United States        Air and Radiation       EPA420-D-02-003
          Environmental Protection                 July 2002
          Agency
&EPA    Draft Regulatory Support
          Document: Control of
          Emissions from
          Spark-Ignition Marine
          Vessels and Highway
          Motorcycles
                                 > Printed on Recycled Paper

-------
                                        EPA420-D-02-003
                                             July 2002
   Draft Regulatory Support Document:
Control of Emissions from Spark-Ignition
Marine Vessels and Highway Motorcycles
           Assessment and Standards Division
          Office of Transportation and Air Quality
          U.S. Environmental Protection Agency

-------
                                 Table of Contents
Overview

CHAPTER 1:  Health and Welfare Concerns
       1.1 - Ozone	1-1
              1.1.1 - General Background	1-1
              1.1.2 - Health and Welfare Effects of Ozone and Its Precursors	1-2
              1.1.3 - Additional Health and Welfare Effects of NOx Emissions  	1-3
              1.1.4 - Ozone Nonattainment	1-4
              1.1.5 - Public Health and Welfare Concerns from Prolonged and Repeated
                    Exposures to Ozone	1-7
       1.2 - Particulate Matter  	1-8
              1.2.1 - General Background	1-8
              1.2.2 - Health and Welfare Effects of PM	1-10
              1.2.3 - PM Nonattainment	1-11
       1.3 - Gaseous Air Toxics	1-13
              1.3.1 - Benzene  	1-13
              1.3.2 - 1,3-Butadiene	1-14
              1.3.3 - Formaldehyde	1-15
              1.3.4 - Acetaldehyde  	1-16
              1.3.5 - Acrolein  	1-17
       1.4 - Inventory Contributions	1-18
              1.4.1 - Inventory Contribution	1-18
              1.4.2 - Inventory Impacts on a Per Vehicle Basis 	1-20
       1.5 - Other Health and Environmental Effects	1-21
              1.5.1 - Carbon Monoxide	1-21
              1.5.2 - Acid Deposition  	1-22
              1.5.3 - Eutrophication and Nitrification	1-23

CHAPTER 2:  Industry Characterization
       2.1- Highway Motorcycles  	2-1
              2.1.1 - Manufacturers	2-1
              2.1.2 - Sales and Fleet Size 	2-2
              2.1.3 -Usage  	2-3
              2.1.4 - Current Trends  	2-4
              2.1.5 - Customer Concerns	2-5
                    2.1.5.1 -Performance	2-5
                    2.1.5.2-Cost 	2-5
                    2.1.5.3 - Consumer Modifications	2-6
                    2.1.5.4 - Safety	2-6
       2.2 - Marine	2-7
              2.2.1 - Gasoline Engine Manufacturers	2-7
                    2.2.1.1 - Identification of Gasoline Engine Manufacturers  	2-7

-------
                     2.2.1.2 - Use of Gasoline Engines	2-8
                     2.2.1.3 - Current Trends	2-8
              2.2.2 - Recreational Boat Builders	2-9
                     2.2.2.1 - Identification of Boat Builders  	2-9
                     2.2.2.2 - Current Trends	2-9
                     2.2.2.3 - Production Practices 	2-9
              2.2.3 - Fuel Tank Manufacturers	2-10
                     2.2.3.1 - Identification of Fuel Tank Manufacturers  	2-10
                     2.2.3.2 - Current Trends	2-11
                     2.2.3.3 - Production Practices 	2-11
              2.2.4 - Hose Manufacturers 	2-11
                     2.2.4.1 - Identification of Hose Manufacturers 	2-11
                     2.2.4.2 - Current Trends	2-11
                     2.2.4.3 - Production Practices 	2-12

CHAPTERS:  Technology
       3.1 - Introduction to Spark-Ignition Engine Technology	3-1
              3.1.1 -Basics of Spark-Ignition Four-Stroke Engines	3-1
              3.1.2 - Engine Calibration  	3-2
                     3.1.2.1 - Air-fuel ratio  	3-2
                     3.1.2.2 - Spark-timing:   	3-3
                     3.1.2.3 - Fuel Metering	3-3
              3.1.3 - Gaseous Fuels	3-4
       3.2 - Exhaust Emissions and Control Technologies 	3-5
              3.2.1 - Combustion chamber design   	3-5
              3.2.2 - Exhaust gas recirculation 	3-5
              3.2.3 - Secondary air  	3-6
              3.2.4 - Catalytic Aftertreatment	3-6
                     3.2.4.1 - System cost  	3-7
                     3.2.4.2 - Packaging constraints  	3-7
              3.2.5 - Multiple valves and variable valve timing	3-7
              3.2.6 - Advanced Emission Controls	3-8
       3.3 - Evaporative Emissions  	3-10
              3.3.1 - Sources of Evaporative Emissions	3-10
                     3.3.1.1 -Diurnal and Running Loss Emissions	3-11
                     3.3.1.2-Hot Soak Emissions	3-12
                     3.3.1.3 - Refueling Emissions 	3-12
                     3.3.1.4 -Permeation	3-12
              3.3.2 - Evaporative Emission Controls  	3-13
                     3.3.2.1 - Sealed System with Pressure Relief 	3-13
                     3.3.2.2 - Limited Flow Orifice	3-14
                     3.3.2.3 - Insulated Fuel Tank	3-14
                     3.3.2.4 - Volume Compensating Air Bag	3-15
                     3.3.2.5 - Collapsible Bladder Fuel  Tank  	3-15
                     3.3.2.6 - Charcoal  Canister  	3-16

-------
                    3.3.2.7 - Floating Fuel and Vapor Separator	3-16
                    3.3.2.8 - Low-permeability Materials  	3-16

CHAPTER 4: Technological Feasibility
       4.1- Highway Motorcycles  	4-1
             4.1.1 - Class I and II Motorcycles 	4-3
                    4.1.1.1 - Class I Motorcycles Above 50cc and Class II Motorcycles . . . 4-3
                    4.1.1.2 - Class I Motorcycles Under 50cc 	4-3
             4.1.2 - Class HI Motorcycles  	4-5
                    4.1.2.1 - Tier-1 Standards 	4-5
                    4.1.2.2 - Analysis of EPA Certification Data	4-8
                    4.1.2.3 - Tier-2 Standards 	4-9
             4.1.3 - Impacts on Noise, Energy, and Safety  	4-13
             4.1.4 - Conclusion 	4-14
                    4.1.4.1 - Tier-1 Standards 	4-14
                    4.1.4.2 - Tier-2 Standards 	4-14
       4.2 - Evaporative Emission Control from Boats 	4-15
             4.2.1 - Diurnal  Evaporative Emissions  	4-15
                    4.2.1.1 - Baseline Emissions  	4-16
                    4.2.1.2 - Diffusion Effect	4-18
                    4.2.1.3 - Sealed System with Pressure Relief 	4-19
                    4.2.1.4 - Insulated Fuel Tank	4-21
                    4.2.1.5 - Volume Compensating Air Bag	4-24
                    4.2.1.6 - Bladder Fuel Tank	4-24
                    4.2.1.7 - Fuel and Vapor Separator  	4-25
             4.2.2 - Permeation Evaporative Emissions  	4-25
                    4.2.2.1 - Fuel Tanks	4-25
                    4.2.2.2 - Hoses	4-30
             4.2.3 - Evaporative Emission Test Procedures  	4-31
                    4.2.3.1 - Diurnal Emission Testing  	4-31
                    4.2.3.2 - Fuel Tank Permeation Testing	4-32
                    4.2.3.3 - Hose Permeation Testing	4-33
       4.3 - Sterndrive/Inboard Marine	4-34
             4.3.1 - Exhaust Emission Data from SD/I Engines 	4-35
                    4.3.1.1 - Baseline Emission Data	4-35
                    4.3.1.2- Emission Data For Catalyst Development Test Program  . . . 4-36
                    4.3.1.3 - Emission Data Using Exhaust Gas Recirculation	4-38
             4.3.2 - Open Issues for Using Catalysts in Marine Applications	4-39
                    4.3.2.1 - Packaging	4-39
                    4.3.2.2 - Durability	4-40
                    4.3.2.3 - Water Reversion 	4-40
                    4.3.2.4 - Safety	4-42

CHAPTER 5: Estimated Costs
       5.1- Methodology 	5-1

-------
       5.2 - Cost of Emission Controls by Engine/Vehicle Type	5-2
             5.2.1 - Evaporative Emission Control from Boats	5-2
                    5.2.1.1 - Technologies and Estimated Costs  	5-2
                    5.2.1.2 - Operating Cost Savings	5-3
                    5.2.1.3 - System Integration and Compliance Costs 	5-4
                    5.2.1.4 - Total Costs	5-4
             5.2.2 - Highway Motorcycles	5-5
                    5.2.2.1 - Technologies and Estimated Costs  	5-6
                    5.2.2.2 - Compliance Costs  	5-10
                    5.2.2.3 - Highway Motorcycle Total Costs 	5-11
                    5.2.2.4 - Highway Motorcycles  Under 50 cc	5-14
                    5.2.2.5 - Highway Motorcycle Aggregate Costs  	5-15
       5.3 - Aggregate Costs	5-16

CHAPTER 6: Emissions Inventory
       6.1- General Methodology  	6-1
             6.1.1 - Highway Motorcycle Emissions	6-1
             6.1.2 - Marine Evaporative Emissions	6-1
                    6.1.2.1 - Diurnal Emissions	6-2
                    6.1.2.2 - Refueling Emissions  	6-3
                    6.1.2.3 - Permeation Emissions	6-4
                    6.1.2.3 - Hot Soak Emissions	6-4
             6.1.3 - SD/I Exhaust Emissions	6-4
       6.2 - Effect of Emission Controls by Engine/Vehicle Type	6-6
             6.2.1 - On-highway Motorcycles	6-6
                    6.2.1.1 - Inputs for the Inventory Calculations  	6-6
                    6.2.1.2 - Reductions Due to the Proposed Standards	6-11
                    6.2.1.3 - Per Equipment Emissions from On-highway Motorcycles  .. 6-13
             6.2.2 - Evaporative Emission Control from Boats	6-14
                    6.2.2.1 - Inputs for the Inventory Calculations  	6-15
                    6.2.2.2 - Reductions Due to the Proposed Standard 	6-19
                    6.2.2.3 - Per Boat Evaporative Emissions	6-20
             6.2.3 - Sterndrive and Inboard Marine	6-21
                    6.2.3.1 - Inputs for the Inventory Calculations  	6-21
                    6.2.3.2 - Baseline Emissions from SD/I Marine Engines  	6-22
                    6.2.3.3 - Analysis of Catalyst-based Approach  	6-22
                    6.2.3.4 - Analysis of EGR-based Approach	6-24

CHAPTER 7 Cost Per Ton
       7.1- Cost Per Ton by Engine Type 	7-1
             7.1.1 - Introduction	7-1
             7.1.2 - Evaporative Emission Control from Boats	7-1
             7.1.3 - On-Highway Motorcycles	7-2
       7.2 - Cost Per Ton for Other Mobile Source Control Programs  	7-4
       7.3 - 20-Year Cost and Benefit Analysis	7-5

-------
CHAPTER 8:  Initial Regulatory Flexibility Analysis
       8.1 - Requirements of the Regulatory Flexibility Act 	8-13
       8.2 - Description of Affected Entities  	8-14
              8.2.1 - Recreational Vehicles (off-highway motorcycles, ATVs, and snowmobiles)
                      	8-14
              8.2.2 - Highway Motorcycles	8-15
              8.2.3 - Marine Vessels	8-15
                    8.2.3.1 - Small Recreational Boat Builders	8-16
                    8.2.3.2 - Small Marine Fuel Tank Manufacturers 	8-16
                    8.2.3.3 - Small Diesel Engine Marinizers 	8-16
                    8.2.3.4 - Small Gasoline Engine Marinizers  	8-17
              8.2.4 - Large Spark Ignition Engines	8-17
       8.3 - Projected Costs of the Proposed Program	8-17
       8.4 - Projected Reporting,  Recordkeeping, and Other Compliance Requirements of the
              Proposed Rule  	8-17
       8.5 - Other Related Federal Rules 	8-17
       8.6 - Regulatory Alternatives	8-18

-------
                                     Overview
       EPA is proposing new standards for emissions of evaporative hydrocarbons from marine
vessels and exhaust hydrocarbons and oxides of nitrogen from highway motorcycles.  This
proposal is the second part of an action for several sources.  On October 5, 2001, we published
proposed standards for large spark-ignition engines, recreational vehicles, and recreational
marine diesel engines (66 FR 51098). That proposal had its own Draft Regulatory Support
Document.

       This Draft Regulatory Support Document provides economic, technical, cost, and
environmental analyses of the proposed emission standards for marine vessels and highway
motorcycles.  The anticipated emission reductions would translate into  significant, long-term
improvements in air quality in many areas of the U.S. Overall, the proposed requirements would
dramatically reduce individual exposure to dangerous pollutants and 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.

       This proposal also discusses potential future emission control of exhaust emission from
sterndrive and inboard marine engines.  This Draft Regulatory Support Document also provides
technical and environmental analyses of potential control strategies for these engines.

       Chapter 1 reviews information related to the health and welfare effects of the pollutants
of concern. Chapter 2 contains an overview of the affected manufacturers, including some
description of the range of engines involved and their place in the market.  Chapter 3 covers a
broad description of engine and evaporative technologies, including a wide variety of approaches
to reducing emissions. Chapter 4 summarizes the available information supporting the specific
standards we are proposing, providing a technical justification for the feasibility of the standards.
Chapter 5 applies cost estimates to the projected technologies. Chapter 6 presents the calculated
contribution of these  sources to the nationwide emission inventory with and without the proposed
standards. Chapter 7 compares the costs and the emission reductions for an estimate of the cost-
effectiveness of the rulemaking.

Market Overview

       This proposed regulation is designed to achieve emission reductions from marine vessels
and highway motorcycles.  Even though there are tangible and intangible benefits associated with
reducing emissions from these sources, significant control has not been brought about by market
forces. From an overall  perspective these  are a relatively small portion of the overall inventory
for HC and NOx.  This document will show that technology exists to achieve these significant
reductions from these sources and demonstrates that this control would be inexpensive on a cost
per ton basis. However, other externalities in the marketplace have deterred this shift to  less

-------
polluting technology. Presented below is a brief summary of the key factors pertaining to this
analysis and reference to more substantive discussions in later chapters when available.

       Emissions Overview

       For marine vessels, there are three primary sources of evaporative emissions. The first
source is diurnal emissions which are caused by fuel vapors escaping from the fuel tank.  The
marine industry currently designs their fuel systems with open vents through hoses that exit the
vessel. The second source of evaporative emissions from marine vessels is permeation through
the walls of the plastic fuel tank. Until the early 1970s, the vast majority of fuel tanks were made
out of metal which does no permeate; now less than a quarter of new fuel tanks are made of
metal. The third source of evaporative emissions from marine  vessels is permeation through the
fuel lines. In this case, there were some improvements to the permeation rate of hoses in the
1980s, but the industry standard for the permeation rate of hoses is still several orders of
magnitude higher than typical automotive fuel lines.

       Exhaust emissions from motorcycle have been the subject of a Federal emission control
program for almost twenty-five years. However, it has been over two decades since EPA last
reviewed these requirements. Technology has progressed over these years and indeed the nature
of the business and market forces are substantially different. The technology used and available
for most highway motorcycles now is far advanced beyond that called for by the 1980 era Federal
standards. Thus, there is a need for an update. Also, today, highway motorcycles are
predominantly an international commodity and importing and  exporting of product is the norm.
Thus, harmonization of emission standards and control requirements is a key need for industry
with the added benefit of lower consumer cost.

Alternatives

       In developing the proposed marine regulation, we looked at several alternatives for
potential emission control strategies and programs. In considering alternatives, we investigated a
wide range of technologies and considered various test procedures and implementation dates. In
addition, we are considering an emission credit program, including early banking, designed to
give manufacturers flexibility in what technology they use to comply with the proposed
standards.

       For diurnal emission control, we are proposing that manufacturers be able to certify to the
proposed standards by design, if they elect to do so, by implementing designs consistent with
available data.   In the preamble we specify several emission control designs and the certification
levels we would assign to these designs.  The technologies we specify include insulation with a
limited flow orifice in the vent, several control levels of sealed systems with pressure relief
valves, a sealed system with a volume compensating air bag and pressure relief, and a bladder
fuel tank.  We are also proposing an averaging, banking, and trading program in conjunction on
this design-based certification for diurnal emission control.

       For permeation emission control of fuel tanks, we identify a number of technologies that

-------
could be used to meet the proposed standards. These technologies include surface treatments
such as fluorination or sulfonation, low permeability barrier materials, and construction using
low permeability materials. For permeation emission control through hoses, we identify a
number of low permeability materials that could be used as either barriers or as construction
materials for the fuel lines.

       With regard to highway motorcycles, the alternatives focus primarily on meeting statutory
requirements while at the same time tailoring the program to the way motorcycles are produced,
sold, and used.  The program alternatives evaluation focused heavily on identifying options that
would lead to the largest emission reductions available at the lowest cost taking into account
these factors. Consultations with industry made it clear that harmonization with the California
program was critical and that international harmonization was also valuable when possible.
From the engineering and users perspective, optimal design meant establishing a program which
permitted manufacturers and users technology choices to be applied within the program. Thu,
the alternatives considered by EPA looked at harmonization, lead time, and emission credit
averaging programs. For under 50cc motorcycles international harmonization and adequate lead
time were key as these drove cost and emission control technology considerations. For over 50
cc and larger motorcycles, California harmonization and emission credit averaging had a first
order impact on cost and consumer choice. The actual program and control technology options
are further in later chapters of this document.
Proposed Emission Standards

Marine vessels powered by spark-ignition engines

       We proposing evaporative emission standards that would apply to any marine vessel
powered by any spark-ignition marine engine. This includes yachts, sport boats, fishing boats, jet
boats, and other types of pleasure craft.  This specifically includes personal watercraft and boats
with outboard engines as well as portable tanks used in any marine vessel.  These evaporative
emissions include diurnal breathing losses and permeation.  The proposed standards are
presented in Table 1. These standards represent more than a 80 percent reduction in total
evaporative emissions from new boats.

                       Table in.C-1: Proposed Evaporative Standards
Evaporative Emission
Component
Diurnal Venting
Fuel Tank Permeation
Hose Permeation
Proposed Emission Standard
1 . 1 g/gallon/day
0.08 g/gallon/day
5 g/m2/day
(15 g/m2/day with 15% methanol blend)
Test Temperature
22.2-3 5. 6°C (72-96°F)
40°C (104°F)
23°C (73°F)

-------
Highway motorcycles
       In addition, we are proposing new exhaust emission standards for highway motorcycles.
Motorcycles come in a variety of two- and three-wheeled configurations and styles, but for the
most part they are two-wheeled, self-powered vehicles.  Federal regulations currently define a
motorcycle as "any motor vehicle with a headlight, taillight, and stoplight and having: two
wheels, or three wheels and a curb mass less than or equal to 793 kilograms (1749 pounds)" (see
40 CFR 86.402-98). Note that if any motorcycle or motorcycle-like vehicle that can't be defined
as a motor vehicle (for example, if its top speed is less than 25 miles per hour), it would fall
under requirements that apply to nonroad recreational vehicles. Highway motorcycles include a
category referred to as "dual use" or "dual-sport," meaning that their designs incorporate features
that allow riders to use them both for street and off-highway application. Highway motorcycles
are operated on public roadways predominantly during warmer weather and often in or near
urban areas where they can contribute to ozone formation and ambient CO and PM levels.
Table 2 shows the proposed standards for highway motorcycles.

                                         Table 2
                 Proposed Highway Motorcycle Exhaust Emission Standards
Class
Class I
Class II
Class III
Engine Size (cc)
0-169
170-279
280 and above
Implementation
Date
2006
2006
2006
2010
HC
(g/km)
1.0
1.0
—
—
HC+NOx
(g/km)
—
—
1.4
0.8
CO (g/km)
12.0
12.0
12.0
12.0
We are also seeking comment on plastic tank and fuel hose permeation emission control
requirements for highway motorcycles, similar to that mentioned above for marine vessels.

Projected Impacts

       The following paragraphs and tables summarize the projected emission reductions and
costs associated with the proposed emission standards.  See the detailed analysis later in this
document for further discussion of these estimates.

       Table 3 contains the projected emissions from the engines subject to this proposal.
Projected figures compare the estimated emission levels with and without the proposed emission
standards for 2020.

-------
                                          Table 3
                  2020 Projected Emissions Inventories (thousand short tons)

Category
Marine SI evap
All Highway
Motorcycles
Total
CO
base with percent
case proposed reduction
standards
0 0 0%
569 569 0%
569 569 0%
NOx
base with percent
case proposed reduction
standards
0 0 0%
14 7 50%
14 7 50%
HC*
base with percent
case proposed reduction
standards
114 50 56%
58 29 50%
172 80 53%
* Evaporative HC for marine SI; exhaust HC for highway motorcycles.
       Table 4 summarizes the projected costs to meet the proposed emission standards. This is
our best estimate of the cost associated with adopting new technologies to meet the proposed
emission standards.  The analysis also considers total operating costs, including maintenance and
fuel consumption. All costs are presented in 2001 dollars.

                                          Table 4
               Estimated Average Cost Impacts of Proposed Emission Standards
Category
Marine SI diurnal
tank permeation
hose permeation
aggregate
Highway motorcycles <50cc
Highway motorcycles >50cc
Highway motorcycles >50cc
Standard
Date
2008
2006
2006
2010
Increased Production
Cost per Vehicle*
$9
$12
$14
$36
$44
$26
$35**
Lifetime Operating Costs
per Engine (NPV)
($4)
($9)
($14)
($27)
($8)
—
—
       * The estimated long-term costs for highway motorcycles >50cc decrease by about 35 percent.
       ** Costs presented are incremental to the first-phase standards.
       We also calculated the cost per ton of emission reductions for the proposed standards.
For both of the proposed programs, we attributed the entire cost of the proposed program to the
control of ozone precursor emissions (HC or NOx or both).  Table 5 presents the discounted cost-
per-ton estimates for the various engines factoring in the effect of reduced operating costs such as
fuel savings.

-------
Estimated Cost-]
Engine Type
Marine SI diurnal
tank permeation
hose permeation
aggregate
Highway Motorcycles <50cc
Highway Motorcycles >50cc
Highway Motorcycles >50cc
Aggregate
Date
2008
2006
2006
2010
—
Table 5
Der-Ton of the Proposed Emission Standards
Pollutant
HC
HC
HC+NOx
HC+NOx
—
Discounted
Reductions
per Vehicle
(short tons)
0.01
0.02
0.04
0.07
0.02
0.03
0.03
—
Discounted Cost per Ton
Without Fuel Savings
$745
$523
$367
$478
$2,130
$970
$1,230
$754
With Fuel Savings
$382
$160
$4
$115
$1,750
$970
$1,230
$515
       Table 6 presents the sum of the costs and emission benefits over the twenty-two year
period after all of the requirements are proposed to take effect, on both a non-discounted basis
and a discounted basis (assuming a seven percent discount rate).  The annualized cost and
emission benefits for the twenty-two year period (assuming the seven percent discount rate) are
also presented. (A twenty-two period is used in this aggregate analysis to cover the first twenty
years of each of the proposed standards which begins in 2006 for on-highway motorcycles and
concludes in 2008 for the proposed evaporative emission requirements for boats.)

                                        Table 6
              Annualized Cost and Emission Benefits for the Period 2006-2027
               due to the Proposed Requirements for All Equipment Categories

Undiscounted 22-
year Value
Discounted 22-
year Value
Annualized Value
HC+NOx
Benefits (tons)
1,550,000
616,000
56,000
Cost w/o Fuel
Savings
(Million $)
$934
$464
$42
Fuel Savings
(Million $)
$369
$147
$13
Cost w/
Fuel Savings
(Million $)
$565
$317
$29

-------
           CHAPTER  1:  Health and Welfare Concerns

       The engines and vehicles that would be subject to the proposed standards generate
emissions of HC, PM and air toxics that contribute to ozone nonattainment as well as adverse
health effects associated with ambient concentrations of PM and air toxics. This section
summarizes the general health effects of these substances. In it, we present information about
these health and environmental effects, air quality modeling results, and inventory estimates pre-
and post-control.

1.1 - Ozone

1.1.1 - General Background

       Ground-level ozone, the main ingredient in smog, is formed by complex chemical
reactions of volatile organic compounds (VOC) and NOx in the presence of heat and sunlight.
Ozone forms readily in the lower atmosphere, usually during hot summer weather. Volatile
organic compounds are emitted from a variety of sources, including motor vehicles, chemical
plants, refineries, factories, consumer and commercial products, and other industrial sources.
Volatile organic compounds also are emitted by natural sources such as vegetation.  Oxides of
nitrogen are emitted largely from motor vehicles, off-highway equipment, power plants, and
other sources of combustion. Hydrocarbons (HC) are a large subset of VOC, and to reduce
mobile source VOC levels we set maximum emissions limits for hydrocarbon as well as
particulate matter emissions.

       The science of ozone formation, transport, and accumulation is complex.  Ground-level
ozone is produced and destroyed in a cyclical set of chemical reactions involving NOx, VOC,
heat, and sunlight.1 As a result,  differences in weather patterns, as well as NOx and VOC levels,
contribute to daily, seasonal, and yearly differences in ozone concentrations and differences from
city to city.  Many of the chemical reactions that are part of the ozone-forming cycle are sensitive
to temperature  and sunlight. When ambient temperatures and sunlight levels remain high for
several days and the air is relatively stagnant,  ozone and its precursors can build up, resulting in
higher ambient ozone levels than typically would occur on a single high temperature day.
Further complicating matters, ozone also can  be transported into an area from pollution sources
found hundreds of miles upwind, resulting in  elevated ozone levels even in areas with low local
VOC or NOx emissions.

       On the  chemical level, NOx and VOC are the principal precursors to ozone formation.
The highest levels of ozone are produced when both VOC and NOx emissions are present in
significant quantities on clear summer days. Relatively small amounts of NOx enable ozone to
form rapidly when VOC levels are relatively high, but ozone production is quickly limited by
removal of the  NOx. Under these conditions, NOx reductions are highly effective in reducing
ozone while VOC reductions have little effect. Such conditions are called "NOx limited."

                                          1-1

-------
Because the contribution of VOC emissions from biogenic (natural) sources to local ambient
ozone concentrations can be significant, even some areas where man-made VOC emissions are
relatively low can be NOx limited.

       When NOx levels are relatively high and VOC levels relatively low, NOx forms
inorganic nitrates but relatively little ozone.  Such conditions are called "VOC limited." Under
these conditions, VOC reductions are effective in reducing ozone, but NOx reductions can
actually increase local ozone under certain circumstances. Even in VOC limited urban areas,
NOx reductions are not expected to increase ozone levels if the NOx reductions are sufficiently
large.

       Rural areas are almost always NOx limited, due to the relatively large amounts of
biogenic VOC emissions in such areas.  Urban areas can be either VOC or NOx limited, or a
mixture of both.

       Ozone concentrations in an area also can be lowered  by the reaction of nitric oxide with
ozone, forming nitrogen dioxide (NO2); as the air moves downwind and the cycle continues, the
NO2 forms additional ozone.  The importance of this reaction depends, in part, on the relative
concentrations of NOx, VOC, and ozone, all  of which change with time and location.

1.1.2 - Health  and Welfare Effects of Ozone and Its Precursors

       Based on a large number of recent studies, EPA has identified several key health effects
caused when people are exposed to levels of ozone found today in many areas of the country.2'3
Short-term exposures (1-3 hours) to high ambient ozone concentrations have been linked to
increased hospital admissions and emergency room visits for respiratory problems. For example,
studies conducted in the northeastern U.S. and Canada show that ozone air pollution is associated
with 10-20 percent of all of the summertime respiratory-related hospital admissions. Repeated
exposure to ozone can make people more susceptible to respiratory infection and lung
inflammation and can aggravate preexisting respiratory diseases, such as asthma. Prolonged (6
to 8 hours), repeated exposure to ozone can cause inflammation of the lung, impairment of lung
defense mechanisms, and possibly irreversible changes in lung structure, which over time could
lead to premature aging of the lungs and/or chronic respiratory illnesses such as emphysema and
chronic bronchitis.

       Children and outdoor workers are most at risk from ozone exposure because they
typically are active outside during the summer when ozone levels are highest.  For example,
summer camp studies in the eastern U.S. and southeastern Canada have reported significant
reductions in lung function in children who are active outdoors. Further, children are more at
risk than adults from ozone exposure because their respiratory systems are still developing.
Adults who are outdoors and are moderately active during the summer months, such as
construction workers and other outdoor workers, also are among those most at risk.  These
individuals, as well as people with respiratory illnesses such as asthma, especially asthmatic
                                           1-2

-------
children, can experience reduced lung function and increased respiratory symptoms, such as
chest pain and cough, when exposed to relatively low ozone levels during prolonged periods of
moderate exertion.

       Evidence also exists of a possible relationship between daily increases in ozone levels
and increases in daily mortality levels. While the magnitude of this relationship is too uncertain
to allow for direct quantification, the full body of evidence indicates the possibility of a positive
relationship between ozone exposure and premature mortality.

       In addition to human health effects, ozone adversely affects crop yield, vegetation and
forest growth,  and the durability of materials. Because ground-level ozone interferes with the
ability of a plant to produce and store food, plants become more susceptible to disease, insect
attack, harsh weather and other environmental stresses.  Ozone causes noticeable foliage damage
in many crops, trees, and ornamental plants (i.e., grass, flowers, shrubs) and causes reduced
growth in plants.  Studies indicate that current ambient levels of ozone are responsible for
damage to forests and ecosystems (including habitat for native animal species).  Ozone
chemically attacks elastomers (natural rubber and certain synthetic polymers), textile fibers and
dyes, and, to a lesser extent, paints. For  example, elastomers become brittle and crack, and dyes
fade after exposure to ozone.

       Volatile organic compounds emissions are detrimental not only for their role in forming
ozone, but also for their role as air toxics. Some VOCs emitted from motor vehicles are toxic
compounds. At elevated concentrations  and exposures, human health effects from air toxics can
range from respiratory effects to cancer.  Other health impacts include neurological
developmental and reproductive effects.  The lexicologically significant VOCs emitted in
substantial quantities from the engines that are the subject of this proposal are discussed in more
detail in Section 1.3, below.

1.1.3 - Additional Health and Welfare Effects of NOx Emissions

       In addition to their role as an ozone precursor, NOx emissions are associated with a wide
variety of other health and welfare effects.4 5  Nitrogen dioxide can irritate the lungs and lower
resistance to respiratory infection (such as influenza). NOx  emissions are an important precursor
to acid rain that may affect both terrestrial and aquatic ecosystems. Atmospheric deposition of
nitrogen leads to excess nutrient enrichment problems ("eutrophication") in the Chesapeake Bay
and several nationally important estuaries along the East and Gulf Coasts. Eutrophication can
produce multiple adverse effects on water quality and the aquatic environment, including
increased algal blooms, excessive phytoplankton growth, and low or no dissolved oxygen in
bottom waters. Eutrophication also reduces sunlight, causing losses in submerged aquatic
vegetation critical for healthy estuarine ecosystems. Deposition of nitrogen-containing
compounds also affects terrestrial ecosystems.  Nitrogen fertilization can alter growth patterns
and change the balance of species in an ecosystem. In extreme cases, this process can result in
nitrogen saturation when additions of nitrogen to soil over time exceed the capacity of plants and
                                           1-3

-------
microorganisms to utilize and retain the nitrogen. These environmental impacts are discussed
further in Sections 1.5.4 and 1.5.5, below.

       Elevated levels of nitrates in drinking water pose significant health risks, especially to
infants. Studies have shown that a substantial rise in nitrogen levels in surface waters are highly
correlated with human-generated inputs of nitrogen in those watersheds.6 These nitrogen inputs
are dominated by fertilizers and atmospheric deposition. Nitrogen dioxide and airborne nitrate
also contribute to pollutant haze, which impairs visibility and can reduce residential property
values and the value placed on scenic views.

1.1.4 - Ozone Nonattainment

       The current primary and secondary ozone National Ambient Air Quality Standard
(NAAQS) is 0.12 ppm daily maximum 1-hour concentration, not to be exceeded more than once
per year on average. The determination that an area is at risk of exceeding the ozone standard in
the future was made for all areas with current design values grater than or equal to 0.125 ppm (or
within a 10 percent margin) and with modeling evidence that exceedances will persist into the
future.

       Ground level ozone today remains a pervasive pollution problem in the United States. In
1999, 90.8 million people (1990 census) lived in 31 areas designated nonattainment under the 1-
hour ozone NAAQS.7 This sharp decline from the 101 nonattainment areas originally identified
under the Clean Air Act Amendments of 1990 demonstrates the effectiveness of the last decade's
worth of emission-control  programs. However, elevated ozone concentrations remain a serious
public health concern throughout the nation.

       Over the last decade, declines in ozone levels were found mostly in urban areas, where
emissions are heavily influenced by controls on mobile sources and their fuels.  Twenty-three
metropolitan areas have realized a decline in ozone levels since 1989, but at the same time ozone
levels in 11 metropolitan areas with 7 million people have increased.8 Regionally, California and
the Northeast have recorded significant reductions in peak ozone levels, while four other regions
(the Mid-Atlantic, the Southeast, the Central and Pacific Northwest) have seen ozone levels
increase.

       The highest ambient concentrations are currently found in suburban areas, consistent with
downwind transport of emissions from urban centers.  Concentrations in rural areas have risen to
the levels previously found only in cities.  Particularly relevant to this proposal, ozone levels at
17 of our National Parks have increased, and in 1998, ozone levels in two parks, Shenandoah
National Park and the Great Smoky Mountains National Park, were 30 to 40 percent higher than
the ozone NAAQS over the last decade.9

       To estimate future ozone levels, we refer to the modeling performed in conjunction with
the final rule for our most recent heavy-duty highway engine and fuel standards.10 We performed
                                           1-4

-------
a series of ozone air quality modeling simulations for nearly the entire Eastern U.S. covering
metropolitan areas from Texas to the Northeast.11 This ozone air quality model was based upon
the same modeling system as was used in the Tier 2 air quality analysis, with the addition of
updated inventory estimates for 2007 and 2030.  The model simulations were performed for
several emission scenarios, and the model outputs were combined with current air quality data to
identify areas expected to exceed the ozone NAAQS in 2007, 2020, and 2030.12  The results of
this modeling are contained in Table 1.1-1. Areas presented in Table 1.1-1 have 1997-99 air
quality data indicating violations of the 1-hour ozone NAAQS, or are within 10 percent of the
standard, are predicted to have exceedance in 2007, 2020, or 2030. An area was considered
likely to have future exceedances if exceedances were predicted by the model, and the area is
currently violating the 1-hour standard, or is within 10 percent of violating the 1-hour standard.
Table 1.1-1 shows that 37 areas with a 1999 population of 91 million people are at risk of
exceeding the 1-hour ozone standard in 2007.
                                           1-5

-------
Table 1.1-1: Eastern Metropolitan Areas with Modeled Exceedances of the 1-Hour Ozone
 Standard in 2007, 2020, or 2030 (Includes all emission controls through HD07 standards)
VISA or CMSA/ State
\tlanta, GA MSA
Barnstable-Yarmouth, MA MSA *
Baton Rouge, LA MSA
Beaumont-Port Arthur, TX MSA
Benton Harbor, MI MSA *
Biloxi-Gulfport-Pascagoula, MS MSA *
Birmingham, AL MSA
Boston- Worcester-Lawrence, MA CMSA
Charleston, WV MSA*
Charlotte-Gastonia-Rock Hill, NC MSA
Chicago-Gary-Kenosha, IL CMSA
Cincinnati-Hamilton, OH-KY-IN CMSA *
Cleveland-Akron, OH CMSA *
Detroit- Ann Arbor-Flint, MI CMSA
Grand Rapids-Muskegon-Holland, MI MSA*
lartford, CT MSA
louma, LA MSA *
rlouston-Galveston-Brazoria, TX CMSA
luntington- Ashland, WV-KY-OH MSA
.ake Charles, LA MSA*
.ouisville, KY-IN MSA
Vlacon, GA MSA
Memphis, TN-AR-MS MSA
Vlilwaukee-Racine, WI CMSA
Nashville, TN MSA
sfew London-Norwich, CT-RI MSA
stew Orleans, LA MSA *
s[ew York-Northern NJ-Long Island, NY-NJ-CT-PA
CMSA
Norfolk- Virginia Beach-Newport News, VA-NC MSA *
Mando, FL MSA *
Densacola, FL MSA
Philadelphia- Wilmington-Atlantic City, PA-NJ-DE-MD
CMSA
Drovidence-Fall River- Warwick,RI-MAMSA*
Richmond-Petersburg, VA MSA
5t. Louis, MO-IL MSA
lampa-St. Petersburg, FL MSA *
Washington-Baltimore
Total number of areas
Copulation
2007 2020 2030 pop (1999)
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
37
91.2
X

X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X

X

X
X
X
X
X
X

X

X
X
X
X
X
X
32
88.5
X

X
X
X
X
X
X

X
X
X
X
X
X
X
X
X
X
X
X

X
X
X
X
X
X
X
X

X
X
X
X

X
32
87.8
3.9
0.2
0.6
0.4
0.2
0.3
0.9
5.7
0.3
1.4
8.9
1.9
2.9
5.4
1.1
1.1
0.2
4.5
0.3
0.2
1
0.3
1.1
1.7
1.2
0.3
1.3
20.2
1.6
1.5
0.4
6
1.1
1
2.6
2.3
7.4

91.4
       * These areas have registered 1997-1999 ozone concentrations within 10 percent of standard.
                                          1-6

-------
       With regard to future ozone levels, our photochemical ozone modeling for 2020 predicts
exceedances of the 1-hour ozone standard in 32 areas with a total of 89 million people (1999
census; see Table 1.1-1).  We expect that the control strategies contained in this proposal for
nonroad engines, marine vessels, and highway motorcycles will further assist state efforts already
underway to attain and maintain the 1-hour ozone standard.

       The inventories that underlie this predictive modeling for 2020 and 2030 include
reductions from all current and committed to federal, state and local control programs, including
the recently promulgated NOx and PM standards for heavy-duty vehicles and low sulfur diesel
fuel. The geographic scope of these areas at risk of future exceedances underscores the need for
additional, nationwide controls of ozone precursors.

       It should be noted that this modeling did not attempt to examine the prospect of areas
attaining or maintaining the ozone standard with possible future controls  (i.e., controls beyond
current or committed federal, State  and local controls).  Therefore, this information should be
interpreted as indicating what areas are at risk of ozone violations in 2007, 2020 or 2030 without
federal or state measures that may be adopted and implemented in the future.  We expect many of
these areas to adopt additional emission reduction programs, but we are unable to quantify or rely
upon future reductions from additional State programs since they have not yet been adopted.

1.1.5 - Public Health and Welfare Concerns from Prolonged and Repeated Exposures to
Ozone

       In addition to the health effects described above, there exists a large body of scientific
literature that shows that harmful effects can occur from sustained levels  of ozone exposure
much lower than 0.125 ppm. Studies of prolonged exposures, those lasting about 7 hours,
showed health effects from  exposures to ozone concentrations as low as 0.08 ppm. Prolonged
and repeated exposures to ozone at  these levels are common in areas that do not attain the 1-hour
NAAQS, and also occur in areas where ambient concentrations of ozone  are in compliance with
the 1-hour NAAQS.

       Prolonged exposure to levels of ozone below the NAAQS have been reported to cause or
be statistically associated with transient pulmonary function responses, transient respiratory
symptoms, effects on exercise performance, increased airway responsiveness, increased
susceptibility to respiratory  infection, increased hospital and emergency room visits, and transient
pulmonary respiratory inflamation.  Such acute health effects have been observed following
prolonged exposures at moderate levels  of exertion at concentrations of ozone as low as 0.08
ppm, the lowest concentration tested.  The effects are more pronounced as concentrations
increase, affecting more subjects or having a greater effect on a given  subject in terms of
functional changes or symptoms. A detailed summary and discussion of the large body of ozone
health effects research may  be found in Chapters 6 through 9 (Volume 3) of the 1996 Criteria
Document for ozone.13 Monitoring data for indicates that 333 counties in 33 states exceed these
levels in  1997-99.14
                                           1-7

-------
       To provide a quantitative estimate of the projected number of people anticipated to reside
in areas in which ozone concentrations are predicted to exceed the 8-hour level of 0.08 to 0.12
ppm or higher for multiple days, we performed regional modeling using the variable-grid Urban
Airshed Model (UAM-V).15  UAM-V is a photochemical grid model that numerically simulates
the effects of emissions, advection, diffusion, chemistry, and surface removal processes on
pollutant concentrations within a 3-dimensional grid. As with the previous modeling analysis,
the inventories that underlie this predictive modeling include reductions from all current and
committed to federal, state and local control programs,  including the recently promulgated NOx
and PM standards for heavy-duty vehicles and low sulfur diesel fuel. This modeling forecast that
111 million people are predicted to live in areas that areas at risk of exceeding these moderate
ozone levels for prolonged periods of time in 2020 after accounting for expected inventory
reductions due to controls on light- and heavy-duty on-highway vehicles; that number is expected
to increase to 125 million in  2030.16 Prolonged and repeated ozone concentrations at these levels
are common in areas throughout the country, and are found both in areas that are exceeding, and
areas that are not exceeding,  the 1-hour ozone  standard. Areas with these high concentrations are
more widespread than those in nonattainment for that 1-hour ozone standard.

       Ozone at these levels can have other welfare effects, with damage to plants being of most
concern. Plant damage affects crop yields, forestry production, and ornamentals.  The adverse
effect of ozone on forests and other natural vegetation can in turn cause damage to associated
ecosystems, with additional resulting economic losses.  Prolonged ozone concentrations of 0.10
ppm can be phytotoxic to a large number of plant species, and can produce acute injury and
reduced crop yield and biomass production.  Ozone concentrations within the range of 0.05 to
0.10 ppm have the potential over a longer duration of creating chronic stress on vegetation that
can result in reduced plant growth and yield, shifts in competitive advantages in mixed
populations, decreased vigor, and injury. Ozone effects on vegetation are presented in more
detail in Chapter 5, Volume II of the 1996 Criteria Document.

1.2 - Particulate Matter

1.2.1 - General Background

       Particulate pollution is a problem affecting urban and non-urban localities in all regions
of the United States. Nonroad engines and vehicles and highway motorcycles that would be
subject to the proposed standards contribute to ambient paniculate matter (PM) levels in two
ways.  First, they contribute through direct emissions of particulate matter.  Second, they
contribute to indirect formation of PM through their emissions of organic carbon, especially HC.
Organic carbon accounts for  between 27 and 36 percent of fine particle mass depending on the
area of the country.

       Particulate matter represents a broad class of chemically and physically diverse
substances. It can be principally characterized as discrete particles that exist in the condensed
(liquid or solid) phase spanning several orders of magnitude in size.  All particles equal to and


                                           1-8

-------
less than 10 microns are called PM10.  Fine particles can be generally defined as those particles
with an aerodynamic diameter of 2.5 microns or less (also known as PM25), and coarse fraction
particles are those particles with an aerodynamic diameter greater than 2.5 microns, but equal to
or less than a nominal 10 microns.

       Manmade emissions that contribute to airborne particulate matter result principally from
combustion sources (stationary  and mobile sources) and fugitive emissions from industrial
processes and non-industrial processes (such as roadway dust from paved and unpaved roads,
wind erosion from cropland, construction, etc.). Human-generated sources of particles include a
variety of stationary sources (including power generating  plants, industrial operations,
manufacturing plants, waste disposal) and mobile sources (light- and heavy-duty on-road
vehicles, and off-highway vehicles such as construction, farming, industrial, locomotives, marine
vessels and other sources). Natural sources also contribute to particulate matter in the
atmosphere and include sources such as  wind erosion of geological material, sea spray, volcanic
emissions, biogenic emanation (e.g., pollen from plants, fungal spores), and wild fires.

       The chemical and physical properties of PM vary greatly with time, region, meteorology,
and source category. Particles may be emitted directly to  the atmosphere (primary particles) or
may be formed by transformations of gaseous emissions of sulfur dioxide, nitrogen oxides or
volatile organic compounds (secondary particles).  Secondary PM is dominated by sulfate in the
eastern U.S. and nitrate in the western U.S.17 The vast majority (>90 percent) of the direct
mobile source PM emissions and their secondary formation products are in the fine PM size
range.  Mobile sources can reasonably be estimated to contribute to ambient secondary nitrate
and sulfate PM in proportion to their contribution to total  NOx and SOx emissions.

               Table 1.2-1:  Percent Contribution to PM75 by Component, 1998

Sulfate
Elemental Carbon
Organic Carbon
Nitrate
Crustal Material
East
56
5
27
5
7
West
33
6
36
8
17
              Source: National Air Quality and Emissions Trends Report, 1998, March, 2000, at 28. This
              document is available at http://www.epa.gov/oar/aqtrnd98/. Relevant pages of this report can be
              found in Memorandum to Air Docket A-2000-01 from Jean Marie Revelt, September 5, 2001,
              Document No. II-A-63.
                                            1-9

-------
1.2.2 - Health and Welfare Effects of PM

       Particulate matter can adversely affect human health and welfare. Discussions of the
health and welfare effects associated with ambient PM can be found in the Air Quality Criteria
for Paniculate Matter.18

       Key EPA findings regarding the health risks posed by ambient PM are summarized as
follows:

a.      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.

b.      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.

c.      Published studies have found statistical associations between PM and several key health
       effects, including 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 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.

d.      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, we have  concluded the following with respect to sensitive populations:

       1.     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.

       2.     Individuals with infectious respiratory disease (e.g., pneumonia) are at greater risk
              of premature mortality and morbidity (e.g., hospitalization, aggravation of
                                           1-10

-------
              respiratory symptoms) due to exposure to ambient PM. Also, exposure to PM
              may increase individuals' susceptibility to respiratory infections.

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

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

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

e.      There are fundamental physical and chemical differences between fine and coarse fraction
       particles. The fine fraction contains acid aerosols, sulfates, nitrates, transition metals,
       diesel exhaust particles, and ultra fine particles; the coarse fraction typically contains high
       mineral concentrations, silica and resuspended  dust. It is reasonable to expect that
       differences may exist in both the nature of potential effects elicited by coarse and fine PM
       and the relative concentrations required to produce such effects.  Both fine and coarse
       particles can accumulate in the respiratory system. Exposure to coarse fraction particles
       is primarily associated with the aggravation of respiratory conditions such as asthma.
       Fine particles are most closely associated  with health effects such as premature death or
       hospital admissions, and for cardiopulmonary diseases.

       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. Particles also
contribute to soiling and materials damage.  Components of paniculate matter (e.g., sulfuric or
nitric acid) also contribute to acid deposition, nitrification of surface soils and water
eutrophication of surface water.

1.2.3 - PM Nonattainment

       The NAAQS for PM10 was established in  1987. According to these standards, the short
term (24-hour) standard of 150 |ig/m3 is not to be exceeded more than once per year on average
over three years. The long-term standard specifies an expected annual arithmetic mean not to
exceed 50 |ig/m3 over three years.

       The most recent PM10 monitoring data indicate that 14  designated PM10 nonattainment
areas with a projected population of 23 million violated the PM10 NAAQS in the period 1997-
1999.  Table 1.2-2 lists the 14 areas, and also indicates the PM10 nonattainment classification,
and 1999 projected population for each PM10 nonattainment area. The projected population in
1999 was based on 1990 population figures which were then increased by the amount of
population growth in the county from 1990 to 1999.
                                           1-11

-------
      Table 1.2-2:  PM,n Nonattainment Areas Violating the PM,n NAAQS in 1997- 1999
Nonattainment Area or County
Anthony, NM (Moderate)8
Clark Co [Las Vegas], NV (Serious)
Coachella Valley, CA (Serious)
El Paso Co, TX (Moderate) A
Hay den/Mi ami, AZ (Moderate)
Imperial Valley, CA (Moderate)
Los Angeles South Coast Air Basin, CA (Serious)
Nogales, AZ (Moderate)
Owens Valley, CA (Serious)
Phoenix, AZ (Serious)
San Joaquin Valley, CA (Serious)
Searles Valley, CA (Moderate)
Wallula, WA (Moderate)8
Washoe Co [Reno], NV (Moderate)
Total Areas: 14
1999 Population
(projected, in millions)
0.003
1.200
0.239
0.611
0.004
0.122
14.352
0.025
0.018
2.977
3.214
0.029
0.052
0.320
23.167
       A EPA has determined that continuing PM10 nonattainment in El Paso, TX is attributable to transport under
       section 179(B).
       B The violation in this area has been determined to be attributable to natural events under section 188(f) of
       the Act.

       In addition to the 14 PM10 nonattainment areas that are currently violating the PM10
NAAQS listed in Table 1.2-2, there are 25 unclassifiable areas that have recently recorded
ambient concentrations of PM10 above the PM10 NAAQS. EPA adopted a policy in 1996 that
allows areas with PM10 exceedances that are attributable to natural events to retain their
designation as unclassifiable if the State is taking all reasonable measures to safeguard public
health regardless of the sources of PM10 emissions. Areas that remain unclassifiable areas are not
required under the Clean Air Act to submit attainment plans, but we work with each of these
areas to understand the nature of the PM10 problem and to determine what best can be done to
reduce it.  With respect to the monitored violations reported in 1997-99 in the 25 areas
designated as unclassifiable, we have not yet excluded the possibility that factors such as a one-
time monitoring upset or natural events, which ordinarily would not result in an area being
designated as nonattainment for PM10, may be responsible for the problem. Emission reductions
from today's action will assist these currently unclassifiable areas to achieve ambient PM10
concentrations below the current PM10 NAAQS.

        Current 1999 PM25 monitored values, which cover about a third of the nation's counties,
indicate that at least 40 million people live in areas where long-term ambient fine particulate
matter levels are at or above 16 |ig/m3 (37 percent of the population in the areas with monitors).19
This 16 |ig/m3 threshold is the low end of the range of long term average PM2 5 concentrations in
cities where statistically significant associations were found with serious health effects, including
premature mortality.20 To estimate the number of people who live in areas where long-term
                                           1-12

-------
ambient fine particulate matter levels are at or above 16 |ig/m3 but for which there are no
monitors, we can use modeling.  According to our national modeled predictions, there were a
total of 76 million people (1996 population) living in areas with modeled annual average PM25
concentrations at or above 16 |ig/m3 (29 percent of the population).21

       To estimate future PM25 levels, we refer to the modeling performed in conjunction with
the final rule for our most recent heavy-duty highway engine and fuel standards using EPA's
Regulatory Model System for Aerosols and Deposition (REMSAD).22 The most appropriate
method of making these projections relies on the model to predict changes between current and
future states. Thus, we have estimated future conditions only for the areas with current PM2 5
monitored data (which covers about a third of the nation's counties). For these counties,
REMSAD predicts the current level of 37 percent of the population living in areas where fine PM
levels are at or above 16 |ig/m3 to increase to 49 percent in 2030.23

1.3 - Gaseous Air  Toxics

       In addition to the human health and welfare impacts described above, emissions from the
engines covered by this proposal also contain several other substances that are known or
suspected human or animal carcinogens, or have serious noncancer health effects.  These include
benzene, 1,3-butadiene, formaldehyde, acetaldehyde, and acrolein.  The health effects of these air
toxics are described in more detail in Chapter 1 of the Draft Regulatory Support Document for
this rule. Additional information can also be found in the Technical Support Document four our
final Mobile Source Air Toxics rule.24

1.3.1 - Benzene

       Benzene is an aromatic hydrocarbon which is present as a gas in both exhaust and
evaporative emissions from motor vehicles. Benzene in the exhaust, expressed as a percentage
of total organic gases (TOG), varies depending on control technology (e.g., type of catalyst) and
the levels of benzene and other aromatics in the fuel, but is generally about three to five percent.
The benzene fraction of evaporative emissions depends on control technology and fuel
composition and characteristics (e.g., benzene level and the  evaporation rate), and is generally
about one percent.25

       EPA has recently reconfirmed that benzene is a known human carcinogen by all routes of
exposure.26 Respiration is the major source of human exposure. Long-term respiratory exposure
to high levels of ambient benzene concentrations has been shown to cause cancer of the tissues
that form white blood cells.  Among these  are acute nonlymphocytic leukemia,27 chronic
lymphocytic leukemia and possibly multiple myeloma (primary malignant tumors  in the bone
marrow), although the evidence for the latter has decreased with more recent studies.28'29
Leukemias, lymphomas, and other tumor types have been observed in experimental animals
exposed to benzene by inhalation or oral administration. Exposure to benzene and/or its
metabolites has also been linked with genetic changes in humans and animals30 and increased


                                          1-13

-------
proliferation of mouse bone marrow cells.31 The occurrence of certain chromosomal changes in
individuals with known exposure to benzene may serve as a marker for those at risk for
contracting leukemia.32

       A number of adverse noncancer health effects, blood disorders such as preleukemia and
aplastic anemia, have also been associated with low-dose, long-term exposure to benzene.33
People with long-term exposure to benzene may experience harmful effects on the blood-forming
tissues, especially the bone marrow. These effects can disrupt normal blood production and
cause a decrease in important blood components, such as red blood cells and blood platelets,
leading to anemia (a reduction in the number of red blood cells), leukopenia (a reduction in the
number of white blood cells), or thrombocytopenia (a reduction in the number of blood platelets,
thus reducing the ability for blood to clot). Chronic inhalation exposure to benzene in humans
and animals results in pancytopenia,34 a condition characterized by decreased numbers of
circulating erythrocytes (red blood cells), leukocytes (white blood cells), and thrombocytes
(blood platelets).35'36 Individuals that develop pancytopenia  and have continued exposure to
benzene may develop aplastic anemia,37 whereas others exhibit both pancytopenia and bone
marrow hyperplasia (excessive cell formation), a condition that may indicate a preleukemic
state.38 39 The most sensitive noncancer effect observed in humans is the depression of absolute
lymphocyte counts in the circulating blood.40

1.3.2 - 1,3-Butadiene

       1,3-Butadiene is formed in vehicle exhaust by the incomplete combustion of fuel.  It is
not present in vehicle evaporative emissions, because it is not present in any appreciable amount
in fuel. 1,3-Butadiene accounts for 0.4 to 1.0 percent of total organic gas exhaust, depending on
control technology and fuel composition.41

       1,3-Butadiene was classified by EPA as a Group B2 (probable human) carcinogen in
1985.42 This classification was based on evidence  from two species of rodents and epidemiologic
data.  In the EPA1998 draft Health Risk Assessment of 1,3-Butadiene, that was reviewed by the
Science Advisory Board (SAB), the EPA proposed that 1,3-butadiene is a known human
carcinogen based on human epidemiologic, laboratory animal data, and supporting data such as
the genotoxicity of 1,3-butadiene metabolites.43 The Environmental Health Committee of EPA's
Scientific Advisory Board (SAB) reviewed the draft document in August 1998 and recommended
that 1,3-butadiene be classified as a probable human carcinogen, stating that designation of 1,3-
butadiene  as a known human carcinogen should be based on observational studies in humans,
without regard to mechanistic or other information.44  In applying the 1996 proposed Guidelines
for Carcinogen Risk Assessment, the Agency relies on both observational studies in humans as
well as experimental evidence demonstrating causality, and therefore the designation of 1,3-
butadiene  as a known human carcinogen remains applicable.45  The Agency has revised the draft
Health Risk Assessment of 1,3-Butadiene based on the SAB and public comments. The draft
Health Risk Assessment of 1,3-Butadiene will undergo the Agency consensus review, during
which time additional changes may be made prior to its public release and placement on the
                                          1-14

-------
Integrated Risk Information System (IRIS).

       1,3-Butadiene also causes a variety of noncancer reproductive and developmental effects
in mice and rats (no human data) when exposed to long-term, low doses of butadiene.46 The
most sensitive effect was reduced litter size at birth and at weaning. These effects were observed
in studies in which male mice exposed to 1,3-butadiene were mated with unexposed females. In
humans,  such an effect might manifest itself as an increased risk of spontaneous abortions,
miscarriages, still births, or very early deaths.  Long-term exposures to 1,3-butadiene should be
kept below its reference concentration of 4.0 microgram/m3 to avoid appreciable risks of these
reproductive and developmental effects.47  EPA has developed a draft chronic, subchronic, and
acute RfC values for 1,3-butadiene exposure as part of the draft risk characterization mentioned
above. The RfC values will be reported on IRIS.

1.3.3 - Formaldehyde

       Formaldehyde is the most prevalent aldehyde in vehicle exhaust. It is formed from
incomplete combustion of both gasoline and diesel fuel and accounts for one to four percent of
total organic gaseous emissions, depending on control technology and fuel composition.  It is not
found in  evaporative emissions.

       Formaldehyde exhibits extremely complex atmospheric behavior.48 It is  formed by the
atmospheric oxidation of virtually all organic species, including biogenic (produced by a living
organism) hydrocarbons.  Mobile sources contribute both primary formaldehyde (emitted directly
from motor vehicles) and secondary formaldehyde (formed from photooxidation of other VOCs
emitted from vehicles).

       EPA has classified formaldehyde as a probable human  carcinogen based on limited
evidence for carcinogenicity in humans and sufficient evidence of carcinogenicity in animal
studies, rats, mice,  hamsters, and monkeys.49 Epidemiological studies in occupationally exposed
workers suggest that long-term inhalation of formaldehyde may be associated with tumors of the
nasopharyngeal cavity (generally the area at the back of the mouth near the nose), nasal cavity,
and sinus. Studies  in experimental animals provide sufficient evidence that long-term inhalation
exposure to formaldehyde causes an increase in the incidence of squamous (epithelial)  cell
carcinomas (tumors) of the nasal cavity. The distribution of nasal tumors in rats suggests that not
only regional exposure but also local tissue susceptibility may be important for the distribution of
formaldehyde-induced tumors.50 Research has demonstrated that formaldehyde  produces
mutagenic activity  in cell cultures.51

       Formaldehyde exposure also causes a range of noncancer health effects.  At low
concentrations (0.05-2.0 ppm), irritation of the eyes (tearing of the eyes and increased blinking)
and mucous membranes is the principal effect observed in humans.  At exposure to 1-11 ppm,
other human upper respiratory effects associated with acute formaldehyde exposure include a dry
or sore throat, and a tingling sensation of the nose.  Sensitive individuals may experience these
                                          1-15

-------
effects at lower concentrations. Forty percent of formaldehyde-producing factory workers
reported nasal symptoms such as rhinitis (inflammation of the nasal membrane), nasal
obstruction, and nasal discharge following chronic exposure.52  In persons with bronchial asthma,
the upper respiratory irritation caused by formaldehyde can precipitate an acute asthmatic attack,
sometimes at concentrations below 5 ppm.53 Formaldehyde exposure may also cause bronchial
asthma-like symptoms in non-asthmatics.54 55

       Immune stimulation may occur following formaldehyde exposure, although conclusive
evidence is not available. Also, little is known about formaldehyde's effect on the central
nervous system.  Several animal inhalation studies have been conducted to assess the
developmental toxicity of formaldehyde. The only exposure-related effect noted in these studies
was decreased maternal body weight gain at the high-exposure level.  No adverse effects on
reproductive outcome of the fetuses that could be  attributed to treatment were noted.  An
inhalation reference concentration (RfC), below which long-term exposures would not pose
appreciable noncancer health risks, is not available for formaldehyde at this time.

1.3.4 - Acetaldehyde

       Acetaldehyde is a saturated aldehyde that is found in vehicle exhaust and is formed as a
result of incomplete combustion of both gasoline and diesel fuel.  It is not a component of
evaporative emissions. Acetaldehyde comprises 0.4 to 1.0 percent of total organic gas exhaust,
depending on control technology and fuel composition.56

       The atmospheric chemistry of acetaldehyde is similar in many respects to that of
formaldehyde.57  Like formaldehyde, it is produced and destroyed by atmospheric chemical
transformation.  Mobile  sources contribute to ambient acetaldehyde levels both by their primary
emissions and by secondary formation resulting from their VOC emissions.  Acetaldehyde
emissions are classified as a probable human carcinogen.  Studies in experimental animals
provide sufficient evidence that long-term inhalation exposure to acetaldehyde causes an increase
in the incidence of nasal squamous cell carcinomas (epithelial tissue)  and adenocarcinomas
(glandular tissue).58  59

       Noncancer effects in studies with rats and mice showed acetaldehyde to be moderately
toxic by the inhalation, oral, and intravenous routes.60  6162 The primary acute effect of exposure
to acetaldehyde vapors is irritation of the eyes, skin, and respiratory tract.  At high
concentrations, irritation and pulmonary effects can occur, which could facilitate the uptake of
other contaminants.  Little research exists that addresses the effects of inhalation of acetaldehyde
on reproductive and developmental effects.  The in vitro and in vivo studies provide evidence to
suggest that acetaldehyde may be the causative factor in birth defects observed in fetal alcohol
syndrome, though evidence is very limited linking these effects to inhalation exposure.  Long-
term exposures  should be kept below the reference concentration of 9 |j.g/m3 to avoid appreciable
risk of these noncancer health effects.63
                                          1-16

-------
1.3.5 - Acrolein

       Acrolein is extremely toxic to humans from the inhalation route of exposure, with acute
exposure resulting in upper respiratory tract irritation and congestion. Although no information
is available on its carcinogenic effects in humans, based on laboratory animal data, EPA
considers acrolein a possible human carcinogen.64
                                           1-17

-------
1.4 - Inventory Contributions

1.4.1 - Inventory Contribution

       The contribution of emissions from the nonroad engines and vehicles and highway
motorcycles that would be subject to the proposed standards to the national inventories of
pollutants that are associated with the health and public welfare effects described in this chapter
are considerable. To estimate nonroad engine and vehicle emission contributions, we used the
latest version of our NONROAD emissions model. This model computes nationwide, state, and
county emission levels for a wide variety of nonroad engines, and uses information on emission
rates, operating data, and population to determine annual emission levels of various pollutants.
Emission estimates for highway motorcycles were developed using information on the emission
levels of current motorcycles and updated information on motorcycle use provided by the
motorcycle industry. A more detailed description of the modeling and our estimation
methodology can be found in the Chapter 6 of this document.

       Baseline emission inventory estimates for the year 2000 for the marine vessels and
highway motorcycles covered by this proposal are summarized in Table 1.4-1. This table shows
the relative contributions of the different mobile-source categories to the overall national mobile-
source inventory.  Of the total emissions from mobile sources, evaporative emissions from spark-
ignition marine vessels contribute about 1.3 percent of HC. Highway motorcycles contribute
about 1.1 percent, 0.1 percent, 0.4 percent, and 0.1 percent of HC, NOx, CO, and PM emissions,
respectively, in the year 2000. When only nonroad emissions are considered, the marine
evaporative and motorcycle emissions would account for a larger share.

       Our draft emission projections for 2020 for the spark-ignition marine vessels and
highway motorcycles that would be subject to the proposed standards show that emissions from
these categories are expected to increase over  time if left uncontrolled. The projections for 2020
are summarized in Table 1.4-2 and indicate that the evaporative emissions from marine vessel
are expected to contribute 1.8 percent of mobile source HC, and motorcycles are expected to
contribute 2.3 percent, 0.2 percent, 0.6 percent, and 0.1 percent of mobile  source HC, NOx, CO,
and PM emissions in the year 2020. Population growth and the effects of other regulatory
control programs are factored into these projections.
                                          1-18

-------
                    Table 1.4-1
        Modeled Annual Emission Levels for
Mobile-Source Categories in 2000 (thousand short tons)
Category
Highway Motorcycles
Marine SI Evaporative
Marine SI Exhaust
Nonroad Industrial SI > 19 kW
Recreational SI
Recreation Marine CI
Nonroad SI< 19 kW
Nonroad CI
Commercial Marine CI
Locomotive
Total Nonroad
Total Highway
Aircraft
Total Mobile Sources
Total Man-Made Sources
Mobile Source percent of Total
Man-Made Sources
NOx
tons
8
0
32
306
13
24
106
2,625
977
1,192
5,275
7,981
178
13,434
24,538
55%
percent
of mobile
source
0.1%
0.0%
0.2%
2.3%
0.1%
0.2%
0.8%
19.5%
7.3%
8.9%
39%
59%
1%
100%
-
-
HC
tons
84
108
708
247
737
1
1,460
316
30
47
3,646
3,811
183
7,640
18,586
41%
percent of
mobile
source
1.1%
1.3%
9.6%
3.2%
9.6%
0.0%
19.1%
4. 1%
0.4%
0.6%
48%
50%
2%
100%
-
-
CO
tons
331
0
2,144
2,294
2,572
4
18,359
1,217
129
119
26,838
49,813
1,017
77,668
99,747
78%
percent of
mobile
source
0.4%
0.0%
2.8%
3.0%
3.3%
0.0%
23.6%
1.6%
0.2%
0.2%
35%
64%
1%
100%
-
-
PM
tons
0.4
0
38
1.6
5.7
1
50
253
41
30
420
240
39
699
3,095
23%
percent
of
mobile
source
0.1%
0.0%
5.4%
0.2%
0.8%
0. 1%
7.2%
36.2%
5.9%
4.3%
60%
34%
6%
100%
-
-
                       1-19

-------
                                       Table 1.4-2
                           Modeled Annual Emission Levels for
                   Mobile-Source Categories in 2020 (thousand short tons)
Category
Highway Motorcycles
Marine SI Evaporative
Marine SI Exhaust
Nonroad Industrial SI > 19 kW
Recreational SI
Recreation Marine CI
Nonroad SI< 19 kW
Nonroad CI
Commercial Marine CI
Locomotive
Total Nonroad
Total Highway
Aircraft
Total Mobile Sources
Total Man-Made Sources
Mobile Source percent of Total
Man-Made Sources
NOx
tons
14
0
58
486
27
39
106
1,791
819
611
3,937
2,050
232
6,219
16,195
38%
percent
of mobile
source
0.2%
0.0%
0.9%
7.8%
0.4%
0.6%
1.7%
28.8%
13.2%
9.8%
63%
33%
4%
100%
-
-
HC
tons
142
114
284
348
1,706
1
986
142
35
35
3,651
2,276
238
6,165
16,234
38%
percent of
mobile
source
2.3%
1.8%
4.6%
5.6%
27.7%
0.0%
16.0%
2.3%
0.6%
0.6%
59%
37%
4%
100%
-
-
CO
tons
572
0
1,985
2,991
5,407
6
27,352
1,462
160
119
39,482
48,906
1,387
89,775
113,443
79%
percent of
mobile
source
0.6%
0.0%
2.2%
3.3%
3.3%
0.0%
30.5%
1.6%
0.2%
0.1%
44%
54%
2%
100%
-
-
PM
tons
0.8
0
28
2.4
7.5
1.5
77
261
46
21
444
145
43
632
3,016
21%
percent
of
mobile
source
0.1%
0.0%
4.4%
0.4%
1.2%
0.2%
12.2%
41.3%
7.3%
3.3%
70%
23%
7%
100%
-
-
1.4.2 - Inventory Impacts on a Per Vehicle Basis

       In addition to the general inventory contributions described above, the engines that would
be subject to the proposed standards are more potent polluters than their highway counterparts in
that they have much higher emissions on a per vehicle basis.  On a typical summer day, an
average sized boat emits as many hydrocarbon emissions to a current passenger car driven 500
miles. A highway motorcycle driven 10 miles emits as many hydrocarbon emissions as a current
passenger car driven for 210 miles.

       The per engine emissions are important because they mean that operators of these engines
and vehicles, as well as those who work in their vicinity, are exposed to high levels of emissions,
many of which are air toxics. These effects are described in more detail in the next section.
                                          1-20

-------
1.5 - Other Health and Environmental Effects

1.5.1 - Carbon Monoxide

       Unlike many gases, CO is odorless, colorless, tasteless, and nonirritating. Carbon
monoxide results from incomplete combustion of fuel and is emitted directly from vehicle
tailpipes.  Incomplete combustion is most likely to occur at low air-to-fuel ratios in the engine.
These conditions are common during vehicle starting when air supply is restricted ("choked"),
when cars are not tuned properly, and at high altitude, where "thin" air effectively reduces the
amount of oxygen available for combustion (except in cars that are designed or adjusted to
compensate for altitude).  Carbon monoxide emissions increase dramatically in cold weather.
This is because engines need more fuel to start at cold temperatures and because some emission
control devices (such as oxygen sensors and catalytic converters) operate less efficiently when
they are cold. Also, nighttime inversion conditions are more frequent in the colder months of the
year.  This is  due to the enhanced stability in the atmospheric boundary layer, which inhibits
vertical mixing of emissions from the surface.

       Carbon monoxide enters the bloodstream through the lungs and forms
carboxyhemoglobin, a compound that inhibits the blood's capacity to carry oxygen to organs and
tissues.  Carbon monoxide has long been known to have substantial adverse effects on human
health, including toxic effects on blood and tissues, and effects on organ functions. Carbon
monoxide has been linked to increased risk for people with heart disease, reduced visual
perception, cognitive functions and aerobic capacity,  and possible fetal effects. Persons with
heart disease  are especially sensitive to carbon monoxide poisoning and may experience chest
pain if they breathe the gas while exercising. Infants, elderly persons, and individuals with
respiratory diseases are also particularly sensitive. Carbon monoxide can affect healthy
individuals, impairing exercise capacity, visual perception, manual dexterity, learning functions,
and ability to perform complex tasks.  More importantly to many individuals is the frequent
exposure of individuals to exhaust emissions from engines operating indoors.  The  Occupational
Safety and Health Administration sets standards regulating the concentration of indoor
pollutants, but high local CO levels are still commonplace.

       Several recent epidemiological studies have shown a link between CO and premature
morbidity (including angina, congestive heart failure, and other cardiovascular diseases).  Several
studies in the United States and Canada have also reported an  association of ambient CO
exposures with frequency of cardiovascular hospital admissions, especially for congestive heart
failure (CHF). An association of ambient CO exposure with mortality has also been reported in
epidemiological studies, though not as consistently or specifically as with CHF admissions. EPA
reviewed these studies as  part of the Criteria Document review process. The CO Criteria
Document (EPA 600/P-99/001F) contains additional  information about the health effects of CO,
human exposure, and air quality.  It was published as a final document and made available  to the
public in August 2000 (www.epa.gov/ncea/co/).
                                          1-21

-------
       The current primary NAAQS for CO are 35 parts per million for the one-hour average
and 9 parts per million for the eight-hour average. These values are not to be exceed more than
once per year.  Air quality carbon monoxide value is estimated using EPA guidance for
calculating design values. In 1999, 30.5 million people (1990 census) lived in 17 areas
designated nonattainment under the CO NAAQS.65

       Nationally, significant progress has been made over the last decade to reduce CO
emissions and ambient CO concentrations. Total CO emissions from all sources have decreased
16 percent from 1989 to 1998, and ambient CO concentrations decreased by 39 percent.  During
that time, while the mobile source CO contribution of the inventory remained steady at about 77
percent, the highway portion decreased from 62 percent of total CO emissions to 56 percent
while the nonroad portion increased from 17 percent to 22 percent.66 Over the next decade, we
would expect there to be a minor decreasing trend from the highway segment due primarily to the
more stringent standards for certain light-duty trucks (LDT2s).67 CO standards for passenger cars
and other light-duty trucks and heavy-duty vehicles did not change as a result of other recent
rulemakings.

1.5.2 - Acid Deposition

       Acid deposition, or acid rain as it is commonly known, occurs when SO2 and NOx react
in the atmosphere with water,  oxygen, and oxidants to form various acidic compounds that later
fall to earth in the form of precipitation or dry deposition of acidic particles.68 It contributes to
damage of trees at high elevations and in extreme cases may cause lakes and streams to become
so acidic that they cannot support aquatic life. In addition, acid deposition accelerates the decay
of building materials and paints, including irreplaceable buildings, statues, and sculptures that are
part of our nation's cultural heritage. To reduce damage to automotive paint caused by acid rain
and acidic dry deposition, some manufacturers use acid-resistant paints, at an average  cost of $5
per vehicle—a total of $61 million per year if applied to all new cars and trucks sold in the U.S.

       Acid deposition primarily affects bodies of water that rest atop soil with a limited ability
to neutralize acidic compounds.  The National Surface Water  Survey (NSWS) investigated the
effects of acidic deposition in  over 1,000 lakes larger than 10 acres and in thousands of miles of
streams. It found that acid deposition was the primary cause of acidity in 75 percent of the acidic
lakes and about 50 percent of the acidic streams, and that the areas most sensitive to acid rain
were the Adirondacks, the mid-Appalachian highlands, the upper Midwest and the high elevation
West. The NSWS found that approximately 580 streams in the Mid-Atlantic Coastal Plain are
acidic primarily due to acidic deposition. Hundreds of the lakes in the Adirondacks surveyed in
the NSWS have acidity levels incompatible with the survival of sensitive fish species. Many of
the over 1,350 acidic streams in the Mid-Atlantic Highlands (mid-Appalachia) region have
already experienced trout losses due to increased stream acidity. Emissions from U.S. sources
contribute to acidic deposition in eastern Canada, where the Canadian government has estimated
that 14,000 lakes are acidic. Acid deposition also has been implicated in contributing to
degradation of high-elevation  spruce forests that populate the ridges of the Appalachian
                                          1-22

-------
Mountains from Maine to Georgia. This area includes national parks such as the Shenandoah
and Great Smoky Mountain National Parks.

1.5.3 - Eutrophication and Nitrification

       Nitrogen deposition into bodies of water can cause problems beyond those associated
with acid rain. The Ecological Society of America has included discussion of the contribution of
air emissions to increasing nitrogen levels in surface waters in a recent major review of causes
and consequences of human alteration of the global nitrogen cycle in its Issues in Ecology
series.69 Long-term monitoring in the United States, Europe, and other developed regions of the
world shows a substantial rise of nitrogen levels in surface waters, which are highly correlated
with human-generated inputs of nitrogen to their watersheds. These nitrogen inputs are
dominated by fertilizers and atmospheric deposition.

       Human activity can increase the flow of nutrients into those waters and result in excess
algae and plant growth. This increased growth can cause numerous adverse ecological effects
and economic impacts, including nuisance  algal blooms, dieback of underwater plants due to
reduced light penetration, and toxic plankton blooms. Algal and plankton blooms can also
reduce the level of dissolved oxygen, which can also adversely affect fish and shellfish
populations.  This problem is of particular concern in coastal areas with poor or stratified
circulation patterns, such as the Chesapeake Bay, Long Island Sound, or the Gulf of Mexico. In
such areas, the "overproduced" algae tends  to sink to the bottom and decay, using all or most of
the available oxygen and thereby reducing or eliminating populations of bottom-feeder fish and
shellfish, distorting the normal population balance between  different aquatic organisms, and in
extreme cases causing dramatic fish kills.

       Collectively, these effects are referred to as eutrophication, which the National Research
Council recently identified as the most serious pollution problem facing the estuarine waters of
the United States (NRC, 1993). Nitrogen is the primary cause of eutrophi cation in  most coastal
waters and estuaries.70  On the New England coast, for example, the number of red  and
browntides and shellfish problems from nuisance and toxic  plankton blooms have increased over
the past two decades, a development thought to be linked to increased nitrogen loadings in
coastal waters.  We believe that airborne NOx contributes from 12 to 44 percent of the total
nitrogen loadings to United States coastal water bodies. For example, some estimates assert that
approximately one-quarter of the nitrogen in the Chesapeake Bay comes from atmospheric
deposition.

       Excessive fertilization with nitrogen-containing compounds can also affect  terrestrial
ecosystems.71 Research suggests that nitrogen fertilization can alter growth patterns and change
the balance of species in an ecosystem, providing beneficial nutrients to plant growth in areas
that do not suffer from nitrogen over-saturation.  In extreme cases, this process can result in
nitrogen saturation when additions of nitrogen to soil over time exceed the capacity of the plants
and microorganisms to utilize and retain the nitrogen.  This  phenomenon has already occurred in
                                          1-23

-------
some areas of the U.S.
                                        1-24

-------
Notes to Chapter 1

1.Carbon monoxide also participates in the production of ozone, albeit at a much slower rate than
most VOC and NOx compounds.

2.U.S. EPA, 1996, Review of National Ambient Air Quality Standards for Ozone, Assessment of
Scientific and Technical Information, OAQPS Staff Paper, EPA-452/R-96-007. A copy of this
document can be obtained from Air Docket A-99-06, Document No. II-A-22.

3.U.S. EPA, 1996, Air Quality Criteria for Ozone and Related Photochemical Oxidants,
EPA/600/P-93/004aF. The document is available on the internet at
http://www.epa.gov/ncea/ozone.htm. A copy can also be obtained from Air Docket A-99-06,
Documents Nos. U-A-15, II-A-16, U-A-17.

4.U.S. EPA, 1995, Review of National Ambient Air Quality Standards for Nitrogen
Dioxide,  Assessment of Scientific and Technical Information, OAQPS Staff Paper,
EPA-452/R-95-005.

5.U.S.EPA, 1993, Air Quality Criteria for Oxides of Nitrogen, EP A/600/8-9 l/049aF.

6.Vitousek, Pert M., John Aber, Robert W. Howarth, Gene E. Likens, et al. 1997. Human
Alteration of the Global Nitrogen Cycle: Causes and Consequences. Issues in Ecology.
Published by Ecological Society of America, Number 1, Spring 1997.

7.National Air Quality and Emissions Trends Report, 1999, EPA, 2001, at Table A-19.  This
document is available at http://www.epa.gov/oar/aqtrnd99/. The data from the Trends report are
the most recent EPA air quality data that has been quality assured. A copy of this table can also
be found in Docket No. A-2000-01, Document No. II-A-64.

S.National Air Quality and Emissions Trends Report, 1998, March, 2000, at 28. This document
is available at http://www.epa.gov/oar/aqtrnd98/.  Relevant pages of this report can be found in a
Memorandum to Air Docket A-2000-01 from Jean Marie Revelt, September 5, 2001,  Document
No. II-A-63.

9.National Air Quality and Emissions Trends Report, 1998, March, 2000, at 32. This document
is available at http ://www.epa. gov/oar/aqtrnd98/.  Relevant pages of this report can be found in
Memorandum to Air Docket A-2000-01 from Jean Marie Revelt, September 5, 2001,  Document
No. II-A-63.

10. Additional information about this modeling can be found in our Regulatory Impact Analysis:
Heavy-Duty Engine and Vehicle Standards and Highway Diesel Fuel Sulfur Control
Requirements, document EPA420-R-00-026, December 2000. Docket No. 1-2000-01,
Document No. U-A-13. This document is also available at
http://www.epa.gov/otaq/diesel.htmtfdocuments.

                                         1-25

-------
11 .We also performed ozone air quality modeling for the western United States but, as described
further in the air quality technical support document, model predictions were well below
corresponding ambient concentrations for our heavy-duty engine standards and fuel sulfur control
rulemaking. Because of poor model performance for this region of the country, the results of the
Western ozone modeling were not relied on for that rule.

12. Additional information about this modeling can be found in our Regulatory Impact Analysis:
Heavy-Duty Engine and Vehicle Standards and Highway Diesel Fuel Sulfur Control
Requirements, document EPA420-R-00-026, December 2000. Docket No. 1-2000-01,
Document No. n-A-13. This document is also available at
http://www.epa.gov/otaq/diesel.htmtfdocuments.

13. Air Quality Criteria Document for Ozone and Related Photochemical Oxidants, EPA National
Center for Environmental Assessment, July 1996, Report No. EPA/600/P-93/004cF.  The
document is available on the internet at http://www.epa.gov/ncea/ozone.htm.  A copy can also be
obtained from Air Docket No. A-99-06, Documents Nos. I-A-15, U-A-16, U-A-17.

14. A copy of this data can be found in Air Docket A-2000-01, Document No. II-A-80.

15.Memorandum to Docket A-99-06 from Eric Ginsburg, EPA, "Summary of Model-Adjusted
Ambient Concentrations for Certain Levels of Ground-Level Ozone over Prolonged Periods,"
November 22, 2000, Docket A-2000-01, Document Number JJ-B-13.

16.Memorandum to Docket A-99-06 from Eric Ginsburg, EPA, "Summary of Model-Adjusted
Ambient Concentrations for Certain Levels of Ground-Level Ozone over Prolonged Periods,"
November 22, 2000, at Table C, Control Scenario - 2020 Populations in Eastern Metropolitan
Counties with Predicted Daily 8-Hour Ozone greater than or equal to 0.080 ppm. Docket A-
2000-01, Document Number II-B-13.

17.U.S. EPA (1998) National Air Quality and Emissions Trends Report,  1997. Office of Air
Quality Planning and Standards. EPA 454/R-99-016.

18.EPA (1996) Review of the National Ambient Air Quality Standards for Particulate Matter:
Policy Assessment of Scientific and Technical Information OAQPS Staff Paper.  EPA-452/R-96-
013. Docket Number A-99-06, Documents Nos. H- A-18, 19, 20, and 23.  The parti culate matter
air quality criteria documents are also available at http://www.epa.gov/ncea/partmatt.htm.

19.Memorandum to Docket A-99-06 from Eric O. Ginsburg, Senior Program  Advisor,
"Summary of 1999 Ambient Concentrations of Fine Particulate Matter," November 15, 2000.
Air Docket A-2000-01, Document No. II-B-12.

20.EPA (1996) Review of the National Ambient Air Quality Standards for Particulate Matter:
Policy Assessment of Scientific and Technical Information OAQPS Staff Paper.  EPA-452/R-96-
013. Air Docket A-99-06, Documents Nos. II-A-18, 19, 20, and 23. The parti culate matter air
quality criteria documents are also available at http://www.epa.gov/ncea/partmatt.htm.

                                        1-26

-------
21.Memorandum to Docket A-99-06 from Eric O. Ginsburg, Senior Program Advisor,
"Summary of Absolute Modeled and Model-Adjusted Estimates of Fine Particulate Matter for
Selected Years," December 6, 2000. Air Docket A-2000-01, Document No. II-B-14.

22. Additional information about the Regulatory Model System for Aerosols and Deposition
(REMSAD) and our modeling protocols can be found in our Regulatory Impact Analysis: Heavy-
Duty Engine and Vehicle Standards and Highway Diesel Fuel Sulfur Control Requirements,
document EPA420-R-00-026, December 2000.  Docket No. A-2000-01, Document No. A-II-13.
This document is also available  at http://www.epa.gov/otaq/disel.htm#documents.

23. Technical Memorandum, EPA Air Docket A-99-06, Eric O. Ginsburg, Senior Program
Advisor, Emissions Monitoring  and Analysis Division, OAQPS, Summary of Absolute Modeled
and Model-Adjusted Estimates of Fine Particulate Matter for Selected Years, December 6, 2000,
Table P-2 Docket A-2000-01, Document I-B-14.

24.See our Mobile Source Air Toxics final rulemaking, 66 FR 17230, March 29, 2001, and the
Technical Support Document for that rulemaking.  Docket No. A-2000-01, Documents Nos. II-
A-42 and H-A-30.

25.U.S. EPA. (1999) Analysis of the Impacts of Control Programs on Motor Vehicle Toxic
Emissions and Exposure in Urban Areas and Nationwide: Volume I. Prepared  for EPA by Sierra
Research, Inc. and Radian International Corporation/Eastern Research Group, November 30,
1999. Report No.  EPA420-R-99-029. http://www.epa.gov/otaq/toxics.htm.

26.U.S. EPA (1998) Environmental Protection Agency, Carcinogenic Effects of Benzene: An
Update, National Center for Environmental Assessment, Washington, DC.  1998. EPA/600/P-
97/001F.  http://www.epa.gov/ncepihom/Catalog/EPA600P97001F.html.

27.Leukemia is a blood disease in which the white blood cells are abnormal in type or number.
Leukemia may be divided into nonlymphocytic (granulocytic) leukemias and lymphocytic
leukemias. Nonlymphocytic leukemia generally involves the types of white blood cells
(leukocytes) that are involved in engulfing, killing, and digesting bacteria and other parasites
(phagocytosis) as well as releasing chemicals involved in allergic and immune responses. This
type of leukemia may also involve erythroblastic cell types (immature red blood cells).
Lymphocytic leukemia involves the lymphocyte type of white bloods cell that are responsible for
the immune responses. Both nonlymphocytic and lymphocytic leukemia may, in turn, be
separated into acute (rapid and fatal) and chronic (lingering, lasting) forms. For example; in
acute myeloid leukemia (AML)  there is diminished production of normal red blood cells
(erythrocytes), granulocytes, and platelets (control clotting) which leads to death by anemia,
infection, or hemorrhage.  These events can be rapid.  In chronic myeloid leukemia (CML) the
leukemic cells retain the ability to differentiate (i.e., be responsive to stimulatory factors) and
perform function; later there is a loss of the ability to respond.
                                         1-27

-------
28.U.S. EPA (1985) Environmental Protection Agency, Interim quantitative cancer unit risk
estimates due to inhalation of benzene, prepared by the Office of Health and Environmental
Assessment, Carcinogen Assessment Group, Washington, DC. for the Office of Air Quality
Planning and Standards, Washington, DC., 1985.

29.Clement Associates, Inc.  (1991) Motor vehicle air toxics health information, for U.S. EPA
Office of Mobile Sources, Ann Arbor, MI, September 1991.  Air Docket A-2000-01, Document
No. II-A-49.

30.International Agency for Research on Cancer (IARC) (1982)  IARC monographs on the
evaluation of carcinogenic risk of chemicals to humans, Volume 29, Some industrial chemicals
and dyestuffs, International Agency for Research on Cancer, World Health Organization, Lyon,
France, p. 345-389.

31.Irons, R.D., W.S. Stillman, D.B. Colagiovanni,  and V.A. Henry (1992) Synergistic action of
the benzene metabolite hydroquinone on myelopoietic stimulating activity of
granulocyte/macrophage colony-stimulating factor in vitro, Proc. Natl. Acad. Sci. 89:3691-3695.

32.Lumley, M., H. Barker, and J.A. Murray (1990) Benzene in petrol, Lancet 336:1318-1319.

33.U.S. EPA (1993) Motor Vehicle-Related Air Toxics Study, U.S. Environmental Protection
Agency, Office of Mobile Sources, Ann Arbor, MI, EPA Report No. EPA 420-R-93-005, April
1993.

34.Pancytopenia is the reduction in the number of all three major types of blood cells
(erythrocytes, or red blood cells, thrombocytes, or platelets, and  leukocytes, or white blood cells).
In adults, all three major types of blood cells are produced in the bone marrow of the vertebra,
sternum, ribs, and pelvis.  The bone marrow contains immature cells, known as multipotent
myeloid stem cells, that later differentiate into the various mature blood cells. Pancytopenia
results from  a reduction in the ability of the red bone marrow to  produce adequate numbers of
these mature blood cells.

35.Aksoy, M (1991) Hematotoxicity, leukemogenicity and carcinogenicity  of chronic exposure
to benzene. In: Arinc, E.; Schenkman, J.B.; Hodgson, E., Eds. Molecular Aspects of
Monooxygenases and Bioactivation of Toxic Compounds.  New York: Plenum Press, pp. 415-
434.

36.Goldstein, B.D. (1988) Benzene toxicity. Occupational medicine.  State of the Art Reviews.
3: 541-554.

37.Aplastic anemia is a more severe blood disease  and occurs when the bone marrow ceases to
function, i.e.,these stem cells never reach maturity. The depression in bone marrow function
occurs in two stages - hyperplasia, or increased synthesis of blood cell elements, followed by
hypoplasia, or decreased synthesis. As the disease progresses, the bone marrow decreases
functioning.  This myeloplastic dysplasia (formation of abnormal tissue) without acute

                                         1-28

-------
leukemiais known as preleukemia. The aplastic anemia can progress to AML (acute mylogenous
leukemia).

38.Aksoy, M., S. Erdem, and G. Dincol. (1974) Leukemia in shoe-workers exposed chronically
to benzene. Blood 44:837.

39.Aksoy, M. and K. Erdem. (1978)  A follow-up study on the mortality and the development of
leukemia in 44 pancytopenic patients associated with long-term exposure to benzene. Blood 52:
285-292.

40.Rothman, N., G.L. Li, M. Dosemeci, W.E. Bechtold, G.E. Marti, Y.Z. Wang, M. Linet, L.Q.
Xi, W. Lu, M.T. Smith, N. Titenko-Holland, L.P. Zhang, W. Blot, S.N. Yin, and R.B. Hayes
(1996) Hematotoxicity among Chinese workers heavily exposed to benzene. Am. J. Ind. Med.
29: 236-246.

41.U.S. EPA. (1999)  Analysis of the Impacts of Control Programs on Motor Vehicle Toxic
Emissions and Exposure in Urban Areas and Nationwide: Volume I.  Prepared for EPA by Sierra
Research, Inc. and Radian International Corporation/Eastern Research Group, November 30,
1999. Report No.  EPA420-R-99-029. http://www.epa.gov/otaq/toxics.htm.

42.U.S. EPA (1985) Mutagenicity and Carcinogenicity Assessment of 1,3-Butadiene.
EPA/600/8-85/004F.  U.S. Environmental Protection Agency, Office of Health and
Environmental Assessment. Washington, DC.

43.U.S. EPA (1998) Draft Health Risk Assessment of  1,3-Butadiene, National Center for
Environmental Assessment, Office of Research and Development, U.S. EPA, EPA/600/P-
98/001 A, February 1998.

44. Scientific Advisory Board. 1998. An SAB Report: Review of the Health Risk Assessment of
1,3-Butadiene. EPA-SAB-EHC-98, August, 1998.

45.EPA 1996. Proposed guidelines for carcinogen risk assessment. Federal Register
61(79):17960-18011.

46.U.S. EPA (1985) Mutagenicity and carcinogenicity assessment of 1,3-butadiene. EPA/600/8-
85/004F. U.S.  Environmental Protection Agency, Office of Health and Environmental
Assessment. Washington, DC. http://www.epa.gov/ngispgm3/iris/subst/0139.htm.

47.U.S. EPA (1985) Mutagenicity and carcinogenicity assessment of 1,3-butadiene. EPA/600/8-
85/004F. U.S.  Environmental Protection Agency, Office of Health and Environmental
Assessment. Washington, DC.  http://www.epa.gov/ngispgm3/iris/subst/0139.htm.

48.Ligocki, M.P., GZ. Whitten, R.R.  Schulhof, M.C. Causley, and GM. Smylie (1991)
Atmospheric transformation of air toxics: benzene, 1,3-butadiene, and formaldehyde, Systems
Applications International, San Rafael, CA (SYSAPP-91/106).

                                         1-29

-------
49.U.S. EPA (1987) Environmental Protection Agency, Assessment of health risks to garment
workers and certain home residents from exposure to formaldehyde, Office of Pesticides and
Toxic Substances, April 1987.  Air Docket A-2000-01, Document No. II-A-48.

50.Clement Associates, Inc. (1991) Motor vehicle air toxics health information, for U.S. EPA
Office of Mobile Sources, Ann Arbor, MI, September 1991.  Air Docket A-2000-01, Document
No. II-A-49.

51.U.S. EPA (1993) Motor Vehicle-Related Air Toxics Study, U.S. Environmental Protection
Agency, Office of Mobile Sources, Ann Arbor, MI, EPA Report No.  EPA 420-R-93-005, April
1993. http://www.epa.gov/otaq/toxics.htm.

52.Wilhelmsson, B. and M. Holmstrom. (1987) Positive formaldehyde PAST after prolonged
formaldehyde exposure by inhalation. The Lancet:l64.

53.Burge, P.S., M.G. Harries, W.K. Lam, I.M. O'Brien, and P.A. Patchett. (1985)  Occupational
asthma due to formaldehyde. Thorax 40:225-260.

54.Hendrick, D.J., RJ. Rando, DJ. Lane, and MJ. Morris (1982) Formaldehyde asthma:
Challenge exposure levels and fate after five years. J. Occup. Med. 893-897.

SS.Nordman, H., H. Keskinen, and M. Tuppurainen. (1985) Formaldehyde asthma - rare or
overlooked? J. Allergy Clin. Immunol.  75:91-99.

56.U.S. EPA.  (1999) Analysis of the Impacts of Control Programs on Motor Vehicle Toxic
Emissions and Exposure in Urban Areas and Nationwide: Volume I.  Prepared for EPA by Sierra
Research, Inc. and Radian International Corporation/Eastern Research Group, November 30,
1999. Report No. EPA420-R-99-029. http://www.epa.gov/otaq/toxics.htm.

57.Ligocki, M.P., G.Z. Whitten (1991) Atmospheric transformation of air toxics: acetaldehyde
and polycyclic organic matter, Systems Applications International, San Rafael, CA, (SYSAPP-
91/113).

58. Environmental Protection Agency, Health assessment document for acetaldehyde, Office of
Health and Environmental Assessment, Environmental Criteria and Assessment Office, Research
Triangle Park, NC, EPA-600/8-86/015A (External Review Draft), 1987.

59.Environmental Protection Agency, Integrated Risk Information System (IRIS), Office of
Health and Environmental Assessment, Environmental Criteria and Assessment Office,
Cincinnati, OH, 1992. Acetaldehyde. http://www.epa.gov/iris/subst/0290.htm.

60.U.S. EPA (1987) Health Assessment Document for Acetaldehyde — External Review Draft.
Office of Health and Environmental Assessment, Research Triangle Park, NC.  Report No. EPA
600/8-86/015 A.
                                         1-30

-------
61.California Air Resources Board (CARB) (1992) Preliminary Draft: Proposed identification of
acetaldehyde as a toxic air contaminant, Part B Health assessment, California Air Resources
Board, Stationary Source Division, August,  1992. Air Docket A-2000-01, Document No. II-A-
34.

62.U.S. EPA (1997) Environmental Protection Agency, Integrated Risk Information System
(IRIS), Office of Health and Environmental  Assessment, Environmental Criteria and Assessment
Office, Cincinnati, OH, 1997. Acetaldehyde. http://www.epa.gov/iris/subst/0290.htm.

63.U.S. EPA (1999) Environmental Protection Agency, Integrated Risk Information System
(IRIS), Office of Health and Environmental  Assessment, Environmental Criteria and Assessment
Office, Cincinnati, OH. 1,3-Butadiene, http://www.epa.gov/iris/subst/0139.htm.

64.U.S. EPA (1993) Environmental Protection Agency, Integrated Risk Information System
(IRIS), Office of Health and Environmental  Assessment, Environmental Criteria and Assessment
Office, Cincinnati, OH. Acrolein.  http://www.epa.gov/ngispgm3/iris/subst/0364.htm.

65.National Air Quality and Emissions Trends Report, 1999, EPA, 2001, at Table A-19. This
document is available at http://www.epa.gov/oar/aqtrnd99/.  The data from the Trends report are
the most recent EPA air quality data that has been quality assured.  A copy of this table can also
be found in Docket No. A-2000-01, Document No. II-A-64.

66. Air Quality Criteria for Carbon Monoxide, US EPA, EPA 600/P-99/001F, June 2000, at 3-
10.

67. LDTs are light-duty trucks greater than 3750 Ibs. loaded vehicle weight, up through 6000
gross vehicle weight rating.

68.Much of the information in this subsection was excerpted from the EPA document, Human
Health Benefits from Sulfate Reduction, written under Title IV of the 1990 Clean Air Act
Amendments, U.S.  EPA, Office of Air and Radiation, Acid Rain Division, Washington, DC
20460, November 1995,  Air Docket A-2000-01, Document No. U-A-32.

69.Vitousek, Peter M., John Aber, Robert W. Howarth, Gene E. Likens, et al.  1997. Human
Alteration of the Global Nitrogen Cycle: Causes and Consequences. Issues in Ecology. Published
by Ecological Society of America, Number 1, Spring 1997.

70.Much of this information was taken from the following EPA document: Deposition of Air
Pollutants to the Great Waters-Second Report to Congress, Office of Air Quality Planning and
Standards, June 1997, EPA-453/R-97-011.

71.Terrestrial nitrogen deposition can act as a fertilizer. In some agricultural areas, this effect can
be beneficial.
                                         1-31

-------
            CHAPTER 2:   Industry Characterization
       To accurately assess the potential impact of this emission control program, it is important
to understand the nature of the affected industries.  This chapter describes relevant background
information related to highway motorcycles, vessels with spark-ignition (SI) marine engines, and
sterndrive and inboard marine engines.

2.1 - Highway Motorcycles

       Motorcycles come in a variety of two- and three-wheeled configurations and styles. For
the most part, however, they are two-wheeled self-powered vehicles. Federal regulations
currently define a motorcycle as "any motor vehicle with a headlight, taillight, and stoplight and
having: two wheels, or three wheels and a curb mass less than or equal to 793 kilograms (1749
pounds)." (See 40 CFR 86.402-98).  Highway motorcycles are those motorcycles that meet this
definition and are not excluded by the provisions that define a "motor vehicle" as, in part, a
vehicle that can exceed 25 mph.1  By definition a highway motorcycle is "street-legal," i.e., it can
be registered for use on public roads under existing state laws. Both EPA and California
regulations sub-divide highway motorcycles into classes based on engine displacement. Table
2.1-1 below shows how these classes are defined.

                                     Table 2.1-1
                                  Motorcycle Classes
Motorcycle
Class
Class I
Class II
Class III
Engine Displacement
(cubic centimeters)
50a-169
170 - 279
280 and greater
                          1 This proposal would extend Class I to include <50cc.
2.1.1 - Manufacturers

       Six companies account for about 95 percent of all motorcycles sold (Honda, Harley-
Davidson/Buell, Yamaha, Kawasaki, Suzuki, and BMW). All of these companies except Harley-
Davidson and BMW also manufacture off-road motorcycles and ATVs for the U.S. market. From
1996 to 2000, Harley-Davidson produced more on-road motorcycles than any other manufacturer
       1 Vehicles that appear to be motorcycle-like (e.g., mopeds, motorized bicycles, etc.) yet
can not exceed 25 mph are considered non-road recreational vehicles, and would be subject to
the proposed provisions for that category of vehicles.
                                         2-1

-------
and accounted for nearly 30 percent of total production by the six largest firms. Honda
accounted for 23.2 percent of total production by the six largest firms and Yamaha accounted for
18.4 percent.  Kawasaki and Suzuki each accounted for approximately 13 percent of production
by the six largest firms, while BMW accounted for 2.7 percent.  As a whole, these firms
increased production steadily in 1997, 1998, 1999, and 2000. In 2001, these firms are projected
to continue producing over 90 percent of all highway motorcycles manufactured for the US
market.

      Dozens of other companies make up the remaining few percent of sales. Many of these
are small U.S. companies manufacturing anywhere from a couple dozen to a couple thousand
motorcycles, although importers and U.S. affiliates of larger international companies also
contribute to the remaining few percent.

      With the exception of Harley-Davidson/Buell, all the major manufacturers produce at
least a portion of their on-road motorcycles for the U.S. market outside of the U.S. BMW,
Suzuki, and Yamaha manufacture all on-road motorcycles destined for the U.S. market outside of
the U.S.  Sixty-five percent of Honda's on-road motorcycles produced for the U.S. market are
manufactured outside the U.S., in either Italy or Japan. Sixty-one percent of Kawasaki's on-road
motorcycles destined for the U.S.  market are produced in Japan.

2.1.2 - Sales and Fleet Size

      Motorcycle sales  are sensitive to economic cycles due to the recreational nature of
motorcycle ownership. While certainly some are used as basic transportation, in general they are
purchased by consumers  who already own one or more cars, and the motorcycle is used for
recreation on summer evenings or weekends.  In much of the nation the motorcycle is not a
practical  means of transportation for one third of the year or more.  Sales were high in the late
1970's and early 1980's, but as the economy neared its downturn that started around 1987-1988
sales clearly began to drop.  Sales bottomed out with that downturn in 1992, then  as the economy
boomed in the 1990's sales once again began to take off, with retail dollars in 1998 approaching
triple the retail dollars often years before.  Historical sales and retail dollars are shown in Table
2.1-2. Data for 2000 indicate that the trend continued, with the retail sales of highway
motorcycles up  by more than 20 percent relative to 1999. The unit sales of highway motorcycles
in calendar year 2000 was approximately 437,000 (including dual-sport motorcycles).2  The
current fleet of highway motorcycles is approximately 4.3 million units.
       2  Dealernews.  Vol. 37, No. 2, Feb. 2001, p. 158.

                                          2-2

-------
                                       Table 2.1-2
                     On-Highway Motorcycle Retail Sales: 1982-1998*
Year
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
1985
1984
1983
1982
Units
(thousands)
387
311
260
242
230
228
217
203
206
226
245
335
465
470
565
605
605
605
Dollars
(millions)
3921
3132
2556
2213
1931
1773
1563
1333
1157
1148
1072
1304
1401
1375
1580
1111
1542
1463
                          * Source: Motorcycle Industry Council,
                          2000 Motorcycle Statistical Annual.
2.1.3 - Usage

       Highway motorcycles are primarily for use on public roads and typically fall into one of
four categories: cruiser, touring, sport bike, or standard. A cruiser motorcycle is designed for
relatively short distances and is the most commonly purchased on-road motorcycle in the U.S.  A
touring motorcycle is designed for long distance travel and extra load capacity. A sport bike is
designed for performance and uses racing technology, and finally a standard motorcycle is a basic

                                           2-3

-------
model. Each of these categories of motorcycles also differs in appearance and styling.  For
instance, a cruiser motorcycle has a heavier appearance, a custom paint job, full-view engines,
and swept-back handlebars. A touring bike is built for rider comfort and includes saddlebags.  In
contrast,  a sportbike is more aerodynamic in appearance, and has low handle bars and high
performance tires.

       The highway motorcycle category also includes motorcycles termed "dual-use" or "dual-
sport," meaning that their designs incorporate features that  enable them to be competent both on
and off road. Dual-sport motorcycles generally can be described as street-legal dirt bikes,
because they often bear a closer resemblance in terms of design features and engines to true off-
road motorcycles than to highway cruisers, touring, or sport bikes.

2.1.4 - Current Trends

       Analyzing production and sales numbers based on type of motorcycle and engine
displacement reveals that heavyweight motorcycles have become particularly popular in recent
years.  In 2000, 72 percent of all on-road motorcycles produced for the U.S. market were
heavyweight motorcycles (651 cc or greater displacement).  Just five years prior, heavyweight
motorcycles made up 67 percent of total production for the U.S. market.  Until relatively
recently,  Harley-Davidson was the only significant manufacturer of American heavyweight
cruiser and touring motorcycles.  In 2000, Harley-Davidson/Buell was still the largest producer
of heavyweight on-road motorcycles  (it produced 34 percent of all heavyweight motorcycles).
However, it has had insufficient production capacity to satisfy increased demand for heavyweight
motorcycles. Estimated shortfall in 1999 is 40,000 motorcycles or 30 percent of total demand.
Estimated annual growth in heavyweight motorcycle sales is 8 percent.  Harley-Davidson is
reported to have plans to increase production by only 9 percent, which allows it to just keep up
with annual growth in the market.

       Of the 175 engine families certified in 2002 by manufacturers for sale in the U.S., 151 fall
in the Class in category (above 279cc), representing more than 90 percent of projected sales.
More than three-quarters of projected 2001 highway motorcycle sales are above 700cc.  The
average displacement of all 2001 certified engine families is about 980cc, and the average
displacement of certified Class HI engine families is above  1 lOOcc.  The sales-weighted average
displacement of 2002 highway motorcycles is about 1 lOOcc.  Class I and U motorcycles, which
make up  less than seven percent of projected 2002 sales and only 24 out of 175 certified 2002
engine families, consist mostly of dual-sport bikes, scooters, and entry-level sportbikes and
cruisers.

       According to the Motorcycle Industry Council (MIC), in 1998 there were about 5.4
million highway motorcycles in use in the United States (565,000 of these were dual-sport).1
Total sales in 1998 of highway motorcycles was estimated to be about 411,000, or about 72
percent of motorcycle sales. About 13,000 of these were dual-sport motorcycles. Recent figures
for the 2000 calendar year show that retail sales approached 438,000 highway motorcycles, about
                                          2-4

-------
19,000 of which were dual-sport bikes.2

       In the second half of the 1990's, motorcycle production at Harley-Davidson fell short of
meeting the rapidly growing demand for their motorcycles.  The result has been the entry of at
least ten new companies into the heavyweight motorcycle market in the last several years. These
companies include two that ceased producing on-road motorcycles more than 50 years ago,
Excelsior Henderson Motorcycles and Indian Motorcycles.  Polaris began manufacturing Victory
motorcycles in 1998. American Eagle Motorcycle Company began commercial production in
1996. Other competitors include companies that custom-build heavyweight motorcycles from
mostly non-proprietary components, including Titan, Big Dog, Pure Steel, American Ironhorse,
and Ultra.

       A number of these companies have not been successful in the heavyweight motorcycle
market. American Quantum, Excelsior-Henderson, and Titan have filed for bankruptcy. One
possible explanation for the difficulty that smaller companies have had in profiting from
insufficient production of heavyweight motorcycle is that Harley increased its production by
more than expected. Harley-Davidson reports a 17.5 percent increase in production from 1998 to
1999. MIC projects that Harley-Davidson increased production by 15.4 percent between 1999
and 2000.

2.1.5 - Customer Concerns

       2.1.5.1 - Performance

       Adequate performance is clearly an important attribute for highway motorcycles. In
particular, buyers of sport or super-sport motorcycles are generally seeking performance that is
high or even extreme, sometimes rivaling the performance of exclusive racing motorcycles.  In
the touring and cruiser segments of the market this kind of outrageous performance is generally
not sought after; these bikes have attributes (such as riding style and position) that make high
performance a less important design factor.  For example, touring motorcycles are designed for
long-term riding comfort and luggage carrying capacity, and cruisers are often more focused on a
retro look, sound, and feel that allows them to be noticed cruising down Main Street, not
sprinting down the freeway.

       2.1.5.2 - Cost

       Motorcycles can range in price from around $1500 - $2000 for inexpensive entry-level or
dual-sport machines and mopeds and scooters to over $50,000 for elite custom machines.
According to Motorcycle Industry Council data the average amount spent by consumers on new
motorcycles in 1999 can be estimated at just over $10,300.3 As with other recreational vehicles,
highway motorcycles are generally discretionary purchases. Significant cost increases could
therefore result in decreased sales of these motorcycles if the increased costs cause consumers to
take their discretionary income elsewhere and into other recreational opportunities.
                                          2-5

-------
       2.1.5.3 - Consumer Modifications

       Many motorcycle owners personalize their motorcycles in a variety of ways.  This is one
of the aspects of motorcycle ownership that is appealing to a large number of motorcycle owners,
and they take their freedom to customize their bikes very seriously. However, there are some
forms of customization that are not legal under the provisions of Clean Air Act section 203(a),
which states that it is illegal:
       "for any person to remove or render inoperative any device or element of design
       installed on or in a motor vehicle or motor vehicle engine in compliance with
       regulations under this title ... after such sale and delivery to the ultimate
       purchaser..."
In other words, owners of motor vehicles cannot legally make modifications that cause the
emissions to exceed the applicable emissions standards, and they cannot remove or disable
emission-control devices installed by the manufacturer.

       We use the term "tampering" to refer specifically to actions that are illegal under Clean
Air Act section 203; the term, and the prohibition, do not apply generally to the wide range of
actions that a motorcycle enthusiast can do to personalize their vehicle, but only to actions that
remove or disable emission control devices or cause the emissions to exceed the standards. We
know, from anecdotal reports and from some data collected from in-use motorcycles, that a
portion of the motorcycle riding population has removed, replaced, or modified the original
equipment on their motorcycles. This customization can include changes that can be detrimental
(or, in some cases, possibly beneficial) to the motorcycle's emission levels.  The ANPRM sought
comments and data that could better help us understand the nature of the issue, such that our
proposal could be made with the best understanding possible of current consumer practices. We
did not intend to suggest that we would be revising the existing tampering restrictions to prohibit
many of the things that motorcycle owners are now doing legally.

       The proposed emissions  standards, if adopted by EPA, would not change this
"tampering" prohibition, which has been in place for more than 20 years. Owners would still be
free generally to customize their motorcycles in any way, as long as they do not  disable emission
controls or cause the motorcycle to exceed the emission standards. They would also be free, as
they are now, to perform routine maintenance on their motorcycles to restore or maintain the
motorcycle engine and related components in their original condition and configuration.

       2.1.5.4 - Safety

       The nature of motorcycling makes riders particularly aware of the many  safety issues that
confront them. Many riders that submitted comments to us following the publication of the
ANPRM in December of 2000 questioned whether catalytic converters could be implemented on
motorcycles without increasing the risk of harm to the rider and/or passenger. The primary
concern is regarding the close proximity of the riders to hot exhaust pipes and the catalytic
converter. Protecting the rider from the excessive heat is a concern for both riders and
                                           2-6

-------
manufacturers.  The current use of catalytic converters on a number of motorcycles (accounting
for tens of thousands in the current fleet) already indicates that these issues are not
insurmountable on a variety of motorcycle styles and engine sizes. A number of approaches to
shielding the rider from the heat of the catalytic converter are possible, such as exterior pipe
covers, shielded foot rests, and similar components. Some manufacturers have found that
placing the converter on the underside of the engine can keep it adequately distant from the rider.
Others may use double-pipe systems that reduce overall heat loss while remaining cooler on the
exterior. Based on the significant lead time that would be allowed for meeting these standards, as
well as on the two years of prior experience in California before meeting the requirements
federally, we believe that these issues can be satisfactorily resolved for the proportion of
motorcycles for which catalytic converters will be required.

2.2 - Marine

       This section gives a general characterization of the segments of the marine industry that
may be impacted by the proposed regulations.  For this discussion, we divide the recreational
marine industry into four segments:  gasoline engine manufacturers, boat builders, fuel tank
manufacturers,  and hose manufacturers. This industry characterization was developed in part
under contract with ICF Consulting4 as well as independent analyses conducted by EPA through
interaction with the industry and other sources.5'6'7

2.2.1 - Gasoline Engine Manufacturers

       2.2.1.1 - Identification of Gasoline Engine Manufacturers

       We have determined that there are at least 25 companies that manufacture sterndrive and
inboard gasoline engines for recreational vessels (including airboats and jetboats). Nineteen of
the identified companies are considered small businesses as defined by Small Business
Administration SIC code 3519 (less than 1000 employees).  These nineteen companies represent
10-20 percent of SD/I engine sales for 2000. Approximately 70-80 percent of gasoline SD/I
engines manufactured in 2000 can be attributed to one company. The next largest company is
responsible for about 10-20 percent of sales. Table 2.2-2 provides a list of gasoline engine
manufacturers identified to date by EPA.
                                           2-7

-------
                           Table 2.2-1 List of SI SD/I Marine
                       Engine Manufacturers Identified by EPA
               Greater than 1000 employees
Less than 1000 employees
                      Bombardier
                      Lamborghini
                       Mercruiser
                        Toyota
                      Volvo Penta
                        Yamaha
       Alumitech
    Boostpower USA
      Cobra Power
     Diamondback
    Flagship Marine
         GTO
    Indmar Products
         KEM
        Kodiak
     Marine Power
       Marshland
        Panther
     PCM/Crusader
     Redline Marine
     Revenge Power
  Rotary Power Marine
     Stump Jumper
   Torque Engineering
       Windmark
       There are currently eight companies that certify their engines to the U.S. standards for
outboards and personal watercraft. None of these companies would qualify as a small business.
These eight companies are Bombardier, Honda, Kawasaki, Mercury Marine, Polaris, Suzuki,
Tohatsu, and Yamaha.

       2.2.1.2 - Use of Gasoline Engines

       Information gathered by ICF suggests that a vast majority of inboard runabouts most
likely run on gasoline because of the small engines in these boats. Additionally, ICF estimates
that approximately 75 to 85 percent of inboard cruisers run on gasoline. Runabouts are
primarily designed for day-use where cruisers are generally designed with an enclosed cabin to
used for extended cruising.  Most cruisers that are less than 40 feet run on gasoline. Almost all
outboards and personal watercraft in operation today run on gasoline.

       2.2.1.3 - Current Trends

       Sales of SD/I marine engines (including jet boats) have grown over the past ten years.
Based on data provided by the National Marine Manufacturers Association (NMMA)8, sales of
                                         2-8

-------
gasoline these engines were 92,400 in 1991 and were 115,800 in 2000.  However, these sales are
down from peak sales years in the mid-to-late 1980's.  Sales of outboards have grown steadily
over the past ten years from 195,000 in 1991  to 241,600 in 2000. However, personal watercraft
sales have seen a steady decline over the past five years from 200,000 in their peak year of 1995
to 92,000 in 2000.

2.2.2 - Recreational Boat Builders

       2.2.2.1 - Identification of Boat Builders

       We have less precise information about recreational boat builders than is available about
engine manufacturers. We used several sources, including trade associations and Internet sites
when identifying entities that build and/or sell recreational boats. We have also worked with an
independent contractor to assist in the characterization  of this segment of the industry.  Finally,
we have obtained a list of nearly 1,700 boat builders known to the U.S. Coast Guard to produce
boats using recreational gasoline and diesel engines. At least 1,200 of these companies install
gasoline-fueled engines and would therefore be subject to the proposed evaporative emission
standards. More that 90 percent of the companies identified so far would be considered small
businesses as defined by SBA SIC code 3732.

       Based on information supplied by a variety of recreational boat builders, fuel tanks for
boats using SI marine engines are usually purchased from fuel tank manufacturers.  However,
some boat builders construct their own fuel tanks.  The boat builder provides the specifications to
the fuel tank manufacturer who helps match the fuel tank for a particular application.  It is the
boat builder's responsibility to install the fuel tank and connections into their vessel design.  For
vessels designed to be used with small outboard engines, the boat builder may not install a fuel
tank; therefore, the end user would use a portable fuel tank with a connection to the engine.

       2.2.2.2 - Current Trends

       Additional information provided by NMMA indicate that an estimated 72 million people
participated in recreational boating in 2000, which is down  slightly from 77 million in  1995.  In
2000, nearly 17 million boats were in use in the United States.

       2.2.2.3 - Production Practices

       Based on information supplied by a variety of recreational boat builders, the following
discussion provides a description of the general production  practices used in this sector of the
marine industry.

       Engines are usually purchased from factory authorized distribution centers.  The boat
builder provides the specifications to the distributor who helps match an engine for  a particular
application. It is the boat builders responsibility to fit the engine into their vessel design. The
                                           2-9

-------
reason for this is that sales directly to boat builders are a very small part of engine manufacturers'
total engine sales.  These engines are not generally interchangeable from one design to the next.
Each recreational boat builder has their own designs.  In general, a boat builder will design one
or two molds that are intended to last 5-8 years.  Very few changes are tolerated in the molds
because of the costs of building and retooling these molds.

       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 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.

2.2.3 - Fuel Tank  Manufacturers

       2.2.3.1 - Identification of Fuel  Tank Manufacturers

       We estimate that total sales of tanks for marine applications is approximately  550,000
units per year. The market is broken into manufacturers that produce plastic tanks and
manufacturers that produce aluminum tanks. We have determined that there are at  least nine
companies that make plastic fuel tanks with total sales of approximately 440,000 units per year.
There are at least four companies that make aluminum fuel tanks with total sales of
approximately 110,000 units per year. Most these fuel tank manufacturers are small businesses
as defined under SBA SIC Code 3713.  Table 2.2-3 provides a list of the diesel engine
manufacturers identified to date by EPA.

                                Table 2.2-3 List of Marine
                       Fuel Tank Manufacturers Identified by EPA
Plastic
Acrotech
Attwood
Chilton
Inca
Kracor
Moeller
Skyline
Aluminum
Coastline
Ezell
Florida Marine
RTS



                                           2-10

-------
       2.2.3.2 - Current Trends

       Until about 20 years ago, the majority of fuel tanks were make of aluminum.  Today,
however, only about a quarter of installed fuel tanks are aluminum. This shift was likely due
largely to the cost savings to boat builders by using plastic tanks.

       2.2.3.3 - Production Practices

       Plastic fuel tanks are either rotationally molded or blow molded.  Generally, portable fuel
tanks are blow molded.  Blow molding involves forming polyethylene in large molds using air
pressure to shape the tank.  Because this has high fixed costs, blow molding is only used where
production volumes are high. This works for portable fuel tanks where the volumes are high and
a single shape can be used for most applications.  For portable tanks, the fuel tank manufacturer
will generally design the tank, then send it out to a blow molder for production.

       Rotational molding is a lower cost alternative for smaller production volumes. In this
method, a mold is filled with a power form of polyethylene with a catalyst material. The mold is
rotated in an oven; the heat melts the plastic and activates the catalyst which causes a strong
cross-link material structure to form. This method is used for installed fuel tanks which
generally need to meet specific size and shape requirements for each boat design.

       Aluminum fuel tanks are also used primarily for installed fuel tanks. These tank
manufacturers generally custom make each tank to meet the boat manufacturers needs.
Generally, sheet aluminum is used and is cut, bent, and welded into the required configuration.

2.2.4 - Hose Manufacturers

       2.2.4.1 - Identification of Hose Manufacturers

       There are two primary U.S. manufacturers of extruded marine fuel hose (fuel  supply and
vent) used in marine applications.  These manufacturers are Parker-Dayco and Goodyear. They
generally sell their hose through distributers such as Lawrence Industries, MPI, Shields Marine
Hoses, and Triton Rubber.  Other hoses are imported from overseas, primarily from Asia; two
such brand names are Thermoid and Flexacol. Lawrence Industries is the only U.S. manufacturer
of marine fuel fill neck hose that we have identified. The majority of marine fuel fill neck hose
is imported.

       2.2.4.2 - Current Trends

       Marine hose is designed to meet the Coast Guard performance requirements as defined by
the Society of Automotive Engineer's recommended practice SAE J 1527.  For fuel supply lines,
this includes a permeation rate of 100 g/m2/day at 23°C (Class 1).  For other fuel hose not
                                          2-11

-------
normally continuously in contact with fuel (vent and fuel fill neck), the permeation standard is
300 g/m2/day (Class 2).  In general, boat builders will use Class 1 hose for both fuel supply and
vent lines for simplicity. Some boat builders use low permeation barrier hose which is well
below the permeation levels in SAE J 1527. For fuel fill necks, boat builders generally use Class
2 hose.

       2.2.4.3 - Production Practices

       Most fuel supply and vent marine hose is extruded nitrile rubber with a coating for better
wear and flame resistance. Because nitrile rubber has a very high permeation rate, this cover also
helps keep the permeation rate below the Coast Guard requirement. Under the SAE J 1527, hose
is defined as being reinforced  with fabric or wire. (In contrast, plastic automotive fuel lines are
extruded without reinforcement and are generally referred to as "tubing.") One hose
manufacturer offers a barrier marine hose that has much better fuel resistance than standard
nitrile hose.  This hose uses a  barrier layer of low permeability material, such as nylon or ethyl
vinyl alcohol, either on the inside surface or sandwiched between layers of nitrile rubber.

       Fuel fill hose is generally manufactured by hand wrapping layers of rubber and
reinforcement materials around a steel mandril. This hose is then heated to cure the rubber. Fuel
fill hose generally has a much larger diameter than fuel supply and vent hose and this process
offers an effective method of producing this larger diameter hose.  Although marine fuel fill
necks are not produced today with low permeation barriers, multi-layer chemical hoses are
produced using this method.  Therefore, we believe that this method could be used to produce
low permeation fill neck hoses.
                                          2-12

-------
Chapter 2 References


1.  Motorcycle Industry Council, "2000 Motorcycle Statistical Annual."

2.  DealerNews, volume 37, no. 2, February 2001.

3.  Motorcycle Industry Council, "2000 Motorcycle Statistical Annual."


4.  "Recreational Boating Industry Characterization," ICF Consulting, Contract No. 68-C-98-170,
WAN 0-5, August, 1999, Docket A-2000-01, Document U-A-9.

5.  Boating Industry Magazine, "1997 Annual Industry Reviews: The Boating Business," January
1998, www.boatbiz.com.

6.  Dunn & Bradstreet.

7.  National Marine Manufacturers Association -http:/www.nmma.org

8.  National Marine Manufacturers Association, "Boating 2000; Facts & Figures at a Glance,"
Prepared by the Marketing Statistics Department, 2000, Docket A-2000-01, Document II-A-95.
                                         2-13

-------
                         CHAPTER 3: Technology

       This chapter describes the current state of spark-ignition technology for engines and
evaporative emission technology as well as the emission control technologies expected to be
available for manufacturers. Chapter 4 presents the technical analysis of the feasibility of the
proposed standards.

3.1 - Introduction to Spark-Ignition Engine Technology

       The two most common types of engines are gasoline-fueled engines and diesel-fueled
engines. These engines have very different combustion mechanisms. Gasoline-fueled engines
initiate combustion using spark plugs, while diesel fueled engines initiate combustion by
compressing the fuel and air to high pressures. Thus these two types of engines are often more
generally referred to as "spark-ignition" and "compression-ignition" (or SI and CI) engines, and
include similar engines that used other fuels.  SI engines include engines fueled with LPG and
CNG.  SI engines may also be four-stroke or two-stroke which refers to the number of piston
strokes per combustion cycle. Motorcycles and SD/I engines are primarily spark-ignition, four-
stroke engines.

3.1.1 - Basics of Spark-Ignition Four-Stroke Engines

       Four-stroke engines are used in many different applications.  Virtually all highway
motorcycles, automobiles, and many trucks are powered by four-stroke SI engines.  Four-stroke
engines are also common in off-road motorcycles, all-terrain vehicles (ATVs), boats, airplanes,
and numerous nonroad applications such as lawn mowers, lawn and garden tractors, and
generators, to name just a few.

       A "four-stroke" engine gets it's name from the fact that the piston makes four passes or
strokes in the cylinder to complete an entire cycle. The strokes are intake, compression, power,
and exhaust. Two of the strokes are downward (intake & power) and two of the strokes are
upward (compression & exhaust).  Valves in the combustion chamber open and close to route
gases into and out of the combustion chamber or create compression.

       The first step of the cycle is for an intake valve in the combustion chamber to open during
the "intake" stroke allowing a mixture of air and fuel to be drawn into the cylinder while an
exhaust valve is closed and the piston moves down the cylinder.  The piston moves from top
dead center (TDC) or the highest piston position to bottom dead center (BDC) or lowest piston
position.  This creates a vacuum or suction in  the cylinder, which draws air and  fuel past the open
intake valve into the combustion chamber.

       The intake valve then closes and the momentum of the crankshaft causes the piston to
move back up the cylinder from BDC to TDC, compressing the air and fuel mixture. This is the
"compression" stroke.  As the piston nears TDC, at the very end of the compression stroke, the

                                          3-1

-------
air and fuel mixture is ignited by a spark from a spark plug and begins to burn. As the air and
fuel mixture burns, increasing temperature and pressure cause the piston to move back down the
cylinder, transmitting power to the crankshaft.  This is referred to as the "power" stroke.  The last
stroke in the four-stroke cycle is the "exhaust"  stroke. At the bottom of the power stroke, an
exhaust valve opens in the combustion chamber and as the piston moves back up the cylinder,
the burnt gases are pushed out through the exhaust valve to the exhaust manifold, and the cycle is
complete.

3.1.2 - Engine Calibration

       For most current SI engines, the two primary variables that manufacturers can control to
reduce emissions are the air and fuel mixture (henceforth referred to as air-fuel ratio) and the
spark timing. For highway motorcycles, these two variables are the most common methods for
controlling exhaust emissions.

             3.1.2.1  - Air-fuel ratio

       The calibration of the air-fuel mixture affects power, fuel consumption (referred to as
Brake Specific Fuel Consumption (BSFC)), and emissions for SI engines.  The effects of
changing the air-fuel mixture are shown in Figure 3-1.l  Traditionally, in most nonroad SI
applications, manufacturers have calibrated their fuel  systems for rich operation for two main
advantages.  First, by running the engine rich, manufacturers can reduce the risk of lean misfire
due to imperfect mixing of the fuel and air and variations in the air-fuel mixture from cylinder  to
cylinder. Second, by making extra fuel available for combustion, it is possible to get more power
from the engine. At the same time, since a rich mixture lacks sufficient oxygen for full
combustion, it results in increased fuel consumption rates and higher HC and CO emissions. As
can be seen from the figure, the best fuel consumption rates occur when the engine is running
lean.

       With the use of more advanced fuel systems, manufacturers would be able to improve
control of the air-fuel mixture in the cylinder. This improved control allows for leaner operation
without increasing the risk of lean misfire.  This reduces HC and CO emissions and fuel
consumption. Leaner  air-fuel mixtures, however, increase NOx emissions due to the higher
temperatures and increased supply of oxygen.
                                           3-2

-------
     Figure 3-1:  Effects of Air-fuel Ratio on Power, Fuel Consumption, and Emissions
 I
 CL
O
LL
OT
CQ
    Lean
Stoichiometric
                                    Rich
         Power
            BSFC
         I
                 I
                               ro
                               o
                               ta
                               CO
                               c
                               g

                               1
                               LJJ
                               "3
                               O
                                               LU
  Lean
Stoichiometric
Rich
                                                      NOx
       I
                  I
   0.7    0.8    0.9    1.0    1.1    1.2
            Fuel/Air Equivalence Ratio

       3.1.2.2 - Spark-timing:
                       1.3
0.7    0.8    0.9    1.0    1.1    1.2
         Fuel/Air Equivalence Ratio
                       1.3
       For each engine speed and air-fuel mixture, there is an optimum spark-timing that results
in peak torque.  If the spark is advanced to an earlier point in the cycle, more combustion occurs
during the compression stroke. If the spark is retarded to a later point in the cycle, peak cylinder
pressure is decreased because too much combustion occurs later in the expansion stroke when it
generates little torque on the crankshaft.  Timing retard may be used as a strategy for reducing
NOx emissions, because it suppresses peak cylinder temperatures that lead to high NOx levels.
Timing retard also results in higher exhaust gas temperatures, because less mechanical work is
extracted from the available energy. This may have the benefit of warming catalyst material to
more quickly reach the temperatures needed to operate effectively during light-load operation.2
Some automotive engine designs rely on timing retard at start-up to reduce cold-start emissions.

       Advancing the spark-timing at higher speeds gives the fuel more time to burn.  Retarding
the spark timing at lower speeds and loads avoids misfire. With a mechanically controlled
engine, a fly-weight or manifold vacuum system adjusts the timing.  Mechanical controls,
however, limit the manufacturer to a single timing curve when calibrating the engine.  This
means that the timing is not completely optimized for most modes of operation.

       3.1.2.3 - Fuel Metering

       Fuel injection has proven to be an effective and durable strategy for controlling emissions
and reducing fuel consumption from highway gasoline engines.  Comparable upgrades are also
available for gaseous fuels.  This section describes a variety of technologies available to improve
fuel metering.

       Throttle-body gasoline injection: A throttle-body system uses the same intake manifold

                                           3-3

-------
as a carbureted engine. However, the throttle body replaces the carburetor.  By injecting the fuel
into the intake air stream, the fuel is better atomized than if it were drawn through with a venturi.
This results in better mixing and more efficient combustion. In addition, the fuel can be more
precisely metered to achieve benefits for fuel economy, performance, and emission control.

       Throttle-body designs have the drawback of potentially large cylinder-to-cylinder
variations.  Like a carburetor, TBI injects the fuel into the intake air at a single location upstream
of all the cylinders.  Because the air-fuel mixture travels different routes to each cylinder, the
amount of fuel that reaches each cylinder will vary. Manufacturers account for this variation in
their design and may make compromises such as injecting extra fuel to ensure that the cylinder
with the leanest mixture will not misfire.  These compromises affect emissions and fuel
consumption.

       Multi-port gasoline injection:  As the name suggests, multi-port fuel injection means that
a fuel injector is placed at each of the intake ports. A quantity of fuel is injected each time the
intake valve opens for each cylinder.  This allows manufacturers to more precisely control the
amount of fuel injected for each combustion event. This control increases the manufacturer's
ability to optimize the air-fuel ratio for emissions,  performance, and fuel consumption. Because
of these benefits, multi-port injection is has been widely used in automotive applications for over
15 years.

       Sequential injection has further improved these systems by more carefully timing the
injection event with the intake valve opening. This improves fuel atomization and air-fuel
mixing, which further improves performance and control of emissions.

       A newer development to improve injector performance is air-assisted fuel injection.  By
injecting high pressure air along with the fuel spray, greater atomization of the fuel droplets can
occur.  Air-assisted  fuel injection is especially helpful in improving engine performance and
reducing emissions at low engine speeds. In  addition, industry studies have shown that the short
burst of additional fuel needed for responsive, smooth transient maneuvers can be reduced
significantly with air-assisted fuel injection due to a decrease in wall wetting in the intake
manifold. On a highway 3.8-liter engine with sequential  fuel injection, the air assist was shown
to reduce HC emissions by 27 percent during cold-start operating conditions.  At wide-open-
throttle with an air-fuel ratio of 17, the HC reduction was 43 percent when compared with a
standard injector.3

3.1.3 - Gaseous Fuels

       Engines operating on LPG or natural gas carry compressed fuel that is gaseous at
atmospheric pressure.  The technical challenges for gasoline related to an extended time to
vaporize the fuel don't apply to gaseous-fuel  engines. Typically, a mixer introduces the fuel into
the intake system. Manufacturers are pursuing new designs to inject the fuel directly into the
intake manifold. This improves control of the air-fuel ratio and the combustion event, similar to
                                           3-4

-------
the improvements in gasoline injection technology.

3.2 - Exhaust Emissions and Control Technologies

       HC and CO emissions from four-stroke engines are primarily the result of poor in-
cylinder combustion.  Higher levels of NOx emissions are the result of leaner air-fuel ratios and
the resulting higher combustion temperatures. Combustion chamber modifications can help
reduce HC emission levels, while using improved air-fuel ratio and spark timing calibrations, as
discussed in sections 3.1.2.1 and 3.1.2.2, can further reduce HC emissions and lower CO
emissions. The conversion from carburetor to EFI will also help reduce HC and CO emissions.
The use of exhaust gas recirculation on large SI engines can reduce NOx emissions, but is not
necessarily needed for recreational vehicles and highway  motorcycles, due to their relatively low
NOx emission levels. The addition of secondary air into the exhaust can significantly reduce HC
and CO emissions. Finally, the use catalytic converters can further reduce all three emissions.

3.2.1 - Combustion chamber design

       Unburned fuel can be trapped momentarily in crevice volumes (especially the space
between the piston and cylinder wall) before being released into the exhaust. Reducing crevice
volumes decreases this amount of unburned fuel, which reduces HC emissions.  One way to
reduce crevice volumes is to design pistons with piston rings closer to the top of the piston. HC
may be reduced by 3 to  10 percent by reducing crevice volumes, with negligible effects on NOx
emissions.4

       HC emissions also come from lubricating oil that  leaks into the combustion chamber.
The  heavier hydrocarbons in the oil generally don't burn completely. Oil in the combustion
chamber can also trap gaseous HC from the fuel and prevent it from burning.  For engines using
catalytic control, some components in lubricating oil can  poison the catalyst and reduce its
effectiveness, which would further increase emissions over time.  To reduce oil  consumption,
manufacturers can tighten tolerances and improve surface finishes for cylinders and pistons,
improve piston ring design and material, and improve exhaust valve stem seals to prevent
excessive leakage of lubricating oil into the combustion chamber.

3.2.2 - Exhaust gas recirculation

       Exhaust gas recirculation (EGR) has been in use in cars and trucks for many years.  The
recirculated gas acts as a diluent in the air-fuel mixture, slowing reaction rates and absorbing heat
to reduce combustion temperatures. These lower temperatures can reduce the engine-out NOx
formation rate by as much as 50 percent.5 HC is increased slightly due to lower temperatures for
HC burn-up during the late expansion and exhaust strokes.

       Depending on the burn rate of the engine and the amount of recirculated gases, EGR can
improve fuel consumption. Although EGR slows the burn rate, it  can offset this effect with some


                                          3-5

-------
benefits for engine efficiency. EGR reduces pumping work since the addition of recirculated gas
increases intake pressure. Because the burned gas temperature is decreased, there is less heat
loss to the exhaust and cylinder walls.  In effect, EGR allows more of the chemical energy in the
fuel to be converted to useable work.6

       For catalyst systems with high conversion efficiencies, the benefit of using EGR becomes
proportionally smaller.  Also, including EGR as a design variable for optimizing the engine adds
significantly to the development time needed to fully calibrate engine models.

3.2.3 - Secondary air

       Secondary injection of air into exhaust ports or pipes after cold start (e.g., the first 40-60
seconds) when the engine is operating rich, coupled with spark retard, can promote combustion
of unburned HC and CO in the exhaust manifold and increase the warm-up rate of the catalyst.
By means of an electrical or mechanical pump, secondary air is injected into the exhaust system,
preferably in close proximity of the exhaust valve.  Together with the oxygen of the secondary air
and the hot exhaust components of HC and CO, oxidation ahead of the catalyst  can bring about
an efficient increase in the exhaust temperature which helps the catalyst to heat  up quicker. The
exothermic reaction that occurs is  dependent on several parameters (secondary air mass, location
of secondary air injection, engine A/F ratio, engine air mass, ignition timing, manifold and
headpipe construction, etc.), and ensuring reproducibility demands detailed individual
application for each vehicle or engine design.

       Secondary air injection was first used as an emission control technique in itself without a
catalyst, and still is used for this purpose in many highway motorcycles and some off-highway
motorcycles to meet federal and California emission standards. For motorcycles, air is usually
provided or injected by a system of check valves which uses the normal pressure pulsations in the
exhaust manifold to draw in air from outside, rather than by a pump.

3.2.4 - Catalytic Aftertreatment

       Over the last several years, there have been tremendous advances in exhaust
aftertreatment systems.  Catalyst manufacturers are progressively moving to palladium (Pd) as
the main precious metal in automotive catalyst applications. Improvements to catalyst thermal
stability and washcoat technologies, the design of higher cell densities, and the use of two-layer
washcoat applications are just some of the advances made in catalyst technology.  There are two
types of catalytic converters commonly used: oxidation and three-way. Oxidation catalysts use
platinum and/or palladium to increase the rate of reaction between oxygen in the exhaust and
unburned HC and CO.  Ordinarily, this reaction woul proceed very slowly at temperatures typical
of engine exhaust.  The effectiveness of the catalyst depends on its temperature, on the air-fuel
ratio of the mixture, and on the mix of HC present.  Highly reactive species such as
formaldehyde and olefms are oxidized more effectively than less-reactive species.  Short-chain
paraffins such as methane, ethane, and propane are among the least reactive HC species, and are
                                           3-6

-------
difficult to oxidize.

       Three-way catalysts use a combination of platinum and/or pallabium and rhodium.  In
addition to promoting oxidation of HC and CO, these metals also promote the reduction of NO to
nitrogen and oxygen. In order for the NO reduction to occur efficiently, an overall rich or
stoichiometric air-fuel ratio is requires.  The NOx efficiency drops rapidly as the ai-fuel ratio
becomes leaner than stoichiometric.  If the air-fuel ratio can be maintained precisely at or just
rich of stoichiometic, a three-way catalyst can simultaneously oxidize HC and CO and reduce
NOx. The window of air-fuel ratios within which this is possible is very narrow and there  is a
trade-off between NOx and HC/CO control even within this window.

       There are several issues involved in designing catalytic control systems for the four-
stroke engines covered by this proposal.  The primary issues are the cost of the system, packaging
constraints, and the durability of the catalyst. This section addresses these issues.

       3.2.4.1 - System cost

       Sales volumes of industrial and recreational equipment are small compared to automotive
sales. Manufacturers therefore have a limited ability to recoup large R&D expenditures for
highway motorcycle engines. For this reason, we believe it is not appropriate to consider highly
refined catalyst systems that are  tailored specifically to nonroad applications. The cost of these
systems will decrease substantially when catalysts become commonplace. Chapter 5 describes
the estimated costs for a nonroad catalyst system.

       3.2.4.2 - Packaging constraints

       Many motorcycles of space constraints for adding a catalyst because they have been fine-
tuned over many years with a very compact fit.  Automotive  catalyst designs typically have one
or two catalyst units upstream of the muffler. This is a viable option for some motorcycles.
However,  if there is no available space to add a separate catalyst, it is possible to build a full
catalyst/muffler combination that fits in the same space as the conventional muffler. With  this
packaging option, even compact applications should have little or no trouble integrating a
catalyst into the equipment design.  The hundreds of catalysts already operating on highway
motorcycles clearly demonstrate this.

3.2.5 - Multiple valves and variable valve timing

       Four-stroke engines generally have two valves  for each cylinder, one for intake of the air-
fuel mixture and the other for exhaust of the combusted mixture. The duration and lift (distance
the valve head is pushed away from its seat) of valve openings is constant regardless of engine
speed. As engine speed increases, the aerodynamic resistance to pumping air in and out of the
cylinder for intake and exhaust also increases. Automotive engines have started to use two
intake and two exhaust valves to reduce pumping losses  and improve their volumetric efficiency
                                           3-7

-------
and useful power output.  Some highway motorcycles have used multiple valves for years,
especially the high-performance sport motorcycles.

       In addition to gains in breathing, 4-valve designs allow the spark plug to be positioned
closer to the center of the combustion chamber, which decreases the distance the flame must
travel inside the chamber.  This decreases the likelihood of flame-out conditions in the areas of
the combustion chamber farthest from the spark plug. In addition, the two streams of incoming
gas can be used to achieve greater mixing of air and fuel, further increasing combustion
efficiency and lowering engine-out emissions.

       Control of valve timing and lift take full advantage of the 4-valve configuration for even
greater improvement in combustion efficiency.  Engines normally use fixed-valve timing and lift
across all engine speeds. If the valve timing is optimized for low-speed torque, it may offer
compromised performance under higher-speed operation.  At light engine loads, for example, it
is desirable to close the intake valve early to reduce pumping losses. Variable-valve timing can
enhance both low-speed and high-speed performance with compromise.  Variable-valve timing
can allow for increased swirl and  intake charge velocity, especially during low-load operating
conditions where this is most problematic. By providing a strong swirl formation in the
combustion chamber, the air-fuel  mixture can mix sufficiently, resulting in a faster, more
complete combustion,  even under lean air-fuel conditions, thereby reducing emissions.

       Variable-valve technology by itself may have somewhat limited effect on reducing
emissions, but combining it with optimized spark plug location and exhaust gas recirculation can
lead to substantial emission reductions.

3.2.6 - Advanced Emission Controls

       On February 10, 2000, EPA published new "Tier 2" emissions standards for all passenger
vehicles, including sport utility vehicles (SUVs), minivans, vans and pick-up trucks.  The new
standards will ensure that exhaust VOC emissions be reduced to less than 0.1 g/mi on average
over the fleet, and that evaporative emissions be reduced by at least 50 percent. Onboard
refueling vapor recovery requirements were also extended to medium-duty passenger vehicles.
By 2020, these standards will reduce VOC emissions from light-duty vehicles by more than 25
percent of the projected baseline inventory.  To achieve these reductions, manufacturers will
need to incorporate advanced emission controls, including: larger and improved close-coupled
catalysts, optimized spark timing  and  fuel control, improved exhaust systems.

       To reduce emissions gasoline-fueled vehicle manufacturers have designed their engines to
achieve virtually complete combustion and have installed catalytic converters in the exhaust
system. In order for these controls to  work well for gasoline-fueled vehicles, it is necessary  to
maintain the mixture of air and fuel at a nearly stoichiometric ratio (that  is, just enough air to
completely burn the fuel).  Poor air-fuel mixture can result in significantly higher emissions of
incompletely combusted fuel.  Current generation highway vehicles are able to maintain
                                           3-8

-------
stoichiometry by using closed-loop electronic feedback control of the fuel systems. As part of
these systems, technologies have been developed to closely meter the amount of fuel entering the
combustion chamber to promote complete combustion.  Sequential multi-point fuel injection
delivers a more precise amount of fuel to each cylinder independently and at the appropriate time
increasing engine efficiency and fuel economy. Electronic throttle control offers a faster
response to engine operational changes than mechanical throttle control can achieve, but it is
currently considered expensive and only used on some higher-price vehicles. The greatest gains
in fuel control can be made through engine calibrations — the algorithms contained in the
powertrain control module (PCM) software that control the operation of various engine and
emission control components/systems. As microprocessor speed becomes faster, it is possible to
perform quicker calculations and to increase response times for controlling engine parameters
such as fuel rate and spark timing.  Other advances in engine design have also been used to
reduce engine-out emissions, including: the reduction of crevice volumes in the combustion
chamber to prevent trapping of unburned fuel; "fast burn"  combustion chamber designs that
promote swirl  and flame propagation; and multiple valves with variable-valve timing to reduce
pumping losses and improve efficiency. These technologies are discussed in more detail in the
RIA for the Tier 2 FRM.3

       As noted above, manufacturers are also using aftertreatment control devices to control
emissions.  New three-way catalysts for highway vehicles  are so effective that once a TWC
reaches its operating temperature, emissions are virtually undetectable.4  Manufacturers are now
working to improve the durability of the TWC and to reduce light-off time (that is, the amount of
time necessary after starting the engine before the catalyst reaches its operating temperature and
is effectively controlling VOCs and other pollutants).  EPA expects that manufacturers will be
able to design  their catalyst systems so that they light off within less than thirty seconds of engine
starting.  Other potential exhaust aftertreatment systems that could further reduce cold-start
emissions are thermally insulated catalysts, electrically heated catalysts, and HC adsorbers (or
traps).  Each of these technologies, which are discussed below, offer the potential for VOC
reductions in the future.  There are technological, implementation, and cost issues that still need
to be addressed, and at this time, it appears that these technologies would not be a cost-effective
means of reducing nonroad emissions on a nationwide basis.

       Thermally insulated catalysts maintain sufficiently high catalyst temperatures by
surrounding the catalyst with an insulating vacuum. Prototypes of this technology have
demonstrated the ability to store heat for more than 12 hours.5 Since ordinary catalysts typically
       3  http://www.epa.gov/otaq/tr2home.htmtfDocuments. EPA 420-R-99-023

       4  McDonald, J., L. Jones, Demonstration of Tier 2 Emission Levels for Heavy Light-Duty Trucks, SAE
2000-01-1957.

       5  Burch, S.D., and J.P. Biel, SULEV and "Off-Cycle" Emissions Benefits of a Vacuum-Insulated Catalytic
Convert, SAE 1999-01-0461.

                                           3-9

-------
cool down below their light-off temperature in less than one hour, this technology could reduce
in-use emissions for vehicles that have multiple cold-starts in a single day.  However, this
technology would have less impact on emissions from vehicles that have only one or two cold-
starts per day.

       Electrically-heated catalysts reduce cold-start emissions by applying an electric current to
the catalyst before the engine is started to get the catalyst up to its operating temperature more
quickly.6 These systems require a modified catalyst, as well as an upgraded battery and charging
system. These can greatly reduce cold-start emissions, but could require the driver to wait until
the catalyst is heated before the engine would start to achieve optimum performance.

       Hydrocarbon adsorbers are designed to trap VOCs while the catalyst is cold and unable to
sufficiently convert them.  They accomplish this by utilizing an adsorbing material which holds
onto the VOC molecules.  Once the catalyst is warmed up, the trapped VOCs are automatically
released from the adsorption material and are converted by the fully functioning downstream
three-way catalyst. There are three principal methods for incorporating an adsorber into the
exhaust system.  The first is to  coat the adsorber directly on the catalyst substrate. The advantage
is that there are no changes to the exhaust system required, but the desorption process cannot be
easily controlled and usually occurs before the catalyst has reached light-off temperature. The
second method locates the adsorber in another exhaust pipe parallel with the main exhaust pipe,
but in front of the catalyst and includes a series of valves that route the exhaust through the
adsorber in the first few seconds after cold  start, switching exhaust flow through the catalyst
thereafter. Under this system, mechanisms to purge the adsorber are also required. The third
method places the trap at the end of the exhaust system, in another exhaust pipe parallel to the
muffler, because of the low thermal tolerance of adsorber material.  Again a purging mechanism
is required to purge the adsorbed VOCs back into the catalyst, but adsorber overheating is
avoided.  One manufacturer who incorporates a zeolite hydrocarbon adsorber in its California
SULEV vehicle found that an electrically heated catalyst was necessary after the adsorber
because the zeolite acts as a heat sink and nearly negates the  cold start advantage of the adsorber.
This approach has been demonstrated to effectively reduce cold start emissions.

3.3 - Evaporative Emissions

3.3.1 - Sources of Evaporative Emissions

       Evaporative emissions from vessels using spark-ignition marine engines can be very high.
This is largely because they have fuel tanks that are vented to the atmosphere and because
materials are used in the construction of the plastic fuel tanks and hoses which have high
permeation rates.  Evaporative  emissions can be grouped into five categories:

       DIURNAL: Gasoline evaporation increases as the temperature rises during the day,
         Laing, P.M., Development of an Alternator-Powered Electrically-Heated Catalyst System, SAE 941042.

                                          3-10

-------
heating the fuel tank and venting gasoline vapors.

       RUNNING LOSSES: The hot engine and exhaust system can vaporize gasoline when the
engine is running.

       HOT SOAK: The engine remains hot for a period of time after the engine is turned off
and gasoline evaporation  continues.

       REFUELING:  Gasoline vapors are always present in typical fuel tanks. These vapors are
forced out when the tank  is filled with liquid fuel.

       PERMEATION:  Gasoline molecules can saturate plastic fuel tanks and rubber hoses,
resulting in a relatively constant rate of emissions as the fuel continues to permeate through these
components.

       Among the factors that affect emission rates are: (1) fuel metering  (fuel injection or
carburetor); (2) the degree to which fuel permeates fuel lines and fuel tanks; (3) the proximity of
the fuel tank to the exhaust system or other heat sources; (4) whether the fuel system is sealed
and the pressure at which fuel vapors are vented; and (5) fuel tank volume.

       3.3.1.1 - Diurnal  and Running Loss Emissions

       In an open fuel tank, the vapor space is at atmospheric pressure (typically about 14.7 psi),
and contains a mixture of fuel vapor and air. At all temperatures below the fuel's boiling point,
the vapor pressure of the fuel is less than atmospheric pressure.  This is also called the partial
pressure of the fuel vapor. The partial pressure of the air is equal to the difference between
atmospheric pressure and the fuel vapor pressure. For example, in an open-vented fuel tank at
60°F, the vapor pressure of typical gasoline would be about 4.5 psi. In this example, the partial
pressure of the air would  be about 10.2 psi.  Assuming that the vapor mixture behaves as an ideal
gas, then the mole fractions (or volumetric fractions)  of fuel vapor and air would be equal to their
respective partial pressures divided by the total pressure; thus, the fuel would be 31 percent of the
mixture (4.5/14.7) and the air would be 69 percent of the mixture (10.2/14.7).

       Diurnal emissions occur when the fuel temperature increases, which increases the
equilibrium vapor pressure of the fuel. For example,  assume that the fuel  in the previous
example was heated to 90°F, where the vapor pressure that same typical fuel would be about 8.0
psi.  To maintain the vapor space at atmospheric pressure, the  partial pressure of the air would
need to decrease to 6.7 psi, which means that the vapor mixture  must expand in volume. This
forces some of the fuel-air mixture to be vented out of the tank.  When the fuel later cools, the
vapor pressure of the fuel decreases, contracting the mixture, and drawing fresh air in through the
vent.  When the fuel is heated again, another cycle of diurnal emissions occurs.  It is important to
note that this is generally  not a rate-limited process. Although the evaporation of the fuel can be
slow, it is generally fast enough to maintain the fuel tank in an essentially  equilibrium  state.
                                          3-11

-------
       Consider a typical fuel use cycle beginning with a full tank. As fuel is used by the engine,
and the liquid fuel volume decreases, air is drawn into the tank to replace the volume of the fuel.
(Note: the decrease in liquid fuel could be offset to some degree by increasing fuel vapor
pressure caused by increasing fuel temperature.) This would continue while the engine was
running.  If the engine was shut off and the tank was left overnight, the vapor pressure of the fuel
would drop as the temperature of the fuel dropped. This would cause a small negative pressure
within the tank that would cause it to fill with more air until the pressure equilibrated. The next
day, the vapor pressure of the fuel would increase as the temperature of the fuel increased.  This
would cause a small positive pressure within the tank that would force a mixture of fuel vapor
and air out. In poorly designed gasoline systems, where the exhaust is very close to the fuel tank,
the fuel can actually begin to boil.  When this happens, large amounts of gasoline vapor can be
vented directly to the atmosphere.

       3.3.1.2 - Hot Soak Emissions

       Hot soak emissions occur after the engine is turned off, especially during the resulting
temperature rise.  The primary source of hot soak emissions is the  evaporation of the fuel left in
the carburetor bowl.  Other sources can include increased permeation and evaporation of fuel
from plastic or rubber fuel lines in the engine compartment.

       3.3.1.3 - Refueling Emissions

       Refueling emissions occur when the  fuel vapors are forced out when the tank is filled
with liquid fuel.  At a given temperature, refueling emissions are proportional to the volume of
the fuel dispensed into the tank. Every gallon of fuel put into the tank forces out one-gallon of
the mixture of air and fuel vapors.  Thus, refueling emissions are highest when the tank is near
empty.  Refueling emissions are also affected by the temperature of the fuel vapors. At low
temperatures, the fuel vapor content of the vapor space that is replaced is lower than it is at
higher temperatures.

       3.3.1.4 - Permeation

       The polymeric material (plastic or rubber) of which many gasoline fuel tanks and fuel
hoses generally have a chemical composition much like that of gasoline. As a result, constant
exposure of gasoline to these surfaces allows the material to continually absorb fuel.  The outer
surfaces of these materials are exposed to ambient air, so the gasoline molecules permeate
through these fuel-system components and are emitted directly into the air. Permeation
emissions continue at a nearly constant rate,  regardless of how much the vehicle  or equipment is
used.  Because of these effects, permeation-related emissions can therefore add up to a large
fraction of the total emissions from recreational boats.
                                          3-12

-------
3.3.2 - Evaporative Emission Controls

       Several emission-control technologies can be used to reduce evaporative emissions. We
expect manufacturers to use a wide variety of these potential technology approaches to meet the
proposed emission standards for marine fuel tanks.  The advantages and disadvantages of the
various possible emission-control strategies are discussed below.  Chapter 4 presents more detail
on how we expect manufacturers to use these technologies to meet proposed emission standards
for the individual applications.

      3.3.2.1 - Sealed System with Pressure Relief

      Evaporative emissions are formed when the fuel heats up, evaporates, and passes through
a vent into the atmosphere. By closing that vent, evaporative emissions are prevented from
escaping.  However, as vapor is generated, pressure builds up in fuel tank. Once the fuel cools
back down, the pressure subsides.

      For marine applications, U.S Coast Guard safety regulations require that fuel tanks be
able to withstand at least 3 psi and must be able to pass a pressure impulse test which cycles the
tank from 0 to 3 psi 25,000 times (33 CFR part 183) .  The Coast Guard also requires that these
fuel tanks must be vented such that the pressure in the tank in-use never exceeds 80 percent of
the pressure that the tank is designed to withstand without leaking. The American Boat and
Yacht Council makes the additional recommendation that the vent line should have a minimum
inner diameter of 7/16 inch (H-24.13).  However, these recommended practices also note that
"there may be EPA or state regulations that limit the discharge of hydrocarbon emissions into the
atmosphere from gasoline fuel systems. The latest version of these regulations should be
consulted."

      To prevent pressure from building too high in marine tanks, we considered a 2 psi
pressure relief valve. This is a typical automotive rating and is below the Coast  Guard
requirements. With this valve, vapors would be retained in the tank until 2 psi of pressure is
built up in the tank due to heating of the fuel. Once the tank pressure reached 2  psi, just enough
of the vapor would be vented to the atmosphere to maintain 2 psi of pressure. As the fuel cooled,
the pressure would decrease.  We estimate that this would achieve about a 55-percent reduction in
evaporative emissions over the proposed test procedure. A 1 psi valve would achieve a
reduction  of about half of this over the proposed test procedure.  In use, this reduction could be
greater because the test procedure is designed to represent a hotter than average day. On a more
mild day there could be less pressure buildup in the tank and the valve may not even need to
open.  With the use of a sealed system, a low pressure vacuum relief valve would also be
necessary so that air could be drawn into the tank to replace fuel drawn from the tank when the
engine is running.

      Manufacturers of plastic fuel tanks have expressed concern that their tanks are not
designed to operate under pressure. For instance, although they will not leak at 3 psi, rotationally
                                          3-13

-------
molded fuel tanks with large flat surfaces could begin deforming at pressures as low as 0.5 psi.
At 2.0 psi, the deformation would be greater. This deformation would affect how the tank is
mounted in the boat. Also, fuel tank manufacturers commented that some of the fittings or
valves used today may not work properly under 2 psi of pressure.  Finally, they commented that
backup pressure-relief valves would be necessary for safety.

       We believe that, with enough lead time, fuel tank manufacturers will be able to redesign
their fuel tanks to be more resistant to deformation under pressure. We also believe that if
certain fittings and valves cannot withstand pressure today, they can be designed to do so. In
addition, designing to meet a pressure of 1  psi valve would require significantly less modification
to current tanks than designing for 2 psi of pressure. Below, we discuss strategies that could be
used in conjunction with a sealed system to minimize  the build-up of pressure in the fuel tank.
Such technologies are insulation, volume-compensating air bags, and bladder fuel tanks.  With
the use of these technologies, the same emission reductions could be achieved with a pressure-
relief valve set to allow lower vent pressures. Finally  the structure of the proposed standards
gives manufacturers the flexibility to meet  the emission limits without building up pressure in the
fuel tank.

       3.3.2.2 - Limited Flow Orifice

       An alternative to using a pressure-relief valve to hold vapors in the fuel tank would be to
use a limited-flow orifice.  However, the orifice size may be so small that there would be a risk
of fouling.  In addition, an orifice designed for a maximum of 2 psi under worst-case conditions
may  not be very effective at lower temperatures. We tested a 17-gallon tank with a 75-micron
diameter limited-flow orifice over the proposed diurnal test procedure  and saw close to a 50
percent reduction in diurnal emissions.  The peak pressure in this test was 1.6 psi.

       3.3.2.3 - Insulated Fuel Tank

       Another option for reducing diurnal emissions is insulating the fuel tank. Rather than
capturing the vapors in the fuel tank, this strategy would minimize the  fuel heating which
therefore minimizes the vapor generation.  However, significant evaporative emissions would
still occur through the vent line  due to diffusion even without temperature gradients. A limited-
flow orifice  could be used to minimize the  to loss of vapor through the vent line due to diffusion.
In this case,  the orifice could be sized to prevent diffusion  losses without causing pressure build-
up in the tank. Additional control could be achieved with the use of a  pressure relief valve or a
smaller limited flow orifice.  Note that an insulated tank could maintain the same emission
control with a lower pressure valve than a tank that was not insulated.  It should be noted that
today's fuel  tanks, when installed in boats,  have some amount of "inherent insulation." This is
especially true for boats that remain in the water and is discussed in more detail in Chapter 4.

       The method of insulation would have to be consistent with U.S. Coast Guard fuel system
requirements specified in 33 CFR 183.  These requirements regulate the resistance to fuels, oils
                                           3-14

-------
and other chemicals, water adsorption, compressive strength, and density of foam used to encase
fuel tanks.  To prevent water from trapping between the fuel tank and foam and causing
corrosion in metal fuel tanks, the Coast Guard requires that the bond between the tank and the
foam be stronger than the sheer strength of the foam.  Corrosion between the tank and foam
would be hidden from view and could be undetected until the tank fails. Also, water must be
able to drain from the fuel tank when a boat is in its static floating position.

       There are several methods that could be used to insulate the fuel tank while maintaining
safe practices. For metal tanks, a honeycomb material could be used as an insulating barrier, but
this may be expensive.  The manufacturer could insulate the compartment that the tank is in
rather than the tank itself. In this case it would be important that no ignition sources were in the
compartment with the tank. In some cases, manufacturers will foam the tank in place by filling
the entire compartment the tank is in. For boats, such as yachts, that are stored in the water, the
tank can be cooled just by placing it below the waterline near the hull.

       3.3.2.4 - Volume Compensating Air Bag

       Another concept for minimizing pressure in a sealed fuel  tank is through the use of a
volume compensating air bag. The purpose of the bag is to fill up the vapor space in the fuel
tank above the fuel itself. By minimizing the vapor space, less air is available to mix with the
heated fuel and less fuel evaporates. As vapor is generated in the small vapor space, air is forced
out of the air bag, which is vented to atmosphere. Because the bag collapses as vapor is
generated, the volume of the vapor space grows and no pressure is generated.  Once the fuel tank
cools as ambient temperature goes down, the resulting vacuum in the fuel tank will open the bag
back up. Depending on the size of the bag, pressure in the tank could be minimized; therefore,
the use of a volume compensating air bag could allow a manufacturer to reduce the pressure limit
on its relief valve.

       We are still investigating materials that would be the most appropriate for the
construction of these bags. The bags would have to hold up in a fuel tank for years and resist
permeation while at the same time be light and flexible. One such material that we are
considering is fluorosilicon fiber. Also, the bag would have to be positioned so that it did not
interfere with other fuel system components such as the fuel pick-up or catch on any sharp edges
in the fuel tank.

       3.3.2.5 - Collapsible Bladder Fuel Tank

       Probably the most effective technology for reducing evaporative emissions from fuel
tanks is through the use of a collapsible fuel bladder.  In this concept, a non-permeable bladder
would be installed in the fuel tank to hold the fuel.  As fuel is drawn from the bladder, the
vacuum created collapses the bladder.  Therefore, there is no vapor space and no pressure build
up. Because the bladder would be sealed, there would be no vapors vented to the atmosphere.
We have received comments that this would be cost prohibitive because it would increase costs
                                          3-15

-------
by 30 to 100 percent depending on tank size. However, bladder fuel tanks have positive safety
implications as well and are already sold by at least one manufacturer to meet market demand in
niche applications.

       3.3.2.6 - Charcoal Canister

       The primary evaporative emission control device used in automotive applications is a
charcoal canister. With this technology, vapor generated in the tank is vented through a charcoal
canister.  The activated charcoal collects and stores the hydrocarbons. Once the engine is
running, purge air is drawn through the canister and the hydrocarbons are burned in the engine.
These charcoal canisters generally are about a liter in size and have the capacity to store three
days of vapor over the test procedure conditions. This technology does not appear to be
attractive for marine fuel tanks because boats may sit for weeks at a time without the engine
running.  Once the canister is saturated, it provides no emission control

       3.3.2.7 - Floating Fuel and Vapor Separator

       Another concept used in some stationary engine applications is a floating fuel and vapor
separator. Generally small, impermeable plastic balls are floated in the fuel tank. The purpose of
these balls is to provide a barrier between the surface of the fuel and the vapor space. However,
this strategy does not appear to be viable for marine fuel tanks.  Because of the motion of the
boat, the fuel sloshes and the barrier would be continuously broken. Even small movements in
the fuel could cause the balls to rotate and transfer fuel to the vapor space. In addition, the
unique geometry of many marine fuel tanks could case the balls to collect in one area of the tank.

       3.3.2.8 - Low-permeability Materials

       Probably the largest source of evaporative emissions is permeation through the walls of
plastic fuel tanks and rubber hoses. We estimate that about a third of the evaporative emissions
from boats with plastic fuel tanks come from permeation through the walls of the fuel tanks and
about a third through the walls of the fuel supply and fill-neck hoses.

       3.3.2.8.1 - Fuel Tank Materials

       In highway applications, non-permeable plastic fuel tanks are typically produced by blow
molding a layer of ethylene vinyl alcohol between two layers of polyethylene. However, blow
molding is expensive and requires high production volumes to be cost effective. For this reason,
this manufacturing technique is generally only used for portable fuel tanks which are generally
produced in  higher volumes.  For these tanks, however, multi-layer fuel tank construction may be
an inexpensive and effective approach to controlling permeation emissions

       Manufacturers  of rotationally molded plastic fuel tanks generally have low production
volumes and have commented that they could not produce their tanks with competitive pricing in
                                          3-16

-------
any other way.  Currently, they use cross-link polyethylene which is a low density material that
has relatively high rate of permeation.  One material that could be used as a low permeation
alternative in the rotational molding process is nylon.  The use of nylon in the construction of
these fuel  tanks would reduce permeation by more than 95 percent when compared to cross-link
polyethylene such as is used today.

       Another type of barrier technology for fuel tanks would be to treat the surfaces of a
plastic fuel tanks with fluorine.  The fluorination process causes a chemical reaction where
exposed hydrogen atoms are replaced by larger fluorine atoms which a barrier on surface of the
fuel tank.  In this process, fuel tanks are be stacked in a steel container.  The container is then be
voided of air and flooded with fluorine gas.  By pulling a vacuum in the container, the fluorine
gas is forced into every crevice in the fuel tanks. As a result of this process, both the inside and
outside surfaces of the fuel tank would be treated. As an alternative, for tanks that are blow
molded, the  inside surface of the fuel tank can be  exposed to fluorine during the blow molding
process. A similar barrier strategy is called sulfonation where sulfur trioxide is used to create the
barrier by reacting with the exposed polyethylene  to form sufonic acid groups on the surface.
Either of these processes can be used to reduce gasoline permeation by more than 95 percent.7

       About a third of fuel tanks installed in boats are made of metal, primarily aluminum.
Hydrocarbons do not permeate through metal.

       3.3.2.8.2 - Fuel Hose Materials

       The majority of fuel hoses used in boats today are made of nitrile rubber (NBR),
reinforced with fabric  and/or wire, and finished with a heat-resistant cover. The Coast Guard
requires that these hoses meet the specifications in SAE J1527 which include fire resistance,
strength, flexibility durability, and permeation requirements.8'9 The permeation requirement for
fuel hoses is 100 g/m2/day at 23°C. Because NBR has a high permeation rate (669 g-
mm/m2/day10) current  marine hoses just meet this  standard.  In contrast, materials used in current
automotive fuel lines are two to three orders of magnitude less permeable.11  By replacing rubber
hoses with low permeability hoses, evaporative emissions through the fuel and vent hoses can be
reduced by more than  95 percent.

       Low permeability hoses produced today are generally constructed in two ways: of low
permeability rubber blends or with a low permeability barrier layer. One hose design, already
used in some marine applications, uses a thermoplastic layer between two rubber layers to
control permeation.  This thermoplastic barrier may either be nylon or ethyl vinyl alcohol.  In
automotive applications, other barrier materials are used such as fluoroelastomers and
fluoroplastics such as  Teflon®. An added benefit  of low permeability lines is that some
fluoropolymers can be made to conduct electricity and therefore can prevent the buildup of static
charges.12
                                          3-17

-------
Chapter 3 References
1.  Heywood, J., "Internal Combustion Engine Fundamentals," McGraw-Hill, Inc., New York,
1988, pp.829-836.

2.  Heywood, pp.827-829.

3.  Saikalis, G., Byers, R., Nogi., T., "Study on Air Assist Fuel Injector Atomization and Effects
on Exhaust Emission Reduction," SAE Paper 930323, 1993, Air Docket A-2000-01, Document
No. II-A-55.

4.  Energy and Environmental Analysis, "Benefits and Cost of Potential Tier 2 Emission
Reduction Technologies", Final Report, November 1997, Docket A-2000-01, Document n-A-01.

5.  Southwest Research Institute, "Three-Way Catalyst Technology for Off-Road Equipment
Powered by Gasoline and LPG Engines," prepared for CARB, CEP A, and SCAQMD, (SwRI
8778), April 1999, Docket A-2000-01, Document H-A-08.

6.  Heywood, pp. 836-839.

7. Kathios, D., Ziff, R., Petrulis, A., Bonczyk, J., "Permeation of Gasoline and Gasoline-alcohol
Fuel Blends Through High-Density Polyethylene Fuel Tanks with Different Barrier
Technologies," SAE Paper 920164, 1992, Docket A-2000-01, Document No. H-A-60.

8. Title 33, Code of Federal Regulations, Part 183, sections 183.540 and 183.558.

9.  SAE Surface Vehicle Standard, "Marine Fuel Hoses," Surface Vehicle Standard, Society of
Automotive Engineers J1527, Issued 1985-12, Revised 1993-02, Docket A-2000-01, Document
No. IV-A-19.

10. Stahl, W., Stevens, R., "Fuel-Alcohol Permeation Rates of Fluoroelastomers, Fluoroplastics,
and other Fuel Resisitant Materials," SAE 920163, 1992, Docket A-2000-01, Document No. IV-
A-20.

11. Denbow, R., Browning, L., Coleman, D., "Report Submitted for WA 2-9, Evaluation of the
Costs and Capabilities of Vehicle Evaporative Emission Control Technologies," ICF, ARCADIS
Geraghty & Miller, March 22, 1999, Docket A-2000-01, Document No. IV-B-05.

12. Denbow, R., Browning, L., Coleman, D., "Report Submitted for WA 2-9, Evaluation of the
Costs and Capabilities of Vehicle Evaporative Emission Control Technologies," ICF, ARCADIS
Geraghty & Miller, March 22, 1999, Docket A-2000-01, Document No. IV-B-05.
                                        3-18

-------
                CHAPTER 4:  Technological Feasibility

       Section 213(a)(3) of the Clean Air Act presents statutory criteria that EPA must evaluate
in determining standards for nonroad engines and vehicles including marine vessels. The
standards must "achieve the greatest degree of emission reduction achievable through the
application of technology which the Administrator determines will be available for the engines or
vehicles to which such standards apply, giving appropriate consideration to the cost of applying
such technology within the period of time available to manufacturers and to noise, energy, and
safety factors associated with the application of such technology."

       We are proposing new motorcycle standards under the authority of section 202 of the
Clean Air Act. Sections 202(a) and (b) of the Act provide EPA with the general authority to
prescribe vehicle standards, subject to any specific limitations otherwise included in the Act.
Section 202(a)(l) of the Act directs us to establish standards regulating the emission of any air
pollutant from an class or classes of new motor vehicles or engines that, in the Administrator's
judgement, cause or contribute to air pollution which may reasonably be anticipated to endanger
public health or welfare. Section 202(a)(2) directs the Administrator to provide lead time
sufficient to "permit the development and application of the requisite technology, giving
appropriate consideration to the cost of compliance within such period."  Section 202(a)(3)(E)
directs the Administrator, in establishing emission standards for highway motorcycles, to
"consider the need to achieve equivalency of emission reductions between motorcycles and
other motor vehicles to the maximum extent practicable."

       This chapter presents the technical analyses and information that form the basis of EPA's
belief that the proposed emission standards are technically achievable accounting for all the
above factors.

4.1 - Highway Motorcycles

       The proposed emission standards for highway motorcycles are summarized in the
Executive Summary. As discussed in Chapter 3, we believe there are several technologies that
can be used to reduce exhaust emissions from highway motorcycles.  This section presents
certification emissions data on a range of emissions levels achieved using different technology
options.  The following sections summarize the data and rationale supporting the proposed
emission standards for highway motorcycles.

       In the development of this proposal following the publication of the ANPRM we
considered several regulatory alternatives. These included: no revision to the  standards,
harmonization with one of the "tiers" of California standards (current, 2004 Tier-1, 2008 Tier-2),
more stringent standards than those in place in California, or possibly different implementation
timing. We also considered various alternatives designed to reduce the burden on small
manufacturers (these are presented in Chapter 8 on the Regulatory Flexibility Act).
                                           4-1

-------
       After considering comments on the ANPRM, we believe that the standards should be
revised. The existing federal standards were established more than twenty years ago, and it is
clear that emission control technology has advanced a great deal in that time.  California has
continued to revise their standards to maintain some contact with current technology, and
manufacturers have generally (but not uniformly) responded by producing motorcycles for sale
nationwide that meet the more stringent California standards. Thus, in large part the existing
federal standards has been superseded because of the preponderance of manufacturers that have
responded in this way. Those arguing against new emission standards often cite the fact that
motorcycles are typically far cleaner than the existing federal standards require. Although we
agree, we see this fact as a reason for improving emission standards and as evidence that the
current federal standards are out of touch with the reality of today's technology.

       We believe it is most appropriate at this time to propose harmonizing with the California
exhaust emission standards, as opposed to other options discussed in the ANPRM. For example,
the dissimilarities between on- and off-highway motorcycles do not encourage a one-size-fits-all
approach for all motorcycles (this opinion is supported by a significant number of those who
commented on the ANPRM).  Off-highway motorcycles are powered predominantly by two-
stroke engines, whereas highway motorcycles are all powered by four-stroke engines as of the
2002 model year.  On- and off-highway motorcycle engines also lie at vastly different ends of the
size spectrum. The average highway motorcycle sold today has a displacement of nearly lOOOcc,
whereas almost 90 percent of off-highway motorcycle engines have an engine displacement  of
less than 350cc. In addition, on- and off-highway motorcycles are used in very different ways;
finding a set of standards and a test procedure that adequately represents the typical range of
operation for both types would therefore be extremely challenging. On-highway motorcycle
manufacturers have commented that, to the extent the standards are revised, harmonization with
California,  rather than a distinctly different set of standards, is preferable because it eliminates
the possibility of needing two distinct product lines for California and Federal regulations.7

       Delaying implementation of the California standards on a nationwide basis by two years
would provide an opportunity for manufacturers to gain some experience with the technology
needed to meet the new standards. Two years provides time for technology optimization and cost
reduction.  Providing a longer delay could potentially provide the option of a further decrease in
the level of the emission standards, given that the technological feasibility of the California
standards has been adequately demonstrated (at least one manufacturer is already selling a
motorcycle meeting the 2008 California standards).  However, this would be a tradeoff against a
more timely introduction of the new standards.

       We also evaluated whether the federal motorcycle program should incorporate averaging
provisions, as the California program does. Given the desire of most manufacturers to
manufacture a motorcycle for nationwide sale, such a program without averaging would not be
       7  See comments on the ANPRM from Harley-Davidson and the Motorcycle Industry
Council, available in the public docket for review.

                                           4-2

-------
desirable because it would not provide the flexibility needed to meet the California and federal
requirements together and could have at least potentially led to a somewhat less stringent Federal
standard.  Therefore, we are proposing to provide an averaging program comparable to
California's.

4.1.1 - Class I and II Motorcycles

       4.1.1.1 - Class I Motorcycles Above 50cc and Class II Motorcycles

       As noted above, we are proposing to adopt the current California standards for Class I
and Class n motorcycles.  These standards have been in place in California since 1982. The
question of whether or not these standards are technically feasible has been answered in the
affirmative, since 21 of the 22 EPA-certified 2001 model year motorcycle engine families in
these classes are already certified to these standards, and all 24 of the 2002 model year engine
families meet these standards. These 24 engine families are all powered by four-stroke engines,
with a variety of emission controls applied, including basic engine modifications on almost all
engine families, secondary air injection on three engine families, and a two-way oxidation
catalyst on one engine family.

       In past model years, but not in the 2002 model year, an engine family that does not meet
the California standards had certified to the existing federal standards and not sold in California.
It was a lOOcc dual-sport motorcycle powered by a two-stroke engine, with an HC certification
level of 3.9 g/km.  This motorcycle no longer appears to be available as of the 2002 model year.
Adopting the California standards for these motorcycle classes could preclude this motorcycle or
others like it from being certified and sold federally, unless the federal program includes
additional flexibility relative to the California program.  As discussed above, we are proposing
that the HC standard for Class I and Class n motorcycles be an averaging standard, in a departure
from California's treatment of these motorcycle classes. This in itself could be of limited use
given the low number of Class I and Class n engine families, but, as discussed in Section V.C.2
above, we are also proposing to allow credits accumulated by certifying Class in engine families
to a level lower than the standard to be used to offset Class I or Class n engine families certified
to levels above the fleet-average standard.8

       4.1.1.2 - Class I Motorcycles Under 50cc

       As we have described earlier we are proposing to apply the current California standard for
Class I motorcycles to motorcycles with displacements of less than 50cc (e.g., most motor
scooters). These motorcycles are currently not subject to regulation by the U.S. EPA or by the
       8  The manufacturer that had certified this two-stroke for highway use has typically certified 4-5 other Class
I or II engine families; therefore, a basic averaging program could enable them to continue to market their two-stroke
dual-sport. However, other manufacturers may not have adequate additional engine families in these classes, making
a basic averaging standard less useful to them.

                                            4-3

-------
State of California. They are, however, subject to emission standards in Europe and much of the
rest of the world.  Historically these motorcycles have been powered by 2-stroke engines, but a
trend appears to be developing that would result in most of these being replaced by 4-stroke
engines or possibly by advanced technology 2-stroke engines, in some cases with catalysts.

       The 4-stroke engine is capable of meeting our proposed standards.  Class I motorcycles
above 50cc are already meeting it, most of them employing nothing more than a 4-stroke engine.
For example, the existing Class I scooters certify at levels ranging from 0.4 to 0.8 grams per
kilometer HC. All of these achieve the standards with 4-stroke engine designs, and only one
incorporates additional technology (a catalyst). These engines range from 80 to 15Ice in
displacement, indicating that a smaller engine should encounter few problems meeting the
proposed standards.

       In order to meet more stringent standards being implemented worldwide, manufacturers
are developing and implementing a variety of options. Honda, perhaps the largest seller of
scooters in the U.S., has entirely eliminated 2-stroke engines from their scooter product lines as
of the 2002 model year.  They continue to offer a 50cc model, but with a 4-stroke engine.  Both
of Aprilia's 49cc scooters available in the U.S. have incorporated electronic direct injection
technology, which, in the case of one model, enables it to meet the "Euro-2" standards of 1.2
grams per kilometer HC and 0.3 grams per kilometer NOx, without use of a catalytic converter.9
Piaggio, while currently selling a 49cc basic 2-stroke scooter in the U.S., expects to begin
production of a direct injection version in 2002, and a 4-stroke 50cc scooter is also in
development. Numerous 49cc models  marketed by Piaggio in Europe are available either as a 4-
stroke or a 2-stroke with a catalyst. Piaggio, also an engine manufacturer and seller, is already
offering a 50cc 4-stroke engine to its customers for incorporation into scooters.

       The U.S. represents a very small portion of the market for small motorcycles and
scooters.  There are few, if any, manufacturers that develop a small-displacement motorcycle
exclusively for the U.S. market; the domestic sales volumes do not appear large enough at this
time to support an industry of this kind. The Italian company Piaggio (maker of the Vespa
scooters), for example, sold about as many scooters worldwide in 2000 (about 480,000) as the
entire volume of highway motorcycles  of all sizes sold in the U.S. in that year. U.S. sales of
Vespas in 2000 amounted to about 4800.  The largest scooter markets today are in South Asia
and Europe, where millions are sold annually.  In Taiwan alone almost 800,000 motorcycles were
sold domestically. More than one third of these were powered by 2-stroke engines.  Two- and
three-wheelers constitute a large portion of the transportation sector in Asia, and in some urban
areas these vehicles - many of them powered by 2-stroke engines - can approach 75 percent of
the vehicle population. According to a World Bank report, two-stroke gasoline engine vehicles
       9  Aprilia website,
http://www.apriliausa.com/ridezone/ing/models/scarabeo50dt/moto.htm.

                                           4-4

-------
are estimated to account for about 60 percent of the total vehicle fleet in South Asia.
                                                                            10
       Many nations are now realizing that the popularity of these vehicles and the high density
of these vehicles in urban areas are contributing to severe air quality problems. As a
consequence, some of the larger small motorcycle markets in Asia and India are now placing
these vehicles under fairly strict regulation. It is clear that actions in these nations will move the
emission control technology on small motorcycles, including those under 50cc, in a positive
direction. For example, according to the World Bank report, beginning in 2000 catalytic
converters will be installed in all new two-stroke engine motorcycles in India, and 2003
standards in Taiwan will effectively ban new two-strokes with emission standards so stringent
that only a four-stroke engine is capable of meeting them.

       Given the emerging  international picture regarding emission standards for scooters, we
believe that scooter manufacturers will be producing  scooters of less than 50cc displacement that
meet our proposed standards well in advance of the 2006 model year, the first year we propose to
subject this category of motorcycle to U.S. emission standards. We request comment on this
assessment.

       There are other numerous factors in the international arena that may affect the product
offerings in the less than 50cc market segment.  For example, Europe recently changed the laws
regarding insurance and helmet use for under 50cc scooters and mopeds. Previously, the
insurance discounts and lack of helmet requirements in Europe provided two relatively strong
incentives to purchasers to consider a 49cc scooter. Recently, however, the provisions were
changed such that helmets are now required and the insurance costs are comparable to larger
motorcycles.  The result was a drop of about 30% in European sales of 49cc scooters in 2001 due
to customers perceiving little benefit from a 49cc scooter relative to a larger displacement engine.

4.1.2 - Class III Motorcycles

       4.1.2.1 - Tier-1 Standards

       In the short term, the proposed Tier-1 HC+NOx standard of 1.4 g/km HC+NOx reflects
the goal of achieving emission reductions that could be met with reasonably available control
technologies, primarily involving technologies less costly and complex,  and more easily
adaptable to the vast majority of motorcycles, than catalytic converters.  As noted earlier, we are
proposing that this standard become effective starting with the 2006 model year.  Based on
current certification data, a number of existing engine families would already comply with this
standard or would need relatively simple modifications to comply. In other cases, the
       10 Improving Urban Air Quality in South Asia by Reducing Emissions from Two-Stroke
Engine Vehicles.  Masami Kojima, Carter Brandon, and Jitendra Shah. December 2000. Prepared
for the World Bank. Available in the public docket for review, or on the internet at:
http ://www. worldbank. org/html/fpd/esmap/publication/airquality .html.

                                           4-5

-------
manufacturers will need to use control technologies that are available but are not yet used on
their particular vehicles (e.g., electronic fuel injection to replace carburetors, changes to cam
lobes/timing, etc.). For the most part, manufacturers will not need to use advanced technologies
such as close-coupled, closed-loop three way catalysts.

       While manufacturers will use various means to meet the Tier-1 standard, there are four
basic types of existing, non-catalyst-based, emission control systems available to manufacturers.
The most important of these  is the use of secondary pulse-air injection.  Other engine
modifications and systems include more precise fuel control, better fuel atomization and delivery,
and reduced engine-out emission levels from engine changes. These technologies  are used in
varying degrees today and are not expected to result in a loss of performance. The combinations
of low-emission technologies ultimately chosen by motorcycle manufacturers are dependent on
the  engine-out emission levels of the vehicle, the effectiveness of the prior emission control
system, and individual manufacturer preferences.

       Secondary pulse-air injection, as demonstrated on current motorcycles, is applied using a
passive system (i.e., no air pump involved) that takes advantage of the flow of gases ("pulse") in
the  exhaust pipes to draw in fresh air that further combusts unburned hydrocarbons in the
exhaust.  Engine modifications include a variety of techniques designed to improve fuel delivery
or atomization; promote "swirl" (horizontal currents) and "tumble" (vertical currents); maintain
tight control on air-to-fuel (A/F) ratios; stabilize combustion (especially in lean A/F mixtures);
optimize valve timing; and retard ignition timing.

       4.1.2.1.1 - Secondary Air Injection

       Secondary pulse air injection involves the introduction of fresh air into the  exhaust pipe
immediately after the gases exist the engine.  The extra air causes further combustion to occur,
thereby controlling more of the hydrocarbons that escape the combustion chamber. This type of
system is relatively inexpensive and uncomplicated because it does not require an air pump; air is
drawn into the exhaust through a one-way reed valve due to the pulses of negative  pressure inside
the  exhaust pipe.  Secondary pulse-air injection is one of the most effective non-catalytic,
emissions control technologies; compared to engines without the system, reductions of 10-40%
for HC are possible with pulse-air injection. Sixty-five of the 151 2001 model year Class in
engine families certified for sale in the U.S employ secondary pulse-air injection to help meet the
current California standards. We anticipate that most of the remaining engine families will use
this technique to help meet the Tier-1 and Tier-2 standards.

       4.1.2.1.2 - Improving Fuel Delivery and Atomization

       Improving fuel delivery and atomization primarily involves the replacement of
carburetors, currently used on most motorcycles, with more precise fuel injection systems. There
are  several types of fuel  injection systems and components manufacturers can choose.  The most
likely type of fuel injection manufacturers will choose to help meet the Tier-1 standard is
                                           4-6

-------
sequential multi-point fuel injection (SFI).

       Unlike conventional multi-point fuel injection systems that deliver fuel continuously or to
paired injectors at the same time, sequential fuel injection can deliver fuel precisely when needed
by each cylinder.  With less than optimum fuel injection timing, fuel puddling and intake-
manifold wall wetting can occur, both of which hinder complete combustion.  Use of sequential-
fuel- injection systems help especially in reducing cold start emissions when fuel puddling and
wall wetting are more likely to occur and emissions are highest.

       Motorcycle manufacturers are already beginning to use sequential fuel injection (SFI).
Improved emission levels are one benefit of a fuel injection system relative to carburetion, but
other advantages include improved reliability and fuel economy. Of the 151 Class in
motorcycle engine families certified for sale this year, 27 employ SFI systems. These 27 engine
families account for about 30 percent of projected 2001 sales, indicating that these engine
families represent some popular motorcycle models. Indeed, three models that are among the
highest-selling - two from Harley-Davidson and one from Honda - are equipped with SFI. We
anticipate increased applications of this or similar fuel injection systems to achieve the more
precise fuel delivery needed to help meet the Tier-1 and Tier-2 standards.

       4.1.2.1.3 - Engine Modifications

       In addition to the techniques mentioned above, various engine modifications can be made
to improve emission levels. Emission performance can be improved, for example, by reducing
crevice volumes in the combustion chamber. Unburned fuel can be trapped momentarily in
crevice volumes before being subsequently released. Since trapped and re-released fuel can
increase engine-out emissions, the elimination  of crevice volumes would be beneficial to
emission performance.  To reduce crevice volumes, manufacturers can  evaluate the feasibility of
designing engines with  pistons that have reduced, top "land heights" (the distance between the
top of the piston and the first ring).

       Lubrication oil which leaks into the combustion chamber also has a detrimental effect on
emission performance since the heavier hydrocarbons in oil do not oxidize as readily as those in
gasoline and some components in lubricating oil may tend to foul the catalyst and reduce its
effectiveness.  Also, oil in the combustion chamber may trap HC and later release the HC
unburned. To reduce oil consumption, manufacturers can tighten the tolerances and improve the
surface finish on cylinders and pistons, piston ring design and materials, and exhaust valve  stem
seals to prevent excessive leakage of lubricating oil into the combustion chamber.

       Increasing valve overlap is another engine modification that can help reduce emissions.
This technique helps reduce NOx generation in the combustion chamber by essentially providing
passive exhaust gas recirculation (EGR).  When the engine is undergoing its pumping cycle,
small amounts of combusted gases flow past the intake valve at the start of the intake cycle. This
creates what is essentially a passive EGR flow, which is then either drawn back into the cylinder
                                           4-7

-------
or into another cylinder through the intake manifold during the intake stroke.  These combusted
gases, when combined with the fresh air/fuel mixture in the cylinder, help reduce peak
combustion temperatures and NOx levels. This technique can be effected by making changes to
cam timing and intake manifold design to optimize NOx reduction while minimizing impacts to
HC emissions.

       4.1.2.2 - Analysis of EPA Certification Data

       Secondary pulse-air injection and engine modifications already play important parts in
reducing emission levels; we expect increased uses of these techniques to help meet the Tier-1
standard. Direct evidence of the extent these technologies can help manufacturers meet the Tier-
1 standard can be found in the ARB on-road motorcycle certification database. This database is
comprised of publicly-available certification emission levels and confidential  data (e.g., projected
sales of each certified engine family) reported by the manufacturers pursuant to existing
requirements.  If one thing is clear from the EPA certification data, it is that the vast majority of
motorcycles are certifying at emission levels well below the existing federal HC standard. The
average certification HC level for 2001 Class HI motorcycles is 1.0 g/km (the  existing EPA
standard is 5.0 g/km). This is due in large part to the fact that manufacturers are designing one
engine family to market in all 50  states;  therefore, it is the California requirements that are
driving the emission levels down nationwide.  In addition, European nations and others around
the world are pursuing lower motorcycle emission levels, contributing further to the trend of
lower overall emissions from companies that want to market a worldwide product.

       The California ARB reports that discussions with manufacturers revealed that typical
NOx levels range from about 0.5  to 0.7 g/km.  Some in-use data collected by the California ARB
indicates an average NOx level for the 109 motorcycles tested of 0.53 g/km. Restricting the data
to 44 1988 and later non-tampered Class HI motorcycles results in an average  NOx level of 0.5
g/km.  Only seven out of these 44 motorcycles had NOx levels that exceeded 0.7 g/km.  In
addition, some recent data from 16 1997-1999 Class HI motorcycles tested by Environment
Canada found an average NOx level of 0.3 g/km, with a maximum of 0.5 g/km.  For the purposes
of our analysis we will the upper  end of the range reported by the California ARB, or 0.7 g/km.

       Of the 151 Class in motorcycle engine families certified for the 2001 model year, 78, or
just over 50 percent, could be certified to HC+NOx levels up to 1.6 g/km today. Although 1.6
g/km would be in excess of the standard, we believe that the reduction required to get below the
standard of 1.4 g/km HC+NOx is minimal enough such that advanced technologies like high-
efficiency two- or three-way catalysts would not be required.  Table 4.1-1 shows the breakdown
of the emission control technologies used by these 78 engine families that could potentially
certify to the Tier-1  standards today. Only seven of these, or less than ten percent, use 3-way
catalysts.  The remaining 74 could be able to certify near the proposed Tier-1 HC+NOx level by
using simpler and less costly engine modifications and secondary air injection. Only 16 of these
74 use a two-way catalyst.
                                           4-8

-------
                                       Table 4.1-1
        Breakdown of HC+NOx Estimated Certification Levels by Technology Use

Estimated HC+NOx
Certification Level
0.9- 1.2
1.3- 1.4
1.5- 1.6
Total
No. of
Engine
Families
26
31
21
78
No. of Engine Families Using Specified Technology
Engine
ModificationsA
18
32
17
67
Pulse Air
Injection
17
10
7
34
2-way Ox-
Cat.
7
6
3
16
3 -way
Catalyst
2
2
0
7
A Includes all forms of fuel injection, electronic control modules, etc.
Source: 2001 U.S. EPA Certification Database
       In addition, we are confident that the two years of experience gained by manufacturers in
meeting the new standard in California prior to having to extend compliance nationwide will help
ensure that the new emission control systems are fully developed and fully capable of meeting
the new standards on a nationwide basis.

       4.1.2.3 - Tier-2 Standards

       In the long term, the proposed Tier-2 HC+NOx standard of 0.8 g/km will ensure that
manufacturers will continue to advance the status of control technologies. We are proposing the
Tier-2 standard to be effective by the 2010 model year. This standard will present some
challenges for manufacturers; however, several manufacturers are already using some of the
technologies that will be needed to meet this standard. In addition, our proposed implementation
time frame gives manufacturers two years of experience  in meeting this standard in California
before having to meet it on a nationwide basis. At least one manufacturer already uses closed-
loop, three-way catalysts on several of its product lines, and another is already marketing a large
touring motorcycle that meets this standard. Depending  upon assumptions regarding NOx levels,
other manufacturers have products on the market today with emission levels that could meet or
almost meet the Tier-2 standards using two-way catalysts, fuel injection, secondary pulse-air
injection, and other engine modifications.  We expect that the significant lead time prior to
meeting these standards on a nationwide basis will allow manufacturers to optimize these and
other technologies to meet the Tier-2 standard.

       To meet the Tier-2 standard for HC+NOx, manufacturers will likely use more advanced
engine modifications and secondary air injection. Specifically, we believe manufacturers will
use computer-controlled secondary pulse-air injection (i.e., the injection valve would be
connected to a computer-controlled solenoid).  In addition to these systems, manufacturers will
                                           4-9

-------
probably need to use catalytic converters on some motorcycles to meet the Tier-2 standards.
There are two types of catalytic converters currently in use: two-way catalysts (which control
only HC and CO) and three-way catalysts (which control HC, CO, and NOx). Under the Tier-2
standard, manufacturers will need to minimize levels of both HC and NOx.  Therefore, to the
extent catalysts are used, manufacturers will likely use a three-way catalyst in addition to engine
modifications and computer-controlled, secondary pulse-air injection.  These types of
technologies are used in varying degrees on current models and are not expected to result in a
loss of performance.

       4.1.2.4.1 - Improving Fuel Control and Delivery

       As discussed previously, improving fuel control and delivery provides emission benefits
by helping to reduce engine-out emissions and minimizing the exhaust variability which the
catalytic converter experiences.  One method for improving fuel control is to provide enhanced
feedback to the computer-controlled fuel injection system through the use of heated oxygen
sensors.  Heated oxygen sensors (HO2S) are located in the exhaust manifold to monitor the
amount of oxygen in the exhaust stream and provide feedback to the electronic control module
(ECM). These  sensors allow the fuel control system to maintain a tighter band around the
stoichiometric A/F ratio than conventional O2 sensors.  In this way, HO2S assist vehicles in
achieving precise control of the A/F ratio and thereby enhance the overall emissions performance
of the engine. At least one manufacturer is currently using this technology on several 2001
engine families.

       In order to further improve fuel control, some motorcycles with electronic controls may
utilize software algorithms to perform individual cylinder fuel control. While dual oxygen
sensor systems are capable of maintaining A/F ratios within a narrow range, some manufacturers
may desire even more precise control to meet their performance needs. On typical applications,
fuel control is modified whenever the O2S determines that the combined A/F of all cylinders in
the engine or engine bank is "too far" from stoichiometric. The  needed fuel modifications (i.e.,
inject more or less fuel) are then applied to all  cylinders simultaneously.  Although this fuel
control method will maintain the "bulk" A/F for the entire engine or engine bank around
stoichiometric,  it would not be capable of correcting for individual cylinder A/F deviations that
can result from  differences in manufacturing tolerances, wear of injectors, or other factors.

       With individual cylinder fuel control, A/F variation among cylinders will be diminished,
thereby  further improving the effectiveness of the emission controls. By modeling the behavior
of the exhaust gases in the exhaust manifold and using software  algorithms to predict individual
cylinder A/F, a  feedback fuel control system for individual cylinders can be developed. Except
for the replacement of the conventional front O2S with  an HO2S sensor and a more powerful
engine control computer, no additional hardware is needed in order to achieve individual cylinder
fuel control.  Software changes and the use of mathematical models of exhaust gas mixing
behavior are required to perform this operation.
                                          4-10

-------
       In order to maintain good driveability, responsive performance, and optimum emission
control, fluctuations of the A/F must remain small under all driving conditions including
transient operation.  Virtually all current fuel systems in automobiles incorporate an adaptive fuel
control system that automatically adjusts the system for component wear, varying environmental
conditions, varying fuel composition, etc., to more closely maintain proper fuel control under
various operating conditions.  For some fuel control systems today, this adaptation process
affects only steady-state operating conditions (i.e., constant or slowly changing throttle
conditions). However, most vehicles are now being introduced with adaptation during
"transient"  conditions (e.g., rapidly changing throttle, purging of the evaporative system).

       Accurate fuel control during transient driving conditions has traditionally been difficult
because of the inaccuracies in predicting the air and fuel flow under rapidly changing throttle
conditions. Because of air and fuel dynamics  (fuel evaporation in the intake manifold and air
flow behavior) and the time delay between the air flow measurement and the injection of the
calculated fuel mass, temporarily lean A/F ratios can occur during transient driving conditions
that can cause engine hesitation, poor driveability and primarily an increase in NOx emissions.
However, by utilizing fuel and air mass modeling, vehicles with adaptive transient fuel control
are more capable of maintaining accurate, precise fuel control under all operating conditions.
Virtually all cars sold in California will incorporate adaptive transient fuel control software;
motorcycles with computer controlled fuel injection can also  benefit from this technique at a
relatively low cost.

       4.1.2.4.2 - Three-way Catalytic Converters

       Three-way catalytic converters traditionally utilize rhodium and platinum as the catalytic
material to control the emissions of all three major pollutants (hydrocarbons (HC), CO, NOx).
Although this type of catalyst is very effective at converting exhaust pollutants, rhodium, which
is primarily used to convert NOx, tends to thermally deteriorate at temperatures significantly
lower than platinum. Recent advances in palladium and tri-metal (i.e.,  palladium-platinum-
rhodium) catalyst technology, however, have improved both the light-off performance (light-off
is defined as the catalyst bed temperature where pollutant conversion reaches 50% efficiency)
and high temperature durability over previous catalysts. In addition, other refinements to catalyst
technology, such as higher cell density substrates and adding  a second layer of catalyst washcoat
to the substrate (dual-layered washcoats), have further improved catalyst performance from just a
few years ago.

       Typical cell densities for conventional catalysts used  in motorcycles are less than 300
cells per square inch (cpsi). To meet the Tier-2 standard, we  expect manufacturers to use
catalysts with cell densities of 300 to 400 cpsi. If catalyst volume is maintained at the same level
(we assume volumes of up to 60% of engine displacement), using a higher density catalyst
effectively  increases the amount of surface area available for  reacting with pollutants.  Catalyst
manufacturers have been  able to increase cell  density by using thinner walls between each cell
without increasing thermal mass (and detrimentally affecting catalyst light-off) or sacrificing
                                           4-11

-------
durability and performance.

       In addition to increasing catalyst volume and cell density, we believe that increased
catalyst loading and improved catalyst washcoats will help manufacturers meet the Tier-2
standard. In general, increased precious metal loading (up to a certain point) will reduce exhaust
emissions because it increases the opportunities for pollutants to be converted to harmless
constituents.  The extent to which precious metal loading is increased will be dependent upon the
precious metals used and other catalyst design parameters. We believe recent developments in
palladium/rhodium catalysts are very promising since rhodium is very efficient at converting
NOx, and catalyst suppliers have been investigating methods to increase the amount of rhodium
in catalysts for improved NOx conversion.

       Double layer technologies allow optimization of each individual precious metal used in
the washcoat. This technology can provide reduction of undesired metal-metal or metal-base
oxide interactions while allowing desirable interactions. Industry studies have shown that
durability and pollutant conversion efficiencies are enhanced with double layer washcoats.  These
recent improvements in catalysts  can help manufacturers meet the Tier-2 standard at reduced cost
relative to older three-way catalysts.

       New washcoat formulations are now thermally stable up to 1050 ° C. This is a significant
improvement from conventional washcoats,  which are stable only up to about 900 °C.  With the
improvements in light-off capability, catalysts may not need to be placed as close to the engine as
previously thought.  However, if placement closer to the engine is required for better emission
performance, improved catalysts based on the enhancements described above would be more
capable of surviving the higher temperature environment without deteriorating. The improved
resistance to thermal degradation will allow  closer placement to the engines where feasible,
thereby providing more heat to the catalyst and allowing them to become effective quickly.

       It is well established that a warmed-up catalyst is very effective at converting exhaust
pollutants.  Recent tests on advanced catalyst systems in automobiles have shown that over 90%
of emissions during the Federal Test Procedure (FTP) are now emitted during the first two
minutes of testing after engine start up.  Similarly, the highest emissions from a motorcycle occur
shortly after start up. Although improvements in catalyst technology have helped reduce catalyst
light-off times, there are several methods to provide additional heat to the catalyst.  Retarding the
ignition spark timing and computer-controlled, secondary air injection have been shown to
increase the heat provided to the catalyst, thereby improving its cold-start effectiveness.

       In addition to using computer-controlled secondary air injection and retarded spark timing
to increase the heat provided to the catalyst,  some vehicles may employ warm-up, pre-catalysts to
reduce the size of their main catalytic converters.  Palladium-only warm-up catalysts (also known
as "pipe catalysts" or "Hot Tubes") using ceramic or metallic substrates may be added to further
decrease warm-up times and improve emission performance.  Although metallic substrates are
usually more expensive than ceramic substrates, some manufacturers and suppliers believe
                                          4-12

-------
metallic substrates may require less precious metal loading than ceramic substrates due to the
reduced light-off times they provide.

       Improving insulation of the exhaust system is another method of furnishing heat to the
catalyst.  Similar to close-coupled catalysts,  the principle behind insulating the exhaust system is
to conserve the heat generated in the engine  for aiding catalyst warm-up. Through the use of
laminated thin-wall exhaust pipes, less heat  will be lost in the exhaust system, enabling quicker
catalyst light-off. As an added benefit, the use of insulated exhaust pipes will also reduce
exhaust noise. Increasing numbers of manufacturers are expected to utilize air-gap exhaust
manifolds (i.e., manifolds with metal inner and outer walls and an insulating layer of air
sandwiched between them) for further heat conservation.

       4.1.2.4.3 -Automotive Technologies

       Besides the hardware modifications described above, motorcycle manufacturers may
borrow from other current automobile techniques. These include using engine calibration
changes such as  a brief period of substantial ignition retard, increased cold idling speed, and
leaner air-fuel mixtures to quickly provide heat to a catalyst after cold-starts. Only software
modifications are required for an engine which already uses a computer to control the fuel
delivery and other engine systems. For these engines, calibration modifications provide
manufacturers with an inexpensive method to quickly achieve light-off of catalytic converters.
When combined with pre-catalysts, computer-controlled secondary air injection, and the other
heat conservation techniques described above, engine calibration techniques may be very
effective at providing the required heat to the catalyst for achieving the Tier-2 standard.  These
techniques are currently in use on most low  emission vehicle (LEV) automobiles and may have
applications in on-road motorcycles.

4.1.3 - Impacts on Noise, Energy, and Safety

       The Clean Air Act directs us to consider potential impacts on noise, energy, and  safety
when establishing the feasibility of emission standards for nonroad engines.

       As automotive technology demonstrates, achieving low emissions from spark-ignition
engines can correspond with greatly  reduced noise levels.  Virtually all highway motorcycles are
equipped with sound suppression systems or mufflers. The four-stroke engines used in highway
motorcycles are considerably more quiet than the  two-stroke engines used by many of their off-
road counterparts.  In addition, highway motorcycles are required to meet existing noise emission
standards.

       Adopting new technologies for controlling fuel metering and air-fuel mixing, particularly
the conversion of some carbureted highway  motorcycles to advanced fuel injection technologies,
will lead to improvements in fuel consumption.
                                          4-13

-------
       Many riders have expressed some concerns regarding the close proximity of the riders to
hot exhaust pipes and the catalytic converter. Protecting the rider from the excessive heat is a
concern for both riders and manufacturers.  The current use of catalytic converters on a number
of motorcycles (accounting for tens of thousands in the current fleet) already indicates that these
issues are not insurmountable on a variety of motorcycle styles and engine sizes. A number of
approaches to shielding the rider from the heat of the catalytic converter are possible, such as
exterior pipe covers,  shielded foot rests, and similar components. Some manufacturers have
found that placing the converter on the underside of the engine can  keep it adequately distant
from the rider. Others may use double-pipe systems that reduce overall heat loss while
remaining cooler on the exterior. Based on the significant lead time that would be allowed for
meeting these standards, as well as on the two years of prior experience in California before
meeting the requirements federally, we believe that these issues can be satisfactorily resolved for
the proportion of motorcycles for which catalytic converters will be required.

       We believe the technology discussed here would have no negative impacts on safety.
Highway motorcycles have been predominantly powered by four-stroke engines for more than
twenty years. Catalytic converters, secondary air injection,  and fuel injection technologies have
been increasingly used on highway motorcycles for a number of years, without any known
adverse safety impacts.

4.1.4 - Conclusion

       4.1.4.1 - Tier-1 Standards

       We expect that the Tier-1 standard will be met with reasonably available control
technologies. A number of existing engine families would already comply with this standard or
would need relatively simple modifications to comply.  In other cases, the manufacturers will
need to use control technologies that are available but are not yet used on their particular vehicles
(e.g., electronic fuel injection to replace carburetors, secondary pulse air injection, changes to
cam lobes/timing, etc.).  For the most part, manufacturers will not need to use advanced
technologies such as  close-coupled, closed-loop three way catalysts. The ultimate combinations
of low-emission technologies ultimately chosen by motorcycle manufacturers are dependent on
the engine-out emission levels of the vehicle, the effectiveness of the prior emission control
system, and individual manufacturer preferences.

       4.1.4.2 - Tier-2 Standards

       We expect that the Tier-2 standard will likely require the use of more  advanced engine
modifications and secondary air injection. Depending upon assumptions regarding NOx levels,
some manufacturers have products on the market today with emission levels that could meet or
almost meet the Tier-2 standard using two-way catalysts, fuel injection, secondary pulse-air
injection, and other engine modifications. We believe that manufacturers will use computer-
controlled secondary pulse-air injection, in addition to using catalytic converters on some
                                          4-14

-------
motorcycles to meet the Tier-2 standards.  To the extent catalysts are used, manufacturers will
likely use a three-way catalyst in addition to engine modifications and computer-controlled,
secondary pulse-air injection.  We expect that the significant lead time prior to meeting these
standards on a nationwide basis will allow manufacturers to optimize these and other
technologies to meet the Tier-2 standard.

4.2 - Evaporative Emission Control from Boats

       The proposed emission standards for evaporative controls from marine vessels are
summarized in the Executive Summary. As discussed in Chapter 3, we believe there are several
technologies that can be used to reduce evaporative emissions from marine vessels as needed to
meet the proposed standards.  This section presents available emissions data on baseline
emissions and on emission reductions achieved through the application of emission control
technology.  In addition, this section provides a description of the proposed test procedures for
evaporative  emission determination.

4.2.1 - Diurnal Evaporative Emissions

       We are engaged in a test program to characterize baseline emissions from marine fuel
tanks and to evaluate several emission reduction strategies. Although this test program is not
complete, we present the initial results below.  As we continue to collect test data, we will refine
our proposed standard bins and our designs for the proposed design-based certification approach.
This section first presents baseline emissions, then discusses several emission control strategies.

       Figure 4.2-1 presents the fuel tanks that have been used for EPA testing of diurnal and
permeation emissions.  The plastic fuel tanks and one of the aluminum fuel tanks were produced
for sale, while the other aluminum fuel tank was build by a tank manufacturer for this testing.
However, this aluminum tank is in a configuration that is representative of fuel tanks used on
marine vessels.  One of the plastic tanks was a portable fuel tank while the other two were
intended to be permanently installed into vessels.  In addition, a bladder fuel tank was assembled
for this testing.
                                          4-15

-------
                     Figure 4.2-1:  Fuel Tanks Used In EPA Testing
         Portable Plastic
                                                    Installed Plastic #1
             Aluminum #1
                                                        Installed Plastic #2
               Aluminum #2
Bladder Tank
       4.2.1.1 - Baseline Emissions

       We tested three marine fuel tanks in their baseline configurations over the proposed
certification test procedure for evaporative emissions in our laboratory. These fuel tanks
included a portable plastic fuel tank and two aluminum fuel tanks.  Aluminum tank #1 was
constructed for this testing, but is representative of a typical fuel tank; aluminum tank # 2 was
removed from an 18 foot runabout.  The portable fuel tank was tested with its vent cracked open
and the aluminum fuel tanks were tested with the venting through a length of 5/8 inch hose. The
advantage of using the aluminum fuel tanks for this testing  was to exclude permeation emissions
from the measured results.

       As described later in Section 4.2.2, the test procedure involves a 24 hour diurnal from 72-
                                          4-16

-------
96°F (22.2-35.6°C) with a fuel tank filled to 40% of capacity with 9RVPk test fuel.  Under these
conditions, the theoretical Wade equations (See Chapter 6) predict an emission level of about 2.3
g/gallon/day.  However, this equation is known to over predict diurnal vapor generation and is
generally corrected.  Although the portable fuel tank was consistent with the Wade equations, we
believe it was due to diffusion of vapor through the vent. This diffusion effect is discussed in
more detail below. Because the plastic fuel tank was new at the time of testing, the material was
not likely saturated with fuel; therefore the permeation component of the measured emissions
should have been small.

          Table 4.2-1:  Baseline Diurnal Evaporative Emission Results (72-96°F)
Tank Type
Portable, polyethylene
Aluminum #1
Aluminum #2
Manu-
facturer
Moeller
Ezell
AFP
Vent Hose
Length [cm]
none
68
137
Fuel Capacity
[gallons]
6
17
30
Evaporative HC
[g/gallon/day]
2.00*
1.39
1.50
   tested with 50% fill (if adjusted to 40% fill using the Wade equations, would be about 2.3 g/gal/day)
       We also tested aluminum tank #2 over three diurnal temperature swings that differ from
the proposed test procedure.  The temperature profiles used in this testing were based fuel
temperatures measured with the whole boat in the SHED over the proposed test procedure.
These profiles are discussed further below in the section on insulated fuel tanks. This tank was
tested at 40% fill with 9RVP fuel.  Table 4.2-2 presents the results compared to the Wade model.
Although the Wade model over predicts the vapor generation, it does show a similar trend with
respect to temperature.

     Table 4.2-2:  Baseline Diurnal Evaporative Emission Results (varied temperature)
Temperatures
24-33°C (74-91°F)
22 - 30°C (71 - 86°F)
25-3TC (77-88°F)
Evaporative HC
[g/gallon/day]
1.13
0.88
0.66
Wade HC
[g/gallon/day]
1.33
1.02
0.88
       k Reid Vapor Pressure (psi). This is a measure of the volatility of the fuel.  9 RVP
represents a typical summertime fuel in northern states.
                                          4-17

-------
       4.2.1.2 - Diffusion Effect

       In testing diurnal emissions from fuel tanks with open vents, the configuration of the vent
can have a significant effect on the measured emissions. This is due to diffusion of vapor out of
the vent line. Depending on the size and configuration of the vent, diffusion can actually occur
when the fuel temperature is cooling. To quantify this diffusion component for a typical fuel
tank we ran the proposed diurnal test procedure for aluminum tank #1 using four configurations
for venting The first was with the fuel cap cracked open and the vent sealed, the second was
with a 68 cm length of vent hose, and the third was with a 1000 micron limiting flow orifice in
the vent opening.  This 1000 micron orifice was large enough to allow venting without any
measurable pressure increase in the fuel tank during the diurnal test. The fourth configuration
was a combination of the limited flow orifice and the vent hose. Table 4.2-3 presents the results
of this testing.
          Table 4.2-3: Diurnal Test Results with Varied Venting Configurations
Vent Configuration
cracked fuel cap
68 cm of 5/8" fuel hose
1000 micron orifice
1000
micron orifice + 68 cm of 5/8" fuel hose
Evaporative HC [g/gallon/day]
2.05
1.39
1.47
1.34
       The above data suggest that, at least for open vent fuel systems, the size and configuration
of the venting system can have a significant effect on evaporative emissions.  Therefore, the
proposed test procedures require that the fuel tank be set up with a vent hose attached.  This data
suggests that diffusion emissions are minimal if the fuel tank is vented through a length of hose;
therefore, we use the data on testing with a vent hose for our emission modeling (see Chapter 6).
To further investigate this diffusion effect, we tested aluminum tank #1 with several venting
configuration, at two constant temperature settings.  Under these conditions, all of the measure
evaporative emissions should be due to diffusion. As seen in Table 4.2-4, diffusion can be very
high with too large of a vent opening unless a vent hose is used. The two lengths of vent hose
tested did not show a significant difference in diffusion emissions We believe that the vent hose
limits diffusion by creating a gradual gradient in fuel vapor concentration.
                                          4-18

-------
          Table 4.2-4: Diffusion Test Results with Varied Venting Configurations
Vent Configuration
1/2M vent opening
68 cm of 5/8" fuel hose
137 cm of 5/8" fuel hose
1000 micron orifice
22 °C (72°F)
Evaporative HC [g/gal/day]
5.65
0.11
0.07
0.28
36°C (96°F)
Evaporative HC [g/gal/day]
10.0
0.18
0.24
0.41
       4.2.1.3 - Sealed System with Pressure Relief

       By sealing a fuel tank, we can capture all of the evaporative emissions. However, this
could cause large pressure build up in the tank. To control the amount of pressure built up into
the tank, we looked at two types of pressure relief strategies: pressure relief valves and limited
flow orifices. Because the Coast Guard requires that fuel systems not exceed 80 percent of their
design capacity of 3 psi, we only looked at pressure relief strategies that would keep the pressure
below 2.4 psi under worst case conditions.

       For the pressure relief valve testing, we looked at several pressures ranging from 0.5 to
2.25 psi. The 2.25  psi valve was an off-the-shelf automotive fuel cap with a nominal 2 psi
pressure relief valve and 0.5 psi vacuum relief valve. For the other pressure settings, we used
another automotive cap modified to allow adjustments to the spring tension in the pressure relief
valve.  We  performed these tests on the aluminum fuel tank to remove the variable of
permeation. As shown in Figure 4.2-2, there was a fairly linear relationship between the pressure
setting of the valve and the emissions measured over the proposed test procedure. At 1 psi, we
believe a level of 1.1 g/gallon/day  could be achieved.
                                          4-19

-------
                Figure 4.2-2: Effect of Pressure Cap on Diurnal Emissions
           0
0.5             1             1.5
        pressure relief setting [psi]
2.5
       Another strategy for maintaining a design pressure is to use a limited flow orifice on the
vent.  In our testing, we are looked at three orifice sizes: 25, 75, and 1,000 microns in diameter.
Again, we are performed these tests, using the proposed test procedures, on aluminum tank #1 to
remove the variable of permeation.  To get these exact orifice sizes, we ordered from a company
that specializes in boring holes with a laser device; however, these orifices were relatively
inexpensive. It should be noted that a smaller tank would need a smaller orifice and a larger tank
could use a larger orifice to build up the same pressure in the tank. The test results are presented
in Table 4.2-5. For all of the tests with the limited flow orifices, no vent hose was attached.

         Table 4.2-5: Diurnal Evaporative Emissions with Limited Flow Orifices
Orifice Diameter (microns)
baseline (open vent with hose)
1000
75
25
Peak Pressure [psi]
0.0
0.0
1.6
3.1
Evaporative HC [g/gallon/day]
1.39
1.47
1.16
0.24
       By limiting the flow of the vapor from the tank, emissions were reduced with some
pressure build up in the tank. However, because the vapor is flowing from the tank even at low
pressure, this strategy is less effective for large emission reductions than a pressure relief valve
                                          4-20

-------
for a given tank pressure. In addition, the limited flow orifice would have to be sized for worst
case conditions to prevent the tank from reaching too high of a pressure. This would limit the
effectiveness of this strategy under typical conditions.

       4.2.1.4 - Insulated Fuel Tank

       By insulating the fuel tank, we can limit the temperature variation that the fuel in the tank
is exposed to.  By reducing the temperature variation, less vapor is formed due to fuel heating. In
our preliminary testing, we insulated the flat, rotationally molded, fuel tank (plastic tank #2, 23
gallon capacity) using 3 inch thick construction foam with an R-value of 15  as defined by 16
CFR 460.5.  This testing was performed with the fuel tank vent open to atmosphere.  Table 4.2-6
presents the fuel  temperatures and evaporative emissions over the three day test.

       Consistent with the proposed test procedures, we tested this fuel tank over three  diurnals
and used the highest grams of the three 24 hour diurnals. This experiment resulted in a  50%
reduction in emissions from baseline on the highest of these three days. The baseline emissions
were measured to be 2.5 g/gallon/day; however it should be noted that for both the baseline test
and the insulated tank tests we did not control for permeation or diffusion.  Over this test, the
emissions decreased for subsequent days.  We believe this was due to the fuel  temperature cycle
stabilizing.

        Table 4.2-6: Evaporative Emission Results for Insulated Flat, Plastic Tank
Test Day
Day#l
Day #2
Day #3
SHED Temperature
22-36°C (72-96°F)
22-36°C (72-96°F)
22-36°C (72-96°F)
Fuel Temperature
22-28°C (72-82°F)
26-30°C (78-86°F)
26-30°C (80-86°F)
Evaporative HC
1.2g/gal/day
1.0 g/gal/day
0.8 g/gal/day
       These results from this preliminary experiment are encouraging; however, we believe that
it will be necessary to find a more appropriate insulation for this application. The construction
foam is not designed to be durable when exposed to gasoline. Also, a thinner insulation would
take up less space in a boat.

       There is a certain amount of inherent insulation for a fuel tank installed in a boat. This is
especially true for a boat that is stored in the water.  The water acts as a cooling medium for the
fuel tank, especially if it is installed in the bottom of the fuel tank.  In addition, the thermal
inertia of the fuel in the tank can act to dampen temperature variation imposed from the diurnal
heating of the ambient air. The following charts present fuel and ambient temperature test data
for two boats on trailers and two boats in the water.
                                          4-21

-------
       Figure 4.2-3:  Temperature Trace for Personal Watercraft on Trailer
a
    12:OOAM    12:OOAM    12:OOAM    12:OOAM   12:OOAM   12:OOAM   12:OOAM   12:OOAM
            Figure 4.2-4:  Temperature Trace for Runabout on Trailer
     12:00 AM
               12:00 PM
                         12:00 AM
                                    12:00 PM
                                              12:00 AM
                                                        12:00 PM
                                                                   12:00 AM
                                                                             12:00 PM
                                      4-22

-------
               Figure 4.2-5: Temperature Trace for Runabout in Water
                                 Fuel Temp
                                 Ambient Temp
                                 Water Temp
  12:OOAM    12:OOAM   12:OOAM    12:OOAM   12:OOAM    12:OOAM   12:OOAM
                                      12:00 AM
                Figure 4.2-6:  Temperature Trace of Deckboat in Water
1
o>
a.
E
0)
Fuel Temp
Ambient Temp
Water Temp
   12:OOAM   12:OOAM  12:OOAM   12:OOAM   12:OOAM   12:OOAM  12:OOAM   12:OOAM   12:OOAM
                                         4-23

-------
       4.2.1.5 - Volume Compensating Air Bag

       Another concept for minimizing pressure in a sealed fuel tank is through the use of a
volume compensating air bag. The purpose of the bag is to fill up the vapor space in the fuel
tank above the fuel itself. By minimizing the vapor space, less air is available to mix with the
heated fuel and less fuel evaporates.  As vapor is generated in the small vapor space, air is forced
out of the air bag, which is vented to atmosphere.  Because the bag collapses as vapor is
generated, the volume of the vapor space grows and no pressure is generated.1 Once the fuel tank
cools as ambient temperature goes down, the resulting vacuum in the fuel tank will open the bag
back up.

       We tested the portable plastic fuel tank with a 1.5 gallon volume compensating bag made
out of tedlar. Tedlar is a light, flexible, clear plastic which we use in our labs for collecting
exhaust emissions samples. In our testing, the pressure relief valve never opened because the
volume compensating bag was able to hold the vapor pressure below 0.8 psi for each of the three
days. This testing supports the theory that a volume compensating bag can be used to minimize
pressure in a fuel tank, which in turn, reduces emissions when used in  conjunction with a
pressure relief valve.

       We did see  an emission rate of about 0.4 g/gal/day over the 3 day test. The emission rate
was fairly constant, even when the ambient temperature was cooling during the test.  This
suggests that the emissions measured were likely permeation through tank. If there were  a leak,
or if the permeation were through the bag,  no hydrocarbons would have been measured during
the cooling periods.

       We are still investigating materials that would be the most appropriate for the
construction of these bags. The bags would have to hold up in a fuel tank for years and resist
permeation while at the same time be light and flexible. One such material that we are
considering is fluorosilicon fiber.

       4.2.1.6 - Bladder Fuel Tank

       A bladder tank is a is installed in the fuel tank to hold the fuel.  Because this bladder is
collapsible, there is no space  above the fuel where vapor can form; therefore, there are essentially
no diurnal emissions. During refueling, because the bladder is collapsed when empty, very little
vapor is displaced into the  atmosphere.  At least one manufacturer is currently manufacturing
bladder fuel  tanks for use in marine applications and information on this system is available in
the docket.1  In addition, because the bladder can be made of a low permeability material,
permeation emissions from a bladder tank  are low compared to  a plastic tank. We tested  a
marine bladder fuel tank in our lab for both diurnal and permeation emissions.  Over the diurnal
       1 The Ideal Gas Law states that pressure and volume are inversely related. By increasing
the volume of the vapor space, the pressure can be held constant.

                                          4-24

-------
test procedure we saw an emission rate of 0.2 g/gal/day.  Based on the results of our permeation
testing (see 4.2.2.1), this measured emission rate was likely due to permeation through the
bladder and not due to diurnal losses. The manufacturer of this bladder tank is now working with
a barrier film that will reduce permeation emissions by more than 90 percent (see 4.2.2.1).

       4.2.1.7 - Fuel and Vapor Separator

       Another technology that we intend to test is a floating fuel and vapor separator. We plan
to test a fuel tank with non-permeable plastic balls covering the surface of the fuel. This strategy
is used in some stationary applications and may show good results in a SHED.  However, we are
concerned that this technology may not be applicable to marine applications where the fuel may
slosh past the floating plastic balls.

4.2.2 - Permeation Evaporative Emissions

       As discussed in Chapter 3, permeation from plastic fuel tanks and rubber hoses makes up
a large fraction of the evaporative emissions from marine fuel systems. This section discusses
current emission rates from these sources as well as potential permeation control technology.

       4.2.2.1 - Fuel Tanks

Baseline emissions

       We are investigating the permeation rates of fuel through the walls of polyethylene fuel
tanks. Initial indications are that polyethylene fuel tanks have very high permeation rates
compared to those used in automotive applications.  The Coast Guard tested three rotationally
molded plastic tanks at 40°C (104°F) for 30 days.2 The results are presented in Table 4.2-7.
Because permeation emissions are a function of surface area and wall thickness, there was some
variation in the permeation rates from the three tanks on a g/gal/day basis.

           Table 4.2-7: Permeation Rates for Plastic Marine Fuel Tanks at 40°C
Tank Capacity
[gallons]
12
18
18
Permeation Loss
[g/gal/day]
1.48
1.39
1.12
Average Wall
Thickness [mm]
5.3
5.6
6.9
       We also tested four marine fuel tanks in our lab for permeation.  They included the two
cross-link fuel tanks and the portable fuel tank in Figure 4.2-1 as well as a bladder fuel tank.
This test was performed over 30 days at 29°C (85°F) with gasoline. Prior to testing, the fuel
                                          4-25

-------
tanks had been stored with fuel in them for more than a month to stabilize the permeation rate.
The permeation rates are presented in Figure 4.2-7.  The flat plastic fuel tank showed a g/gal/day
permeation rate that is consistent with the Coast Guard data in Table 4.2-7 (once corrected for
temperature). However, the V-shaped tank showed about half of the permeation.  This was likely
due to the difference in geometry, primarily the thickness of the tank. Doubling the thickness of
the tank would half the permeation rate.  The portable fuel tank showed similar permeation rate
as the flat cross-link tank.  Although it is constructed of high-density polyethylene which has
better resistance to fuel, the walls are considerably thinner because it is a smaller tank. For the
same reason, the bladder tank showed g/gal/day permeation rates close to the plastic tanks.
Although the bladder is constructed out of an aromatic polyester polyurethane sheet with a
permeation which is considerably more resistant to fuel than polyethylene, the bladder thickness
is less than a tenth  of the thickness of the standard fuel tank.
     Figure 4.2-7: Permeation Test Results on Four Marine Fuel Tanks at 29°C (85°F)
OK nn

on nn
c
— 1 K nn
re
•?
"w
E m nn
5
O)
K nn
Onn r
0


• portable, 0.61 g/gal/day
• flat, 0.64 g/gal/day
A vee 0 31 g/gal/day
• bladder, 0.46 g/gal/day

1
S«
.
• ' ! ^
0 5.0 10.0



. J
:• , •
• • •
fi' A *
• •'
AA A
15.0 20.0 25.0 3C
days








).0
Permeation barrier technology

       We are looking into low permeability tank materials.  We are already aware that plastic
automotive fuel tanks are capable achieving low permeation rates. This emission control is
accomplished by forming multiple layers in the fuel tank through the blow molding process.
Generally a low permeation barrier, such as nylon or ethyl vinyl alcohol, is sandwiched between
layers of polyethylene.  Another emission control strategy for blow molded tanks is the use of
laminar barriers. In this technology, a polar polymer is mixed with the polyethylene in the blow
molded process.  As a result, non continuous barrier platelets form in the plastic which slow
permeation by creating a longer path length for hydrocarbons passing through the tank wall. We
believe that these strategies could be used with portable fuel tanks because these tanks are
                                          4-26

-------
generally blow-molded today.
       As discussed in Chapter 3, the majority of plastic fuel tanks installed on boats are
rotationally molded using a low density polyethylene.  We are interested in testing barrier
technologies and/or alternative materials that could be used with the rotational molding process.
Barrier technologies include fluorination and sulfonation. The results from one study 3 of these
techniques on automotive high density polyethylene fuel tanks are shown below in Table 4.2-8.
Because alcohol blended in the fuel can reduce the effectiveness of some strategies, this study
looked at both gasoline and methanol blends. Table 4.2-8 presents the results with both gasoline
and a 20 percent methanol blend. The  20 percent methanol blend is presented here because it
was generally the worst case for each of these technologies.  Fluorination consistently showed
more than a 90 percent reduction in permeation emissions. We are planning to have a marine
tank fluorinated so that we can test it for permeation in our lab.

                            Table 4.2-8: Permeation Rates for
                   Low Permeation Fuel Tanks at Room  Temperature
Technology
High density polyethylene
Barrier platelets (Selar®)
Fluorination
Sulfonation
Multi-layer continuous barrier
Gasoline
g/gal/day
0.13
0.06
0.01
0.01
0.01
20% Methanol
g/gal/day
0.12
0.07
0.01
0.03
0.03
       To compare the above permeation rates to those in Table 4.2-7, they would have to be
adjusted for temperature.  As temperature increases, the permeation rates increase exponentially.
For most materials, permeability increases by a factor of 2 or more for every 10°C increase in
temperature.4 Therefore, the permeation rates in Table 4.2-8 may be four times as high at 40°C.

       The California Air Resources Board (ARB) also investigated barrier platelets,
fluorination, and sulfonation on portable fuel  containers. This data is compiled in twelve data
reports on their web site and is included in our docket.5 Table 4.2-9 presents a summary of this
data which was collected using ARB Test Method 513.6  This permeation test method includes
three stages: durability cycling, preconditioning,  and testing in a variable temperature SFtED.  To
ensure durability of the fuel container and surface treating, the fuel containers are cycled a
minimum of 1000 times, over 8 hours, between 5 psi and -1 psi. The containers are then filled
with gasoline and soaked for a minimum of four weeks to ensure that the permeation rate is
stabilized. Actual permeation testing takes place in a variable temperature SFtED.  The fuel
container is sealed and 24 hour variable temperature profile is applied that runs from 65°F to
105°Fandbackto65°F.
                                          4-27

-------
       The results of the ARB testing show that all three barrier technologies can be used to
achieve significant reductions in permeation from plastic fuel containers. However, fluorination
appeared to show the most consistent results. This variability is likely due to the quality of the
barrier treatments. For instance, the effectiveness of the fluorination treatment would depend on
the concentration of fluorine, temperature profile, vacuum, and length of time of the treatment
process.  We believe that it may be necessary to define certain parameters to ensure that these
processes are applied appropriately for meeting the proposed standards.

           Table 4.2-9: Permeation Rates for Low Permeation Fuel Containers
      Filled with California Reformulated Certification Fuel over a 65-105°F Diurnal
Technology
Low density polyethylene
High density polyethylene
Barrier platelets (Selar®)
Fluorination
Sulfonation
Number of
Tanks Tested
1
12
16
37
32
Average
g/gal/day
0.99
0.85
0.44
0.08
0.35
Minimum
g/gal/day
0.99
0.49
0.05
0.03
0.01
Maximum
g/gal/day
0.99
1.63
0.98
0.25
1.87
       Over the first month or so of use, polyethylene fuel tanks can expand by as much as three
percent due to saturation of the plastic with fuel. Manufacturers have raised the concern that this
hydrocarbon expansion could affect the effectiveness of surface treatments like fluorination or
sulfonation.  We believe that this will not have a significant effect on the effectiveness of these
surface treatments. During the treatment process, the tank is heated and therefore would be
expanded. Also, as shown by the ARB data above, more than 95 percent reductions were
achievable in some tanks even after being subjected to the pressure cycling and fuel soak.
Fluorination, which showed the most consistent results resulted in more than a 90 percent
reduction on average.  As methods improve, even better permeation control could be achieved.
The variability in the barrier processes can be reduced by  specifically controlling the treatment
process specifically for the tanks being treated.  The tanks included in the ARB testing were
generally included in larger batches of other products where the process was optimized for those
other products.  Also, more consistent fuel tank construction, including the makeup of the plastic
used, would make it easier for the fluorinater or sulfonater to optimize their processes and reduce
variability.

       The U.S. Coast Guard has raised the issue that any process applied to marine fuel tanks to
reduce permeation would also need to pass  Coast Guard flame resistance requirements (33 CFR
183). We are not aware of any reason that the barrier technologies discussed above would affect
the flame resistance of a marine fuel tank.  In response to  this concern, we sent a fluorinated fuel
tank to a testing lab which certified the fuel tank as passing the U.S. Coast Guard flame
resistance test.7
                                          4-28

-------
Material properties
       We are also investigating the possibility of using nylon as a low permeation alternative in
the rotational molding process.  Based on our understanding of material properties, the use of
nylon in the construction of these fuel tanks would reduce permeation by more than 95 percent
when compared to cross-link polyethylene such as is used today. Table 4.2-10 presents
permeation rates of various plastics (including barrier materials) on an ASTM specified gasoline
(Fuel C) and a blend with 15% methanol (M15).8 Table 4.2-11 presents permeation rates for
several materials used as barriers in automotive applications at 40°C.9  Table 4.2-12 presents
data from the same study for three different temperatures using Fuel C. We are also assessing the
durability of surface treatments.  Specifically, we are evaluating the durability of sulfonation and
fluorination on crosslink tanks and are assessing the viability of replacing crosslink materials
with FIDPE as an option if needed to  address durability concerns associated with sloshing.

	Table 4.2-10: Fuel System Material Permeation Rates at 23°C	
           Material
Fuel C, g-mm/m2/day
Ml 5, g-mm/m2/day
 High density polyethylene
 Nylon 12, rigid
 Ethylene vinyl alcohol
 Polyacetal
 Polybutylene terephthalate
 Polyvinylidene fluoride
         35
         0.2
        35
        64
        10
        3.1
        0.4
        0.2
                                           4-29

-------
              Table 4.2-11:  Fuel System Material Permeation Rates at 40°C
Material Name
CARILON
EVOH-F101
EVOH-XEP380
HOPE
LDPE
Nylon 12 (L2101F)
Nylon 12 (L2140)
CELCON
THV - 500
DyneonE14659
DyneonE14944
ETFE Aflon COP
m-ETFE
ETFE Aflon LM730 AP
FKM-70 16286
GFLT 19797
Composition
aliphatic poly-ketone thermoplastic
ethylene-vinyl alcohol
ethylene-vinyl alcohol
high density polyethylene
low density polyethylene
plasticized fuel line grade
plasticized fuel line grade
acetal copolymer
terpolymer*
fluoropolymer film
fluoropolymer film
ethylene-tetrafluoro-ethylene
ethylene-tetrafluoro-ethylene
ethylene-tetrafluoro-ethylene
fluoroelastomer
fluoroelastomer
FuelC
g-mm/m2/day
0.06
O.OOOl
O.OOOl
90
420
2.0
1.8
0.38
0.31
0.25
0.14
0.24
0.27
0.41
11
13
M15
g-mm/m2/day
13
3.5
5.3
71
330
250
-
-
3.0
2.1
1.7
1.8
1.6
2.6
-
-
* tetra-fluoro-ethylene, hexa-fluoro-propylene, and vinyledene fluoride
Table 4.2-12: Fuel System Material Material Permeation Rates for Fuel C by Temperature
Material Name
CARILON
HOPE
Nylon 12 (L2140)
CELCON
FKM-70 16286
GFLT 19797
40°C
g-mm/m2/day
0.06
90
1.8
0.38
11
13
50°C
g-mm/m2/day
0.2
190
4.9
0.76
25
28
60°C
g-mm/m2/day
0.55
310
9.5
1.7
56
60
       The manufacturer who supplied the bladder tank for our test program is now working
with a lower permeability material known as THV. THV is a Teflon® based material that can be
used to achieve more than a 95 percent reduction in permeation from current bladder fuel tanks
made from polyurethane.10  In addition, THV is fairly resistant to ethanol. Table 4.2-13
compares fuel permeation rates for gasoline and a  10 percent ethanol at 60°C.
                                          4-30

-------
                Table 4.2-13: Bladder Material Permeation Rates at 60°C
Barrier Film Material
polyurethane
THV-200
THV-500
Thickness
mm
0.52
0.35
0.32
Gasoline
g-mm/m2/day
285
8.2
10%Ethanol
g-mm/m2/day
460
54
10.5
       4.2.2.2 - Hoses

       As discussed in Chapter 3, the majority of marine fuel hoses are constructed primarily of
nitrile rubber with a chloroprene cover and are designed to meet the Coast Guard specification of
100 g/m2/day at 23°C for hoses where liquid fuel is normally continuously in the hose. We
believe that permeation emissions from hoses can be reduced by more than 95 percent by using
either thermoplastics blended into the rubber or a multi-layer hose construction with low
permeation barrier layers. In fact, at least one company produces  hoses of this type for marine
applications. Table 4.2-14 presents permeation rates for various materials that could be used in
hose construction.  These hoses were tested at 23°C on ASTM Fuel C and a 15% methanol
blend.11 Fuel fill neck hoses are subject to a less stringent permeation standard (300 g/m2/day at
23°C) under the Coast Guard specifications because they are not normally continuously in
contact with fuel.  However, we are proposing to control permeation rates in fill neck hoses
because hose exposed to air saturated with fuel vapor will permeate as much as hose exposed to
liquid fuel. Although marine fill neck hoses are wrapped rather than extruded, we believe that
barrier technology could be applied here as well. In fact, there are chemical hoses that employ
several different material layers, which are manufactured using the same process as marine fill
neck hoses.  We intend to test several standard marine hoses and barrier hoses for permeation.
We also intend to investigate fluorination as an option for reducing permeation through hoses.
                                          4-31

-------
                  Table 4.2-14: Hose Material Permeation Rates at 23°C
Material Name
NBR(33%ACN)
HNBR (44%ACN)
FVMQ
FKM Viton A200 (66%F)
FKM Viton B70 (66%F)
FKM Viton GLT (65%F)
FKM Viton B200 (68%F)
FKM Viton GF (70%F)
FKM Viton GFLT (67%F)
Teflon FEE 1000L
Teflon PF A 1000LP
TefzelETFElOOOLZ
Nylon 12 (GM grade)
Composition
nitrile rubber
hydrogenated nitrile rubber
flourosilicone
fluoroelastomer
fluoroelastomer
fluoroelastomer
fluoroelastomer
fluoroelastomer
fluoroelastomer
fluoroplastic
fluoroplastic
fluoroplastic
plasticized fuel line grade
FuelC
g-mm/m2/day
669
230
455
0.80
0.80
2.60
0.70
0.70
1.80
0.03
0.18
0.03
6.0
M15
g-mm/m2/day
1,188
828
635
36
32
60
12
3.0
14
0.03
0.13
0.20
83
4.2.3 - Evaporative Emission Test Procedures

       This section discusses test procedures for measuring evaporative emissions from fuel
tanks and hoses.

       4.2.3.1 - Diurnal Emission Testing

       The evaporative emission test is intended to be representative of a single 22.2-35.6°C
(72-96°F) day. Emissions would be measured in a Sealed Housing for Evaporative
Determination (SFLED) over a 72 hour period. The fuel tank is set up in the SFLED with all of the
hoses, seals and other components attached.  The fuel tank is then filled and drained to 40
percent capacity with 9 RVPm test fuel and soaked with an open vent until the fuel reached
22.2°C.  Immediately after the fuel reaches this temperature, the SFLED would be purged, and the
diurnal temperature cycling  would begin. The temperature cycle actually three repeats of a 24
hour diurnal trace and shown in Figure 4.2-8. The final g/gal/day result would be based on the
highest grams of these three 24 hour cycles divided by the fuel tank capacity.
       m Reid Vapor Pressure (psi). This is a measure of the volatility of the fuel.  9 RVP
represents a typical summertime fuel in northern states.
                                          4-32

-------
             Figure 4.2-8: Three Day Temperature Trace for Diurnal Testing
          0
12
24          36
    Test Time [hours]
48
60
72
       This diurnal cycle is consistent with the test requirements for highway vehicles.
However, the test procedure for highway vehicles includes engine operation and hot soaks (fuel
system heating from the engine). The purpose of the engine operation is to purge the charcoal
canister that collects evaporative emissions in highway applications. However, we are excluding
engine operation from the evaporative test procedures for boats using SI marine engines because
we do not anticipate the use of charcoal canisters in these applications.

       For plastic fuel tanks,  the tank would have to be filled and soaked for an appropriate
period of time to ensure that permeation emissions are accurately reflected in the test procedure.
We are proposing that the fuel tank be filled to capacity with 9RVP fuel and soaked for 30 days
at a temperature between 20-30°C (68-86°F). The fuel would then be drained just prior to the
beginning of the test described above.

       4.2.3.2 - Fuel Tank Permeation Testing

       In 33 CFR 183.620(i), Coast Guard sets a fuel tank permeation standard of 1.2 grams/day
per cubic foot of compartment space. Testing to determine this permeation rate must be
performed at 40°C (104°F) using ASTM reference fuel "C" (D-471-1979).  In  1986, the Coast
Guard performed a test program to measure permeation rates from three plastic fuel tanks.12 As a
part of this  program, Coast Guard investigated several existing test methods and found them to
be insufficient for their test program. In the end they decided to test the entire  fuel tank in a
closed chamber with tight temperature control.  The closed chamber had a nitrogen purge to
prevent saturation of the gas in the chamber with hydrocarbons.  Fuel loss was measured using a
gravimetric approach in which the tank was sitting on a sensitive scale.
                                          4-33

-------
       We also are investigating whether we should require specific durability test procedures
for fuel tanks.  Such durability tests could include pressure-vacuum cycle testing, slosh testing,
and temperature cycling. Examples of these tests are included in the docket.13

       4.2.3.3 - Hose Permeation Testing

       There are three primary methods that are used today for measuring permeation from fuel
lines. The first is used for marine hoses while the other two are used for low permeability
automotive fuel lines.

       Coast Guard standards for marine fuel hoses (33 CFR 183) cite SAE recommended
practice J152714 which, among other things, includes test procedures for measuring permeation
from marine fuel hoses. In this test procedure, a short section of hose is attached to a
nonpermeable container (i.e. metal fuel can) and plugged.  Fuel is added to the container and the
mass of the entire unit is measured every 24 hours for 15 days and the peak fuel loss is
determined.  This testing is performed at 23 ± 2°C on both reference fuel "C" for the version of
the SAE standard referenced in 33 CFR 183.  However, SAE J1527 was revised in 1993 to
include permeation standards for hoses tested on a fuel blend with 15% methanol. This test
procedure is  simple; however, it is sufficient for marine hoses because they have high permeation
rates ranging from 100 to 600 g/m2/day depending on the hose class and the fuel used.

       Recommended practice for automotive fuel tubing is  defined in SAE J2260.15 The
permeation requirements in this standard are one to two orders of magnitude lower than those
defined for marine hoses.  These permeation requirements  are based on the same fuels as the
revised SAE J 1527, but at a much higher temperature (60°C).  At 60°C, permeation rates for a
given material may be 16 times as high or higher than at 23 °C based on the rule of thumb that
permeation doubles for every 10°C increase in temperature.  SAE J2260 refers to the permeation
test procedures in SAE J1737.16

       The procedures in SAE J1737 were designed to measure the low permeation  rates needed
in automotive applications to meet EPA evaporative emission requirements. There was concern
that the weight loss measurement, such as used in SAE J1527, was not sensitive enough to
measure these low permeation rates. In addition, this procedure requires exposing the material to
be tested for hundreds of hours, depending on the material  and fuel, to reach a steady-state
permeation rate.  In this procedure, fuel is heated to 60°C and circulated through a tube running
through a glass test cell. Nitrogen around the tube in this test cell is used to carry the permeate to
activated charcoal canisters. The canisters are weighed to determine their capture. Because the
canister is much lighter than the reservoir/hose in the SAE J1527 configuration, a much more
accurate measurement of the permeation loss can be made.

       Some manufacturers of low permeability product are  finding that as their emission rates
decrease, they need more refined test procedures to accurately measure permeation.  These
manufacturers are finding that the weight of the charcoal canisters are much higher than the
                                          4-34

-------
permeate being measured.  As an alternative to the gravimetric approach used in the above two
procedures, even very low permeation emissions can be measured by a flame ionization detector
and a SHED. As discussed earlier, SHED testing is generally used to measure evaporative
emissions from whole automobile systems as well.

4.2.4 - Impacts 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 new emission standards for marine vessels. In this
case, we would not expect evaporative emission controls to have any impact on noise from a
marine vessel because noise from the fuel system is insignificant.

       We anticipate that the proposed evaporative emission standards will have a positive
impact on energy. By capturing or preventing the loss of fuel through evaporation, we estimate
that the average boat will save about 44 gallons of fuel over its lifetime. This translates to a fuel
savings of about 31 million gallons in 2030  when most boats used in the U. S. are expected to
have evaporative emission  control.

       We believe that the proposed evaporative emission standards will have no negative
impacts on safety, and may even have some benefits. All of the evaporative emission control
technologies discussed in this document are capable of meeting the U.S. Coast Guard safety
requirements for fuel systems described in 33 CFR 183. Even by using a sealed system with an
automotive fuel cap, manufacturers would be well under the Coast Guard pressure requirements.
We recognize that some fittings and valves may have to be redesigned if required to operate
under pressure, but believe that this could easily be done by 2008. Several of the technologies
that we believe could be used to meet the proposed standards would  require low or no pressure
build up in the fuel system.  In addition, the  reduction of fuel vapor in and around the boat could
provide some benefits to boaters.

4.3 - Sterndrive/Inboard Marine

       At this time we are  not proposing exhaust emission  standards for spark-ignition
sterndrive and inboard (SD/I) marine engines. However, we are continuing our efforts to develop
and demonstrate catalytic control on SD/I marine engines in the laboratory and in-use, and will
place new information in the docket when it is available.  In fact, we intend to follow with
another rulemaking in the future that will address emissions from SD/I engines, and possibly
other marine engines, once we have collected more information.  This section presents
information that we have collected on an SD/I engines and discusses issues associated with using
catalysts in marine applications.
                                          4-35

-------
4.3.1 - Exhaust Emission Data from SD/I Engines

       This section presents emission data that we have collected on SD/I marine engines. All
of this data was collected during laboratory tests over the ISO E4 duty cycle.  The first part of
this section presents baseline emission data while the second part presents results from the first
step of our catalyst development program.

       4.3.1.1 - Baseline Emission Data

       The vast majority of SD/I engines four-stroke reciprocating piston engines similar to
those used in automotive applications. The exceptions are small sales of air boats using aircraft
piston-type engines and at least one marinizer that uses Mazda rotary engines. More than half of
the new engines sold are equipped with electronic fuel injection while the rest still use
carburetors. The majority of the electronic fuel injection systems are multi-port injection;
however, throttle-body injection is  also widely used, especially on smaller engines.

       Table 4.3-1 presents baseline  emissions for four-stroke SD/I engines built up from
automotive engine blocks. 17>18'19>20>21 Five of these engines are carbureted, one uses throttle-body
fuel injection, and three use multi-port fuel injection. One of the multi-port fuel injected engines
was tested with three calibrations.  Note that without emissions calibrations performed
specifically for low emissions, the HC+NOx emissions are roughly equal for the carbureted and
fuel injected engines.  Using the straight average, HC+NOx  from the carbureted engines is 15.6
g/kW-hr while it is 15.9 g/kW-hr from the fuel injected  engines (14.8 g/kW-hr if the low HC
calibration outlier is excluded).
                                           4-36

-------
                    Table 4.3-1: Baseline SD/I Exhaust Emission Data
Engine
#
1
2
3
4
5
6
7
8
8
8
9
10
Power
[kW]
79
91
121
158
167
196
159
185
181
191
219
229
Fuel Delivery System
carburetor
carburetor
carburetor
carburetor
carburetor
carburetor
throttle-body fuel injection
multi-point electronic fuel injection
#8, low CO calibration
#8, low HC calibration
multi-point electronic fuel injection
multi -point electronic fuel injection
HC
[g/kW-hr]
11.2
4.4
8.5
7.3
8.0
4.4
2.9
5.2
5.8
3.3
4.7
2.7
NOx
[g/kW-hr]
8.0
13.9
6.0
6.0
5.7
10.3
8.7
9.7
11.7
18.2
9.4
13.1
CO
[g/kW-hr]
281
98
247
229
174
101
42
149
48
72
160
44
       4.3.1.2 - Emission Data For Catalyst Development Test Program

       In a joint effort with the California Air Resources Board, we contracted with Southwest
Research Institute to perform catalyst development and emission testing on a SD/I marine
engine.22 This test program was performed on a 7.4 L electronically controlled Mercruiser engine
with multi-port fuel injection. Figure 4.3-1 illustrates the three primary catalyst packaging
configurations we used. The upper right-hand picture shows a catalyst packaged in a riser
extension which would be placed between the lower exhaust manifold and the exhaust elbow.
This riser had the same outer dimensions as the stock riser extension produced by Mercury
Marine. The upper left-hand picture shows a catalyst packaged in the elbow. The lower picture
shows a larger catalyst that was packaged downstream of the exhaust elbow. All of these catalyst
configurations were water jacketed to prevent high surface temperatures.
                                          4-37

-------
         Figure 4.2-1:  Three Catalyst Configurations Used in SD/I Test Program
       Table 4.3-2 presents the exhaust emission results for the baseline test and three catalyst
packaging configurations. For the riser catalyst we tested the engine with two cell densities, 60
and 300 cells per square inch (cpsi), to investigate the effects of back-pressure on power.  The
catalysts reduced in HC+NOx in the range of 42% to 77% and reduced CO in the range of 46%
to 54%. There were no significant impacts on power, and fuel consumption actually improved
due to the closed-loop engine calibrations necessary to optimize the catalyst effectiveness. At the
full power mode, we left the engine controls in open-loop and allowed it to operate rich to protect
the catalysts from over-heating.
                                          4-38

-------
       Table 4.3-2: Exhaust Emission Data on a SD/I Engine with Various Catalysts
Catalyst Scenario
baseline (no catalyst)
60 cpsi catalyst in riser
300 cpsi catalyst in riser
400 cpsi catalyst in elbow
200 cpsi catalyst downstream
EC
[g/kW-hr]
4.7
2.5
1.7
2.8
2.1
NOx
[g/kW-hr]
9.4
5.7
1.9
1.1
1.2
CO
[g/kW-hr]
160
81
87
81
83
Power
[kW]
219
214
213
217
221
BSFC
[g/kW-hr]
357
345
349
337
341
       The testing described above is our first step in developing and demonstrating catalysts
that can reduce emissions from marine engines. However, this program only looked at catalysts
operating in a laboratory. We are now engaged in a new test program which involves operating
an engine in a boat with catalysts. The intent of this program is to address open issues associated
with using catalysts in marine applications.  These open issues are described below.

       4.3.1.3 - Emission Data Using Exhaust Gas Recirculation

       We have tested two engines over the ISO E4 marine test cycle with and without the use of
exhaust gas recirculation (EGR).23'24  The first engine was a Ford heavy-duty highway engine.
Although this was not a marine engine, it uses the same basic technology as SD/I engines. The
second engine was a Mercruiser SD/I engine which  is a marinized version of a GM heavy-duty
highway engine. This is the same engine on which we performed the catalyst work; however, the
baseline emissions are a little different because we rebuilt the head of the engine prior to the
catalyst development work described above.  This test data suggests that, through the use of EGR
on a SD/I marine engine, a 40-50% reduction in NOx (30-40% reduction in HC+NOx) can be
achieved.  EGR was not applied at peak power in this testing because the throttle is wide open at
this point and displacing fresh  air with exhaust gas at this mode of operation would reduce
power.  We also did not apply EGR at idle because the  idle mode does not contribute
significantly to the cycle weighted NOx.
                                         4-39

-------
       Table 4.3-3: Exhaust Emission Data Using EGR on the E4 Marine Duty Cycle
EGR Scenario

First Engine: baseline
with EGR
Second Engine: baseline
with EGR
HC
[g/kW-hr]
2.7
2.7
4.5
4.5
NOx
[g/kW-hr]
13.4
7.1
8.4
4.8
CO
[g/kW-hr]
26.5
24.3
171
184
Power
[kW]
145
145
209
209
BSFC
[g/kW-hr]
326
360
349
356
4.3.2 - Open Issues for Using Catalysts in Marine Applications

       Although we have demonstrated that catalysts can be used to reduce emissions in the
laboratory, we believe there are several open issues with the application of catalysts in the marine
environment. These issues include durability concerns, packaging constraints, water reversion,
and safety considerations that are unique to marine applications.

       4.3.2.1 - Packaging

       Due to the design of marine exhaust systems, fitting a catalyst into the exhaust system
may be difficult for many boat/engine designs.  Often boat builders will strive to minimize the
space taken up in the boat by the engine compartment. In addition, these exhaust systems are
designed, for safety reasons, to avoid hot surface temperatures. For most SD/I engines, the
surface temperature is kept low by running raw water through a jacket around the exhaust
system. This raw water is then mixed with the exhaust before being passed out of the engine. To
avoid a major redesign of the exhaust system, the catalyst must be placed upstream of where the
water and exhaust mix. In addition, the catalyst must be insulated and/or water-jacketed to keep
the surface temperatures of the exhaust low.
       The above design constraints led to the
catalyst packaging designs presented above in
Figure 4.3-1. All of these catalysts were water-
jacketed and placed in the exhaust manifold
upstream of where the water and exhaust mix.
Of these catalyst packaging configurations, it
appears that the riser and elbow catalysts
would be the easiest to package in a boat.  The
larger catalyst may be too long to fit in most
applications, especially sterndrives where
exhaust passes through the drive unit attached
to the engine. Figure 4.3-2 illustrates where
the riser catalyst would be fit in a typical
marine exhaust system.
Figure 4.3-2:  Marine Riser Catalyst
                                          4-40

-------
       Another issue is maintaining high enough temperatures with a water-jacketed catalyst for
the catalyst to react properly.  The light-off temperature of these advanced catalysts is in the
range of 250 to 270°C which was low enough for the catalysts to work effectively in our
laboratory tests.  However, it could be necessary for manufacturers to retard the spark timing at
idle and low load for some engines to maintain this minimum temperature in the catalyst.

       4.3.2.2 - Durability

       Two aspects of marine applications that could affect catalyst durability are thermal load
and vibration. Because the catalyst would be coupled close to the exhaust ports, it would likely
see temperatures as high as 750 to 850°C when the engine is operated at full power. The bed
temperature of the  catalyst would be higher due to the reactions in the catalyst. However, even at
full power, the bed temperature of the catalyst most likely wouldn't exceed the exhaust
temperature by more than 50-100°C. In our laboratory  testing, we minimized the temperature at
full load by operating the engine with a rich air-fuel mixture.  The temperatures seen were well
within the operating range of new Pd-only catalysts which are capable of withstanding prolonged
exposure to temperatures approaching 1100°C.25

       In on-highway applications, catalysts are designed to operate in gasoline vehicles  for
more than 100,000 miles.  This translates to about 4,000-5,000 hours of use on the
engine/catalyst. We estimate that, due to low annual hours of operation, the average useful life
of SD/I engines is only about  10-20 percent of this value. This suggests that catalysts designed
for automotive use should be durable over the useful life of a marine engine. However, we do
not know the relative stress on the catalyst from an hour of automotive versus marine use. The
testing that we are currently engaged in will include operation of the catalyst over a thermal aging
cycle.

       Use of catalysts in automotive, motorcycle, and hand-held equipment applications
suggests that catalysts can be packaged to withstand the vibration in the exhaust manifold. We
will learn more about the effects of vibration on a catalyst in our boat testing.

       4.3.2.3 - Water Reversion

       A third aspect of marine applications that could affect catalyst durability is the effect of
water contact with  the catalyst. There is  concern that water could creep back up the exhaust
passages, due to pressure pulses in the exhaust, and damage the catalyst and oxygen sensor. This
damage could be due to thermal shock from cold water coming into contact with a hot catalyst
ore due to salt deposition on the catalyst. One study was performed to investigate these catalyst
durability effects on a two-stroke outboard.26 The results of this study are summarized in Table
4.3-4.
                                          4-41

-------
               Table 4.3-4: Summary of Marine Catalyst Durability Study
Issue
high catalyst
temperatures
salt water effects
fresh water effects
thermal shock of
hot catalyst with
cold water
deterioration
factor
Investigation
- compared base catalyst to catalysts
aged for 10 hrs at 900 and 1050°C
- soaked catalysts in two seawater
solutions and compared to base catalyst
- used intake air with a salt-water mist
- soaked catalyst in fresh water and
compared to base catalyst
- flushed out catalyst with fresh water
that was soaked in salt water
- as part of the catalyst soaking tests,
900°C catalysts were soaked in both
salt and fresh water
- operated engine with catalyst for 300
hours of E4 operation
Result
- little change in conversion
efficiency observed
- large drop in conversion
efficiency observed
- no effect on catalyst
- little change in conversion
efficiency observed
- washing catalyst removes salt
and restores some performance
- no damage to the catalysts was
reported
- only a 20% loss in conversion
efficiency for a 2-stroke engine
       The above study on catalysts in marine applications was performed supplemental to an
earlier study.27 The earlier study also showed that immersing the catalysts in salt water would
hurt the conversion efficiency of the catalyst, but that operating in a marine environment would
not.  In addition, this earlier study showed that much of the efficiency loss due to salt on the
catalyst could be reversed by flushing the catalyst with water. This paper also showed that with
the catalyst activated, temperatures at full power were less than at mid power because the space
velocity of the exhaust gases at rated speed was high enough to reduce the conversion efficiency
of the catalyst.

       Because salt deposition severely reduces the conversion efficiency of the catalyst, it is
critical that salt water not reach the catalyst. This design problem is the primary goal of the
catalyst and boat testing that we are currently performing.  Manufacturers claim that the cooling
water in the exhaust manifold jacket will work its way back up the exhaust stream and poison the
catalyst if it is not packaged correctly.  Although this water reversion may happen in some
instances, we believe that this is an engineering design problem that can be solved given proper
resources and time. Manufacturers already design their exhaust systems to prevent water from
reaching the exhaust ports. If water did reach the exhaust ports in today's designs, significant
durability problems would result from corrosion. Also, it would not take much water back-flow
into the cylinder after engine shut-down to cause the engine to hydraulically lock the next time it
was started.  However, we believe that it would be valuable for us to demonstrate a catalyst on a
                                          4-42

-------
boat before proposing catalyst-forcing standards.

       4.3.2.4 - Safety

       We also want to ensure that the use of catalysts on boats will not impose new hazards to
boaters. We have been working with the U.S. Coast Guard to identify potential safety problems
with using catalysts in marine applications. The Coast Guard has stated that they have two
concerns.  First, they want to make sure that any additional heat load in the engine compartment
will not add to the risk of fires, other safety hazards, or other detrimental impacts on the engine
or components. Second, they want to make sure that the exhaust systems with catalysts will not
lead to CO leaks due to additional joints in or maintenance of the exhaust system.  These issues
will be addressed in the upcoming boat testing.
                                          4-43

-------
Chapter 4 References
1.  Letter from Kevin Madison, Top Dog Systems, to Mike Samulski, U.S. EPA, regarding the
Top Dog vaporless fuel system, April 23, 2001, Docket A-2000-01, Document U-G-106.

2.  Allen, S.J., "Fuel Tank Permeability Test Procedure Development; Final Report," U.S. Coast
Guard, December 1986, Docket A-2000-01, Document U-A-17.

3.  Kathios, D., Ziff, R., Petrulis,  A., Bonczyk, J., "Permeation of Gasoline and Gasoline-alcohol
Fuel Blends Through High-Density Polyethylene Fuel Tanks with Different Barrier
Technologies," SAE Paper 920164, 1992, Docket A-2000-01, Document No. H-A-60.

4.  Lockhart, M., Nulman, M., Rossi, G., "Estimating Real Time Diurnal Permeation from
Constant Temperature Measurements," SAE Paper 2001-01-0730, 2001, Docket A-2000-01,
Document No. IV-A-21.

5.  www.arb.ca.gov/msprog/spillcon/reg.htm,  Updated March 26, 2001, Copy of linked data
reports available in Docket A-2000-01, Document IV-A-09.

6.  "Test Method 513; Determination of Permeation Rate for Spill-Proof Systems," California
Air Resources Board, Adopted July 6, 2000, Docket A-2000-01, Document IV-A-08.

7.  EVIANNA Laboratory Inc, "Certification Test Report 15707-1; Fire Resistance Test of
Moeller  Test Tank for EPA," March 28, 2002, Docket A-2000-01, Document IV-A-33.

8.  Hopf, G., Ries, H., Gray, E., "Development of Multilayer Thermoplastic Fuel Lines with
Improved Barrier Properties," SAE Paper 940165, 1994, Docket A-2000-01, Document No. IV-
A-22.

9.  Nulman, M., Olejnik, A., Samus, M., Fead, E., Rossi, G., "Fuel Permeation Performance of
Polymeric Materials," SAE Paper 2001-01-1999, 2001, Docket A-2000-01, Document No. IV-A-
23.

10. Facsimile from Bob Hazekamp, Top Dog Systems, to Mike  Samulski, U.S. EPA,
"Permeation of Polyurethate versus THV Materials @ 60°C," January 14, 2002, Docket A-2000-
01, Document U-B-30.

11. Stahl, W., Stevens, R., "Fuel-Alcohol Permeation Rates of Fluoroelastomers,  Fluoroplastics,
and Other Fuel Resistant Materials," SAE Paper 920163, 1992, Docket A-2000-01, Document
No. IV-A-20.

12. Allen, S.J., "Fuel Tank Permeability Test Procedure Development; Final Report," U.S.  Coast
Guard, December 1986, Docket A-2000-01, Document U-A-17.
                                        4-44

-------
13.  DRAFT SAE Information Report J1769, "Test Protocol for Evaluation of Long Term
Permeation Barrier Durability on Non-Metallic Fuel Tanks," Docket A-2000-01, Document IV-
A-24.

14.  SAE Surface Vehicle Standard, "Marine Fuel Hoses," Society of Automotive Engineers J
1527, Issued 1985-12, Revised 1993-02, Docket A-2000-01, Document No. IV-A-19.

15.  SAE Surface Vehicle Standard, "Nonmetallic Fuel System Tubing with One or More
Layers," Society of Automotive Engineers J 2260, Issued 1996-11, Docket A-2000-01,
Document No. IV-A-18.

16.  SAE Surface Vehicle Recommended Practice, "Test Procedure to Determine the
Hydrocarbon Losses from Fuel Tubes, Hoses, Fittings, and Fuel Line Assemblies by
Recirculation," Society of Automotive Engineers J 1737, Issued 1997-08, Docket A-2000-01.

17.  Michigan Automotive Research Corporation, "Exhaust Emissions from Marine Engines
Using Alternate Fuels," Prepared for NMMA, October 8, 1992, Docket A-2000-01, Document H-
A-96.

18.  Mike Samulski, "Effects of Transience on Emissions from Inboard Marine Engines," memo
to Docket #A-92-28, May 30, 1996, Docket A-2000-01, Document II-B-07.

19. Letter from Jeff Carmody, Santa Barbara Air Quality management District, to Mike
Samulski, U.S.  EPA, July 21, 1997, Docket A-2000-01, Document U-A-21.

20.  "National Marine Manufacturers Association's Small Business Boat Builder and Engine
Manufacturers Comments in Response to EPA's Initial Regulatory Flexibility Analysis
Regarding EPA's Plans to Propose Emission Regulations for Recreational Marine Gas and
Diesel Powered Sterndrive/Inboard Engines," July 12, 1999, Docket A-2000-01, Document II-G-
253.

21.  Mace, et. al., "Emissions from Marine Engines with Water Contact in the Exhaust System,"
SAE Paper 980681, 1998, Docket A-2000-01, Document II-A-97.

22. Carroll, J., White, J., "Marine Gasoline Engine Testing," Southwest Research Institute,
Prepared for the U.S. EPA, September, 2001, Docket A-2000-01, Document II-A-91.

23.  Memorandum from Joe McDonald and Mike Samulski, "EGR Test Data from  a Heavy-Duty
Gasoline Engine on the E4 Duty Cycle," July 12, 1999, Docket A-2000-01, Document U-A-92.

24. Carroll, J., White, J., "Marine Gasoline Engine Testing," Southwest Research Institute,
Prepared for the U.S. EPA, September, 2001, Docket A-2000-01, Document II-A-91.

25.  Manufacturers of Emission Controls Association, "Overview of Recent Emission Control
Technology Developments," November 18, 1997, Docket A-2000-01, Document U-A-23.


                                        4-45

-------
26.  Fujimoto, H., Isogawa, A., Matsumoto, N., "Catalytic Converter Applications for Two
Stroke, Spark-Ignited Marine Engines," SAE Paper 951814, 1995, Docket A-2000-01, Document
II-A-94.

27.  Fujimoto, H., Isogawa, A., Matsumoto, N., "Catalytic Converter Applications for Two
Stroke, Spark-Ignited Marine Engines," SAE Paper 941786, 1994, Docket A-2000-01, Document
II-A-93.
                                        4-46

-------
                      CHAPTER 5: Estimated Costs
       This chapter describes our approach to estimating the cost of complying with emission
standards. We start with a general description of the approach to estimating costs, then describe
the technology changes we expect and assign costs to them.  We also present an analysis of the
estimated aggregate cost to society.

5.1 - Methodology

       We developed the costs for individual technologies using information provided by ICF,
Incorporated and Arthur D. Little, as cited below.  The technology characterization and cost
figures reflect our current best judgment based on  engineering analysis, information from
manufacturers, and the published literature.  The analysis combines cost figures including
markups to the retail level.

       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 the engine or vessel manufacturers' overhead and
profit.1 For technologies sold by a supplier to the engine manufacturers, an additional 29 percent
markup is included for the supplier's overhead and profit.  All costs are in 2001 dollars.

       The analysis presents an estimate of costs that would apply in the first year of new
emission standards and the corresponding long-term costs. Long-term costs decrease due to  two
principal factors. First, fixed costs are assessed for five years, after which they are fully
amortized and are therefore no longer part of the cost calculation. Second, manufacturers are
expected to learn over time to produce the engines with the new technologies at a lower cost.
Because of relatively low  sales volumes, manufacturers are less likely to put in the extra R&D
effort for low-cost manufacturing.  As production  starts, assemblers and production engineers
will then be expected to find significant improvements in fine-tuning the designs and production
processes.  Consistent with analyses from other programs, we reduce estimated variable costs by
20 percent beginning with the third year of production and an additional 20 percent beginning
with the sixth year of production.2   We believe it is appropriate to apply this factor here, given
that the industries are facing emission regulations for the first time and it is reasonable to expect
learning to occur with the experience of producing and improving emission-control technologies.

       Even though the analysis does not reflect all the possible technology variations and
options that are available to manufacturers,  we believe the projections presented here provide a
cost estimate  representative of the different approaches manufacturers may ultimately take.  We
expect manufacturers in many cases to find and develop approaches to achieve the emission
standards at a lower cost than we describe in this analysis.
                                           5-1

-------
5.2 - Cost of Emission Controls by Engine/Vehicle Type

5.2.1 - Evaporative Emission Control from Boats

       5.2.1.1 - Technologies and Estimated Costs

       As discussed in earlier chapters, we believe that there is a wide range of technology that
could be used to meet the proposed evaporative emission standards for marine vessels.  Table
5.2.1-1 presents our best estimates of the costs of applying various evaporative emission control
technologies to a boat with a thirty gallon fuel tank. These costs would generally be higher for a
larger boat and lower for a smaller boat, but this is illustrative of a typical marine configuration.

       The range of potential  costs for bladder tanks are based on  confidential conversations
with a manufacturer of this product.  This section looks at the costs for an average sized tank (32
gallons) with typical fuel hose lengths. Non-permeable material costs are based on conversations
with manufacturers, current material costs,  and engineering judgement.  Tank insulation costs are
based on typical costs for commercial insulating sheets with an R value of 15.  To cost out the
pressurized tank, we estimated that the material cost the tank is about ten percent of the total
value. For a 32 gallon tank, approximately 25 to 35 pounds of plastic is used at a cost of about
$0.50 per pound.  We estimate that if a manufacturer were to make changes to the geometry of
the fuel tank to help withstand 1 psi of pressure without significant deflection, it could increase
the material needed by 10 to 30 percent.  Pressure relief valve and  limited flow orifice costs are
based on products available in automotive parts stores.  Barrier hose technologies are based on
incremental costs of existing products.3 Surface treatment costs are based on price quotes from a
company that specializes in this fluorination costs4; shipping costs  are also included.  Finally,
nylon tank costs are based on  a material price of $2.00 per pound.  Finally, the cost of the
pressure-free tank is based on  the cost of materials that could be used for a volume compensating
bag and the consideration of adding a new fitting to a fuel tank.  The costs in Table 5.2.1-1
include a 29 percent manufacturer markup.
                                           5-2

-------
     Table 5.2.1-1 Technologies Under Consideration and Costs for a 32 Gallon Tank
Technologies Under Consideration
bladder tank (commercial technology)
tank insulation
1 psi pressurized tank: material
relief valve/orifice
pressure-free tank (volume compensating bag)
low-permeable barrier in fuel lines (8 ft.)
low-permeable barrier in fill neck (2 ft.)
fluorination or sulfonation of plastic fuel tank
low-permeable material (nylon) for plastic fuel tank
In-Use Reductions
90% diurnal and permeation
25%+ of diurnal
25 -50% of diurnal
50- 100% of diurnal
95%+ of permeation
95%+ of permeation
95%+ of permeation
95%+ of permeation
Cost Increase
$65 - 194
$13-26
$1 -6
$4-9
$26 - 52
$6-13
$3-6
$13-19
$52 - 65
       5.2.1.2 - Operating Cost Savings

       Evaporative emissions are essentially fuel that is lost to the atmosphere. Over a the
lifetime of a typical boat, this can result in a significant loss in fuel. The anticipated 70 percent
reduction in evaporative emissions due to the proposed standards will result in significant fuel
savings.  Table 5.2.1-2 presents the value of the fuel savings for control of diurnal and
permeation emissions. These numbers are calculated using an estimated fuel cost of $1.10 per
gallon (based on 1997-2002 pre-tax fuel prices)5 and fuel density of 6 Ibs/gallon (for lighter
hydrocarbons which evaporate  first). The figures in Table 5.2.1-2 are for an average boat which
we estimate has a 32 gallon fuel tank and a 20 year life.  The total savings over 20 years is 44
gallons which translates  to a discounted lifetime savings of 27 dollars.

            Table 5.2.1-2 Fuel Savings Per Boat Due to the Proposed  Standards
Average Parameters

Evaporative HC reduced [grams/year]
Evaporative HC reduced [gallons/year]
Fuel savings [gallons/life]
Undiscounted savings [$/life]
Lifetime fuel savings (NPV, 7%)
Diurnal

990
0.4
7
$8
$4
Tank
Permeation
1,930
0.7
14
$15
$9
Hose
Permeation
3,190
1.2
23
$25
$14
                                           5-3

-------
       5.2.1.3 - System Integration and Compliance Costs

       We are proposing to allow manufacturers to use design based certification to the proposed
performance standard as an alternative to evaporative emission testing.  However, they will still
need to integrate insulation or fluorination into their designs and there will be some engineering
and clerical effort need to submit the required information for certification. We estimate a cost
of $10,000 for a manufacturer producing 50,000 fuel tanks per year. After the first year of
certifying to our standards, manufacturers will be able to carry over the majority of the
certification effort from year to year. Therefore, this cost will be greatly reduced after the first
year of the program.  If we treat the $10,000 as a fixed cost and amortize it over five years, we
get a cost per fuel tank  of about four cents.

       5.2.1.4 - Total Costs

       To determine the total costs, we use the scenario that all boat builders will use low
permeation fuel tanks and hoses and use a zero pressure limited flow orifice to control diurnal
emissions.  Because these technologies are well established, we do not apply  a learning curve to
future costs.

       For a boat with  a 32 gallon plastic fuel tank, we based cost of reducing fuel tank
permeation on the fuel tank being fluorinated or sulfonated.  This translates to a technology cost
in the range of $13 to $19. In addition we consider the cost of low permeation barrier hose and
fill neck. This adds another $9 to $19 to the hardware cost. We also estimate the cost of diurnal
emission control by addition a cost of $4-9 for the pressure relief valve and a material cost of $1-
6 to minimize deflection under 1  psi of pressure. Using the average costs, we get a total cost of
$43.

       For a boat with  a 32 gallon aluminum fuel tank, we estimate our costs based only on low
permeation barrier hose and fill neck and the use of a pressure relief valve. No costs are included
for fluorination or sulfonation because the aluminum tanks do not permeate fuel emissions.
Using the average of the costs, we get a total cost of $21.

       Weighting the aluminum and plastic fuel tank costs we get an average cost of $36 dollars
for an average 32 gallon marine tank.  Coupled with the discounted fuel savings of $27 and
certification costs, we get a discounted lifetime cost per vessel of $9.

       5.2.1.6 - Aggregate Costs for Marine Evaporative Emission Control

       The above analyses developed incremental per-vessel cost estimates for boats powered by
spark-ignition engines.  Using these per-vessel costs and projections of future annual sales, we
have estimated total aggregate annual costs for proposed emission standards.  The aggregate
costs are presented on a cash-flow basis, with hardware and fixed costs incurred in the year the
vehicle is sold and fuel savings occurring as the vehicle is operated over its life. This may
                                           5-4

-------
understate the time-value of the fixed costs because they are likely to be incurred before the
vehicle is sold; however, this has a negligible effect on the results of this analysis. Table 5.2.1-3
presents a summary of this analysis.  As shown in the table, aggregate net costs are $22.5 million
in 2008, but are then projected to decline as fuel savings continue to ramp-up as more vessels
meeting the standards are sold. Long-term projections show that, starting in 2023, fuel savings
offset the costs of meeting emission standards.

                                      Table 5.2.1-3
           Summary of Annual Aggregate Costs and Fuel Savings for Marine SI
                                   (millions of dollars)

Total Costs
Fuel Savings
Net Costs
2008
$24.4
($1.8)
$22.5
2010
$24.7
($5.5)
$19.2
2015
$25.6
($14.6)
$11.0
2020
$26.5
($23.1)
$3.4
2025
$27.4
($29.7)
($2.3)
       To project annual sales, we started with the 1998 population estimates presented in
Chapter 6.  We then backed out sales based on the useful life and growth estimates in Chapter 6.
Table 5.2.2-4 provides a summary of the sales estimates used in the aggregate cost analysis. We
then use the relationship between engine type and size to fuel tank size discussed in Chapter 6.

                                     Table 5.2.1-4
                    Estimated Annual Sales of Marine SI Fuel Tanks
Application
Outboards
Personal Watercraft
Sterndrive/Inboards
1999
422,000
115,000
91,000
2006
443,000
121,000
95,000
2010
456,000
124,000
98,000
2020
490,000
133,000
105,000
       To calculated annual aggregate costs, the sales estimates have been multiplied by the per-
unit costs discussed above. This calculation takes into consideration vessel sales and scrappage
rates. The year-by-year results of the analysis are provided in Chapter 7.

5.2.2 - Highway Motorcycles

       Costs estimated for Class 3 highway motorcycles were developed in cooperation with ICF
Incorporated and Arthur D. Little - Acurex Environmental. The analysis was based upon an
analysis completed by the California Air Resources Board (ARB) when they proposed new
emission standards in October of 1998.  The ARB standards, finalized in  1999 and applicable
only to Class 3 motorcycles (280 cc and greater), will occur in two phases. A Tier-1 standard of
                                           5-5

-------
1.4 g/km HC+NOx will apply to Class 3 motorcycles for the 2004 through 2007 model years, and
a Tier-2 standard of 0.8 g/km HC+NOx will apply to Class 3 motorcycles for the 2008 and later
model years.  These new standards may be met on a corporate-average basis. Our analysis, while
following the general methodology used by the ARB, incorporates some changes to the
methodology as well as some component cost estimates that differ from those used by the ARB.6

       The ARB Staff Report estimated costs for two displacement ranges for Class 3
motorcycles: 280 to 699 cc and 700 cc and greater. Currently the smaller range is subject to an
HC standard of 1.0 g/km in California, while the larger displacement range is subject to an HC
standard of 1.4 g/km in California. In order to simplify the calculations, two specific engine
displacements were used to compare with the ARB displacement categories.  The ARB Staff
Report indicates that the sales-weighted average for the 280cc to 699cc class was 600cc and the
sales weighted average for the greater than 700cc category was 1200cc. EPA certification data
shows that the national averages for the two displacement categories are 593cc and  1260cc, thus
demonstrating that the ARB analysis is reasonably representative of national sales.  These sales
weighted averages were used in developing the cost estimates. The costs include a mark-up to the
retail level.

       The costs also account for the difference in production levels between the two
displacement categories as represented by the 600 cc and 1200 cc engines. The California ARB
used an average annual production level of 15,000 units for both categories. In this analysis we
use 25,000 units per year for the 600 cc engine and 64,000 units per year for the 1200 cc engine.7

       The analysis for Class 3 motorcycles combines the individual technology costs into a total
estimated cost package. The composite analysis weights the costs by the projected percentage of
use of the technologies both in the baseline and control scenarios to project industry-wide
average per vehicle costs. The full analysis for Class 3 is followed by an analysis of costs for
motorcycles under 50cc in Section 5.2.2.4.

       5.2.2.1 - Technologies and Estimated Costs

       Highway motorcycles are currently powered mostly by carbureted four-stroke engines.
However, even in the absence of new regulation, the penetration of fuel-injected models is
increasing, most likely due to the improvements in reliability, performance, and fuel economy
that fuel injection can offer. EPA's motorcycle certification database for 2001 and 2002 model
year motorcycles indicates that the most prevalent emission controls used to meet the current
standards are engine modifications and mechanically-controlled secondary air injection. To an
increasingly greater extent manufacturers are currently incorporating fuel injection and catalytic
converters on some 2001 and 2002 models. Table 5.2.2-9 shows the increased use of these
technologies in the 2002 model year relative to the 2001 model year, based on EPA certification
data. For example, catalyst usage increased from 13 to 20 percent, and fuel injection usage
increased from 37 to 47 percent, for the larger Class 3 motorcycles.
                                          5-6

-------
       While manufacturers will use various means to meet the future standards, there are four
basic types of existing emission control technologies on which we have based our cost analysis;
engine modifications, electronic fuel injection, pulse air systems, and catalyst converters with
oxygen (O2) sensors. These technologies are being used to varying degrees already and we
would expect manufacturers to increase the use of these approaches to meet future standards. In
general, we would expect emissions control strategies to include more precise fuel control, better
fuel atomization and delivery, reduced engine-out emission levels from engine changes, and
increased use of catalysts.

       We have included the costs for increased use of engine modifications to meet the Tier 1
standards.  We would expect this approach to be used on essentially all engine models. Engine
modifications include changes such as improved cylinder honing for better oil control, modified
cam profiles to provide increased valve overlap (providing internal exhaust gas recirculation),
and piston modifications to improve ring land height, tumble and squish for reduced hydrocarbon
emissions and better combustion.  As shown in Table 5.2.2-1, we have estimated the per unit cost
of engine modifications to be in the $6 to $8 range.

                         Table 5.2.2-1 Engine Modification Costs
Engine Size
600cc
1200cc
Variable Costs
mproved Pistons
Number Required
Hardware costs
Markup @ 29%
Warranty Markup @ 5%
Total Component Costs
$2
2
$4
$1
$0
$5
$3
2
$6
$2
$0
$8
Fixed Costs
l&D Costs
Fooling Costs
Jnits/yr.
^ears to recover
Fixed cost/unit
Total Costs m
$62,292
$30,000
25,000
5
$1
Sfi
$62,292
$35,000
64000
5
$0
ss
       The combinations of low-emission technologies ultimately chosen by motorcycle
manufacturers are dependent on the engine-out emission levels of the vehicle, the effectiveness
of the prior emission control system, and individual manufacturer preferences.  We believe
manufacturers will increase their use of electronic fuel injection in order to achieve better fuel
delivery control. We project that the use of electronic fuel injection will increase to meet the
Tier-1 and Tier-2 standards, with an accompanying decline in the use of carburetion. Tables
5.2.2-2 and 5.2.2-3 provide estimates of the costs of carburetor systems and electronic fuel
injection systems.  To project the incremental costs of going to electronic fuel injection we have
subtracted the costs of the carburetor systems. We have estimated the incremental costs of
                                           5-7

-------
electronic fuel control to be in the range of $183 and $191.
                             Table 5.2.2-2. Carburetor Costs
Cnpine Size
Carburetor
Number Required
Hardware Cost to Manufacturer
^abor @ $28 per hour
^abor overhead @ 40%
Vlarkup @ 29%
Total Comnonent Costs
600cc
$60
2
$120
$1
$1
$35
$157
1 200cc
$60
2
$120
$1
$1
$35
$157
                      Table 5.2.2-3. Electronic Fuel Injection Costs
Engine Size
600cc
1200cc
Variable Costs
injectors (each)
Number Required
3ressure Regulator
intake Manifold
Throttle Body/Position Sensor
7uel Pump
3CM
\ir Intake Temperature Sensor
Vlanifold Air Pressure Sensor
injection Timing Sensor
Wiring/Related Hardware
Hardware Cost to Manufacturer
^abor @ $28 per hour
^abor overhead @ 40%
Vlarkup @ 29%
Warranty Markup @ 5%
Total Component Costs
$12
2
$10
$30
$35
$20
$100
$5
$10
$5
$10
$249
$4
$2
$72
$12
$339
$15
2
$10
$30
$35
$20
$100
$5
$10
$5
$10
$255
$4
$2
$74
$13
$348
Fixed Cost to Manufacturer
^&D Costs
Tooling Costs
Jnits/yr.
fears to recover
Fixed cost/unit
Total Costs (^
$62,292
$10,000
25,000
5
$1
S340
$62,292
$12,000
64000
5
$0
S348
       We are also projecting the increased use of pulse air systems and catalyst systems for both
the Tier 1 and Tier 2 standards. We have estimated the cost of both a mechanical and an
                                           5-8

-------
electronically controlled pulse air system and have projected the use of the electronic system for
motorcycles equipped with electronic fuel control systems.  For catalysts, we would expect an
increase in use both for Tier 1  and Tier 2. While we do not expect catalysts to be used on all
models, they will likely remain a key tool for emissions control. We are also projecting the
increased use of oxygen sensors and have tied the increase in their use to the use of catalysts.  It
seems reasonable to expect manufacturers to use oxygen sensors with catalysts to ensure
stoichiometric engine operation.  Tables 5.2.2-4 through 5.2.2-7 provide estimates for these
technologies.

                           Table 5.2.2-4  Pulse Air Valve Costs
Fnpine Size
Pulse Air Valve Type
600cc
Mechanical
Electronic
1200cc
Mechanical
Electronic
Variable Costs
-'ulse Air
^abor @ $28 per hour
^abor overhead @ 40%
Vlarkup @ 29%
Warranty Markup @ 5%
Total Component Costs
$8
$1
$0
$3
$0
$12
$12
$1
$0
$4
$1
$18
$8
$1
$0
$3
$0
$12
$12
$1
$0
$4
$1
$18
Fixed Costs
l&D Costs
Fooling Costs
Jnits/yr.
^ears to recover
Fixed cost/unit
Total Costs (^
$0
$8,000
25,000
5
$0
$12
$0
$8,000
25,000
5
$0
$18
$0
$10,000
64,000
5
$0
$12
$0
$10,000
64,000
5
$0
$18
                      Table 5.2.2-5.  Catalyst Costs to Manufacturer
Emission Level
Catalyst Volume (L)
Vletallic Substrate
Washcoat
3recious Metals
^an (18 gauge 304 SS)
TOTAL MAT. COST
LABOR
^abor Overhead @ 40%
supplier Markup @ 29%
Manufacturer Price

0.3
$8.80
$0.54
$5.65
$1.09
$16.08
$11.20
$4.48
$9.21
$40.98
0.60
$11.60
$1.09
$11.30
$1.52
$25.51
$11.20
$4.48
$11.94
$53.13
                                            5-9

-------
                               Table 5.2.2-6  Catalyst Costs
Enmne Size
600cc
1200cc
Variable Costs
Dxidation Catalyst
^abor @ $28 per hour
^abor overhead @ 40%
DEM markup @ 29%
Warranty Markup @ 5%
Total Component Costs
$41
$1
$1
$12
$2
$57
$45
$1
$1
$14
$0
$61
Fixed Costs
l&D Costs
Fooling Costs
Jnits/yr.
^ears to recover
Fixed cost/unit
Total Costs (^
$54,750
$10,000
25,000
5
$1
$58
$54,750
$10,000
64000
5
$0
$61
                            Table 5.2.2-7 Oxygen Sensor Costs
Engine Size
600cc
1 200cc
Variable Costs
Dxygen Sensors
Number Required
Hardware costs
Markup @ 29%
Warranty Markup @ 5%
Total Component Costs
$10
2
$20
$6
$1
$27
$10
2
$20
$6
$1
$27
Fixed Costs
l&D Costs
Fooling Costs
Jnits/yr.
^ears to recover
Fixed cost/unit
Total Costs ($}
$0
$5,000
25,000
5
$0
$27
$0
$5,000
64,000
5
$0
$27
       5.2.2.2 - Compliance Costs

       We estimate highway motorcycle chassis-based certification to cost about $25,000 per
engine line, including $10,000 for engineering and clerical work and $15,000 for durability and
certification testing.  As with other fixed costs, we amortized the cost over 5 years of engine sales
to calculate per unit certification costs shown in Table 5.2.2-8. We are not projecting increased
facility costs because manufacturers currently are required to certify and therefore have adequate
test facilities. In addition, because California standards are schedule to be implemented 2 years
                                           5-10

-------
earlier than the EPA standards would be, we would expect actual average certification costs to be
much lower than those estimated here.  It is likely that data from the California program would
be used by manufacturers for EPA certification.

                   Table 5.2.2-8 Estimated Per Unit Certification Costs

units/year
certification costs
Highway Motorcycles
25,000
$0.26
64,000
$0.10
       5.2.2.3 - Highway Motorcycle Total Costs

       The analysis below combines the costs estimated above into a total composite or average
cost. The composite analysis weights the costs by projecting the percentage of their use, both in
the baseline and control scenario, to project industry-wide average per vehicle costs. Baseline
estimates were made based on the sales weighted current usage rate indicated by our certification
database. The weighting between the two engine displacement ranges is also based on projected
sales information submitted by manufacturers at time of certification. The table below presents
the baseline technology usage for the most recent two model years for which we have data (2001
and 2002).  The analysis uses the 2002 technology usage baseline, but the 2001 penetration of the
various technologies is presented to illustrate the progress that already appears to be underway in
implementation of technologies such as engine modifications, fuel injection, and catalyst-based
systems.  For example, almost half of the larger motorcycles (where most U.S. sales are) are fuel-
injected as of the most recent model year, a trend that we expect to continue as manufacturers
implement designs intended to meet the upcoming California and European standards, both of
which are large motorcycle markets.

       For electronic fuel injection, we are attributing half the cost to emissions control and the
other half to improved performance.  We believe this is appropriate because the technology
provides substantial benefits in performance and reliability in addition to improved emissions.
Discounting the cost by one-half also helps account for the current trend toward increased use of
electronic fuel injection in the absence of more stringent federal emissions standards.

       A summary of the estimated near-term and long-term per unit average incremental costs
for highway motorcycles is provided in Tables 5.2.2-9 and 5.2.2-10.  Long-term costs do not
include fixed  costs, which are retired,  and include cost reductions due to the learning curve.
                                          5-11

-------
                              Table 5.2.2-9  Estimated Average Costs For Motorcycles (Tier 1)

600 cc (17%)
1200 cc (83%)
engine modifications
electronic fuel injection*
mechanical pulse air valve
electronic pulse air valve
catalyst
O2 sensors
total
engine modifications
electronic fuel injection*
mechanical pulse air valve
electronic pulse air valve
catalyst
O2 sensors
total
Near Term Composite Incremental Cost
Long Term Composite Incremental Cost
Cost
$6
$92
$12
$17
$58
$27
--
$8
$96
$12
$17
$60
$27
--
--
--
Baseline Usage Rate
2001
60%
20%
53%
0%
10%
0%
--
54%
37%
39%
0%
13%
5%
--
--
--
2002
66%
23%
48%
0%
15%
1%
--
53%
47%
37%
0%
20%
5%
--
--
--
Tier 1 Control Usage Rate
100%
40%
60%
40%
15%
15%
--
100%
50%
50%
50%
25%
25%
--
--
--
Incremental Cost
$2
$16
$1
$7
$0
$4
$30
$4
$3
$2
$9
$3
$5
$26
$26
$17
* The electronic fuel injection costs have been discounted by 50 percent to reflect the portion of the cost attributed to emissions
control.
                                                           5-12

-------
                              Table 5.2.2-10 Estimated Average Costs For Motorcycles (Tier 2)

600 cc (17%)
1200 cc (83%)
engine modifications
electronic fuel injection*
mechanical pulse air valve
electronic pulse air valve
catalyst
O2 sensors
total
engine modifications
electronic fuel injection*
mechanical pulse air valve
electronic pulse air valve
catalyst
O2 sensors
total
Near Term Composite Incremental Cost
Long Term Composite Incremental Cost
Cost
$6
$92
$12
$17
$58
$27

$8
$96
$12
$17
$60
$27

--
--
Baseline Usage Rate
100%
40%
60%
40%
15%
15%

100%
50%
50%
50%
25%
25%

--
--
Tier 2 Control Usage Rate
100%
60%
40%
60%
50%
50%

100%
60%
40%
60%
50%
50%

--
--
Incremental Cost
$0
$18
($2)
$3
$20
$10
$49
$0
$10
($1)
$2
$15
$7
$33
$35
$22
* The electronic fuel injection costs have been discounted by 50 percent to reflect the portion of the cost attributed to emissions
control.
                                                           5-13

-------
       5.2.2.4 - Highway Motorcycles Under 50 cc

       We are establishing standards that are in line with standards already established in the
countries and regions that represent the major scooter markets in the world, and which produce
millions of scooters.  The US scooter market is tiny in comparison to those in other parts of the
world and represents a very small market for scooter manufacturers. In costing out changes in
technology, it is therefore reasonable to project that the research and development and tooling
necessary to meet standards will occur in response to the standards in other countries, rather than
EPA standards. US standards would ensure that clean scooters developed for major markets  also
are the ones brought into the US market rather than traditional 2-strokes.  It is also reasonable to
expect that the fixed costs of research and development and tooling are spread over the
worldwide unit sales for those scooters. Therefore, we would expect those costs to be small on a
per unit basis.

       To establish cost estimates for under 50 cc motorcycles, we first looked at the detailed
cost analysis performed for our Small SI Engine Phase 2 Standards rulemaking, in which the cost
for converting from 2-stroke to 4-stroke was estimated to be about $14 per unit for a small engine
(about 25 cc).8 The unit basis for this analysis was 90,000 units. EPA's estimate was supported
in comments received from one Small  SI manufacturer that estimated a cost of about $15 for
production less than 1 million units and another that estimated a unit cost of about $10.

       We also searched  available literature and found a report prepared for the World Bank that
estimated the difference in cost between a 2-stroke to a 4-stroke 50 cc scooter to be about $60 per
unit.9 We believe the difference in the two cost estimates can be accounted for in the difference
in engine displacement and manufacturer mark-up. The $60 estimate appears to be reasonable
for a baseline 4-stroke engine. We are assuming that the cost represents a long-term stabilized
cost rather than the initial cost of production. To the extent the $60 estimate reflects costs of
initial production, the long-term costs may be somewhat lower due to the learning curve effect.
Also, to the extent that some manufacturers currently may have initial costs that are  somewhat
higher than $60, we believe that costs reductions are likely to take place prior to 2006, due to the
learning curve effect and a substantial increase in  production in response to the world markets.
Costs are likely to be stabilized by 2006.

       In addition to the cost of the scooter, manufacturers will also incur costs for certification.
Our estimate of average certification costs is $1.52 per vehicle based on average U.S sales per
engine family of 4,300 units. We have used the same methodology and costs for certification
used for motorcycles above 50 cc (see  section 5.2.2.2 - Compliance Costs) because the
requirements are the same. We have established average  sales per vehicle line using U.S. sales
information provided by the Motorcycle Industry Council for motorcycles under 50cc.

       The average cost for motorcycles under 50 cc must account for the fact that some scooters
are already equipped with 4-stroke engines and a portion of the market is likely to remain 4-
stroke prior to 2006.  For  those, the costs for meeting the  standards would essentially be the cost
                                          5-14

-------
of certification. Based on current our estimated sales split between two and four-strokes in the
less than 50 cc market, we are projecting that 4-strokes will account for about 30 percent of sales
prior to 2006." Sales weighting the estimated cost for a 2-stroke of $61.52 and the estimated cost
for the 4-stroke of $1.52 (for certification) results in an estimated average cost for motorcycles
under 50cc of about $44.

       Switching from conventional 2-stroke engines to 4-stroke engines results in a fuel
economy savings of at least 30 percent.  We have estimated a net present value per vehicle
lifetime savings of about $8, using the factors provided in the table below.10

              Table 5.2.2-11 Estimate Per vehicle Lifetime Fuel Use and Cost

Annual Miles
Average Life
fuel consumption (miles per gallon)
Fuel cost
Discounted Lifetime Fuel Consumption
Discounted Lifetime Fuel Cost
2-stroke
650
6
70
$1.10
43.45 gallons
$47.80
4-stroke
650
6
91
$1.10
36.43 gallons
$40.07
       5.2.2.5 - Highway Motorcycle Aggregate Costs

       The above analyses developed incremental per vehicle cost estimates for highway
motorcycles.  Using these per vehicle costs and projections of future annual sales, we have
estimated total aggregate annual costs for the highway motorcycle standards.  To estimate future
sales for Class 3 motorcycles, we started with 1999 sales of 387,000 units and projected out
using a annual growth rate of 1 percent.11  For motorcycles below 50cc, we estimated 2001 sales
of about 30,000 units and also applied a compound growth rate of 1 percent.12 Aggregate fuel
savings for the motorcycles less than 50cc, has also been estimated based on the usage and
fraction of the fleet converted to 4-stroke engines.13  Table 5.2.2-12 presents the results of this
analysis. As shown in the table, annualized aggregate costs increase from about $16 million in
2006 to about $26 million in 2010 when the program is fully phased in.  Costs are projected to
then decline somewhat to about $20 million as fixed costs are retired,  after which costs are
       n The approximate sales split is based on available information from MIC, discussions
with industry, and the number of 4-stroke 50 cc models currently offered in the US by non MIC
manufacturers and importers.
                                          5-15

-------
projected to gradually increase over time due to growth in vehicle sales.
           Table 5.2.2-12 Annualized Aggregate Costs for Highway Motorcycles
Calendar
Year
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
Total Cost
(Millions ^
0
0
12.16
12.28
10.24
10.34
23.61
23.60
20.81
21.02
18.76
18.84
19.03
19.22
19.41
19.60
19.80
20.00
20.20
20.40
20.60
20.81
Fuel Savings
riUillions ^
0
0
0.02
0.07
0.11
0.14
0.17
0.20
0.22
0.24
0.25
0.26
0.27
0.28
0.29
0.30
0.30
0.31
0.32
0.32
0.33
0.34
5.3 - Aggregate Costs

       The above analysis presents unit cost estimates for each category of equipment included
in the proposal. These costs represent the total set of costs the manufacturers will bear to comply
with emission standards. With current and projected estimates of engine and equipment sales,
we translate these costs into projected direct costs to the nation for the new emission standards in
any year. The costs to manufacturers generally range from $12 to $42 million annually.  The cost
savings due to fuel savings start slowly, then increase as greater numbers of compliant engines
and equipment enter the fleet.  The net cost (i.e., the sum of the cost to manufacturers and the
fuel savings) peaks at about $36 million in 2010 and decreases to about $11 million by 2025.
Table 7-3.5 shows these aggregate costs for all years during the period between 2006 and 2027.
                                          5-16

-------
Chapter 5 References
1. "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-2000-01, Document II-A-54.

2.For further information on learning curves, see Chapter 5 of the Economic Impact, from
Regulatory Impact Analysis - Control of Air Pollution from New Motor Vehicles Emissions
Standards and Gasoline Sulfur Control Requirements, EPA420-R-99-023, December 1999. A
copy of this document is included in Air Docket A-2000-01, at Document No. II-A-83.  The
interested reader should also refer to previous final rules for Tier 2 highway vehicles (65 FR
6698, February 10, 2000), marine diesel engines (64 FR 73300, December 29, 1999), nonroad
diesel engines (63 FR 56968, October 23, 1998), and highway diesel engines (62 FR 54694,
October 21, 1997).

3.  Trident Marine Hose, "Retail Price List 2001," Docket A-2000-01, Document No. IV-A-15.

4. "Information on Costs and Effectiveness of Fluorination Received from Fluoroseal,"
Memorandum from Mike Samulski to Docket A-2000-1, March 27, 2002, Docket A-2000-01,
Document IV-B-03.

5. Energy Information Administration, www.eia.doe.gov/neic/hisotoric/hpetroleum.htm,
February  27, 2002, Docket A-2000-01, Document IV-A-16.

6.  Arthur D. Little-Acurex Environmental. Memorandum from Lou Browning to Chris Lieske:
On-Road  Motorcycle Draft Final Cost Comparisons, June 22, 2001, Docket A-2000-01,
Document U-G-145.

7.  Arthur D. Little-Acurex Environmental. Memorandum from Lou Browning to Chris Lieske:
On-Road  Motorcycle Draft Final Cost Comparisions, June 22, 2001, Docket A-2000-01,
Document H-G-145.

8. EPA Final Regulatory Impact Analysis, Phase 2 Final Rule: New Nonroad Handheld Spark-
Ignition Engines At or Below 19 Kilowatts, March 2000, EPA420-R-00-004

9. "Air Pollution from Motor Vehicles, Standards and Technology for Controlling Emissions",
Asif Faiz, Christopher S. Weaver, Michael Walsh, The World Bank, Washington DC,
November 1996, Docket A-2000-01, Document IV-A-29.

10. "Spreadsheet for Modeling the Emissions Inventories from On-Highway Motorcycles <50cc
under the Proposed Rule", Memorandum from Phil  Carlson to Docket A-2000-01, April 12,
2002, Docket A-2000-01, Document IV-B-09.
                                        5-17

-------
11. Annual sales taken from the "2000 Motorcycle Statistical Annual", Motorcycle Industry
Council, Docket A-2000-01, Document H-D-03.

12. The Motorcycle Industry Council provided 2001 sales estimates for member companies of
17,166 for scooters below 50 cc. We were unable to find statistics for non-MIC companies
which make up a sizable share of the market. To attempt to account for sales for these non-MIC
companies we asked several companies for their estimate of the overall 50cc scooter market and
have based our final sales estimates on  the information they provided.

13. "Spreadsheet for Modeling the Emissions Inventories from On-Highway Motorcycles <50cc
under the Proposed Rule", Memorandum from Phil Carlson to Docket A-2000-01, April 12,
2002, Docket A-2000-01, Document IV-B-09.
                                         5-18

-------
                  CHAPTER 6:  Emissions Inventory
6.1 - General Methodology

       The following chapter presents our analysis of the emission impact of the proposed
standards for marine vessels and highway motorcycles. This chapter also discusses the possible
impact of exhaust emission control from sterndrive and inboard marine engines.  We first present
an overview of the methodology used to generate the emissions inventories, followed by a
discussion of the specific information used in generating the inventories for each of the regulated
categories of engines as well as the emission inventories.  Emissions from a typical piece of
equipment are also presented. This analysis does not monetize the emission reductions or health
benefits.

6.1.1 - Highway Motorcycle Emissions

       The modeling of highway motorcycle >50cc emissions is typically done with our
MOBILE model. However, the inputs for motorcycles >50cc used in the MOBILE model have
not been updated in many years. In developing the proposal for highway motorcycles, we have
come across new information on current emission levels, revised scrappage estimates, and
revised mileage accumulation rates for such vehicles.  Because of this new information, we
developed a spreadsheet for modeling the effect of the proposed standards for highway
motorcycles >50cc that incorporates this new information. In addition, a similar spreadsheet was
developed for modeling the effect of the proposed standards for highway motorcycles <50cc
(currently unregulated). A copy of both spreadsheets developed for modeling the effect of the
proposed standards on highway motorcycles has been placed in the docket for this rulemaking.1'2

6.1.2 - Marine Evaporative Emissions

       Evaporative emissions refer to hydrocarbons released into the atmosphere when gasoline,
or other volatile fuels, evaporate from a vehicle. For this analysis, we model four types of
evaporative emissions:

       - diurnal: These emissions are due to temperature changes throughout the day.  As the
       day gets warmer, the fuel heats up and begins to evaporate.

       - refueling: These emissions are the vapors displaced from the fuel tank when fuel is
       dispensed into the tank.

       - permeation:  These emissions are due to fuel that works its way through the material
       used in the fuel system. Permeation is most common through plastic fuel tanks and
       rubber hoses.
                                          6-1

-------
       -hot soak:  These emissions are due to temperature changes due to heat from the engine.


       Copies of the spreadsheets developed for modeling marine evaporative emissions are
included in the docket.3 The calculations in these spreadsheets are described below.

       6.1.2.1 - Diurnal Emissions

       We are currently in the process of revising the inputs to the calculations for evaporative
emissions in the draft NONROAD model.  The analysis for this rule includes the inputs that we
anticipate will be used in the draft NONROAD model. Because diurnal and refueling emissions
are dependent on ambient temperatures and fuel properties which vary through the nation and
through the year, we divided the nation into six regions and modeled each region individually for
each day of the year.  The daily temperatures by region are based on a report which summarizes a
survey of dispensed fuel and ambient temperatures in the United States.4

       For diurnal emission estimates, we used the Wade equations5'6 to calculate grams of
hydrocarbons emitted per day per volume of fuel tank capacity.  The Wade equations are well
established and are used in both the MOBILE and draft NONROAD models with an adjustment
based on empirical data. These calculations are a function of vapor space, fuel vapor pressure,
and daily temperature variation and are as follows:

  Vapor space (ft3) = ((1.15 - tank fill) x tank size) / 7.841                           (Eq. 6-1)

       where:
              tank fill = fuel in tank/fuel tank capacity0
              tank size = fuel tank capacity in gallons

  TI (°F) = (Tmax - Tmin) x 0.922 + Tmin                                            (Eq. 6-2)

       where:
              Tmax = maximum diurnal temperature (°F) of the fuel
              Tmin = minimum diurnal temperature (°F) of the fuel

  V100 (psi) = 1.0223 x RVP + [(0.0357 X RVP)/(1-0.0368 x RVP)]                   (Eq. 6-3)

       where:
              V100 = vapor pressure at 100°F
              RVP = Reid Vapor Pressure of the fuel
       0 We use 50% fill for our calculations.

                                          6-2

-------
  E100 (%) = 66.401-12.718 x V100 +1.3067 x V1002 -  0.077934 x V1003
                + 0.0018407 x V1004                                             (Eq. 6-4)

  D^ (%) = E100 + [(262 / (0.1667 * E100 + 560) - 0.113] x (100 - Tmin)              (Eq. 6-5a)

  Dmax (%) = E100 + [(262 / (0.1667 * E100 + 560) - 0.113] x (100 - T\)               (Eq. 6-5b)

       where:
             Dmin/max = distillation percent at the max/min temperatures in the fuel tank
             E100 = percent of fuel evaporated at 100°F from equation 6-4

  P! (psi) = 14.697 - 0.53089 x D^H- 0.0077215 x Dmin2 - 0.000055631 x Dmin3
                + 0.0000001769 x Dmin4                                         (Eq. 6-6a)

  PF (psi) = 14.697 - 0.53089 x Dmax + 0.0077215 x Dmax2 - 0.000055631 x Dmax3
                + 0.0000001769 x Dmax4                                         (Eq. 6-6a)

  Density (Ib/gal) = 6.386 - 0.0186 x RVP                                         (Eq. 6-7)

  MW (Ib/lb mole) = (73.23 -  1.274 x RVP) + [0.5 x( Tmin + Tx) - 60] x 0.059          (Eq. 6-8)

  Diurnal emissions (grams) = vapor space x 454 x density x [520 / (690 - 4 x MW)]
                 x 0.5 x [Pj / (14.7 - PT) + PF / (14.7 - PF)]
                 x [(14.7 - Pj) / (Tmin + 460) - (14.7 - PF) / (Tx + 460)]                (Eq. 6-9)

       where:
             MW = molecular weight of hydrocarbons from equation 6-8
             PI/F = initial and final pressures from equation 6-6

       We use these same equations in our modeling of evaporative emissions from on-highway
vehicles. However for on-highway applications we make a correction of 0.78 based on empirical
data.7 Because this correction is based on automotive applications we do not apply this
correction factor here.  Instead we use a correction factor of 0.65 which is based on the data
presented in Chapter 4 for non-permeable fuel tanks with a vent hose.

       6.1.2.2 - Refueling Emissions

       We used the draft NONROAD model (discussed below) to determine the amount of fuel
consumed by spark-ignition marine engines.  To calculate refueling emissions, we used an
empirical equation to calculate grams of vapor displaced during refueling events. This equation
was developed based on testing of 22 highway vehicles under various refueling scenarios and in
the benefits calculations for our onboard refueling vapor recovery rulemaking for cars and
trucks.8 These calculations are a function of fuel vapor pressure, ambient temperature, and
                                          6-3

-------
dispensed fuel temperature.  The refueling vapor generation equation is as follows:

  Refueling vapor (g/gal) = EXP(-1.2798 - 0.0049 x (Td - Ta) + 0.0203 x Td
                              + 0.1315xRVP)                                (Eq. 6-10)
       where:
             Td = dispensed fuel temperature (°F)
             Ta = ambient fuel temperature (°F)
             RVP = Reid Vapor Pressure of the fuel (psi)

       Title 40, Section 80.27 of the Code of Federal Regulations specifies the maximum
allowable fuel vapor pressure allowed for each state in the U.S. for each month of the year. We
used these limits as an estimate of fuel vapor pressure in our calculations.

       6.1.2.3 - Permeation Emissions

       For our permeation emissions modeling, we used the emission data presented in Chapter
4 to determine the mass of hydrocarbons permeated through plastic fuel tanks and rubber fuel
hoses on boats.  Because permeation is very sensitive to pressure, we used Arrhenius'
relationship9 to adjust the emission factors by temperature:

  P(T) = P0 x EXP(-a / T)                                                       (Eq. 6-11)

       where:
             T = absolute temperature
             P(T) = permeation rate at T
             P0 and a are constants

       We determined the constants by relating the equation to the known properties of materials
used in fuel tanks and hoses (presented in Chapter 4). In general, permeation doubles with every
10°C increase in temperature.10

       6.1.2.3 - Hot Soak Emissions

       To get a rough estimate of hot soak emissions from marine vessels we used data collected
on 612 light duty vehicles.11 We then applied this emission factor to our estimates of boat
operation discussed below.

6.1.3 - SD/I Exhaust Emissions

       We are in the  process of developing an emission model that will calculate emissions
inventories for most off-highway vehicle categories, including those in this rule.  This draft
model is called NONROAD. For this effort we use the most recent version of the draft
NONROAD model publicly available with some updates that we anticipate will be included in
                                          6-4

-------
the next draft release. This section gives a brief overview of the calculation methodology used in
NONROAD for calculating exhaust emission inventories. Inputs and results specific to each of
the off-highway categories in this rule are discussed in more detail later in this chapter.  For more
detailed information on the draft NONROAD model, see our website at
www. epa.gov/otaq/nonrdmdl. htm.

                      Figure 6.1.1-1: Normalized Scrappage Curve
1
0
o
0 0.8
0)
0 0.6
c
en
m 0.4
M—
0
c
|0.2
ro
LL
0
(

	 ^,







*>v^
^\
\
\








V
\
V
^







	 	




) 0.5 1 1.5 2
Engine Age Normalized by Average Useful Life
       For the inventory calculations in this rule, SD/I marine engines were divided into power
ranges to distinguish between technology or usage differences in each category. 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 exhaust emissions
inventory, for a single year, from off-highway engines is shown below:
Emissions =

                   population * power x load x annual use x emissionfactor]
(Eq.6- 12
       This equation sums the total emissions for each of the power ranges for a given calendar
year.  "Population" refers to the number of engines 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 engines operating in a given calendar year. Exhaust emission
inventories were calculated for HC, CO, and NOx.
                                          6-5

-------
       To be able to determine the mix between baseline and controlled engines, we need to
determine the turnover of the fleet. Through the combination of historical population and
scrappage rates, historical sales and retirement of engines can be estimated. We use a normalized
scrappage rate and fit it to the data for each engine type on average operating life.  Figure 6.1.1-1
presents the normalized scrappage curve used in the draft NONROAD model.  Further discussion
of this scrappage curve is available in the NONROAD documentation.12

6.2 - Effect of Emission Controls by Engine/Vehicle Type

       The remainder of this chapter discusses the inventory results for highway motorcycles,
marine vessels, and SD/I engines. Also, this section describes inputs and methodologies used for
the inventory calculations that are specific to each engine/vehicle class.

6.2.1 - On-highway Motorcycles

       As noted above, we projected the annual tons of exhaust HC, CO, and NOx, from on-
highway motorcycles using one spreadsheet for on-highway motorcycles <50cc and a second
spreadsheet for on-highway motorcycles >50cc.  Both of the spreadsheets are based on the
MOBILE model  and incorporate new information on current emission levels, updated scrappage
rates, and updated mileage accumulation rates for on-highway motorcycles. This section
describes inputs to the calculations that are specific to on-highway motorcycles then presents the
resulting emissions inventories.  These results are for the nation as a whole and include baseline
and control inventory projections.

       6.2.1.1 - Inputs for the Inventory Calculations

       Several usage inputs are specific to the calculations for on-highway motorcycles exhaust
emissions.  These inputs are annual use, average operating life, and population. Based on data
received from an industry trade group, we developed annual usage rates (i.e., mileage
accumulation rates), and average operating life estimates for on-highway motorcycles.13'14 The
average operating lifetimes were estimated to be 6.0 years for on-highway motorcycles <50cc
and 12.5 years for on-highway motorcycles >50cc. Due to limited information on motorcycles
<50cc, an  average mileage accumulation rate of 650 miles/year was used for each year of
operation (based on survey information provided by the industry trade group).  The annual
mileage accumulation rates for motorcycles >50cc used in the analysis (based on the results of a
motorcycle user survey provided by the industry trade group)  are contained in Table 6.2.1-1.  The
average mileage accumulation rate over the lifetime of a typical motorcycle >50cc is estimated to
be 2,907 miles/year.
                                          6-6

-------
                                      Table 6.2.1-1
              Mileage Accumulation Rates for On-highway Motorcycles >50cc
Age (yrs)
1
2
3
4
5
6
7
8
9
10
11+
Annual Mileage
3,308
3,320
3,751
3,330
1,920
2,884
3,021
3,475
4,131
3,977
2,032
                           Source: Motorcycle Industry Council User Survey
       In order to generate emission inventories for on-highway motorcycles, the spreadsheets
developed for this analysis calculate a fleet-average emission factor (in grams per mile) and
multiply the result by the total vehicle miles traveled (VMT) estimate for the appropriate portion
of the highway motorcycle fleet (i.e., <50cc and >50cc) in a given year. The on-highway
motorcycle fleet VMT estimates used for this analysis are summarized in Table 6.2.1-2. The on-
highway motorcycle >50cc fleet VMT estimates were developed for our recent rulemaking for
model year 2007 and later heavy-duty engines and vehicles standards.15 The 2001 VMT estimate
for on-highway motorcycles <50cc was calculated based on the estimated 2001 population of on-
highway motorcycles <50cc multiplied by the average mileage accumulation rate of 650
miles/year. VMT estimates for future year on-highway motorcycles  <50cc were grown at the
same rate as on-highway motorcycles >50cc.
                                          6-7

-------
                                      Table 6.2.1-2
       On-Highway Motorcycle Fleet VMT in Specific Calendar Year (Millions of Miles)
Category
On-Highway
Motorcycles
<50cc
On-Highway
Motorcycles
>50cc
2000
75


11,469


2005
88


13,466


2010
102


15,562


2020
130


19,885


2030
158


24,208


       The current fleet of on-highway motorcycles <50cc is powered by mostly two-stroke
engines.  Beginning in 2002, a number of new vehicles (estimated to be about 30% of the new
sales) are being powered by four-stroke engines.  The baseline (i.e., pre-control) emission factors
for two-strokes used in the spreadsheet analysis for on-highway motorcycles <50cc are based
primarily on the results of testing uncontrolled two-stroke mopeds.16 (Because the emission
factors for baseline two-stroke mopeds are the average of over one hundred vehicles of different
ages, the emission levels are assumed to be a fleet average level and so no deterioration is added
to the average levels cited in the report.)  The baseline four-strokes now being introduced are
assumed to have emission factors at the levels described below for Tier 1 vehicles.  (As noted in
Chapter 4, the Tier 1 standards are expected to result in the conversion of two-strokes to four-
strokes. We believe that existing four-stroke designs would meet the proposed Tier 1 standards.)
The HC and CO emission factors for Tier 1 on-highway motorcycles <50cc are based on the
proposed HC and CO standards factoring in the effect of deterioration under certification
conditions. The estimated emission factors also assume that manufacturers will include a
compliance margin (estimated to be 20 percent) when certifying.  (For NOx, where we are not
proposing a standard, we  have assumed that four-strokes emit at the level of uncontrolled small
4-stroke motorcycles taken from a separate report.17) The deterioration factors for Tier 1 on-
highway motorcycles <50cc are based on the data for the smallest motorcycles used in the
MOBILE model. Table 6.2.1.-3 contains the emission factors and deterioration rates for on-
highway motorcycles <50cc used in the spreadsheet analysis.
                                           6-8

-------
                                      Table 6.2.1-3
Zero-Mile Level Emission Factors and Deterioration Rates for On-Highway Motorcycles <50cc
Control Category


Baseline (Pre-control)
- two- strokes
Baseline (Pre-control)
- four- strokes
Tier 1
THC
ZML,
g/mi
9.66

1.27

1.27
DR,
g/mi/ 10k mi
0

1.31

1.31
CO
ZML,
g/mi
16.1

15.5

15.5
DR,
g/mi/ 10k mi
0

2.53

2.53
NOx
ZML,
g/mi
0.10

0.32

0.32
DR,
g/mi/ 10k mi
0

0

0
       The baseline (i.e., Tier 0) emission factors used in the spreadsheet analysis for on-
highway motorcycles >50cc are based on the average certification levels of current on-highway
motorcycles >50cc.  The baseline deterioration rates were taken from the MOBILES model. In
order to estimate the control HC+NOx zero-mile level emission level for Tier 1 on-highway
motorcycles, we took the proposed Tier 1 HC+NOx standard of 2.26 grams per mile (g/mi) and
multiplied it by 0.68, which is the ratio of the baseline zero-mile HC level based on certification
data compared to the baseline (i.e., Tier 0) HC  standard.  (We do not currently have a NOx
standard for on-highway motorcycles >50cc.) Because we are proposing an HC+NOx standard
for on-highway motorcycles >50cc, we have assumed that the Tier 1 HC/NOx split will remain
the same as the baseline HC/NOx split. Because we do not currently regulate NOx for on-
highway motorcycles >50cc, we based the HC/NOx split on current  on-highway motorcycle
>50cc certification data from the California Air Resources Board which does have a NOx
standard. For the Tier 1 deterioration rates, we applied a factor of 0.67, which is the ratio of the
pre-control HC zero-mile level compared to the estimate Tier 1 HC zero-mile level. For Tier 1
on-highway motorcycles >50cc, the NOx deterioration rate was assumed to be zero (i.e., no
deterioration in NOx emissions).  (The MOBILE model currently estimates no deterioration in
NOx emissions for on-highway motorcycles.) To estimate the emission factors and deterioration
rates for Tier 2 on-highway motorcycles >50cc, the Tier 1 emission factors and  deterioration
rates were multiplied by a factor of 0.57, which is the ratio of the proposed Tier 1 to proposed
Tier 2 HC+NOx standards.  Table 6.2.1 -4 presents the emission factors and deterioration rates
used in the spreadsheet analysis for on-highway motorcycles >50cc.  (Because we are not
proposing to change the CO standard for on-highway motorcycles >50cc, the CO emission
factors and deterioration rates are the same for baseline and control cases.)
                                          6-9

-------
                                      Table 6.2.1-4
 Zero-Mile Level Emission Factors and Deterioration Rates for On-Highway Motorcycles >50cc
Control Category
Baseline (Tier 0)
Tier 1
Tier 2
THC
ZML,
g/mi
1.44
1.02
0.58
DR,
g/mi/ 10k mi
0.70
0.49
0.28
CO
ZML,
g/mi
17.4
17.4
17.4
DR,
g/mi/ 10k mi
2.46
2.46
2.46
NOx
ZML,
g/mi
0.70
0.52
0.30
DR,
g/mi/ 10k mi
0
0
0
       The Tier 1 standards for on-highway motorcycles are proposed to take effect in 2006.
The Tier 2 standards for on-highway motorcycles >50cc are proposed to take effect in 2010.

       Another piece of information needed to develop the fleet average gram per mile emission
factors is information on the scrappage/survival rates of on-highway motorcycles.  For our
spreadsheet analyses, we used scrappage/survival rate information provided by an industry trade
group.18'19 Table 6.2.1-5 presents the scrappage/survival rate information used in the spreadsheet
models for on-highway motorcycles.
                                          6-10

-------
                                     Table 6.2.1-5
                   Scrappage/Survival Rates for On-highway Motorcycles
Age
1
2
3
4
5
6
7
8
9
10
11
12
13
Fraction Surviving
<50cc
0.94
0.88
0.79
0.71
0.60
0.50
0.36
0.28
0.23
0.13
0.08
0.05
-
>50cc
0.99
0.97
0.96
0.94
0.91
0.88
0.82
0.75
0.71
0.66
0.60
0.55
0.45
Age
14
15
16
17
18
19
20
21
22
23
24
25
-
Fraction Surviving
<50cc
-
-
-
-
-
-
-
-
-
-
-
-
-
>50cc
0.40
0.34
0.29
0.25
0.18
0.12
0.09
0.06
0.04
0.03
0.01
0
-
                    Source: Motorcycle Industry Council
       One final adjustment included in the spreadsheets, is an adjustment to account for
temperature effects on emissions.  In order to account for these impacts, the current MOBILES
model was run at ambient temperature conditions of 75°F and at typical summertime temperature
ambient conditions.  The MOBILES outputs (in grams per mile) were compared for the two runs.
The results showed that HC emissions decreased by 1 percent under summertime conditions, CO
increased by 20 percent, and NOx emissions decreased by approximately 8 percent. These
adjustments were applied to the emission factors in the spreadsheets.

       6.2.1.2 - Reductions Due to the Proposed Standards

       We anticipate that the proposed standards for on-highway motorcycles will result in a
50% reduction in both exhaust HC inventories and NOx inventories by the year 2020. Tables
6.2.1-6 through 6.2.1.-8 present our projected exhaust HC, CO, and NOx, emission inventories
for on-highway motorcycles  (including both <50cc and >50cc vehicles) and the anticipated
                                         6-11

-------
emission reductions from the proposed standards. (Because we do not expect CO reductions
from the proposal, only the baseline CO inventories are shown below.)

                                     Table 6.2.1-6
  Projected Exhaust HC Inventories and Reductions for On-Highway Motorcycles (short tons)
Calendar Year
2000
2005
2010
2020
2030
Baseline
35,000
40,000
46,000
58,000
71,000
Control
35,000
40,000
41,000
29,000
28,000
Reduction
0
0
5,000
29,000
43,000
% Reduction
0%
0%
9%
50%
61%
                                     Table 6.2.1-7
         Projected Exhaust CO Inventories for On-Highway Motorcycles (short tons)
Calendar Year
2000
2005
2010
2020
2030
Baseline
331,000
387,000
448,000
572,000
697,000
                                     Table 6.2.1-8
  Projected Exhaust NOx Inventories and Reductions for On-Highway Motorcycles (short tons)
Calendar Year
2000
2005
2010
2020
2030
Baseline
8,000
10,000
11,000
14,000
17,000
Control
8,000
10,000
10,000
7,000
7,000
Reduction
0
0
1,000
7,000
10,000
% Reduction
0%
0%
9%
50%
59%
                                         6-12

-------
       6.2.1.3 - Per Equipment Emissions from On-highway Motorcycles

       The following section describes the development of the emission estimates on a per piece
of equipment basis over the average lifetime of a typical on-highway motorcycle. The emission
estimates were developed to estimate the cost per ton of the proposed standards as presented in
Chapter 7.

       In order to estimate the emissions from an on-highway motorcycle, information on the
emission level of the vehicle, the annual usage rate of the vehicle, and the lifetime of the vehicle
are needed. The values used to predict the per piece of equipment emissions for this analysis and
the methodology for determining the values are described below.

       The information necessary to calculate the emission levels of a piece of equipment over
the lifetime of a typical on-highway motorcycle were presented in Table 6.2.1-3 and Table 6.2.1-
4.  A brand new on-highway motorcycle emits at the zero-mile level presented in the tables.  As
the on-highway motorcycle ages, the emission levels increase based on the pollutant-specific
deterioration rate.  Deterioration continues throughout the lifetime of the vehicle and the amount
of deterioration is directly proportional to the number of miles accumulated on the on-highway
motorcycle.

       As described earlier in this section, the annual usage rate for an on-highway motorcycle
<50cc is estimated to be 650 miles per year and the average lifetime is estimated to be 6.0 years.
For on-highway motorcycles >50cc, the annual usage rate is estimated to be 2,907 miles per year
and the average lifetime is estimated to be 12.5 years.

       Using the information described above, we calculated the lifetime HC or HC+NOx
emissions from typical on-highway motorcycles for both current (i.e., pre-control for <50cc and
Tier 0 for >50cc) vehicles and vehicles meeting the proposed standards. Table 6.2.1-9 presents
the lifetime HC or HC+NOx emissions for typical on-highway motorcycles on both an
undiscounted and  discounted basis (using a discount rate of 7 percent). Table 6.2.1-10 presents
the corresponding lifetime HC or HC+NOx emission reductions for the proposed standards.  (HC
estimates are  shown for on-highway motorcycles <50cc because we are not proposing a NOx
standard. HC+NOx estimates are shown for on-highway motorcycles >50cc because we are
proposing combined HC+NOx standards.)
                                          6-13

-------
                                      Table 6.2.1-9
            Lifetime Emissions from a Typical On-highway motorcycle (short tons)
Control Level
Current (Pre-control for <50cc)
Current (Tier 0 for >50cc)
Tier 1
Tier 2
<50cc Motorcycles
Exhaust HC
Undiscounted
0.031
-
0.007
-
Discounted
0.026
-
0.006
-
>50cc Motorcycles
Exhaust HC+NOx
Undiscounted
-
0.141
0.100
0.057
Discounted
-
0.093
0.066
0.038
                                     Table 6.2.1-10
       Lifetime Emission Reductions from a Typical On-highway motorcycle (short tons)
Control Increment
Current (Pre-control) to Tier 1
Current (Tier 0) to Tier 1
Tier 1 to Tier 2
<50cc Motorcycles
Exhaust HC
Undiscounted
0.024
-
-
Discounted
0.020
-
-
>50cc Motorcycles
Exhaust HC+NOx
Undiscounted
-
0.041
0.043
Discounted
-
0.027
0.029
6.2.2 - Evaporative Emission Control from Boats

       We projected the annual tons of hydrocarbons evaporated into the atmosphere from boats
using spark-ignition (SI) marine engines using the methodology discussed above in Section 6.1.2.
These evaporative emissions include diurnal, refueling, hot soak, and permeation emissions.
Although the proposed standards do not specifically require the control of refueling and
permeation emissions, we model them here because some of the technology that could be used to
meet the proposed standards could result in reductions in refueling and permeation evaporative
emissions. This section describes inputs to the calculations that are specific to boats and presents
our baseline  and controlled national inventory projections for evaporative emissions.
                                          6-14

-------
       6.2.2.1 - Inputs for the Inventory Calculations
       Several usage inputs are specific to the calculations for SI marine evaporative emissions.
These inputs are fuel tank sizes, population, and distribution throughout the nation. The draft
NONROAD model includes current and projected engine populations for each state.  We
assumed that the national fuel tank distribution would be a function of the engine distribution.
Table 6.2.2-1 presents the SI marine engine population estimates for 1998.  To predict sales
growth in future years we use the growth rates in the draft NONROAD model.20 This is a linear
growth rate, but is roughly equivalent to a compound growth rate of 0.7 percent from 2008 to
2028. These growth projections are consistent with historical  boat sales recorded by the National
Marine Manufacturers Association.21  The equivalent compound growth rate of recorded sales of
motorboats from 1980 to 2000 is  about 1 percent.

                                      Table 6.2.2-1
                  1998 Spark-Ignition Marine Engine Population by Region
Region

Northeast
Southeast
Southwest
Midwest
West
Northwest
Total
Outboards and
Personal Water craft
3,530,000
1,420,000
355,000
626,000
661,000
3,530,000
10,130,000
Sterndrive and
Inboards
651,000
261,000
65,400
115,000
122,000
650,000
1,865,000
Total

4,180,000
1,680,000
420,000
741,000
783,000
4,180,000
11,990,000
       The draft NONROAD model breaks this engine distribution further into ranges of engine
sizes. Outboards are divided into 10 power ranges, personal watercraft are divided into 5 power
ranges, and SD/I are divided into 9 power ranges.  For each of these power ranges we apply a fuel
tank size for our evaporative emission calculations.  To determine fuel capacity as a function of
engine power we looked at data on 870 boats using outboard and SD/I engines.22 This data
included total engine power for the boat and fuel tank capacity for boats ranging from 10 to 58
feet in length. From this data we determined a relationship between engine horsepower and fuel
tank size which will be used in the next draft version of NONROAD and is used for this analysis.
For personal watercraft we use the existing draft NONROAD default of gallons = 0.24 x rated
horsepower, which is consistent with sales literature published for personal watercraft.

       6.2.2.1.1 -Diurnal

       For diurnal emission calculations, we also  needed to make a distinction between portable
and installed fuel tanks. This is because most, if not all, portable fuel tanks have valves that the
user can close when the boat is not in use. We estimated that if these valves can be closed
                                          6-15

-------
whenever the boat is not in use, that about 90 percent of the diurnal vapor would be contained.
This 90 percent estimate is based on the assumption that most users close the fuel vents when the
boat is not in use. Also we assumed that all fuel tanks under 12 gallons are portable tanks.

       Another effect that we consider in our diurnal emission modeling is the relationship
between ambient temperature and fuel tank temperature throughout the day. Based the limited
test data presented in Chapter 4, we have found that the fuel temperature swings are generally not
as large as the associated ambient temperature swings.  This is especially true for boats stored in
the water. For boats stored out of the water (i.e. trailer boats) we modeled the fuel temperature
swing to be 65 percent of ambient and for boats stored in the water we used a fuel temperature
swing of 20 percent of ambient.  For example, on a 72-96°F day, we modeled fuel temperature to
be 76-92°F for a trailer boat and 82-86°F for a boat in the water. In our current modeling, we
treat portable tanks as exposed to ambient air.  We treat the smallest installed tanks as being on
trailer boats and the largest tanks as being on boats stored in the water. For the sizes in between
we use a linear interpolation of the fraction of boats stored in the water. We are interested in
collecting better information on the fractions of boats stored in and out of the water.

       Table 6.2.2-2 presents baseline and controlled emission factors for the proposed
certification test conditions and for a typical summer day with low vapor pressure fuel and a 50
percent full tank. (This comparison  is for illustrative purposes; as discussed above,  we modeled
daily temperature for 365 days over 6 regions of the U.S.)  Decreasing temperature  and fuel RVP
and increasing fill level all have the effect of reducing the diurnal emission factor.  In addition,
Table 6.2.2-2 shows the effect of tank installation and boat location on the emission factors.
Effectively insulating the tank through installation in a boat or exposing the tank to a cooling
medium by storing the boat in the water reduce the fuel temperature swing compared to ambient.
This, in turn, reduces the emission factors compared to test conditions.

                                      Table 6.2.2-2
         Evaporative Emission Factors for Test Conditions and Typical Summer Day
Tank Type
Boat
Location
Certification Test Conditions
72-96°F, 9 RVP* Fuel, 40% fill
Typical Summer Day
60-84°F, 8 RVP* Fuel, 50% fill
Baseline Scenario (open vent)
exposed
installed
installed
trailer
water
1.50 g/gallon/day
0.90 g/gallon/day
0.26 g/gallon/day
0.55 g/gallon/day
0.34 g/gallon/day
0.10 g/gallon/day
Control Scenario (closed vent with 1.0 psi pressure relief valve)
exposed
installed
installed
trailer
water
1.09 g/gallon/day
0.49 g/gallon/day
0.00 g/gallon/day
0.39 g/gallon/day
0.19 g/gallon/day
0.00 g/gallon/day
       * Reid Vapor Pressure [psi]
                                          6-16

-------
       The proposed certification test procedure is designed to represent a day on which high
amounts of ozone formation would be likely.  These are generally the days when the need for
hydrocarbon reductions is the greatest, but the contribution due to evaporative emissions is the
highest. For automotive applications, where activated charcoal is used to collect hydrocarbons,
designing for a three day test under these conditions ensures that equivalent emission control will
be gained under less severe circumstances.  This is also true for bladder fuel tanks, and to  some
extent, insulated fuel tanks. For sealed tanks with limited flow orifices, less of a percent
emission reduction may be seen in-use than over the certification test. This is because the orifice
must be sized to prevent over-pressure of the system under worst case conditions, so it may not
be effective under typical conditions.  However, for sealed tanks with pressure relief valves (with
or without volume compensating bags), we would expect the percent hydrocarbon reduction in
use to be much better than  over the test procedure.  This is because a valve designed for limited
hydrocarbon breakthrough on the proposed test procedure may not have any breakthrough under
less severe conditions when less pressure is formed in the tank. Therefore, although the closed
vent with a 1 psi pressure relief valve results in little more than a 25 percent reduction in diurnal
emissions over the proposed test procedure, we estimate that this technology would result in
more than twice this reduction on average in use.  The effects of ambient conditions and fuel tank
exposure with an open vent and with a 1 psi valve in the vent are illustrated above in Table
6.2.2-2.

       6.2.2.1.2-Refueling

       We used the draft NONROAD model to determine the amount of fuel  consumed by
spark-ignition marine engines.  This draft model assumes annual operation of 34.8 hours for
outboards, 77.3 hours for personal watercraft, and 47.6 hours for SD/I. Table  6.2.2-3 presents
the fuel consumption estimates we used in our modeling.  For 1998, the draft NONROAD model
estimated that spark-ignition marine engines consumed about 1.3 billion gallons of gasoline.

                                      Table 6.2.2-3
           Fuel Consumption Estimates used in Refueling Calculations [kg/kW-hr]
Technology
carbureted two- stroke
fuel -injected two-stroke
carbureted four-stroke
fuel -injected four-stroke
OB/PWC
0.79
0.59
0.59
-
SD/I
	
0.43
0.34
       For the refueling emission calculations, we had to consider the differences in dispensed
fuel temperature for an above ground versus below ground fuel storage system.  At most marinas,
the fuel is stored above ground due to the water table level while most gas stations store their fuel
underground. For our calculations, we assumed that the dispensed fuel temperature at marinas


                                          6-17

-------
was the same as ambient temperature. Our understanding is that most trailer boats are filled up at
local gas stations, while larger boats need to fill up at marinas. To account for this we modeled
boats with engines greater than 37 kW as marina fill-ups while the rest we modeled as gas station
fill-ups. While this is a simplified approach, the results are fairly insensitive to this assumption.
Table 6.2.2-4 presents emission factors for two conditions.

                                      Table 6.2.2-4
                Refueling Emission Factors for Two Summer Day Conditions
Evaporative Control
underground fuel tank
marina (above ground)
84°F dispensed fuel.
78°F ambient temperature,
9RVPFuel, 20% fill
4.25 g/gallon
3.88g/gallon
72 °F dispensed fuel.
68°F ambient temperature,
9RVPFuel, 20% fill
3. 84 g/gallon
3.61 g/gallon
       * Reid Vapor Pressure
       6.2.2.1.3 - Permeation

       We are in the process of collecting information on permeation rates from marine fuel
tanks and hoses. Based on the data presented in Chapter 4, we developed the emission factors
presented in Table 6.2.2-5. Fuel tank permeation was based on emission rates at 40°C is
expressed in terms of g/gallon/day. Hose permeation was based on emission rates at 23°C and is
based on g/m2/day. To show the effect of temperature on permeation rates, we present emission
rates at both temperatures. For aluminum fuel tanks, we modeled no permeation through the
tank.

                                      Table 6.2.2-5
                     Fuel Tank and Hose Permeation Emission Factors
Material
Polyethylene fuel tanks
Fluorinated polyethylene fuel tank
Class 1 rubber hose (fuel lines)
Class 2 rubber hose (fill necks)
Multi-layer barrier hose
23 °C (73°F)
0.40 g/gal/day
0.02 g/gal/day
100g/m2/day
190g/m2/day
5 g/m2/day
40°C (104°F)
1.33 g/gal/day
0.07 g/gal/day
330 g/m2 /day
630 g/m2/day
16 g/m2/day
       Based on the above emission factors, and a distribution of fuel tank sizes, materials, and
hose lengths, we were able to estimate evaporative emissions due to permeation.  Our
distribution of fuel tank sizes is discussed later in this chapter and our distribution of plastic
versus aluminum tanks is based on the sales information presented in Chapter 2.  Our estimates
                                          6-18

-------
of hose lengths are based on the few boats that we have used in our testing; we are interested in
collecting more information on the length of fuel hoses used in boats.

       6.2.2.1.4-HotSoak

       Our estimate of the average hot soak emission level was 0.48 grams per hot soak event
for automotive sized fuel tanks (roughly 20 gallons on average). Based on the hours per year of
operation discussed above, we estimate about 25 hot soak events per boat per year.

       6.2.2.2 - Reductions Due to the Proposed Standard

       We anticipate that the proposed standards will result in about an 82 percent reduction in
total evaporative emissions from new boats.  This is based on reductions from new boats of 95
percent fuel tank permeation, 95 percent hose permeation, and 57 percent diurnal venting.  The
total percent reduction increases as older boats are replaced with boats meeting the proposed
standards. Refueling emissions are projected to decrease at in early years due to better fuel
economy through the introduction of new engine technology (presented in Table 6.2.2-3).
However, projected sales growth eventually offsets this effect. Figure 6.2.2-1  and the following
tables present our emission inventory projections.

 Figure 6.2.2-1:  Projected National Evaporative HC from Boats Using SI Marine Engines
   re
   0)
   .>
   "35
   I
   •e
   o
   o
   I
   SS
   o
   Q.
   CO
140,000 -,

120,000 -

100,000

 80,000 -

 60,000 -

 40,000 -

 20,000 -

       0
• Baseline
• Control
              2000      2005      2010     2015     2020
                                         calendar year
                                                             2025
                                                  2030
                                     Table 6.2.2-6
                                         6-19

-------
               Projected Baseline Evaporative Emissions for Boats [short tons]
Calendar
Year
2000
2005
2010
2020
2030
Diurnal
22,700
23,600
24,500
26,300
28,100
Tank
Permeation
26,600
27,700
28,700
30,800
32,900
Hose
Permeation
43,200
44,900
46,500
49,900
53,300
Refueling
6,700
6,600
6,400
6,400
6,700
Hot Soak
260
270
280
300
320
Total
100,000
103,000
106,000
113,000
121,000
                                      Table 6.2.2-7
      Projected Controlled Evaporative Emissions Due to Proposed Standards [short tons]
Calendar
Year
2000
2005
2010
2020
2030
Diurnal
22,700
23,600
22,000
15,600
12,800
Tank
Permeation
26,600
27,700
23,800
10,000
3,300
Hose
Permeation
43,200
44,900
39,000
17,700
5,200
Refueling
6,700
6,600
6,400
6,400
6,700
Hot Soak
260
270
280
300
320
Total
100,000
103,000
91,400
49,900
28,300
                                      Table 6.2.2-8
     Projected Reductions in Evaporative Emissions Due to Proposed Standards [short tons]
Calendar
Year
2010
2020
2030
Diurnal

2,540
10,700
15,300
Tank
Permeation
4,930
20,800
29,600
Hose
Permeation
7,590
32,200
48,100
Total

15,100
63,800
93,000
Percent
Reduction
14%
56%
77%
       6.2.2.3 - Per Boat Evaporative Emissions

       In developing the cost per ton estimates in Chapter 7, we need to know the lifetime
emissions per boat.  The lifetime emissions are based on the projected lives of 21 years, 10 years,
and 20 years for OB, PWC, and SD/I respectively. We determine annual per boat evaporative
emissions by dividing the total annual evaporative emissions for 2000 in by the boat populations
shown in Table 6.2.2-1  (grown to 2000).  Using this approach allows us to essentially weight the
per boat emission factors for variation in tank sizes and materials. Per boat emission reductions
are based on the modeling described above. Table 6.2.2-9 presents these results with and without
the consideration of a 7 percent per year discount on the value of emission reductions.
                                          6-20

-------
                                      Table 6.2.2-9
                   Typical Lifetime Evaporative Emissions Per Boat (tons)
Engine
Class
OB
PWC
SD/I
Baseline
Undiscounted
0.145
0.068
0.311
All Boats U 0.158
Discounted
0.080
0.051
0.176
0.091
Control
Undiscounted
0.027
0.017
0.047
0.028
Discounted
0.015
0.013
0.027
0.016
Reduction
Undiscounted
0.118
0.051
0.264
0.130
Discounted
0.065
0.039
0.150
0.074
6.2.3 - Sterndrive and Inboard Marine

       We projected the annual tons of exhaust HC, CO, and NOx from SD/I marine engines
using the draft NONROAD model discussed above.  This section describes inputs to the
calculations that are specific to SD/I marine engines then presents the results. Even though we
are not proposing to regulate these engines in this rulemaking, we intend to revisit these engines
once in a future rulemaking in the time frame of the first technology review to be performed by
the California Air Resources Board (2005). Therefore, we present the baseline inventory here
and perform an  analysis of what the emission reductions could be if we were to finalize catalyst-
based or EGR-based standards in the future.

       6.2.3.1 - Inputs for the Inventory Calculations

       Several usage inputs are specific to the calculations for SD/I exhaust emissions. These
inputs are load factor, annual use, average operating life, and population.  Based on data
collected in developing the draft NONROAD model, we use a load factor of 20.7 percent, an
annual usage factor of 47.6 hours, and an average operating life 20 years for all SD/I engines.
The draft NONROAD model includes current and projected engine populations.  Table 6.2.3-1
presents these population estimates for selected years.

                                      Table 6.2.3-1
                      Projected Recreational SD/I Population by Year
Year
population
2000
1,870,000
2005
1,940,000
2010
2,010,000
2020
2,160,000
2030
2,310,000
       We used the data presented in Chapter 4 to develop the baseline emission factors.
Because the industry is currently making a transition from carburetors to electronic fuel injection
(EFI) for most SD/I engines, we present baseline emission factors for both technologies. We
                                          6-21

-------
included this phase-in of EFI technology in our baseline estimates. More detail on the emission
factors, deterioration factors, and phase-in of electronic fuel injection used in the draft
NONROAD model may be found in the docket.23 Table 6.2.3-2 presents the emission factors
used in this analysis for new engines and for engines deteriorated to the regulatory useful life (10
years).

                                      Table 6.2.3-2
                        Emission Factors for SD/I Marine Engines
Engine Technology
baseline carbureted
baseline EFI
HC [g/kW-hr]
new lOyrs
7.9 8.9
4.0 4.6
NOx [g/kW-hr]
new lOyrs
7.2 7.3
11.4 11.5
CO [g/kW-hr]
new lOyrs
210 240
96 110
       6.2.3.2 - Baseline Emissions from SD/I Marine Engines

       Table 6.2.3-3 presents our projected exhaust emission inventories for SD/I.  There is
some decrease in HC and CO over time due to the introduction of electronic fuel injection.
Through the use of electronic fuel injection, manufacturers are able to better control their fuel
metering which allows them to calibrate for leaner operation to get better fuel economy.
However, there  is an increase in NOx over time (beyond the projected growth) due  to the tradeoff
between calibrating an engine for HC and NOx. The net effect of these leaner calibrations on
combined HC+NOx is minimal in the range of air-fuel ratios that we expect manufacturers to use
in their designs.

                                      Table 6.2.3-3
             Projected Exhaust Emissions from SD/I Marine Engines [short tons]
Calendar Year
2000
2005
2010
2020
2030
HC
23,100
21,800
20,200
17,600
17,500
NOx
19,900
22,700
25,800
31,700
35,100
HC+NOx
43,000
44,400
46,000
49,300
52,600
CO
629,000
589,000
540,000
456,000
446,000
       6.2.3.3 - Analysis of Catalyst-based Approach

       The California Air Resources Board (CARB) is going ahead with catalyst-forcing
standards of 5 g/kW-hr HC+NOx.  These standards are contingent on a two technology reviews,
but are planned to be implemented on the following phase-in schedule: 45% in 2007, 75% in
2008, and 100% in 2009. In the near term, CARB also has a engine cap of 16 g/kW-hr HC+NOx
                                          6-22

-------
in 2005 for SD/I marine engines.  Although we are not proposing these standards today, we will
participate in the technology reviews and consider setting similar standards nationally based on
data from these reviews and other sources. This section presents our analysis of what the
emission benefits would be if the California approach was implemented nationally.

       For the emission factors under the cap, we assumed that the manufacturers would design
their engines to meet the proposed standard at regulatory useful life with a small compliance
margin beginning with the 2005 model year.  To determine the HC and NOx split for a standard
of 16 g/kW-hr HC+NOx, we used the HC and NOx data presented in Chapter 4 from SD/I
engines that are near this level.  Also, we used the deterioration factors in the draft NONROAD
model.
       For the catalyst control scenario, we considered the CARB standards and implementation
schedule.  To determine the HC and NOx levels individually under the combined standard, we
used the data presented in Chapter 4 from catalysts on SD/I engines. Consistent with our
modeling of Large SI catalyst-based standards, we assumed a compliance margin of 10 percent.
However, we used deterioration factors based on the rates in the MOBILE model for trucks with
catalysts. Table 6.2.3-4 presents the emission factors used in this analysis for new engines and
for engines deteriorated to the regulatory useful life (10 years).

                                      Table 6.2.3-4
                  Emission Factors for SD/I Marine Engines with Catalysts
Engine Technology
emission calibrated EFI
marine catalyst
HC [g/kW-hr]
new lOyrs
4.0 4.6
1.7 2.3
NOx [g/kW-hr]
new lOyrs
9.8 9.9
2.1 2.3
CO [g/kW-hr]
new lOyrs
96 110
83 92
       The following three tables present our projected impacts of a catalyst-based approach.
We believe that the application of a catalyst will result in a reduction of 67% HC+NOx from new
engines compared to baseline. We also anticipate that there would also be some CO benefits
from using a catalyst.

                                      Table 6.2.3-5
 Projected Exhaust HC Reductions for SD/I Marine Engines with Catalyst Approach [short tons]
Calendar Year
2000
2005
2010
2020
2030
Baseline
23,100
21,800
20,200
17,600
17,500
Control
23,100
21,800
19,200
12,700
8,720
Reduction
0
0
1,040
4,920
8,750
% Reduction
0%
0%
5%
28%
50%
                                          6-23

-------
                                      Table 6.2.3-6
    Projected NOx Reductions for SD/I Marine Engines with Catalyst Approach [short tons]
Calendar Year
2000
2005
2010
2020
2030
Baseline
19,900
22,700
25,800
31,700
35,100
Control
19,900
22,400
21,200
14,700
8,570
Reduction
0
200
4,610
17,000
26,600
% Reduction
0%
1%
18%
54%
76%
                                      Table 6.2.3-7
     Projected CO Reductions for SD/I Marine Engines with Catalyst Approach [short tons]
Calendar Year
2000
2005
2010
2020
2030
Baseline
629,000
589,000
540,000
456,000
446,000
Control
629,000
589,000
534,000
427,000
393,000
Reduction
0
0
5,920
28,800
52,500
% Reduction
0%
0%
1%
6%
12%
       6.2.3.4 - Analysis of EGR-based Approach

       Because the California standards are subject to two technology reviews, it is possible that
the catalyst-based standards may not go into place in the 2007-2009 time frame.  If that were the
case, an alternative to catalyst-based emissions reductions could be achieved through the use of
exhaust gas recirculation (EGR). This section presents our analysis of what the emission benefits
would be if an EGR-based approach was implemented nationally in 2009.

       For the EGR control scenario, we use a standard of 10 g/kW-hr HC+NOx in 2009.  To
determine the HC and NOx levels individually under the combined standard, we  used the data
presented in Chapter 4 from EGR on SD/I engines.  We used the same compliance margin and
multiplicative deterioration rates as the above analysis.  Table 6.2.3-8 presents the emission
factors used in this analysis for new engines and for engines deteriorated to the regulatory useful
life (10 years).

                                     Table 6.2.3-8
                   Emission Factors for SD/I Marine Engines with EGR
Engine Technology
exhaust gas recirculation
HC [g/kW-hr]
new lOyrs
3.3 3.7
NOx [g/kW-hr]
new lOyrs
5.4 5.5
CO [g/kW-hr]
new lOyrs
96 110
                                          6-24

-------
       The following two tables present our projected impacts of an EGR-based approach on HC
and NOx.  We believe that the application of EGR would result in a reduction of nearly 40%
HC+NOx from new engines compared to baseline.  We would not expect any significant changes
in CO due to the use of EGR.

                                    Table 6.2.3-9
  Projected Exhaust HC Reductions for SD/I Marine Engines with EGR Approach [short tons]
Calendar Year
2000
2005
2010
2020
2030
Baseline
23,100
21,800
20,200
17,600
17,500
Control
23,100
21,800
20,000
16,200
14,800
Reduction
0
0
200
1,400
2,700
% Reduction
0%
0%
1%
8%
15%
                                    Table 6.2.3-10
     Projected NOx Reductions for SD/I Marine Engines with EGR Approach [short tons]
Calendar Year
2000
2005
2010
2020
2030
Baseline
19,900
22,700
25,800
31,700
35,100
Control
19,900
22,700
24,200
22,000
18,500
Reduction
0
0
1,600
9,600
16,700
% Reduction
0%
0%
6%
30%
47%
                                        6-25

-------
Chapter 6 References
1. "Updated Spreadsheet for Modeling the Emission Inventories from On-Highway Motorcycles
>50cc under the Proposed Rule," EPA memo from Phil Carlson to Docket A-2000-01, April 12,
2002, Document IV-B-10.

2. "Spreadsheet for Modeling the Emission Inventories from On-Highway Motorcycles <50cc
under the Proposed Rule," EPA memo from Phil Carlson to Docket A-2000-01, April 12, 2002,
Document IV-B-09.

3.  "Draft Evaporative Emission Calculations for SI Marine, Snowmobiles, Nonroad
Motorcycles, All Terrain Vehicles, and Highway Motorcycles," EPA memo from Mike Samulski
to Docket A-2001, April 23, 2002, Document IV-B-08.

4. API Publication No. 4278, "Summary and Analysis of Data from Gasoline Temperature
Survey Conducted at Service Stations by American Petroleum Institute," Prepared by Radian
Corporation for American Petroleum Institute, November 11, 1976, Docket A-2000-01,
Document II-A-16.

5. D. T. Wade,  "Factors Influencing Vehicle Evaporative Emissions," SAE Paper 670126, 1967,
Docket A-2000-01, Document II-A-59.

6. Wade et. al.,  "Mathematical Expressions Relating Evaporative Emissions from Motor
Vehicles without Evaporative Loss-Control Devices to Gasoline Volatility," SAE Paper 720700,
1972, Docket A-2000-01, Document II-A-58.

7. S. Raghuma Reddy, "Prediction of Fuel Vapor Generation from a Vehicle Fuel Tank as a
Function of Fuel RVP and Temperature," SAE Paper 892089,  1989, Docket A-2000-01,
Document II-A-61.

8.  "Final Regulatory Impact Analysis: Refueling Emission Regulations for Light Duty Vehicles
and Trucks and Heavy Duty Vehicles," U.S. EPA, January 1994,  Docket A-2000-01, Document
II-A-79.

9. Nulman, M., Olejnik, A., Samus, M., Fead, E., Rossi,  G., "Fuel Permeation Performance of
Polymeric Materials," SAE Paper 2001-01-1999, 2001, Docket A-2000-01, Document No. IV-A-
23.

10.  Lockhart, M., Nulman, M, Rossi, G., "Estimating Real Time Diurnal Permeation from
Constant Temperature Measurements," SAE Paper 2001-01-0730, 2001, Docket A-2000-01,
Document No. IV-A-21.

11.  "Control of Vehicular Evaporative Emissions, Final Regulatory Impact Analysis and
Summary and Analysis of Comments," US EPA, February 1993, Docket A-2000-01, Document

                                       6-26

-------
No. IV-A-14.

12. U.S. EPA, "Calculation of Age Distributions in the Nonroad Model: Growth and
Scrappage," Report No. NR-007, February 19, 1998, Docket A-2000-01, Document II-A-69.

13. "Information on On-Highway Motorcycle Mileage Accumulation Rates and Survival
Fractions Supplied by the Motorcycle Industry Council," EPA memo from Phil Carlson to
Docket A-2000-01, September 12, 2001, Docket A-2000-01, Document U-B-22.

14. "Additional Information on Mileage Accumulation, Survival Fraction, and Sales Supplied by
the Motorcycle Industry Council for On-Highway Motorcycles <50cc," EPA memo from Phil
Carlson to Docket A-2000-01, April 12, 2002, Document IV-B-11.

15. "Procedures for Developing Base Year and Future Year Mass and Modeling Inventories for
the Heavy-Duty Engine and Vehicle Standards and Highway Diesel Fuel Rulemaking," U.S.
Environmental Protection Agency, April 2000. Pages VI-2 and VI-15, Docket A-2000-01,
Document II-A-76.

16. "Air Pollution from Motor Vehicles, Standards and Technologies for Controlling
Emissions," prepared by Asif Faiz, Christopher Weaver, and Michael Walsh for the World Bank,
1996, Docket A-2000-01, Document IV-A-29.

17. "Exhaust Pollution Abatement Technologies and the Requirements for a world-wide
Motorcycle Emissions Test Cycle," Rudolf Rijkeboer and Cornells Havenith.

18. "Information on On-Highway Motorcycle Mileage Accumulation Rates and Survival
Fractions Supplied by the Motorcycle Industry Council," EPA memo from Phil Carlson to
Docket A-2000-01, September 12, 2001, Docket A-2000-01, Document U-B-22.

19. "Additional Information on Mileage Accumulation, Survival Fraction, and Sales Supplied by
the Motorcycle Industry Council for On-Highway Motorcycles <50cc," EPA memo from Phil
Carlson to Docket A-2000-01, April 12, 2002, Document IV-B-11.

20. Dolce, G., "Nonroad Engine Growth Estimates," Report No. NR-008, U.S. EPA, March 6,
1998, Docket A-2000-01, Document IV-A-10.

21. National Marine Manufacturers Association, "Boating 2000; Facts & Figures at a Glance,"
Prepared by the Marketing  Statistics Department, 2000, Docket A-2000-01,  Document II-A-95.

22.  "Boat Show on CD," - available at www.boatshow.com/CD-ROM.html.

23."Revisions to the June 2000 Release of NONROAD to Reflect New Information and Analysis
on Marine and Industrial Engines," EPA memorandum from Mike Samulski to Docket A-98-01 ,
November 2, 2000, Docket A-2000-01, Document II-B-08.
                                        6-27

-------
                        CHAPTER 7 Cost Per Ton

7.1 - Cost Per Ton by Engine Type

7.1.1 - Introduction

       This chapter presents our estimate of the cost per ton of the standards contained in this
proposal. The analysis relies on the costs estimates presented in Chapter 5 and the estimated
lifetime emissions reductions using the information presented in Chapter 6.  The chapter also
presents a summary of the cost per ton of other recent EPA mobile source rulemakings for
comparison purposes.  Finally, this chapter presents the estimated costs and emission reductions
as incurred  over the first twenty years after the proposed standards are implemented.

       In calculating net present values that were used in our cost-per-ton estimates, we used a
discount rate of 7 percent, consistent with the 7 percent rate reflected in the cost-per-ton analyses
for other recent mobile source  programs.  OMB Circular A-94 requires us to generate benefit and
cost estimates reflecting a 7  percent rate. Using the 7 percent rate allows us to make direct
comparisons of cost-per-ton estimates with estimates for other, recently adopted, mobile source
programs.

       However, we consider that the cost and cost-per-ton estimates for future proposed mobile
source programs could reflect a 3 percent rate.  The 3  percent rate is in the 2 to 3 percent range
recommended by the Science Advisory Board's Environmental Economics Advisory Committee
for use in EPA social benefit-cost analyses,  a recommendation incorporated in EPA's new
Guidelines for Preparing Economic Analyses (November 2000).  Therefore, we have also
calculated the overall cost-effectiveness of today's rule based on a 3 percent rate to facilitate
comparison of the cost-per-ton of this rule with future proposed rules which might use the 3
percent rate. The results using both a 3 percent and 7 percent discount rate are provided in this
Chapter.

7.1.2 - Evaporative Emission Control from Boats

       This section provides our estimate of the cost per ton of evaporative emissions reduced
for boats. The analysis relies on the per vessel costs estimated in Chapter 5 and the estimated
lifetime emissions reductions (tons) presented in Chapter 6.  All costs and emission reductions
are discounted to the year of sale of the boats at a rate of 7 percent. Table 7.1.2-1 presents the
cost per ton with and without consideration of the significant fuel savings that will result from
evaporative emission control assuming a 7 percent discount rate.  The cost per ton results
assuming a 3 percent discount  rate are presented in Table 7.1.2-2. As shown in these tables, the
fuel savings more than offset the cost of the evaporative emission control technology.
                                          7-1

-------
                                     Table 7.1.2.-1
                 Estimated Marine Vessel Cost Per Ton of HC Reduced
                                (7 percent discount rate)

dirunal
tank permeation
hose permeation
aggregate
Total
Cost Per
Boat
$9
$12
$14
$36
Lifetime Fuel
Savings Per
Boat (NPV)
($4)
($9)
($14)
($27)
Lifetime
Reductions
Per Boat
(NPV tons)
0.012
0.023
0.039
0.074
Discounted Per Boat
Cost Per Ton without
Fuel Savings
($/ton)
$745
$523
$367
$478
Discounted Per Boat
Cost Per Ton with
Fuel Savings
($/ton)
$382
$160
$4
$115
                                     Table 7.1.2.-2
                 Estimated Marine Vessel Cost Per Ton of HC Reduced
                                (3 percent discount rate)

dirunal
tank permeation
hose permeation
aggregate
Total
Cost Per
Boat
$9
$12
$14
$36
Lifetime Fuel
Savings Per
Boat (NPV)
($6)
($11)
($19)
($36)
Lifetime
Reductions
Per Boat
(NPV tons)
0.016
0.031
0.052
0.100
Discounted Per Boat
Cost Per Ton without
Fuel Savings
($/ton)
$555
$390
$273
$356
Discounted Per Boat
Cost Per Ton with
Fuel Savings
($/ton)
$192
$27
($90)
($7)
7.1.3 - On-Highway Motorcycles

       This section provides our estimate of the cost per ton of emissions reduced for on-
highway motorcycles. For on-highway motorcycles <50 cc, we have calculated cost per ton on
the basis of HC only because we are only proposing an HC standard. For on-highway
motorcycles >50 cc, we have calculated cost per ton on the basis of HC plus NOx because we are
proposing HC plus NOx standards. The analysis relies on the per vehicle costs estimated in
Chapter 5 and the estimated net present value of the per vehicle lifetime emissions reductions
(tons) presented in Chapter 6.

       Table 7.1.3.-1  presents the cost per ton estimates for the proposed standards  for on-
highway motorcycles  <50cc. As described in Chapter 5, we expect a decrease in operating costs
(i.e., decreased fuel costs) as manufacturers  convert from 2-stroke to 4-stroke designs to meet the
proposed standards. Therefore, Table 7.1.3-1 presents cost per ton estimates both without and
                                          7-2

-------
with the fuel savings included for both a 7 percent and 3 percent discount rate.
                                    Table 7.1.3.-1
         Estimated On-Highway Motorcycle <50cc Cost Per Ton of HC Reduced
Discount
Rate
7%
3%
Cost per
Vehicle
(NPV)
$43.50
$43.50
Lifetime
Fuel Savings
(NPV)
$7.70
$8.50
Lifetime
Reductions
(NPV, tons)
0.020
0.022
Discounted Per Vehicle
Cost Per Ton without
Fuel Savings
($/ton)
$2,130
$1,950
Discounted Per Vehicle
Cost Per Ton with Fuel
Savings
($/ton)
$1,750
$1,570
       As described in Chapter 5, the estimated per vehicle costs for on-highway motorcycles
>50cc change over time, with reduced costs in the long term. We have estimated both a near-
term and long-term  cost per ton for both the Tier 1 and Tier 2 standards, with the Tier 2 estimates
incremental to Tier  1.  The results of the analysis are presented in Table 7.1.3.-2 assuming a 7
percent discount rate.  The cost per ton results assuming a 3 percent discount rate as presented in
Table 7.1.3.-3.

                                    Table 7.1.3.-2
       Estimated On-Highway Motorcycle >50cc Cost Per Ton of HC+NOx Reduced
                               (7 percent discount rate)
Standard
Tier 1 - Near-term
Tier 1 - Long-term
Tier 2 - Near-term
Tier 2 - Long-term
Cost per Vehicle
(NPV)
$26
$17
$35
$22
Lifetime Reductions
(NPV tons)
0.027
0.029
Discounted Per Vehicle Cost Per Ton
($/ton)
$970
$640
$1,230
$770
                                    Table 7.1.3.-3
      Estimated On-Highway Motorcycle >50cc Cost Per Ton of HC+NOx Reduced
                               (3 percent discount rate)
Standard
Tier 1 - Near-term
Tier 1 - Long-term
Tier 2 - Near-term
Tier 2 - Long-term
Cost per Vehicle
(NPV)
$26
$17
$35
$22
Lifetime Reductions
(NPV tons)
0.034
0.036
Discounted Per Vehicle Cost Per Ton
($/ton)
$770
$510
$980
$620
                                         7-3

-------
7.2 - Cost Per Ton for Other Mobile Source Control Programs

       Because the primary purpose of cost-effectiveness is to compare our program to
alternative programs, we made a comparison between the cost per ton values presented in this
chapter and the cost-effectiveness of other programs.  Table 7.2-1 summarizes the cost
effectiveness of several recent EPA actions for controlled emissions from mobile sources.

                                      Table 7.2-1
                      Cost-effectiveness of Previously Implemented
               Mobile Source Programs (Costs Adjusted to 2001 Dollars)1
                    Program
                    Tier 2 vehicle/gasoline sulfur
                    2007 Highway HD diesel
                    2004 Highway HD diesel
                    Off-highway diesel engine
                    Tier 1 vehicle
                    NLEV
                    Marine SI engines
                    On-board diagnostics
                    Marine CI engines	
    mon
1,437-2,423
 1,563-2,002
  227 - 444
  456 - 724
2,202 - 2,993
   2,069
1,255 - 1,979
   2,480
  26- 189
       By comparing the cost per ton values presented in earlier in this chapter to those in Table
7.2-1, we can see that the cost effectiveness of the proposed standards for this rulemaking are in
or below the range of these other programs. It is true that some previous programs have been
more cost efficient than the motorcycle program we are proposing today. However, it should be
expected that the next generation of standards will be more expensive than the last, because the
least costly means for reducing emissions is generally pursued first.

       The primary advantage of making comparisons to previously implemented programs is
that their cost-effectiveness values were based on a rigorous analysis and are generally accepted
as representative of the efficiency with which those programs reduce emissions. Unfortunately,
previously implemented programs can be poor comparisons because they may not be
representative of the cost-effectiveness of potential future programs. Therefore, in evaluating the
cost-effectiveness of our program, we also considered whether our proposal is cost-effective in
comparison with potential future means of controlling emissions.  In the context of the Agency's
rulemaking which would have revised the ozone and PM NAAQSP, the Agency compiled a list of
       p This rulemaking was remanded by the D.C. Circuit Court on May 14, 1999. However,
the analyses completed in support of that rulemaking are still relevant, since they were designed
                                          7-4

-------
additional known technologies that could be considered in devising new emission reductions
strategies.2 Through this broad review, over 50 technologies were identified that could reduce
NOx, VOC, or PM. The cost-effectiveness of these technologies averaged approximately
$5,000/ton for VOC, $13,000/ton for NOx, and $40,000/ton for PM.  Although a $10,000/ton
limit was actually used in the air quality analysis presented in the NAAQS revisions rule, these
values clearly indicate that, not only are future emission control strategies likely to be more
expensive (less  cost-effective) than past strategies, but the cost-effectiveness of our program falls
within the range of potential  future strategies.

       In summary, given the array of controls that will have to be implemented to make
progress toward attaining and maintaining the NAAQS, we believe that the weight of the
evidence from alternative means of providing substantial NOx + NMHC emission reductions
indicates that our proposed program is cost-effective.  This is true from the perspective of other
mobile source control programs or from the perspective of other stationary source technologies
that might be considered.

7.3 - 20-Year Cost and Benefit Analysis

       The following section presents the year-by-year cost and emission benefits associated
with the proposed standards for the 20-year period after implementation of the proposed
standards.  For the evaporative requirements for boats, where we expect a reduction in fuel
consumption due to the proposed standards, the fuel savings are presented separately. The
overall cost, incorporating the impact of the fuel savings is also presented.

       Table 7.3-1 presents the year-by-year cost and emission benefits for the proposed
evaporative emission controls from boats. (The numbers presented in Table 7.3-1  are not
discounted.)
to investigate the cost-effectiveness of a wide variety of potential future emission control
strategies.

                                           7-5

-------
                                       Table 7.3-1
   Cost and Emission Benefits of the Proposed Evaporative Emission Requirements for Boats
Year
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Evaporative HC
Benefits
(short tons)
4,981
10,008
15,051
20,093
25,130
30,173
35,176
40,177
45,139
50,050
54,781
59,310
63,758
67,965
71,954
75,543
78,823
81,837
84,675
87,295
Cost w/o
Fuel Savings
$24,352,501
$24,529,461
$24,706,422
$24,887,920
$25,069,418
$25,250,916
$25,432,414
$25,613,912
$25,793,142
$25,972,371
$26,151,600
$26,330,830
$26,510,059
$26,689,288
$26,868,517
$27,047,747
$27,226,976
$27,406,205
$27,586,569
$27,766,933
Fuel Savings
$1,808,029
$3,632,959
$5,463,449
$7,293,723
$9,122,336
$10,952,755
$12,768,878
$14,584,108
$16,385,542
$18,168,163
$19,885,677
$21,529,628
$23,144,174
$24,671,245
$26,119,472
$27,422,051
$28,612,596
$29,706,991
$30,736,918
$31,687,988
Cost w/
Fuel Savings
$22,544,472
$20,896,502
$19,242,973
$17,594,197
$15,947,082
$14,298,161
$12,663,536
$11,029,804
$9,407,600
$7,804,208
$6,265,923
$4,801,202
$3,365,885
$2,018,043
$749,045
($374,304)
($1,385,620)
($2,300,786)
($3,150,349)
($3,921,055)
       Table 7.3-2 presents the sum of the costs and emission benefits over the twenty year
period after the evaporative emission requirements for boats are proposed to take effect, on both
a non-discounted basis and a discounted basis (assuming a seven percent discount rate). The
annualized cost and emission benefits for the twenty-year period (assuming the seven percent
discount rate) are also presented.  It should be noted that these cost per ton figures are a little
higher than those presented above 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 2027 will not  achieve benefits until after 2027.
                                           7-6

-------
                                      Table 7.3-2
              Annualized Cost and Emission Benefits for the Period 2008-2027
              due to the Proposed Evaporative Emission Requirements for Boats

Undiscounted 20-
year Value
Discounted 20-
year Value
Annualized Value
Evaporative HC
Benefits
(short tons)
1,002,000
456,100
43,100
Cost w/o Fuel
Savings
(Million $)
$521.2
$290.9
$27.5
Fuel Savings
(Million $)
$363.7
$165.6
$15.6
Cost w/
Fuel Savings
(Million $)
$157.5
$125.4
$11.8
       Table 7.3.-3 presents the year-by-year cost and emission benefits for the proposed on-
highway motorcycle requirements. (The numbers presented in Table 7.3-3 are not discounted
and include the benefits and savings for all on-highway motorcycles, including those <50cc and
those >50cc.)
                                          7-7

-------
                                       Table 7.3-3
      Cost and Emission Benefits of the Proposed On-Highway Motorcycle Requirements
Year
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
909^
HC+NOx
Benefits
(short tons)
503
1,622
3,051
4,478
5,985
8,629
11,721
15,006
18,082
21,664
24,516
27,585
30,953
34,249
36,513
38,701
40,726
42,603
44,314
4^ Old
Cost w/o
Fuel Savings
$12,164,486
$12,283,740
$10,241,347
$10,341,370
$23,608,149
$23,601,656
$20,812,513
$21,019,555
$18,760,779
$18,838,943
$19,027,333
$19,217,606
$19,409,782
$19,603,880
$19,799,919
$19,997,918
$20,197,897
$20,399,876
$20,603,875

-------
                                      Table 7.3-4
              Annualized Cost and Emission Benefits for the Period 2006-2025
                 due to the Proposed On-Highway Motorcycle Requirements

Undiscounted 20-
year Value
Discounted 20-
year Value
Annualized Value
HC+NOx
Benefits
(short tons)
457,000
193,000
18,200
Cost w/o Fuel
Savings
(Million $)
$370.7
$199.6
$18.8
Fuel Savings
(Million $)
$4.7
$2.3
$0.2
Cost w/
Fuel Savings
(Million $)
$366.0
$197.3
$18.6
       Table 7.3-5 presents the aggregate year-by-year cost and emission benefits for both of the
programs contained in the proposal. (The numbers presented in Table 7.3-5 are not discounted.)
                                          7-9

-------
                                       Table 7.3-5
                  Cost and Emission Benefits of the Proposed Requirements
                    for All Equipment Categories covered by the Proposal
Year
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
HC+NOx
Benefits
(short tons)
503
1,622
8,032
14,486
21,036
28,722
36,851
45,179
53,258
61,841
69,655
77,635
85,734
93,559
100,271
106,666
112,680
118,146
123,137
127,751
132,089
136,056
Cost w/o
Fuel Savings
$12,164,486
$12,283,740
$34,593,848
$34,870,831
$48,314,571
$48,489,576
$45,881,931
$46,270,471
$44,193,193
$44,452,855
$44,820,475
$45,189,977
$45,561,382
$45,934,710
$46,309,978
$46,687,206
$47,066,414
$47,447,623
$47,830,851
$48,216,119
$48,604,582
$48,995,126
Fuel Savings
$24,176
$67,997
$1,914,753
$3,774,424
$5,634,793
$7,491,221
$9,340,087
$11,188,275
$13,020,693
$14,848,092
$16,659,643
$18,451,134
$20,175,230
$21,825,762
$23,446,889
$24,980,541
$26,435,349
$27,744,509
$28,941,635
$30,042,611
$31,079,119
$32,036,770
Cost w/
Fuel Savings
$12,140,310
$12,215,743
$32,679,096
$31,096,406
$42,679,777
$40,998,355
$36,541,844
$35,082,196
$31,172,500
$29,604,763
$28,160,832
$26,738,843
$25,386,153
$24,108,948
$22,863,089
$21,706,665
$20,631,065
$19,703,114
$18,889,216
$18,173,508
$17,525,463
$16,958,356
       Table 7.3-6 presents the sum of the costs and emission benefits over the twenty-two year
period after all of the requirements are proposed to take effect, on both a non-discounted basis
and a discounted basis (assuming a seven percent discount rate). The annualized cost and
emission benefits for the twenty-two year period (assuming the  seven percent discount rate) are
also presented.  (A twenty-two period is used in this aggregate analysis to cover the first twenty
years of each of the proposed standards which begins in 2006 for on-highway motorcycles and
concludes in 2008 for the proposed evaporative emission requirements for boats.)
                                          7-10

-------
                        Table 7.3-6
Annualized Cost and Emission Benefits for the Period 2006-2027
due to me Proposed Kec

Undiscounted 22-
year Value
Discounted 22-
year Value
Annualized Value
HC+NOx
Benefits
(short tons)
1,555,000
615,800
55,700
uirements tor All Equipment Categories
Cost w/o Fuel
Savings
(Million $)
$934.2
$464.3
$42.0
Fuel Savings
(Million $)
$369.1
$147.1
$13.3
Cost w/
Fuel Savings
(Million $)
$565.1
$317.2
$28.7
                           7-11

-------
Chapter 7 References
1.  Gross Domestic Product Implicit Price Deflator, U.S. Department of Commerce, Bureau of
Economic Analysis, http://www.stls.frb.org/fred/data/gdp/gdpdef, April 12, 2002, Docket A-
2000-01, Document IV-A-31.

2."Regulatory Impact Analyses for the Particulate Matter and Ozone National Ambient Air
Quality Standards and Regional Haze Rule," Appendix B, "Summary of control measures in the
PM, regional haze, and ozone partial attainment analyses," Innovative Strategies and Economics
Group, Office of Air Quality Planning and Standards, U.S. Environmental Protection Agency,
Research Triangle Park, NC, July 17, 1997, Docket A-2000-01, Document II-A-77.
                                         7-12

-------
      CHAPTER 8:  Initial Regulatory Flexibility Analysis

       This section presents our Initial Regulatory Flexibility Analysis (IRFA) which evaluates
the impacts of our proposed program on small businesses. In preparing this IRFA, we looked at
both the effect of this proposal for marine vessels and highway motorcycles and an earlier
proposal for nonroad large SI engines, recreational vehicles, and CI recreational marine engines
(October 5, 2001, 66 CFR 51098). Prior to issuing these proposals, we analyzed the potential
impacts of our program on small businesses. As a part of this analysis, we convened a Small
Business Advocacy Review Panel, as required under the Regulatory Flexibility Act as amended
by the Small Business Regulatory Enforcement Fairness Act of 1996 (SBREFA). Through the
Panel process, we gathered advice and recommendations from small entity representatives
(SERs) who would be affected by our proposed vehicle and fuel standards. The report of the
Panel has been placed in the rulemaking record.

8.1 - Requirements of the Regulatory Flexibility Act

       When proposing and promulgating rules subject to notice and comment under the Clean
Air Act, we are generally required under the Regulatory Flexibility Act (RFA) to conduct a
regulatory flexibility analysis unless we certify that the requirements of a regulation will not
cause a significant impact on a substantial number of small entities. The key elements of the
FRFA include:

             the number of affected small entities;

             the projected reporting, record keeping, and other compliance requirements of the
             proposed rule, including the classes of small entities that would be affected and
             the type of professional  skills necessary for preparation of the report or record;

             other federal rules that may duplicate, overlap, or conflict with the proposed rule;
             and,

       •      any significant alternatives to the proposed rule that accomplish the stated
             objectives of applicable statutes and which minimize significant economic
             impacts of the proposed rule on small entities.

       The RFA was amended by SBREFA to ensure that concerns regarding small  entities are
adequately considered during the development of new regulations that affect them. Although we
are not required by the CAA to provide special treatment to small businesses, the RFA requires
us to carefully consider the economic impacts that our rules will have on small entities.
Specifically, the RFA requires us to determine, to the extent feasible, our rule's economic impact
on small entities, explore regulatory options for reducing any significant economic impact on  a
substantial number of such entities, and explain our ultimate choice of regulatory approach.

                                         8-13

-------
       In developing the NPRM, we concluded that the program under consideration for
recreational vehicles would likely have a significant impact on a substantial number of small
entities.

8.2 - Description of Affected Entities

       The following table (Table 1) provides an overview of the primary SB A small business
categories potentially affected by this regulation.  EPA is in the process of developing a more
detailed industry characterization of the entities potentially subject to this regulation.

                                       Table 8.2-1
 Primary SBA Small Business Categories Potentially Affected by this Proposed Regulation
Industry
Motorcycles and motorcycle parts
manufacturers
Snowmobile and ATV
manufacturers
Independent Commercial
Importers of Vehicles and parts
Nonroad SI engines
Internal Combustion Engines
Boat Building and Repairing
Fuel Tank Manufacturers
NAICSa Codes
336991
336999
421110
333618
333618
336612
336211
Defined by SBA as a
Small Business If:b
<500 employees
<500 employees
<100 employees
< 1,000 employees
< 1000 employees
< 500 employees
<1000 employees
NOTES:
a. North American Industry Classification System
b. According to SBA's regulations (13 CFR 121), businesses with no more than the listed number of employees or
dollars in annual receipts are considered "small entities" for purposes of a regulatory flexibility analysis.
8.2.1 - Recreational Vehicles (off-highway motorcycles, ATVs, and snowmobiles)

       The ATV sector has the broadest assortment of manufacturers. There are seven
companies representing over 95 percent of total domestic ATV sales. The remaining 5 percent
come from importers who tend to import inexpensive, youth-oriented ATVs from China and
other Asian nations. EPA has identified 21 small companies (as defined in Table 4.1, above) that
offer off-road motorcycles, ATVs, or both products.  Annual unit sales for these companies can
range from a few hundred to several thousand units per year.
                                          8-14

-------
       Based on available industry information, four major manufacturers, Arctic Cat,
Bombardier (also known as Ski-Doo), Polaris, and Yamaha, account for over 99 percent of all
domestic snowmobile sales.  The remaining one percent comes from very small manufacturers
who tend to specialize in unique and high performance designs .

       We have identified three small manufacturers of snowmobiles and one potential small
manufacturer who hopes to produce snowmobiles within the next year. Two of these
manufacturers (Crazy Mountain and Fast), plus the potential newcomer (Redline) specialize in
high performance versions of standard recreational snowmobile types (i.e., travel and mountain
sleds).  The other manufacturer (Fast Trax) produces a unique design, which is a scooter-like
snowmobile designed to be ridden standing up. Most of these manufacturers build less than 50
units per year.

8.2.2 - Highway Motorcycles

       Of the numerous manufacturers supplying the U.S. market for highway motorcycles,
Honda, Harley Davidson, Yamaha, Kawasaki, Suzuki, and BMW are the largest, accounting for
95 percent or more of the total  U.S. sales. All of these companies except Harley-Davidson and
BMW also manufacture off-road motorcycles and ATVs for the U.S. market. Harley-Davidson
is the only company manufacturing highway motorcycles exclusively in the U.S. for the U.S.
market.

       Since highway motorcycles have had to meet emission  standards for the last twenty years,
EPA has good information on the number of companies that manufacture or market highway
motorcycles for the U.S. market in each model year.  In addition to the big six manufacturers
noted above, EPA finds as many as several dozen more companies that have operated in the U.S.
market in the last couple of model years.  Most of these are U.S. companies that are either
manufacturing or importing motorcycles, although a few are U.S. affiliates of larger companies
in Europe or Asia. Some of the U.S. manufacturers employ only a few people and produce only
a handful of custom motorcycles per year, while others may employ several hundred and produce
up to several thousand motorcycles per year.

8.2.3 - Marine Vessels

       Marine vessels include  the boat, engine, and fuel system.  The evaporative emission
controls discussed above may affect the boat builders and/or the fuel tank manufacturers.
Exhaust emission controls including NTE requirements, as addressed in the August 29, 1999
SBREFA Panel  Report, would affect the  engine manufacturers and may affect boat builders.
                                         8-15

-------
       8.2.3.1 - Small Recreational Boat Builders

       EPA has less precise information about recreational boat builders than is available about
engine manufacturers. EPA has utilized several sources, including trade associations and Internet
sites when identifying entities that build and/or sell recreational boats. EPA has also worked
with an independent contractor to assist in the characterization of this segment of the industry.
Finally, EPA has obtained a list of nearly 1,700 boat builders known to the U.S. Coast Guard to
produce boats using engines for propulsion.  At least 1,200 of these companies install engines
that use gasoline fueled engines and would therefore be subject to the evaporative emission
control program discussed above. More than 90 percent of the companies identified so far would
be considered small businesses as defined by SBA SIC code 3732. EPA continues to develop a
more complete picture of this segment of the industry and will provide additional information as
it becomes available.

       Based on information supplied by a variety of recreational boat builders, fuel tanks for
boats using SI marine engines are usually purchased from fuel tank manufacturers.  However,
some boat builders construct their own fuel tanks. The boat builder provides the specifications to
the fuel tank manufacturer who helps match  the fuel tank for a particular application. It is the
boat builder's responsibility to install the fuel tank and connections into their vessel design. For
vessels designed to be used with small outboard engines, the boat builder may not install a fuel
tank; therefore, the end user would use a portable fuel tank with a connection to the engine.

       8.2.3.2 - Small Marine Fuel Tank Manufacturers

       EPA has determined that total sales of tanks for gasoline marine applications is
approximately 550,000 units per year. The market is broken into manufacturers that produce
plastic tanks and manufacturers that produce aluminum tanks. EPA has determined that there are
at least seven companies that make plastic fuel tanks with total sales of approximately 440,000
units per year.  EPA has determined that there at least four companies that make aluminum fuel
tanks with total sales of approximately 110,000 units per year. All but one of these plastic and
aluminum fuel tank manufacturers is a small business as defined under SBA SIC Code 3713.

       8.2.3.3 - Small Diesel Engine Marinizers

       EPA has determined that there are at least 16 companies that manufacture CI diesel
engines for recreational vessels. Nearly 75 percent of diesel engines sales for recreational vessels
in 2000 can be attributed to three large companies.  Six of the 16 identified companies are
considered small businesses as  defined by SBA SIC code 3519. Based on sales estimates for
2000, these six companies represent approximately 4 percent of recreational marine diesel engine
sales. The remaining companies each comprise between two and  seven percent of sales for 2000.
                                          8-16

-------
       8.2.3.4 - Small Gasoline Engine Marinizers

       EPA has determined that there are at least 24 companies that manufacture SD/I gasoline
engines (including airboats and jet boats) for recreational vessels. Seventeen of the identified
companies are considered small businesses as defined by SBA SIC code 3519.  These 17
companies represent approximately 6 percent of recreational gasoline marine engines sales for
2000. Approximately 70-80 percent of gasoline SD/I engines manufactured in 2000  can be
attributed to one company.  The next largest company is responsible for about 10-20  percent of
2000 sales.

8.2.4 - Large Spark Ignition Engines

       The Panel is aware of one engine manufacturer of Large SI engines that qualifies as a
small business.  This company plans to produce engines that meet the standards adopted by
CARB in 2004, with the possible exception of one engine family.   If EPA adopts long-term
standards, this would require manufacturers to do additional calibration and testing work. If EPA
adopts new test procedures (including transient operation), there may also be a cost associated
with upgrading test facilities.

8.3 - Projected Costs of the Proposed Program

       The costs associated with the proposed program can be found in Chapter 5 of the Draft
Regulatory  Support Document. Chapter 5 includes a description of our approach to estimating
the cost of complying with emission standards. We start with a general  description of the
approach to estimating costs, then describe the technology changes we expect and assign costs to
them. We also present an analysis of the estimated aggregate cost to society.

8.4 - Projected Reporting, Recordkeeping, and Other Compliance
Requirements  of the Proposed Rule

       For  any emission control program,  EPA must have assurances that the regulated engines
will meet the standards.  Historically, EPA programs have included provisions placing
manufacturers responsible for providing these assurances.  The program that EPA is  considering
for manufacturers subject to this proposal may include testing, reporting, and record keeping
requirements. Testing requirements for some manufacturers may include certification (including
deterioration testing), and production line testing. Reporting requirements would likely include
test data and technical data on the engines  including defect reporting. Manufacturers would
likely have  to keep records of this information.

8.5 - Other Related Federal Rules

       We  are aware of several other current Federal rules that relate to the proposed rule under
development. During the Panel's outreach meeting, SERs specifically pointed to Consumer

                                         8-17

-------
Product Safety Commission (CPSC) regulations covering ATVs, and noted that they may be
relevant to crafting an appropriate definition for a competition exclusion in this category. The
Panel recommends that EPA continue to consult with the CPSC in developing a proposed and
final rule in order to better understand the scope of the Commission's regulations as they may
relate to the competition exclusion.

       Other SERs, representing manufacturers of marine engines, noted that the U.S. Coast
Guard regulates vessel tanks, most notably tank pressure and anti-siphoning requirements for
carburetted engines. Tank manufacturers would have to take these requirements into account in
designing evaporative control systems. The Panel recommends that EPA continue to work with
the Coast Guard to evaluate the safety implications of any proposed evaporative emissions
standards and to avoid interference with Coast Guard safety regulations.

       The Panel is also aware of other Federal rules that relate to the categories that EPA would
address with the proposed rule, but are not likely to affect policy considerations in the rule
development process. For example, there are now EPA noise standards covering off-road
motorcycles; however, EPA expects that most emission control devices are likely to reduce,
rather than increase, noise, and that therefore the noise standards are not likely to be important in
developing a proposed rule.

8.6 - Regulatory Alternatives

       The Panel considered a wide range of options and regulatory alternatives for providing
small businesses with flexibility in complying with the proposed emissions standards and related
requirements. As part of the process, the Panel requested and received comment on several ideas
for flexibility that were suggested by SERs  and Panel members.  The major options
recommended by the Panel can be found in Section 9 of the Panel's full Report.

       Many of the flexible approaches recommended by the Panel can be applied to several of
the equipment categories that would potentially be affected by the proposed rule EPA is
developing.  These approaches are identified in Table 1.  First Tier Flexibilities: Based on
consultations with SERs, the Panel believes that the first four provisions in Table 1  are likely to
provide the greatest flexibility for many small entities. These provisions are likely to be most
valuable because they either provide more time for compliance (e.g., additional leadtime  and
hardship  provisions) or allow for certification of engines based on particular engine designs or
certification to other EPA programs.  Second Tier Flexibilities: The remaining four approaches
have the potential to reduce near-term and even long-term costs once a small entity  has a product
it is preparing to certify. These are important in that the costs of testing multiple engine
families,  testing a fraction of the production line, and/or developing deterioration factors can be
significant. Small businesses could also meet an emission standard on average or generate
credits for producing engines which emit at levels below the standard; these credits could then be
sold to other manufacturers for compliance  or banked for use in future model years.
                                          8-18

-------
       During the consultation process, it became evident that, in a few situations, it could be
helpful to small entities if unique provisions were available. Five such provisions are described
below.

       (a) Snowmobiles: The Panel recommends EPA seek comment on a provision which
       would allow small snowmobile manufacturers to petition EPA for a relaxed standard for
       one or more engine families, up to 300 engines per year, until the family is retired or
       modified, if such a standard is justifiable based on the criteria described in the Panel
       report.

       (b) ATVs and Off-road Motorcycles: The Panel recommends that the hardship provision
       for ATVs and off-road motorcycles allow hardship relief to be reviewed annually for a
       period that EPA anticipates will likely be no more than two years in order for importers to
       obtain complying products.

       (c) Large SI: The Panel recommends that small entities be granted the flexibility initially
       to reclassify a small number of their small displacement engines into EPA's small spark-
       ignition engine program (40 CFR part 90).  Small entities would be allowed to use those
       requirements in lieu of the requirements EPA intends to propose for large entities.

       (d) Marine Vessel Tanks: Most of this sector involves small fuel tank manufacturers and
       small boat builders. The Panel recommends that the program be structured with longer
       lead times and an early credit generation program to enable the fuel tank manufacturers to
       implement controls on tanks on a schedule consistent with their normal turnover of fuel
       tank molds.

       (e) Highway Motorcycles: California ARB has found that California's Tier 2 standard is
       potentially infeasible for small manufacturers. Therefore, the Panel recommends that
       EPA delay making decisions on the applicability to small businesses of Tier 2 or other
       such revisions to the federal regulations until California's 2006 review is complete.

       Table 1  describes the flexibilities that the Panel is generally recommending for each of
the sectors where appropriate as indicated in the table.

       The Panel also crafted recommendations to address SERs' concerns that ATV and off-
road motorcycle standards that essentially required manufacturers to switch to four-stroke
engines might increase costs to the point that many small importers and manufacturers could
experience significant adverse effects.  The Panel recommends that EPA request comment in its
proposed rule on the effect of the proposed standard on these  small entities, with the specific
intent of developing information—including the extent to which sales of their products would
likely to be reduced in response to changes in product price attributable to the proposed
standards—that could be used to inform a decision in the final rule as to whether EPA should
provide additional flexibility beyond that considered by the Panel.
                                          8-19

-------