by-side comparison of cost-effectiveness values with
and without accounting for the non-emission benefits.  Cost-effectiveness figures in other tables in
this chapter are presented similarly.

                                       Table 6-1
         Discounted Cost-Effectiveness ($/short ton) of the Tier 2 HC+NOx Standards
Nominal
Power
(kW)
100
400
750
1500
3000
Model
Year
Grouping
Ito5
6 +
Ito5
6 +
Ito5
6 +
Ito5
6 +
Ito5
6 +
NPV
Benefits
(short tons)
4.3
26
80
267
750
Emission Benefits Only
Operating &
Compliance
NPV Costs
$442
$704
$206
$636
$12,430
Engine &
Vessel
Costs
$1,806
$486
$3,208
$846
$25,395
$856
$22,818
$1,120
$54,192
$13,019
Discounted
Cost-
Effectiveness
$521
$215
$151
$60
$319
$13
$88
$7
$89
$34
Accounting for Non-Emission Benefits
Operating &
Compliance
NPV Costs
$221
$352
$103
$318
$6,215
Engine &
Vessel
Costs
$1,175
$272
$1,920
$458
$22,562
$772
$20,039
$860
$41,494
$7,214
Discounted
Cost-
Effectiveness
$470
$164
$137
$46
$318
$12
$87
$5
$81
$26
    Category 1, Category 2, and aggregate marine diesel engine cost-effectiveness figures for the
Tier 2 HC+NOx standards were also calculated. To accomplish this, each nominal power rating was
assumed to represent a range of engine power ratings. Each of the ratings was then weighted using
1997 marine engine populations. Table 6-2 shows how the nominal power ratings were applied to
the population as a whole.
                                          121

-------
Regulatory Impact Analysis
                                       Table 6-2
                 1997 Marine Diesel Engine Populations by Power Category
Category
Category 1
Category 2
Nominal Power (kW)
100
400
750
1500
3000
Power Range (kW)
37 - 225
225 - 560
560 - 1000
1000 +
all
Population
42,690
23,691
2,414
2,596
1,569
    Table 6-3 presents the Tier 2 aggregate cost-effectiveness for Category 1 and 2 both individually
and combined. Population weighted costs and were divided by the population weighted discounted
benefits to determine the aggregate cost-effectiveness figures.  These figures suggest that the Tier
2 standards for HC+NOx from marine diesel engines are very cost-effective. This result is largely
due to the high hours of operation performed annually by these engines and their long useful lives.

                                        Table 6-3
                     Discounted Cost-Effectiveness ($/shortton) of the
              Tier 2 HC+NOx Standards for Category 1 and Category 2 Engines
Engine
Class
Category 1
Category 2
Aggregate
Model
Year
Grouping
Ito5
6 +
Ito5
6 +
Ito5
6 +
NPV
Benefits
(short tons)
24
750
39
Emission Benefits Only
Operating &
Compliance
NPV Costs
$528
$12,430
$784
Engine &
Vessel
Costs
$3,833
$641
$54,192
$13,019
$5,979
$1,163
Discounted
Cost-
Effectiveness
$185
$50
$89
$34
$172
$50
Accounting for Non-Emission Benefits
Operating &
Compliance
NPV Costs
$264
$6,215
$392
Engine &
Vessel
Costs
$2,831
$372
$41,494
$7,214
$3,663
$519
Discounted
Cost-
Effectiveness
$131
$27
$64
$18
$103
$23
B. PM

    EPA has also estimated the cost-effectiveness of the Tier 2 PM emission standards for each of
the nominal power ratings.  The per-engine PM emission reduction estimates were developed in
Chapter 5.  For costs, EPA assumed that half of the increased engine and vessel costs projected in
                                          122

-------
                                                        Chapter 6: Cost-Effectiveness
Chapter 4 were allocated for PM control. Because the entire cost of the program was included in the
HC+NOx analysis,  this analysis for PM cost-effectiveness is only a supplement to the cost-
effectiveness calculations.  EPA believes that this analysis is conservative, given the stringency of
the HC+NOx standards and results in an upper end estimate of the cost-effectiveness of the new PM
standards. In addition, it should be noted that the HC+NOx cost-effectiveness calculations discussed
above include the entire cost of the engine modifications that EPA believes will result from this final
rule. Table 6-4 contains the resulting cost-effectiveness for each of the nominal power ratings.

                                        Table 6-4
            Discounted Cost-Effectiveness ($/short ton) of the New PM Standards
Nominal
Power
(kW)
100
400
750
1500
3000
Model
Year
Grouping
Ito5
6 +
Ito5
6 +
Ito5
6 +
Ito5
6 +
Ito5
6 +
NPV
Benefits
(short tons)
0.13
0.93
2.87
6.20
6.58
Emission Benefits Only
Operating &
Compliance
NPV Costs
$221
$352
$103
$318
$6,215
Engine &
Vessel
Costs
$903
$243
$1,604
$423
$12,698
$428
$11,409
$560
$27,096
$6,510
Discounted
Cost-
Effectiveness
$8,600
$3,550
$2,112
$837
$4,465
$185
$1,892
$142
$5,062
$1,934
Accounting for Non-Emission Benefits
Operating &
Compliance
NPV Costs
$111
$176
$52
$159
$3,108
Engine &
Vessel
Costs
$588
$136
$960
$229
$11,281
$386
$10,020
$430
$20,747
$3,607
Discounted
Cost-
Effectiveness
$5,341
$1,886
$1,227
$437
$3,953
$153
$1,642
$95
$3,625
$1,020
    Table 6-5 presents the aggregate of these numbers for Category 1 and 2 both separately and
together. This aggregation of the cost-effectiveness figures was performed in the same way as for
the HC+NOx cost-effectiveness analysis. The results of this analysis show low cost-effectiveness
numbers for PM control. These  costs would probably also be considered low even if all of the
engine and vessel costs associated with this rulemaking were applied to the PM cost-effectiveness
analysis.
                                           123

-------
Regulatory Impact Analysis
                                        Table 6-5
                        Discounted Cost-Effectiveness ($/short ton)
                     of the PM Standards for Category 1 and 2 Engines
Engine
Class
Category 1
Category 2
Aggregate
Model
Year
Grouping
Ito5
6 +
Ito5
6 +
Ito5
6 +
NPV
Benefits
(short tons)
0.77
6.20
0.89
Emission Benefits Only
Operating &
Compliance
NPV Costs
$264
$6,215
$392
Engine &
Vessel
Costs
$1,916
$321
$27,096
$6,510
$2,989
$581
Discounted
Cost-
Effectiveness
$2,848
$763
$5,374
$2,053
$3,797
$1,093
Accounting for Non-Emission Benefits
Operating &
Compliance
NPV Costs
$132
$3,108
$196
Engine &
Vessel
Costs
$1,416
$186
$20,747
$3,607
$1,831
$260
Discounted
Cost-
Effectiveness
$2,021
$415
$3,849
$1,083
$2,276
$511
C. Comparison with Cost-Effectiveness of Other Control Programs

    In an effort to evaluate  the cost-effectiveness  of the new emission standards, EPA  has
summarized the cost-effectiveness results for four other recent EPA mobile source rulemakings that
required reductions in NOx emissions. Where HC+NOx cost-effectiveness was not reported, NOx
cost-effectiveness-figures are reported. NOx cost-effectiveness figures should be close to HC+NOx
cost-effectiveness figures since NOx is the primary focus of the new standards and because NOx
emissions are generally much higher than HC emissions for diesel engines. Table 6-6 summarizes
the cost-effectiveness results from the heavy-duty vehicle portion of the Clean Fuel Fleet Vehicle
Program, the most recent HC+NOx engine standards for highway heavy-duty diesel engines, and the
nonroad Tier 2 standards.

    A comparison of the cost-effectiveness numbers in Table 6-6 with the cost-effectiveness results
presented throughout this chapter for marine diesel engines shows that the cost-effectiveness of the
Tier 2 HC+NOx standards are more favorable than the cost-effectiveness results of any of these
recent rules.  To be consistent with the cost-effectiveness values for other programs, the marine
diesel numbers shown in Tables 6-6  and 6-7 reflect the aggregate cost-effectiveness for the program
rather than the high and low for individual engines.
                                          124

-------
                                                       Chapter 6: Cost-Effectiveness
                                       Table 6-6
           Summary of Cost-Effectiveness for Recent EPA NOx Control Programs
EPA Rule
Clean Fuel Fleet Vehicle Program
(Heavy-duty)
2.5 g/hp-hrNMHC*+NOx Standard for
Highway Heavy-Duty Engines
Locomotive Engine Standards
Nonroad Tier 2 Standards
Pollutants Considered
in Calculations
NOx
NMHC*+NOx
NOx
NMHC*+NOx
Cost-Effectiveness
($/ton)
$1,300 -$1,500
$100 - $600
$160 - $250
$480 - $540
 : nonmethane hydrocarbons (roughly equivalent to total HC for diesel engines)
    For comparison purposes, EPA has also summarized the cost-effectiveness results for three
other recent EPA mobile source rulemakings that required reductions in PM emissions. Table 6-7
summarizes the cost-effectiveness results for the most recent urban bus engine PM standard and the
urban bus retrofit/rebuild program. The PM cost-effectiveness results presented earlier in Table 6-5
are more favorable than either of the urban bus programs and comparable to the nonroad Tier 2
standards.

                                       Table 6-7
            Summary of Cost-Effectiveness for Recent EPA PM Control Programs
EPA Rule
0.05 g/hp-hr Urban Bus PM Standard
Urban Bus Retrofit/Rebuild Program
Nonroad Tier 2 Standards
Cost-Effectiveness ($/ton)
$10,000 - $16,000
$25,500
$700 - $2,320
II. 30-Year Cost-Effectiveness of the New Standards

    Another tool that can be used to evaluate the cost-effectiveness of a regulatory program is to
look at the costs incurred and the emissions benefits achieved over a fixed period of time.  The
standard period of time associated with this type of rulemaking is 30 years. In this analysis, the net
present value (NPV) of the costs incurred from 2000 to 2030 is divided by the NPV of the benefits
                                          125

-------
Regulatory Impact Analysis
achieved over this same time period. NPV is calculated using a seven percent time value of money
and a seven percent time value of emission reductions. The streams of costs presented here are based
on the cost figures developed in Chapter 4. Non-emission benefits are not accounted for here.

A. HC+NOx

     Table 6-8 presents the stream of costs and benefits associated with the new HC+NOx standards.
The discounted cost-effectiveness of the HC+NOx standards over a 30-year period is $106/ton for
Category 1, $46/ton for Category 2, and $92/ton for the aggregate. It should be noted that these
figures are a little higher than those presented above for year 6 on a per-engine basis.  This difference
is because the per-engine analysis relates  costs to their resulting benefits while the  stream of costs
analysis compares costs incurred to benefits achieved in a fixed time frame. In other words, many
of the costs incurred prior to 2030 will not achieve benefits until after 2030. This is, in part, due to
the long lives and slow turnover of marine diesel engines.

B. PM

     Table 6-9 presents the stream of costs and benefits associated with the new PM standards. As
with the per-engine analysis, EPA conservatively applied half of the incurred costs to PM control.
The discounted cost-effectiveness of the new PM standards over a 30-year period is $l,272/ton for
Category 1. No PM benefits are claimed for Category 2. The same relationship exists for these PM
cost-effectiveness figures compared to the per-engine PM cost-effectiveness figures as does for the
HC+NOx analysis.
                                          126

-------
                                      Chapter 6: Cost-Effectiveness
                       Table 6-8
30-Year Stream of Costs and Benefits for the HC+NOx Standards
Calendar
Year
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
NPV
Category 1
Costs (106 $)
$10.1
$10.2
$23.2
$26.2
$17.7
$9.1
$9.2
$6.6
$4.0
$4.0
$4.1
$4.1
$4.1
$4.2
$4.2
$4.3
$4.3
$4.4
$4.4
$4.4
$4.5
$4.5
$4.6
$4.6
$4.7
$4.7
$2.4
$2.4
$2.4
$2.5
$2.5
$112.2
Benefits (103 tons)
7.0
14.3
21.7
31.3
41.0
50.7
60.5
70.2
79.9
89.6
99.1
108.5
116.7
124.2
131.0
136.3
141.0
145.1
148.7
151.8
154.7
157.3
159.7
162.0
164.1
166.2
168.1
169.8
171.5
173.2
175.0
1063.1
Category 2
Costs (106 $)
$0
$0
$0
$2.9
$2.9
$3.0
$3.0
$3.0
$0.5
$0.5
$0.5
$0.6
$0.6
$0.6
$0.6
$0.6
$0.6
$0.6
$0.6
$0.6
$0.6
$0.6
$0.6
$0.6
$0.6
$0.6
$0.7
$0.7
$0.7
$0.7
$0.7
$13.8
Benefits (103 tons)
0
0
0
2.7
5.5
8.3
11.1
14.0
17.0
20.0
23.0
26.0
29.1
32.2
35.4
38.5
41.7
44.9
48.1
51.4
54.6
57.8
61.1
64.3
67.1
69.7
72.1
74.4
76.6
78.3
80.0
301.8
                          127

-------
Regulatory Impact Analysis
                                    Table 6-9
                30-Year Stream of Costs and Benefits for the PM Standards
Calendar
Year
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
NPV
Category 1
Costs (106 $)
$5.0
$5.1
$11.6
$13.1
$8.9
$4.5
$4.6
$3.3
$2.0
$2.0
$2.0
$2.1
$2.1
$2.1
$2.1
$2.1
$2.2
$2.2
$2.2
$2.2
$2.2
$2.3
$2.3
$2.3
$2.3
$2.4
$1.2
$1.2
$1.2
$1.2
$1.2
$56.1
Benefits (103 tons)
0.3
0.6
0.9
1.3
1.7
2.1
2.5
2.9
3.3
3.7
4.1
4.5
4.8
5.1
5.4
5.6
5.8
6.0
6.2
6.3
6.4
6.5
6.6
6.7
6.8
6.9
7.0
7.1
7.1
7.2
7.3
44.1
                                       128

-------