[Federal Register Volume 72, Number 96 (Friday, May 18, 2007)]
[Proposed Rules]
[Pages 28098-28393]
From the Federal Register Online via the Government Publishing Office [www.gpo.gov]
[FR Doc No: 07-1998]



[[Page 28097]]

-----------------------------------------------------------------------

Part II





Environmental Protection Agency





-----------------------------------------------------------------------



40 CFR Parts 60, 63, et al.



Control of Emissions from Nonroad Spark-Ignition Engines and Equipment; 
Proposed Rule

Federal Register / Vol. 72, No. 96 / Friday, May 18, 2007 / Proposed 
Rules

[[Page 28098]]


-----------------------------------------------------------------------

ENVIRONMENTAL PROTECTION AGENCY

40 CFR Parts 60, 63, 85, 89, 90, 91, 1027, 1045, 1048, 1051, 1054, 
1060, 1065, 1068, and 1074

[EPA-HQ-OAR-2004-0008; FRL-8303-7]
RIN 2060-AM34


Control of Emissions from Nonroad Spark-Ignition Engines and 
Equipment

AGENCY: Environmental Protection Agency (EPA).

ACTION: Proposed rule.

-----------------------------------------------------------------------

SUMMARY: We are proposing emission standards for new nonroad spark-
ignition engines that will substantially reduce emissions from these 
engines. The proposed exhaust emission standards would apply in 2009 
for new marine spark-ignition engines, including first-time EPA 
standards for sterndrive and inboard engines. The proposed exhaust 
emission standards would apply starting in 2011 and 2012 for different 
sizes of new land-based, spark-ignition engines at or below 19 
kilowatts (kW). These small engines are used primarily in lawn and 
garden applications. We are also proposing evaporative emission 
standards for vessels and equipment using any of these engines. In 
addition, we are making other minor amendments to our regulations. We 
estimate that by 2030, the proposed standards would result in 
significant annual reductions of pollutant emissions from regulated 
engine and equipment sources nationwide, including 631,000 tons of 
volatile organic hydrocarbon emissions, 98,200 tons of NOX 
emissions, and 6,300 tons of direct particulate matter 
(PM2.5) emissions. These reductions correspond to 
significant reductions in the formation of ground-level ozone. We also 
expect to see annual reductions of 2,690,000 tons of carbon monoxide 
emissions, with the greatest reductions in areas where there have been 
problems with individual exposures. The requirements in this proposal 
would result in substantial benefits to public health and welfare and 
the environment. We estimate that by 2030, on an annual basis, these 
emission reductions would prevent 450 PM-related premature deaths, 
approximately 500 hospitalizations, 52,000 work days lost, and other 
quantifiable benefits every year. The total estimated annual benefits 
of this rule in 2030 are approximately $3.4 billion. Estimated costs in 
2030 are many times less at approximately $240 million.

DATES: Comments: Comments must be received on or before August 3, 2007. 
Under the Paperwork Reduction Act, comments on the information 
collection provisions must be received by OMB on or before June 18, 
2007.

ADDRESSES: Submit your comments, identified by Docket No. EPA-HQ-OAR-
2004-0008, by one of the following methods:
    www.regulations.gov: Follow the on-line instructions for submitting 
comments.
    E-mail: [email protected].
    Fax: (202) 260-4400.
    Mail: Environmental Protection Agency, Air Docket, Mail-code 6102T, 
1200 Pennsylvania Ave., NW., Washington, DC 20460. In addition, please 
mail a copy of your comments on the information collection provisions 
to the Office of Information and Regulatory Affairs, Office of 
Management and Budget (OMB), Attn: Desk Officer for EPA, 725 17th St., 
NW., Washington, DC 20503.
    Hand Delivery: EPA Docket Center (EPA/DC), EPA West, Room 3334, 
1301 Constitution Ave., NW., Washington, DC, Attention Docket No. EPA-
HQ-OAR-2004-0008. Such deliveries are accepted only during the Docket's 
normal hours of operation, special arrangements should be made for 
deliveries of boxed information.
    Instructions: Direct your comments to Docket ID No. EPA-HQ-OAR-
2004-0008. EPA's policy is that all comments received will be included 
in the public docket without change and may be made available online at 
www.regulations.gov, including any personal information provided, 
unless the comment includes information claimed to be Confidential 
Business Information (CBI) or other information whose disclosure is 
restricted by statute. Do not submit information that you consider to 
be CBI or otherwise protected through www.regulations.gov or e-mail. 
The www.regulations.gov Web site is an ``anonymous access'' system, 
which means EPA will not know your identity or contact information 
unless you provide it in the body of your comment. If you send an e-
mail comment directly to EPA without going through www.regulations.gov, 
your e-mail address will be automatically captured and included as part 
of the comment that is placed in the public docket and made available 
on the Internet. If you submit an electronic comment, EPA recommends 
that you include your name and other contact information in the body of 
your comment and with any disk or CD-ROM you submit. If EPA cannot read 
your comment due to technical difficulties and cannot contact you for 
clarification, EPA may not be able to consider your comment. Electronic 
files should avoid the use of special characters, any form of 
encryption, and be free of any defects or viruses. For additional 
instructions on submitting comments, go to Unit XIII of the 
SUPPLEMENTARY INFORMATION section of this document.
    Docket: All documents in the docket are listed in the 
www.regulations.gov index. Although listed in the index, some 
information is not publicly available, such as CBI or other information 
whose disclosure is restricted by statute. Certain other material, such 
as copyrighted material, will be publicly available only in hard copy. 
Publicly available docket materials are available either electronically 
in www.regulations.gov or in hard copy at the ``Control of Emissions 
from Nonroad Spark-Ignition Engines, Vessels and Equipment'' Docket, 
EPA/DC, EPA West, Room 3334, 1301 Constitution Ave., NW., Washington, 
DC. The Public Reading Room is open from 8:30 a.m. to 4:30 p.m., Monday 
through Friday, excluding legal holidays. The telephone number for the 
Public Reading Room is (202) 566-1744 and the telephone number for the 
``Control of Emissions from Nonroad Spark-Ignition Engines, Vessels, 
and Equipment'' Docket is (202) 566-1742.
    Hearing: A hearing will be held at 9:30 a.m. on Tuesday, June 5, 
2007 at the Sheraton Reston Hotel. The hotel is located at 11810 
Sunrise Valley Drive in Reston, Virginia; their phone number is 703-
620-9000. For more information on these hearings or to request to 
speak, see Section XIII.

FOR FURTHER INFORMATION CONTACT: Carol Connell, Environmental 
Protection Agency, Office of Transportation and Air Quality, Assessment 
and Standards Division, 2000 Traverwood Drive, Ann Arbor, Michigan 
48105; telephone number: 734-214-4349; fax number: 734-214-4050; e-mail 
address: [email protected].

SUPPLEMENTARY INFORMATION:

Does This Action Apply to Me?

    This action will affect you if you produce or import new spark-
ignition engines intended for use in marine vessels or in new vessels 
using such engines. This action will also affect you if you produce or 
import new spark-ignition engines below 19 kilowatts used in nonroad 
equipment, including agricultural and construction equipment, or 
produce or import such nonroad vehicles.

[[Page 28099]]

    The following table gives some examples of entities that may have 
to follow the regulations; however, since these are only examples, you 
should carefully examine the proposed regulations. Note that we are 
proposing minor changes in the regulations that apply to a wide range 
of products that may not be reflected in the following table (see 
Section XI). If you have questions, call the person listed in the FOR 
FURTHER INFORMATION CONTACT section of this preamble:

----------------------------------------------------------------------------------------------------------------
                                                                               Examples of potentially regulated
                Category                    NAICS codes a      SIC codes b                 entities
----------------------------------------------------------------------------------------------------------------
Industry                                            333618              3519  Manufacturers of new engines.
Industry                                            333111              3523  Manufacturers of farm machinery
                                                                               and equipment.
Industry                                            333112              3524  Manufacturers of lawn and garden
                                                                               tractors (home).
Industry                                            336612        3731, 3732  Manufacturers of marine vessels.
Industry                                    811112, 811198        7533, 7549  Commercial importers of vehicles
                                                                               and vehicle components.
----------------------------------------------------------------------------------------------------------------
a North American Industry Classification System (NAICS).
b Standard Industrial Classification (SIC) system code.

What Should I Consider as I Prepare My Comments for EPA?

    Submitting CBI. Do not submit this information to EPA through 
www.regulations.gov or e-mail. Clearly mark the part or all of the 
information that you claim to be CBI. For CBI information in a disk or 
CD ROM that you mail to EPA, mark the outside of the disk or CD ROM as 
CBI and then identify electronically within the disk or CD ROM the 
specific information that is claimed as CBI. In addition to one 
complete version of the comment that includes information claimed as 
CBI, a copy of the comment that does not contain the information 
claimed as CBI must be submitted for inclusion in the public docket. 
Information so marked will not be disclosed except in accordance with 
procedures set forth in 40 CFR part 2.
    Tips for Preparing Your Comments. When submitting comments, 
remember to:
     Identify the rulemaking by docket number and other 
identifying information (subject heading, Federal Register date and 
page number).
     Follow directions--The agency may ask you to respond to 
specific questions or organize comments by referencing a Code of 
Federal Regulations (CFR) part or section number.
     Explain why you agree or disagree; suggest alternatives 
and substitute language for your requested changes.
     Describe any assumptions and provide any technical 
information and/or data that you used.
     If you estimate potential costs or burdens, explain how 
you arrived at your estimate in sufficient detail to allow for it to be 
reproduced.
     Provide specific examples to illustrate your concerns and 
suggest alternatives.
     Explain your views as clearly as possible, avoiding the 
use of profanity or personal threats.
     Make sure to submit your comments by the comment period 
deadline identified.

Table of Contents

I. Introduction
    A. Overview
    B. Why Is EPA Taking This Action?
    C. What Regulations Currently Apply to Nonroad Engines or 
Vehicles?
    D. Putting This Proposal into Perspective
    E. What Requirements Are We Proposing?
    F. How Is This Document Organized?
II. Public Health and Welfare Effects
    A. Ozone
    B. Particulate Matter
    C. Air Toxics
    D. Carbon Monoxide
III. Sterndrive and Inboard Marine Engines
    A. Overview
    B. Engines Covered by This Rule
    C. Proposed Exhaust Emission Standards
    D. Test Procedures for Certification
    E. Additional Certification and Compliance Provisions
    F. Small-Business Provisions
    G. Technological Feasibility
IV. Outboard and Personal Watercraft Engines
    A. Overview
    B. Engines Covered by This Rule
    C. Proposed Exhaust Emission Standards
    D. Changes to Existing OB/PWC Test Procedures
    E. Additional Certification and Compliance Provisions
    F. Other Adjustments to Regulatory Provisions
    G. Small-Business Provisions
    H. Technological Feasibility
V. Small SI Engines
    A. Overview
    B. Engines Covered by This Rule
    C. Proposed Requirements
    D. Testing Provisions
    E. Certification and Compliance Provisions for Small SI Engines 
and Equipment
    F. Small Business Provisions
    G. Technological Feasibility
VI. Evaporative Emissions
    A. Overview
    B. Fuel Systems Covered by This Rule
    C. Proposed Evaporative Emission Standards
    D. Emission Credit Programs
    E. Testing Requirements
    F. Certification and Compliance Provisions
    G. Small-Business Provisions
    H. Technological Feasibility
VII. General Concepts Related to Certification and Other 
Requirements
    A. Scope of Application
    B. Emission Standards and Testing
    C. Demonstrating Compliance
    D. Other Concepts
VIII. General Nonroad Compliance Provisions
    A. Miscellaneous Provisions (Part 1068, subpart A)
    B. Prohibited Acts and Related Requirements (Part 1068, subpart 
B)
    C. Exemptions (Part 1068, subpart C)
    D. Imports (Part 1068, subpart D)
    E. Selective Enforcement Audit (Part 1068, subpart E)
    F. Defect Reporting and Recall (Part 1068, subpart F)
    G. Hearings (Part 1068, subpart G)
IX. General Test Procedures
    A. Overview
    B. Special Provisions for Nonroad Spark-Ignition Engines
X. Energy, Noise, and Safety
    A. Safety
    B. Noise
    C. Energy
XI. Proposals Affecting Other Engine and Vehicle Categories
    A. State Preemption
    B. Certification Fees
    C. Amendments to General Compliance Provisions in 40 CFR Part 
1068
    D. Amendments Related to Large SI Engines (40 CFR Part 1048)
    E. Amendments Related to Recreational Vehicles (40 CFR Part 
1051)
    F. Amendments Related to Heavy-Duty Highway Engines (40 CFR Part 
85)
    G. Amendments Related to Stationary Spark-Ignition Engines (40 
CFR Part 60)
XII. Projected Impacts
    A. Emissions from Small Nonroad and Marine Spark-Ignition 
Engines
    B. Estimated Costs
    C. Cost per Ton
    D. Air Quality Impact
    E. Benefits
    F. Economic Impact Analysis
XIII. Public Participation
XIV. Statutory and Executive Order Reviews
    A. Executive Order 12866: Regulatory Planning and Review
    B. Paperwork Reduction Act
    C. Regulatory Flexibility Act
    D. Unfunded Mandates Reform Act
    E. Executive Order 13132: Federalism

[[Page 28100]]

    F. Executive Order 13175: Consultation and Coordination With 
Indian Tribal Governments
    G. Executive Order 13045: Protection of Children from 
Environmental Health and Safety Risks
    H. Executive Order 12898: Federal Actions to Address 
Environmental Justice in Minority Populations and Low-Income 
Populations.
    I. Executive Order 13211: Actions that Significantly Affect 
Energy Supply, Distribution, or Use
    J. National Technology Transfer Advancement Act

I. Introduction

A. Overview

    Air pollution is a serious threat to the health and well-being of 
millions of Americans and imposes a large burden on the U.S. economy. 
Ground-level ozone is linked to potentially serious health problems, 
especially respiratory effects, and environmental degradation. Carbon 
monoxide emissions are also related to health problems. Over the past 
quarter century, state and federal agencies have established emission 
control programs that make significant progress in addressing these 
concerns.
    This proposal includes steps that would reduce the mobile-source 
contribution to air pollution in the United States. In particular, we 
are proposing standards that would require manufacturers to 
substantially reduce emissions from marine spark-ignition engines and 
from nonroad spark-ignition engines below 19 kW that are generally used 
in lawn and garden applications.\1\ We refer to these as Marine SI 
engines and Small SI engines, respectively. The proposed standards are 
a continuation of the process of establishing standards for nonroad 
engines and vehicles as required by Clean Air Act section 213. All the 
nonroad engines subject to this proposal are already regulated under 
existing emission standards, except sterndrive and inboard marine 
engines, which will be subject to EPA emission standards for the first 
time.
---------------------------------------------------------------------------

    \1\ Otto-cycle engines (referred to here as spark-ignition or SI 
engines) typically operate on gasoline, liquefied petroleum gas, or 
natural gas. Diesel-cycle engines, referred to simply as ``diesel 
engines'' in this document, may also be referred to as compression-
ignition or CI engines. These engines typically operate on diesel 
fuel, but other fuels may also be used.
---------------------------------------------------------------------------

    Nationwide, emissions from Marine SI engines and Small SI engines 
contribute significantly to mobile source air pollution. By 2020 
without the proposed requirements these engines will account for about 
27 percent (1,352,000 tons) of mobile source volatile organic 
hydrocarbon compounds (VOC) emissions, 31 percent (16,374,000 tons) of 
mobile source carbon monoxide (CO) emissions, 4 percent (202,000 tons) 
of mobile source oxides of nitrogen (NOX) emissions, and 16 
percent (39,000 tons) of mobile source particulate matter 
(PM2.5) emissions. The proposed standards will reduce 
exposure to these emissions and help avoid a range of adverse health 
effects associated with ambient ozone, CO, and PM levels. In addition, 
the proposed standards will help reduce acute exposure to CO, air 
toxics, and PM for persons who operate or who work with or are 
otherwise active in close proximity to these engines. They will also 
help address other environmental problems associated with Marine SI 
engines and Small SI engines, such as visibility impairment in our 
national parks and other wilderness areas. These effects are described 
in more detail in subsequent sections of this Preamble.

B. Why Is EPA Taking This Action?

    Clean Air Act section 213(a)(1) directs us to study emissions from 
nonroad engines and vehicles to determine, among other things, whether 
these emissions ``cause, or significantly contribute to, air pollution 
which may reasonably be anticipated to endanger public health or 
welfare.'' Section 213(a)(2) further requires us to determine whether 
emissions of CO, VOC, and NOX from all nonroad engines 
significantly contribute to ozone or CO concentrations in more than one 
nonattainment area. If we determine that emissions from all nonroad 
engines do contribute significantly to these nonattainment areas, 
section 213(a)(3) then requires us to establish emission standards for 
classes or categories of new nonroad engines and vehicles that cause or 
contribute to such pollution. We may also set emission standards under 
section 213(a)(4) regulating any other emissions from nonroad engines 
that we find contribute significantly to air pollution which may 
reasonably be anticipated to endanger public health or welfare.
    Specific statutory direction to propose standards for nonroad 
spark-ignition engines comes from section 428(b) of the 2004 
Consolidated Appropriations Act, which requires EPA to propose 
regulations under the Clean Air Act ``that shall contain standards to 
reduce emissions from new nonroad spark-ignition engines smaller than 
50 horsepower.'' \2\ As highlighted above and more fully described in 
Section II, these engines emit pollutants that contribute to ground-
level ozone and ambient CO levels. Human exposure to ozone and CO can 
cause serious respiratory and cardiovascular problems. Additionally, 
these emissions contribute to other serious environmental degradation. 
This proposal implements Congress' mandate by proposing new 
requirements for particular nonroad engines and equipment that are 
regulated as part of EPA's overall nonroad emission control program.
---------------------------------------------------------------------------

    \2\ Pub. L. 108-199, Div G, Title IV, Sec.  428(b), 118 Stat. 
418 (January 23, 2004).
---------------------------------------------------------------------------

    We are proposing this rule under the procedural authority of 
section 307(d) of the Clean Air Act.

C. What Regulations Currently Apply to Nonroad Engines or Vehicles?

    EPA has been setting emission standards for nonroad engines and/or 
vehicles since Congress amended the Clean Air Act in 1990 and included 
section 213. These amendments have led to a series of rulemakings to 
reduce the air pollution from this widely varying set of products. In 
these rulemakings, we divided the broad group of nonroad engines and 
vehicles into several different categories for setting application-
specific requirements. Each category involves many unique 
characteristics related to the participating manufacturers, technology, 
operating characteristics, sales volumes, and market dynamics. 
Requirements for each category therefore take on many unique features 
regarding the stringency of standards, the underlying expectations 
regarding emission control technologies, the nature and extent of 
testing, and the myriad details that comprise the implementation of a 
compliance program.
    At the same time, the requirements and other regulatory provisions 
for each engine category share many characteristics. Each rulemaking 
under section 213 sets technology-based standards consistent with the 
Clean Air Act and requires annual certification based on measured 
emission levels from test engines or vehicles. As a result, the broader 
context of EPA's nonroad emission control programs demonstrates both 
strong similarities between this rulemaking and the requirements 
adopted for other types of engines or vehicles and distinct differences 
as we take into account the unique nature of these engines and the 
companies that produce them.
    We completed the Nonroad Engine and Vehicle Emission Study to 
satisfy Clean Air Act section 213(a)(1) in

[[Page 28101]]

November 1991.\3\ On June 17, 1994, we made an affirmative 
determination under section 213(a)(2) that nonroad emissions are 
significant contributors to ozone or CO in more than one nonattainment 
area (56 FR 31306). Since then we have undertaken several rulemakings 
to set emission standards for the various categories of nonroad 
engines. Table I-1 highlights the different engine or vehicle 
categories we have established and the corresponding cites for emission 
standards and other regulatory requirements. Table I-2 summarizes the 
series of EPA rulemakings that have set new or revised emission 
standards for any of these nonroad engines or vehicles. These actions 
are described in the following sections, with additional discussion to 
explain why we are not proposing more stringent standards for certain 
types of nonroad spark-ignition engines below 50 horsepower.
---------------------------------------------------------------------------

    \3\ This study is available on EPA's web site at http://www.epa.gov/otaq/equip-ld.

                        Table I-1.--Nonroad Engine Categories for EPA Emission Standards
----------------------------------------------------------------------------------------------------------------
                                             CFR cite for regulationse
            Engine categories                  establishing emission          Cross reference to Table I.C-2
                                                     standards
----------------------------------------------------------------------------------------------------------------
1. Locomotives engines...................  40 CFR Part 92..............  d
2. Marine diesel engines.................  40 CFR Part 94..............  g, i, j
3. Other nonroad diesel engines..........  40 CFR Parts 89 and 1039....  a, e, k
4. Marine SI engines \4\.................  40 CFR Part 91..............  c
5. Recreational vehicles.................  40 CFR Part 1051............  i
6. Small SI engines \5\..................  40 CFR Part 90..............  b, f, h
7. Large SI engines \4\..................  40 CFR Part 1048............  i
----------------------------------------------------------------------------------------------------------------


                                Table I-2.--EPA's Rulemakings for Nonroad Engines
----------------------------------------------------------------------------------------------------------------
 Nonroad engines (categories and sub-
              categories)                         Final rulemaking                           Date
----------------------------------------------------------------------------------------------------------------
a. Land-based diesel engines >=37 kW    56 FR 31306........................  June 17, 1994.
 Tier 1.
b. Small SI engines--Phase 1..........  60 FR 34581........................  July 3, 1995.
c. Marine SI engines--outboard and      61 FR 52088........................  October 4, 1996.
 personal watercraft.
d. Locomotives........................  63 FR 18978........................  April 16, 1998.
e. Land-based diesel engines--Tier 1    63 FR 56968........................  October 23, 1998.
 and Tier 2 for engines <37 kW--Tier 2
 and Tier 3 for engines >=37 kW.
f. Small SI engines (Nonhandheld)--     64 FR 15208........................  March 30, 1999.
 Phase 2.
g. Commercial marine diesel <30 liters  64 FR 73300........................  December 29, 1999.
 per cylinder.
h. Small SI engines (Handheld)--Phase   65 FR 24268........................  April 25, 2000.
 2.
i. Recreational vehicles, Industrial    67 FR 68242........................  November 8, 2002.
 spark-ignition engines >19 kW, and
 Recreational marine diesel.
j. Marine diesel engines >=2.5 liters/  68 FR 9746.........................  February 28, 2003.
 cylinder.
k. Land-based diesel engines--Tier 4..  69 FR 38958........................  June 29, 2004.
----------------------------------------------------------------------------------------------------------------

(1) Small SI Engines
    We have previously adopted emission standards for nonroad spark-
ignition engines at or below 19 kW in two phases. The first phase of 
these standards introduced certification and an initial level of 
emission standards for both handheld and nonhandheld engines. On March 
30, 1999 we adopted a second phase of standards for nonhandheld 
engines, including both Class I and Class II engines, which are almost 
fully phased-in today (64 FR 15208).\6\ These standards involved 
emission reductions based on improving engine calibrations to reduce 
exhaust emissions and added a requirement that emission standards must 
be met over the engines' entire useful life as defined in the 
regulations. We believe catalyst technology has now developed to the 
point that it can be applied to all nonhandheld Small SI engines to 
reduce exhaust emissions. Various emission control technologies are 
similarly available to address the different types of fuel evaporative 
emissions we have identified.
---------------------------------------------------------------------------

    \4\ The term ``Marine SI,'' used throughout this document, 
refers to all spark-ignition engines used to propel marine vessels. 
This includes outboard engines, personal watercraft engines, and 
sterndrive/inboard engines. See Section III for additional 
information.
    \5\ The terms ``Small SI'' and ``Large SI'' are used throughout 
this document. All nonroad spark-ignition engines not covered by our 
programs for Marine SI engines or recreational vehicles are either 
Small SI engines or Large SI engines. Small SI engines include those 
engines with maximum power at or below 19 kW, and Large SI engines 
include engines with maximum power above 19 kW.
    \6\ Handheld engines generally include those engines for which 
the operator holds or supports the equipment during operation; 
nonhandheld engines are Small SI engines that are not handled 
engines (see Sec.  1054.801). Class I refers to nonhandheld engines 
with displacement below 225 cc; Class II refers to larger 
nonhandheld engines.
---------------------------------------------------------------------------

    For handheld engines, we adopted Phase 2 exhaust emission standards 
in April 25, 2000 (65 FR 24268). These standards were based on the 
application of catalyst technology, with the expectation that 
manufacturers would have to make considerable investments to modify 
their engine designs and production processes. A technology review we 
completed in 2003 indicated that manufacturers were making progress 
toward compliance, but that additional implementation flexibility was 
needed if manufacturers were to fully comply with the regulations by 
2010. This finding and a change in the rule were published in the 
Federal Register on January 12, 2004 (69FR1824). At this point, we have 
no information to suggest that manufacturers can uniformly apply new 
technology or make design improvements to reduce exhaust emissions 
below the Phase 2 levels. We therefore believe the Phase 2 standards 
continue to represent the greatest degree of emission reduction 
achievable for these engines.\7\ However, we believe it is appropriate 
to apply evaporative emission standards to the handheld engines similar 
to those we are

[[Page 28102]]

proposing for the nonhandheld engines. Manufacturers can control 
evaporative emissions in a way that has little or no impact on exhaust 
emissions.
---------------------------------------------------------------------------

    \7\ Note that we refer to the handheld exhaust emission 
standards in 40 CFR part 1054 as Phase 3 standards. This is intended 
to maintain consistent terminology with the comparable standards in 
California rather than indicating an increase in stringency.
---------------------------------------------------------------------------

(2) Marine SI Engines
    On October 4, 1996 we adopted emission standards for spark-ignition 
outboard and personal watercraft engines that have recently been fully 
phased in (61 FR 52088). We decided not to finalize emission standards 
for sterndrive or inboard marine engines at that time. Uncontrolled 
emission levels from sterndrive and inboard marine engines were already 
significantly lower than the outboard and personal watercraft engines. 
We did, however, leave open the possibility of revisiting the need for 
emission standards for sterndrive and inboard engines in the future. 
See Section III for further discussion of the scope and background of 
past and current rulemakings for these engines.
    We believe existing technology can be applied to all Marine SI 
engines to reduce emissions of harmful pollutants, including both 
exhaust and evaporative emissions. Manufacturers of outboard and 
personal watercraft engines can continue the trend of producing four-
stroke engines and advanced-technology two-stroke engines to further 
reduce emissions. For sterndrive/inboard engines, manufacturers can add 
technologies, such as fuel injection and aftertreatment, that can 
safely and substantially improve the engines' emission control 
capabilities.
    (3) Large SI Engines
    We adopted emission standards for Large SI engines on November 8, 
2002 (67 FR 68242). This includes Tier 1 standards for 2004 through 
2006 model years and Tier 2 standards starting with 2007 model year 
engines. Manufacturers are today facing a considerable challenge to 
comply with the Tier 2 standards, which are already substantially more 
stringent than any of the standards proposed or contemplated for the 
other engine categories in this proposal. The Tier 2 standards also 
include evaporative emission standards, new transient test procedures, 
and additional exhaust emission standards to address off-cycle 
emissions, and diagnostic requirements. Stringent standards for this 
category of engines, and in particular, engines between 25 and 50 
horsepower (19 to 37 kW), have been completed in the recent past, and 
are currently being implemented. Because of that we do not have 
information on the actual Tier 2 technology that manufacturers will use 
and do not have information at this time on possible advances in 
technology beyond Tier 2. We therefore believe the evidence provided in 
the recently promulgated rulemaking continues to represent the best 
available information regarding the appropriate level of standards for 
these engines under section 213 at this time. California Air Resources 
Board (ARB) has adopted an additional level of emission control for 
Large SI engines starting with the 2010 model year. However, as 
described in Section I.D.1, their new standards would not increase 
overall stringency beyond that reflected in the federal standards. As a 
result, we believe it would be inappropriate to pursue more stringent 
emission standards for these engines in this rulemaking.
    Note that the Large SI standards apply to nonroad spark-ignition 
engines above 19 kW. However, we adopted a special provision for engine 
families where production engines have total displacement at or below 
1000 cc and maximum power at or below 30 kW, allowing these engine 
families to instead certify to the applicable standards for Small SI 
engines.
(4) Recreational Vehicles
    We adopted exhaust and evaporative emission standards for 
recreational vehicles in our November 8, 2002 final rule (67FR68242). 
These standards apply to all-terrain vehicles, off-highway motorcycles, 
and snowmobiles.\8\ These exhaust emission standards will be fully 
phased in starting with the 2007 model year. The evaporative emission 
standards apply starting with the 2008 model year.
---------------------------------------------------------------------------

    \8\ Note that we treat certain high-speed off-road utility 
vehicles as all-terrain vehicles (see 40 CFR part 1051).
---------------------------------------------------------------------------

    Recreational vehicles will soon be subject to permeation 
requirements that are very similar to the requirements proposed in this 
rulemaking. We have also learned more about controlling running losses 
and diffusion emissions that may eventually lead us to propose 
comparable standards for recreational vehicles. We expect to revisit 
these questions in the context of a rulemaking to modify the duty cycle 
for all-terrain vehicles, as described below. Considering these new 
requirements for recreational vehicles in this later rulemaking would 
give us additional time to collect information to better understand the 
feasibility, costs, and benefits of applying these requirements to 
recreational vehicles.
    The following sections describe the state of technology and 
regulatory requirements for the different types of recreational 
vehicles.
(a) All-Terrain Vehicles
    The regulations for all-terrain vehicles (ATV) specify testing 
based on a chassis-based transient procedure. However, on an interim 
basis, we are permitting manufacturers the option to use a steady-state 
engine-based procedure to allow manufacturers an opportunity to develop 
the field operating data needed to determine if ATV operation is 
dominantly steady state or transient in nature and to develop an 
appropriate emission test cycle from that information. The emissions 
test procedure and duty cycle are critical to getting the degree of 
emission control expected from these engines. We are continuing to work 
toward a resolution of this test cycle development initiative in a 
separate action. The anticipated changes to the test cycle raise new 
questions we will need to work through before we are prepared to change 
the existing regulation and perhaps pursue new emission control 
requirements. In particular, we will need to further explore the extent 
to which the new duty cycle represents in-use operation and whether 
engine or chassis testing is more appropriate in simulating in-use 
operation for accurate emission characterization and measurements. We 
believe it is appropriate to consider more stringent exhaust emission 
standards for these engines after we have had the opportunity to 
address the emission test cycle issue and to thus establish a long-term 
testing protocols and related requirements.
(b) Off-Highway Motorcycles
    For off-highway motorcycles, manufacturers are in many cases making 
a substantial transition to move away from two-stroke engines in favor 
of four-stroke engines. This transition is now underway. While it may 
eventually be appropriate to apply aftertreatment or other additional 
emission control technologies to off-highway motorcycles, we need more 
time for this transition to be completed and to assess the success of 
aftertreatment technologies such as catalysts on similar applications 
such as highway motorcycles. As EPA and manufacturers learn more in 
implementing emission standards, we would expect to be able to better 
judge the potential for broadly applying new technology to achieve 
further emission reductions from off-highway motorcycles.
(c) Snowmobiles
    In our November 8, 2002 final rule we set three phases of exhaust 
emission standards for snowmobiles (67 FR

[[Page 28103]]

68242). Environmental and industry groups challenged the third phase of 
these standards. The court decision upheld much of EPA's reasoning for 
the standards, but vacated the NOX standard and remanded the 
CO and HC standards to clarify the analysis and evidence upon which the 
standards are based. See Bluewater Network, et al v. EPA, 370 F 3d 1 
(D.C. Cir. 2004). A large majority of snowmobile engines are rated 
below 50 hp and there is still a fundamental need for time to pass to 
allow us to assess the success of 4 stroke engine technology in the 
market place. This is an important of the assessment we need to conduct 
with regard to 2012 and later model year emission standards. Thus we 
believe is appropriate to address this in a separate rulemaking.\9\ We 
expect to complete that work with sufficient lead time for 
manufacturers to meet any revised Phase 3 standards that we might adopt 
for the 2012 model year, consistent with the original rulemaking 
requirements.
---------------------------------------------------------------------------

    \9\ Only about 3 percent of snowmobiles are rated below 50 
horsepower.
---------------------------------------------------------------------------

(5) Nonroad Diesel Engines
    The 2004 Consolidated Appropriations Act providing the specific 
statutory direction for this rulemaking focuses on nonroad spark-
ignition engines. Nonroad diesel engines are therefore not included 
within the scope of that Congressional mandate. However, we have gone 
through several rulemakings to set standards for these engines under 
the broader authority of Clean Air Act section 213. In particular, we 
have divided nonroad diesel engines into three groups for setting 
emission standards. We adopted a series of standards for locomotives on 
April 16, 1998, including requirements to certify engines to emission 
standards when they are rebuilt (63 FR 18978). We also adopted emission 
standards for marine diesel engines over several different rulemakings, 
as described in Table I-2. These included separate actions for engines 
below 37 kW, engines installed in oceangoing vessels, engines installed 
in commercial vessels involved in inland and coastal waterways, and 
engines installed in recreational vessels. We have recently proposed 
new emission standards for both locomotive and marine diesel engines 
(72 FR 15938, April 3, 2007).
    Finally, all other nonroad diesel engines are grouped together for 
EPA's emission standards. We have adopted multiple tiers of 
increasingly stringent standards in three separate rulemakings, as 
described in Table I-2. We most recently adopted Tier 4 standards based 
on the use of ultra-low sulfur diesel fuel and the application of 
exhaust aftertreatment technology (69 FR 38958, June 29, 2004).

D. Putting This Proposal Into Perspective

    Most manufacturers that will be subject to this rulemaking are also 
affected by regulatory developments in California and in other 
countries. Each of these is described in more detail below.
(1) State Initiatives
    Clean Air Act section 209 prohibits California and other states 
from setting emission standards for new motor vehicles and new motor 
vehicle engines, but authorizes EPA to waive this prohibition for 
California, in which case other states may adopt California's 
standards. Similar preemption and waiver provisions apply for emission 
standards for nonroad engines and vehicles, whether new or in-use. 
However for new locomotives, new engines used in locomotives, and new 
engines used in farm or construction equipment with maximum power below 
130 kW, California and other states are preempted and there is no 
provision for a waiver of preemption. In addition, in section 428 of 
the amendment to the 2004 Consolidated Appropriations Act, Congress 
further precluded other states from adopting new California standards 
for nonroad spark-ignition engines below 50 horsepower. In addition, 
the amendment required that we specifically address the safety 
implications of any California standards for these engines before 
approving a waiver of federal preemption. We are proposing to codify 
these changes to preemption in this rule.
    California ARB has adopted requirements for five groups of nonroad 
engines: (1) Diesel- and Otto-cycle small off-road engines rated under 
19 kW; (2) spark-ignition engines used for marine propulsion; (3) land-
based nonroad recreational engines, including those used in all-terrain 
vehicles, off-highway motorcycles, go-carts, and other similar 
vehicles; (4) new nonroad spark-ignition engines rated over 19 kW not 
used in recreational applications; and (5) new land-based nonroad 
diesel engines rated over 130 kW. They have also approved a voluntary 
registration and control program for existing portable equipment.
    In the 1990s California ARB adopted Tier 1 and Tier 2 standards for 
Small SI engines consistent with the federal requirements. In 2003, 
they moved beyond the federal program by adopting exhaust 
HC+NOX emission standards of 10 g/kW-hr for Class I engines 
starting in the 2007 model year and 8 g/kW-hr for Class II engines 
starting in the 2008 model year. In the same rule they adopted 
evaporative emission standards for nonhandheld equipment, requiring 
control of fuel tank permeation, fuel line permeation, diurnal 
emissions, and running losses.
    California ARB has adopted two tiers of exhaust emission standards 
for outboard and personal watercraft engines beyond EPA's original 
standards. The most recent standards, which apply starting in 2008, 
require HC+NOX emission levels as low as 16 g/kW-hr. For 
sterndrive and inboard engines, California has adopted a 5 g/kW-hr 
HC+NOX emission standard for 2008 and later model year 
engines, with testing underway to confirm the feasibility of standards. 
California ARB's marine programs include no standards for exhaust CO 
emissions or evaporative emissions.
    The California emission standards for recreational vehicles have a 
different form than the comparable EPA standards but are roughly 
equivalent in stringency. The California standards include no standards 
for controlling evaporative emissions. Another important difference 
between the two programs is California ARB's reliance on a provision 
allowing noncompliant vehicles to be used in certain areas that are 
less environmentally sensitive as long as they have a specified red 
sticker that would identify their lack of emission controls to prevent 
them from operating in other areas.
    California ARB in 1998 adopted requirements that apply to new 
nonroad engines rated over 25 hp produced for California, with 
standards phasing in from 2001 through 2004. Texas has adopted these 
initial California ARB emission standards statewide starting in 2004. 
More recently, California ARB has proposed exhaust emission standards 
and new evaporative emission standards for these engines, consistent 
with EPA's 2007 model year standards. Their proposal also included an 
additional level of emission control for Large SI engines starting with 
the 2010 model year. However, their proposed standards would not 
increase overall stringency beyond that reflected in the federal 
standards. Rather, they aim to achieve reductions in HC+NOX 
emissions by removing the flexibility incorporated into the federal 
standards allowing manufacturers to have higher HC+NOX 
emissions by certifying to a more stringent CO standard.

[[Page 28104]]

(2) Actions in Other Countries
    While the proposed emission standards will apply only to engines 
sold in the United States, we are aware that manufacturers in many 
cases are selling the same products into other countries. To the extent 
that we have the same emission standards as other countries, 
manufacturers can contribute to reducing air emissions without being 
burdened by the costs associated with meeting differing or inconsistent 
regulatory requirements. The following discussion describes our 
understanding of the status of emission standards in countries outside 
the United States.

    Regulations for spark ignition engines in handheld and 
nonhandheld equipment are included in the ``Directive 97/68/EC of 
the European Parliament and of the Council of 16 December 1997 on 
the approximation of the laws of the Member States relating to 
measures against the emission of gaseous and particulate pollutants 
from internal combustion engines to be installed in non-road mobile 
machinery (OJ L 59, 27.2.1998, p. 1)'', as amended by ``Directive 
2002/88/EC of the European Parliament and of the Council of 9 
December 2002''. The Stage I emission standards are to be met by all 
handheld and nonhandheld engines by 24 months after entry into force 
of the Directive (as noted in a December 9, 2002 amendment to 
Directive 97/68/EC). The Stage I emission standards are similar to 
the U.S. EPA's Phase 1 emission standards for handheld and 
nonhandheld engines. The Stage II emission standards are implemented 
over time for the various handheld and nonhandheld engine classes 
from 2005 to 2009 with handheld engines >= 50cc on August 1, 2008. 
The Stage II emission standards are similar to EPA's Phase 2 
emission standards for handheld and nonhandheld engines. Six months 
after these dates Member States shall permit placing on the market 
of engines, whether or not already installed in machinery, only if 
they meet the requirements of the Directive.

    The European Commission has adopted emission standards for 
recreational marine engines, including both diesel and gasoline 
engines. These requirements apply to all new engines sold in member 
countries and began in 2006 for four-stroke engines and in 2007 for 
two-stroke engines. Table I-3 presents the European standards for 
diesel and gasoline recreational marine engines. The numerical emission 
standards for NOX are based on the applicable standard from 
MARPOL Annex VI for marine diesel engines (See Table I-3). The European 
standards are roughly equivalent to the nonroad diesel Tier 1 emission 
standards for HC and CO. Emission measurements under the European 
standards rely on the ISO D2 duty cycle for constant-speed engines and 
the ISO E5 duty cycle for other engines.

                     Table I-3.--European Emission Standards for Recreational Marine Engines
                                                    [g/kW-hr]
----------------------------------------------------------------------------------------------------------------
                 Engine Type                           HC               NOX             CO              PM
----------------------------------------------------------------------------------------------------------------
Two-Stroke Spark-Ignition....................   30 + 100/P\0.75\            10.0     150 + 600/P  ..............
Four-Stroke Spark-Ignition...................     6 + 50/P\0.75\            15.0     150 + 600/P  ..............
Compression-Ignition.........................     1.5 + 2/P\0.5\             9.8             5.0            1.0
----------------------------------------------------------------------------------------------------------------
\*\ P = rated power in kilowatts (kW)

E. What Requirements Are We Proposing?

    EPA's emission control provisions require engine, vessel and 
equipment manufacturers to design and produce their products to meet 
the emission standards we adopt. To ensure that engines, vessels and 
equipment meet the expected level of emission control, we also require 
compliance with a variety of additional requirements, such as 
certification, labeling engines, and meeting warranty requirements. The 
following sections provide a brief summary of the new requirements we 
are proposing in this rulemaking. See the later sections for a full 
discussion of the proposal.
(1) Marine SI Engines and Vessels
    We are proposing a more stringent level of emission standards for 
outboard and personal watercraft engines starting with the 2009 model 
year. The proposed standards for engines above 40 kW are 16 g/kW-hr for 
HC+NOX and 200 g/kW-hr for CO. For engines below 40 kW, the 
standards increase gradually based on the engine's maximum power. We 
expect manufacturers to meet these standards with improved fueling 
systems and other in-cylinder controls. The levels of the standards are 
consistent with the requirements recently adopted by California ARB 
with the advantage of a simplified form of the standard for different 
power ratings and with a CO emission standard. We are not pursuing 
catalyst-based emission standards for outboard and personal watercraft 
engines. As is discussed later in this preamble, the application of 
catalyst-based standards to the marine environment creates special 
technology challenges that must be addressed. Unlike the sterndrive/
inboard engines discussed in the next paragraph, outboard and personal 
watercraft engines are not built from automotive engine blocks and are 
not as easily amenable to the fundamental engine modifications, fuel 
system upgrades, and other engine control modifications needed to get 
acceptable catalyst performance. This proposal is an appropriate next 
step in the evolution of technology-based standards for outboard and 
personal watercraft engines as they are likely to lead to the 
elimination of carbureted two-stroke engines in favor of direct-
injection two-stroke engines and to encourage the fuel system upgrades 
and related engine modifications needed to achieve the required 
reductions and to potentially set the stage for future considerations.
    We are proposing new exhaust emission standards for sterndrive and 
inboard marine engines. The proposed standards are 5.0 g/kW-hr for 
HC+NOX and 75.0 g/kW-hr for CO starting with the 2009 model 
year. We expect manufacturers to meet these standards with three-way 
catalysts and closed-loop fuel injection. To ensure proper functioning 
of these emission control systems in use, we are proposing a 
requirement that engines have a diagnostic system for detecting a 
failure in the emission control system. For sterndrive and inboard 
marine engines at or above 373 kW with high-performance characteristics 
(generally referred to as ``SD/I high-performance engines''), we are 
proposing an HC+NOX emission standard of 5.0 g/kW-hr and a 
CO standard of 350 g/kW-hr. We are also proposing a variety of other 
special provisions for these engines to reflect unique operating 
characteristics and to make it feasible to meet emission standards 
using emission credits. These standards are consistent with the 
requirements recently adopted by California ARB, with some adjustment 
to the provisions for SD/I high-performance engines and with a CO 
emission standard.
    The emission standards described above relate to engine operation 
over a

[[Page 28105]]

prescribed duty cycle for testing in the laboratory. We are also 
proposing not-to-exceed (NTE) standards that establish emission limits 
when engines operate under normal speed-load combinations that are not 
included in the duty cycles for the other engine standards.
    We are proposing new standards to control evaporative emissions for 
all Marine SI vessels. The new standards include requirements to 
control fuel tank permeation, fuel line permeation, and diurnal 
emissions, including provisions to ensure that refueling emissions do 
not increase.
    We are proposing to place these new regulations for Marine SI 
engines in 40 CFR part 1045 rather than changing the current 
regulations in 40 CFR part 91. This new part will allow us to improve 
the clarity of regulatory requirements and update our regulatory 
compliance program to be consistent with the provisions we have 
recently adopted for other nonroad programs. We are also making a 
variety of changes to 40 CFR part 91 to make minor adjustments to the 
current regulations and to prepare for the transition to 40 CFR part 
1045.
(2) Small SI Engines and Equipment
    We are proposing HC+NOX exhaust emission standards of 
10.0 g/kW-hr for Class I engines starting in the 2012 model year and 
8.0 g/kW-hr for Class II engines starting in the 2011 model year. For 
both classes of nonhandheld engines, we are proposing to maintain the 
existing CO standard of 610 g/kW-hr. We expect manufacturers to meet 
these standards by improving engine combustion and adding catalysts. 
These standards are consistent with the requirements recently adopted 
by California ARB.
    For spark-ignition engines used in marine generators, we are 
proposing a more stringent Phase 3 CO emission standard of 5.0 g/kW-hr. 
This would apply equally to all sizes of engines subject to the Small 
SI standards.
    We are proposing new evaporative emission standards for both 
handheld and nonhandheld engines. The new standards include 
requirements to control permeation from fuel tanks and fuel lines. For 
nonhandheld engines we are also proposing to require control of 
diffusion emissions and running losses.
    We are proposing to place the new regulations for Small SI engines 
from 40 CFR part 90 to 40 CFR part 1054. This new part will allow us to 
improve the clarity of regulatory requirements and update our 
regulatory compliance program to be consistent with the provisions we 
have recently adopted for other nonroad programs.

F. How Is This Document Organized?

    Since this proposal covers a broad range of engines and equipment 
that vary in design and use, many readers may be interested only in 
certain aspects of the proposal. We have therefore attempted to 
organize this preamble in a way that allows each reader to focus on the 
material of particular interest. The Air Quality discussion in Section 
II, however, is general in nature and applies to all the categories 
covered by this proposal.
    The next several sections contain our proposal for Small SI engines 
and equipment and Marine SI engines and vessels. Sections III through V 
describe the proposed requirements related to exhaust emission 
standards for each of the affected engine categories, including 
standards, effective dates, testing information, and other specific 
requirements. Section VI details the proposed requirements related to 
evaporative emission requirements for all categories. Sections VII 
through IX contain some general concepts that are relevant to all of 
the engines, vessels and equipment covered by this proposal, such as 
certification requirements and general testing procedures and 
compliance provisions. Section X discusses how we took energy, noise, 
and safety factors into consideration for the proposed standards.
    Section XI describes a variety of proposed provisions that affect 
other categories of engines besides those that are the primary subject 
of this proposal. This includes the following changes:
     We are proposing to reorganize the regulatory language 
related to preemption of state standards and to clarify certain 
provisions. We are also requesting comment regarding a petition to 
reconsider some of the provisions including the extent to which states 
may regulate the use and operation of nonroad engines and vehicles.
     We are incorporating new provisions related to 
certification fees for newly regulated products covered by this 
proposal. This involves some restructuring of the regulatory language. 
We are also proposing various technical amendments, such as identifying 
an additional payment method, that would apply broadly to our 
certification programs.
     We are proposing changes to 40 CFR part 1068 to clarify 
how the provisions apply with respect to evaporative emission 
standards. We are also proposing various technical amendments. These 
changes would apply to all types of nonroad engines that are subject to 
the provisions of part 1068.
     We are proposing several technical amendments for Large SI 
engines and recreational vehicles, largely to maintain consistency 
across programs for different categories of engines and vehicles.
     We are proposing to amend provisions related to the 
delegated-assembly exemption for heavy-duty highway engines as part of 
the effort to apply these provisions to Small SI engines, as described 
in Section V.E.2.
     We are proposing to apply the new standards for Small SI 
engines to the comparable stationary engines.
    Section XII summarizes the projected impacts and benefits of this 
proposal. Finally, Sections XIII and XIV contain information about 
public participation and how we satisfy our various administrative 
requirements.

II. Public Health and Welfare Effects

    The engines, vessels and equipment that would be subject to the 
proposed standards generate emissions of hydrocarbons (HC), nitrogen 
oxides (NOX), particulate matter (PM) and carbon monoxide 
(CO) that contribute to nonattainment of the National Ambient Air 
Quality Standards (NAAQS) for ozone, PM and CO. These engines, vessels 
and equipment also emit hazardous air pollutants (air toxics) that are 
associated with a host of adverse health effects. Emissions from these 
engines, vessels and equipment also contribute to visibility impairment 
and other welfare and environmental effects.
    The health and environmental effects associated with emissions from 
Small SI engines and equipment and Marine SI engines and vessels are a 
classic example of a negative externality (an activity that imposes 
uncompensated costs on others). With a negative externality, an 
activity's social cost (the cost on society imposed as a result of the 
activity taking place) exceeds its private cost (the cost to those 
directly engaged in the activity). In this case, as described in this 
section, emissions from Small SI engines and equipment and Marine SI 
engines and vessels impose public health and environmental costs on 
society. The market system itself cannot correct this externality. The 
end users of the equipment and vessels are often unaware of the 
environmental impacts of their use for lawn care or recreation. Because 
of this, consumers fail to send the market a signal to provide cleaner 
equipment and vessels. In addition, producers of these engines, 
equipment, and vessels are rewarded for emphasizing other aspects of 
these

[[Page 28106]]

products (e.g., total power). To correct this market failure and reduce 
the negative externality, it is necessary to give producers social cost 
signals. The standards EPA is proposing will accomplish this by 
mandating that Small SI engines and equipment and Marine SI engines and 
vessels reduce their emissions to a technologically feasible limit. In 
other words, with this proposed rule the costs of the services provided 
by these engines and equipment will account for social costs more 
fully.
    This section summarizes the general health and welfare effects of 
these emissions. Interested readers are encouraged to refer to the 
Draft RIA for more in-depth discussions.

A. Ozone

    Ground-level ozone pollution is formed by the reaction of volatile 
organic compounds (VOC), of which HC are the major subset, and 
NOX in the lower atmosphere in the presence of heat and 
sunlight. These pollutants, often referred to as ozone precursors, are 
emitted by many types of pollution sources, such as highway and nonroad 
motor vehicles and engines (including those subject to this proposed 
rule), power plants, chemical plants, refineries, makers of consumer 
and commercial products, industrial facilities, and smaller area 
sources. The engine, vessel and equipment controls being proposed will 
reduce VOCs and NOX.
    The science of ozone formation, transport, and accumulation is 
complex.\10\ Ground-level ozone is produced and destroyed in a cyclical 
set of chemical reactions, many of which 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 and result in more ozone than typically would 
occur on a single high-temperature day. 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 VOC or 
NOX emissions.
---------------------------------------------------------------------------

    \10\ U.S. EPA. Air Quality Criteria for Ozone and Related 
Photochemical Oxidants (Final). U.S. Environmental Protection 
Agency, Washington, DC, EPA/600/R-05/004aF-cF, 2006. This document 
is available in Docket EPA-HQ-OAR-2004-0008.
---------------------------------------------------------------------------

    The current ozone NAAQS, established by EPA in 1997, has an 8-hour 
averaging time.\11\ The 8-hour ozone NAAQS is based on well-documented 
science demonstrating that more people were experiencing adverse health 
effects at lower levels of exertion, over longer periods, and at lower 
ozone concentrations than addressed by the previous one-hour ozone 
NAAQS. The current ozone NAAQS addresses ozone exposures of concern for 
the general population and populations most at risk, including children 
active outdoors, outdoor workers, and individuals with pre-existing 
respiratory disease, such as asthma. The 8-hour ozone NAAQS is met at 
an ambient air quality monitoring site when the average of the annual 
fourth-highest daily maximum 8-hour average ozone concentration over 
three years is less than or equal to 0.084 parts per million (ppm).
---------------------------------------------------------------------------

    \11\ EPA's review of the ozone NAAQS is underway and a proposal 
is scheduled for June 2007 with a final rule scheduled for March 
2008.
---------------------------------------------------------------------------

(1) Health Effects of Ozone
    The health and welfare effects of ozone are well documented and are 
assessed in the EPA's 2006 ozone Air Quality Criteria Document (ozone 
AQCD) and staff paper.12 13 Ozone can irritate the 
respiratory system, causing coughing, throat irritation, and/or 
uncomfortable sensation in the chest. Ozone can reduce lung function 
and make it more difficult to breathe deeply, and breathing may become 
more rapid and shallow than normal, thereby limiting a person's 
activity. Ozone can also aggravate asthma, leading to more asthma 
attacks that require a doctor's attention and/or the use of additional 
medication. Animal toxicologic evidence indicates that with repeated 
exposure, ozone can inflame and damage the lining of the lungs, which 
may lead to permanent changes in lung tissue and irreversible 
reductions in lung function. People who are more susceptible to effects 
associated with exposure to ozone include children, the elderly, and 
individuals with respiratory disease such as asthma. There is also 
suggestive evidence that certain people may have greater genetic 
susceptibility. Those with greater exposures to ozone, for instance due 
to time spent outdoors (e.g., outdoor workers), are also of concern.
---------------------------------------------------------------------------

    \12\ U.S. EPA. Air Quality Criteria for Ozone and Related 
Photochemical Oxidants (Final). U.S. Environmental Protection 
Agency, Washington, DC, EPA/600/R-05/004aF-cF, 2006. This document 
is available in Docket EPA-HQ-OAR-2004-0008.
    \13\ U.S. EPA (2007) Review of National Ambient Air Quality 
Standards for Ozone, Assessment of Scientific and Technical 
Information, OAQPS Staff Paper, EPA-452/R-07-003. This document is 
available in Docket EPA-HQ-OAR-2004-0008.
---------------------------------------------------------------------------

    The recent ozone AQCD also examined relevant new scientific 
information that has emerged in the past decade, including the impact 
of ozone exposure on such health effects as changes in lung structure 
and biochemistry, inflammation of the lungs, exacerbation and causation 
of asthma, respiratory illness-related school absence, hospital 
admissions and premature mortality. Animal toxicologic studies have 
suggested potential interactions between ozone and PM with increased 
responses observed to mixtures of the two pollutants compared to either 
ozone or PM alone. The respiratory morbidity observed in animal studies 
along with the evidence from epidemiologic studies supports a causal 
relationship between acute ambient ozone exposures and increased 
respiratory-related emergency room visits and hospitalizations in the 
warm season. In addition, there is suggestive evidence of a 
contribution of ozone to cardiovascular-related morbidity and non-
accidental and cardiopulmonary mortality.
    EPA typically quantifies ozone-related health impacts in its 
regulatory impact analyses (RIAs) when possible. In the analysis of 
past air quality regulations, ozone-related benefits have included 
morbidity endpoints and welfare effects such as damage to commercial 
crops. EPA has not recently included a separate and additive mortality 
effect for ozone, independent of the effect associated with fine 
particulate matter. For a number of reasons, including (1) Advice from 
the Science Advisory Board (SAB) Health and Ecological Effects 
Subcommittee (HEES) that EPA consider the plausibility and viability of 
including an estimate of premature mortality associated with short-term 
ozone exposure in its benefits analyses and (2) conclusions regarding 
the scientific support for such relationships in EPA's 2006 Air Quality 
Criteria for Ozone and Related Photochemical Oxidants (the CD), EPA is 
in the process of determining how to appropriately characterize ozone-
related mortality benefits within the context of benefits analyses for 
air quality regulations. As part of this process, we are seeking advice 
from the National Academy of Sciences (NAS) regarding how the ozone-
mortality literature should be used to quantify the reduction in 
premature mortality due to diminished exposure to ozone, the amount of 
life expectancy to be added and the monetary value of this increased 
life expectancy in the context of health benefits analyses associated 
with regulatory assessments. In addition, the Agency has sought advice 
on characterizing and communicating the uncertainty associated with 
each of these aspects in health benefit analyses.
    Since the NAS effort is not expected to conclude until 2008, the 
agency is currently deliberating how best to

[[Page 28107]]

characterize ozone-related mortality benefits in its rulemaking 
analyses in the interim. We do not quantify an ozone mortality benefit 
for the analysis of the proposed emission standards. So that we do not 
provide an incomplete picture of all of the benefits associated with 
reductions in emissions of ozone precursors, we have chosen not to 
include an estimate of total ozone benefits in the proposed RIA. By 
omitting ozone benefits in this proposal, we acknowledge that this 
analysis underestimates the benefits associated with the proposed 
standards. For more information regarding the quantified benefits 
included in this analysis, please refer to Chapter 8 of the Draft RIA.
(2) Plant and Ecosystem Effects of Ozone
    Ozone contributes to many environmental effects, with impacts to 
plants and ecosystems being of most concern. Ozone can produce both 
acute and chronic injury in sensitive species depending on the 
concentration level and the duration of the exposure. Ozone effects 
also tend to accumulate over the growing season of the plant, so that 
even lower concentrations experienced for a longer duration have the 
potential to create chronic stress on vegetation. Ozone damage to 
plants includes visible injury to leaves and a reduction in food 
production through impaired photosynthesis, both of which can lead to 
reduced crop yields, forestry production, and use of sensitive 
ornamentals in landscaping. In addition, the reduced food production in 
plants and subsequent reduced root growth and storage below ground, can 
result in other, more subtle plant and ecosystems impacts. These 
include increased susceptibility of plants to insect attack, disease, 
harsh weather, interspecies competition and overall decreased plant 
vigor. The adverse effects of ozone on forest and other natural 
vegetation can potentially lead to species shifts and loss from the 
affected ecosystems, resulting in a loss or reduction in associated 
ecosystem goods and services. Lastly, visible ozone injury to leaves 
can result in a loss of aesthetic value in areas of special scenic 
significance like national parks and wilderness areas. The 2006 ozone 
AQCD presents more detailed information on ozone effects on vegetation 
and ecosystems.
(3) Current and Projected 8-Hour Ozone Levels
    Currently, ozone concentrations exceeding the level of the 8-hour 
ozone NAAQS occur over wide geographic areas, including most of the 
nation's major population centers.\14\ As of October, 2006 there are 
approximately 157 million people living in 116 areas designated as not 
in attainment with the 8-hour ozone NAAQS. There are 461 full or 
partial counties that make up the 116 8-hour ozone nonattainment areas. 
These numbers do not include the people living in areas where there is 
a potential risk of failing to maintain or achieve the 8-hour ozone 
NAAQS in the future.
---------------------------------------------------------------------------

    \14\ A map of the 8-hour ozone nonattainment areas is included 
in the RIA for this proposed rule.
---------------------------------------------------------------------------

    EPA has already adopted many emission control programs that are 
expected to reduce ambient ozone levels. These control programs include 
the Clean Air Interstate Rule (70 FR 25162, May 12, 2005), as well as 
many mobile source rules, some of which are described in Section I of 
this preamble. As a result of these programs, the number of areas that 
fail to meet the 8-hour ozone NAAQS in the future is expected to 
decrease.
    Based on the recent ozone modeling performed for the CAIR analysis, 
barring additional local ozone precursor controls, we estimate 37 
eastern counties (where 24 million people are projected to live) will 
exceed the 8-hour ozone NAAQS in 2010.15 16 An additional 
148 eastern counties (where 61 million people are projected to live) 
are expected to be within 10 percent of the 8-hour ozone NAAQS in 2010.
---------------------------------------------------------------------------

    \15\ Technical Support Document for the Final Clean Air 
Interstate Rule Air Quality Modeling. This document is available in 
Docket EPA-HQ-OAR-2004-0008, Document  EPA-HQ-OAR-2004-
0008-0484.
    \16\ We expect many of the 8-hour ozone nonattainment areas to 
adopt additional emission reduction programs but we are unable to 
quantify or rely upon future reductions from additional state and 
local programs that have not yet been adopted.
---------------------------------------------------------------------------

    States with 8-hour ozone nonattainment areas will be required to 
take action to bring those areas into compliance in the future. Based 
on the final rule designating and classifying 8-hour ozone 
nonattainment areas (69 FR 23951, April 30, 2004), most 8-hour ozone 
nonattainment areas will be required to attain the 8-hour ozone NAAQS 
in the 2007 to 2014 time frame and then be required to maintain the 8-
hour ozone NAAQS thereafter.\17\ Emissions of ozone precursors from the 
engines, vessels and equipment subject to the proposed standards 
contribute to ozone in many, if not all, of these areas. Therefore, the 
expected HC and NOX reductions from the standards proposed 
in this action will be useful to states in attaining or maintaining the 
8-hour ozone NAAQS.
---------------------------------------------------------------------------

    \17\ The Los Angeles South Coast Air Basin 8-hour ozone 
nonattainment area will have until June 15, 2021 to reach 
attainment.
---------------------------------------------------------------------------

    EPA's review of the ozone NAAQS is currently underway and a 
proposed decision in this review is scheduled for June 2007 with a 
final rule scheduled for March 2008. If the ozone NAAQS is revised then 
new nonattainment areas could be designated. While EPA is not relying 
on it for purposes of justifying this rule, the emission reductions 
from this rulemaking would also be helpful to states if there is an 
ozone NAAQS revision.
(4) Air Quality Modeling for Ozone
    To model the ozone air quality benefits of this rule we used the 
Comprehensive Air Quality Model with Extension (CAMx). CAMx simulates 
the numerous physical and chemical processes involved in the formation, 
transport, and destruction of ozone. This model is commonly used in 
developing attainment demonstration State Implementation Plans (SIPs) 
as well as estimating the ozone reductions expected to occur from a 
reduction in emitted pollutants. Meteorological data are developed by a 
separate program, the Regional Atmospheric Modeling System (RAMS), and 
input into CAMx. The simulation periods modeled by CAMx include several 
multi-day periods when ambient measurements were representative of 
ozone episodes over the eastern United States: June 12-24, July 5-15 
and August 7-21, 1995. The modeling domain we used includes the 37 
eastern states modeled in the Clean Air Interstate Rule (CAIR). More 
detailed information is included in the Air Quality Modeling Technical 
Support Document (TSD), which is located in the docket for this rule.
    Note that the emission control scenarios used in the air quality 
and benefits modeling are slightly different than the emission control 
program in this proposal reflecting further refinement of the 
regulatory program since we performed the air quality modeling for this 
proposal. Additional detail on the difference between the modeled and 
proposed inventories is included in Section 3.6 of the Draft RIA.
(5) Results of the Air Quality Modeling for Ozone
    According to air quality modeling performed for this proposal, the 
proposed controls for emissions from the engines, vessels and equipment 
subject to the proposed standards are expected to provide nationwide 
improvements in ozone levels. On a population-weighted basis, the 
average modeled future-year 8-hour ozone design values would decrease 
by 0.7

[[Page 28108]]

ppb in 2020 and 0.8 ppb in 2030.\18\ Within areas predicted to have 
design values greater than 85 ppb the average decrease would be 
somewhat higher: 0.8 ppb in 2020 and 1.0 ppb in 2030.
---------------------------------------------------------------------------

    \18\ A design value is the monitored reading used by EPA to 
determine an area's air quality status; e.g., for ozone, the fourth 
highest reading measured over the most recent three years is the 
design value. (http://www.epa.gov/OCEPAterms/dterms.html).
---------------------------------------------------------------------------

B. Particulate Matter

    Particulate matter (PM) 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. PM is further described 
by breaking it down into size fractions. PM10 refers to 
particles generally less than or equal to 10 micrometers ([mu]m) in 
diameter. PM2.5 refers to fine particles, those particles 
generally less than or equal to 2.5 [mu]m in diameter. Inhalable (or 
``thoracic'' ) coarse particles refer to those particles generally 
greater than 2.5 [mu]m but less than or equal to 10 [mu]m in diameter. 
Ultrafine PM refers to particles with diameters generally less than 100 
nanometers (0.1 [mu]m). Larger particles (>10 [mu]m) tend to be removed 
by the respiratory clearance mechanisms, whereas smaller particles are 
deposited deeper in the lungs.
    Fine particles are produced primarily by combustion processes and 
by transformations of gaseous emissions (e.g., SOx, 
NOX and VOCs) in the atmosphere. The chemical and physical 
properties of PM2.5 may vary greatly with time, region, 
meteorology and source category. Thus, PM2.5, may include a 
complex mixture of different pollutants including sulfates, nitrates, 
organic compounds, elemental carbon and metal compounds. These 
particles can remain in the atmosphere for days to weeks and travel 
through the atmosphere hundreds to thousands of kilometers.
    EPA's final rule to amend the PM NAAQS addressed revisions to the 
primary and secondary NAAQS for PM to provide increased protection of 
public health and welfare, respectively (71 FR 61144, October 17, 
2006). The primary PM2.5 NAAQS include a short-term (24-
hour) and a long-term (annual) standard. The level of the 24-hour 
PM2.5 NAAQS has been revised from 65[mu]g/m 3 to 
35[mu]g/m 3 to provide increased protection against health 
effects associated with short-term exposures to fine particles. The 
current form of the 24-hour PM2.5 standard was retained 
(e.g., based on the 98th percentile concentration averaged over three 
years). The level of the annual PM2.5 NAAQS was retained at 
15[mu]g/m 3, continuing protection against health effects 
associated with long-term exposures. The current form of the annual 
PM2.5 standard was retained as an annual arithmetic mean 
averaged over three years, however, the following two aspects of the 
spatial averaging criteria were narrowed: (1) The annual mean 
concentration at each site shall be within 10 percent of the spatially 
averaged annual mean, and (2) the daily values for each monitoring site 
pair shall yield a correlation coefficient of at least 0.9 for each 
calendar quarter. With regard to the primary PM10 standards, 
the 24-hour PM10 NAAQS was retained at a level of 150[mu]g/m 
3 not to be exceeded more than once per year on average over 
a three-year period. Given that the available evidence does not suggest 
an association between long-term exposure to coarse particles at 
current ambient levels and health effects, EPA has revoked the annual 
PM10 standard.
    With regard to the secondary PM standards, EPA has revised these 
standards to be identical in all respects to the revised primary 
standards. Specifically, EPA has revised the current 24-hour 
PM2.5 secondary standard by making it identical to the 
revised 24-hour PM2.5 primary standard, retained the annual 
PM2.5 and 24-hour PM10 secondary standards, and 
revoked the annual PM10 secondary standards. This suite of 
secondary PM standards is intended to provide protection against PM-
related public welfare effects, including visibility impairment, 
effects on vegetation and ecosystems, and material damage and soiling.
(1) Health Effects of PM
    Scientific studies show ambient PM is associated with a series of 
adverse health effects. These health effects are discussed in detail in 
the 2004 EPA Particulate Matter Air Quality Criteria Document (PM AQCD) 
as well as the 2005 PM Staff Paper.19 20 Further discussion 
of health effects associated with PM can also be found in the Draft 
RIA.
---------------------------------------------------------------------------

    \19\ U.S. EPA (2004) Air Quality Criteria for Particulate Matter 
(Oct 2004), Volume I Document No. EPA600/P-99/002aF and Volume II 
Document No. EPA600/P-99/002bF. This document is available in Docket 
EPA-HQ-OAR-2004-0008. This document is available electronically at: 
http://cfpub2.epa.gov/ncea/cfm/recordisplay.cfm?deid=87903.
    \20\ U.S. EPA (2005) Review of the National Ambient Air Quality 
Standard for Particulate Matter: Policy Assessment of Scientific and 
Technical Information, OAQPS Staff Paper. EPA-452/R-05-005. This 
document is available electronically at http://www.epa.gov/ttn/naaqs/standards/pm/s_pm_cr_sp.html and in Docket EPA-HQ-OAR-2004-
0008.
---------------------------------------------------------------------------

    Health effects associated with short-term exposures (e.g. hours to 
days) in ambient PM2.5 include premature mortality, 
increased hospital admissions, heart and lung diseases, increased 
cough, adverse lower-respiratory symptoms, decrements in lung function 
and changes in heart rate rhythm and other cardiac effects. Studies 
examining populations exposed to different levels of air pollution over 
a number of years, including the Harvard Six Cities Study and the 
American Cancer Society Study, show associations between long-term 
exposure to ambient PM2.5 and both total and 
cardiorespiratory mortality. In addition, the reanalysis of the 
American Cancer Society Study shows an association between fine 
particle and sulfate concentrations and lung cancer mortality. The 
engines, vessels and equipment covered in this proposal contribute to 
both acute and chronic PM2.5 exposures. Additional 
information on acute exposures is available in Section 2.5 of the Draft 
RIA.
    Recently, several studies have highlighted the adverse effects of 
PM specifically from mobile sources.21 22 Studies have also 
focused on health effects due to PM exposures on or near roadways.\23\ 
Although these studies include all air pollution sources, including 
both spark-ignition (gasoline) and diesel powered vehicles, they 
indicate that exposure to PM emissions near roadways, thus dominated by 
mobile sources, are associated with health effects. The proposed 
controls may help to reduce exposures, and specifically exposures near 
the source, to mobile source related PM2.5.
---------------------------------------------------------------------------

    \21\ Laden, F.; Neas, L.M.; Dockery, D.W.; Schwartz, J. (2000) 
Association of Fine Particulate Matter from Different Sources with 
Daily Mortality in Six U.S. Cities. Environmental Health 
Perspectives 108: 941-947.
    \22\ Janssen, N.A.H.; Schwartz, J.; Zanobetti, A.; Suh, H.H. 
(2002) Air Conditioning and Source-Specific Particles as Modifiers 
of the Effect of PM10 on Hospital Admissions for Heart 
and Lung Disease. Environmental Health Perspectives 110: 43-49.
    \23\ Riediker, M.; Cascio, W.E.; Griggs, T.R..; Herbst, M.C.; 
Bromberg, P.A.; Neas, L.; Williams, R.W.; Devlin, R.B. (2003) 
Particulate Matter Exposures in Cars is Associated with 
Cardiovascular Effects in Healthy Young Men. Am. J. Respir. Crit. 
Care Med. 169: 934-940.
---------------------------------------------------------------------------

(2) Visibility
    Visibility can be defined as the degree to which the atmosphere is 
transparent to visible light.\24\ Visibility impairment

[[Page 28109]]

manifests in two principal ways: as local visibility impairment and as 
regional haze.\25\ Local visibility impairment may take the form of a 
localized plume, a band or layer of discoloration appearing well above 
the terrain as a result from complex local meteorological conditions. 
Alternatively, local visibility impairment may manifest as an urban 
haze, sometimes referred to as a ``brown cloud.'' This urban haze is 
largely caused by emissions from multiple sources in the urban areas 
and is not typically attributable to only one nearby source or to long-
range transport. The second type of visibility impairment, regional 
haze, usually results from multiple pollution sources spread over a 
large geographic region. Regional haze can impair visibility over large 
regions and across states.
---------------------------------------------------------------------------

    \24\ National Research Council, 1993. Protecting Visibility in 
National Parks and Wilderness Areas. National Academy of Sciences 
Committee on Haze in National Parks and Wilderness Areas. National 
Academy Press, Washington, DC. This document is available in Docket 
EPA-HQ-OAR-2004-0008. This book can be viewed on the National 
Academy Press Website at http://www.nap.edu/books/0309048443/html/.
    \25\ See discussion in U.S. EPA , National Ambient Air Quality 
Standards for Particulate Matter; Proposed Rule; January 17, 2006, 
Vol71 p 2676. This information is available electronically at http://epa.gov/fedrgstr/EPA-AIR/2006/January/Day-17/a177.pdf.
---------------------------------------------------------------------------

    Visibility is important because it has direct significance to 
people's enjoyment of daily activities in all parts of the country. 
Individuals value good visibility for the well-being it provides them 
directly, where they live and work, and in places where they enjoy 
recreational opportunities. Visibility is also highly valued in 
significant natural areas such as national parks and wilderness areas, 
and special emphasis is given to protecting visibility in these areas. 
For more information on visibility see the 2004 PM AQCD as well as the 
2005 PM Staff Paper.26 27
---------------------------------------------------------------------------

    \26\ U.S. EPA (2004) Air Quality Criteria for Particulate Matter 
(Oct 2004), Volume I Document No. EPA600/P-99/002aF and Volume II 
Document No. EPA600/P-99/002bF. This document is available in Docket 
EPA-HQ-OAR-2004-0008.
    \27\ U.S. EPA (2005) Review of the National Ambient Air Quality 
Standard for Particulate Matter: Policy Assessment of Scientific and 
Technical Information, OAQPS Staff Paper. EPA-452/R-05-005. This 
document is available in Docket EPA-HQ-OAR-2004-0008.
---------------------------------------------------------------------------

    Fine particles are the major cause of reduced visibility in parts 
of the United States. To address the welfare effects of PM on 
visibility, EPA set secondary PM2.5 standards that would act 
in conjunction with the establishment of a regional haze program. In 
setting this secondary standard, EPA concluded that PM2.5 
causes adverse effects on visibility in various locations, depending on 
PM concentrations and factors such as chemical composition and average 
relative humidity. The secondary (welfare-based) PM2.5 NAAQS 
was established as equal to the suite of primary (health-based) NAAQS. 
Furthermore, section 169 of the Act provides additional authorities to 
remedy existing visibility impairment and prevent future visibility 
impairment in the 156 national parks, forests and wilderness areas 
categorized as mandatory class I Federal areas (62 FR 38680-81, July 
18, 1997).\28\ In July 1999 the regional haze rule (64 FR 35714) was 
put in place to protect the visibility in mandatory class I federal 
areas. Visibility can be said to be impaired in both PM2.5 
nonattainment areas and mandatory class I federal areas.
---------------------------------------------------------------------------

    \28\ These areas are defined in section 162 of the Act as those 
national parks exceeding 6,000 acres, wilderness areas and memorial 
parks exceeding 5,000 acres, and all international parks which were 
in existence on August 7, 1977.
---------------------------------------------------------------------------

(a) Current Visibility Impairment
    Recently designated PM2.5 nonattainment areas indicate 
that, as of October 2006, almost 90 million people live in 
nonattainment areas for the 1997 PM2.5 NAAQS. Thus, at least 
these populations would likely be experiencing visibility impairment, 
as well as many thousands of individuals who travel to these areas. In 
addition, while visibility trends have improved in mandatory Class I 
federal areas, the most recent data show that these areas continue to 
suffer from visibility impairment. In summary, visibility impairment is 
experienced throughout the U.S., in multi-state regions, urban areas, 
and remote mandatory class I federal areas.29 30 The 
mandatory class I federal areas are listed in Chapter 2 of the RIA for 
this action. The areas that have design values above the 1997 
PM2.5 NAAQS are also listed in Chapter 2 of the RIA for this 
action.
---------------------------------------------------------------------------

    \29\ US EPA, Air Quality Designations and Classifications for 
the Fine Particles (PM2.5) National Ambient Air Quality 
Standards, December 17, 2004. (70 FR 943, Jan 5. 2005) This document 
is also available on the web at: http://www.epa.gov/pmdesignations/.
    \30\ US EPA. Regional Haze Regulations, July 1, 1999. (64 FR 
35714, July 1, 1999).
---------------------------------------------------------------------------

(b) Future Visibility Impairment
    Recent modeling for the CAIR was used to project visibility 
conditions in mandatory class I federal areas across the country in 
2015. The results for the mandatory class I federal areas suggest that 
these areas are predicted to continue to have annual average deciview 
levels above background in the future.\31\ Modeling done for the PM 
NAAQS projected PM2.5 levels in 2015. These projections 
include all sources of PM2.5, including the engines, vessels 
and equipment covered in this rule, and suggest that PM2.5 
levels above the NAAQS will persist into the future.
---------------------------------------------------------------------------

    \31\ The deciview metric describes perceived visual changes in a 
linear fashion over its entire range, analogous to the decibel scale 
for sound. A deciview of 0 represents pristine conditions. The 
higher the deciview value, the worse the visibility, and an 
improvement in visibility is a decrease in deciview value.
---------------------------------------------------------------------------

    The engines, vessels and equipment that would be subject to these 
proposed standards contribute to visibility concerns in these areas 
through both their primary PM emissions and their VOC and 
NOX emissions, which contribute to the formation of 
secondary PM2.5. Reductions in these direct and secondary PM 
emissions will help to improve visibility across the nation, including 
mandatory class I federal areas.
(3) Atmospheric Deposition
    Wet and dry deposition of ambient particulate matter delivers a 
complex mixture of metals (e.g., mercury, zinc, lead, nickel, aluminum, 
cadmium), organic compounds (e.g., POM, dioxins, furans) and inorganic 
compounds (e.g., nitrate, sulfate) to terrestrial and aquatic 
ecosystems. The chemical form of the compounds deposited is impacted by 
a variety of factors including ambient conditions (e.g., temperature, 
humidity, oxidant levels) and the sources of the material. Chemical and 
physical transformations of the particulate compounds occur in the 
atmosphere as well as the media onto which they deposit. These 
transformations in turn influence the fate, bioavailability and 
potential toxicity of these compounds. Atmospheric deposition has been 
identified as a key component of the environmental and human health 
hazard posed by several pollutants including mercury, dioxin and 
PCBs.\32\
---------------------------------------------------------------------------

    \32\ U.S. EPA (2000) Deposition of Air Pollutants to the Great 
Waters: Third Report to Congress. Office of Air Quality Planning and 
Standards. EPA-453/R-00-0005.
---------------------------------------------------------------------------

    Adverse impacts on water quality can occur when atmospheric 
contaminants deposit to the water surface or when material deposited on 
the land enters a waterbody through runoff. Potential impacts of 
atmospheric deposition to waterbodies include those related to both 
nutrient and toxic inputs. Adverse effects to human health and welfare 
can occur from the addition of excess particulate nitrate nutrient 
enrichment, which contributes to toxic algae blooms and zones of 
depleted oxygen, which can lead to fish kills, frequently in coastal 
waters. Particles contaminated with heavy metals or other toxins may 
lead to the ingestion of contaminated fish, ingestion of contaminated 
water, damage to the marine ecology, and limited recreational uses. 
Several

[[Page 28110]]

studies have been conducted in U.S. coastal waters and in the Great 
Lakes Region in which the role of ambient PM deposition and runoff is 
investigated.33 34 35 36 37
---------------------------------------------------------------------------

    \33\ U.S. EPA (2004) National Coastal Condition Report II. 
Office of Research and Development/ Office of Water. EPA-620/R-03/
002.
    \34\ Gao, Y., E.D. Nelson, M.P. Field, et al. 2002. 
Characterization of atmospheric trace elements on PM2.5 
particulate matter over the New York-New Jersey harbor estuary. 
Atmos. Environ. 36: 1077-1086.
    \35\ Kim, G., N. Hussain, J.R. Scudlark, and T.M. Church. 2000. 
Factors influencing the atmospheric depositional fluxes of stable 
Pb, 210Pb, and 7Be into Chesapeake Bay. J. Atmos. Chem. 36: 65-79.
    \36\ Lu, R., R.P. Turco, K. Stolzenbach, et al. 2003. Dry 
deposition of airborne trace metals on the Los Angeles Basin and 
adjacent coastal waters. J. Geophys. Res. 108(D2, 4074): AAC 11-1 to 
11-24.
    \37\ Marvin, C.H., M.N. Charlton, E.J. Reiner, et al. 2002. 
Surficial sediment contamination in Lakes Erie and Ontario: A 
comparative analysis. J. Great Lakes Res. 28(3): 437-450.
---------------------------------------------------------------------------

    Adverse impacts on soil chemistry and plant life have been observed 
for areas heavily impacted by atmospheric deposition of nutrients, 
metals and acid species, resulting in species shifts, loss of 
biodiversity, forest decline and damage to forest productivity. 
Potential impacts also include adverse effects to human health through 
ingestion of contaminated vegetation or livestock (as in the case for 
dioxin deposition), reduction in crop yield, and limited use of land 
due to contamination.
(4) Current and Projected PM2.5 Levels
    In 2005 EPA designated 39 nonattainment areas for the 1997 
PM2.5 NAAQS based on air quality design values (using 2001-
2003 or 2002-2004 measurements) and a number of other factors (70 FR 
943, January 5, 2005).\38\ These areas are comprised of 208 full or 
partial counties with a total population exceeding 88 million. As 
mentioned in Section II.B.2, the 1997 PM2.5 NAAQS was 
recently revised and the 2006 PM2.5 NAAQS became effective 
on December 18, 2006. Table II-1 presents the number of counties in 
areas currently designated as nonattainment for the 1997 
PM2.5 NAAQS as well as the number of additional counties 
that have monitored data that is violating the 2006 PM2.5 
NAAQS. Nonattainment areas will be designated with respect to the new 
2006 PM2.5 NAAQS in early 2010.
---------------------------------------------------------------------------

    \38\ The full details involved in calculating a PM2.5 
design value are given in Appendix N of 40 CFR part 50.

  Table II-1.--Fine Particle Standards: Current Nonattainment Areas and
                        Other Violating Counties
------------------------------------------------------------------------
 Nonattainment areas/other violating      Number of
              counties                    counties       Population \1\
------------------------------------------------------------------------
1997 PM2.5 Standards: 39 areas                     208        88,394,000
 currently designated...............
2006 PM2.5 Standards: counties with                 49        18,198,676
 violating monitors \2\.............
                                     -----------------------------------
    Total...........................               257      106,592,676
------------------------------------------------------------------------
\1\ Population numbers are from 2000 census data.
\2\ This table provides an estimate of the counties violating the 2006
  PM2.5 NAAQS based on 2003-05 air quality data. The areas designated as
  nonattainment for the 2006 PM2.5 NAAQS will be based on 3 years of air
  quality data from later years. Also, the county numbers in the summary
  table include only the counties with monitors violating the 2006 PM2.5
  NAAQS. The monitored county violations may be an underestimate of the
  number of counties and populations that will eventually be included in
  areas with multiple counties designated nonattainment.

    Based on modeling performed for the PM NAAQS analysis, we estimate 
that 52 counties (where 53 million people are projected to live) will 
exceed the 2006 PM2.5 standard in 2015.\39\ In addition, 54 
counties (where 27 million people are projected to live) are expected 
to be within 10 percent of the 2006 PM2.5 NAAQS in 2015.
---------------------------------------------------------------------------

    \39\ US EPA (2006). Regulatory Impact Analysis for the 2006 
NAAQS for Particle Pollution. This document is available in Docket 
EPA-HQ-OAR-2004-0008.
---------------------------------------------------------------------------

    Areas designated as not attaining the 1997 PM2.5 NAAQS 
will need to attain these standards in the 2010 to 2015 time frame, and 
then be required to maintain the NAAQS thereafter. The attainment dates 
associated with the potential new 2006 PM2.5 nonattainment 
areas would likely be in the 2015 to 2020 timeframe. The emission 
standards being proposed in this action would become effective as early 
as 2009 making the expected HC, NOX and PM inventory 
reductions from this rulemaking useful to states in attaining or 
maintaining the PM2.5 NAAQS.
(5) Current PM10 Levels
    As of October 2006 approximately 28.5 million people live in 46 
designated PM10 nonattainment areas, which include all or 
part of 46 counties. These population numbers do not include the people 
living in areas where there is a potential risk of failing to maintain 
or achieve the PM10 NAAQS in the future. The expected PM, HC 
and NOX inventory reductions from these proposed standards 
would be useful to states in maintaining the PM10 NAAQS.

C. Air Toxics

    Emissions from the engines, vessels and equipment subject to the 
proposed standards contribute to ambient levels of gaseous air toxics 
known or suspected as human or animal carcinogens, or that have non-
cancer health effects. These compounds include benzene, 1,3-butadiene, 
formaldehyde, acetaldehyde, acrolein, polycyclic organic matter (POM), 
and naphthalene. All of these compounds, except acetaldehyde, were 
identified as national or regional risk drivers in the 1999 National-
Scale Air Toxics Assessment (NATA) and have significant inventory 
contributions from mobile sources. That is, for a significant portion 
of the population, these compounds pose a significant portion of the 
total cancer risk from breathing outdoor air toxics. The reductions in 
the emissions from these engines, vessels and equipment would help 
reduce exposure to these harmful substances.
    Air toxics can cause a variety of cancer and noncancer health 
effects. A number of the mobile source air toxic pollutants described 
in this section are known or likely to pose a cancer hazard in humans. 
Many of these compounds also cause adverse noncancer health effects 
resulting from chronic,\40\ subchronic,\41\ or acute \42\ inhalation 
exposures. These include neurological, cardiovascular, liver, kidney, 
and respiratory effects as well as effects on the immune and 
reproductive systems.
---------------------------------------------------------------------------

    \40\ Chronic exposure is defined in the glossary of the 
Integrated Risk Information (IRIS) database (http://www.epa.gov/iris) as repeated exposure by the oral, dermal, or inhalation route 
for more than approximately 10% of the life span in humans (more 
than approximately 90 days to 2 years in typically used laboratory 
animal species).
    \41\ Defined in the IRIS database as exposure to a substance 
spanning approximately 10 of the lifetime of an organism.
    \42\ Defined in the IRIS database as exposure by the oral, 
dermal, or inhalation route for 24 hours or less.

---------------------------------------------------------------------------

[[Page 28111]]

    Benzene. The EPA's Integrated Risk Information (IRIS) database 
lists benzene as a known human carcinogen (causing leukemia) by all 
routes of exposure, and that exposure is associated with additional 
health effects, including genetic changes in both humans and animals 
and increased proliferation of bone marrow cells in 
mice.43 44 45 EPA states in its IRIS database that data 
indicate a causal relationship between benzene exposure and acute 
lymphocytic leukemia and suggests a relationship between benzene 
exposure and chronic non-lymphocytic leukemia and chronic lymphocytic 
leukemia. A number of adverse noncancer health effects including blood 
disorders, such as preleukemia and aplastic anemia, have also been 
associated with long-term exposure to benzene.46 47 The most 
sensitive noncancer effect observed in humans, based on current data, 
is the depression of the absolute lymphocyte count in 
blood.48 49 In addition, recent work, including studies 
sponsored by the Health Effects Institute (HEI), provides evidence that 
biochemical responses are occurring at lower levels of benzene exposure 
than previously known.50 51 52 53 EPA's IRIS program has not 
yet evaluated these new data.
---------------------------------------------------------------------------

    \43\ U.S. EPA (2000). Integrated Risk Information System File 
for Benzene. This material is available electronically at http://www.epa.gov/iris/subst/0276.htm.
    \44\ International Agency for Research on Cancer, 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, 1982.
    \45\ Irons, R.D.; Stillman, W.S.; Colagiovanni, D.B.; Henry, 
V.A. (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.
    \46\ Aksoy, M. (1989). Hematotoxicity and carcinogenicity of 
benzene. Environ. Health Perspect. 82: 193-197.
    \47\ Goldstein, B.D. (1988). Benzene toxicity. Occupational 
medicine. State of the Art Reviews. 3: 541-554.
    \48\ 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.
    \49\ EPA 2005 ``Full IRIS Summary for Benzene (CASRN 71-43-2)'' 
Environmental Protection Agency, Integrated Risk Information System 
(IRIS), Office of Health and Environmental Assessment, Environmental 
Criteria and Assessment Office, Cincinnati, OH http://www.epa.gov/iris/subst/0276.htm.
    \50\ Qu, O.; Shore, R.; Li, G.; Jin, X.; Chen, C.L.; Cohen, B.; 
Melikian, A.; Eastmond, D.; Rappaport, S.; Li, H.; Rupa, D.; 
Suramaya, R.; Songnian, W.; Huifant, Y.; Meng, M.; Winnik, M.; Kwok, 
E.; Li, Y.; Mu, R.; Xu, B.; Zhang, X.; Li, K. (2003). HEI Report 
115, Validation & Evaluation of Biomarkers in Workers Exposed to 
Benzene in China.
    \51\ Qu, Q., R. Shore, G. Li, X. Jin, L.C. Chen, B. Cohen, et 
al. (2002). Hematological changes among Chinese workers with a broad 
range of benzene exposures. Am. J. Industr. Med. 42: 275-285.
    \52\ Lan, Qing, Zhang, L., Li, G., Vermeulen, R., et al. (2004). 
Hematotoxically in Workers Exposed to Low Levels of Benzene. Science 
306: 1774-1776.
    \53\ Turtletaub, K.W. and Mani, C. (2003). Benzene metabolism in 
rodents at doses relevant to human exposure from Urban Air. Research 
Reports Health Effect Inst. Report No.113.
---------------------------------------------------------------------------

    1,3-Butadiene. EPA has characterized 1,3-butadiene as carcinogenic 
to humans by inhalation.54 55 The specific mechanisms of 
1,3-butadiene-induced carcinogenesis are unknown. However, it is 
virtually certain that the carcinogenic effects are mediated by 
genotoxic metabolites of 1,3-butadiene. Animal data suggest that 
females may be more sensitive than males for cancer effects, but there 
are insufficient data in humans from which to draw conclusions about 
sensitive subpopulations. 1,3-Butadiene also causes a variety of 
reproductive and developmental effects in mice; no human data on these 
effects are available. The most sensitive effect was ovarian atrophy 
observed in a lifetime bioassay of female mice.\56\
---------------------------------------------------------------------------

    \54\ U.S. EPA. (2002). Health Assessment of 1,3-Butadiene. 
Office of Research and Development, National Center for 
Environmental Assessment, Washington Office, Washington, DC. Report 
No. EPA600-P-98-001F.
    \55\ U.S. EPA (1998). A Science Advisory Board Report: Review of 
the Health Risk Assessment of 1,3-Butadiene. EPA-SAB-EHC-98.
    \56\ Bevan, C.; Stadler, J.C.; Elliot, G.S.; et al. (1996) 
Subchronic toxicity of 4-vinylcyclohexene in rats and mice by 
inhalation. Fundam. Appl. Toxicol. 32:1-10.
---------------------------------------------------------------------------

    Formaldehyde. Since 1987, EPA has classified formaldehyde as a 
probable human carcinogen based on evidence in humans and in rats, 
mice, hamsters, and monkeys.\57\ EPA is currently reviewing recently 
published epidemiological data. For instance, recently released 
research conducted by the National Cancer Institute (NCI) found an 
increased risk of nasopharyngeal cancer and lymphohematopoietic 
malignancies such as leukemia among workers exposed to 
formaldehyde.58 59 NCI is currently performing an update of 
these studies. A recent National Institute of Occupational Safety and 
Health (NIOSH) study of garment workers also found increased risk of 
death due to leukemia among workers exposed to formaldehyde.\60\ Based 
on the developments of the last decade the working group of the 
International Agency for Research on Cancer (IARC) concluded in 2004 
that formaldehyde is carcinogenic to humans (Group 1), a higher 
classification than previous IARC evaluations, on the basis of 
sufficient evidence in humans and sufficient evidence in experimental 
animals.
---------------------------------------------------------------------------

    \57\ U.S. EPA (1987). Assessment of Health Risks to Garment 
Workers and Certain Home Residents from Exposure to Formaldehyde, 
Office of Pesticides and Toxic Substances, April 1987.
    \58\ Hauptmann, M.; Lubin, J.H.; Stewart, P.A.; Hayes, R.B.; 
Blair, A. 2003. Mortality from lymphohematopoetic malignancies among 
workers in formaldehyde industries. Journal of the National Cancer 
Institute 95: 1615-1623.
    \59\ Hauptmann, M.; Lubin, J.H.; Stewart, P.A.; Hayes, R.B.; 
Blair, A. 2004. Mortality from solid cancers among workers in 
formaldehyde industries. American Journal of Epidemiology 159: 1117-
1130.
    \60\ Pinkerton, L.E. 2004. Mortality among a cohort of garment 
workers exposed to formaldehyde: an update. Occup. Environ. Med. 61: 
193-200.
---------------------------------------------------------------------------

    Formaldehyde exposure also causes a range of noncancer health 
effects, including irritation of the eyes (tearing of the eyes and 
increased blinking) and mucous membranes.
    Acetaldehyde. Acetaldehyde is classified in EPA's IRIS database as 
a probable human carcinogen, based on nasal tumors in rats, and is 
considered toxic by the inhalation, oral, and intravenous routes.\61\ 
The primary acute effect of exposure to acetaldehyde vapors is 
irritation of the eyes, skin, and respiratory tract.\62\ The agency is 
currently conducting a reassessment of the health hazards from 
inhalation exposure to acetaldehyde.
---------------------------------------------------------------------------

    \61\ U.S. EPA (1988). Integrated Risk Information System File of 
Acetaldehyde. This material is available electronically at http://www.epa.gov/iris/subst/0290.htm.
    \62\ U.S. EPA (1988). Integrated Risk Information System File of 
Acetaldehyde. This material is available electronically at http://www.epa.gov/iris/subst/0290.htm.
---------------------------------------------------------------------------

    Acrolein. Acrolein is intensely irritating to humans when inhaled, 
with acute exposure resulting in upper respiratory tract irritation and 
congestion. EPA determined in 2003 using the 1999 draft cancer 
guidelines that the human carcinogenic potential of acrolein could not 
be determined because the available data were inadequate. No 
information was available on the carcinogenic effects of acrolein in 
humans and the animal data provided inadequate evidence of 
carcinogenicity.\63\
---------------------------------------------------------------------------

    \63\ U.S. EPA. 2003. Integrated Risk Information System File of 
Acrolein. Research and Development, National Center for 
Environmental Assessment, Washington, DC. This material is available 
electronically at http://www.epa.gov/iris/subst/0364.htm.
---------------------------------------------------------------------------

    Polycyclic Organic Matter (POM). POM is generally defined as a 
large class of organic compounds with multiple benzene rings and a 
boiling point greater than 100 degrees Celsius. One of these compounds, 
naphthalene, is discussed separately below. Polycyclic aromatic 
hydrocarbons (PAH) are a class of POM that contain only hydrogen and 
carbon atoms. A number of PAHs are known or suspected carcinogens.

[[Page 28112]]

    Recent studies have found that maternal exposures to PAHs in a 
population of pregnant women were associated with several adverse birth 
outcomes, including low birth weight and reduced length at birth, as 
well as impaired cognitive development at age three.64 65 
EPA has not yet evaluated these recent studies.
---------------------------------------------------------------------------

    \64\ Perera, F.P.; Rauh, V.; Tsai, W-Y.; et al. (2002) Effect of 
transplacental exposure to environmental pollutants on birth 
outcomes in a multiethnic population. Environ Health Perspect. 111: 
201-205.
    \65\ Perera, F.P.; Rauh, V.; Whyatt, R.M.; Tsai, W.Y.; Tang, D.; 
Diaz, D.; Hoepner, L.; Barr, D.; Tu, Y.H.; Camann, D.; Kinney, P. 
(2006) Effect of prenatal exposure to airborne polycyclic aromatic 
hydrocarbons on neurodevelopment in the first 3 years of life among 
inner-city children. Environ Health Perspect 114: 1287-1292.
---------------------------------------------------------------------------

    Naphthalene. Naphthalene is found in small quantities in gasoline 
and diesel fuels but is primarily a product of combustion. EPA recently 
released an external review draft of a reassessment of the inhalation 
carcinogenicity of naphthalene.\66\ The draft reassessment recently 
completed external peer review.\67\ Based on external peer review 
comments, additional analyses are being considered. California EPA has 
released a new risk assessment for naphthalene, and the IARC has 
reevaluated naphthalene and re-classified it as Group 2B: possibly 
carcinogenic to humans.\68\ Naphthalene also causes a number of chronic 
non-cancer effects in animals, including abnormal cell changes and 
growth in respiratory and nasal tissues.\69\
---------------------------------------------------------------------------

    \66\ U.S. EPA. 2004. Toxicological Review of Naphthalene 
(Reassessment of the Inhalation Cancer Risk), Environmental 
Protection Agency, Integrated Risk Information System, Research and 
Development, National Center for Environmental Assessment, 
Washington, DC. This material is available electronically at http://www.epa.gov/iris/subst/0436.htm.
    \67\ Oak Ridge Institute for Science and Education. (2004). 
External Peer Review for the IRIS Reassessment of the Inhalation 
Carcinogenicity of Naphthalene. August 2004. http://cfpub2.epa.gov/ncea/cfm/recordisplay.cfm?deid=86019.
    \68\ International Agency for Research on Cancer (IARC). (2002). 
Monographs on the Evaluation of the Carcinogenic Risk of Chemicals 
for Humans. Vol. 82. Lyon, France.
    \69\ U.S. EPA. 1998. Toxicological Review of Naphthalene, 
Environmental Protection Agency, Integrated Risk Information System, 
Research and Development, National Center for Environmental 
Assessment, Washington, DC. This material is available 
electronically at http://www.epa.gov/iris/subst/0436.htm.
---------------------------------------------------------------------------

    In addition to reducing VOC, NOX, CO and 
PM2.5 emissions from these engines, vessels and equipment, 
the standards proposed in this document would also reduce air toxics 
emitted from these engines, vessels and equipment, thereby helping to 
mitigate some of the adverse health effects associated with operation 
of these engines, vessels and equipment.

D. Carbon Monoxide

    Carbon monoxide (CO) is a colorless, odorless gas produced through 
the incomplete combustion of carbon-based fuels. The current primary 
NAAQS for CO are 35 ppm for the 1-hour average and nine ppm for the 8-
hour average. These values are not to be exceeded more than once per 
year.
    We have already found that emissions from nonroad engines 
contribute significantly to CO concentrations in more than one 
nonattainment area (59 FR 31306, June 17, 1994). We have also 
previously found that emissions from Small SI engines contribute to CO 
concentrations in more than one nonattainment area. We propose to find 
here, based on the information in this section of the preamble and 
Chapters 2 and 3 of the Draft RIA, that emissions from Marine SI 
engines and vessels likewise contribute to CO concentrations in more 
than one CO nonattainment area.
    Carbon monoxide enters the bloodstream through the lungs, forming 
carboxyhemoglobin and reducing the delivery of oxygen to the body's 
organs and tissues. The health threat from CO is most serious for those 
who suffer from cardiovascular disease, particularly those with angina 
or peripheral vascular disease. Healthy individuals also are affected, 
but only at higher CO levels. Exposure to elevated CO levels is 
associated with impairment of visual perception, work capacity, manual 
dexterity, learning ability and performance of complex tasks. Carbon 
monoxide also contributes to ozone nonattainment since carbon monoxide 
reacts photochemically in the atmosphere to form ozone.\70\ Additional 
information on CO related health effects can be found in the Carbon 
Monoxide Air Quality Criteria Document (CO AQCD).\71\
---------------------------------------------------------------------------

    \70\ U.S. EPA (2000). Air Quality Criteria for Carbon Monoxide, 
EPA/600/P-99/001F. This document is available in Docket EPA-HQ-OAR-
2004-0008.
    \71\ U.S. EPA (2000). Air Quality Criteria for Carbon Monoxide, 
EPA/600/P-99/001F. This document is available in Docket EPA-HQ-OAR-
2004-0008.
---------------------------------------------------------------------------

    In addition to health effects from chronic exposure to ambient CO 
levels, acute exposures to higher levels are also a problem, see the 
Draft RIA for additional information. In recent years a substantial 
number of CO poisonings and deaths have occurred on and around 
recreational boats across the nation.\72\ The actual number of deaths 
attributable to CO poisoning while boating is difficult to estimate 
because CO-related deaths in the water may be labeled as drowning. An 
interagency team consisting of the National Park Service, the U.S. 
Department of the Interior, and the National Institute for Occupational 
Safety and Health maintains a record of published CO-related fatal and 
nonfatal poisonings.\73\ Between 1984 and 2004, 113 CO-related deaths 
and 458 non-fatal CO poisonings have been identified based on hospital 
records, press accounts and other information. Deaths have been 
attributed to exhaust from both onboard generators and propulsion 
engines. Houseboats, cabin cruisers, and ski boats are the most common 
types of boats associated with CO poisoning cases. These incidents have 
prompted other federal agencies, including the United States Coast 
Guard and National Park Service, to issue advisory statements and other 
interventions to boaters to avoid excessive CO exposure.\74\
---------------------------------------------------------------------------

    \72\ Mott, J.S.; Wolfe, M.I.; Alverson, C.J.; Macdonald, S.C.; 
Bailey, C.R.; Ball, L.B.; Moorman, J.E.; Somers, J.H.; Mannino, 
D.M.; Redd, S.C. (2002) National Vehicle Emissions Policies and 
Practices and Declining US Carbon Monoxide-Related Mortality. JAMA 
288:988-995.
    \73\ National Park Service; Department of the Interior; National 
Institute for Occupational Safety and Health. (2004) Boat-related 
carbon monoxide poisonings. This document is available 
electronically at http://safetynet.smis.doi.gov/thelistbystate10-19-04.pdf and in docket EPA-HQ-OAR-2004-0008.
    \74\ U.S Department of the Interior. (2004) Carbon monoxide 
dangers from generators and propulsion engines. On-board boats--
compilation of materials. This document is available online at 
http://safetynet.smis.doi.gov/COhouseboats.htm and in docket EPA-HQ-
OAR-2004-0008.
---------------------------------------------------------------------------

    As of October 2006, there were approximately 15 million people 
living in 6 areas (which include 10 counties) designated as 
nonattainment for CO. The CO nonattainment areas are presented in the 
Draft RIA.
    EPA previously determined that emissions from nonroad engines and 
equipment contribute significantly to ozone and CO concentrations in 
more than one nonattainment area (59 FR 31306, June 17, 1994). EPA also 
determined that the categories of small land-based SI engines cause or 
contribute to ambient ozone and CO in more than one nonattainment area 
(65 FR 76790, Dec. 7, 2000). With regard to Marine SI engines and 
vessels, our NONROAD model indicates that these engines are present in 
each of the CO nonattainment areas and thus contribute to CO 
concentrations in those nonattainment areas. The CO contribution from 
Marine SI engines in classified CO nonattainment areas is presented in 
Table II-2.

[[Page 28113]]



        Table II-2.--CO Emissions from Marine SI Engines and Vessels in Classified CO Nonattainment Areas
----------------------------------------------------------------------------------------------------------------
                                                                                                 CO (short tons
                  Area                             County                     Category              in 2005)
----------------------------------------------------------------------------------------------------------------
Missoula, MT...........................  Missoula..................  Marine SI................                94
Las Vegas, NV..........................  Clark.....................  Marine SI................             3,016
Reno, NV...............................  Washoe....................  Marine SI................             3,494
El Paso, TX............................  El Paso...................  Marine SI................                37
South Coast Air Basin..................  Los Angeles...............  Marine SI................             4,615
                                         Riverside.................  Marine SI................             1,852
                                         Orange....................  Marine SI................             5,360
                                         San Bernardino............  Marine SI................            2,507
----------------------------------------------------------------------------------------------------------------
Source: U.S. EPA, NONROAD 2005 model.

    Based on the national inventory numbers in Chapter 3 of the Draft 
RIA and the local inventory numbers described in this section of the 
preamble, we propose to find that emissions of CO from Marine SI 
engines and vessels contribute to CO concentrations in more than one CO 
nonattainment area.

III. Sterndrive and Inboard Marine Engines

A. Overview

    This section applies to sterndrive and inboard marine (SD/I) 
engines. Sterndrive and inboard engines are spark-ignition engines 
typically derived from automotive engine blocks for which a 
manufacturer will take steps to ``marinize'' the engine for use in 
marine applications. This marinization process includes choosing and 
optimizing the fuel management system, configuring a marine cooling 
system, adding intake and exhaust manifolds, and adding accessory 
drives and units. These engines typically have water-jacketed exhaust 
systems to keep surface temperatures low. Ambient surface water 
(seawater or freshwater) is generally added to the exhaust gases before 
the mixture is expelled under water.
    As described in Section I, the initial rulemaking to set standards 
for Marine SI engines did not include final emission standards for SD/I 
engines. In that rulemaking, we finalized the finding under Clean Air 
Act section 213(a)(3) that all Marine SI engines cause or contribute to 
ozone concentrations in two or more ozone nonattainment areas in the 
United States. However, because uncontrolled SD/I engines appeared to 
be a low-emission alternative to outboard and personal watercraft 
engines in the marketplace, even after the emission standards for these 
engines were fully phased in, we decided to set emission standards only 
for outboard and personal watercraft engines. At that time, outboard 
and personal watercraft engines were almost all two-stroke engines with 
much higher emission rates compared to the SD/I engines, which were all 
four-stroke engines. We pointed out in that initial rulemaking that we 
wanted to avoid imposing costs on SD/I engines that could cause a 
market shift to increased use of the higher-emitting outboard engines, 
which would undermine the broader goal of achieving the greatest degree 
of emission control from the full set of Marine SI engines.
    We believe now is an appropriate time to set standards for SD/I 
engines, for several reasons. First, the available technology for SD/I 
engines has developed significantly, so we are now able to anticipate 
substantial emission reductions. With the simultaneous developments in 
technology for outboard and personal watercraft engines, we can set 
standards that achieve substantial emission reductions from all Marine 
SI engines. Second, now that California has adopted standards for SD/I 
engines, the cost impact of setting new standards for manufacturers 
serving the California market is generally limited to the hardware 
costs of adding emission control technology; these manufacturers will 
be undergoing a complete redesign effort for these engines to meet the 
California standards. Third, we believe SD/I engines meeting the 
proposed standards will in many cases have performance advantages over 
pre-control engines, which will allow manufacturers of SD/I engines to 
promote their engines as having a greater value to justify any price 
increases. As a result, we believe we can achieve the maximum emission 
reductions from Marine SI engines by setting standards for SD/I engines 
based on the use of catalyst technology at the same time that we adopt 
more stringent standards for outboard and personal watercraft engines.
    As described in Section II, we are proposing to make the finding 
under Clean Air Act section 213(a)(3) that Marine SI engines cause or 
contribute to CO concentrations in two or more nonattainment areas of 
the United States. We believe the proposed CO standards will also 
reduce the exposure of individual boaters and bystanders to potentially 
dangerous CO levels.
    We believe catalyst technology is available for achieving these 
proposed standards. Catalysts have been used for decades in automotive 
applications to reduce emissions, and catalyst manufacturers have 
continued to develop and improve this technology. Design issues for 
using catalysts in marine applications are primarily centered on 
packaging catalysts in the water-jacketed, wet exhaust systems seen on 
most SD/I engines. Section III.G discusses recent development work that 
has shown success in packaging catalysts in SD/I applications. In 
addition, there are ongoing efforts in evaluating catalyst technology 
in SD/I engines being sponsored by the marine industry, U.S. Coast 
Guard, and California ARB.

B. Engines Covered by This Rule

(1) Definition of Sterndrive and Inboard Engines
    For the purpose of this regulation, SD/I engines encompass all 
spark-ignition marine propulsion engines that are not outboard or 
personal watercraft engines. A discussion of the proposed new 
definitions for outboard and personal watercraft engines is in Section 
IV.B. We consider all the following to be SD/I engines: inboard, 
sterndrive (also known as inboard/outboard), airboat engines, and jet 
boat engines.
    The existing definitions for sterndrive and inboard engines from 40 
CFR part 91 are presented below:
     Sterndrive engine means a four stroke Marine SI engine 
that is designed such that the drive unit is external to the hull of 
the marine vessel, while the engine is internal to the hull of the 
marine vessel.
     Inboard engine means a four stroke Marine SI engine that 
is designed such that the propeller shaft penetrates the

[[Page 28114]]

hull of the marine vessel while the engine and the remainder of the 
drive unit is internal to the hull of the marine vessel.
    We are proposing to amend the above definitions for determining 
which exhaust emission standards apply to spark-ignition marine engines 
in 2009. The new proposed definition would be a single term to include 
sterndrive and inboard engines together as a single engine category. 
The proposed definition for sterndrive/inboard also is drafted to 
include all engines not otherwise classified as outboard or personal 
watercraft engines. Note that we are proposing to revise the 
definitions of outboard and personal watercraft engines as described in 
Section IV.B.
    The proposed definition has several noteworthy impacts. First, it 
removes a requirement that only four-stroke engines can qualify as 
sterndrive/inboard engines. We believe limiting the definition to 
include only four-stroke engines is unnecessarily restrictive and could 
create an incentive to use two-stroke (or rotary) engines to avoid the 
proposed catalyst-based standards. Second, it removes limitations 
caused by reference to propellers. The definition should not refer 
specifically to propellers, because there are other propulsion drives 
on marine vessels, such as jet drives, that could be used with SD/I 
engines. Third, as explained in the section on the OB/PWC definitions, 
the proposed definitions treat engines installed in open-bay vessels 
(e.g. jet boats) and in vessels over 4 meters long as SD/I engines. 
Finally, the existing definition does not clearly specify how to treat 
specialty vessels such as airboats or hovercraft that use engines that 
are similar to those in conventional SD/I applications. Under the 
discretion in the regulation allowing EPA to make judgments about the 
scope of the SD/I engine definition, we have classified airboats as SD/
I engines. See 40 CFR 91.3 for the existing definitions of the marine 
engine classes. We continue to believe these engines share fundamental 
characteristics with traditional SD/I engines and should therefore be 
treated the same way. However, we believe the definitions should 
address these applications expressly to make clear which standards 
apply.
    We request comment on the following proposed definition:
     Sterndrive/inboard engine means a spark-ignition engine 
that is used to propel a marine vessel, but is not an outboard engine 
or a personal watercraft engine. This includes engines on propeller-
driven vessels, jet boats, airboats, and hovercraft.
    High-performance SD/I engines are generally characterized by high-
speed operation, supercharged air intake, customized parts, very high 
power densities, and a short time until rebuild (50 to 200 hours). 
Based on current SD/I product offerings, we are proposing to define a 
high-performance engine as an SD/I engine with maximum power at or 
above 373 kW (500 hp) that has design features to enhance power output 
such that the expected operating time until rebuild is substantially 
shorter than 480 hours.
(2) Exclusions and Exemptions
    We are proposing to extend our basic nonroad exemptions to the SD/I 
engines and vessels covered by this proposal. These include the testing 
exemption, the manufacturer-owned exemption, the display exemption, and 
the national-security exemption. If the conditions for an exemption are 
met, then the engine is not subject to the exhaust emission standards. 
These exemptions are described in more detail under Section VIII.
    In the rulemaking for recreational vehicles, we chose not to apply 
standards to hobby products by exempting all reduced-scale models of 
vehicles that are not capable of transporting a person (67 FR 68242, 
November 8, 2002). We are proposing to extend that same provision to 
SD/I marine engines (see Sec.  1045.5).
    The Clean Air Act provides for different treatment of engines used 
solely for competition. Rather than relying on engine design features 
that serve as inherent indicators of dedicated competitive use, as 
specified in the current regulations, we have taken the approach in 
more recent programs of more carefully differentiating competition and 
noncompetition models in ways that reflect the nature of the particular 
products. In the case of Marine SI engines, we do not believe there are 
engine design features that allow us to differentiate between engines 
that are used in high-performance recreational applications and those 
that are used solely for competition. We are therefore proposing that, 
starting January 1, 2009, Marine SI engines meeting all the following 
criteria would be considered to be used solely for competition, except 
in other cases where information is available indicating that engines 
are not used solely for competition (see Sec.  1045.620):
     The engine (or a vessel in which the engine is installed) 
may not be displayed for sale in any public dealership or otherwise 
offered for sale to the general public.
     Sale of the vessel in which the engine is installed must 
be limited to professional racers or other qualified racers.
     The engine must have performance characteristics that are 
substantially superior to noncompetitive models (e.g. higher power-to-
weight ratio).
     The engines must be intended for use only in racing events 
sanctioned (with applicable permits) by the Coast Guard or other public 
organization, with operation limited to racing events, speed record 
attempts, and official time trials.
    Engine manufacturers would make their request for each new model 
year, and we would deny a request for future production if there are 
indications that some engines covered by previous requests are not 
being used solely for competition. Competition engines are produced and 
sold in very small quantities, so manufacturers should be able to 
identify which engines qualify for this exemption. We are also 
proposing to apply the same criteria to outboard and personal 
watercraft engines and vessels. We request comment on this approach to 
qualifying for a competition exemption.
    We are proposing a new exemption to address individuals who 
manufacture recreational marine vessels for personal use (see Sec.  
1045.630). Under the proposed exemption, these vessels and their 
engines could be exempt from standards, subject to certain limitations. 
For example, an individual may produce one such vessel over a ten-year 
period, the vessel may not be used for commercial purposes, and any 
exempt engines may not be sold for at least five years. The vessel must 
generally be built from unassembled components, rather than simply 
completing assembly of a vessel that is otherwise similar to one that 
will be certified to meet emission standards. This proposal addresses 
the concern that hobbyists who make their own vessels could otherwise 
be a manufacturer subject to the full set of emission standards by 
introducing these vessels into commerce. We expect this exemption to 
involve a very small number of vessels.

C. Proposed Exhaust Emission Standards

    We are proposing technology-based exhaust emission standards for 
new SD/I engines. These standards are similar to the exhaust emission 
standards that California ARB recently adopted (see Section I). This 
section describes the proposed requirements for SD/I engines for 
controlling exhaust emissions. See

[[Page 28115]]

Section V for a description of the proposed requirements related to 
evaporative emissions.
(1) Standards and Dates
    We are proposing exhaust emission standards of 5 g/kW-hr 
HC+NOX and 75 g/kW-hr CO for SD/I engines, starting with the 
2009 model year (see Sec.  1045.105). On average, this represents about 
a 70 percent reduction in HC+NOX and a 50 percent reduction 
in CO from baseline engine configurations. Due to the challenges of 
controlling CO emissions at high load, the expected reduction in CO 
emissions from low to mid-power operation is expected to be more than 
80 percent. We are proposing additional lead time for small businesses 
as discussed in Section III.F.2. The proposed standards would be based 
on the same duty cycle that currently is in place for outboard and 
personal watercraft engines, as described in Section III.D. Section 
III.F discusses the technological feasibility of these standards in 
more detail. We request comment on the feasibility and appropriateness 
of the proposed standards.
    The proposed standards are largely based on the use of small 
catalytic converters that can be packaged in the water-cooled exhaust 
systems typical for these applications. California ARB also adopted an 
HC+NOX standard of 5 g/kW-hr, but they did not adopt a 
standard for CO emissions. We believe the type of catalyst used to 
achieve the HC+NOX standard will also be effective in 
reducing CO emissions enough to meet the proposed standard, so no 
additional technology will be needed to control CO emissions.
    Manufacturers have expressed concern that the proposed 
implementation dates may be difficult to meet, for certain engines, due 
to anticipated changes in engine block designs produced by General 
Motors. As described in the Draft RIA and in the docket, the vast 
majority of SD/I engines are based on automotive engine blocks sold by 
General Motors.\75\ There are five basic engine blocks used, and 
recently GM has announced that it will discontinue production of the 
4.3L and 8.1L engine blocks in 2009. GM anticipates that it will offer 
a 4.1L engine block and a 6.0L supercharged engine block to the marine 
industry as replacements. Full run production of these new blocks is 
anticipated in mid to late 2009. SD/I engine manufacturers have 
expressed concern that they will not be able to begin the engineering 
processes related to marinizing these engines, including the 
development of catalyst-equipped exhaust manifolds, until mid-2007, 
when they are expecting to see the first prototypes of the two 
replacement engine models. In addition, they are concerned that they do 
not have enough remaining years of sales of the 4.3L and 8.1L engines 
to justify the cost of developing catalyst-equipped exhaust manifolds 
for these engines and amortizing the costs of the required tooling 
while also developing the two new engine models.
---------------------------------------------------------------------------

    \75\ ``GM Product Changes Affecting SD/I Engine Marinizers,'' 
memo from Mike Samulski, EPA, to Docket EPA-HQ-QAR-2004-0008-0528.
---------------------------------------------------------------------------

    The SD/I requirements begin in earnest in California in the 2008 
model year. Manufacturers have indicated that they plan to use 
catalysts to meet the California standards in 2008 for three or four of 
the five engine models used in SD/I applications but to potentially 
have limited availability of the 4.3L or 8.1L engines until the 
catalyst-equipped versions of the two new engine models (4.1L and 6.0L) 
have been marinized and meet the new California emission standards. At 
this point, the manufacturers project that the two new engine models 
would be available for sale in California in 2010. Some 4.3L and 8.1L 
engines may be available in California during the phase-out based on 
the possibility of some use of catalyst for one or both of these 
displacements and the use of transitional flexibilities.
    These are unique circumstances because the SD/I engine 
manufacturers' plans and products depend on the manufacture of the base 
engine by a company not directly involved in marine engine 
manufacturing. The SD/I sales represent only a small fraction of total 
engine sales and thus did not weigh heavily in GM's decision to replace 
the existing engine blocks with two comparable versions during the 
timeframe when the SD/I manufacturers are facing new emission 
standards. SD/I manufacturers have stated that alternative engine 
blocks that meet their are not available in the interim, and that it 
would be cost-prohibitive for them to produce their own engine blocks.
    EPA is proposing that the Federal SD/I standards take effect for 
the 2009 model year, one year after the same standards apply in 
California. We believe a requirement to extend the California standards 
nationwide after a one-year delay allows manufacturers adequate time to 
incorporate catalysts across their product lines as they are doing in 
California. Once the technology is developed for use in California, it 
would be available for use nationwide soon thereafter. In fact, one 
company currently certified to the California standards is already 
offering catalyst-equipped SD/I engines nationwide. However, we request 
comment on whether an additional year of lead time would be appropriate 
for engines not using catalysts in California in 2008. This is 
potentially the 4.3L or 8.1L SD/I engines. Under this alternative, 
engines based on the three engine blocks not being changed would be 
required to meet the standards in 2009. Also, engines built from the 
4.3L and/or the 8.1L GM blocks would be required to meet the EPA 
standards if sold in California in 2008 or 2009. Otherwise the new 
standards for these engines could be delayed for an additional model 
year (until 2010). Assuming product plans follow through as projected, 
the two new engine blocks would be required to meet the standards in 
the 2010 model year.
    Another possibility would be to address this issue through the 
combination of the flexibilities provided through an ABT program and a 
phase-in of the standards over two model years (2009/2010) instead of 
implementation in one model year (2009). Under this approach, 
manufacturers could certify and sell the 4.3L and 8.1L engines in the 
2009 model year without catalysts or with limited use of catalysts 
through emissions averaging. This approach would have the advantage of 
giving manufacturers flexibility in how they choose to phase in their 
catalyst-equipped engines. However, engine manufacturers have expressed 
concern that, even though they will be offering limited configurations 
of catalyzed engines in California in 2008, that the lead time is short 
and they will not have the ability to fully catalyze their entire line 
of engines for 2009. Thus, if the rule is structured in a manner to 
permit it, marine engine manufacturers would sell a mix of catalyzed 
and non-catalyzed engines in 2009. Since boat builders can determine 
which engines are purchased and can choose either catalyzed or non-
catalyzed versions of the engines if available, manufacturers are 
concerned that it would be difficult for SD/I engine manufacturers to 
ensure compliance with standards based on sales and horsepower 
weightings. Engine manufacturers, not boat builders, are subject to 
exhaust emission standards. Thus, a phase-in approach, which would be 
based on a projection that a certain number of catalyzed engines would 
be sold, may not be a feasible approach for this industry. The industry 
would thus prefer a mandatory implementation date as discussed below 
without a phase-in that uses averaging. The industry's concerns 
notwithstanding, there are benefits to

[[Page 28116]]

this approach. Therefore, we are requesting comment on phasing in the 
proposed standards over the 2009-2010 timeframe. Under this approach, 
the standards would be 10 g/kW-hr HC+NOX and 100 g/kW-hr CO 
in 2009. The proposed standards would then go into effect in 2010. 
During the phase-in period, the proposed family emission limit (FEL) 
caps (see Section III.C.3) would still apply.
    A third alternative, preferred by the two large SD/I manufacturers, 
would be full compliance with the 5 g/kW-hr standard in 2010 except for 
the 4.1L engine and the 6.0L supercharged engine and requiring those 
engines to comply with the standards in 2011. Manufacturers have 
expressed the view that there is value in limiting production volumes 
of catalyst-equipped engines only to California for two years to gain 
in-use experience before selling these engines nationwide. Under this 
approach, any technical issues that may arise with catalyst designs or 
in-use performance would affect only a small portion of the fleet, 
which would help minimize in-use concerns and costs associated with 
warranty claims. This approach would also provide additional lead time 
for those configurations not modified for California and the two new 
engine displacements. In addition, as discussed above, manufacturers 
stated that an averaging-based phase-in program that required the 
introduction of catalyst-equipped engines outside of California before 
2010 is problematic because of marketplace and competitive issues as 
discussed above. For these reasons, we request comment on whether the 
proposed standards for SD/I engines should be delayed to 2010 for the 
three engine models that are not being modified and with an additional 
model year (2011) for the 4.1L and 6.0L supercharged engines.
    Under stoichiometric or lean conditions, catalysts are effective at 
oxidizing CO in the exhaust. However, under very rich conditions, 
catalysts are not effective for reducing CO emissions. In contrast, 
NOX emissions are effectively reduced under rich conditions. 
SD/I engines often run at high power modes for extended periods of 
time. Under high-power operation, engine marinizers must calibrate the 
engine to run rich as an engine-protection strategy. If the engine were 
calibrated for a stoichometric air-fuel ratio at high power, high 
temperatures could lead to failures in exhaust valves and engine heads. 
In developing the proposed CO standard for SD/I engines, we considered 
an approach where test Mode 1 (full power) would be excluded from the 
weighted CO test level and the other four test modes would be re-
weighted accordingly. Under this approach, the measured CO emissions 
from catalyst-equipped engines were observed to be 65-85 percent lower 
without Mode 1, even though the weighting factor for Mode 1 is only 6 
percent of the total cycle weighting. These test results are presented 
in Chapter 4 of the Draft RIA. We request comment on finalizing a CO 
standard of 25 g/kW-hr based on a four-mode duty cycle that excludes 
Mode 1 instead of the proposed CO standard. Under this approach, we 
also request comment on CO cap, such as 350 g/kW-hr, specific to Mode 
1. Manufacturers would still measure CO emissions at Mode 1 to 
demonstrate compliance with this cap.
    Controlling CO emissions at high power may be a more significant 
issue with supercharged 6.0L engines due to uncertainty with regard to 
the air fuel ratio of the engine at high power. Engine manufacturers 
have not yet received prototype engines; however, they have expressed 
concern that these engines may need to be operated with a rich air-fuel 
ratio even at Mode 2 as an engine-protection strategy.\76\ This concern 
is based on previous experience with other supercharged engines. If 
this is the case, it may affect the potential CO emission reductions 
from these engines. To address the uncertainties related to the two new 
SD/I engines (4.1L and 6.0L supercharged) we are asking for comment on 
a CO averaging standard with a maximum family emission limit to cap 
high CO emissions. Specifically, we request comment on averaging 
standard of 25 g/kW-hr CO based on a four-mode test, as discussed 
above, with a maximum family emission limit for the four-mode test of 
75 g/kW-hr.
---------------------------------------------------------------------------

    \76\ 80 percent of maximum engine test speed and 71.6 percent of 
maximum torque at maximum test speed.
---------------------------------------------------------------------------

    Engines used on jet boats may have been classified under the 
existing definitions as personal watercraft engines. As described 
above, engines used in jet boats or personal watercraft-like vessels 4 
meters or longer would be classified as SD/I engines under the proposed 
definitions. Such engines subject to part 91 today would therefore need 
to continue meeting EPA emission standards as personal watercraft 
engines through the 2008 model year under part 91, after which they 
would need to meet the new SD/I standards under the proposed part 1045. 
This is another situation where the transition period discussed above 
may be helpful. In contrast, as discussed above, air boats have been 
classified as SD/I engines under EPA's discretionary authority and are 
not required to comply with part 91.
    As described above, engines used solely for competition would not 
be subject to the proposed regulations, but many SD/I high-performance 
engines are sold for recreational use. High-performance SD/I engines 
have very high power outputs, large exhaust gas flow rates, and 
relatively high concentrations of hydrocarbons and carbon monoxide in 
the exhaust gases. From a conceptual perspective, the application of 
catalytic converter technology to these engines is feasible. As is the 
case in similar heavy-duty highway gasoline engines, these catalytic 
converters would have to be quite large in volume, perhaps on the order 
of the same volume as the engine displacement, and would involve 
significant heat rejection issues. Highway heavy-duty gasoline engine 
certification information from the late 1970s and early 1980s suggests 
that it is possible to achieve HC and CO emission reductions around 20 
to 40 percent by adding an air pump to increase the level of oxygen in 
the exhaust stream. This would be a relatively low-cost and durable 
method of oxidizing HC and CO when the exhaust gases are hot enough to 
support further oxidation reactions. California ARB has implemented the 
same HC+NOX standards we are proposing but is expecting 
manufacturers to rely on emissions averaging within the SD/I class. 
This is not viable for small business manufacturers who do not have 
other products with which to average.
    Even if manufacturers use catalysts to control HC+NOX 
emissions from high-performance engines, controlling CO emissions 
continues to present a technological challenge. Since these engines 
generally operate with fuel-rich combustion, there is little or no 
oxygen in the exhaust stream. As a result, any oxidation of hydrocarbon 
compounds in the catalyst would likely increase CO levels, rather than 
oxidizing all the way to CO2. We are therefore proposing a 
CO standard for high-performance engines of 350 g/kW-hr. We believe 
this is achievable with more careful control of fueling under idle 
conditions. Control of air-fuel ratios at idle should result in 
improved emission control even after multiple rebuilds. Basing 
standards on non-catalyst hardware such as an air pump could enable 
lower CO levels.
    We are proposing a variety of provisions to simplify the 
requirements for exhaust emission certification and compliance for 
these engines, as described in Section IV.F. We are also proposing not 
to apply the not-to-exceed

[[Page 28117]]

emission standards to high-performance SD/I marine engines.
    We also request comment on two alternative approaches to define 
emission standards for high-performance engines. First, we could set 
the HC+NOX standard at 5 g/kW-hr and allow for emission 
credits as described above, but allow small-volume manufacturers of 
high-performance engines to meet a HC+NOX emission standard 
in the range of 15 to 22 g/kW-hr. See Section III.F.2 for our proposed 
definition of small-volume SD/I engine manufacturers. We would also 
need to adopt an FEL cap of 22 g/kW-hr for HC+NOX for all 
manufacturers under this approach to avoid the situation where only 
small-volume manufacturers of high-performance engines need to make 
design changes to reduce these emissions. Our concern is that a large 
manufacturer would otherwise be able to use emission credits to avoid 
making design changes to their high-performance engines. This emission 
level is consistent with measured HC+NOX emission values 
from these engines showing a range of emission levels with different 
types of fuel systems and different calibrations, as shown in the Draft 
RIA. Treating small-volume manufacturers of high-performance engines 
differently may be appropriate because they have little or no access to 
emission credits.
    Second, we could alternatively set the high-performance engine 
HC+NOX standard in the range of 15 to 22 g/kW-hr for all 
companies and disallow the use of emission credits for meeting this 
standard. This would require all companies to redesign their engines, 
rather than use emission credits, to reduce emissions to a standard 
that is tailored to high-performance engines.
    We request comment on the primary approach as well as the two 
alternatives for high-performance engine standards. Comment is 
requested on the costs and general positives and negatives of each 
approach. Comment is also requested on the technology required if a 
level above the proposed standards is supported, as well as information 
on safety and energy implications of the alternative emission 
standards. If a commenter supports either of the two alternative 
approaches, information and data are requested to assist EPA in setting 
the appropriate HC+NOX and CO emission standards within the 
15 to 22 g/kW-hr range.
    We are also aware that there may be some very small sterndrive or 
inboard engines. In particular, sailboats may have small propulsion 
engines for backup power. These engines would fall under the proposed 
definition of sterndrive/inboard engines, even though they are much 
smaller and may experience very different in-use operation. These 
engines may have more in common with marine auxiliary engines that are 
subject to land-based standards. Nevertheless, these engines share some 
important characteristics with bigger SD/I engines, such as reliance on 
four-stroke technology and access to water-based cooling. It is also 
true that emission standards are based on specific emission levels 
expected from engines of comparable sizes, so the standards adjust 
automatically with the size of the engine to require a relatively 
constant level of stringency. These engines are not like the very small 
outboard engines that are subject to less stringent standards because 
of their technical limitations in controlling emissions. Accordingly, 
we believe these engines can incorporate the same technologies as the 
bigger marine propulsion engines and meet the same emission standards. 
However, we request comment on the need for adjusting the emission 
standards for these engines to accommodate any technology constraints 
related to their unique designs. Specifically, we request comment on 
allowing manufacturers the option of certifying small SD/I engines to 
the proposed standards for auxiliary marine engines discussed in 
Section V.C.1. We also request comment on the possibility that some 
other small engines may inappropriately fall into the category of 
sterndrive/inboard engines. We request comment on the engine size for 
which any special accommodations must be made. Such comments should 
also address any issues that may exist for these engines with regard to 
meeting the proposed standards, or identify any other appropriate way 
of differentiating these engines from conventional sterndrive/inboard 
engines.
(2) Not-To-Exceed Standards
    We are proposing emission standards for an NTE zone representing a 
multiplier times the duty cycle standard for HC+NOX and for 
CO (see Sec.  1045.105). Section III.D.2 describes the proposed NTE 
test procedures and gives an overview of the proposed NTE provisions. 
In addition, Section III.D.2 presents the specific multipliers for the 
proposed NTE standards.
    The NTE approach is consistent with the concept of a weighted modal 
emission test such as the steady-state tests included in this rule. The 
proposed duty cycle standard itself is intended to represent the 
average emissions under steady-state conditions. Because it is an 
average, manufacturers design their engines with emission levels at 
individual points varying as needed to maintain maximum engine 
performance and still meet the engine standard. The NTE limit would be 
an additional requirement. It is intended to ensure that emission 
controls function with relative consistency across the full range of 
expected operating conditions.
(3) Emission Credit Programs
(a) Averaging, Banking, and Trading
    We are proposing averaging, banking, and trading of emission 
credits for sterndrive and inboard marine engines for meeting 
HC+NOX and CO standards (see Sec.  1045.105 and part 1045, 
subpart H). See Section VII.C.5 for a description of general provisions 
related to averaging, banking, and trading programs. Emission credit 
calculations would be based on the maximum engine power for an engine 
family, as described in Section IV.F.
    As with previous emission control programs, we are also proposing 
not to allow an emission family to earn credits for one pollutant if it 
is using credits to meet the standard for another pollutant. In other 
words, an engine family that does not meet the CO standard would not be 
able to earn HC+NOX emission credits, or vice versa. This 
should rarely be an issue for SD/I engines, because the same catalyst 
technology is effective for controlling HC+NOX and CO 
emissions. In addition, as with previous emission control programs, we 
are proposing that engines sold in California would not be included in 
this ABT program because they are already subject to California 
HC+NOX requirements.
    Credit generation and use is calculated based on the family 
emission limit (FEL) of the engine family and the standard. We are 
proposing FEL caps to prevent the sale of very-high emitting engines. 
For HC+NOX, the proposed FEL cap is 16 g/kW-hr for 
HC+NOX emissions from engines below 373 kW; this emission 
level is equal to the first phase of the California SD/I standards. We 
are proposing an FEL cap of 150 g/kW-hr for CO emissions from engines 
below 373 kW. These FEL caps represent the average baseline emission 
levels of SD/I engines, based on data described in the Draft RIA. The 
analogous figures for high-performance engines are 30 g/kW-hr for 
HC+NOX and 350 g/kW-hr for CO, as described in Section 
III.C.(d).
    Except as specified below for jet boat engines, we are proposing to 
keep OB/PWC engines and SD/I engines in separate averaging sets. This 
means that credits earned by SD/I and OB/PWC engines are counted 
separately and may not be exchanged to demonstrate

[[Page 28118]]

compliance with emission standards. Most of the engine manufacturers 
building SD/I engines do not also build OB/PWC engines. The exception 
to this is the largest manufacturer in both categories. We are 
concerned that allowing averaging, banking, or trading between OB/PWC 
engines and SD/I engines would not provide the greatest achievable 
reductions, because the level of the standard we are proposing is 
premised on the use of aftertreatment technology in SD/I engines, and 
is based on what is feasible for SD/I engines. We did not set the SD/I 
level based on the reductions achievable between OB/PWC and SD/I, but 
instead based on what is achievable by SD/I engines alone. The proposed 
limitation on ABT credits is consistent with this approach to setting 
the level of the SD/I standard. In addition, allowing such credit usage 
could provide an incentive to avoid the use of aftertreatment 
technologies in SD/I engines. This could create a competitive 
disadvantage for the many small manufacturers of SD/I engines that do 
not also produce OB/PWC engines.
    We propose that emission credits for SD/I engines have an unlimited 
credit life with no discounting. We consider these emission credits to 
be part of the overall program for complying with the proposed 
standards. Given that we may consider further reductions beyond these 
standards in the future, we believe it will be important to assess the 
ABT credit situation that exists at the time any further standards are 
considered. We would need to set such future emission standards based 
on the statutory direction that emission standards must represent the 
greatest degree of emission control achievable, considering cost, 
safety, lead time, and other factors. Emission credit balances will be 
part of the analysis for determining the appropriate level and timing 
of new standards. If we were to allow the use of credits generated 
under this proposed program for future, more stringent, standards, we 
may, depending on the level of emission credit banks, need to adopt 
emission standards at more stringent levels or with an earlier start 
date than we would absent the continued or limited use of existing 
emission credits. Alternatively, we could adopt future standards 
without allowing the use of existing emission credits.
    We are requesting comment on one particular issue regarding credit 
life. As proposed, credits earned under the exhaust ABT program would 
have an unlimited lifetime. This could result in a situation where 
credits generated by an engine sold in a model year are not used until 
many years later when the engines generating the credits have been 
scrapped and are no longer part of the fleet. EPA believes there may be 
value to limiting the use of credits to the period that the credit-
generating engines exist in the fleet. For this reason, EPA requests 
comment on limiting the lifetime of the credits to five years or, 
alternatively, to the regulatory useful life of the engine.
(b) Early-Credit Approaches
    We are proposing an early-credit program in which a manufacturer 
could earn emission credits before 2009 with early introduction of 
emission controls designed to meet the proposed standards (see Sec.  
1045.145). For engines produced by small-volume SD/I manufacturers that 
are eligible for the proposed two-year delay described in Section 
III.F.2, early credits could be earned before 2011. While we believe 
adequate lead time is provided to meet the proposed standards, we 
recognize that flexibility in timing could help some manufacturers--
particularly small manufacturers--to meet the new standards. Other 
manufacturers that are able to comply early on certain models would be 
better able to transition their full product line to the new standards 
by spreading out the transition over two years or more. Under this 
approach, we anticipate that manufacturers would generate credits 
through the use of catalysts.
    Manufacturers would generate these credits based on the difference 
between the measured emission level of the clean engines and an 
assigned baseline level (16 g/kW-hr HC+NOX and 150 g/kW-hr 
CO). These assigned baseline levels are based on data presented in 
Chapter 4 of the Draft RIA representing the average level observed for 
uncontrolled engines. We are also proposing to provide bonus credits to 
any manufacturer that certifies early to the proposed standard to 
provide a further incentive for introducing catalysts in SD/I engines. 
The bonus credits would take the form of a multiplier times the earned 
credits. The proposed multipliers are 1.25 for one year early, 1.5 for 
two years early, and 2.0 for three years early. For example, a small-
volume manufacturer certifying an engine to 5.0 g/kW-hr 
HC+NOX in 2009 (2 years early) would get a bonus multiplier 
of 1.5. Therefore, early HC+NOX credits would be calculated 
using the following equation: credits [grams] = (16-5) x Power [kW] x 
Useful Life [hours] x Load Factor x 1.5. We are proposing to use a load 
factor of 0.207, that is currently used in the OB/PWC calculations.
    To earn these credits, the engine would have to meet both the 
proposed HC+NOX and CO standards. These early credits would 
be treated the same as emission credits generated after the emission 
standards start to apply. This approach would provide an incentive for 
manufacturers to pull ahead significantly cleaner technologies. We 
believe such an incentive would lead to early introduction of catalysts 
on SD/I and help promote earlier market acceptance of this technology. 
Because of the proposed credit life, these credits would only be able 
to be used during the transition period to the new standards. We 
believe this proposed early credit program will allow manufactures to 
comply to the proposed standards in an earlier time frame than they 
would otherwise because it allows them to spread out their development 
resources over multiple years. To ensure that manufacturers do not 
generate credits for already required activities, no credits would be 
generated for the proposed federal program for engines that are 
produced for sale in California. We request comment on this approach.
    Alternatively, we request comment on the alternative of an early 
``family banking'' approach. Under this approach, we would allow 
manufacturers to certify an engine family early to the proposed 
standards. For each year of certifying engines early, the manufacturer 
would be able to delay certification of a comparable number of engines 
by one year, taking into account the relative power ratings of the 
different engine families. This would be based on the actual sales and 
would require no calculation or accounting of emission credits. This 
approach would not provide the same degree of precision as the early-
credit program described above, but it may be an effective way of 
helping manufacturers make the transition to new emission standards. 
See 40 CFR 1048.145(a) for an example of regulations that implement 
such a family banking program.
    We request comment on the above early-credit approaches or any 
other approach that would help manufacturers bring the product lines 
into compliance with the proposed standards without compromising 
overall emission reductions. Any allowance for high-emitting or late-
compliant engines should be offset by emission controls that achieve 
emission reductions beyond that required by the new standards. We 
request comment on the merits of the various approaches noted above and 
others that commenters may wish to suggest. We request that commenters 
provide detailed comments on how the approaches described above

[[Page 28119]]

should be set up, enhanced, or constrained to ensure that they serve 
their purpose without diminishing the overall effectiveness of the 
standards.
(c) Jet Boats
    Sterndrive and inboard vessels are typically propelled by 
traditional SD/I engines based on automotive engine blocks. As 
explained in Section IV, we are proposing to amend the definition of 
personal watercraft engine to ensure that engines used on jet boats 
would no longer be classified as personal watercraft engines but 
instead as SD/I engines because jet boats are more comparable to SD/I 
vessels. However, manufacturers in some cases make these jet boats by 
installing an engine also used in outboard or personal watercraft 
applications (less than 4 meters in length) and coupling the engine to 
a jet drive for propelling the jet boat. Thus, manufacturers of 
outboard or personal watercraft engines may also manufacture the same 
or similar engine for use on what we would propose here to be 
considered a jet boat (whose engine we would therefore proposed to be 
subject to SD/I standards).
    We are proposing to allow some flexibility in meeting new emission 
standards for jet boat engines because they are currently designed to 
use engines derived from OB/PWC applications and because of their 
relatively low sales volumes. We are also proposing to allow 
manufacturers to use emission credits generated from outboard and 
personal watercraft engines to demonstrate that their jet boat engines 
meet the proposed HC+NOX and CO standards for SD/I engines 
(see Sec.  1045.660 and Sec.  1045.701). We further propose that such 
engine manufacturers may only use this provision if the engines are 
certified as outboard or personal watercraft engines, and if the 
majority of units sold in the United States from those related engine 
families are sold for use as outboard or personal watercraft engines. 
We would decide whether a majority of engine units are sold for use as 
outboard or personal watercraft engines based on projected sales 
volumes from the application for certification. Manufacturers would 
need to group SD/I engines used for jet boats in a separate engine 
family from the outboard or personal watercraft engine to ensure proper 
labeling and calculation of emission credits, but manufacturers could 
rely on emission data from the same prototype engine for certifying 
both engine families. Finally, we propose that manufacturers of jet 
boat engines subject to SD/I standards and using credits from outboard 
or personal watercraft engines must certify these jet boat engines to 
an FEL that meets or exceed the standards for outboard and personal 
watercraft engines. This limits the degree to which manufacturers may 
take advantage of emission credits to produce engines that are emitting 
at higher levels than competitive engines. As such, the FELs for these 
engines must therefore be at or below the proposed emission standards 
for outboard and personal watercraft engines.
(d) SD/I High-Performance Engines
    We are proposing that the ABT program described above (III.C.3(a) 
through (c)) would also include SD/I high-performance engines. 
Manufacturers would be able to use emission credits from conventional 
SD/I engines to offset credit deficits from higher-emitting SD/I high-
performance engines. Although SD/I high-performance engines represent 
fewer than 1 percent of total SD/I engine sales, there are many more 
companies producing SD/I high-performance engines than conventional SD/
I engines. Because of the relatively small sales of these engines, a 
large manufacturer with a broad product line could readily offset a 
potential credit deficit by using credits from high-volume SD/I 
engines. In contrast, most manufacturers of SD/I high-performance 
engines are small businesses that do not also produce conventional SD/I 
engines. Section III.F discusses special provisions intended to reduce 
the burden for small businesses to meet the proposed standards. We 
request comment on whether this ABT program would create a competitive 
disadvantage for small businesses.
    We are proposing an approach in which manufacturers can use default 
emission factor of 30 g/kW-hr for HC+NOX emissions and 350 
g/kW-hr for CO emissions in lieu of testing for certification. For 
purposes of this ABT program these default emission factors, if used in 
lieu of testing, would be used for certification to an FEL at these 
levels. Thus, the emission credits needed would be the difference 
between the default levels and the applicable standard (see Sec.  
1045.240). These default emission levels represent the highest emission 
rates observed on uncontrolled engines. Manufacturers would always have 
the option of conducting tests to establish a measured emission rate to 
reduce or eliminate the need to use emission credits. While this 
testing may require additional setup and preparation, we believe it 
would be possible even for the most high-powered engines. To avoid the 
possibility of manufacturers selectively taking advantage of the 
default values, we would require them to rely on measured values for 
both HC+NOX and CO emissions if they do testing.
    For the purposes of the credit calculations, we are proposing to 
use an hours term longer than the proposed useful life for these 
engines. The proposed useful life for traditional SD/I engines is 
intended to reflect the full useable life of the engine. For high-
performance engines the proposed useful life is intended to reflect the 
expected time until the engine is rebuilt. High-performance engines are 
typically rebuilt several times. In fact, manufacturers have indicated 
that it is common for the boat owner to own two pairs of engines so 
that they can use one pair while the other is being rebuilt. Therefore, 
the proposed useful life does not reflect the full life of the engine, 
including rebuilds, over which emission credits would be used (or 
generated). We are proposing, for purposes of the credit calculations, 
that a life of 480 hours would be used for high-performance SD/I 
engines at or below 485 kW and 250 hours for engines above 485 kW. We 
request comment on the number of times that high-performance engines 
are typically rebuilt and how the number of rebuilds should be 
addressed in the credit calculations.
(4) Crankcase Emissions
    Due to blowby of combustion gases and the reciprocating action of 
the piston, exhaust emissions can accumulate in the crankcase. 
Uncontrolled engine designs route these vapors directly to the 
atmosphere. Closed crankcases have become standard technology for 
automotive engines and for outboard and personal watercraft engines. 
Manufacturers generally do this by routing crankcase vapors through a 
valve into the engine's air intake system. We propose to require 
manufacturers to prevent crankcase emissions from SD/I marine engines 
(see Sec.  1045.115). Because automotive engine blocks are already 
tooled for closed crankcases, the cost of adding a valve for positive 
crankcase ventilation is small for SD/I engines. Even with non-
automotive blocks, the tooling changes necessary for closing the 
crankcase are straight-forward.
(5) Durability Provisions
    We rely on pre-production certification, and other programs, to 
ensure that engines control emissions throughout their intended 
lifetime of operation. Section VII describes how we are proposing to 
require manufacturers to incorporate laboratory aging in the 
certification process, how we limit the

[[Page 28120]]

extent of maintenance that manufacturers may specify to keep engines 
operating as designed, and other general provisions related to 
certification. The following sections describe additional provisions 
that are specific to SD/I engines.
(a) Useful Life
    We are proposing to specify a useful life period of 480 hours or 
ten years, whichever comes first. The engines would be subject to the 
emission standards during this useful life period. This is consistent 
with the requirements adopted by California ARB (see Sec.  1045.105). 
We are further proposing that the 480-hour useful life period is a 
baseline value, which may be extended if data show that the average 
service life for engines in the family is longer. For example, we may 
require that the manufacturer certify the engine over a longer useful 
life period that more accurately represents the engines' expected 
operating life if we find that in-use engines are typically operating 
substantially more than 480 hours. This approach is similar to what we 
adopted for recreational vehicles.
    For high-performance SD/I engines (at or above 373 kW), we are 
proposing a useful life of 150 hours or 3 years for engines at or below 
485 kW and a useful life of 50 hours or 1 year for engines above 485 
kW. Due to the high power and high speed of these engines, mechanical 
parts are often expected to wear out quickly. For instance, one 
manufacturer indicated that some engines above 485 kW have scheduled 
head rebuilds between 50 and 75 hours of operation. These proposed 
useful life values are consistent with the California ARB regulations 
for high-performance SD/I engines. We request comment on the proposed 
useful life requirements for high performance marine engines.
    Some SD/I engines below 373 kW may be designed for high power 
output even though they do not reach the power threshold to qualify as 
SD/I high-performance engines. Because they do not qualify for the 
shorter useful life that applies to SD/I high-performance engines, they 
would be subject to the default value of 480 hours for other SD/I 
engines. However, to address the limited operating life for engines 
that are designed for especially high power output, we are proposing to 
allow manufacturers to request a shorter useful life for such an engine 
family based on information showing that engines in the family rarely 
operate beyond the requested shorter period. For example, if engines 
designed for extremely high performance are typically rebuilt after 250 
hours of operation, this would form the basis for establishing a 
shorter useful life period for those engines. See the proposed 
regulations for additional detail in establishing a shorter useful 
life.
(b) Warranty Periods
    We are proposing that manufacturers must provide an emission-
related warranty during the first 3 years or 480 hours of engine 
operation, whichever comes first (see Sec.  1045.120). This warranty 
period would apply equally to emission-related electronic components on 
SD/I high-performance engines. However, we are proposing shorter 
warranty periods for emission-related mechanical components on SD/I 
high-performance engines because these parts are expected to wear out 
more rapidly than comparable parts on traditional SD/I engines. 
Specifically, we are proposing a warranty period for emission-related 
mechanical components of 3 years or 150 hours for engines between 373 
and 485 kW, and 1 year or 50 hours for engines above 485 kW. These 
proposed warranty periods are the same as those adopted by the 
California ARB.
    If the manufacturer offers a longer warranty for the engine or any 
of its components at no additional charge, we propose that the 
emission-related warranty for the respective engine or component must 
be extended by the same amount. The emission-related warranty includes 
components related to controlling exhaust, evaporative, and crankcase 
emissions from the engine. This approach to setting warranty 
requirements is consistent with provisions that apply in most other 
programs for nonroad engines.
(6) Engine Diagnostics
    We are proposing to require that manufacturers design their SD/I 
engines to diagnose malfunctioning emission control systems starting 
with the introduction of the proposed standards (see Sec.  1045.110). 
As discussed in the Draft RIA, three-way catalyst systems with closed-
loop fueling control work well only when the air-fuel ratios are 
controlled to stay within a narrow range around stoichiometry. Worn or 
broken components or drifting calibrations over time can prevent an 
engine from operating within the specified range. This increases 
emissions and can lead to significantly increased fuel consumption and 
engine wear. The operator may or may not notice the change in the way 
the engine operates. We are not proposing to require similar diagnostic 
controls for OB/PWC or Small SI engines because the anticipated 
emission control technologies for these other applications are 
generally less susceptible to drift and gradual deterioration. We have 
adopted similar diagnostic requirements for Large SI engines operating 
in forklifts and other industrial equipment that also use three-way 
catalysts to meet emission standards.
    This diagnostic requirement focuses solely on maintaining 
stoichiometric control of air-fuel ratios. This kind of design detects 
problems such as broken oxygen sensors, leaking exhaust pipes, fuel 
deposits, and other things that require maintenance to keep the engine 
at the proper air-fuel ratio.
    Diagnostic monitoring provides a mechanism to help keep engines 
tuned to operate properly, with benefits for both controlling emissions 
and maintaining optimal performance. There are currently no inspection 
and maintenance programs for marine engines, so the most important 
variable in making the emission control and diagnostic systems 
effective is in getting operators to repair the engine when the 
diagnostic light comes on. This calls for a relatively simple design to 
avoid signaling false failures as much as possible. The diagnostic 
requirements in this rule therefore focus on detecting inappropriate 
air-fuel ratios, which is the most likely failure mode for three-way 
catalyst systems. The malfunction indicator light must go on when an 
engine runs for a full minute under closed-loop operation without 
reaching a stoichiometric air-fuel ratio.
    California ARB has adopted diagnostic requirements for SD/I engines 
that involve a more extensive system for monitoring catalyst 
performance and other parameters. We would accept a California-approved 
system as meeting EPA requirements. However, we believe the simpler 
system described above is better matched to the level of emission 
control involved, and is more appropriate in the context of 
recreational boating by consumers who are not subject to any systematic 
requirements for inspecting or maintaining their engines.
    The proposed regulations direct manufacturers to follow standard 
practices defined in documents adopted by the International 
Organization for Standardization (ISO) that establish protocols for 
automotive systems. The proposed regulations also state that we may 
approve variations from these industry standards, because individual 
manufacturers may have systems with unique operating parameters that 
warrant a deviation from the automotive approach. Also, if a new 
voluntary consensus standard is adopted to define appropriate practices 
for marine

[[Page 28121]]

engines, we would expect to incorporate that new standard into our 
regulations. See Sec.  1045.110 of the draft regulations for more 
information.

D. Test Procedures for Certification

(1) General Provisions
    The proposed test procedures are generally the same for both SD/I 
and OB/PWC engines. This involves laboratory measurement of emissions 
while the engine operates on the ISO E4 duty cycle. This is a five-mode 
steady-state duty cycle including an idle mode and four modes lying on 
a propeller curve with an exponent of 2.5, as shown in Appendix II to 
part 1045 of the draft regulations. The International Organization for 
Standardization (ISO) intended for this cycle to be used for 
recreational spark-ignition marine engines installed in vessels up to 
24 m in length. Because most or all vessels over 24 m have diesel 
engines, we believe the E4 duty cycle is most appropriate for SD/I 
engines covered by this rule. There may be some spark-ignition engines 
installed in vessels somewhat longer than 24 m, but we believe the E4 
duty cycle is no less appropriate in these cases. See Section IV.D for 
a discussion of adjustments to the test procedures related to the 
migration to 40 CFR part 1065, testing with a ramped-modal cycle, 
determining maximum test speed for denormalizing the duty cycle, and 
testing at higher altitudes.
    The E4 duty cycle is gives a weighting of 40 percent for idle. 
High-performance engine manufacturers have expressed their belief that 
the E4 duty cycle overstates the idle fraction of operation of high-
performance engines. They stated that these engines are rarely operated 
at idle and are therefore primarily designed for mid-range and high-
power operation at the expense of rough idle operation. We request 
comment on whether the modes for the proposed duty cycle should be 
reweighted toward higher power for high-performance engines. Commenters 
should support their assertions with data on high-performance engine 
use. If constructive data are forthcoming, we may finalize an 
alternative cycle weighting for high-performance engines based on this 
data.
(2) Not-to-Exceed Test Procedures and Standards
    We are proposing not-to-exceed (NTE) requirements similar to those 
established for marine diesel engines. Engines would be required to 
meet the NTE standards during normal in-use operation. We request 
comment on applying the proposed NTE requirements to spark-ignition 
marine engines and on the application of the requirements to these 
engines.
(a) Concept
    Our goal is to achieve control of emissions over a wide range of 
ambient conditions and over the broad range of in-use speed and load 
combinations that can occur on a marine engine. This would ensure real-
world emission control, rather than just controlling emissions under 
certain laboratory conditions. An important tool for achieving this 
goal is an in-use testing program with an objective standard and an 
easily implemented test procedure. Our traditional approach has been to 
set a numerical standard on a specified test procedure and rely on the 
additional prohibition of defeat devices to ensure in-use control over 
a broad range of operation not included in the test procedure.
    We are proposing to apply the same prohibition on defeat devices 
for OB/PWC and SD/I engines (see Sec.  1045.115).
    No single test procedure or test cycle can cover all real-world 
applications, operations, or conditions. Yet to ensure that emission 
standards are providing the intended benefits in use, we must have a 
reasonable expectation that emissions under real-world conditions 
reflect those measured on the test procedure. The defeat device 
prohibition is designed to ensure that emission controls are employed 
during real-world operation, not just under laboratory testing 
conditions. However, the defeat device prohibition is not a quantified 
standard and does not have an associated test procedure, so it does not 
have the clear objectivity and ready enforceability of a numerical 
standard and test procedure. We believe using the traditional approach, 
i.e., using only a standardized laboratory test procedure and test 
cycle, makes it difficult to ensure that engines will operate with the 
same level of control in use as in the laboratory.
    Because the proposed duty cycle uses only five modes on an average 
propeller curve to characterize marine engine operation, we are 
concerned that an engine designed to the proposed duty cycle would not 
necessarily perform the same way over the range of speed and load 
combinations seen on a boat. This proposed duty cycle is based on an 
average propeller curve, but a marine propulsion engine may never be 
fitted with an ``average propeller.'' For instance, an engine fit to a 
specific boat may operate differently based on how heavily the boat is 
loaded.
    To ensure that engines control emissions over the full range of 
speed and load combinations seen on boats, we propose to establish a 
zone under the engine's power curve where the engine may not exceed a 
specified emission limit (see Sec.  1045.105 and Sec.  1045.515). This 
limit would apply to all regulated pollutants during steady-state 
operation. In addition, we propose that a wide range of real ambient 
conditions be included in testing with this NTE zone. The NTE zone, 
limit, and ambient conditions are described below.
    We believe there are significant advantages to establishing NTE 
standards. The proposed NTE test procedure is flexible, so it can 
represent the majority of in-use engine operation and ambient 
conditions. The NTE approach thus takes all the benefits of a numerical 
standard and test procedure and expands it to cover a broad range of 
conditions. Also, laboratory testing makes it harder to perform in-use 
testing because either the engines would have to be removed from the 
vessel or care would have to be taken to achieve laboratory-type 
conditions on the vessel. With the NTE approach, in-use testing and 
compliance become much easier since emissions may be sampled during 
normal boating. By establishing an objective measurement, this approach 
makes enforcement of defeat device provisions easier and provides more 
certainty to the industry.
    Even with the NTE requirements, we believe it is still appropriate 
to retain standards based on the steady-state duty cycle. This is the 
standard that we expect the certified marine engines to meet on average 
in use. The NTE testing is focused more on maximum emissions for 
segments of operation and, in most cases, would not require additional 
technology beyond what is used to meet the proposed standards. In some 
cases, the calibration of the engine may need to be adjusted. We 
believe that basing the emission standards on a distinct cycle and 
using the NTE zone to ensure in-use control creates a comprehensive 
program.
    We believe the technology used to meet the standards over the five-
mode duty cycle will meet the caps that apply across the NTE zone. We 
therefore do not expect the proposed NTE standards to cause 
manufacturers to need additional technology. We believe the NTE 
standard will not result in a large amount of additional testing, 
because these engines should be designed to perform as well in use as 
they do over the five-mode test. However, our cost analysis in the 
Draft RIA accounts for some additional testing, especially in the early 
years, to provide

[[Page 28122]]

manufacturers with assurance that their engines would meet the proposed 
NTE requirements.
(b) Shape of NTE Zone
    Figure III-1 illustrates our proposed NTE zone for SD/I engines. We 
developed this zone based on the range of conditions that these engines 
typically see in use. Manufacturers collected data on several engines 
installed on vessels and operated under light and heavy load. Chapter 4 
of the Draft RIA presents this data and describes the development of 
the boundaries and conditions associated with the proposed NTE zone. 
Although significant in-use engine operation occurs at low speeds, we 
are excluding operation below 40 percent of maximum test speed because 
brake-specific emissions increase dramatically as power approaches 
zero. An NTE limit for low-speed or low-power operation would be very 
hard for manufacturers and EPA to implement in a meaningful way. We are 
proposing NTE limits for the subzones shown in Figure III-1, as 
described below. We request comment on the proposed NTE zone and 
subzones.
[GRAPHIC] [TIFF OMITTED] TP18MY07.000

    We propose to allow manufacturers to request approval for 
adjustments to the size and shape of the NTE zone for certain engines, 
if they can show that the engine will not see operation outside of the 
revised NTE zone in use (see Sec.  1045.515). We would not want 
manufacturers to go to extra lengths to design and test their engines 
to control emissions for operation that will not occur in use. However, 
manufacturers would still be responsible for all operation of an engine 
on a vessel that would reasonably be expected to be seen in use, and 
they would be responsible for ensuring that their specified operation 
is indicative of real-world operation. In addition, if a manufacturer 
designs an engine for operation at speeds and loads outside of the 
proposed NTE zone, the manufacturer would be responsible for notifying 
us so the NTE zone can be modified appropriately to include this 
operation for that engine family.
(c) Excluded Operation
    As with marine diesel engines, we are proposing that only steady-
state operation be included for NTE testing (see Sec.  1045.515). 
Steady-state operation would generally mean setting the throttle (or 
speed control) in a fixed position. We believe most operation with 
Marine SI engines involves nominally steady-state operator demand. It 
is true that boats often experience rapid accelerations, such as with 
water skiing. However, boats are typically designed for planing 
operation at relatively high speeds. This limits the degree to which we 
would expect engines to experience frequent accelerations during 
extended operation. Also, because most of the transient events involve 
acceleration from idle to reach a planing condition, most transient 
engine operation is outside the NTE zone and would therefore not be 
covered by NTE testing anyway. Moreover, we believe OB/PWC and SD/I 
engines designed to comply with steady-state NTE requirements will be 
using technologies that also work effectively under the changing speed 
and load conditions that may occur. If we find there is substantial 
transient operation within the NTE zone that causes significantly 
increased emissions from installed engines, we will revisit

[[Page 28123]]

this provision in the future. We request comment on the appropriateness 
of excluding transient operation from NTE requirements.
    We are aware that SD/I engines may not be able to meet emission 
standards under all conditions, such as times when emission control 
must be compromised for startability or safety. We are proposing to 
specify that NTE testing excludes engine starting and warm-up. We would 
allow manufacturers to design their engines to utilize engine 
protection strategies that would not be covered by defeat device 
provisions or NTE standards. This is analogous to the tampering 
exemptions incorporated into 40 CFR 1068.101(b)(1) to address 
emergencies. We believe it is appropriate to allow manufacturers to 
design their engines with ``limp-home'' capabilities to prevent a 
scenario where an engine fails to function, leaving an operator on the 
water without any means of propulsion.
(d) NTE Emission Limits
    We are proposing NTE limits for the subzones shown in Figure III-1 
above based on data collected from several SD/I engines equipped with 
catalysts. These data and our analysis are presented in Chapter 4 of 
the Draft RIA. See Section IV.C for a discussion of NTE limits for OB/
PWC engines.
    Because the proposed NTE zone does not include the idle point, 
which is weighted at 40 percent of the certification duty cycle, brake-
specific emissions throughout most of the proposed NTE zone are less 
than the weighted average from the steady-state testing. For most of 
the NTE zone, we are therefore proposing a limit equal to the duty 
cycle standard (i.e., NTE multiplier = 1.0). However, data on low-
emission engines show that brake-specific emissions increase for engine 
speeds below 50 percent of maximum test speed (Subzone 4). We are 
therefore proposing an HC+NOX cap of 1.5 times the 
certification level in Subzone 4. Emission data on catalyst-equipped 
engines also show higher emissions near full-power operation. We 
understand that richer air-fuel ratios are needed under high-power 
operation to protect the engines from overheating. We are therefore 
proposing higher NTE limits for engine speeds at or above 90 percent of 
rated test speed and at or above 100 percent of peak torque measured at 
the rated test speed (Subzone 1). Specifically, we are proposing an 
HC+NOX cap of 1.5 times the duty cycle standard and a CO cap 
of 3.5 times the duty cycle standard for Subzone 1. We request comment 
on the proposed NTE limits for SD/I engines. These limits are 
summarized in Table III-1.

                          Table III-1.--Proposed NTE Limits by Subzone for SD/I Engines
----------------------------------------------------------------------------------------------------------------
                    Pollutant                        Subzone 1       Subzone 2       Subzone 3       Subzone 4
----------------------------------------------------------------------------------------------------------------
HC+NOX..........................................             1.5             1.0             1.0             1.5
CO..............................................             3.5             1.0             1.0             1.0
----------------------------------------------------------------------------------------------------------------

    SD/I engine manufacturers have begun developing prototype engines 
with catalysts, and one manufacturer is currently selling SD/I engines 
equipped with catalysts. These manufacturers have indicated that they 
begin moving to richer air-fuel ratio calibrations at torque values 
greater than 80 percent of maximum. These richer air-fuel ratios give 
more power but because more fuel is burned also lead to higher 
hydrocarbon and carbon monoxide emission rates. Part of the 
manufacturers' rationale in selecting the appropriate air-fuel ratio in 
this type of operation is to protect the engine by minimizing excess 
air, which would lead to greater engine temperatures as increased 
combustion of fuel and exhaust gases. To avoid the adverse effects of 
this potential for overheating, we request comment on whether subzone 1 
should be expanded to accommodate the engine-protection strategies 
needed for SD/I engines at high power. In addition, we request comment 
on the proposed NTE limits in subzone 1 with respect to open-loop 
engine operation, especially for carbon monoxide.
    Marine engine manufacturers have suggested alternative approaches 
to setting NTE limits for marine engines, which are discussed in 
Section IV.C.2. Largely, these suggestions have been made to address 
the emission variability between test modes seen in direct-injection 
two-stroke outboard and PWC engines. However, we request comment on 
alternative approaches for SD/I engines as well.
(e) Ambient Conditions
    Variations in ambient conditions can affect emissions. Such 
conditions include air temperature, water temperature, and barometric 
pressure, and humidity. We are proposing to apply the comparable ranges 
for these variables as for marine diesel engines (see Sec.  1045.515). 
Within the ranges, there is no calculation to correct measured 
emissions to standard conditions. Outside of the ranges, emissions 
could be corrected back to the nearest end of the range using good 
engineering practice. The proposed ranges are 13 to 35 [deg]C (55 to 95 
[deg]F) for ambient air temperature, 5 to 27 [deg]C (41 to 80 [deg]F) 
for ambient water temperature, and 94.0 to 103.325 kPa for atmospheric 
pressure. We do not specify a range of humidity values, but propose 
only to require that laboratory testing be conducted at humidity levels 
representing in-use conditions.
(f) Measurement Methods
    While it may be easier to test outboard engines in the laboratory, 
there is a strong advantage to using portable measurement equipment to 
test SD/I engines and personal watercraft without removing the engine 
from the vessel. Field testing would also provide a much better means 
of measuring emissions to establish compliance with the NTE standards, 
because it is intended to ensure control of emissions during normal in-
use operation that may not occur during laboratory testing over the 
specified duty cycle. We propose to apply the field testing provisions 
for all SD/I engines. These field-testing procedures are described 
further in Section IV.E.2.d. We request comment on any ways the field 
testing procedures should be modified to address the unique operating 
characteristics of marine engines.
    A parameter to consider is the minimum sampling time for field 
testing. A longer period allows for greater accuracy, due mainly to the 
smoothing effect of measuring over several transient events. On the 
other hand, an overly long sampling period can mask areas of engine 
operation with poor emission control characteristics. To balance these 
concerns, we are applying a minimum sampling period of 30 seconds. This 
is consistent with the requirement for marine diesel engines. Spark-
ignition engines generally don't have turbochargers and they control 
emissions largely by maintaining air-fuel ratio. Spark-ignition engines 
are therefore much less prone to consistent

[[Page 28124]]

emission spikes from off-cycle or unusual engine operation. We believe 
the minimum 30 second sampling time will ensure sufficient measurement 
accuracy and will allow for meaningful measurements.
    We do not specify a maximum sampling time. We expect manufacturers 
testing in-use engines to select an approximate sampling time before 
measuring emissions; however, the standards apply for any sampling time 
that meets the minimum.
(g) Certification
    We propose to require that manufacturers state in their application 
for certification that their engines will comply with the NTE standards 
under any nominally steady-state combination of speeds and loads within 
the proposed NTE zone (see Sec.  1045.205). The manufacturer would also 
provide a detailed description of all testing, engineering analysis, 
and other information that forms the basis for the statement. This 
statement would be based on testing and, if applicable, other research 
that supports such a statement, consistent with good engineering 
judgment. We would be able to review the basis for this statement 
during the certification process. For marine diesel engines, we have 
provided guidance that manufacturers may demonstrate compliance with 
NTE standards by testing their engines at a number of standard points 
throughout the NTE zone. In addition, manufacturers must test at a few 
random points chosen by EPA prior to the testing. We request comment on 
this approach for Marine SI engines.

E. Additional Certification and Compliance Provisions

(1) Production Line Testing
    We are proposing to require that manufacturers routinely test 
engines at the point of production to ensure that production 
variability does not affect the engine family's compliance with 
emission standards (see part 1045, subpart D). These proposed testing 
requirements are the same as we are proposing for outboard and personal 
watercraft engines and are very similar to those already in place in 
part 91. See Section VII.C.7 and the draft regulations for a detailed 
description of these requirements. We may also require manufacturers to 
perform production line testing under the selective enforcement 
auditing provisions described in Section VIII.E.
(2) In-Use Testing
    Manufacturers of OB/PWC engines have been required to test in-use 
engines to show that they continue to meet emission standards. We 
contemplated a similar requirement for SD/I engines, but have decided 
not to propose a requirement for a manufacturer-run in-use testing 
program at this time. Manufacturers have pointed out that it would be 
very difficult to identify a commercial fleet of boats that could be 
set up to operate for hundreds of hours, because it is very uncommon 
for commercial operators to have significant numbers of SD/I vessels. 
Where there are commercial fleets of vessels that may be conducive to 
accelerated in-use service accumulation, these vessels generally use 
outboard engines. Manufacturers could instead hire drivers to operate 
the boats, but this may be cost-prohibitive. We request comment on any 
other alternative approaches that might be available for accumulating 
operating hours with SD/I engines. For example, to the extent that boat 
builders maintain a fleet of boats for product development or 
employees' recreational use, those engines may be available for 
emission testing after in-use operation.
    There is also a question about access to the engines for testing. 
If engines need to be removed from vessels for testing in the 
laboratory, it is unlikely that owners would cooperate. However, we are 
proposing test procedures with specified portable equipment that would 
potentially allow for testing engines that remain installed in boats. 
This is described in Section IV.E.2.d.
    While we are not proposing a program to require manufacturers to 
routinely test in-use engines, the Clean Air Act allows us to perform 
our own testing at any time with in-use engines to evaluate whether 
they continue to meet emission standards throughout the useful life. 
This may involve either laboratory testing or in-field testing with 
portable measurement equipment. For laboratory tests, we could evaluate 
compliance with either the duty cycle standards or the not-to-exceed 
standards. For testing with engines that remain installed on marine 
vessels, we would evaluate compliance with the not-to-exceed standards. 
In addition, we may require the manufacturer to conduct a reasonable 
degree of testing under Clean Air Act section 208 if we have reason to 
believe that an engine family does not conform to the regulations. This 
testing may take the form of a Selective Enforcement Audit, or we may 
require the manufacturer to test in-use engines.
(3) Certification Fees
    Under our current certification program, manufacturers pay a fee to 
cover the costs for various certification and other compliance 
activities associated with implementing the emission standards. As 
explained below, we are proposing to assess EPA's compliance costs 
associated with SD/I engines based on EPA's existing fees regulation. 
Section VI describes our proposal to establish a new fees category, 
based on the cost study methodology used in establishing EPA's existing 
fees regulation, for costs related to the proposed evaporative emission 
standards for both vessels and equipment that would be subject to 
standards under this proposal.
    EPA established a fee structure by grouping together various 
manufacturers and industries into fee categories, with an explanation 
that separation of industries into groups was appropriate to tailor the 
applicable fee to the level of effort expected for EPA to oversee the 
range of certification and compliance responsibilities (69 FR 26222, 
May 11, 2004). As part of this process, EPA conducted a cost analysis 
to determine the various compliance activities associated with each fee 
category and EPA's associated annual cost burden. Once the total EPA 
costs were determined for each fee category, the total number of 
certificates involved within a fee category was added together and 
divided into the total costs to determine the appropriate assessment 
for each anticipated certificate.\77\ One of the fee categories created 
was for ``Other Engines and Vehicles,'' which includes marine engines 
(both compression-ignition and spark-ignition), nonroad spark-ignition 
engines (above and below 19 kW), locomotive engines, recreational 
vehicles, heavy-duty evaporative systems, and heavy-duty engines 
certified only for sale in California. These engine and vehicle types 
were grouped together because EPA planned a more basic certification 
review than, for example, light-duty vehicles.
---------------------------------------------------------------------------

    \77\ See Cost Analysis Document at p. 21 associated with the 
proposed fees rule (http://www.epa.gov/otaq/fees.htm).
---------------------------------------------------------------------------

    EPA determined in the final fees rulemaking that it would be 
premature to assess fees for the SD/I engines since they were not yet 
subject to emission standards. The fee calculation nevertheless 
includes a projection that there will eventually be 25 certificates of 
conformity annually for SD/I engines. We are proposing to now formally 
include SD/I engines in the ``Other Engines and Vehicles'' category and

[[Page 28125]]

assess a fee of $839 for each certificate of conformity in 2006. Note 
that we will continue to update assessed fees each year, so the actual 
fee in 2009 and later model years will depend on these annual 
calculations (see Sec.  1027.105).
(4) Special Provisions Related to Partially Complete Engines
    It is common practice for Marine SI engines for one company to 
produce the base engine for a second company to modify for the final 
application. Since our regulations prohibit the sale of uncertified 
engines, we are proposing provisions to clarify the status of these 
engines and defining a path by which these engines can be handled 
without violating the regulations. See Section XI for more information.
(5) Use of Engines Already Certified to Other Programs
    In some cases, manufacturers may want to use engines already 
certified under our other programs. Engines certified to the emission 
standards for highway applications in part 86 or Large SI applications 
in part 1048 are meeting more stringent standards. We are therefore 
proposing to allow the pre-existing certification to be valid for 
engines used in marine applications, on the condition that the engine 
is not changed from its certified configuration in any way (see Sec.  
1045.605). Manufacturers would need to demonstrate that fewer than five 
percent of the total sales of the engine model are for marine 
applications. There are also a few minor notification and labeling 
requirements to allow for EPA oversight of this provision.
(6) Import-Specific Information at Certification
    We are proposing to require additional information to improve our 
ability to oversee compliance related to imported engines (see Sec.  
1045.205). In the application for certification, we are proposing to 
require the following additional information: (1) The port or ports at 
which the manufacturer will import the engines, (2) the names and 
addresses of the agents the manufacturer has authorized to import the 
engines, and (3) the location of the test facilities in the United 
States where the manufacturer will test the engines if we select them 
for testing under a selective enforcement audit.

F. Small-Business Provisions

(1) Small Business Advocacy Review Panel
    On June 7, 1999, we convened a Small Business Advocacy Review Panel 
under section 609(b) of the Regulatory Flexibility Act as amended by 
the Small Business Regulatory Enforcement Fairness Act of 1996. The 
purpose of the Panel was to collect the advice and recommendations of 
representatives of small entities that could be affected by this 
proposed rule and to report on those comments and the Panel's findings 
and recommendations as to issues related to the key elements of the 
Initial Regulatory Flexibility Analysis under section 603 of the 
Regulatory Flexibility Act. We convened a Panel again on August 17, 
2006 to update our review for this new proposal. The Panel reports have 
been placed in the rulemaking record for this proposal. Section 609(b) 
of the Regulatory Flexibility Act directs the review Panel to report on 
the comments of small entity representatives and make findings as to 
issues related to identified elements of an initial regulatory 
flexibility analysis (IRFA) under RFA section 603. Those elements of an 
IRFA are:
     A description of, and where feasible, an estimate of the 
number of small entities to which the proposed rule will apply;
     A description of projected reporting, recordkeeping, and 
other compliance requirements of the proposed rule, including an 
estimate of the classes of small entities that will be subject to the 
requirements and the type of professional skills necessary for 
preparation of the report or record;
     An identification, to the extent practicable, of all 
relevant Federal rules that may duplicate, overlap, or conflict with 
the proposed rule; and
     A description of any significant alternative to the 
proposed rule that accomplishes the stated objectives of applicable 
statutes and that minimizes any significant economic impact of the 
proposed rule on small entities.
    In addition to the EPA's Small Business Advocacy Chairperson, the 
Panel consisted of the Director of the Assessment and Standards 
Division of the Office of Transportation and Air Quality, the 
Administrator of the Office of Information and Regulatory Affairs 
within the Office of Management and Budget, and the Chief Counsel for 
Advocacy of the Small Business Administration.
    Using definitions provided by the Small Business Administration 
(SBA), companies that manufacture internal-combustion engines and that 
employ fewer than 1000 employees are considered small businesses for a 
Small Business Advocacy Review (SBAR) Panel. Equipment manufacturers, 
boat builders, and fuel system component manufacturers that employ 
fewer than 500 people are considered small businesses for the SBAR 
Panel. Based on this information, we asked 25 companies that met the 
SBA small business thresholds to serve as small entity representatives 
for the duration of the Panel process. Of these 25 companies, 13 were 
involved in the marine industry. These companies represented a cross-
section of SD/I engine manufacturers, boat builders, and fuel system 
component manufacturers.
    With input from small entity representatives, the Panel reports 
provide findings and recommendations on how to reduce potential burden 
on small businesses that may occur as a result of this proposed rule. 
The Panel reports are included in the rulemaking record for this 
proposal. In light of the Panel reports, and where appropriate, the 
agency has made changes to the provisions anticipated for the proposed 
rule. The proposed options recommended to us by the Panel are described 
below.
(2) Proposed Burden Reduction Approaches for Small-Volume SD/I Engine 
Manufacturers
    We are proposing several options for small-volume SD/I engine 
manufacturers. For purposes of determining which engine manufacturers 
are eligible for the small business provisions described below for SD/I 
engine manufacturers, we are proposing criteria based on a production 
cut-off of 5,000 SD/I engines per year. Under this approach, we would 
allow engine manufacturers that exceed the production cut-off level 
noted above to request treatment as a small business if they have fewer 
than the number of employees specified above. In such a case, the 
manufacturer would provide information to EPA demonstrating the number 
of employees in their employ. The proposed options would be used at the 
manufacturers' discretion. We request comment on the appropriateness of 
these options, which are described in detail below.
(a) Additional Lead Time
    One small business marine engine manufacturer is already using 
catalytic converters on some of its production SD/I marine engines 
below 373 kW. These engines have been certified to meet standards 
adopted by California ARB that are equivalent to the proposed 
standards. However, other small businesses producing SD/I engines have 
stated that they are not as far along in their catalyst development 
efforts. These manufacturers support the concept of receiving 
additional time for

[[Page 28126]]

compliance, beyond the implementation date for large manufacturers.
    High-performance SD/I engine manufacturers are typically smaller 
businesses than other SD/I engine manufacturers. The majority of high-
performance engine manufacturers produce fewer than 100 engines per 
year for sale in the United States, and some produce only a few engines 
per year. Due to these very low sales volumes, additional lead time may 
be useful to the manufacturers to help spread out the compliance 
efforts and costs.
    As recommended in the SBAR Panel report, EPA is proposing an 
implementation date of 2011 for SD/I engines below 373 kW produced by 
small business marine engine manufacturers and a date of 2013 for small 
business manufacturers of high-performance (at or above 373 kW) marine 
engines (see Sec.  1045.145). As discussed earlier, we have requested 
comment on alternative non-catalyst based standard of 22 g/kW-hr for 
high-performance SD/I marine engines. In the case of an alternative 
non-catalyst based standard, less lead time may be necessary. EPA 
requests comments on the proposed additional lead time in the 
implementation of the proposed SD/I exhaust emission standards for 
small businesses.
(b) Exhaust Emission ABT
    As discussed above, we are proposing an averaging, banking, and 
trading (ABT) credit program for exhaust emissions from SD/I marine 
engines (see part 1045, subpart H). Small businesses expressed some 
concern that ABT could give a competitive advantage to large 
businesses. Specifically, there was an equity concern that if credits 
generated by SD/I engines below 373 kW could be used for high-
performance SD/I engines, that one large manufacturer could use these 
credits to meet the high-performance SD/I engine standards without 
making any changes to their engines. EPA requests comment on the 
desirability of credit trading between high-performance and other SD/I 
marine engines and the impact it could have on small businesses.
(c) Early Credit Generation for ABT
    The SBAR Panel recommended an early banking program and expressed 
belief that bonus credits will provide greater incentive for more small 
business engine manufacturers to introduce advanced technology earlier 
across the nation than would otherwise occur. As discussed above, we 
are proposing an early banking program in which bonus credits could be 
earned for certifying early (see Sec.  1045.145). This program, 
combined with the additional lead time for small businesses, would give 
small-volume SD/I engine manufacturers ample opportunity to bank 
emission credits prior to the proposed implementation date of the 
standards.
(d) Assigned Emission Rates for High-Performance SD/I Engines
    Small businesses commented that certification may be too costly to 
amortize effectively over the small sales volumes for high-performance 
SD/I engines. One significant part of certification costs is engine 
testing. This includes testing for emissions over the specified duty 
cycle, deterioration testing, and not to exceed (NTE) zone testing. 
Even in the case where an engine manufacturer is using emission credits 
to comply with the standard, the manufacturer would still need to test 
engines to calculate how many emission credits are needed. One way of 
minimizing this testing burden would be to allow manufacturers to use 
assigned baseline emission rates for certification based on previously 
generated emission data. As discussed earlier in this preamble, we are 
proposing assigned baseline HC+NOX and CO emission rates for 
all high-performance SD/I engines. These assigned emission rates are 
based on test data presented in Chapter 4 of the Draft RIA.
(e) Alternative Standards for High-Performance SD/I Engines
    Small businesses expressed concern that catalysts have not been 
demonstrated on high-performance engines and that they may not be 
practicable for this application. In addition, the concern was 
expressed that emission credits may not be available at a reasonable 
price. As discussed earlier, we are requesting comment on the need for 
and level of alternative standards for high-performance marine engines.
    The proposed NTE standards discussed above would likely require 
additional certification and development testing. The SBAR Panel 
recommended that NTE standards not apply to any high-performance SD/I 
engines, as it would minimize the costs of compliance testing for small 
businesses. For these reasons, we are not proposing to apply NTE 
standards to high-performance SD/I engines (See Sec.  1045.105).
(f) Broad Engine Families for High-Performance SD/I Engines
    Testing burden could be reduced by using broader definitions of 
engine families. Typically in EPA engine and equipment programs, 
manufacturers are able to group their engine lines into engine families 
for certification to the standards. Engines in a given family must have 
many similar characteristics including the combustion cycle, cooling 
system, fuel system, air aspiration, fuel type, aftertreatment design, 
number of cylinders and cylinder bore sizes. A manufacturer would then 
perform emission tests only on the engine in that family that would be 
most likely to exceed an emission standard. We are proposing to allow 
small businesses to group all of their high performance SD/I engines 
into a single engine family for certification, subject to good 
engineering judgment (see Sec.  1045.230).
(g) Simplified Test Procedures for High-Performance SD/I Engines
    Existing testing requirements include detailed specifications for 
the calibration and maintenance of testing equipment and tolerances for 
performing the actual tests. For laboratory equipment and testing, 
these specifications and tolerances are intended to achieve the most 
repeatable results feasible given testing hardware capabilities. For 
in-use testing, EPA allows for different equipment than is specified 
for the laboratory and with arguably less restrictive specifications 
and tolerances. The purpose of separate requirements for in-use testing 
is to account for the variability inherent in testing outside of the 
laboratory. These less restrictive specifications allow for lower cost 
emission measurement devices, such as portable emission measurement 
units. For high performance SD/I engines, it may be difficult to hold 
the engine at idle or high power within the tolerances currently 
specified by EPA in the laboratory test procedure. Therefore, we are 
proposing less restrictive specifications and tolerances, for testing 
high performance SD/I engines, which would allow the use of portable 
emission measurement equipment (see Sec.  1065.901(b)). This would 
facilitate less expensive testing for these small businesses without 
having a negative effect on the environment.
(h) Reduced Testing Requirements
    We are proposing that small-volume engine manufacturers may rely on 
an assigned deterioration factor to demonstrate compliance with the 
standards for the purposes of certification rather than doing service 
accumulation and additional testing to measure deteriorated emission 
levels at the end of the regulatory useful life (see Sec.  1045.240). 
EPA is not proposing actual

[[Page 28127]]

levels for the assigned deterioration factors with this proposal. EPA 
intends to analyze available emission deterioration information to 
determine appropriate deterioration factors for SD/I engines. The data 
will likely include durability information from engines certified to 
California ARB's standards and may also include engines certified early 
to EPA's standards. Prior to the implementation date for the SD/I 
standards, EPA will provide guidance to engine manufacturers specifying 
the levels of the assigned deterioration factors for small-volume 
engine manufacturers.
    We are also proposing that small-volume engine manufacturers would 
be exempt from the production-line testing requirements (see Sec.  
1045.301). While we are proposing to exempt small-volume engine 
manufacturers from production line testing, we believe requiring 
limited production-line testing could be beneficial to implement the 
ongoing obligation to ensure that production engines are complying with 
the standards. Therefore, we request comment on the alternative of 
applying limited production-line testing to small-volume engine 
manufacturers with a requirement to test one production engine per 
year.
(i) Hardship Provisions
    We are proposing two types of hardship provisions for SD/I engine 
manufacturers consistent with the Panel recommendations. The first type 
of hardship is an unusual circumstances hardship, which would be 
available to all businesses regardless of size. The second type of 
hardship is an economic hardship provision, which would be available to 
small businesses only. Sections VIII.C.8 and VIII.C.9 provide a 
description of the proposed hardship provisions that would apply to SD/
I engine manufacturers.
    Because boat builders in many cases will depend on engine 
manufacturers to supply certified engines in time to produce complying 
boats, we are also proposing a hardship provision for all boat 
builders, regardless of size, that would allow the builder to request 
more time if they are unable to obtain a certified engine and they are 
not at fault and would face serious economic hardship without an 
extension (see Sec.  1068.255). Section VIII.C.10 provides a 
description of the proposed hardship provisions that would apply to 
boat builders.

G. Technological Feasibility

(1) Level of Standards
    Over the past few years, developmental programs have demonstrated 
the capabilities of achieving significant reductions in exhaust 
emissions from SD/I engines. California ARB has acted on this 
information to set an HC+NOX emission standard of 5 g/kW-hr 
for SD/I engines, starting in 2008. Chapter 4 of the Draft RIA presents 
data from several SD/I engines with catalysts packaged within water-
cooled exhaust manifolds. Four of these engines were operated with 
catalysts in vessels for 480 hours. The remaining engines were tested 
with catalysts that had been subjected to a rapid-aging cycle in the 
laboratory. Data from these catalyst-equipped engines generally show 
emission levels below the proposed standards.
(2) Implementation Dates
    We anticipate that manufacturers will use the same catalyst designs 
to meet the proposed standards that they will use to meet the 
California ARB standards for SD/I engines in 2008. We believe a 
requirement to extend the California standards nationwide after a one-
year delay allows manufacturers adequate time to incorporate catalysts 
across their product lines. Once the technology is developed for use in 
California, it would be available for use nationwide. In fact, one 
company currently certified to the California standards is already 
offering catalyst-equipped SD/I engines nationwide. As discussed above, 
we request comment on the effect that anticipated product changes for 
specific General Motors engine blocks may have on the proposed 
implementation dates.
(3) Technological Approaches
    Engine manufacturers can adapt readily available technologies to 
control emissions from SD/I engines. Electronically controlled fuel 
injection gives manufacturers more precise control of the air/fuel 
ratio in each cylinder, thereby giving them greater flexibility in how 
they calibrate their engines. With the addition of an oxygen sensor, 
electronic controls give manufacturers the ability to use closed-loop 
control, which is especially valuable when using a catalyst. In 
addition, manufacturers can achieve HC+NOX reductions 
through the use of exhaust gas recirculation. However, the most 
effective technology for controlling emissions is a three-way catalyst 
in the exhaust stream.
    In SD/I engines, the exhaust manifolds are water-jacketed and the 
water mixes with the exhaust stream before exiting the vessel. 
Manufacturers add a water jacket to the exhaust manifold to meet 
temperature-safety protocol. They route this cooling water into the 
exhaust to protect the exhaust couplings and to reduce engine noise. 
Catalysts must therefore be placed upstream of the point where the 
exhaust and water mix--this ensures the effectiveness and durability of 
the catalyst. Because the catalyst must be small enough to fit in the 
exhaust manifold, potential emission reductions are not likely to 
exceed 90 percent, as is common in land-based applications. However, as 
discussed in Chapter 4 of the Draft RIA, demonstration programs have 
shown that emissions may be reduced by 70 to 80 percent for 
HC+NOX and 30 to 50 percent for CO over the proposed test 
cycle. Larger reductions, especially for CO, have been achieved at 
lower-speed operation.
    There have been concerns that aspects of the marine environment 
could result in unique durability problems for catalysts. The primary 
aspects that could affect catalyst durability are sustained operation 
at high load, saltwater effects on catalyst efficiency, and thermal 
shock from cold water coming into contact with a hot catalyst. Modern 
catalysts perform well at temperatures up to 1100[deg] C, which is much 
higher than would be seen in a marine exhaust manifold. These catalysts 
have also been shown to withstand the thermal shock of being immersed 
in water. More detail on catalyst durability is presented in the Draft 
RIA. In addition, use of catalysts in automotive, motorcycle, and 
handheld equipment has shown that catalysts can be packaged to 
withstand vibration in the exhaust manifold.
    Manufacturers already strive to design their exhaust systems to 
prevent water from reaching the exhaust ports. If too much water 
reaches the exhaust ports, significant durability problems would result 
from corrosion or hydraulic lock. As discussed in the Draft RIA, 
industry and government worked on a number of cooperative test programs 
in which several SD/I engines were equipped with catalysts and 
installed in vessels to prove out the technology. Early in the 
development work, a study was performed on an SD/I engine operating in 
a boat to see if water was entering the part of the manifold where 
catalysts would be installed. Although some water was collected in the 
exhaust manifold, it was found that this water came from water vapor 
that condensed out of the combustion products. This was easily 
corrected using a thermostat

[[Page 28128]]

to prevent overcooling from the water jacket.
    Four SD/I engines equipped with catalysts were operated in vessels 
for 480 hours on fresh water. This time period was intended to 
represent the full expected operating life of a typical SD/I engine. No 
significant deterioration was observed on any of these catalysts, nor 
was there any evidence of water reaching the catalysts. In addition, 
the catalysts were packaged such that the exhaust system met industry 
standards for maximum surface temperatures.
    Testing has been performed on one engine in a vessel on both fresh 
water and saltwater over a test protocol designed by industry to 
simulate the worst-case operation for water reversion. No evidence was 
found of water reaching the catalysts. After the testing, the engine 
had emission rates below the proposed HC+NOX standard. We 
later engaged in a test program to evaluate three additional engines 
with catalysts in vessels operating on saltwater for extended periods. 
Early in the program, two of the three manifolds experienced corrosion 
in the salt-water environment resulting in water leaks and damage to 
the catalyst. These manifolds were rebuilt with guidance from experts 
in the marine industry and additional hours have been accumulated on 
the boats. Although the accumulated hours are well below the 480 hours 
performed on fresh water, the operation completed has shown no visible 
evidence of water reversion or damage to the catalysts.
    One SD/I engine manufacturer began selling engines equipped with 
catalysts in Summer 2006. They have certified their engines to the 
California ARB standards, and are selling their catalyst-equipped 
engines nationwide. This manufacturer indicated that they have 
successfully completed durability testing, including extended in-use 
testing on saltwater. Other manufacturers have indicated that they will 
have catalyst-equipped SD/I engines for sale in California by the end 
of this year.
(4) Regulatory Alternatives
    In developing the proposed emission standards, we considered both 
what was achievable without catalysts and what could be achieved with 
larger, more efficient catalysts than those used in our test programs. 
Chapter 4 of the Draft RIA presents data on SD/I engines equipped with 
exhaust gas recirculation (EGR). HC+NOX emission levels 
below 10 g/kW-hr were achieved for each of the engines. CO emissions 
ranged from 25 to 185 g/kW-hr. We believe EGR would be a 
technologically feasible and cost-effective approach to reducing 
emissions from SD/I marine engines. However, we believe greater 
reductions could be achieved through the use of catalysts. We 
considered basing an interim standard on EGR, but were concerned that 
this would divert manufacturers' resources away from catalyst 
development and could have the effect of delaying emission reductions 
from this sector.
    Several of the marine engines with catalysts that were tested as 
part of the development of the proposed standards had HC+NOX 
emission rates in the 3-4 g/kW-hr range, even with consideration of 
expected in-use emissions deterioration associated with catalyst aging. 
However, we believe a standard of 5 g/kW-hr is still appropriate given 
the potential variability in in-use performance and in test data. The 
test programs described in Chapter 4 of the Draft RIA did not 
investigate larger catalysts for SD/I applications. The goal of the 
testing was to demonstrate catalysts that would work within the 
packaging constraints associated with water jacketing the exhaust and 
fitting the engines into engine compartments on boats. However, we did 
perform testing on engines equipped with both catalysts and EGR. These 
engines showed emission results in the 2-3 g/kW-hr range. We expect 
that these same reductions could be achieved more simply through the 
use of larger catalysts or catalysts with higher precious metal 
loading. Past experience indicates that most manufacturers will strive 
to achieve emission reductions well below the proposed standards to 
give them certainty that they will pass the standards in-use, 
especially as catalysts on SD/I engines are a new technology. 
Therefore, we do not believe it is necessary at this time to set a 
lower standard for these engines.
(5) Our Conclusions
    We believe the proposed 2009 exhaust emission standards for SD/I 
engines represent the greatest degree of emission reduction feasible in 
this time frame. Manufacturers could meet the proposed standards 
through the use of three-way catalysts packaged in the exhaust systems 
upstream of where the water and exhaust mix. One manufacture is already 
selling engines with this technology and by 2009 many other 
manufacturers will have experience in producing engines with catalysts 
for sale in California.
    As discussed in Section X, we do not believe the proposed standards 
would have negative effects on energy, noise, or safety and may lead to 
some positive effects.

IV. Outboard and Personal Watercraft Engines

A. Overview

    This section applies to spark-ignition outboard and personal 
watercraft (OB/PWC) marine engines and vessels. OB/PWC engines are 
currently required to meet the HC+NOX exhaust emissions and 
other related requirements under 40 CFR part 91. As a result of these 
standards, manufacturers have spent the last several years developing 
new technologies to replace traditional, carbureted, two-stroke engine 
designs. Many of these technologies are capable of emission levels well 
below the current standards. We are proposing new HC+NOX and 
CO exhaust emission standards for OB/PWC marine engines.
    For outboard and personal watercraft engines, the current emission 
standards regulate only HC+NOX emissions. As described in 
Section II, we are proposing in this notice to make the finding under 
Clean Air Act section 213(a)(3) that Marine SI engines cause or 
contribute to CO nonattainment in two or more areas of the United 
States.
    We believe manufacturers can use readily available technological 
approaches to design their engines to meet the proposed standards. In 
fact, as discussed in Chapter 4 of the Draft RIA, manufacturers are 
already producing several models of four-stroke engines and direction-
injection two-stroke engines that meet the proposed standards. The most 
important compliance step for the proposed standards will be to retire 
high-emitting designs that are still available and replace them with 
these cleaner engines. We are not proposing standards based on the use 
of catalytic converters in OB/PWC engines. While this may be an 
attractive technology in the future, we do not believe there has been 
sufficient development work on the application of catalysts to OB/PWC 
engines to use as a basis for standards at this time.
    Note that we are proposing to migrate the regulatory requirements 
for marine spark-ignition engines from 40 CFR part 91 to 40 CFR part 
1045. This gives us the opportunity to update the details of our 
certification and compliance program to be consistent with the 
comparable provisions that apply to other engine categories and 
describe regulatory requirements in plain language. Most of the change 
in regulatory text provides improved clarity without substantially 
changing procedures or compliance obligations. Where there is a change 
that warrants further attention, we describe the need for the change 
below.

[[Page 28129]]

B. Engines Covered by This Rule

(1) Definition of Outboard and Personal Watercraft Engines and Vessels
    The proposed standards are intended to apply to outboard marine 
engines and engines used to propel personal watercraft. We are 
proposing to change the existing definitions of outboard and personal 
watercraft to reflect this intent. The existing definitions of outboard 
engine and personal watercraft marine engine are presented below:
     Outboard engine is a Marine SI engine that, when properly 
mounted on a marine vessel in the position to operate, houses the 
engine and drive unit external to the hull of the marine vessel.
     Personal watercraft engine (PWC) is a Marine SI engine 
that does not meet the definition of outboard engine, inboard engine, 
or sterndrive engine, except that the Administrator in his or her 
discretion may classify a PWC as an inboard or sterndrive engine if it 
is comparable in technology and emissions to an inboard or sterndrive 
engine.
    With the proposed implementation of catalyst-based standards for 
sterndrive and inboard marine engines, we believe the above definitions 
could be problematic. Certain applications using SD/I engines and able 
to apply catalyst control would not be categorized as SD/I under the 
existing definitions in at least two cases. First, an airboat engine, 
which is often mounted well above the hull of the engine and used to 
drive an aircraft-like propeller could be misconstrued as an outboard 
engine. However, like traditional sterndrive and inboard engines, 
airboat engines are typically derived from automotive-based engines 
without substantial modifications for marine application. Airboat 
engines can use the same technologies that are available to sterndrive 
and inboard engines, so we believe they should be subject to the same 
standards. To address the concerns about classifying airboats, we are 
proposing to change the outboard definition to specify that the engine 
and drive unit be a single, self-contained unit that is designed to be 
lifted out of the water. This clarifies that air boats are not outboard 
engines; air boats do not have engines and drive units that are 
designed to be lifted out of the water. We are proposing the following 
definition:
     Outboard engine means an assembly of a spark-ignition 
engine and drive unit used to propel a marine vessel from a properly 
mounted position external to the hull of the marine vessel. An outboard 
drive unit is partially submerged during operation and can be tilted 
out of the water when not in use.
    Second, engines used on jet boats (with an open bay for passengers) 
have size, power, and usage characteristics that are very similar to 
sterndrive and inboard applications, but these engines may be the same 
as OB/PWC engines, rather than the marinized automotive engines 
traditionally used on sterndrive vessels. We believe classifying such 
engines as personal watercraft engines is inappropriate because it 
would subject the jet boats to less stringent emission standards than 
other boats with similar size and power characteristics. This different 
approach could lead to increased use of high-emitting engines in these 
vessels. Under the current regulations, engines powering jet boats 
could be treated as SD/I engines at the discretion of the Agency, 
because they are comparable in technology to conventional SD/I engines. 
We are proposing definitions that would explicitly exclude jet boats 
and their engines from being treated as personal watercraft engines or 
vessels. Instead, we are proposing to classify jet boat engines as SD/
I.
    The proposed definitions conform to the existing definition of 
personal watercraft established by the International Organization for 
Standardization (ISO 13590). This ISO standard excludes open-bay 
vessels and specifies a maximum vessel length of 4 meters. The ISO 
standard therefore excludes personal watercraft-like vessels 4 meters 
or greater and jet boats. Thus, engines powering such vessels would be 
classified as sterndrive/inboard engines. We believe this definition 
effectively serves to differentiate vessels in a way that groups 
propulsion engines into categories that are appropriate for meeting 
different emission standards. This approach is shown below with the 
corresponding proposed definition of personal watercraft engine. We are 
proposing one change to the ISO definition for domestic regulatory 
purposes; we propose to remove the word ``inboard'' to prevent 
confusion between PWC and inboard engines and state specifically that a 
vessel powered by an outboard marine engine is not a PWC. We are 
proposing the following definition:
     Personal watercraft means a vessel less than 4.0 meters 
(13 feet) in length that uses an installed internal combustion engine 
powering a water jet pump as its primary source of propulsion and is 
designed with no open load carrying area that would retain water. The 
vessel is designed to be operated by a person or persons positioned on, 
rather than within, the confines of the hull. A vessel using an 
outboard engine as its primary source of propulsion is not a personal 
watercraft.
     Personal watercraft engine means a spark-ignition engine 
used to propel a personal watercraft.
    Section III.C.2 describes special provisions that would allow 
manufacturers extra flexibility with emission credits if they want to 
continue using outboard or personal watercraft engines in jet boats. 
These engines would need to meet the standards for sterndrive/inboard 
engines, but we believe it is appropriate for them to make this 
demonstration using emission credits generated by other outboard and 
personal watercraft engines because these vessels are currently using 
these engine types. We request comment on this approach to defining 
personal watercraft, especially as it relates to vessels 4 meters or 
longer and jet boats.
(2) Exclusions and Exemptions
    We are proposing to maintain the existing exemptions for OB/PWC 
engines. These include the testing exemption, the manufacturer-owned 
exemption, the display exemption, and the national-security exemption. 
If the conditions for an exemption are met, the engine is not subject 
to the exhaust emission standards. These exemptions are described in 
more detail under Section VIII.
    The Clean Air Act provides for different treatment of engines used 
solely for competition. In the initial rulemaking to set standards for 
OB/PWC engines, we adopted the conventional definitions that excluded 
engines from the regulations if they had features that would be 
difficult to remove and that would make it unsafe, impractical, or 
unlikely to be used for noncompetitive purposes. We have taken the 
approach in other programs of more carefully differentiating 
competition and noncompetition models, and are proposing these kinds of 
changes in this rule. The proposed changes to the existing provisions 
relating to competition engines would apply equally to all types of 
Marine SI engines. See Section III and Sec.  1045.620 of the 
regulations for a full discussion of the proposed approach.
    We are proposing a new exemption to address individuals who 
manufacture recreational marine vessels for personal use (see Sec.  
1045.630). Under the proposed exemption, these vessels and their 
engines could be exempt from standards, subject to certain limitations. 
For example, an individual may produce one such vessel over a ten-year

[[Page 28130]]

period, the vessel may not be used for commercial purposes, and any 
exempt engines may not be sold for at least five years. The vessel must 
generally be built from unassembled components, rather than simply 
completing assembly of a vessel that is otherwise similar to one that 
will be certified to meet emission standards. This proposal addresses 
the concern that hobbyists who make their own vessels would otherwise 
be manufacturers subject to the full set of emission standards by 
introducing these vessels into commerce. We expect this exemption to 
involve a very small number of vessels.
    In the rulemaking for recreational vehicles, we chose not to apply 
standards to hobby products by exempting all reduced-scale models of 
vehicles that are not capable of transporting a person (67 FR 68242, 
November 8, 2002). We are proposing to extend that same provision to 
OB/PWC marine engines (see Sec.  1045.5).

C. Proposed Exhaust Emission Standards

    We are proposing more stringent exhaust emission standards for new 
OB/PWC marine engines. These proposed standards can be met through the 
expanded reliance on four-stroke engines and two-stroke direct-
injection engines. This section describes the proposed requirements for 
OB/PWC engines for controlling exhaust emissions. See Section V for a 
description of the proposed requirements related to evaporative 
emissions.
(1) Standards and Dates
    We are proposing new HC+NOX standards for OB/PWC engines 
starting in model year 2009 that would achieve more than a 60 percent 
reduction from the existing 2006 standards. We are also proposing new 
CO emission standards. These proposed standards would result in 
meaningful CO reductions from many engines and prevent CO from 
increasing from engines that already use technologies with lower CO 
emissions. The proposed emission standards are largely based on 
certification data from cleaner-burning Marine SI engines, such as 
four-stroke engines and two-stroke direct-injection engines. Section 
IV.F discusses the technological feasibility of these standards in more 
detail. Table IV-1 presents the proposed exhaust emission standards for 
OB/PWC. We are also proposing to apply not-to-exceed emission standards 
over a range of engine operating conditions, as described in Section 
IV.C.2. (See Sec.  1045.103.)

  Table IV-1--Proposed OB/PWC Exhaust Emission Standards [g/kW-hr] for
                             2009 Model Year
------------------------------------------------------------------------
             Pollutant                P\a\ <= 40 kW       P\a\ > 40 kW
------------------------------------------------------------------------
HC+NOX............................         28-0.3 x P                 16
CO................................        500-5.0 x P               300
------------------------------------------------------------------------
\a\ P = maximum engine power in kilowatts (kW).

    The proposed emission standards for HC+NOX are similar 
in stringency to the 2008 model year standards adopted in California, 
and we expect that the same technology anticipated to be used in 
California can be used to meet these proposed standards. However, we 
are proposing to simplify the form of the standards. The existing EPA 
2006 and California ARB 2008 requirements use a functional relationship 
to set the emission standard for each engine family depending on the 
power rating--the numerical value of the standard increases with 
decreasing power ratings, especially for the smallest engines. However, 
as described in Chapter 4 of the Draft RIA, certification data show 
that brake-specific emission rates (in g/kW-hr) are relatively constant 
for engines with maximum engine power above 40 kW. We are therefore 
proposing a single standard for engines with maximum engine power above 
40 kW. For smaller engines, the relationship between brake-specific 
emissions and maximum engine power is pronounced. We are proposing a 
simple linear function for the standards for these engines, as shown in 
Table IV-1. While this approach differs slightly from the California 
ARB standards, we believe it provides a good match for establishing a 
comparable level of stringency while simplifying the form of the 
regulatory standard.
    The proposed implementation date gives an additional year beyond 
the implementation date of the California standards of similar 
stringency. Manufacturers generally sell their lower-emission engines, 
which are already meeting the 2008 California standards, nationwide. 
However, the additional year would give manufacturers time to address 
any models that may not meet the upcoming California standards or are 
not generally sold in California. We request comment on additional 
regulatory flexibility that manufacturers may need to transition to the 
proposed standards. For instance, a modest phase-in of the standards 
may be useful to manufacturers to complete an orderly turnover of high-
emitting engines. This phase-in could take the form of giving an extra 
year for compliance with the proposed standards for a small percentage 
of engines (e.g., 10 percent of projected sales) or phasing-in the 
level of the standard (e.g., 20-25 g/kW-hr HC+NOX). Any 
comments on proposed transitional flexibility should give details that 
fully describe the recommended program.
    The proposed standards include the same general provisions that 
apply today. For example, engines must control crankcase emissions. The 
regulations also require compliance over the full range of adjustable 
parameters and prohibit the use of defeat devices. (See Sec.  
1045.115.)
(2) Not-to-Exceed Standards
    Section III.D.2 describes NTE standards for sterndrive and inboard 
engines. We are proposing to apply the same NTE testing provisions to 
OB/PWC engines, including the same NTE zone and subzones and ambient 
conditions (see Sec.  1045.515). However, data presented in Chapter 4 
of the Draft RIA suggest that different emission limits would be 
appropriate for OB/PWC engines. For instance, we are proposing higher 
limits at full power for SD/I engines equipped with catalysts because 
the engines must operate rich at this mode to protect catalysts and 
exhaust valves. Because we are not anticipating the use of catalysts on 
OB/PWC to meet the exhaust emission standards, we believe it is not 
necessary to adopt such high limits for OB/PWC engines.
    The Draft RIA describes the available emission data that allow us 
to specify appropriate modal caps for OB/PWC engines based on four-
stroke engine technology. The available data for direct-injection two-
stroke engines showed two different distinct patterns in modal emission 
rates. We are therefore proposing two alternative sets of NTE limits--
manufacturers could use either set of NTE limits for their OB/PWC 
engines. To offset the relaxed

[[Page 28131]]

limits for certain subzones, we are proposing more stringent limits for 
other subzones for these alternative approaches. Table IV-2 presents 
the proposed sets of NTE limits for the subzones described in Section 
III.D.2. We request comment on the proposed NTE limits for OB/PWC 
engines.

                          Table IV-2--Proposed NTE Limits by Subzone for OB/PWC Engines
----------------------------------------------------------------------------------------------------------------
           Approach                 Pollutant        Subzone 4       Subzone 3       Subzone 2       Subzone 1
----------------------------------------------------------------------------------------------------------------
Primary.......................  HC+NOX..........             1.6             1.2             1.2             1.2
                                CO..............             1.5             1.5             1.5             1.5
Alternative 1.................  HC+NOX..........             2.0             0.8             0.8             2.0
                                CO..............             1.0             1.0             1.5             3.0
Alternative 2.................  HC+NOX..........             3.0             1.0             1.0             1.0
                                CO..............             2.0             1.0             1.0             1.5
----------------------------------------------------------------------------------------------------------------

    Marine engine manufacturers indicated that they are concerned that 
the differences in engine designs, especially for direct-injection two-
stroke engines, may result in emission variation that would make it 
difficult to meet a fixed set of NTE limits for all engines. To address 
this variability, they have suggested two alternative approaches to 
setting NTE limits for marine engines. The first approach would be to 
base the NTE limits on the modal test results from the certification 
test rather than fixed values that would apply to all engines. NTE 
limits would then be linearly interpolated between the modes as a 
function of speed and load. For example, if the modal results were 2.0 
g/kW-hr at Mode 3 and 4.0 g/kW-hr at Mode 4, the interpolated value 
half way between these modal test points would be 3 g/kW-hr. A 
multiplier would then be applied to this interpolated value to create 
the NTE limit. This multiplier would be intended to account for testing 
and production variability. The multiplier would not likely need to be 
as large as the proposed general multipliers for the subzones presented 
above because it would be applied to a surface generated from each 
manufacturer's actual modal data. Because the NTE cap would be 
calculated from the individual test modes in the steady-state test, it 
may be necessary for the manufacturers to assign family emission limits 
for each of the test modes in the proposed NTE zone.
    The second conceptual approach would be to use a weighted average 
approach to the NTE limit rather than to have individual NTE limits for 
each subzone. Under this approach, an emission measurement would be 
made in each of the subzones plus idle. These measurements could be 
made at any operation point within each subzone. The measured emissions 
would then be combined using the weighting factors for the modal test. 
This weighted average emission level would be required to be below the 
standard (or family emission limit) times a multiplier (under this 
approach, only a single multiplier would be needed). The purpose of the 
multiplier would be to allow for some variability within each subzone. 
Because the weighted average emissions from the subzones would have the 
tendency of approaching the steady-state test value, this multiplier 
would not be expected to be much higher than 1.0. However, one drawback 
to this approach is that there is no specific cap for each mode and a 
weighted average approach may not be as effective in capping modal 
emissions as would be specific limits for each subzone. More detail on 
this concept is available in the docket.\78\
---------------------------------------------------------------------------

    \78\ ``Marine NTE Zones,'' Presentation to EPA by BRP on October 
26, 2006, Docket EPA-HQ-OAR-2004-0008-0508.
---------------------------------------------------------------------------

    We request comment on the two alternative NTE limit approaches 
described above. Specifically, commenters should provide detail on what 
advantages (and disadvantages) these alternatives may provide and what 
effect they may have on in-use emissions and the potential for 
improving the manufacturer in-use testing program. In addition, 
commenters should describe what emission limits or multipliers would be 
appropriate for the alternative approaches and provide test data 
supporting these conclusions.
(3) Emission Credit Programs
    Engine manufacturers may use emission credits to meet OB/PWC 
standards under part 91. See Section VII.C.5 for a description of 
general provisions related to averaging, banking, and trading programs.
    We propose to adopt an ABT program for the new HC+NOX 
emission standards that is similar to the existing program (see part 
1045, subpart H). Credits may be used interchangeably between outboard 
and personal watercraft engine families. Credits earned under the 
current program may also be used to comply with the new OB/PWC 
standards as described below.
    We are proposing an unlimited life for emission credits earned 
under the proposed new standards for OB/PWC engines. We consider these 
emission credits to be part of the overall program for complying with 
proposed standards. Given that we may consider further reductions 
beyond the proposed standards in the future, we believe it will be 
important to assess the ABT credit situation that exists at the time 
any future standards are considered. We would need to set such future 
emission standards based on the statutory direction that emission 
standards must represent the greatest degree of emission control 
achievable, considering cost, safety, lead time, and other factors. 
Emission credit balances will be part of the analysis for determining 
the appropriate level and timing of new standards. If we were to allow 
the use of existing emission credits for meeting future standards, we 
may, depending on the level of emission credit banks, need to adopt 
emission standards at more stringent levels or with an earlier start 
date than we would absent the continued or limited use of existing 
emission credits. Alternatively, we could adopt future standards 
without allowing the use existing credits. The proposal described in 
this notice describes a middle path in which we allow the use of 
existing credits to meet the proposed new standards, with provisions 
that limit the use of these credits based on a three-year credit life.
    We are requesting comment on one particular issue regarding credit 
life. As proposed, credits earned under the new exhaust ABT program 
would have an unlimited lifetime. This could result in a situation 
where credits generated by an engine sold in a model year are not used 
until many years later when the engines generating the credits have 
been scrapped and are no longer part of the fleet. EPA believes there 
may be value to limiting the use of credits to the period that the 
credit-generating engines

[[Page 28132]]

exist in the fleet. For this reason, EPA requests comment on limiting 
the lifetime of the credits generated under the proposed exhaust ABT 
program to five years or, alternatively, to the regulatory useful life 
of the engine.
    We are interested in using a common emission credit calculation 
methodology across our programs. Therefore, we are proposing to use the 
same emission credit equation for OB/PWC engines that is common in many 
of our other programs. This equation results in a simpler calculation 
than is currently used for OB/PWC engines. The primary difference is 
that the regulatory useful life would be used in the credit calculation 
rather than a discounted useful life function based on engine type and 
power rating. In addition, the emission credits would be reported in 
units of kilograms rather than grams. We anticipate that this change in 
the credit calculation would directionally increase the relative value 
of emission credits generated under the existing ABT program. However, 
due to the proposed limit on credit life and the proposed FEL cap for 
OB/PWC engines, we do not believe that this increase in relative value 
will significantly hamper the introduction of clean engine technology. 
We request comment on the new credit calculation and on whether credits 
generated under the existing OB/PWC standards should be adjusted to be 
more equivalent to credits generated under the proposed ABT program.
    We are proposing an averaging program for CO emissions. Under this 
program, manufacturers could generate credits with engine families that 
have FELs below the CO emission standard to be used for engine families 
in their product line in the same model year that are above the CO 
standard. However, we are proposing to disallow banking for CO 
emissions. We are concerned that a banking program could result in a 
large accumulation of credits based on a given company's mix of engine 
technologies. If banking were allowed, the proposed CO standard would 
need to be substantially more stringent to reflect the capability for 
industry-wide average CO emission levels. We generally allow trading 
only with banked credits, so we are also proposing to disallow trading 
of CO emission credits.
    As with previous emission control programs, we are also proposing 
not to allow manufacturers to earn credits for one pollutant for an 
emission family that is using credits to meet the standard for another 
pollutant. In other words, an engine family that does not meet the CO 
standard would not be able to earn HC+NOX emission credits, 
or vice versa. In addition, as with the current standards, we are 
proposing that engines sold in California would not be included in this 
ABT program because they are already subject to California 
requirements.
    Under the existing standards, no cap is set on FELs for certifying 
engine families. This was intended to allow manufacturers to sell old-
technology two-stroke engines by making up the emissions deficit with 
credits under the ABT program. For engines subject to the new emission 
standards, we are proposing FEL caps to prevent the sale of very high-
emitting engines. For HC+NOX, the proposed FEL cap is based 
on the existing 2006 standards. For CO, the proposed FEL cap is 150 g/
kW-hr above the proposed standard. We believe this will still allow a 
great deal of flexibility for manufacturers using credits, but will 
require manufacturers to stop producing engines that emit pollutants at 
essentially uncontrolled levels.
    Except as specified in Section III.C.2 for jet boats, we are 
proposing to specify that OB/PWC engines and SD/I engines are in 
separate averaging sets. This means that credits earned by OB/PWC 
engines may be used only to offset higher emissions from other OB/PWC 
engines, and credits earned by SD/I engines may be used only to offset 
higher emissions from other SD/I engines. We are allowing jet boats to 
use OB/PWC credits because there are currently small sales of these 
engines currently using OB/PWC engines. Most of the engine 
manufacturers building SD/I engines do not also build OB/PWC engines. 
The exception to this is the largest manufacturer in both categories. 
We are concerned that allowing averaging, banking, and trading between 
OB/PWC engines and SD/I engines would not provide the greatest 
achievable reductions, because the level of the standard we are 
proposing is premised on the technology used in OB/PWC engines, and is 
based on what is feasible for these engines. We did not set the OB/PWC 
level based on the reductions achievable between OB/PWC and SD/I, but 
instead based on what is achievable by OB/PWC itself. The proposed 
limitation on ABT credits is consistent with this approach to setting 
the level of the OB/PWC standards. We are also concerned that allowing 
trading between OB/PWC and SD/I could create a competitive disadvantage 
for the many small manufacturers of SD/I engines that do not also 
produce OB/PWC engines. In addition, we are proposing SD/I emission 
standards that would likely require the use of aftertreatment. We would 
not want to provide an incentive to use credits from the OB/PWC marine 
sector to avoid the use of aftertreatment technologies in SD/I engines.
    We request comment on the structure of the proposed ABT program, 
including the new provisions related to CO emissions. For any 
commenters suggesting that we include banking or trading for CO 
emissions, we solicit further comment on what the appropriate CO 
standard should be to account for the greater regulatory flexibility 
and therefore greater degree of control achievable using emissions 
credits. We also request comment on the use and level of the proposed 
FEL caps and on the approach to defining averaging sets.
(4) Durability Provisions
    We are proposing to keep the existing useful life periods from 40 
CFR part 91. The specified useful life for outboard engines is 10 years 
or 350 hours of operation, whichever comes first. The useful life for 
personal watercraft engines is 5 years or 350 hours of operation, 
whichever comes first. (See Sec.  1045.103.)
    We are proposing to update the specified emissions warranty periods 
for outboard and personal watercraft engines to align with our other 
emission control programs (see Sec.  1045.120). Most nonroad engines 
have emissions warranty periods that are half of the total useful life 
period. As a result, we are proposing a warranty period for outboard 
engines of five years or 175 hours of operation, whichever comes first. 
The proposed warranty period for personal watercraft engines is 30 
months or 175 hours, whichever comes first. This contrasts somewhat 
with the currently specified warranty period of 200 hours or two years 
(or three years for specified major emission control components). The 
proposed approach would slightly decrease the warranty period in terms 
of hours, but would somewhat increase the period in terms of calendar 
years (or months). We request comment on this revised approach to 
defining warranty periods.
    If the manufacturer offers a longer mechanical warranty for the 
engine or any of its components at no additional charge, we propose 
that the emission-related warranty for the respective engine or 
component must be extended by the same amount. The emission-related 
warranty includes components related to controlling exhaust, 
evaporative, and crankcase emissions from the engine. This approach to 
setting warranty requirements is consistent with provisions that apply 
in

[[Page 28133]]

most other programs for nonroad engines.
    We are proposing to keep the existing requirements related to 
demonstrating the durability of emission controls for purposes of 
certification (see Sec.  1045.235, Sec.  1045.240, and Sec.  1045.245). 
Manufacturers must run engines long enough to develop and justify full-
life deterioration factors. This allows manufacturers to generate a 
deterioration factor that helps ensure that the engines will continue 
to control emissions over a lifetime of operation. The new requirement 
to generate deterioration factors for CO emissions is the same as that 
for HC+NOX emissions. For the HC+NOX standard, we 
propose to specify that manufacturers use a single deterioration factor 
for the sum of HC and NOX emissions. However, if 
manufacturers get our approval to establish a deterioration factor on 
an engine that is tested with service accumulation representing less 
than the full useful life for any reason, we would require separate 
deterioration factors for HC and NOX emissions. The 
advantage of a combined deterioration factor is that it can account for 
an improvement in emission levels with aging. However, for engines that 
have service accumulation representing less than the full useful life, 
we believe it is not appropriate to extrapolate measured values 
indicating that emission levels for a particular pollutant will 
decrease.
    Under the current regulations, emission-related maintenance is not 
allowed during service accumulation to establish deterioration factors. 
The only maintenance that may be done must be (1) Regularly scheduled, 
(2) unrelated to emissions, and (3) technologically necessary. This 
typically includes changing engine oil, oil filter, fuel filter, and 
air filter. In addition, we are proposing to specify that manufacturers 
may not schedule critical emission-related maintenance during the 
useful life period (see Sec.  1045.125). This would prevent 
manufacturers from designing engines with emission controls that depend 
on scheduled maintenance that is not likely to occur with in-use 
engines. We request comment on all aspects of our provisions related to 
manufacturers' prescribed maintenance.

D. Changes to Existing OB/PWC Test Procedures

    We are proposing a number of minor changes to the test procedures 
for OB/PWC to make them more consistent with the test procedures for 
other nonroad spark-ignition engines. These test provisions would apply 
to SD/I marine engines as well.
(1) Duty Cycle
    A duty cycle is the set of modes (engine speed and load) over which 
an engine is operated during a test. For purposes of exhaust emission 
testing, we are proposing to keep the existing duty cycle specified for 
OB/PWC engines, with two adjustments (see Sec.  1045.505). First, we 
are proposing that manufacturers may choose to run the specified duty 
cycle as a ramped-modal cycle, as described in Section IX.B. Second, we 
are proposing to change the low-power test mode from a specified 25 
percent load condition to 25.3 percent load, which would complete the 
intended alignment with the E4 duty cycle adopted by the International 
Organization for Standardization.
    We request comment on the appropriateness of changing part 91 to 
include the correction to the duty cycle described above. We request 
comment regarding whether a change in the specification for the current 
standards may cause some existing test data to be considered invalid. 
For example, testing from an earlier model year may have involved 
measurements that were slightly below 25 percent load, but within the 
specified tolerance for testing. These measurements may be used for 
carryover engine families today, but increasing the load point in the 
regulation could cause some measurements to be outside the tolerance 
once it shifts to a nominal value of 25.3 percent.
(2) Maximum Test Speed
    The definition of maximum test speed, where speed is the angular 
velocity of an engine's crankshaft (usually expressed in revolutions 
per minute, or rpm), is an important aspect of the duty cycles for 
testing. Engine manufacturers currently declare the rated speeds for 
their engines and then used the rated speed as the maximum speed for 
testing. However, we have established an objective procedure for 
measuring this engine parameter to have a clearer reference point for 
an engine's maximum test speed. This is important to ensure that 
engines are tested at operating points that correspond with in-use 
operation. This also helps ensure that the NTE zone is appropriately 
matched to in-use operating conditions.
    We propose to define the maximum test speed for any engine to be 
the single point on an engine's maximum-power versus speed curve that 
lies farthest away from the zero-power, zero-speed point on a 
normalized maximum-power versus speed plot. In other words, consider 
straight lines drawn between the origin (speed = 0, load = 0) and each 
point on an engine's normalized maximum-power versus speed curve. 
Maximum test speed is defined at that point where the length of this 
line reaches its maximum value. This change would apply to testing of 
OB/PWC engines as well as SD/I engines. We request comment on the use 
and definition of maximum test speed.
(3) 40 CFR Part 1065
    We are proposing to specify that OB/PWC engines certified to the 
proposed exhaust emission standards use the test procedures in 40 CFR 
part 1065 instead of those in 40 CFR part 91.\79\ We are proposing that 
the new procedures would apply starting with the introduction of 
proposed exhaust standards, though we allow manufacturers to start 
using these new procedures earlier as an alternative procedure. The 
procedures in part 1065 include updated provisions to account for newer 
measurement technologies and improved calculation and corrections 
procedures. Part 1065 also specifies more detailed provisions related 
to alternate procedures, including a requirement to conduct testing 
representative of in-use operation. In many cases, we allow carryover 
of emission test data from one year to another. After the 
implementation of the proposed standards, we are proposing to allow 
carryover of any test data generated prior to 2009 under the test 
procedures in 40 CFR part 91.
---------------------------------------------------------------------------

    \79\ See our previous rulemakings related to 40 CFR part 1065 
for more information about the changes in test provisions (70 FR 
40420, July 13, 2005 and 67 FR 68242, November 8, 2002).
---------------------------------------------------------------------------

(4) Altitude
    EPA emission standards generally apply at a wide range of 
altitudes, as reflected in the range of barometric pressures in the 
specified test procedures. For marine spark-ignition engines, it is 
clear that the large majority of operation is at sea level or at inland 
lakes that are not at high altitude. We are therefore proposing a 
specific range of barometric pressures from 94.0 to 103.325 kPa, which 
corresponds to all altitudes up to about 2,000 feet (see Sec.  
1045.501). Manufacturers are expected to design emission control 
systems that continue to function effectively at lower barometric 
pressures (i.e., higher altitudes), but we would not require that 
engines meet emission standards when tested at altitudes more than 
2,000 feet above sea level.
(5) Engine Break-in
    Testing new engines requires a period of engine operation to 
stabilize emission

[[Page 28134]]

levels. The regulations specify two separate figures for break-in 
periods. First, for certification, we establish a limit on how much an 
engine may operate and still be considered a ``low-hour'' engine. The 
results of testing with the low-hour engine are compared with a 
deteriorated value after some degree of service accumulation to 
establish a deterioration factor. For Large SI engines, we require that 
low-hour test engines have no more than 300 hours of engine operation. 
However, given the shorter useful life for marine engines, this would 
not make for a meaningful process for establishing deterioration 
factors, even if there is a degree of commonality between the two types 
of engines. We are proposing for all marine spark-ignition engines that 
low-hour engines generally have no more than 30 hours of engine 
operation (see Sec.  1045.801). This allows some substantial time for 
break-in, stabilization, and running multiple tests, without 
approaching a significant fraction of the useful life. The current 
regulation in part 91 specifies that manufacturers perform the low-hour 
measurement after no more than 12 hours of engine operation (see Sec.  
91.408(a)(1)). The proposed approach, 30 hours of engine operation, is 
consistent with what we have done for recreational vehicles and would 
give manufacturers more time to complete a valid low-hour test.
    For production-line testing there is also a concern about how long 
an engine should operate to reach a stabilized emission level. We are 
proposing to keep the provision in part 91 that allows for a presumed 
stabilization period of 12 hours (see Sec.  90.117(a)). We believe 12 
hours is sufficient to stabilize the emissions from the engine.
    We request comment on these specified values for stabilizing new 
engines for emission measurements.

E. Additional Certification and Compliance Provisions

(1) Production-Line Testing
    We are proposing to continue to require that manufacturers 
routinely test engines at the point of production to ensure that 
production variability does not affect the engine family's compliance 
with emission standards. This is largely based on the existing test 
requirements, but includes a variety of changes. See Section VII.C.7 
for a detailed description of these requirements. We may also require 
manufacturers to perform production line testing under the selective 
enforcement auditing provisions described in Section VIII.E.
(2) In-Use Testing
    We are also proposing to continue the requirements related to the 
manufacturer-run in-use testing program. Under this program, 
manufacturers test field-aged engines to determine whether they 
continue to meet emission standards (see part 1045, subpart E). We are 
proposing to make a variety of changes and clarifications to the 
existing requirements, as described in the following sections.
(a) Adjustments Related to Engine Selection
    Both EPA and manufacturers have gained insights from implementing 
the current program. Manufacturers have expressed a concern that engine 
families are selected rather late in the model year, which makes it 
harder to prepare a test fleet for fulfilling testing obligations. On 
the other hand, we have seen that manufacturers certify some of their 
engine families well into the model year. By making selections early in 
the model year, we would generally be foregoing the opportunity to 
select engine families for which manufacturers don't apply for 
certification until after the selections occur.
    To address these competing interests, we are proposing an approach 
that allows for early selection of engine families, while preserving 
the potential to require testing for engines that are certified later 
in the model year. For applications we receive by December 31 of a 
given calendar year for the following model year, we would expect to 
select engine families for testing by the end of February of the 
following year. If we have not made a complete selection of engine 
families by the end of February, manufacturers would have the option of 
making their own selections for in-use testing. The proposed 
regulations include criteria to serve as guidance for manufacturers to 
make appropriate selections. For example, we would expect manufacturers 
to most strongly consider those engine families with the highest 
projected sales volume and the smallest compliance margins. 
Manufacturers may also take into account past experience with engine 
families if they have already passed an in-use testing regimen and have 
not undergone significant design changes since that time.
    We propose to treat engine families differently for in-use testing 
if we receive the application after December 31. This would apply, for 
example, if manufacturers send an application for a 2009 engine family 
in February 2009. In these cases, we are proposing that all these 
engine families are automatically subject to in-use testing, without 
regard to the 25 percent limitation that would otherwise dictate our 
selections. This may appear to increase the potential test burden, but 
the clear majority of applications for certification are completed 
before the end of the calendar year for the following model year. This 
proposed provision would eliminate the manufacturers' ability to game 
the testing system by delaying a family of potential concern until the 
next calendar year. We would expect to receive few new applications 
after the end of the calendar year. This would be consistent with the 
manufacturers' interest in early family selections, without 
jeopardizing EPA's interest in being able to select from a 
manufacturer's full product lineup.
    We request comment on the approach to selecting engine families for 
in-use testing.
(b) Crankcase Emissions
    Because the crankcase requirements are based on a design 
specification rather than emission measurements, the anticipated 
crankcase technologies are best evaluated simply by checking whether or 
not they continue to function as designed. As a result, we intend for 
an inspection of in-use engines to show whether these systems continue 
to function properly throughout the useful life, but are not proposing 
to require manufacturers to include crankcase measurements as part of 
the in-use testing program described in this section. This is 
consistent with the approach we have taken in other programs.
(c) In-Use Emission Credits
    Clean Air Act section 213 requires engines to comply with emission 
standards throughout their regulatory useful lives, and section 207 
requires a manufacturer to remedy in-use nonconformity when we 
determine that a substantial number of properly maintained and used 
engines fail to conform with the applicable emission standards (42 
U.S.C. 7541). As described in the original rulemaking, manufacturers 
could use a calculation of emission credits generated under the in-use 
testing program to avoid a recall determination if an engine family's 
in-use testing results exceeded emission standards (61 FR 52095, 
October 4, 1996).
    We are proposing a more general approach to addressing potential 
noncompliance under the in-use testing program than is specified in 40 
CFR part 91. The proposed regulations do not specify how manufacturers 
would

[[Page 28135]]

generate emission credits to offset a nonconforming engine family. The 
proposed approach is preferred for two primary reasons. First, 
manufacturers will be able to use emission data generated from field 
testing to characterize an engine family's average emission level. This 
becomes necessarily more subjective, but allows us to consider a wider 
range of information in evaluating the degree to which manufacturers 
are complying with emission standards across their product line. 
Second, this approach makes clearer the role of the emission credits in 
our consideration to recall failing engines. We plan to consider, among 
other information, average emission levels from multiple engine 
families in deciding whether to recall engines from a failing engine 
family. We therefore believe it is not appropriate to have a detailed 
emission credit program defining precisely how and when to calculate, 
generate, and use credits that do not necessarily have value elsewhere.
    Not specifying how manufacturers generate emission credits under 
the in-use testing program gives us the ability to consider any 
appropriate test data in deciding what action to take. In generating 
this kind of information, some general guidelines would apply. For 
example, we would expect manufacturers to share test data from all 
engines and all engine families tested under the in-use testing 
program, including nonstandard tests that might be used to screen 
engines for later measurement. This allows us to understand the 
manufacturers' overall level of performance in controlling emissions to 
meet emission standards. Average emission levels should be calculated 
over a running three-year period to include a broad range of testing 
without skewing the results based on old designs. Emission values from 
engines certified to different tiers of emission standards or tested 
using different measurement procedures should not be combined to 
calculate a single average emission level. Average emission levels 
should be calculated according to the following equation, rounding the 
results to 0.1 g/kW-hr:

Average EL = [Sigma]i[(STD-CL)i x 
(UL)i x (Sales)i x Poweri x 
LFi] / [Sigma]i [(UL)i x 
(Sales)i x Poweri x LFi]

Where:

Average EL = Average emission level in g/kW-hr.
Salesi = The number of eligible sales, tracked to the 
point of first retail sale in the U.S., for the given engine family 
during the model year.
(STD-CL)i = The difference between the emission standard 
(or Family Emission Limit) and the average emission level for an in-
use testing family in g/kW-hr.
ULi = Useful life in hours.
Poweri = The sales-weighted average maximum engine power 
for an engine family in kW.
LFi = Load factor or fraction of maximum engine power 
utilized in use; use 0.50 for engine families used only in constant-
speed applications and 0.32 for all other engine families.
    We have adopted this same approach for the in-use testing program 
that applies for Large SI engines in 40 CFR part 1048.
(3) Optional Procedures for Field Testing
    Outboard engines are inherently portable, so it may be easier to 
test them in the laboratory than in the field. However, there is a 
strong advantage to using portable measurement equipment to test 
personal watercraft and SD/I engines while the engine remains installed 
to avoid the effort of taking the engine out and setting it up in a 
laboratory. Field testing would also provide a much better means of 
measuring emissions to establish compliance with the NTE standards, 
because it is intended to ensure control of emissions during normal in-
use operation that may not occur during laboratory testing over the 
specified duty cycle. We propose to apply the field testing provisions 
described below as an option for all OB/PWC and SD/I engines. We 
request comment on any ways the field testing procedures should be 
modified to address the unique operating characteristics of OB/PWC or 
SD/I engines.
    The regulations at 40 CFR part 1065, subpart J, specify how to 
measure emissions using portable measurement equipment. To test engines 
while they remain installed, analyzers are connected to the engine's 
exhaust to detect emission concentrations during normal operation. 
Exhaust volumetric flow rate and continuous power output are also 
needed to convert the analyzer responses to units of g/kW-hr for 
comparing to emission standards. These values can be calculated from 
measurements of the engine intake flow rate, the exhaust air-fuel ratio 
and the engine speed, and from torque information.
    Available small analyzers and other equipment may be adapted for 
measuring emissions from field equipment. A portable flame ionization 
detector can measure total hydrocarbon concentrations. A portable 
analyzer based on zirconia technology can measure NOX 
emissions. A nondispersive infrared (NDIR) unit can measure CO. We are 
proposing to require manufacturers to specify how they would allow for 
drawing emission samples from in-use engines for testing installed 
engines. For example, emission samples can be drawn from the exhaust 
flow directly upstream of the point at which water is mixed into the 
exhaust flow. This should minimize collection of water in the extracted 
sample, though a water separator may be needed to maintain a 
sufficiently dry sample. Mass flow rates also factor into the torque 
calculation; this may be measured either in the intake or exhaust 
manifold.
    Calculating brake-specific emissions depends on determining 
instantaneous engine speed and torque levels. We propose to require 
that manufacturers must therefore design their engines to be able to 
continuously monitor engine speed and torque. We have already adopted 
this requirement for other mobile source programs where electronic 
engine control is used. Monitoring speed values is straightforward. For 
torque, the onboard computer needs to convert measured engine 
parameters into useful units. Manufacturers generally will need to 
monitor a surrogate value such as intake manifold pressure or throttle 
position (or both), then rely on a look-up table programmed into the 
onboard computer to convert these torque indicators into Newton-meters. 
Manufacturers may also want to program the look-up tables for torque 
conversion into a remote scan tool. Part 1065 specifies the performance 
requirements for accuracy, repeatability, and noise related to speed 
and torque measurements. These tolerances are taken into account in the 
selection of the proposed NTE standards.
(4) Other Changes for In-use Testing
    A question has been raised regarding the extent of liability if an 
engine family is found to be noncompliant during in-use testing. 
Because it can take up to two years to complete the in-use testing 
regimen for an engine family, we want to clarify the status of engines 
produced under that engine family's certificate, and under the 
certificates of earlier and later engine families that were effectively 
of the same design. For example, manufacturers in many cases use 
carryover data to continue certifying new engine families for a 
subsequent model year; this avoids the need to produce new test data 
for engines whose design does not change from year to year. For these 
cases, absent any contrary information from the manufacturer, we will 
maintain the discretion to include other applicable

[[Page 28136]]

engine families in the scope of any eventual recall, as allowed by the 
Act.
    There are a variety of smaller changes to the in-use testing 
provisions as a result of updating the regulatory language to reflect 
the language changes that we adopted for similar testing with Large SI 
engines. First, we are proposing to remove the requirement to select 
engines that have had service accumulation representing less than 75 
percent of the useful life. This will allow manufacturers the 
flexibility to test somewhat older engines if they want to. Second, we 
are proposing to slightly adjust the description of the timing of the 
test program, specifying that the manufacturer must submit a test plan 
within 12 months of EPA selecting the family for testing, with a 
requirement to complete all testing within 24 months. This contrasts 
with the current requirement to complete testing within 12 months after 
the start of testing, which in turn must occur within 12 months of 
family selection. We believe the modified approach allows additional 
flexibility without delaying the conclusion of testing. Third, we are 
proposing to require that manufacturers explain why they excluded any 
particular engines from testing. Finally, we are proposing to require 
manufacturers to report any noncompliance within 15 days after 
completion of testing for a family, rather than 15 days after an 
individual engine fails. This has the advantage for manufacturers and 
the Agency of a more unified reporting after testing is complete, 
rather than piecemeal reporting before conclusions can be drawn.
(5) Use of Engines Already Certified to Other Programs
    In some cases, manufacturers may want to use engines already 
certified under our other programs. Engines certified to the emission 
standards for highway applications in part 86 or Large SI applications 
in part 1048 are meeting more stringent standards. We are therefore 
proposing to allow the pre-existing certification to be valid for 
engines used in marine applications, on the condition that the engine 
is not changed from its certified configuration in any way (see Sec.  
1045.605). For outboard and personal watercraft engines, we are also 
proposing to allow this for engines certified to the Phase 3 emission 
standards for Small SI engines. Manufacturers would need to demonstrate 
that fewer than five percent of the total sales of the engine model are 
for marine applications. There are also a few minor notification and 
labeling requirements to allow for EPA oversight of this provision.
(6) Import-Specific Information at Certification
    We are proposing to require additional information to improve our 
ability to oversee compliance related to imported engines (see Sec.  
1045.205). In the application for certification, we are proposing to 
require the following additional information: (1) The port or ports at 
which the manufacturer will import the engines, (2) the names and 
addresses of the agents the manufacturer has authorized to import the 
engines, and (3) the location of the test facilities in the United 
States where the manufacturer will test the engines if we select them 
for testing under a selective enforcement audit.

F. Other Adjustments to Regulatory Provisions

    We are proposing to migrate the regulatory requirements for marine 
spark-ignition engines from 40 CFR part 91 to 40 CFR part 1045. This 
gives us the opportunity to update the details of our certification and 
compliance program to be consistent with the comparable provisions that 
apply to other engine categories. The following paragraphs highlight 
some of the changes in the new language that may involve noteworthy 
changes from the existing regulations. All these provisions apply 
equally to SD/I engines, except that they are not subject to the 
current requirements in 40 CFR part 91.
    We are proposing some adjustments to the criteria for defining 
engine families (see Sec.  1045.230). The fundamental principle behind 
engine families is to group together engines that will have similar 
emission characteristics over the useful life. We are proposing that 
engines within an engine family must have the same approximate bore 
diameter and all use the same method of air aspiration (for example, 
naturally aspirated vs. turbocharged). Under the current regulation, 
manufacturers may consider bore and stroke dimensions and aspiration 
method if they want to subdivide engine families beyond what would be 
required under the primary criteria specified in Sec.  91.115. We 
believe engines with substantially different bore diameters will have 
combustion and operating characteristics that must be taken into 
account with unique engineering. Similarly, adding a turbocharger or 
supercharger to an engine changes the engine's combustion and emission 
control in important ways. Finally, we are proposing that all the 
engines in an engine family use the same type of fuel. This may have 
been a simple oversight in the current regulations, since all OB/PWC 
engines operate on gasoline. However, if a manufacturer would produce 
an engine model that runs on natural gas or another alternative fuel, 
that engine model should be in its own engine family.
    The proposed regulatory language related to engine labels remains 
largely unchanged (see Sec.  1045.135). However, we are including a 
provision to allow manufacturers to print labels that have a different 
company's trademark. Some manufacturers in other programs have 
requested this flexibility for marketing purposes.
    The proposed warranty provisions are described above. We are 
proposing to add an administrative requirement to describe the 
provisions of the emission-related warranty in the owners manual (see 
Sec.  1045.120). We expect that many manufacturers already do this, but 
believe it is appropriate to require this as a routine practice.
    Certification procedures depend on establishing deterioration 
factors to predict the degradation in emission controls that occurs 
over the course of an engine's useful life. This typically involves 
service accumulation in the laboratory to simulate in-use operation. 
Since manufacturers do in-use testing to further characterize this 
deterioration rate, we are proposing to specify that deterioration 
factors for certification must take into account any available data 
from in-use testing with similar engines. This provision applies in 
most of our emission control programs that involve in-use testing. To 
the extent that this information is available, it should be factored 
into the certification process. For example, if in-use testing shows 
that emission deterioration is substantially higher than that 
characterized by the deterioration factor, we would expect the 
manufacturer to factor the in-use data into a new deterioration factor, 
or to revise durability testing procedures to better represent the 
observed in-use degradation.
    Maximum engine power for an engine family is an important 
parameter. For engines below 40 kW, the maximum engine power determines 
the applicable standard. For bigger engines, emission credits are 
calculated based on total power output. As a result, we are proposing 
to specify that manufacturers determine their engines' maximum engine 
power as the point of maximum engine power on the engine map the 
manufacturers establish with their test engines (see Section VII.C.6 
and

[[Page 28137]]

Sec.  1045.140). This value would be based on the measured maximum 
engine power, without correction to some standard ambient conditions.
    The proposed requirements related to the application for 
certification would involve some new information, most of which is 
described above, such as installation instructions and a description of 
how engines comply with not-to-exceed standards (see Sec.  1045.205). 
In addition, we are proposing to require that manufacturers submit 
projected sales volumes for each family, rather than requiring that 
manufacturers keep these records and make them available upon request. 
Manufacturers already do this routinely and it is helpful to have ready 
access to this information to maintain compliance oversight of the 
program for Marine SI engines for such things as emission credit 
calculations. We are also proposing that each manufacturer identify an 
agent for service in the United States. For companies based outside the 
United States, this ensures that we will be able to maintain contact 
regarding any official communication that may be required. We have 
adopted these same requirements for other nonroad programs.
    We are proposing to require that manufacturers use good engineering 
judgment in all aspects of their effort to comply with regulatory 
requirements. The regulations at Sec.  1068.5 describe how we would 
apply this provision and what we would require of manufacturers where 
we disagree with a manufacturer's judgment.
    We are also proposing new defect-reporting requirements. These are 
requirements are described in Section VIII.
    It is common practice for Marine SI engines for one company to 
produce the base engine for a second company to modify for the final 
application. Since our regulations prohibit the sale of uncertified 
engines, we are proposing provisions to clarify the status of these 
engines and defining a path by which these engines can be handled 
without violating the regulations. See Section XI for more information.
    We request comment on all these changes to the regulations. Where 
there is an objection to any of the proposed provisions, we request 
comment on alternative provisions that would best address the concern 
on which the proposed provisions are based. Also, aside from the items 
described in this section, there are many minor adjustments in the 
regulatory text. While most of these changes are intended to improve 
the clarity of the regulations without imposing new requirements, we 
request comment on any of these changes that may be inappropriate. We 
also request comment on any additional changes that may be helpful in 
making the regulations clear or addressing the administration or 
implementation of the regulatory requirements.

G. Small-Business Provisions

    The OB/PWC market has traditionally been made up of large 
businesses. In addition, we anticipate that the OB/PWC standards will 
be met through the expanded use of existing cleaner engine 
technologies. Small businesses certifying to standards today are 
already using technologies that could be used to meet the proposed 
standards. As a result, we are proposing only three small business 
regulatory relief provisions for small business manufacturers of OB/PWC 
engines. We are proposing to allow small business OB/PWC engine 
manufacturers to be exempt from PLT testing and to use assigned 
deterioration factors for certification. (EPA will provide guidance to 
engine manufacturers on the assigned deterioration factors prior to 
implementation of the new OB/PWC standards.) We are also proposing to 
extend the economic hardship relief for small businesses described in 
Section VIII.C.9 to small-business OB/PWC engine manufacturers (see 
Sec.  1068.250). We are proposing small business eligibility criteria 
for OB/PWC engine manufacturers based on a production cut-off of 5,000 
OB/PWC engines per year. We would also allow OB/PWC engine 
manufacturers that exceed the production cut-off level noted above but 
have fewer than 1,000 employees to request treatment as a small 
business.
    In addition to the flexibilities noted above, all OB/PWC engine 
manufacturers, regardless of size, would be able to apply for the 
unusual circumstances hardship described in Section VIII.C.8 (see Sec.  
1068.245). Finally, all OB/PWC vessel manufacturers, regardless of 
size, that rely on other companies to provide certified engines or fuel 
system components for their product would be able to apply for the 
hardship provisions described in Section VIII.C.10 (see Sec.  
1068.255).

H. Technological Feasibility

(1) Level of Standards
    Over the past several years, manufacturers have demonstrated their 
ability to achieve significant HC+NOX emission reductions 
from outboard and personal watercraft engines. This has largely been 
accomplished through the introduction of two-stroke direct injection 
engines and conversion to four-stroke engines. Current certification 
data for these types of engines show that these technologies may be 
used to achieve emission levels significantly below the existing 
exhaust emission standards. In fact, California has adopted standards 
requiring a 65 percent reduction beyond the current federal standards 
beginning in 2008.
    Our own analysis of recent certification data show that most four-
stroke outboard engines and many two-stroke direct injection outboard 
engines can meet the proposed HC+NOX standard. Similarly, 
although PWC engines tend to have higher HC+NOX emissions, 
presumably due to their higher power densities, many of these engines 
can also meet the proposed HC+NOX standard. Although there 
is currently no CO standard for OB/PWC engines, OB/PWC manufacturers 
are required to report CO emissions from their engines (see Sec.  
91.107(d)(9)). These emissions are based on test data from new engines 
and do not consider deterioration or compliance margins. Based on this 
data, all of the two-stroke direct injection engines show emissions 
well below the proposed standards. In addition, the majority of four-
stroke engines would meet the proposed CO standards as well.
    We therefore believe the proposed HC+NOX and CO emission 
standards can be achieved by phasing out conventional carbureted two-
stroke engines and replacing them with four-stroke engines or two-
stroke direct injection engines. This has been the market-driven trend 
over the last five years. Chapter 4 of the Draft RIA presents charts 
that compare certification data to the proposed standards.
(2) Implementation Dates
    We are proposing to implement the new emission standards beginning 
with the 2009 model year. This gives an additional year beyond the 
implementation date of the California standards of similar stringency. 
This additional year may be necessary for manufacturers that don't sell 
engine models in California or that sell less than their full product 
lineup into the California market. We believe the same technology used 
to meet the 2008 standards in California could be used nationwide with 
the additional year allowed for any engine models not sold in 
California. Low-emission engines sold in California are generally sold 
nationwide as part of manufacturer compliance strategies for the 
Federal 2006 standards. Manufacturers have

[[Page 28138]]

indicated that they are calibrating their four-stroke and direct-
injection two-stroke engines to meet the California requirements. To 
meet the proposed standards, manufacturers' efforts would primarily 
center on phasing out their higher-emission carbureted two-stroke 
engines and producing more of their lower emission engines.
(3) Technological Approaches
    Conventional two-stroke engines add a fuel-oil mixture to the 
intake air with a carburetor, and use the crankcase to force this mixed 
charge air into the combustion chamber. In the two-stroke design, the 
exhaust gases must be purged from the cylinder while the fresh charge 
enters the cylinder. With traditional two-stroke designs, the fresh 
charge, with unburned fuel and oil, would push the exhaust gases out of 
the combustion chamber as the combustion event concludes. As a result, 
25 percent or more of the fresh fuel-oil could pass through the engine 
unburned. This is known as scavenging losses. Manufacturers have phased 
out sales of the majority of their traditional two-stroke engines to 
meet the federal 2006 OB/PWC exhaust emission standards. However, many 
of these engines still remain in the product mix as a result of 
emission credits.
    One approach to minimizing scavenging losses in a two-stroke engine 
is through the use of direct fuel injection into the combustion 
chamber. The primary advantage of direct injection for a two-stroke is 
that the exhaust gases can be scavenged with fresh air and fuel can be 
injected into the combustion chamber after the exhaust port closes. As 
a result, hydrocarbon emissions, fuel economy, and oil consumption are 
greatly improved. Some users prefer two-stroke direct injection engines 
over four-stroke engines due to the higher power-to-weight ratio. Most 
of the two-stroke direct injection engines currently certified to the 
current OB/PWC emission standards have HC+NOX emissions 
levels somewhat higher than certified four-stroke engines. However, 
these engines also typically have lower CO emissions due to the nature 
of a heterogeneous charge. By injecting the fuel directly into a charge 
of air in the combustion chamber, localized areas of lean air/fuel 
mixtures are created where CO is efficiently oxidized.
    OB/PWC manufacturers are also achieving lower emissions through the 
use of four-stroke engine designs. Because the combustion cycle takes 
place over two revolutions of the crankshaft, the fresh fuel-air charge 
can enter the combustion chamber after the exhaust valve is closed. 
This prevents scavenging losses. Manufacturers currently offer four-
stroke marine engines with maximum engine power ranging from 1.5 to 224 
kW. These engines are available with carburetion, throttle-body fuel 
injection, or multi-point fuel injection. Based on the certification 
data, whether the engine is carbureted or fuel-injected does not have a 
significant effect on combined HC+NOX emissions. For PWC 
engines, the HC+NOX levels are somewhat higher, primarily 
due to their higher power-to-weight ratio. CO emissions from PWC 
engines are similar to those for four-stroke outboard engines.
    One manufacturer has certified two PWC engine models with oxidation 
catalysts. One engine model uses the oxidation catalyst in conjunction 
with a carburetor while the other uses throttle-body fuel injection. In 
this application, the exhaust system is shaped in such a way to protect 
the catalyst from water. The exhaust system is relatively large 
compared to the size of the engine. We are not aware of any efforts to 
develop a three-way catalyst system for PWC engines. We are also not 
aware of any development efforts to package a catalyst into the exhaust 
system of an outboard marine engine. In current designs, water and 
exhaust are mixed in the exhaust system to help cool the exhaust and 
tune the engine. Water can work its way up through the exhaust system 
because the lower end is under water and varying pressures in the 
exhaust stream can draw water against the prevailing gas flow. As 
discussed in Chapter 4 of the Draft RIA, saltwater can be detrimental 
to catalyst performance and durability. In addition, outboard engines 
are designed with lower units that are designed to be as thin as 
possible to improve the ability to turn the engine on the back of the 
boat and to reduce drag on the lowest part of the unit. This raises 
concerns about the placement and packaging of catalysts in the exhaust 
stream. Certainly, the success of packaging catalysts in sterndrive and 
inboard boats in recent development efforts (see Section III) suggests 
that catalysts may be feasible for outboards with additional effort. 
However, this has not yet been demonstrated and significant development 
efforts would be necessary. We request comment on the feasibility of 
using catalysts on OB and PWC engines.
(4) Regulatory Alternatives
    We considered a level of 10 g/kW-hr HC+NOX for OB/PWC 
engines above 40 kW with an equivalent percent reduction below the 
proposed standards for engines below 40 kW. This second tier of 
standards could apply in the 2012 or later time frame. Such a standard 
would be consistent with currently certified emission levels from a 
significant number of four-stroke outboard engines. We have three 
concerns with adopting this second tier of OB/PWC standards. First, 
while some four-stroke engines may be able to meet a 10 g/kW-hr 
standard with improved calibrations, it is not clear that all engines 
could meet this standard without applying catalyst technology. As 
described in Section IV.H.3, we believe it is not appropriate to base 
standards in this rule on the use of catalysts for OB/PWC engines. 
Second, certification data for personal watercraft engines show 
somewhat higher exhaust emission levels, so setting the standard at 10 
g/kW-hr would likely require catalysts for many models. Third, it is 
not clear that two-stroke engines would be able to meet the more 
stringent standard, even with direct injection and catalysts. These 
engines operate with lean air-fuel ratios, so reducing NOX 
emissions with any kind of aftertreatment is especially challenging.
    Therefore, unlike the proposed standards for sterndrive and inboard 
engines, we are not adopting OB/PWC standards that will require the use 
of catalysts. Catalyst technology would be necessary for significant 
additional control of HC+NOX and CO emissions. While there 
is good potential for eventual application of catalyst technology to 
outboard and personal watercraft engines, we believe the technology is 
not adequately demonstrated at this point. Much laboratory and in-water 
work is needed.
(5) Our Conclusions
    We believe the proposed emission standards can be achieved by 
phasing out conventional carbureted two-stroke engines in favor of 
four-stroke engines or two-stroke direct injection engines. The four-
stroke engines or two-stroke direct injection engines are already 
widely available from marine engine manufacturers. One or both of these 
technologies are currently in place for the whole range of outboard and 
personal watercraft engines.
    The proposed exhaust emission standards represent the greatest 
degree of emission control achievable in the contemplated time frame. 
While manufacturers can meet the proposed standards with their full 
product line in 2009, requiring full compliance with a nationwide 
program earlier, such as in the same year that California introduces 
new emission standards, would pose an unreasonable requirement. 
Allowing

[[Page 28139]]

one year beyond California's requirements is necessary to allow 
manufacturers to certify their full product line to the new standards, 
not only those products they will make available in California. Also, 
as described above, we believe the catalyst technology that would be 
required to meet emission standards substantially more stringent than 
we are proposing has not been adequately demonstrated for outboard or 
personal watercraft engines. As such, we believe the proposed standards 
for HC+NOX and CO emissions are the most stringent possible 
in this rulemaking. More time to gain experience with catalysts on 
sterndrive and inboard engines and a substantial engineering effort to 
apply that learning to outboard and personal watercraft engines may 
allow us to pursue more stringent standards in a future rulemaking.
    As discussed in Section X, we do not believe the proposed standards 
would have negative effects on energy, noise, or safety and may lead to 
some positive effects.

V. Small SI Engines

A. Overview

    This section applies to new nonroad spark-ignition engines with 
rated power at or below 19 kW (``Small SI engines''). These engines are 
most often used in lawn and garden applications, typically by 
individual consumers; they are many times also used by commercial 
operators and they provide power for a wide range of other home, 
industrial, farm, and construction applications. The engines are 
typically air-cooled single-cylinder models, though Class II engines 
(with displacement over 225 cc) may have two or three cylinders, and 
premium models with higher power may be water-cooled.
    We have already adopted two phases of exhaust standards for Small 
SI engines. The first phase of standards for nonhandheld engines 
generally led manufacturers to convert any two-stroke engines to four-
stroke engines. These standards applied only to engines at the time of 
sale. The second phase of standards for nonhandheld engines generally 
led manufacturers to apply emission control technologies such as in-
cylinder controls and improved carburetion, with the additional 
requirement that manufacturers needed to meet emission standards over a 
useful life period.
    As described in Section I, this proposal is the result of a 
Congressional mandate that springs from the new California ARB 
standards. In 2003, the California ARB adopted more stringent standards 
for nonhandheld engines. These standards target emission reductions of 
approximately 35 percent below EPA's Phase 2 standards and are based on 
the expectation that manufacturers will use relatively low-efficiency 
three-way catalysts to control HC+NOX emissions. California 
ARB did not change the applicable CO emission standard.\80\
---------------------------------------------------------------------------

    \80\ California ARB also adopted new fuel evaporative emission 
standards for equipment using handheld and nonhandheld engines. 
These included tank permeation standards for both types of equipment 
and hose permeation, running loss, and diurnal emission standards 
for nonhandheld equipment. See Section VI for additional information 
related to evaporative emissions.
---------------------------------------------------------------------------

    We are proposing to place these new regulations for Small SI 
engines in 40 CFR part 1054 rather than changing the current 
regulations in 40 CFR part 90. This gives us the opportunity for 
proposing updates to the details of our certification and compliance 
program that are consistent with the comparable provisions that apply 
to other engine categories and describe regulatory requirements in 
plain language. Most of the change in regulatory text provides improved 
clarity without changing procedures or compliance obligations. Where 
there is a change that warrants further attention, we describe the need 
for the change below.

B. Engines Covered by This Rule

    This action includes proposed exhaust emission standards for new 
nonroad engines with rated power at or below 19 kW that are sold in the 
United States. The exhaust standards are for nonhandheld engines 
(Classes I and II). As described in Section I, handheld Small SI 
engines (Classes III, IV, and V) are also subject to standards, but we 
are not proposing changes to the level of exhaust emission standards 
for these engines. As described in Section VI, we are also proposing 
standards for controlling evaporative emissions from Small SI engines, 
including both handheld and nonhandheld engines. Certain of the 
provisions discussed in this Section V apply to both handheld and 
nonhandheld engines, as noted. Reference to both handheld and 
nonhandheld engines also includes marine auxiliary engines subject to 
the Small SI standards for that size engine.
(1) Engines Covered by Other Programs
    The Small SI standards do not apply to recreational vehicles 
covered by EPA emission standards in 40 CFR part 1051. The regulations 
in part 1051 apply to off-highway motorcycles, snowmobiles, all-terrain 
vehicles, and high-speed offroad utility vehicles. However, if an 
amphibious vehicle with an engine at or below 19 kW is not subject to 
standards under part 1051, its engine would need to meet the Small SI 
standards. We also do not consider vehicles such as go karts or golf 
carts to be recreational vehicles because they are not intended for 
high-speed operation over rough terrain; these engines are also subject 
to Small SI standards. The Small SI standards do not apply to engines 
used in scooters or other vehicles that qualify as motor vehicles.
    Consistent with the current regulation under 40 CFR part 90, Small 
SI standards apply to spark-ignition engines used as generators or for 
other auxiliary power on marine vessels, but not to marine propulsion 
engines. As described below, we are proposing more stringent exhaust 
emission standards that would apply uniquely to marine generator 
engines.
    Engines with rated power above 19 kW are subject to emission 
standards under 40 CFR part 1048. However, we adopted a special 
provision under part 1048 allowing engines with total displacement at 
or below 1000 cc and with rated power at or below 30 kW to meet the 
applicable Small SI standards instead of the standards in part 1048. 
For any engines that are certified using this provision, any emission 
standards that we adopt for Class II engines and equipment in this 
rulemaking will also apply at the same time. Since these engines are 
not required to meet the Small SI standards we have not included them 
in the analyses associated with this proposal.
(2) Maximum Engine Power and Engine Displacement
    Under the current regulations, rated power and power rating are not 
defined terms, which leaves manufacturers to determine their values. We 
are proposing to establish an objective approach to establishing 
``maximum engine power'' under the regulations (see Section VII.C.6 and 
Sec.  1054.140). This value has regulatory significance for Small SI 
engines only to establish whether or not engines are instead subject to 
Large SI standards. Determining maximum engine power is therefore 
relevant only for those engines that are approaching the line 
separating these two engine categories. We are proposing to require 
that manufacturers determine and report maximum engine power if their 
emission-data engine has a maximum modal power at or above 15 kW.
    Similarly, the regulations depend on engine displacement to 
differentiate engines for the applicability of different standards. The 
regulations currently provide no objective direction or

[[Page 28140]]

restriction regarding the determinations of engine displacement. We are 
proposing to define displacement as the intended swept volume of the 
engine to the nearest cubic centimeter, where the engine's swept volume 
is the product of the internal cross-section area of the cylinders, the 
stroke length, and the number of cylinders. As described Section 
VII.C.6 for maximum engine power, we are proposing that the intended 
swept volume must be within the range of the actual swept volumes of 
production engines considering normal production variability. If 
production engines are found to have different swept volumes, this 
should be noted in a change to the application for certification.
(3) Exempted or Excluded Engines
    Under the Clean Air Act, engines that are used in stationary 
applications are not nonroad engines. States are generally preempted 
from setting emission standards for nonroad engines but this preemption 
does not apply to stationary engines. EPA recently adopted emission 
standards for stationary compression-ignition engines sold or used in 
the United States (71 FR 39154, July 11, 2006). In addition, EPA has 
proposed emission standards for stationary spark-ignition engines in a 
separate action (71 FR 33804, June 12, 2006). In pursuing emission 
standards for stationary engines, we have attempted to maintain 
consistency between stationary and nonroad requirements as much as 
possible. As explained in the proposal for stationary spark-ignition 
engines, since stationary spark-ignition engines below 19 kW are almost 
all sold into residential applications, we believe it is not 
appropriate to include requirements for owners or operators that would 
normally be part of a program for implementing standards for stationary 
engines. As a result, in that proposal we indicated that it is most 
appropriate to set exhaust and evaporative emission standards for 
stationary spark-ignition engines below 19 kW as if they were nonroad 
engines. This would allow manufacturers to make a single product that 
meets all applicable EPA standards for both stationary and nonroad 
applications.
    The Clean Air Act provides for different treatment of engines used 
solely for competition. Rather than relying on engine design features 
that serve as inherent indicators of dedicated competitive use, we have 
taken the approach in other programs of more carefully differentiating 
competition and noncompetition models in ways that reflect the nature 
of the particular products. In the case of Small SI engines, we do not 
believe there are engine design features that allow us to differentiate 
between engines that are used solely for competition from those with 
racing-type features that are not used solely for competition. We are 
proposing that handheld and nonhandheld equipment with engines meeting 
all the following criteria would be considered to be used solely for 
competition, except in other cases where information is available 
indicating that engines are not used solely for competition:
     The engine (or equipment in which the engine is installed) 
may not be displayed for sale in any public dealership;
     Sale of the equipment in which the engine is installed 
must be limited to professional competitors or other qualified 
competitors;
     The engine must have performance characteristics that are 
substantially superior to noncompetitive models;
     The engines must be intended for use only in competition 
events sanctioned (with applicable permits) by a state or federal 
government agency or other widely recognized public organization, with 
operation limited to competition events, performance-record attempts, 
and official time trials.
    Engine manufacturers would make their request for each new model 
year and we would deny a request for future production if there are 
indications that some engines covered by previous requests are not 
being used solely for competition. Competition engines are produced and 
sold in very small quantities so manufacturers should be able to 
identify which engines qualify for this exemption. We request comment 
on this approach to qualifying for a competition exemption. (See Sec.  
1054.620.)
    In the rulemaking for recreational vehicles, we chose not to apply 
standards to hobby products by exempting all reduced-scale models of 
vehicles that were not capable of transporting a person (67 FR 68242, 
November 8, 2002). We are proposing to extend that same provision to 
handheld and nonhandheld Small SI engines. (See Sec.  1054.5.)
    In the rulemaking to establish Phase 2 emission standards, we 
adopted an exemption for handheld and nonhandheld engines used in 
rescue equipment. The regulation does not require any request, 
approval, or recordkeeping related to the exemption but we discovered 
while conducting the SBAR Panel described in Section VI.F that some 
companies are producing noncompliant engines under this exemption. We 
are proposing to keep this exemption but add several provisions to 
allow us to better monitor how it is used (see Sec.  1054.625). We are 
proposing to keep the requirement that equipment manufacturers use 
certified engines if they are available. We are proposing to update 
this provision by adding a requirement that equipment manufacturers use 
an engine that has been certified to less stringent Phase 1 or Phase 2 
standards if such an engine is available. We are proposing to 
explicitly allow engine manufacturers to produce engines for this 
exemption (with permanent labels identifying the particular exemption), 
but only if they have a written request for each equipment model from 
the equipment manufacturer. We are further proposing that the equipment 
manufacturer notify EPA of the intent to produce emergency equipment 
with exempted engines. Also, to clarify the scope of this provision, we 
are proposing to define ``emergency rescue situations'' as firefighting 
or other situations in which a person is retrieved from imminent 
danger. Finally, we are proposing to clarify that EPA may discontinue 
the exemption on a case-by-case basis if we find that engines are not 
used solely for emergency and rescue equipment or if we find that a 
certified engine is available to power the equipment safely and 
practically. We propose to apply the provisions of this section for new 
equipment built on or after January 1, 2009.
    The current regulations also specify an exemption allowing 
individuals to import up to three nonconforming handheld or nonhandheld 
engines one time. We are proposing to keep this exemption with three 
adjustments (see Sec.  1054.630). First, we are proposing to allow this 
exemption only for used equipment. Allowing importation of new 
equipment under this exemption is not consistent with the intent of the 
provision, which is to allow people to move to the United States from 
another country and continue to use lawn and garden equipment that may 
already be in the person's possession. Second, we are proposing to 
allow such an importation once every five years but require a statement 
that the person importing the exempted equipment has not used this 
provision in the preceding five years. The current regulations allow 
only one importation in a person's lifetime without including any way 
of making that enforceable. We believe the proposed combination of 
provisions represents an appropriate balance between preserving the 
enforceability of the exemption within the normal flow

[[Page 28141]]

of personal property for people coming into the country. Third, we are 
proposing to no longer require submission of the taxpayer 
identification number since this is not essential for ensuring 
compliance.

C. Proposed Requirements

    A key element of the proposed new requirements for Small SI engines 
is the more stringent exhaust emission standards for nonhandheld 
engines. We are also proposing several changes to the certification 
program that would apply to both handheld and nonhandheld engines. For 
example, we are proposing to clarify the process for selecting an 
engine family's useful life, which defines the length of time over 
which manufacturers' are responsible for meeting emission standards. We 
are also proposing several provisions to update the program for 
allowing manufacturers to use emission credits to show that they meet 
emission standards. The following sections describe the elements of 
this proposed rule.
    The timing for implementation of the new exhaust emission standards 
is described below. Unless we specify otherwise, all the additional 
proposed regulatory changes would apply when engines are subject to the 
emission standards and the other provisions under 40 CFR part 1054. 
This would be model year 2012 for Class I engines and model year 2011 
for Class II engines. For handheld engines, we propose to require 
compliance with the provisions of part 1054, including the 
certification provisions, starting in the 2010 model year. These 
proposed requirements apply to handheld engines unless stated 
otherwise. For convenience we refer to the handheld emission standards 
in part 1054 as Phase 3 standards even though the numerical values 
remain unchanged.
(1) Emission Standards
    Extensive testing and dialogue with manufacturers and other 
interested parties has led us to a much better understanding of the 
capabilities and limitations of applying emission control technologies 
to Small SI nonhandheld engines. As described in the Draft RIA, we have 
collected a wealth of information related to the feasibility, 
performance characteristics, and safety implications of applying 
catalyst technology to these engines. We have concluded within the 
context of Clean Air Act section 213 that it is appropriate to propose 
emission standards that are consistent with those adopted by California 
ARB. We are proposing HC+NOX emission standards of 10.0 g/
kW-hr for Class I engines starting in the 2012 model year, and 8.0 g/
kW-hr for Class II engines starting in the 2011 model year (see Sec.  
1054.105). For both classes of nonhandheld engines we are proposing to 
maintain the existing CO standard of 610 g/kW-hr.
    We are proposing to eliminate the defined subclasses for the 
smallest sizes of nonhandheld engines starting with implementation of 
the Phase 3 standards. Under the current regulations in part 90, Class 
I-A is designated for engines with displacement below 66 cc that may be 
used in nonhandheld applications. To address the technological 
constraints of these engines, all the current requirements for these 
engines are the same as for handheld engines. Class I-B is similarly 
designated for engines with displacement between 66 and 100 cc that may 
be used in nonhandheld applications. These engines are currently 
subject to a mix of provisions that result in an overall stringency 
that lies between handheld and nonhandheld engines. We are proposing to 
revise the regulations such that engines below 80 cc are subject to the 
Phase 3 handheld engine standards in part 1054 starting in the 2010 
model year. We are also proposing to allow engines below 80 cc to be 
used without restriction in nonhandheld equipment. Identifying the 
threshold at 80 cc aligns with the California ARB program. For 
nonhandheld engines at or above 80 cc, we are proposing to treat them 
in every way as Class I engines. Based on the fact that it is more 
difficult for smaller displacement engines to achieve the same g/kW-hr 
emission level as larger displacement engines, it will be more of a 
challenge for manufacturers to achieve a 10.0 g/kW-hr HC+NOX 
level on these smallest Class I engines. However, for those engines 
unable to achieve the level of the proposed standards (either with or 
without a catalyst), manufacturers may elect to rely on emissions 
averaging to comply with emission standards. We believe all 
manufacturers producing engines formerly included in Class I-B also 
have a wide enough range of engine models that they should be able to 
generate sufficient credits to meet standards across the full product 
line. (See Sec.  1054.101 and Sec.  1054.801.)
    We are proposing another slight change to the definition of 
handheld engines that may affect whether an engine is subject to 
handheld or nonhandheld standards. The handheld definition relies on a 
weight threshold for certain engines. As recently as 1999, we affirmed 
that the regulation should allow for the fact that switching to a 
heavier four-stroke engine to meet emission standards might 
inappropriately cause an engine to no longer qualify as a handheld 
engine (64 FR 5252, February 3, 1999). The regulation accordingly 
specifies that the weight limit is 20 kilograms for one-person augers 
and 14 kilograms for other types of equipment, based on the weight of 
the engine that was in place before applying emission control 
technologies. We believe it is impractical to base a weight limit on 
product specifications that have become difficult to establish. We are 
therefore proposing to increase each of the specified weight limits by 
1 kilogram, representing the approximate additional weight related to 
switching to a four-stroke engine, and applying the new weight limit to 
all engines and equipment (see Sec.  1054.801). We request comment on 
this adjustment to the handheld engine definition.
    The regulations in part 90 allow manufacturers to rely on altitude 
kits to comply with emission requirements at high altitude. We are 
proposing to continue with this approach but to clarify that all 
nonhandheld engines must comply with Phase 3 standards without altitude 
kits at barometric pressures above 94.0 kPa, which corresponds to 
altitudes up to about 2,000 feet above sea level (see Sec.  1054.115). 
This would ensure that all areas east of the Rocky Mountains and most 
of the populated areas in Pacific Coast states would have compliant 
engines without depending on engine modifications. This becomes 
increasingly important as we anticipate manufacturers relying on 
technologies that are sensitive to controlling air-fuel ratio for 
reducing emissions. Engine manufacturers must identify the altitude 
ranges for proper engine performance and emission control that are 
expected with and without the altitude kit in the owners manual. The 
owners manual must also state that operating the engine with the wrong 
engine configuration at a given altitude may increase its emissions and 
decrease fuel efficiency and performance. See Section V.E.5 for further 
discussion related to the deployment of altitude kits where the 
manufacturers rely on them for operation at higher altitudes.
    We are proposing a slightly different approach for handheld engines 
with respect to altitude. Since we are not adopting more stringent 
exhaust emission standards, we believe it is appropriate to adopt 
provisions that are consistent with current practice at this time. We 
are therefore proposing to require handheld engines to comply with the 
current standards without altitude kits at barometric pressures

[[Page 28142]]

above 96.0 kPa, which would allow for testing in most weather 
conditions at all altitudes up to about 1,100 feet above sea level.
    Spark-ignition engines used for marine auxiliary power are covered 
by the same regulations as land-based engines of the same size. 
However, the marine versions of Small SI engines are able to make use 
of ambient water for enhanced cooling of the engine and exhaust system. 
Exhaust systems for these engines are water-jacketed to maintain low 
surface temperatures to minimize the risk of fires on boats where the 
generator is often installed in small compartments within the boat. 
Recently, auxiliary marine engine manufacturers have developed advanced 
technology in an effort to improve fuel consumption and CO emission 
rates for marine generators. This advanced technology includes the use 
of electronic fuel injection and three-way catalysts. As a result, 
manufacturers are offering new products with more than a 99 percent 
reduction in CO and have expressed their intent to offer only these 
advanced technology engines in the near future. They have stated that 
these low CO engines are due to market demand. We are proposing a CO 
standard of 5.0 g/kW-hr CO for marine generator engines to reflect the 
recent trend in marine generator engine design (see Sec.  1054.105). 
For other auxiliary marine engines, we are proposing the same CO 
emission limits as for land-based engines. We believe this cap is 
necessary to prevent backsliding in CO emissions that could occur if 
new manufacturers were to attempt to enter the market with cheaper, 
high-CO designs. See Section II for a discussion of air quality 
concerns related to CO emissions. We request comment on the 
appropriateness of setting a separate standard for marine auxiliary 
engines and on the most appropriate level of such a standard.
    At this time, we are planning to continue the current regulatory 
approach for wintertime engines (e.g., engines used exclusively to 
power equipment such as snowthrowers and ice augers). Under this 
proposal, the HC+NOX exhaust emission standards would be 
optional for wintertime engines. However, if a manufacturer chooses to 
certify its wintertime engines to such standards, those engines would 
be subject to all the requirements as if the optional standards were 
mandatory. We are adding a definition of wintertime engines to clarify 
which engines qualify for these special provisions. We are also 
proposing to require that manufacturers identify these as wintertime 
engines on the emission control information label to prevent someone 
from inappropriately installing these engines (either new or used) in 
equipment that would not qualify for the wintertime exemption.
    All engines subject to standards must continue to control crankcase 
emissions.
(2) Useful Life
    The Phase 2 standards for Small SI engines included the concept 
that manufacturers are responsible for meeting emission standards over 
a useful life period. The useful life defines the design target for 
ensuring the durability of emission controls under normal in-use 
operation for properly maintained engines. Given the very wide range of 
engine applications, from very low-cost consumer products to commercial 
models designed for continuous operation, we determined that a single 
useful life value for all products, which is typical for other engine 
programs, was not appropriate for Small SI engines. We proposed at that 
time to determine the useful life for an engine family based on 
specific criteria, but commenters suggested that such a requirement was 
overly rigid and unnecessary. The final rule instead specified three 
alternative useful life values, giving manufacturers the responsibility 
to select the useful life that was most appropriate for their engines 
and the corresponding types of equipment. The preamble to the final 
rule expressed a remaining concern that manufacturers might not select 
the most appropriate useful life value, both for ensuring effective in-
use emission control and for maintaining the integrity of emission-
credit calculations. The preamble also stated our intent to 
periodically review the manufacturers' decisions to determine whether 
modifications to these rules are appropriate.
    The regulations in Sec.  90.105 provide a benchmark for determining 
the appropriate useful life value for an engine family. The regulations 
direct manufacturers to select the useful life value that ``most 
closely approximates the expected useful lives of the equipment into 
which the engines are anticipated to be installed.'' To maintain a 
measure of accountability, we included a requirement that manufacturers 
document the basis for their selected useful life values. The suggested 
data included, among other things: (1) Surveys of the life spans of the 
equipment in which the subject engines are installed; (2) engineering 
evaluations of field-aged engines to ascertain when engine performance 
deteriorates to the point where utility and/or reliability is impacted 
to a degree sufficient to necessitate overhaul or replacement; and (3) 
failure reports from engine customers. These regulatory provisions 
identify the median time to retirement for in-use equipment as the 
marker for defining the useful life period. This allows manufacturers 
to consider that equipment models may fail before the engine has 
reached the point of failure and that engines may be installed in 
different types of equipment with varying usage patterns. Engines used 
in different types of equipment, or even engines used in the same 
equipment models used by different operators, may experience widely 
varying usage rates. The manufacturer is expected to make judgments 
that take this variability into account when estimating the median life 
of in-use engines and equipment.
    Several manufacturers have made a good faith effort to select 
appropriate useful life values for their engine families, either by 
selecting only the highest value, or by selecting higher values for 
families that appear more likely to be used in commercial applications. 
At the same time, we have observed several instances in which engine 
models are installed in commercial equipment and marketed as long-life 
products but are certified to the minimum allowable useful life period. 
As described in the Phase 2 final rule, we are considering 
modifications to the regulations to address this recurring problem.
    After assessing several ideas, we are proposing an approach that 
preserves the fundamental elements of the current provisions related to 
useful life but clarifies and enhances its implementation (see Sec.  
1054.107). Manufacturers will continue to select the most appropriate 
useful life from the same nominal values to best match the expected in-
use lifetime of the equipment into which the engines in the engine 
family will be installed. Manufacturers must continue to document the 
information supporting their selected useful life. We are considering 
three approaches to address remaining concerns with the process of 
selecting useful life values.
    First, for manufacturers not selecting the highest available 
nominal value for useful life, we would expect to routinely review the 
information to confirm that it complies with the regulation. Where our 
review indicates that the selected useful life may not be appropriate 
for an engine family, we may request further justification. If we 
determine from available information that a longer useful life is 
appropriate, the manufacturer must either provide additional 
justification or select a longer

[[Page 28143]]

useful life for that engine family. We would encourage manufacturers to 
use the proposed provisions related to preliminary approval in Sec.  
1054.210 if there is any uncertainty related to the useful life 
selection. We would rather work to establish this together early in the 
certification process rather than reviewing a completed application for 
certification to evaluate whether the completed durability 
demonstration is sufficient.
    Second, we believe it is appropriate to modify the regulations to 
allow nonhandheld engine manufacturers to select a useful life value 
that is longer than the three specified nominal values. Manufacturers 
may choose to do this for the marketing advantage of selling a long-
life product or they may want to generate emission credits that 
correspond to an expected lifetime that is substantially longer than we 
would otherwise allow. We are proposing to allow manufacturers to 
select longer useful life values in 100-hour increments. Durability 
testing for certification would need to correspond to the selected 
useful life period. We have considered the possibility that a 
manufacturer might overstate an engine family's useful life to generate 
emission credits while knowing that engines may not operate that long. 
We believe the inherent testing burden and compliance liability is 
enough to avoid such a problem, but we are specifying maximum values 
corresponding with the applicable useful life for comparable diesel 
engines or Large SI engines. We are not proposing to allow for longer 
useful life values for handheld engines.
    We are also proposing to require that engines and equipment be 
labeled to identify the applicable useful life period. The current 
requirement allows manufacturers to identify the useful life with code 
letters on the engine's emission control information label, with the 
numerical value of the useful life spelled out in the owners manual. We 
believe it is important for equipment manufacturers and consumers to be 
able to find an unambiguous designation showing the manufacturer's 
expectations about the useful life of the engine. There has also been 
some interest in using descriptive terms to identify the useful life on 
the label. We believe any terminology would communicate less 
effectively than the numerical value of the useful life. However, we 
request comment on allowing or requiring manufacturers to also include 
descriptive terms. We believe it would be most appropriate to 
characterize the three useful life values in increasing order as 
Residential, Premium Residential (or General Purpose), and Commercial. 
Any useful life values beyond the three nominal values would 
appropriately be identified as Heavy Commercial. Handheld engine 
manufacturers have suggested using the terms Light Use, Medium Use, and 
Heavy Use to characterize the three useful life categories applicable 
to handheld engines.
    In all of our other engine programs, useful life is defined in 
terms of years of use or extent of engine operation, whichever comes 
first. Under the current regulations, manufacturers are responsible for 
meeting emission standards for any in-use engine that is properly 
maintained and used over the full useful life period. Since the useful 
life is defined in operating hours without regard to calendar years, 
some engines that accumulate operating hours very slowly could remain 
within the useful life period for ten years or more. We request comment 
regarding the appropriateness of revising the useful life to limit the 
useful life period to five years or the specified number of operating 
hours, whichever comes first. Adding a five-year limit on the useful 
life would not change the certification process.
(3) Averaging, Banking, and Trading
    EPA has included averaging, banking, and trading (ABT) programs in 
almost all of its recent mobile source emissions control programs. 
EPA's existing Phase 2 regulations for Small SI engines include an 
exhaust ABT program (40 CFR 90.201 through 90.211). We propose to adopt 
an ABT program for the Phase 3 HC+NOX exhaust emission 
standards that is similar to the existing program (see part 1054, 
subpart H in the proposed regulations). The proposed exhaust ABT 
program is intended to enhance the ability of engine manufacturers to 
meet the emission standards for the proposed model years. The proposed 
exhaust ABT program is also structured to avoid delay of the transition 
to the new exhaust emission controls. As described in Section VI, we 
are proposing a separate evaporative ABT program for fuel tanks used in 
Small SI equipment (and for fuel lines used in handheld equipment). We 
are proposing that credits cannot be exchanged between the exhaust ABT 
program and the evaporative ABT program.
    The exhaust ABT program has three main components. Averaging means 
the exchange of emission credits between engine families within a given 
engine manufacturer's product line for a specific model year. Engine 
manufacturers divide their product line into ``engine families'' that 
are comprised of engines expected to have similar emission 
characteristics throughout their useful life. Averaging allows a 
manufacturer to certify one or more engine families at levels above the 
applicable emission standard, but below a set upper limit. This level 
then becomes the applicable standard for all of the engines in that 
engine family, for purposes of certification, in-use testing, and the 
like. However, the increased emissions must be offset by one or more 
engine families within that manufacturer's product line that are 
certified below the same emission standard, such that the average 
standard from all the manufacturer's engine families, weighted by 
engine power, regulatory useful life, and production volume, is at or 
below the level of the emission standard. Banking means the retention 
of emission credits by the engine manufacturer for use in future model 
year averaging or trading. Trading means the exchange of emission 
credits between engine manufacturers which can then be used for 
averaging purposes, banked for future use, or traded to another engine 
manufacturer.
    Because we are not proposing any change in the general equation 
under which emission credits are calculated, EPA is proposing to allow 
manufacturers to use Phase 2 credits generated under the part 90 ABT 
program for engines that are certified in the Phase 3 program under 
part 1054, within the limits described below. As with the existing 
exhaust ABT program for Phase 2 engines in part 90, we are proposing 
that engines sold in California which are subject to the California ARB 
standards would not be included in the proposed exhaust ABT program 
because they are subject to California's requirements and not EPA's 
requirements. Furthermore, even though we are not proposing new exhaust 
emission standards for handheld engines, the handheld engine 
regulations are migrating to part 1054. Therefore, handheld engines 
will be included in the proposed ABT program under part 1054 with one 
change in the overall program as described below.
    Under an ABT program, averaging is allowed only between engine 
families in the same averaging set, as defined in the regulations. For 
the exhaust ABT program, we are proposing to separate handheld engines 
and nonhandheld engines into two distinct averaging sets starting with 
the 2011 model year. Under the proposed program, credits may generally 
be used interchangeably between Class I and Class II engine families, 
with a limited restriction on Phase 3 credits during model years 2011

[[Page 28144]]

and 2012 as noted below. Likewise, credits will be able to be used 
interchangeably between all three handheld engine classes (Classes III, 
IV, and V). Because the Phase 2 exhaust ABT program allowed exchange 
across all engine classes (i.e., allowing exchanges between handheld 
engines and nonhandheld engines), manufacturers using credits beginning 
with the 2011 model year would need to show that the credits were 
generated within the allowed category of engines. For many companies, 
especially those in the handheld market, this will potentially be 
straightforward since they are primarily in the handheld market. For 
companies that have a commingled pool of emission credits generated by 
both handheld engines and nonhandheld engines, this will take some more 
careful accounting. Because manufacturers are aware of this already at 
the time of this proposal, keeping records to distinguish handheld 
credits and nonhandheld credits will be relatively straightforward for 
2006 and later model years.
    We are proposing two exceptions to the provision restricting credit 
exchanges between handheld engines and nonhandheld engines. Currently, 
some companies that are primarily nonhandheld engine manufacturers also 
sell a relatively limited number of handheld engines. Under the Phase 2 
program, these engine manufacturers can use credits from nonhandheld 
engines to offset the higher emissions of their handheld engines. 
Because we are not proposing new exhaust requirements for handheld 
engines, we are proposing to address this existing practice by 
specifying that an engine manufacturer may use emission credits from 
their nonhandheld engines for their handheld engines under the 
following conditions. A manufacturer may use credits from their 
nonhandheld engines for their handheld engines but only where the 
handheld engine family is certified in 2008 and later model years 
without any design changes from the 2007 model year and the FEL of the 
handheld engine family does not increase above the level that applied 
in the 2007 model year unless such an increase is based on emission 
data from production engines. We believe this allows for engine 
manufacturers to continue producing these handheld engines for use in 
existing handheld models of low-volume equipment applications while 
preventing new high-emitting handheld engine families from entering the 
market through the use of nonhandheld engine credits. As discussed 
below, we are proposing to prohibit the use of Phase 2 nonhandheld 
engine credits after 2013 to demonstrate compliance with the Phase 3 
nonhandheld engine standards. For this reason, we request comment on 
whether we should allow only Phase 3 nonhandheld engine credits to be 
used under this handheld engine credit provision after 2013 as well.
    A second exception to the provision restricting credit exchanges 
between handheld engines and nonhandheld engines arises because of our 
proposed handling of engines below 80cc. Under the proposed Phase 3 
program, all engines below 80cc are considered handheld engines for the 
purposes of the emission standards. However, a few of these engines are 
used in nonhandheld applications. Therefore, EPA will allow a 
manufacturer to generate nonhandheld ABT credits from engines below 
80cc for those engines a manufacturer has determined are used in 
nonhandheld applications. (The credits would be generated against the 
applicable handheld engine standard.) These nonhandheld credits could 
be used within the Class I and Class II engine classes to demonstrate 
compliance with the Phase 3 exhaust standards (subject to applicable 
restrictions). The credits generated by engines below 80cc used in 
handheld applications could only be used for other handheld engines.
    Under an ABT program, a manufacturer establishes a ``family 
emission limit'' (FEL) for each participating engine family. This FEL 
may be above or below the standard. The FEL becomes the enforceable 
emissions limit for all the engines in that family for purposes of 
compliance testing. FELs that are established above the standard may 
not exceed an upper limit specified in the ABT regulations. For 
nonhandheld engines we are proposing FEL caps to prevent the sale of 
very high-emitting engines. Under the proposed FEL cap, manufacturers 
would need to establish FELs at or below the levels of the Phase 2 
HC+NOX emission standards of 16.1 g/kW-hr for Class I 
engines and 12.1 g/kW-hr for Class II engines. (The Phase 3 FEL cap for 
Class I engines with a displacement between 80 cc and 100 cc would be 
40.0 g/kW-hr since these engines would have been Class I-B engines 
under the Phase 2 regulations and subject to this higher level.) For 
handheld engines, where we are not proposing new exhaust emission 
standards, we are maintaining the FEL caps as currently specified in 
the part 90 ABT regulations.
    For nonhandheld engines we are proposing two special provisions 
related to the transition from Phase 2 to Phase 3 standards. First, we 
are proposing incentives for manufacturers to produce and sell engines 
certified at or below the Phase 3 standards before the standards are 
scheduled to be implemented. Second, we are proposing provisions to 
allow the use of Phase 2 credits for a limited period of time under 
specific conditions. The following discussions describes each of these 
provisions in more detail for Class I engines and Class II engines 
separately.
    For Class I, engine manufacturers could generate early Phase 3 
credits by producing engines with an FEL at or below 10.0 g/kW-hr prior 
to 2012. These early Phase 3 credits would be calculated and 
categorized into two distinct types of credits, Transitional Phase 3 
credits and Enduring Phase 3 credits. For engines certified with an FEL 
at or below 10.0 g/kW-hr, the manufacturer would earn Transitional 
Phase 3 credits. The Transitional Phase 3 credits would be calculated 
based on the difference between 10.0 g/kW-hr and 15.0 g/kW-hr. (The 
15.0 g/kW-hr level is the production-weighted average of Class I FEL 
values under the Phase 2 program.) Manufacturers could use the 
Transitional Phase 3 credits from Class I engines in 2012 through 2014 
model years. For engines certified with an FEL below 10.0 g/kW-hr, 
manufacturers would earn Enduring Phase 3 credits in addition to the 
Transitional Phase 3 credits described above. The Enduring Phase 3 
credits would be calculated based on the difference between the FEL for 
the engine family and 10.0 g/kW-hr (i.e., the applicable Phase 3 
standard). The Enduring Phase 3 credits could be used once the Phase 3 
standards are implemented without the model year restriction noted 
above for Transitional Phase 3 credits.
    For Class I, engine manufacturers may use Phase 2 credits generated 
by nonhandheld engines for the first two years of the Phase 3 standards 
(i.e., model years 2012 and 2013) under certain conditions. The 
manufacturer must first use all of its available Phase 3 credits to 
demonstrate compliance with the Phase 3 standards. This would include 
all early Phase 3 credits (Transitional and Enduring) as well as all 
other Phase 3 credits, subject to the cross-class credit restriction 
noted below which applies prior to model year 2013. If these Phase 3 
credits are sufficient to demonstrate compliance, the manufacturer may 
not use Phase 2 credits. If these Phase 3 credits are insufficient to 
demonstrate compliance, the manufacturer could use Phase 2 credits to a 
limited degree (under the conditions described below) to cover the

[[Page 28145]]

remaining amount of credits needed to demonstrate compliance.
    The maximum number of Phase 2 HC+NOX exhaust emission 
credits a manufacturer could use for their Class I engines would be 
calculated based on the characteristics of Class I engines produced 
during the 2007, 2008, and 2009 model years. For each of those years, 
the manufacturer would calculate a Phase 2 credit allowance using the 
ABT credit equation and inserting 1.6 g/kW-hr for the ``Standard--FEL'' 
term, and basing the rest of the values on the total production of 
Class I engines, the production-weighted power for all Class I engines, 
and production-weighted useful life value for all Class I engines 
produced in each of those years. Manufacturers would not include their 
wintertime engines in the calculations unless the engines are certified 
to meet the otherwise applicable HC+NOX emission standard. 
The maximum number of Phase 2 HC+NOX exhaust emission 
credits a manufacturer could use for their Class I engines (calculated 
in kilograms) would be the average of the three values calculated for 
model years 2007, 2008, and 2009. The calculation described above 
allows a manufacturer to use Phase 2 credits to cover a cumulative 
shortfall over the first two years for their Class I engines of 1.6 g/
kW-hr above the Phase 3 standard.
    The Phase 2 credit allowance for Class I engines could be used all 
in 2012, all in 2013, or partially in either or both model year's ABT 
compliance calculations. Because ABT compliance calculations must be 
done annually, the manufacturer will know its 2013 remaining allowance 
based on its 2012 calculation. For example, if a manufacturer uses all 
of its Phase 2 credit allowance in 2012, it will have no use of Phase 2 
credits for 2013. Conversely, if a manufacturer doesn't use any Phase 2 
credits in 2012, it will have all of its Phase 2 credit allowance 
available for use in 2013. And of course, if a manufacturer uses less 
than its calculated total credits based on the 1.6 g/kW-hr limit in 
2012, the remainder would be available for use in 2013. This provision 
allows for some use of Phase 2 emission credits to address the 
possibility of unanticipated challenges in reaching the Phase 3 
emission levels in some cases or selling Phase 3 compliant engines 
early nationwide, without creating a situation that would allow 
manufacturers to substantially delay the introduction of Phase 3 
emission controls.
    For Class II, engine manufacturers could generate early Phase 3 
credits by producing engines with an FEL at or below 8.0 g/kW-hr prior 
to 2011. These early Phase 3 credits would be calculated and 
categorized as Transitional Phase 3 credits and Enduring Phase 3 
credits. For engines certified with an FEL at or below 8.0 g/kW-hr, the 
manufacturer would earn Transitional Phase 3 credits. The Transitional 
Phase 3 credits would be calculated based on the difference between 8.0 
g/kW-hr and 11.0 g/kW-hr. (The 11.0 g/kW-hr level is the production-
weighted average of Class II FEL values under the Phase 2 program.) 
Manufacturers could use the Transitional Phase 3 credits from Class II 
engines in 2011 through 2013 model years. For engines certified with an 
FEL below 8.0 g/kW-hr, manufacturers would earn Enduring Phase 3 
credits in addition to the Transitional Phase 3 credits described 
above. The Enduring Phase 3 credits would be calculated based on the 
difference between the FEL for the engine family and 8.0 g/kW-hr (i.e., 
the applicable Phase 3 standard). The Enduring Phase 3 credits could be 
used once the Phase 3 standards are implemented without the model year 
restriction noted above for Transitional Phase 3 credits.
    For Class II, engine manufacturers may use Phase 2 credits 
generated by nonhandheld engines for the first three years of the Phase 
3 standards (i.e., model years 2011, 2012 and 2013) under certain 
conditions. The manufacturer must first use all of its available Phase 
3 credits to demonstrate compliance with the Phase 3 standards. This 
would include all early Phase 3 credits (Transitional and Enduring) as 
well as all other Phase 3 credits, subject to the cross-class credit 
restriction noted below which applies prior to model year 2013. If 
these credits are sufficient to demonstrate compliance, the 
manufacturer may not use Phase 2 credits. If these Phase 3 credits are 
insufficient to demonstrate compliance, the manufacturer could use 
Phase 2 credits to a limited degree (under the conditions described 
below) to cover the remaining amount of credits needed to demonstrate 
compliance.
    The maximum number of Phase 2 HC+NOX exhaust emission 
credits a manufacturer could use for their Class II engines would be 
calculated based on the characteristics of Class II engines produced 
during the 2007, 2008, and 2009 model years. For each of those years, 
the manufacturer would calculate a Phase 2 credit allowance using the 
ABT credit equation and inserting 2.1 g/kW-hr for the ``Standard--FEL'' 
term, and basing the rest of the values on the total production of 
Class II engines, the production-weighted power for all Class II 
engines, and production-weighted useful life value for all Class II 
engines produced in each of those years. Manufacturers would not 
include their wintertime engines in the calculations unless the engines 
are certified to meet the otherwise applicable HC+NOX 
emission standard. The maximum number of Phase 2 HC+NOX 
exhaust emission credits a manufacturer could use for their Class II 
engines (calculated in kilograms) would be the average of the three 
values calculated for model years 2007, 2008, and 2009. The calculation 
described above allows a manufacturer to use Phase 2 credits to cover a 
cumulative shortfall over the first three years for their Class II 
engines of 2.1 g/kW-hr above the Phase 3 standard.
    The Phase 2 credit allowance for Class II engines could be used all 
in 2011, all in 2012, all in 2013, or partially in any or all three 
model year's ABT compliance calculations. Because ABT compliance 
calculations must be done annually, the manufacturer will know its 
remaining allowance based on its previous calculations. For example, if 
a manufacturer uses all of its Phase 2 credit allowance in 2011, it 
will have no Phase 2 credits for 2012 or 2013. However, if a 
manufacturer uses less than its calculated total credits based on the 
2.1 g/kW-hr limit in 2011, it will have the remainder of its allowance 
available for use in 2012 and 2013. This provision allows for some use 
of Phase 2 emission credits to address the possibility of unanticipated 
challenges in reaching the Phase 3 emission levels in some cases or 
selling Phase 3 engines nationwide, without creating a situation that 
would allow manufacturers to substantially delay the introduction of 
Phase 3 emission controls.
    Engine manufacturers have raised concerns that despite all of their 
planning, they may not be able to accurately predict their use of 
credits at the beginning of the year. They are concerned that they may 
end up in a credit deficit situation if sales do not materialize as 
projected, potentially needing to use more Phase 2 credits than they 
have available to them. In order to prevent such a non-compliance 
situation from occurring, manufacturers have suggested that we allow 
manufacturers to carry a limited credit deficit during the initial 
years of the Phase 3 program. EPA has allowed such provisions in other 
rules, including deficit provisions for handheld engines in the Phase 2 
regulations in which the manufacturer was required to cover the deficit 
in the next four model years with a penalty applied that increased over 
time depending how soon the deficit

[[Page 28146]]

was repaid. EPA requests comment on providing some type of credit 
deficit provisions for the Phase 3 exhaust standards for nonhandheld 
engines including what limits and penalties would be appropriate if 
such provisions were adopted.
    To avoid the use of credits to delay the introduction of Phase 3 
technologies, we are also proposing that manufacturers may not use 
Phase 3 credits from Class I engines to demonstrate compliance with 
Class II engines in the 2011 and 2012 model years. Similarly, we are 
proposing that manufacturers may not use Phase 3 credits from Class II 
engines to demonstrate compliance with Class I engines in the 2012 
model year. The 1.6 kW-hr and 2.1 g/kW-hr allowances discussed above 
may not be traded across engine classes or among manufacturers.
    We are proposing to make two additional adjustments related to the 
exhaust ABT program for engines subject to the new emission standards. 
As with all our other emission control programs, we are proposing that 
engine manufacturers identify an engine's FEL on the emission control 
information label (see Sec.  1054.135). This is important for readily 
establishing the enforceable level of emission control that applies for 
each engine. Recent experience has shown that this is also necessary in 
cases where the engine's build date is difficult to determine. We are 
proposing to require that lowering an FEL after the start of production 
may occur only if the manufacturer has emission data from production 
engines justifying the lower FEL (see Sec.  1054.225). This prevents 
manufacturers from making FEL changes late in the model year to 
generate more emission credits (or use fewer emission credits) when 
there is little or no opportunity to verify whether the revised FEL is 
appropriate for the engine family. This provision is common in EPA's 
emission control programs for other engine categories. We are also 
proposing that the any revised FEL can apply only for engines produced 
after the FEL change. This is necessary to prevent manufacturers from 
recalculating emission credits in a way that leaves no way of verifying 
that the engines produced prior to the FEL change met the applicable 
requirements. It is also consistent with the proposal to require 
identification of the FEL on the emission control information label. 
Manufacturers have raised concerns that this approach sets up an 
inappropriate incentive to set FELs with the smallest possible 
compliance margin to avoid foregone emission credits in case 
production-line testing shows that actual emission levels were below 
that represented by the emission-data engine for certification. 
However, it is not clear why manufacturers should not perform 
sufficient testing early in the model year to be confident that the FEL 
is properly matched to the emission levels from production engines. 
Nevertheless, we request comment on any appropriate methods to use the 
results of production-line testing to revise FELs retroactively such 
that the past production is clearly compliant with respect to the 
modified FEL. An important element of our compliance program involves 
the responsibility to meet standards with production-line testing, not 
just with a backward-looking calculation, but with a real-time 
evaluation at the point of testing. We would therefore not consider 
allowing revised FELs to apply for more than the first half of the 
production for a given model year.
    As described below in Section V.E.3., we are proposing that a 
limited number of Class II engines certified by engine manufacturers 
with a catalyst as Phase 3 engines, may be installed by equipment 
manufacturers in equipment without the catalyst. (This would only be 
allowed when the engine is shipped separately from the exhaust system 
under the provisions described in Section V.E.2.) Because engine 
manufacturers may be generating emission credits from these catalyst-
equipped engines, EPA is concerned that engine manufacturers could be 
earning exhaust ABT credits for engines that are sold but never have 
the catalyst installed. In discussions with EPA, engine manufacturers 
expressed concern about the difficulty of tracking the eventual use of 
these engines by equipment manufacturers (i.e., whether the catalyst-
equipped exhaust system was installed or not). Therefore, instead of 
requiring engine manufacturers to track whether equipment manufacturers 
install the catalyst-equipped exhaust system into the equipment, EPA is 
proposing for model years 2011 through 2014 that all Class II engine 
families which are offered for sale under the separate shipment 
provisions must decrease the number of ABT credits generated by the 
engine family by 10 percent. This adjustment would only apply to 
engines generating credits because those are the engines most likely to 
be equipped with catalysts. We believe the 10 percent decrease from 
credit generating engines should provide an emission adjustment 
commensurate with the potential use of the equipment manufacturer 
flexibility provisions described in Section V.E.3. We request comment 
on this approach to addressing the concern related to engines involving 
delegated-assembly provisions. In particular, we request comment 
regarding the amount of the credit adjustment, and whether there might 
be alternative approaches that would address this concern.
    For all emission credits generated by engines under the Phase 3 
exhaust ABT program, we are proposing an unlimited credit life. We 
consider these emission credits to be part of the overall program for 
complying with Phase 3 standards. Given that we may consider further 
reductions beyond the Phase 3 standards in the future, we believe it 
will be important to assess the ABT credit situation that exists at the 
time any post-Phase 3 standards are considered. We will need to set 
such future emission standards based on the statutory direction that 
emission standards must represent the greatest degree of emission 
control achievable, considering cost, safety, lead time, and other 
factors. Emission credit balances will be part of the analysis for 
determining the appropriate level and timing of new standards. If we 
were to allow the use of Phase 3 credits for meeting post-Phase 3 
standards, we may, depending on the level of Phase 3 credit banks, need 
to adopt emission standards at more stringent levels or with an earlier 
start date than we would absent the continued or limited use of Phase 3 
credits. Alternatively, we could adopt future standards without 
allowing the use of Phase 3 credits. The proposal described in this 
notice describes a middle path in which we allow the use of Phase 2 
credits to meet the Phase 3 standards, with provisions that limit the 
extent and timing of using these credits.
    We are requesting comment on one particular issue regarding credit 
life. As proposed, credits earned under the Phase 3 exhaust ABT program 
would have an unlimited lifetime. This could result in a situation 
where credits generated by an engine sold in a model year are not used 
until many years later when the engines generating the credits have 
been scrapped and are no longer part of the fleet. EPA believes there 
may be value to limiting the use of credits to the period that the 
credit-generating engines exist in the fleet. For this reason, EPA 
requests comment on limiting the lifetime of the credits generated 
under the Phase 3 exhaust ABT program to five years. The five-year 
period is intended to be similar to the typical median life of Small SI 
equipment and is consistent with the contemplated specification for 
defining the useful life in years in addition to

[[Page 28147]]

operating hours (see Section V.C.2 for more information).

D. Testing Provisions

    The test procedures provide an objective measurement for 
establishing whether engines comply with emission standards. The 
following sections describe a variety of proposed changes to the 
current test procedures. Except as identified in the following 
sections, we are proposing to preserve the testing-related regulatory 
provisions that currently apply under 40 CFR part 90. Note that we will 
approve any appropriate alternatives, deviations, or interpretations of 
the new testing requirements on a case-by-case basis rather than 
operating under any presumption that any such judgments made under the 
Phase 1 or Phase 2 programs will continue to apply.
(1) Migrating Procedures to 40 CFR Part 1065
    Manufacturers have been using the procedures in 40 CFR part 90 to 
test their engines for certification of Phase 1 and Phase 2 engines. As 
part of a much broader effort, we have adopted comprehensive testing 
specifications in 40 CFR part 1065 that are intended to serve as the 
basis for testing all types of engines. The procedures in part 1065 
include updated information reflecting the current state of available 
technology. We are proposing to apply the procedures in part 1065 to 
nonhandheld engines starting with the applicability of the Phase 3 
standards as specified in 40 CFR part 1054, subpart F. As described in 
Section IX, the procedures in part 1065 identifies new types of 
analyzers and updates a wide range of testing specifications, but 
leaves intact the fundamental approach for measuring exhaust emissions. 
There is no need to shift to the part 1065 procedures for nonhandheld 
engines before the proposed Phase 3 standards apply. See Section IX for 
additional information.
    We are not proposing new exhaust emission standards for handheld 
engines so there is no natural point in time for shifting to the part 
1065 procedures. For the reasons described above and in Section IX, we 
nevertheless believe handheld engines should also use the part 1065 
procedures for measuring exhaust emissions. We propose to require 
manufacturers to start using the part 1065 procedures in the 2012 model 
year. Manufacturers would be allowed to continue certifying engines 
using carryover data generated under the part 90 procedures, but any 
new certification testing would be subject to the part 1065 procedures.
    Engine manufacturers have raised one issue related to the specified 
test procedures in part 1065. The calculations for determining mass 
emissions depend on a simplifying assumption that combustion is at 
stoichiometry or is in fuel-lean environment. This is not the case for 
many Small SI engines. The equation with the simplifying assumption 
does not take into account the equilibrium reaction between hydrogen 
and water. As a result, engines with fuel-rich operation would have 
detectable hydrogen concentrations in the exhaust, which would cause 
the analyzers to have a reading for hydrocarbon emissions that is 
somewhat higher than the actual value. To the extent there is a 
concern, we believe it would always be appropriate to rely on the 
reference equations without the simplifying assumptions made for the 
equations published in part 1065. We request comment on this approach 
to measurements from Small SI engines.
(2) Duty Cycle
    The regulations under part 90 currently specify duty cycles for 
testing engines for exhaust emissions. The current requirements specify 
how to control speeds and loads and describe the situations in which 
the installed engine governor controls engine speed. We are proposing 
to extend these provisions to testing under the new standards with a 
few adjustments described below. For engines equipped with an engine 
speed governor, the current regulations at 40 CFR 90.409(a)(3) state:

    For Phase 2 Class I, Phase 2 Class I-B, and Phase 2 Class II 
engines equipped with an engine speed governor, the governor must be 
used to control engine speed during all test cycle modes except for 
Mode 1 or Mode 6, and no external throttle control may be used that 
interferes with the function of the engine's governor; a controller 
may be used to adjust the governor setting for the desired engine 
speed in Modes 2-5 or Modes 7-10; and during Mode 1 or Mode 6 fixed 
throttle operation may be used to determine the 100 percent torque 
value.

    In addition the current regulations at 40 CFR 90.410(b) state:

    For Phase 2 Class I, I-B, and II engines equipped with an engine 
speed governor, during Mode 1 or Mode 6 hold both the specified 
speed and load within  five percent of point, during 
Modes 2-3, or Modes 7-8 hold the specified load with  
five percent of point, during Modes 4-5 or Modes 9-10, hold the 
specified load within the larger range provided by  0.27 
Nm ( 0.2 lb-ft), or  ten (10) percent of 
point, and during the idle mode hold the specified speed within 
 ten percent of the manufacturer's specified idle engine 
speed (see Table 1 in Appendix A of this subpart for a description 
of test Modes).

    Manufacturers have raised some questions about the interpretation 
of these provisions. Our intent is that the current requirements 
specify that testing be conducted as follows:
     Full-load testing (Mode 1) occurs at wide-open throttle to 
maintain engines at rated speed, which is defined as the speed at which 
the engine's maximum power occurs (as declared by the manufacturer).
     Idle testing (Mode 6) occurs at the manufacturer's 
specified idle speed with a maximum load of five percent of maximum 
torque. The regulation allows adjustment to control speeds that are 
different than would be maintained by the installed governor.
     The installed governor must be used to control engine 
speed for testing at all modes with torque values between idle and 
full-load modes. The regulation allows adjustments for nominal speed 
settings that are different than would be maintained by the installed 
governor without modification.
    We are proposing adjustments to the current regulatory requirements 
in 40 CFR part 90 (see Sec.  1054.505). Since each of these proposed 
adjustments may have some effect on measured emission levels, we 
believe it is appropriate to implement these changes concurrent with 
the Phase 3 standards. To the extent the proposed adjustments apply to 
handheld engines, we believe it is appropriate to apply the changes for 
new testing with 2012 and later model year engines for the reasons 
described above for adopting the test procedures in part 1065.
    First, we are proposing to require engine speed during the idle 
mode to be controlled by the engine's installed speed governor. We 
believe there is no testing limitation that would call for engine 
operation at idle to depart from the engine's governed speed. Allowing 
manufacturers to arbitrarily declare an idle speed only allows 
manufacturers to select an idle speed that gives them an advantage in 
achieving lower measured emission results, but not in a way that 
corresponds to in-use emission control. We are also aware that some 
production engines have a user-selectable control for selecting high-
speed or low-speed idle (commonly identified as ``rabbit/turtle'' 
settings). We believe this parameter adjustment may have a significant 
effect on emissions that should be captured in the certification test 
procedure. As a result, we are proposing a requirement that 
manufacturers conduct testing with user-selectable controls set to keep 
the

[[Page 28148]]

engine operating at low-speed idle if any production engines in the 
engine family have such an option.
    Second, we are proposing an option in which manufacturers would 
test their nonhandheld engines using a ramped-modal version of the 
specified duty cycle, as described in Section IX. We expect this 
testing to be equivalent to the modal testing described above but would 
have advantages for streamlining test efforts by allowing for a single 
result for the full cycle instead of relying on a calculation from 
separate modal results. Under the proposal we would allow manufacturers 
the option to select this type of testing. EPA's testing would 
generally involve ramped-modal testing only if the engine manufacturer 
selected this option for certification.
    Third, the part 90 regulations currently specify two duty cycles 
for nonhandheld engines: (1) Testing at rated speed; and (2) testing at 
85 percent of rated speed. The regulations direct manufacturers simply 
to select the most appropriate cycle and declare the rated speed for 
their engines. We believe it is appropriate to make this more objective 
by stating that rated speed is 3600 rpm and intermediate speed is 3060 
rpm, unless the manufacturer demonstrates that a different speed better 
represents the in-use operation for their engines. This is consistent 
with the most common in-use settings and most manufacturers' current 
practice.
    In addition, we are proposing regulatory provisions to clarify how 
nonhandheld engines are operated to follow the prescribed duty cycle. 
As described in part 90, we are proposing to require that the engines 
operate ungoverned at wide-open throttle for the full-power mode. This 
test mode is used to denormalize the rest of the duty cycle. Testing at 
other modes occurs with the governor controlling engine speed. Before 
each test mode, manufacturers may adjust the governor to target the 
same nominal speed used for the full-power mode, with a tolerance 
limiting the variation in engine speed at each mode. Alternatively, 
testing may be done by letting the installed governor control engine 
speed, in which case only the torque value would need to be controlled 
within an established range.
    A different duty cycle applies to handheld engines, which are 
generally not equipped with governors to control engine speed. The 
current regulations allow manufacturers to name their operating speed 
for testing at each of the test modes. We are proposing to continue the 
allowance for manufacturers to select an appropriate engine speed for 
idle operation. However, we are concerned that this approach allows 
manufacturers too much discretion for selecting a rated speed for high-
load testing. Manufacturers are encouraged to select a speed that best 
represents in-use operation for the engine family, but there is no 
requirement to prevent a manufacturer from selecting a rated speed that 
results in lower emissions, independent of the speeds at which in-use 
engines operate. We are proposing to specify that manufacturers select 
a value for rated speed that matches the most common speed for full-
load operation within the engine family. Engine manufacturers generally 
also make their own equipment, so this information should be readily 
available. We would expect manufacturers to identify the range of 
equipment models covered by a given engine family, identify the in-use 
operating speeds for those models, and select the full-load speed 
applicable for the greatest number of projected unit sales. We further 
propose to require manufacturers to describe in their application for 
certification how they selected the value for rated speed.
(3) Test Fuel
    We are proposing to require Phase 3 testing with a standard test 
fuel consistent with the requirements under 40 CFR part 90 (see 40 CFR 
part 1065, subpart H). In particular, we do not believe it is 
appropriate to create a flexibility to allow for testing using 
oxygenated fuel since this could affect an engine's air-fuel ratio, 
which in turn could affect the engine's combustion and emission 
characteristics. However, we understand that engine manufacturers may 
have emission data from some model years before the Phase 3 standards 
take effect. We would allow for continued use of this pre-existing data 
as long as it is appropriate to use carryover data for demonstrating 
compliance with current standards.
    Ethanol is commonly blended into in-use gasoline and is anticipated 
to be more widely used in the future. However, we are not proposing a 
test fuel containing ethanol for two reasons. First, the technical 
feasibility of this rule is based on certification gasoline. If an 
ethanol fuel blend were used as the certification fuel, the standards 
would need to be adjusted to account for the effects of this fuel on 
emissions. Second, manufacturers may not use ethanol blends to certify 
Small SI engines in California. The use of an ethanol blend would 
require manufacturers to test their engines separately for the 
California and Federal testing.
    The test fuel specifications apply to all testing. However, we may 
be able to allow for testing with oxygenated fuel for production-line 
testing if manufacturers first establish the appropriate correction to 
account for the fuel's effect on emissions. We request comment on an 
appropriate approach that would allow for production-line testing with 
oxygenated fuel.
    We are similarly proposing test fuel specifications for liquefied 
petroleum gas (LPG) and natural gas. Since natural gas has a very high 
methane content and methane is generally nonreactive in the atmosphere, 
we are proposing to apply the same emission standards for natural gas 
engines but not count methane emissions toward the total hydrocarbon 
measurement.

E. Certification and Compliance Provisions for Small SI Engines and 
Equipment

(1) Deterioration Factors
    As part of the certification process, manufacturers generate 
deterioration factors to demonstrate that their engines meet emission 
standards over the full useful life. We are proposing some changes from 
the procedures currently included in part 90 (see Sec.  1054.240 and 
Sec.  1054.245). Much of the basis for these changes comes from the 
experience gained in testing many different engines in preparation for 
this proposal. First, we are proposing to discontinue bench aging of 
emission components. Testing has shown that operating and testing the 
complete engine is necessary to get accurate deterioration factors. 
Second, we are proposing to allow for assigned deterioration factors 
for a limited number of small-volume nonhandheld engine families. 
Manufacturers could use assigned deterioration factors for multiple 
small-volume nonhandheld engine families as long as the total 
production for all of the nonhandheld engine families for which the 
manufacturer is using assigned deterioration factors is estimated at 
the time of certification to be no more than 10,000 units per year. 
Third, we are proposing to allow for assigned deterioration factors for 
all engines produced by small-volume nonhandheld engine manufacturers.
    For the HC+NOX standard, we propose to specify that 
manufacturers use a single deterioration factor for the sum of HC and 
NOX emissions. However, if manufacturers get approval to 
establish a deterioration factor on an engine that is tested with 
service accumulation representing less than the full useful life for 
any reason, we would require separate deterioration factors for

[[Page 28149]]

HC and NOX emissions. The advantage of a combined 
deterioration factor is that it can account for an improvement in 
emission levels with aging. However, for engines that have service 
accumulation representing less than the full useful life, we believe it 
is not appropriate to extrapolate measured values indicating that 
emission levels for a particular pollutant will decrease. This is the 
same approach we adopted for recreational vehicles.
    EPA is not proposing the values for the assigned deterioration 
factors for small-volume nonhandheld engine manufacturers in this 
proposal. In an effort to develop deterioration factors that are 
appropriate for Small SI engines, we plan to evaluate certification 
data from Phase 3 engines certified early with EPA and from engines 
certified under California ARB's Tier 3 standards (which begin in 2007 
and 2008). Because we are not proposing new exhaust standards for 
handheld engines, the assigned deterioration factor provisions adopted 
for Phase 2 handheld engines are being retained.
    Although we are not proposing new exhaust standards for handheld 
engines, handheld engine manufacturers noted that California ARB has 
approved certain durability cycles for accumulating hours on engines 
for the purpose of demonstrating emissions durability. The durability 
cycles approved by California ARB vary from a 30-second cycle for 
chainsaws to a 20-minute cycle for blowers, with 85 percent of the time 
operated at wide open throttle and 15 percent of the time operated at 
idle. Engine manufacturers can run the durability cycles over and over 
until they accumulate the hours of operation equivalent to the useful 
life of the engine family. Our current regulations state that ``service 
accumulation is to be performed in a manner using good judgment to 
ensure that emissions are representative of production engines.'' While 
we are not proposing to change the regulatory language regarding 
service accumulation, we believe the California ARB-approved durability 
cycles are appropriate and acceptable to EPA for accumulating hours on 
handheld engines for demonstrating emissions durability.
    Manufacturers have pointed out that they are developing a testing 
protocol that would allow manufacturers to develop deterioration 
factors for catalysts through a bench-aging procedure. A fundamental 
factor in evaluating the appropriateness of any bench-aging procedure 
is the extent to which it simulates representative exhaust gas 
composition and other in-use operating parameters. We request comment 
on any appropriate procedures, or limitations on the use of such 
procedures, for certifying Small SI engines.
(2) Delegated Final Assembly
    The current practice of attaching exhaust systems to engines 
varies. Class I engines are typically designed and produced by the 
engine manufacturer with complete emission control systems. Equipment 
manufacturers generally buy these engines and install them in their 
equipment, adjusting equipment designs if necessary to accommodate the 
mufflers and the rest of the exhaust system from the engine 
manufacturer.
    Engine manufacturers generally produce Class II engines without 
exhaust systems, relying instead on installation instructions to ensure 
that equipment manufacturers get mufflers that fall within a specified 
range of backpressures that is appropriate for a given engine model. 
Equipment manufacturers are free to work with muffler manufacturers to 
design mufflers that fit into the space available for a given equipment 
model, paying attention to the need to stay within the design 
specifications from the engine manufacturers. A similar situation 
applies for air filters, where equipment manufacturers in some cases 
work with component manufacturers to use air filters that are tailored 
to the individual equipment model while staying within the design 
specifications defined by the engine manufacturer.
    The existing regulations require that certified engines be in their 
certified configuration when they are introduced into commerce. We 
therefore need special provisions to address the possibility that 
engines will need to be produced and shipped without exhaust systems or 
air intake systems that are part of the certified configuration. We 
have adopted such provisions for heavy-duty highway engines and for 
other nonroad engines in 40 CFR 85.1713 and 40 CFR 1068.260, 
respectively. These provisions generally require that engine 
manufacturers establish a contractual arrangement with equipment 
manufacturers and take additional steps to ensure that engines are in 
their certified configuration before reaching the ultimate purchaser.
    We are proposing to apply delegated-assembly provisions for 
nonhandheld engines that are similar to those adopted for heavy-duty 
highway engines, with a variety of adjustments to address the unique 
situation for Small SI engines (see Sec.  1054.610). This would require 
that engine manufacturers apply for certification in the normal way, 
identifying all the engine parts that make up the engine configurations 
covered by the certification. Equipment manufacturers would be able to 
work with muffler manufacturers to get mufflers with installed 
catalysts as specified in the engine manufacturer's application for 
certification. If equipment manufacturers would need a muffler or 
catalyst that is not covered by the engine manufacturer's 
certification, the engine manufacturer would need to amend the 
application for certification. This may require new testing if the data 
from the original emission-data engine are not appropriate for showing 
that the new configuration will meet emission standards, as described 
in Sec.  1054.225. (Alternatively, the equipment manufacturer may take 
on the responsibility for certifying the new configuration, as 
described in Sec.  1054.612.) Engine manufacturers would also identify 
in the application for certification their plans to sell engines 
without emission-related components. We are proposing several 
provisions to ensure that engines will eventually be in their certified 
configuration. For example, engine manufacturers would establish 
contracts with affected equipment manufacturers, include installation 
instructions to make clear how engine assembly should be completed, 
keep records of the number of engines produced under these provisions, 
and obtain annual affidavits from affected equipment manufacturers to 
confirm that they are installing the proper emission-related components 
on the engines and that they have ordered a number of components that 
corresponds to the number of engines involved.
    While the delegated-assembly provisions are designed for direct 
shipment of engines from engine manufacturers to equipment 
manufacturers, we are aware that distributors play an important role in 
providing engines to large numbers of equipment manufacturers. We are 
proposing that these provisions apply to distributors in one of two 
ways. First, engine manufacturers may have an especially close working 
relationship with primary distributors. In such a case, the engine 
manufacturer would be able to establish a contractual arrangement 
allowing the distributor to act as the engine manufacturer's agent for 
all matters related to compliance with the delegated-assembly 
provisions. This would allow the distributor to make arrangements with 
equipment manufacturers to address design needs

[[Page 28150]]

and perform oversight functions. We would hold the engine manufacturer 
directly responsible if the distributor failed to meet the regulatory 
obligations that would otherwise apply to the engine manufacturer. 
Second, other distributors may receive shipment of engines without 
exhaust systems, but they would need to add any aftertreatment 
components before sending the engines on to equipment manufacturers. 
Engine manufacturers would treat these distributors as equipment 
manufacturers for the purposes of delegated assembly. Equipment 
manufacturers buying engines from such a distributor would not have the 
option of separately obtaining mufflers from muffler manufacturers. In 
both of these scenarios, the engine manufacturer continues to be 
responsible for the in-use compliance of all their engines.
    Engine manufacturers would need to affix a label to the engine to 
clarify that it needs certain emission-related components before it is 
in its certified configuration. This labeling information is important 
for alerting assembly personnel to select mufflers with installed 
catalysts; the label would also give in-house inspectors or others with 
responsibility for quality control a tool for confirming that all 
engines have been properly assembled and installed. Given the large 
numbers of engine and equipment models and the interchangeability of 
mufflers with and without catalysts, we believe proper labeling will 
reduce the possibility that engines will be misbuilt.
    This labeling may be done with any of three approaches. First, a 
temporary label may be applied such that it would not be removed 
without a deliberate action on the part of the equipment manufacturer. 
We believe it is not difficult to create a label that will stay on the 
engine until it is deliberately removed. Second, manufacturers may add 
the words ``delegated assembly'' to the engine's permanent emission 
control information label. Third, manufacturers may create a unique 
alphanumeric code to apply to the engine's permanent emission control 
information label. This code would be identified in the application for 
certification. Creating a unique code would not provide a clear enough 
communication to equipment manufacturers that they are responsible for 
bringing the engine into its certified configuration. Engine 
manufacturers taking this approach would therefore need to add features 
to the label to make this clear. For example, creating labels with a 
different color or shading would make it easy to identify that an 
engine needs to be properly assembled before it is in its certified 
configuration.
    Any of these labeling approaches would properly identify the 
engines as needing emission-related components from the equipment 
manufacturer. We have a remaining concern that the approaches involving 
permanent labels do not identify that an engine is not yet in its 
certified configuration. Since there is no change in the label to show 
the engine's status, we believe these approaches may not be as 
effective as the temporary labels in preventing misbuilt engines. We 
are also concerned that imported engines with manufacturer-specific 
codes will lead to confusion with Customs inspectors. With no 
standardized approach for identifying which engines do not need 
catalysts, there is a significant risk that engines will be held up 
while inspectors confirm their status. We request comment on the best 
way of requiring labeling information for these engines. For example, 
we request comment on adding a requirement for equipment manufacturers 
to add some identifying mark to the permanent label to show that the 
engine is in its certified configuration. We also request comment on 
replacing the provision allowing for a manufacturer-specific code to 
some standardized abbreviation for ``delegated assembly'' that would 
allow for unambiguous identification of the engine's status with a 
minimum burden in terms of requiring larger labels.
    In addition, engine manufacturers would need to perform or arrange 
for audits to verify that equipment manufacturers are properly 
assembling engines. Engine manufacturers may rely on third-party agents 
to perform auditing functions. Since the purpose of the audit is to 
verify that equipment manufacturers are properly assembling products, 
they may not perform audits on behalf of engine manufacturers. We are 
proposing to require that audits must involve at a minimum reviewing 
the equipment manufacturer's production records and procedures, 
inspecting the equipment manufacturer's production operations, or 
inspecting the final assembled products. Inspection of final assembled 
products may occur at any point in the product distribution system. For 
example, products may be inspected at the equipment manufacturer's 
assembly or storage facilities, at regional distribution centers, or at 
retail locations. The audit must also include confirmation that the 
number of aftertreatment devices shipped was sufficient for the number 
of engines involved. We would typically expect engine manufacturers to 
perform more than the minimum auditing steps identified above. For 
example, equipment manufacturers with low order volumes, an unclear 
history of compliance, or other characteristics that would cause some 
concern may prompt us to require a more extensive audit to ensure 
effective oversight in confirming that engines are always built 
properly. Moreover, in the early years of this program, engine 
manufacturers should consider nearly all participating equipment 
manufacturers to be unfamiliar with the regulatory requirements and the 
mechanics of meeting their responsibilities and obligations as 
contracted manufacturers of certified engines. Engine manufacturers 
would describe in the application for certification their plan for 
taking steps to ensure that all engines will be in their certified 
configuration when installed by the equipment manufacturer. EPA 
approval of a manufacturer's plan for delegated assembly would be 
handled as part of the overall certification process. We request 
comment on appropriate requirements related to specific auditing 
procedures that would be appropriate to address these concerns and to 
provide adequate assurance that engines are routinely assembled in 
their certified configuration.
    We are proposing that engine manufacturers annually audit twelve 
equipment manufacturers, or fewer if they are able to audit all 
participating equipment manufacturers on average once every four years. 
These audits would be divided over different equipment manufacturers 
based on the number of engines sold to each equipment manufacturer. We 
further propose that these auditing rates may be reduced after the 
first eight years, or after the engine manufacturer has audited all 
affected equipment manufacturers. This reduced auditing rate would be 
based on an expectation that all participating equipment manufacturers 
would be audited on average once every ten years.
    To facilitate auditing related to catalysts, we are proposing to 
require engine manufacturers to establish an alphanumeric designation 
to identify each unique catalyst design (including size, washcoat, 
precious metal loading, supplier, and any other appropriate factors) 
and instruct equipment manufacturers to use stamping or other means to 
permanently display this designation on the external surface of the 
exhaust system, making it readily visible as much as possible when the 
equipment is fully assembled, consistent with the objective of 
verifying the identity of the installed

[[Page 28151]]

catalyst. This designation could be the same as the code applied to the 
emission control information label as described above.
    We are proposing that all the same requirements apply for separate 
shipment related to air filters if they are part of an engine's 
certified configuration, except for the auditing. We would require 
auditing related to air filters only if engine manufacturers are 
already performing audits related to catalysts. We believe there is 
much less incentive or potential for problems with equipment 
manufacturers producing engines with noncompliant air filters so we 
believe a separate auditing requirement for air filters would be 
unnecessary.
    The draft regulation specifies that the exemption expires when the 
equipment manufacturer takes possession of the engine and the engine 
reaches the point of final equipment assembly. We would understand the 
point of final equipment assembly for purposes of delegated assembly 
for aftertreatment components to be the point at which the equipment 
manufacturer attaches a muffler to the engine. Engines observed in 
production or inventory assembled with improper mufflers would be 
considered to have been built contrary to the engine manufacturer's 
installation instructions. Catalysts are invariably designed as part of 
the muffler, so we would understand that there would be no reason to 
install a different muffler once a given muffler has been installed 
using normal production procedures. If equipment manufacturers sell 
equipment without following these instructions, they would be 
considered in violation of the prohibited acts (i.e., selling 
uncertified engines). If there is a problem with any given equipment 
manufacturer, we would hold the engine manufacturer responsible for 
those noncompliant engines and require the engine manufacturer to 
discontinue the practice of delegated assembly for that equipment 
manufacturer. We request comment on the need to more explicitly 
identify the meaning of the point of final equipment assembly in the 
regulations, as described above.
    We are aware that the proposed approach of allowing equipment 
manufacturers to make their own arrangements to order mufflers results 
in a situation in which the equipment manufacturer must spend time and 
money to fulfill their responsibilities under the regulations. This 
introduces a financial incentive to install mufflers with inferior 
catalysts, or to omit the catalyst altogether. To address this concern 
for heavy-duty highway engines, we adopted a requirement for engine 
manufacturers to confirm that a vehicle manufacturer has ordered the 
appropriate aftertreatment devices before they ship an engine. 
Equipment manufacturers' purchasing practices for Small SI engines, 
especially considering the order volumes, makes this approach 
impractical. We are instead proposing to require that engine 
manufacturers get written confirmation from each equipment manufacturer 
before an initial shipment of engines in a given model year for a given 
engine model. This confirmation would document the equipment 
manufacturer's understanding that they are using the appropriate 
aftertreatment components. The written confirmation would be due within 
30 days after shipping the engines and would be required before 
shipping any additional engines from that engine family to that 
equipment manufacturer.
    The shipping confirmation included in the rule for heavy-duty 
highway engines is a very substantial provision to address the fact 
that vehicle manufacturers would gain a competitive advantage by 
producing noncompliant products, and that engines in commerce would be 
labeled as if they were fully compliant even though they are not yet in 
their certified configuration. This is especially problematic when a 
muffler with no catalyst can easily be installed and can perform 
without indicating a problem. To address this concern for Small SI 
engines, we are including a requirement that equipment manufacturers 
include in their annual affidavits an accounting for the number of 
aftertreatment components they have ordered relative to the number of 
engines shipped without the catalysts that the mufflers would otherwise 
require.
    Production-line testing normally involves building production 
engines using normal assembly procedures. For engines shipped without 
catalysts under the delegated-assembly provisions, it is not normally 
possible to do this at the engine manufacturer's facility, where such 
testing would normally occur. To address this, we are proposing to 
specify that engine manufacturers must arrange to get a randomly 
selected catalyst that will be used with the engine. The catalyst may 
come from any point in the normal distribution from the aftertreatment 
component manufacturer to the equipment manufacturer. The catalyst may 
not come from the engine manufacturer's own inventory. Engine 
manufacturers would keep records to show how they randomly selected 
catalysts.
    As described above, we believe this is a very significant 
compliance issue since it allows manufacturers to introduce into 
commerce engines that are labeled as meeting current emission standards 
even though they are not in their certified configuration. This is 
especially true for Small SI engines where many high-volume products 
are handled by many different manufacturers such that the final 
assembly requires equipment manufacturers to properly install otherwise 
indistinguishable products to keep products in the certified 
configuration. Also, an equipment manufacturer may install multiple 
engine models in a single type of equipment, some of which may need 
catalyzed mufflers while others would use a conventional muffler. The 
appearance and function of such mufflers with and without catalysts 
would be virtually indistinguishable, which increases the likelihood of 
accidentally installing the wrong muffler.
    The provisions described above are intended to minimize the risks 
associated with this practice. However, this concern is heightened for 
companies that would use the delegated-assembly provisions to import 
noncompliant engines with the expectation that equipment manufacturers 
in the United States would add catalyzed mufflers as specified in the 
engine manufacturer's application for certification. This raises two 
potential problems. First, this practice could create a loophole in 
EPA's enforcement program that would allow for widespread importation 
of noncompliant engines, with the financial incentive for equipment 
manufacturers to complete assembly with noncompliant mufflers. Since 
all engines have mufflers, and since proper catalyst installation 
generally can be confirmed only with an emission test or a destructive 
inspection, it would be very difficult to find and correct any problems 
that might occur. Second, engine manufacturers outside the United 
States may be willing to take risks with noncompliant products based on 
their limited exposure to EPA enforcement. As described in Section VI.F 
we are considering bonding requirements for imported engines to ensure 
that we will be able to fully resolve compliance or enforcement issues 
with companies that have little or no presence or selling history in 
the United States. We would expect to specify an increased bond payment 
for importation of engines using the delegated-assembly provisions. 
Increasing the per-engine bond value by 20 percent corresponds roughly 
with the

[[Page 28152]]

value of catalyzed mufflers that would be required. We believe this 
would be an appropriate additional bond value to address the concerns 
for noncompliance from imported engines.
    While this section describes the compliance provisions we believe 
are necessary for addressing the practice of delegating assembly of 
emission-related components to equipment manufacturers, providing a 
broader view of the context for delegated assembly is also appropriate 
for understanding our concern regarding the duplicative aspects of 
delegated assembly with other provisions in this rulemaking. Recent 
evaluation of a wide range of equipment models powered by Small SI 
engines has led to several important observations. Many equipment 
models have mufflers installed away from all other components such that 
they have no space or packaging constraints. Other equipment models 
with mufflers that are installed inside a cage or compartment generally 
include substantial space around the muffler, which is necessary to 
isolate the muffler's high surface temperatures and radiant heat from 
operators and any heat-sensitive components. Another important 
observation was the striking uniformity of muffler geometries, even 
where equipment manufacturers obtained mufflers directly from muffler 
manufacturers. Most mufflers on Class II engines are cylindrical models 
with the size varying to correspond with the size of the engines. Other 
Class II engine models use a box-shaped muffler design, but these 
mufflers also exhibited little variation across models. These 
observations have fundamental implications for the regulatory 
provisions we are proposing for ensuring a smooth transition to the 
Phase 3 emission standards.
    For example, in situations that limit equipment manufacturers to 
standardized muffler configurations, they would at most need to make 
modest changes to their equipment to accommodate somewhat different 
muffler geometries. We have taken these equipment design changes into 
account with the Transition Program for Equipment Manufacturers 
described below. We are therefore concerned that the proposed 
provisions for delegated assembly and the Transition Program for 
Equipment Manufacturers may be duplicative in providing additional time 
and/or flexibilities for equipment manufacturers to redesign their 
equipment for accommodating engines that meet the Phase 3 standards. If 
this is the case, the proposed provisions for delegated assembly merely 
serve to preserve the current business arrangements for the different 
types of manufacturers. We request comment on the need for these 
delegated-assembly provisions in light of the Transition Program for 
Equipment Manufacturers. We also request comment on the appropriateness 
of adopting these delegated-assembly provisions for Class I engines 
since these engine manufacturers already install complete exhaust 
systems for the large majority of their engines. Finally, we request 
comment on the need to allow for the use of the more restrictive 
delegated-assembly provisions in Sec.  1068.260 in the event that we do 
not finalize the delegated-assembly provisions described above.
(3) Transition Program for Equipment Manufacturers
    Given the level of the proposed Phase 3 exhaust emission standards 
for Class II engines, we believe there may be situations where the use 
of a catalyzed muffler could require equipment manufacturers to modify 
their equipment. We are therefore proposing a set of provisions to 
provide equipment manufacturers with reasonable lead time for 
transition to the proposed standards. The proposed provisions are 
similar to the program we adopted for nonroad diesel engines (69 FR 
38958, June 29, 2004).
    Equipment manufacturers would not be obligated to use any of these 
provisions, but all equipment manufacturers that produce Class II 
equipment would be eligible to do so. We are also proposing that all 
entities under the control of a common entity would have to be 
considered together for the purposes of applying these allowances. 
Manufacturers would be eligible for the allowances described below only 
if they have primary responsibility for designing and manufacturing 
equipment, and if their manufacturing procedures include installing 
engines in the equipment.
(a) General Provisions
    Under the proposed approach, beginning in the 2011 model year and 
lasting through the 2014 model year, each equipment manufacturer may 
install Class II engines not certified to the proposed Phase 3 emission 
standards in a limited number of equipment applications produced for 
the U.S. market (see Sec.  1054.625). We refer to these here as ``flex 
engines.'' These flex engines would need to meet the Phase 2 standards. 
The maximum number of ``allowances'' each manufacturer could use would 
be based on 30 percent of an average year's production of Class II 
equipment. The number of ``allowances'' would be calculated by 
determining the average annual U.S.-directed production of equipment 
using Class II engines produced from January 1, 2007 through December 
31, 2009. Thirty percent of this average annual production level would 
be the total number of ``allowances'' under this transition program 
over four years. Manufacturers could use these allowances for their 
Class II equipment over four model years from 2011 through 2014, with 
the usage spread over these model years as determined by the equipment 
manufacturer. Equipment produced under these provisions could use 
engines that meet the Phase 2 emission standards instead of the Phase 3 
standards. If an equipment manufacturer newly enters the Class II 
equipment market during 2007, 2008 or 2009, the manufacturer would 
calculate its average annual production level based only on the years 
during which it actually produced Class II equipment. Equipment 
manufacturers newly entering the Class II equipment market after 2009 
would not receive any allowances under the transition program and would 
need to incorporate Phase 3 compliant engines into the Class II 
equipment beginning in 2011.
    Equipment using engines built before the effective date of the 
proposed Phase 3 standards would not count toward an equipment 
manufacturer's allowances. Equipment using engines that are exempted 
from the Phase 3 standards for any reason would also not count toward 
an equipment manufacturer's allowances. For example, we are proposing 
that small-volume engine manufacturers may continue to produce Phase 2 
engines for two model years after the Phase 3 standards apply. All 
engines subject to the Phase 3 standards, including those engines that 
are certified to FELs at higher levels than the standard, but for which 
an engine manufacturer uses exhaust ABT credits to demonstrate 
compliance, would count as Phase 3 complying engines and would not be 
included in an equipment manufacturer's count of allowances.
    The choice of the allowances based on 30 percent of one year's 
production is based on our best estimate of the degree of reasonable 
lead time needed by the largest equipment manufacturers to modify their 
equipment designs as needed to accommodate engines and exhaust systems 
that have changed as a result of more stringent emission standards. We 
believe the proposed level of allowances responds to the need for lead 
time to accommodate the workload related to redesigning

[[Page 28153]]

equipment models to incorporate catalyzed mufflers while ensuring a 
significant level of emission reductions in the early years of the 
proposed program.
    Equipment manufacturers may face similar challenges in 
transitioning to rotational-molded fuel tanks that meet the proposed 
permeation standards. We are therefore proposing to allow equipment 
manufacturers to use noncompliant rotational-molded fuel tanks with any 
equipment that is counted under the allowances described in this 
section which use engines meeting Phase 2 exhaust emission standards 
(see Sec.  1054.627). As part of this expanded rotational-molded fuel 
tank allowance, we are requiring that equipment manufacturers first use 
up any available credits or allowances generated from early compliance 
with the fuel tank permeation requirements (see Section VI.D.4).
    A similar concern applies for controlling running losses. As 
described in Section VI, technologies for controlling running losses 
may involve a significant degree of integration between engine and 
equipment designs. In particular, routing a vapor line from the fuel 
tank to the engine's intake system depends on engine modifications that 
would allow for this connection. As a result, we are proposing that any 
equipment using flex engines would not need to meet running loss 
standards.
(b) Coordination Between Engine and Equipment Manufacturers
    We are proposing two separate paths for complying with 
administrative requirements related to the proposed transition program, 
depending on how the engine manufacturer chooses to make flex engines 
available under the transition program. Engine manufacturers choosing 
to use the delegated-assembly provisions described above would be 
enabling equipment manufacturers to make the decision whether to 
complete the engine assembly in the Phase 3 configuration or to use a 
noncatalyzed muffler such that the engine would meet Phase 2 standards 
and would therefore need to be counted as a flex engine. If engine 
manufacturers do not use the delegated-assembly provisions, equipment 
manufacturers would need to depend on engine manufacturers to produce 
and ship flex engines that are already in a configuration meeting Phase 
2 standards and labeled accordingly. Each of these scenarios involves a 
different set of compliance provisions, which we describe below.
    (i) Compliance based on engine manufacturers. Engine manufacturers 
will in many cases produce complete engines. This would be the case if 
the engine does not require a catalyst or if the engine manufacturer 
chooses to design their own exhaust systems and ship complete engine 
assemblies to equipment manufacturers.
    Under this scenario, we propose to require that equipment 
manufacturers request a certain number of flex engines from the engine 
manufacturer. The proposed regulatory provisions would specifically 
allow engine manufacturers to continue to build and sell Phase 2 
engines needed to meet the market demand created by the transition 
program for equipment manufacturers provided they receive the written 
assurance from the equipment manufacturer that such engines are being 
procured for this purpose. We are proposing to require that engine 
manufacturers keep copies of the written assurance from equipment 
manufacturers for at least five years after the final year in which 
allowances are available.
    Engine manufacturers are currently required to label their 
certified engines with a variety of information. We are proposing that 
engine manufacturers producing complete flex engines under this program 
identify on the engine label that they are flex engines. In addition, 
equipment manufacturers would be required to apply an Equipment 
Flexibility Label to the engine or piece of equipment that identifies 
the equipment as using an engine produced under the Phase 3 transition 
program for equipment manufacturers. These proposed labeling 
requirements would allow EPA to easily identify flex engines and 
equipment, verify which equipment manufacturers are using these flex 
engines, and more easily monitor compliance with the transition 
provisions. Labeling of the equipment could also help U.S. Customs to 
quickly identify equipment being imported lawfully using the Transition 
Program for Equipment Manufacturers.
    While manufacturers would need to meet Phase 2 standards with their 
flex engines, they would not need to certify them for the current model 
year. We are proposing instead to apply the requirements in 40 CFR 
1068.260, which requires that manufacturers keep records showing that 
they meet emission standards without requiring submission of an 
application for certification. We request comment on these requirements 
and whether these engines should be certified annually along with the 
Phase 3 engines.
    (ii) Compliance based on equipment manufacturers. We are proposing 
to set up a different set of compliance provisions for engine 
manufacturers that ship the engine separately from the exhaust system. 
Under this scenario, as discussed above, the engine manufacturers must 
establish a relationship with the equipment manufacturers allowing the 
equipment manufacturer to install catalysts to complete engine assembly 
for compliance with Phase 3 standards.
    In this case, engine manufacturers would design and produce their 
Phase 3 engines and label them accordingly. The normal path for these 
engines covered by the delegated-assembly provisions would involve 
shipment of the engine without an exhaust system to the equipment 
manufacturer, the equipment manufacturer would then follow the engine 
manufacturer's instructions to add the exhaust system including the 
catalyst to bring the engine into a certified Phase 3 configuration. 
Under the proposed transition program, equipment manufacturers would 
choose for each of these engines to either follow the engine 
manufacturer's instructions to install a catalyst to make it compliant 
with Phase 3 standards or follow a different set of instructions to 
install a non-catalyzed muffler to make it compliant with Phase 2 
standards. Any such engines downgraded to Phase 2 standards would count 
toward the equipment manufacturer's total number of allowances under 
the transition program.
    To make this work, engine manufacturers would need to take certain 
steps to ensure overall compliance. First, engine manufacturers would 
need to include emission data in the application for certification 
showing that the engine would meet Phase 2 standards without any 
modification other than installing a non-catalyzed exhaust system. This 
may include a specified range of backpressures that equipment 
manufacturers would need to meet in procuring a non-catalyst muffler. 
If the Phase 3 engine without a catalyst would otherwise still be 
covered by the emission data from engines produced in earlier model 
years under the Phase 2 standards, manufacturers could rely on 
carryover emission data to make this showing. Second, the installation 
instructions we specify under the delegated-assembly provisions would 
need to describe the steps equipment manufacturers would need to take 
to make either Phase 3 engines or Phase 2 flex engines. Third, for 
engine families that generate positive emission credits under the 
exhaust ABT

[[Page 28154]]

program, engine manufacturers must decrease the number of ABT credits 
generated by the engine family by 10 percent. We believe the 10 percent 
decrease should provide an emission adjustment commensurate with the 
potential use of the equipment manufacturer flexibility provisions.
    Equipment manufacturers using allowances under these provisions 
would need to keep records that would allow EPA or engine manufacturers 
to confirm that equipment manufacturers followed appropriate procedures 
and produced an appropriate number of engines without catalysts. In 
addition, we are proposing to require that equipment manufacturers 
place a label on the engine as close as possible to the engine 
manufacturer's emission control information label to identify it as a 
flex engine. This could be the full label described above or it could 
be a simplified label that has only the equipment manufacturer's name 
and a simple statement that this is a flex engine. The location of this 
label is important since it effectively serves as an extension of the 
engine manufacturer's label, clarifying that the engine meets Phase 2 
standards, not the Phase 3 standards referenced on the original label. 
This avoids the problematic situation of changing or replacing labels, 
or requiring engine manufacturers to send different labels. We request 
comment on an approach in which we would require the full label for 
equipment manufacturers to be placed on the engine adjacent to the 
engine manufacturer's label to prevent confusion and the risks 
associated with multiple labels.
    Engine manufacturers might choose to produce Phase 3 engines before 
the 2011 model year and set up arrangements for separate shipment of 
catalyzed mufflers as described in Section V.E.2. We would expect any 
engine manufacturers producing these early Phase 3 engines to continue 
production of comparable engine models that meet Phase 2 standards 
rather than forcing all equipment manufacturers to accommodate the new 
engine design early. We believe it would not be appropriate for 
equipment manufacturers to buy Phase 3 engines in 2010 or earlier model 
years and downgrade them to meet Phase 2 emission standards as 
described above. We are therefore proposing to allow the downgrading of 
Phase 3 engines only for 2011 and later model years.
    Because equipment manufacturers in many cases depend on engine 
manufacturers to supply certified engines in time to produce complying 
equipment, we are also proposing a hardship provision for all equipment 
manufacturers (see Sec.  1068.255). An equipment manufacturer would be 
required to use all of its allowances under the transition program 
described above before being eligible to use this hardship. See Section 
VIII.C.9 for further discussion of this proposed hardship provision for 
equipment manufacturers.
    As described in Section V.E.2, we are concerned that the Transition 
Program for Equipment Manufacturers and the provisions related to 
delegated assembly may be redundant approaches to address the need to 
design equipment models to accommodate upgraded engines. The transition 
program is intended to give equipment manufacturers four years to make 
the design changes needed to reach a point of being able to accommodate 
low-emission Phase 3 engines, even for the most challenging equipment 
models. If equipment manufacturers are able to continue to 
independently source their exhaust systems based on the catalyst 
specifications determined by the engine manufacturer, it is not clear 
that allowances for additional lead time would be needed. We request 
comment on the relative advantages of these two approaches and, more 
specifically, which approach we should adopt in the final rule to 
address equipment manufacturers' needs for designing and producing 
equipment with Phase 3 engines. We request comment on an alternative 
approach of relying on the delegated-assembly provisions in Sec.  
10654.610 and the equipment-manufacturer hardship provisions in Sec.  
1068.255. This combination of tools would still allow for substantial 
flexibility in helping equipment manufacturers transition to Phase 3 
engines. The hardship provisions of Sec.  1068.255 were an important 
element of the successful transition to new emission standards for 
Large SI engines.
    (iii) Reporting and recordkeeping requirements. Equipment 
manufacturers choosing to participate in the transition program would 
be required to keep records of the U.S-directed production volumes of 
Class II equipment in 2007 through 2009 broken down by equipment model 
and calendar year. Equipment manufacturers would also need to keep 
records of the number of flex engines they use under this program.
    We are also proposing some notification requirements for equipment 
manufacturers. Under this proposal, equipment manufacturers wishing to 
participate in the transition provisions would need to notify EPA by 
June 30, 2010 that they plan to participate. They must submit 
information on production of Class II equipment over the three-year 
period from 2007 through 2009, calculate the number of allowances 
available, and provide basic business information about the company. 
For example, we would want to know the names of related companies 
operating under the same parent company that would be required to count 
engines together under this program. This early notification will not 
be a significant burden to the equipment manufacturer and will greatly 
enhance our ability to ensure compliance. Indeed, equipment 
manufacturers would need to have the information required in the 
notification to know how to use the allowances.
    We are proposing an ongoing reporting requirement for equipment 
manufacturers participating in the Phase 3 transition program. Under 
this proposal, participating equipment manufacturers would be required 
to submit an annual report to EPA that shows its annual number of 
equipment produced with flex engines under the transition provisions in 
the previous year. Each report would include a cumulative count of the 
number of equipment produced with flex engines for all years. To ease 
the reporting burden on equipment manufacturers, EPA intends to work 
with the manufacturers to develop an electronic means for submitting 
information to EPA.
(c) Additional Allowances for Small- and Medium-Sized Companies
    We believe small-volume equipment manufacturers would need a 
greater degree of lead time than manufacturers that sell large volumes 
of equipment. The small companies are less likely to have access to 
prototype engines from engine manufacturers and generally have smaller 
engineering departments for making the necessary design changes. 
Allowances representing thirty percent of annual U.S.-directed 
production provide larger companies with substantial lead time to plan 
their product development for compliance but smaller companies may have 
a product mix that requires extensive work to redesign products in a 
short amount of time. We are therefore proposing to specify that small-
volume equipment manufacturers may use this same transition program 
with allowances totaling 200 percent of the average annual U.S.-
directed production of equipment using Class II engines from 2007 
through 2009. For purposes of this program, a small-volume equipment 
manufacturer would be a manufacturer that produces fewer than 5,000 
pieces of nonhandheld equipment

[[Page 28155]]

per year subject to EPA regulations in each of the three years from 
2007 through 2009 or meets the SBA definition of small business 
equipment manufacturer (i.e., generally fewer than 500 employees for 
manufacturers of most types of equipment). These allowances would be 
spread over the same four-year period between 2011 and 2014. For 
example, a small-volume equipment manufacturer could potentially use 
Phase 2 engines on all their Class II equipment for two years or they 
might sell half their Class II equipment with Phase 2 engines for four 
years assuming production stayed constant over the four years.
    Medium-sized equipment manufacturers, i.e., companies that produce 
too much equipment to be considered a small-volume equipment 
manufacturer but produce fewer than 50,000 pieces of Class II 
equipment, may also face difficulties similar to that of small-volume 
equipment manufacturers. These companies may be like small-volume 
manufacturers if they have numerous product lines with varied 
approaches to installing engines and mufflers. Other companies may be 
more like bigger companies if they produce most of their equipment in a 
small number of high-volume models or have consistent designs related 
to engine and muffler installations. We are therefore proposing to 
create special provisions that would enable us to increase the number 
of transition allowances that are available to these medium-sized 
companies that have annual U.S.-directed production of Class II 
equipment of between 5,000 and 50,000 in each of the three years from 
2007 through 2009. To obtain allowances greater than 30 percent of 
average annual production, a medium-sized manufacturer would need to 
notify us by January 31, 2010 if they believe the standard allowances 
based on 30 percent of average annual production of Class II equipment 
would not provide adequate lead time starting in the 2011 model year. 
Additional allowances could be requested only if the equipment 
manufacturer can show they are on track to produce a number of 
equipment models representing at least half of their total U.S.-
directed production volume of Class II equipment in the 2011 model year 
compliant with all exhaust and evaporative emission standards. As part 
of their request, the equipment manufacturer would need to describe why 
more allowances are needed to accommodate anticipated changes in engine 
designs resulting from engine manufacturers' compliance with changing 
exhaust emission standards. The equipment manufacturer would also 
request a specific number of additional allowances needed with 
supporting information to show why that many allowances are needed. We 
may approve additional allowances up to 70 percent of the average 
annual U.S.-directed production of Class II equipment from 2007 through 
2009. If a medium-sized company were granted the full amount of 
additional allowances, they would have allowances equivalent to 100 
percent of the average annual production volume of Class II equipment.
    As noted above, the determination of whether a company is a small- 
or medium-sized manufacturer will be based primarily on production data 
over the 2007 through 2009 period submitted to EPA during 2010. After a 
company's status as a small- or medium-sized company has been 
established based on that data, EPA is proposing that manufactures 
would keep that status even if a company's production volume grows 
during the next few years, such that the company would no longer 
qualify as a small- or medium-sized company. EPA believes that 
equipment manufacturers need to know at the beginning of the transition 
program (i.e., 2011) how many allowances they will receive under the 
program. Changing a company's size determination during the program, 
which could affect the number of allowances available, would make it 
difficult for companies to plan and could lead to situations where a 
company is in violation of the provisions based on the use of 
allowances that were previously allowed. Likewise, if a company is 
purchased by another company or merges with another company after the 
determination of small- or medium-size status is established in 2010, 
EPA is proposing that the combined company could, at its option, keep 
the status for the individual portions of the combined company. If the 
combined company chooses to keep the individual designations, the 
combined company would submit the annual reports on the use of 
allowances broken down for each of the previously separate companies.
    (i) Requirements for foreign equipment manufacturers and importers. 
Under this proposal, only companies that manufacture equipment would 
qualify for the relief provided under the Phase 3 transition 
provisions. Foreign equipment manufacturers who comply with the 
compliance related provisions discussed below would enjoy the same 
transition provisions as domestic manufacturers. Foreign equipment 
manufacturers that do not comply with the compliance-related provisions 
discussed below would not receive allowances. Importers that do not 
manufacture equipment would not receive any transition relief directly, 
but could import equipment with a flex engine if it is covered by an 
allowance or transition provision associated with a foreign equipment 
manufacturer. This would allow transition provisions to be used by 
foreign equipment manufacturers in the same way as domestic equipment 
manufacturers, at the option of the foreign manufacturer, while 
avoiding the potential for importers to inappropriately use allowances. 
For the purposes of this proposal, a foreign equipment manufacturer 
would include any equipment manufacturer that produces equipment 
outside of the United States that is eventually sold in the United 
States.
    All foreign equipment manufacturers wishing to use the transition 
provisions would have to comply with all requirements discussed above. 
Along with the equipment manufacturer's notification described earlier, 
a foreign equipment manufacturer would have to comply with various 
compliance related provisions similar to those adopted for nonroad 
diesel engines (see Sec.  1054.626).\81\ As part of the notification, 
the foreign equipment manufacturer would have to:
---------------------------------------------------------------------------

    \81\ See, for example, 40 CFR 80.410 concerning provisions for 
foreign refiners with individual gasoline sulfur baselines.
---------------------------------------------------------------------------

     Agree to provide EPA with full, complete and immediate 
access to conduct inspections and audits;
     Name an agent in the District of Columbia for service;
     Agree that any enforcement action related to these 
provisions would be governed by the Clean Air Act;
     Submit to the substantive and procedural laws of the 
United States;
     Agree to additional jurisdictional provisions;
     Agree that the foreign equipment manufacturer will not 
seek to detain or to impose civil or criminal remedies against EPA 
inspectors or auditors for actions performed within the scope of EPA 
employment related to the provisions of this program;
     Agree that the foreign equipment manufacturer becomes 
subject to the full operation of the administrative and judicial 
enforcement powers and provisions of the United States without 
limitation based on sovereign immunity; and
     Submit all reports or other documents in the English 
language, or include an English language translation.

[[Page 28156]]

    In addition to these proposed requirements, we are proposing to 
require foreign equipment manufacturers that participate in the 
transition program to comply with a bond requirement for equipment 
imported into the United States. We describe a bond program below that 
we believe could be an important tool for ensuring that foreign 
equipment manufacturers are subject to the same level of enforcement as 
domestic equipment manufacturers. Specifically, we believe a bonding 
requirement for the foreign equipment manufacturer is an important 
enforcement tool for ensuring that EPA has the ability to collect any 
judgments assessed against a foreign equipment manufacturer for 
violations of these transition provisions. We request comments on all 
aspects of the specific program we describe here, but also on 
alternative measures that would achieve the same goal.
    Under a bond program, the participating foreign equipment 
manufacturer would have to maintain a bond in the proper amount that is 
payable to satisfy judgments that result from U.S. administrative or 
judicial enforcement actions for conduct in violation of the Clean Air 
Act. The foreign equipment manufacturer would generally obtain a bond 
in the proper amount from a third party surety agent that has been 
listed with the Department of the Treasury. As discussed in Sections 
V.E.6.c and V.E.6.d, EPA is proposing other bond requirements as well. 
An equipment manufacturer required to post a bond under any of these 
provisions would be required to obtain only one bond of the amount 
specified for those sections.
    In addition to the foreign equipment manufacturer requirements 
discussed above, EPA also proposes to require importers of equipment 
with flex engines from a complying foreign equipment manufacturer to 
comply with certain provisions. EPA believes these importer provisions 
are essential to EPA's ability to monitor compliance with the 
transition provisions. EPA proposes that the regulations would require 
each importer to notify EPA prior to their initial importation of 
equipment with flex engines. Importers would be required to submit 
their notification prior to the first calendar year in which they 
intend to import equipment with flex engines from a complying foreign 
equipment manufacturer. The importer's notification would need to 
include the following information:
     The name and address of importer (and any parent company);
     The name and address of the manufacturers of the equipment 
and engines the importer expects to import; and
     Number of units of equipment with flex engines the 
importer expects to import for each year broken down by equipment 
manufacturer.
    In addition, EPA is proposing that any importer electing to import 
to the United States equipment with flex engines from a complying 
foreign equipment manufacturer would have to submit annual reports to 
EPA. The annual report would include the number of units of equipment 
with flex engines the importer actually imported to the United States 
in the previous calendar year; and identify the equipment manufacturers 
and engine manufacturers whose equipment and engines were imported.
(4) Equipment Manufacturer Recertification
    Generally, it has been engine manufacturers who certify with EPA 
for exhaust emissions because the standards are engine-based. However, 
because the Phase 3 nonhandheld standards under consideration are 
expected to result in the use of catalysts, a number of equipment 
manufacturers, especially those that make low-volume models, believe it 
may be necessary to produce their own unique engine/muffler designs, 
but using the same catalyst substrate already used in a muffler 
certified by the engine manufacturer. In this situation, the engine 
would not be covered by the engine manufacturer's certificate, as the 
engine/muffler design is not within the specifications for the 
certified engine. The equipment manufacturer is therefore producing a 
new distinct engine which is not certified and needs to be certified 
with EPA. In order to allow the possibility of an equipment 
manufacturer certifying an engine/muffler design with EPA, we are 
proposing a simplified engine certification process for nonhandheld 
equipment manufacturers (see Sec.  1054.612). Under this simplified 
certification process, the nonhandheld equipment manufacturer would 
need to demonstrate that it is using the same catalyst substrate as the 
approved engine manufacturer's engine family, provide information on 
the differences between their engine/exhaust system and the engine/
exhaust system certified by the engine manufacturer, and explain why 
the emissions deterioration data generated by the engine manufacturer 
would be representative for the equipment manufacturer's configuration. 
The equipment manufacturer would need to perform low-hour emission 
testing on an engine equipped with their modified exhaust system and 
demonstrate that it meets the emission standards after applying the 
engine manufacturer's deterioration factors for the certified engine 
family. We would not require production-line testing for these engines. 
The equipment manufacturer would be responsible to meet all of the 
other requirements of an engine manufacturer under the regulations, 
including labeling, warranty, defect reporting, payment of 
certification fees, and other things. EPA requests comments on the 
usefulness of such a provision. EPA also requests comments on whether 
such a simplified certification provision should expire after a period 
of time, for example, after five years. If the provision were to 
expire, an equipment manufacturer could continue to certify, but they 
would have to follow the general certification regulations at that 
point.
(5) Special Provisions Related to Altitude
    As described in Section V.C.1, we allow manufacturers of handheld 
and nonhandheld engines to comply with emission standards at high 
altitudes using an altitude kit. We are proposing to keep the 
provisions that already apply in part 90 related to descriptions of 
these altitude kits in the application for certification. This would 
include a description of how engines comply with emission standards at 
varying atmospheric pressures, a description of the altitude kits, and 
the associated part numbers. The manufacturer would also identify the 
altitude range for which it expects proper engine performance and 
emission control with and without the altitude kit, state that engines 
will comply with applicable emission standards throughout the useful 
life with the altitude kit installed according to instructions, and 
include any supporting information. Finally, manufacturers would need 
to describe a plan for making information and parts available such that 
altitude kits would reasonably be expected to be widely used in high-
altitude areas. For nonhandheld engines, this would involve all 
counties with elevations substantially above 4,000 feet (see Appendix 
III to part 1054). This includes all U.S. counties where 75 percent of 
the land mass and 75 percent of the population are above 4,000 feet 
(see 45 FR 5988, January 24, 1980 and 45 FR 14079, March 4, 1980). For 
handheld engines, this would involve all areas at an elevation at or 
above that which they identify in their application

[[Page 28157]]

for certification for needing an altitude kit to meet emission 
standards.
    We are also proposing to require information related to altitude 
kits to be on the emission control information label, unless space 
limitations prevent it. We believe it is important for operators to 
know that engines may need to be modified to run properly at high 
elevations.
    We request comment on all aspects of this approach for compliance 
at high-altitude conditions. (See Sec. Sec.  1054.115, 1054.135, 
1054.205, and 1054.655.)
(6) Special Provisions for Compliance Assurance
    EPA's experiences in recent years have highlighted the need for 
more effective tools for preventing the introduction into commerce of 
noncompliant engines. These include noncompliant engines sold without 
engine labels or with counterfeit engine labels. We are proposing the 
special provisions in the following sections to help us address these 
problems.
(a) Importation Form
    Importation of engines is regulated both by EPA and U.S. Customs. 
The current regulations for U.S. Customs specify that anyone importing 
a nonroad engine (or equipment containing a nonroad engine) must 
complete a declaration form before importation. EPA has created 
Declaration Form 3520-21 for this purpose. Customs requires this in 
many cases, but there are times when they allow engines to be imported 
without the proper form. It would be an important advantage for EPA's 
own compliance efforts to be able to enforce this requirement. We are 
therefore proposing to modify part 90 to mirror the existing Customs 
requirement (and the EPA requirement in Sec.  1068.301) for importers 
to complete and retain the declaration form before importing engines 
(see Sec.  90.601). This would facilitate a more straightforward 
processing of cases in which noncompliant products are brought to a 
U.S. port for importation because currently no requirement exists for 
measuring emissions or otherwise proving that engines are noncompliant 
at the port facility. Since this is already a federal requirement, we 
are proposing to make this effective immediately with the final rule.
(b) Assurance of Warranty Coverage
    Manufacturers of Small SI engines subject to the standards are 
required to provide an emission-related warranty so owners are able to 
have repairs done at no expense for emission-related defects during an 
initial warranty period. Established companies are able to do this with 
a network of authorized repair facilities that can access replacement 
parts and properly correct any defects. In contrast, we are aware that 
some manufacturers are selling certified engines in the United States 
without any such network for processing warranty claims. As such, 
owners who find that their engines have an emission-related defect are 
unable to properly file a warranty claim or get repairs that should be 
covered by the warranty. In effect, this allows companies to certify 
their engines and agree to provide warranty coverage without ever 
paying for legitimate repairs that should be covered by the warranty. 
We are therefore proposing to require that manufacturers demonstrate 
several things before we will approve certification for their engines 
(see Sec.  90.1103 and Sec.  1054.120). The following provisions would 
apply to manufacturers who certify engines, and would include importers 
who certify engines. First, we are proposing to require manufacturers 
to provide and monitor a toll-free telephone number and an e-mail 
address for owners to receive information about how to make a warranty 
claim and how to make arrangements for authorized repairs. Second, we 
are proposing to require manufacturers to provide a source of 
replacement parts within the United States. For imported parts, this 
would require at least one distributor within the United States.
    Finally, we are proposing to require manufacturers to have a 
network of authorized repair facilities or to take one of several 
alternate approaches to ensure that owners will be able to get free 
repair work done under warranty. If warranty-related repairs are 
limited to authorized repair facilities, we are proposing to require 
that manufacturers have enough such facilities that owners do not have 
to go more than 100 miles for repairs. An exception would be made for 
remote areas where we would allow for approval of greater travel 
distances for getting repairs as long as the longer travel distance 
applies to no more than 10 percent of affected owners. For small 
businesses, start-up companies, or importers, it may not be realistic 
to maintain a national repair network. We are proposing a variety of 
alternative methods for such companies to meet their warranty 
obligations. Manufacturers would be able to meet warranty obligations 
by informing owners that free shipping to and from an authorized 
service center is available, a service technician will be provided to 
come to the owner to make the warranty repair, or repair costs at a 
local nonauthorized service center will be reimbursed.
    We believe these proposed requirements are both necessary and 
effective for ensuring proper warranty coverage for all owners. At the 
same time, we are proposing a flexible approach that allows companies 
to choose from widely varying alternatives to provide warranty service. 
We therefore believe these proposed requirements are readily achievable 
for any company. We are therefore proposing to implement these 
requirements starting with the 2009 model year. This should allow time 
for the administrative steps necessary to arrange for any of the 
allowable compliance options described above. We request comment on 
these provisions to ensure proper warranty coverage. We also request 
comment on alternative means of demonstrating effective warranty 
coverage comparable to that described above.
(c) Bond Requirements Related to Enforcement and Compliance Assurance
    Certification initially involves a variety of requirements to 
demonstrate that engines and equipment are designed to meet applicable 
emission standards. After certification is complete, however, several 
important obligations apply to the certifying manufacturer or importer. 
For example, we require ongoing testing of production engines, warranty 
coverage for emission-related defects, reporting of recurring defects, 
and payment of penalties if there is a violation. For companies 
operating within the United States, we are generally able to take steps 
to communicate clearly and insist on compliance with applicable 
regulations. For companies without staff or assets in the United 
States, this is not the case. Accordingly, we have limited ability to 
enforce these requirements or recover any appropriate penalties, which 
increases the risk of environmental problems as well as problems for 
owners. This creates the potential for a company to gain a competitive 
advantage if they do not operate in the United States by avoiding some 
of the costs of complying with EPA regulations.
    We request comment on a requirement for importers of certified 
engines and equipment to post a bond to cover any potential compliance 
or enforcement actions under the Clean Air Act. Importers would be 
exempt from the bond requirement if they were able to sufficiently 
demonstrate an assurance that they would meet any compliance-or 
enforcement-related obligations. We

[[Page 28158]]

would consider adopting provisions to waive the bonding requirement 
based on a variety of specific criteria. For example, importers might 
show that they have physical assets in the United States with a value 
equal to the retail value of the engines that they will import during 
the model year (or equipment that they will import during the model 
year if they import equipment). Also, we may be able to establish an 
objective measure for a company to demonstrate long-term compliance 
with applicable regulations. Another alternative might involve a 
showing that an importer has been certified under certain industry 
standards for production quality and regulatory compliance. Finally, we 
may be able to rely on a company's commitment to periodically perform 
voluntary in-use testing in the United States to show that engines 
comply with emission standards. In addition to these specific criteria, 
we would consider adopting a provision that allows an individual 
importer to request a waiver from bonding requirements based on that 
importer's particular circumstances. If we adopt a bonding requirement, 
we would expect to apply that starting with the 2009 model year.
    We would expect the per-engine bond amount to be $25 for handheld 
engines and Class I engines. Class II engines cover a much wider range 
of applications, so we further differentiate the bond for those 
engines. The proposed per-engine bond amounts for Class II engines 
would be $50 for engines between 225 and 740 cc, $100 for engines 
between 740 and 1,000 cc, and $200 for engines above 1,000 cc. These 
values are generally scaled to be approximately 10 to 15 percent of the 
retail value. In the case of handheld engines, this is based on the 
retail value of equipment with installed engines, since these products 
are generally traded that way. Class II engines are very often sold as 
loose engines to equipment manufacturers, so the corresponding per-
engine bond values are based on the retail value of the engine alone. 
This approach is similar to the bond requirements that apply for 
nonroad diesel engines (see Sec.  1039.626).
    The total bond amount would be based on the value of imported 
products over a one-year period. If an importer's bond would be used to 
satisfy a judgment, the importer would then be required to increase the 
amount of the bond within 90 days of the date the bond is used to cover 
the amount that was used. Also, we would require the bond to remain in 
place for five years after the importer no longer imports Small SI 
engines.
    (d) Bond Requirements Related to Recall
    Recall is another potential compliance obligation. The Clean Air 
Act specifies that EPA must require the manufacturer to conduct a 
recall if EPA determines that a substantial number of engines do not 
conform to the regulations. We have experience with companies that have 
faced compliance-related problems where it was clear that they did not 
have the resources to conduct a recall if that were necessary. Such 
companies benefit from certification without bearing the full range of 
associated obligations. We believe it is appropriate again to add a 
requirement to post a bond to ensure that a company can meet their 
recall obligations. The concern for being able to meet these 
obligations applies similarly to domestic and foreign manufacturers. 
The biggest indicator of a manufacturer's ability to make recall 
repairs relates to the presence of repair facilities in the United 
States. We are therefore proposing a bond requirement starting with the 
2009 model year for all manufacturers (including importers) that do not 
have assembly facilities in the United States that are available for 
processing recall repairs or a repair network in the United States 
capable of processing recall repairs (see Sec.  90.1007 and Sec.  
1054.685). Note that a single bond payment would be required for 
companies that must post bond for compliance-related obligations, as 
described above, in addition to the recall-related obligations. Such a 
repair network would need to involve at least 100 authorized repair 
facilities in the United States or at least one such facility for each 
5,000 engines sold in the United States, whichever is less. Companies 
not meeting these criteria would need to post a bond as described above 
for compliance assurance. We would allow these companies to arrange for 
any applicable recall repairs to be done at independent facilities.
(e) Restrictions Related to Naming Model Years
    New exhaust emission standards apply based on the date of engine 
assembly. We similarly require that equipment manufacturers use engines 
meeting emission standards in the same model year as equipment based on 
the equipment assembly date. For example, a manufacturer of a 2007 
model year piece of equipment must generally use a 2007 model year 
engine. However, we allow equipment manufacturers to deplete their 
normal inventories of engines from the previous model year as long as 
there is no stockpiling of those earlier engines. We also note that 
this restriction does not apply if emission standards are unchanged for 
the current model year. We have found many instances where companies 
will import new engines usually installed in equipment and claim that 
the engine was built before emission standards took effect, even if the 
start date for emission standards was several years earlier. We believe 
many of these engines were in fact built later than the named model 
year, but it is difficult to prove the date of manufacture, which then 
makes it difficult to properly enforce these requirements. Now that 
emission standards have been in place for Small SI engines for almost 
ten years, we believe it is appropriate to implement a provision that 
prevents new engines manufactured several years previously to be 
imported when more recent emission standards have been adopted. This 
would prevent companies from importing noncompliant products by 
inappropriately declaring a manufacture date that precedes the point at 
which the current standards started to apply. It would also put a time 
limit on our existing provisions that allow for normal inventory 
management to use the supply of engines from previous model years when 
there has been a change in standards.
    Starting January 1, 2009, we are proposing to specify that engines 
and equipment will be treated as having a model year at most one year 
earlier than the calendar year in which the importation occurs when 
there is a change in emission standards (see Sec.  90.616 and Sec.  
1054.695). For example, for new standards starting in the 2011 model 
year, beginning January 1, 2012, all imported new products would be 
considered 2011 or later model year engines and would need to comply 
with new 2011 standards, regardless of the actual build date of the 
engines or equipment. (Engines or equipment would be considered new 
unless the importer demonstrates that the engine or equipment had 
already been placed into service, as described below.) This would allow 
a minimum of twelve months for manufactured engines to be shipped to 
equipment manufacturers, installed in equipment and imported into the 
United States. This time interval would be substantially longer for 
most engines because the engine manufacturer's model year typically 
ends well before the end of the calendar year. Also, engines produced 
earlier in the model year would have that much more time to be shipped, 
installed, and imported.
    Manufacturers have expressed concern that the one-year limitation 
on imported products may be too short

[[Page 28159]]

since there are often delays related to shipping, inventory, and 
perhaps most significantly, unpredictable fluctuations in actual sales 
volumes. We do not believe it is appropriate to maintain long-term 
inventories of these products outside the United States for eventual 
importation when it is clear several years ahead that the new standards 
are scheduled to take effect. Companies may be able to import these 
products shortly after manufacturing and keep their inventories in a 
U.S. distribution network to avoid the situation of being unable to 
sell these products. We request comment on the need to extend the one-
year limit to account for the business dynamics. We also request 
comment on any narrower provisions that would allow for exceptions in 
certain circumstances. For example, should we consider allowing an 
additional year for products if manufacturers let us know ahead of time 
that they have certain numbers of engines or equipment that will not be 
imported in time, and they can demonstrate that they are not 
stockpiling or circumventing regulatory requirements?
    In years where the standards do not change, this proposed provision 
would have no practical effect because, for example, a 2004 model year 
engine meets the 2006 model year standards. We would treat such an 
engine as compliant based on its 2004 emission label, any emission 
credit calculations for the 2004 model year, and so on. These engines 
could therefore be imported anytime until the end of the calendar year 
in which new standards take effect. Also, because the changes do not 
affect importation until there is a change in the standards, we are 
proposing to implement these provisions starting with the Phase 3 
standards.
    We do not intend for these proposed provisions to delay the 
introduction of emission standards by one year. It is still a violation 
to produce an engine in the 2011 calendar year and call it a 2010 model 
year engine to avoid being subject to 2011 standards.
    Importation of equipment that is not new is handled differently. 
These products would not be required to be upgraded to meet new 
emission standards that started to apply after the engine and equipment 
were manufactured. However, to avoid the situation where companies 
simply declare that they are importing used equipment to avoid new 
standards, we are proposing to require that they provide clear and 
convincing evidence that such engines have been placed into service 
prior to importation. Such evidence would generally include documentary 
evidence of purchase and maintenance history and visible wear that is 
consistent with the reported manufacture date. Importing products for 
resale or importing more than one engine or piece of equipment at a 
time would generally call for closer evaluation to determine that this 
degree of evidence has been met.
(f) Import-Specific Information at Certification
    We are proposing to require additional information to improve our 
ability to oversee compliance related to imported engines (see Sec.  
90.107 and Sec.  1054.205). In the application for certification, we 
are proposing to require the following additional information: (1) The 
port or ports at which the manufacturer intends to import the engines, 
(2) the names and addresses of the agents the manufacturer has 
authorized to import the engines, and (3) the location of the test 
facilities in the United States where the manufacturer would test the 
engines if we select them for testing under a selective enforcement 
audit. This information should be readily available so we propose to 
require it for the 2009 model year. The current regulations in part 90 
do not include these specific requirements; however, we do specify 
already that we may select imported engines at a port of entry. In such 
a case, we would generally direct the manufacturer to do testing at a 
facility in the United States. The proposed provision allows the 
manufacturers to make these arrangements ahead of time rather than 
relying on EPA's selection of a test lab. The current regulations also 
state clearly in Sec.  90.119 that EPA may conduct testing at any 
facility to determine whether engines meet emission standards.
(g) Counterfeit Emission Labels
    We have observed that some importers attempt to import noncompliant 
products by creating an emission control information label that is an 
imitation of a valid label from another company. We are not proposing 
to require that certifying manufacturers take steps to prevent this, 
but we are proposing to include a provision that specifically allows 
manufacturers to add appropriate features to prevent counterfeit 
labels. This may include the engine's serial number, a hologram, or 
some other unique identifying feature. We propose to apply this 
provision immediately upon completion of the final rule since it is an 
allowance and not a requirement (see Sec.  1054.135).
(h) Partially Complete Engines
    As described in Section XI, we are proposing to clarify engine 
manufacturers' responsibilities for certification with respect to 
partially complete engines. While this is intended to establish a path 
for secondary engine manufacturers to get their engines from the 
original engine manufacturer, we are aware that this will also prevent 
manufacturers from selling partially complete engines as a strategy to 
circumvent certification requirements. If long blocks or engines 
without fuel systems are introduced into U.S. commerce, either the 
original manufacturer or the company completing engine assembly would 
need to hold a certificate for that engine.
(7) Using Certified Small SI Engines in Marine Applications
    Manufacturers have described situations in which Small SI engines 
are used in marine applications. As described in Section III.E.5, we 
are proposing to allow certified Small SI engines to be used in 
outboard or personal watercraft applications without certifying to the 
Marine SI emission standards in part 1045. We request comment on the 
appropriateness of this provision. In particular, we request comment on 
the extent to which the proposed provisions will address the unique 
situations that apply for swamp boats and other unusual configurations.
(8) Other Provisions
    We are also proposing a variety of changes in the provisions that 
make up the certification and compliance program. Most of these changes 
serve primarily to align with the regulations we have started to apply 
to other types of engines.
    The proposed warranty provisions are based on the requirements that 
already apply under 40 CFR part 90. We are proposing to add an 
administrative requirement to describe the provisions of the emission-
related warranty in the owners manual. We expect that many 
manufacturers already do this but believe it is appropriate to require 
this as a routine practice. (See Sec.  1054.120.) Testing new engines 
requires a period of engine operation to stabilize emission levels. The 
regulations specify two separate figures for break-in periods for 
purposes of certification testing. First, engines are generally 
operated long enough to stabilize emission levels. Second, we establish 
a limit on how much an engine may operate and still be considered a 
``low-hour'' engine. The results of testing with the low-hour engine 
are compared with a deteriorated

[[Page 28160]]

value after some degree of service accumulation to establish a 
deterioration factor. For Marine SI engines, we are proposing that the 
engine can be presumed to have stabilized emission levels after 12 
hours of engine operation, with a provision allowing approval for more 
time if needed, and we generally require that low-hour test engines 
have no more than 30 hours of engine operation. However, given the 
shorter useful life for many Small SI engines, this would not make for 
a meaningful process for establishing deterioration factors. For 
example, emission levels in Small SI engines may not stabilize before 
deterioration begins to affect emission levels, which would prevent the 
engine from ever truly having stabilized emission levels. Also, the 
low-hour emission test should occur early enough to adequately 
represent the deterioration over the engine's lifetime.
    We are proposing that Small SI engines with a useful life above 300 
hours can be presumed stable after 12 hours with low-hour testing 
generally occurring after no more than 24 hours of engine operation. 
For Small SI engines with useful life below 300 hours, we are proposing 
a combination of provisions to address this concern. First, we are 
proposing to allow manufacturers to establish a stabilization period 
that is less than 12 hours without showing that emission levels have 
fully stabilized (see Sec.  1054.501). Second, we propose to specify 
that low-hour testing must generally occur after no more than 15 hours 
of engine operation (see Sec.  1054.801). This allows some substantial 
time for break-in, stabilization, and running multiple tests, without 
approaching a significant fraction of the useful life. Third, we are 
proposing that manufacturers consistently test low-hour production-line 
engines (and emission-data engines in the case of carryover 
deterioration factors for certification) using the same degree of 
service accumulation to avoid inaccurate application of deterioration 
factors (see Sec.  1054.301).
    As described in Section VII.C, we are proposing to clarify the 
maintenance that manufacturers may perform during service accumulation 
as part of the certification process. The general approach is to allow 
any amount of maintenance that is not emission-related, but to allow 
emission-related maintenance only if it is a routine practice with in-
use engines. In most of our emission control programs we specify that 
80 percent of in-use engines should undergo a particular maintenance 
step before manufacturers can do that maintenance during service 
accumulation for certification testing. We are aware that Small SI 
engines are predominantly operated by homeowners with widely varying 
practices in servicing their lawn and garden equipment. As such, 
achieving a rate of 80 percent may be possible only for the most 
obvious maintenance steps. We are therefore proposing a more 
accommodating approach for Small SI engines. In particular, we are 
proposing to allow manufacturers to perform a maintenance step during 
certification based on information showing that 60 to 80 percent of in-
use engines get the specified maintenance at the recommended interval. 
We would approve the use of such maintenance based on the relative 
effect on performance and emissions. For example, we may allow 
scheduled fuel-injector replacement if survey data show this is done at 
the recommended interval for 65 percent of engines and performance 
degradation is shown to be roughly proportional to the degradation in 
emission control for engines that do not have their fuel injectors 
replaced.
    One maintenance step of particular interest will be replacement of 
air filters. In larger spark-ignition engines, we don't treat 
replacement of air filters as critical emission-related maintenance, 
largely because those engines have feedback controls to compensate for 
changes in varying pressure drop across the air filter. However, for 
Small SI engines varying air flow through the air filter has a direct 
effect on the engine's air-fuel ratio, which in turn directly affects 
the engine's emission rates for each of the regulated pollutants. 
Service accumulation generally takes place in laboratory conditions 
with far less debris, dust, or other ambient particles that would cause 
filter loading, so filter changes should be unnecessary to address this 
conventional concern. We are concerned that the greater affect is from 
fuel and oil that may deposit on the back side of the filter, 
especially from crankcase ventilation into the intake. If filters are 
changed before an emission test, this effect will go undetected. If 
filter changes are disallowed before emission testing, manufacturers 
would need to design their intake systems to prevent internal filter 
contamination. We request comment on the need for replacing air 
filters, the effect on emission levels, and on the extent of change 
that would be needed to prevent filter contamination from recirculating 
crankcase gases. We also request comment on the extent to which air 
filters are changed with in-use engines. While this is clearly done 
with many engines, it is not clear that the experience is common enough 
that we would consider it to be routine, and therefore appropriate for 
certification engines. Since the cost of equipment, the types of jobs 
performed, and the operating lifetime varies dramatically for Class I 
and Class II engines, commenters should distinguish between in-use 
maintenance that is done by engine class as much as possible. We may, 
for example, conclude that owners of riding mowers and other Class II 
equipment routinely replace air filters to keep their equipment 
operating properly, while owners of walk-behind mowers and other Class 
I equipment are more likely to treat their equipment as a disposable 
product and therefore not replace the air filter.
    We are proposing to define criteria for establishing engine 
families that are very similar to what is currently specified in 40 CFR 
part 90. We are proposing to require that engines with turbochargers be 
in a different family than naturally aspirated engines since that would 
be likely to substantially change the engine's emission 
characteristics. Very few if any Small SI engines are turbocharged 
today so this change will not be disruptive. We are also specifying 
that engines must have the same number, arrangement, and approximate 
bore diameter of cylinders. This will help us avoid the situation where 
manufacturers argue that engines with substantially different engine 
blocks should be in the same engine family. We would expect to 
implement this provision consistent with the approach adopted by 
California ARB in which they limit engine families to include no more 
than 15 percent variation in total engine displacement. Similarly, the 
current regulations in part 90 do not provide a clear way of 
distinguishing engine families by cylinder dimensions (bore and stroke) 
so we are also proposing to change part 90 to limit the variation in 
displacement within an engine family to 15 percent. (See Sec.  1054.230 
and Sec.  90.116.)
    The test procedures for Small SI engines are designed for engines 
operating in constant-speed applications. This covers the large 
majority of affected equipment; however, we are aware that engines 
installed in some types of equipment, such as small utility vehicles or 
go carts, are not governed to operate only at a single rated speed. 
These engines would be certified based on their emission control over 
the constant-speed duty cycle even though they do not experience 
constant-speed operation in use. We are not prepared to propose a

[[Page 28161]]

new duty cycle for these engines but we are proposing to require engine 
manufacturers to explain how their emission control strategy is not a 
defeat device in the application for certification. For example, if 
engines will routinely experience in-use operation that differs from 
the specified duty cycle for certification, the manufacturer should 
describe how the fuel-metering system responds to varying speeds and 
loads not represented by the duty cycle. We are also proposing to 
require that engine distributors and equipment manufacturers that 
replace installed governors must have a reasonable technical basis for 
believing that the effectiveness of the modified engine's emission 
controls over the expected range of in-use operation will be similar to 
that measured over the specified duty cycle (see Sec.  1054.650). This 
may require test data. While this does not require a new certificate of 
conformity, it may require testing to confirm that the engine 
modification should not be considered tampering. In addition, we would 
require that engine distributors and equipment manufacturers notify the 
engine manufacturer before modifying the engine, follow any 
instructions from the engine manufacturer related to the emission 
control system, and avoid making any other changes to the engine that 
would remove it from its certified configuration. We request comment on 
these provisions.

F. Small Business Provisions

(1) Small Business Advocacy Review Panel
    On August 17, 2006, we convened a Small Business Advocacy Review 
Panel (SBAR Panel or the Panel) under section 609(b) of the Regulatory 
Flexibility Act (RFA), as amended by the Small Business Regulatory 
Enforcement Fairness Act of 1996 (SBREFA). The purpose of the Panel was 
to collect the advice and recommendations of representatives of small 
entities that could be affected by this proposed rule and to prepare a 
report containing the Panel's recommendations for small entity 
flexibilities based on those comments, as well as on the Panel's 
findings and recommendations regarding the elements of the Initial 
Regulatory Flexibility Analysis (IRFA) under section 603 of the RFA. 
Those elements of an IRFA are:
     A description of, and where feasible, an estimate of the 
number of small entities to which the proposed rule will apply;
     A description of projected reporting, recordkeeping, and 
other compliance requirements of the proposed rule, including an 
estimate of the classes of small entities that will be subject to the 
requirements and the type of professional skills necessary for 
preparation of the report or record;
     An identification, to the extent practicable, of all 
relevant Federal rules that may duplicate, overlap, or conflict with 
the proposed rule; and
     A description of any significant alternative to the 
proposed rule that accomplishes the stated objectives of applicable 
statutes and that minimizes any significant economic impact of the 
proposed rule on small entities.
    The report of the Panel has been placed in the rulemaking record 
for this proposal.
    In addition to EPA's Director of the Office of Regulatory 
Management and Information who acted as chairperson, the Panel 
consisted of the Director of the EPA's Assessment and Standards 
Division of the Office of Transportation and Air Quality, the 
Administrator of the Office of Management and Budget's Office of 
Information and Regulatory Affairs, and the Chief Counsel for Advocacy 
of the Small Business Administration.
    Using definitions provided by the Small Business Administration 
(SBA), companies that manufacture internal-combustion engines and that 
employ fewer than 1,000 people are considered small businesses for the 
SBAR Panel. Companies that manufacture equipment and that employ fewer 
than 500 people, or fewer than 750 people for manufacturers of 
construction equipment, or fewer than 1,000 people for manufacturers of 
generators, are considered small businesses for the SBAR Panel. Based 
on this information, we asked 25 companies that met the SBA small 
business thresholds to serve as small entity representatives for the 
duration of the Panel process. Of these 25 companies, 14 of them 
represented a cross-section of Small SI engine manufacturers, equipment 
manufacturers, and fuel system component manufacturers. (The rest of 
the companies were involved in the Marine SI market.)
    With input from small entity representatives, the Panel drafted a 
report providing findings and recommendations to us on how to reduce 
the potential burden on small businesses that may occur as a result of 
this proposed rule. The Panel report is included in the rulemaking 
record for this proposal. In light of the Panel report, and where 
appropriate, we have identified provisions anticipated for the proposed 
rule. The proposed flexibility options, based on the recommendations of 
the Panel, are described below.
(2) Proposed Burden Reduction Approaches for Small-Volume Nonhandheld 
Engine Manufacturers
    We are proposing several provisions for small business nonhandheld 
engine manufacturers. The purpose of these provisions is to reduce the 
burden on companies for which fixed costs cannot be distributed over a 
large number of engines. We request comment on the appropriateness of 
these provisions which are described in detail below.
    Under EPA's current Phase 2 regulations, EPA provided a number of 
provisions for small-volume engine manufacturers. For the Phase 2 
regulations, the criteria for determining if a company was a ``small-
volume engine manufacturer'' was based on whether the company projected 
at time of certification to have production of no more than 10,000 
nonhandheld engines per year (excluding engines sold in California that 
are subject to the California ARB standards). Based on past experience, 
EPA believes that determining the applicability of the provisions based 
on number of employees, as compared to volume of products, can be more 
problematic given the nature of the workforce in terms of full-time, 
part-time, contract, overseas versus domestic, and parent companies. 
EPA believes it can avoid these potential complications and still 
provide relief to nearly all small businesses by continuing to use the 
annual sales criteria for determining which entities qualify as a small 
volume engine manufacturer under the Phase 3 program. For these 
reasons, EPA is proposing to retain the current production-based 
criteria for determining who is a small-volume engine manufacturer and, 
as a result, eligible for the Phase 3 flexibilities described below 
(see Sec.  1054.801).
    Based on confidential sales data provided to EPA by engine 
manufacturers, the 10,000 unit cut-off for engine manufacturers would 
include all of the small business engine manufacturers currently 
identified using SBA's employee-based definition. To ensure all small 
businesses have access to the flexibilities described below, EPA is 
also proposing to allow engine manufacturers which exceed the 
production cut-off level noted above but have fewer than 1,000 
employees to request treatment as a small-volume engine manufacturer 
(see Sec.  1054.635). In such a case, the manufacturer would need to 
provide information to EPA demonstrating that the manufacturer has

[[Page 28162]]

fewer employees than the 1,000 cut-off level.
    If a small-volume engine manufacturer grows over time and exceeds 
the production volume limit of 10,000 nonhandheld engines per year, the 
engine manufacturer would no longer be eligible for the small volume 
flexibilities. However, because some of the flexibilities described 
below provide manufacturers with the ability to avoid certain testing 
such as durability testing or production line testing, it may be 
difficult for a manufacturer to fully comply with all of the testing 
requirements immediately upon losing its small-volume status. In such 
cases, EPA is proposing that the engine manufacturer would be able to 
contact EPA and request additional time, subject to EPA approval, to 
meet the testing requirements that generally apply to engine 
manufacturers.
(a) Assigned Deterioration Factors
    We are proposing that small-volume engine manufacturers may rely on 
an assigned deterioration factor to demonstrate compliance with the 
standards for the purposes of certification rather than doing service 
accumulation and additional testing to measure deteriorated emission 
levels at the end of the regulatory useful life (see Sec.  1054.240). 
EPA is not proposing actual levels for the assigned deterioration 
factors with this proposal. EPA intends to analyze emissions 
deterioration information that becomes available over the next few 
years to determine what deterioration factors would be appropriate for 
nonhandheld engines. This is likely to include deterioration data for 
engines certified to comply with California ARB's Tier 3 standards and 
engines certified early to EPA's Phase 3 standards. Prior to the 
implementation date for the Phase 3 standards, EPA will provide 
guidance to engine manufacturers specifying the levels of the assigned 
deterioration factors for small-volume engine manufacturers.
(b) Exemption From Production-Line Testing
    We are proposing that small-volume engine manufacturers would be 
exempt from the production-line testing requirements (see Sec.  
1054.301). While we are proposing to exempt small-volume engine 
manufacturers from production line testing, we believe requiring 
limited production-line testing could be beneficial to implement the 
ongoing obligation to ensure that production engines are complying with 
the standards. Therefore, we request comment on the alternative of 
applying limited production-line testing to small-volume engine 
manufacturers with a requirement to test one production engine per 
year.
(c) Additional Lead Time
    We are proposing that small-volume engine manufacturers could delay 
implementation of the Phase 3 exhaust emission standards for two years 
(see Sec.  1054.145). Small-volume engine manufacturers would be 
required to comply with the Phase 3 exhaust emission standards 
beginning in model year 2014 for Class I engines and model year 2013 
for Class II engines. Under this approach, manufacturers would be able 
to apply this delay to all of their nonhandheld engines or to just a 
portion of their production. For those engine families that are 
certified to meet the Phase 3 standards prior to these delayed dates by 
selecting an FEL at or below the Phase 3 standards, small volume engine 
manufacturers could generate early Phase 3 credits (as discussed in 
Section V.C.3) through the 2013 model year for Class I engines and 
through the 2012 model years for Class II engines. This option provides 
more lead time for small-volume engine manufacturers to redesign their 
products. They would also be able to learn from some of the hurdles 
overcome by larger manufacturers.
(d) Broad Engine Families
    We are also proposing that small-volume engine manufacturers may 
use a broader definition of engine family for certification purposes. 
Under the existing engine family criteria specified in the regulations, 
manufacturers group their various engine lines into engine families 
that have similar design characteristics including the combustion 
cycle, cooling system, cylinder configuration, number of cylinders, 
engine class, valve location, fuel type, aftertreatment design, and 
useful life category. We are proposing to allow small-volume engine 
manufacturers to group all of their Small SI engines into a single 
engine family for certification by engine class and useful life 
category, subject to good engineering judgment (see Sec.  1054.230).
(e) Hardship Provisions
    We are also proposing two types of hardship provisions for 
nonhandheld engine manufacturers consistent with the Panel 
recommendations. The first type of hardship is an unusual circumstances 
hardship which would be available to all businesses, regardless of 
size. The second type of hardship is an economic hardship provision 
which would be available to small businesses only. Sections VIII.C.8 
and VIII.C.9 provide a description of the proposed hardship provisions 
that would apply to nonhandheld engine manufacturers.
(3) Proposed Burden Reduction Approaches for Small-Volume Nonhandheld 
Equipment Manufacturers
    We are proposing three provisions for small-volume nonhandheld 
equipment manufacturers. The purpose of these provisions is to reduce 
the burden on companies for which fixed costs cannot be distributed 
over large sales volumes. We are offering these provisions because 
equipment manufacturers may need more lead time to redesign their 
equipment to accommodate the new Phase 3 engine designs. We request 
comment on the appropriateness of the flexibilities described below.
    Under EPA's current Phase 2 regulations, EPA provided a number of 
lead time provisions for small-volume equipment manufacturers. For the 
Phase 2 regulations, the criteria for determining if a company was a 
``small-volume equipment manufacturer'' was based on whether the 
company produced fewer than 5,000 nonhandheld pieces of equipment per 
year (excluding equipment sold in California that are subject to the 
California ARB standards). For the same reasons noted above for engine 
manufacturers, EPA is proposing to retain the current production-based 
criteria for determining who is a small-volume equipment manufacturer 
and, as a result, eligible for the Phase 3 flexibilities described 
below (see Sec.  1054.801). The determination of which companies 
qualify as small-volume equipment manufacturers for the purposes of the 
flexibilities described below would be based on the annual U.S.-
directed production of nonhandheld equipment in each of the three years 
from 2007 through 2009.
    Based on estimated sales data for equipment manufacturers, EPA 
believes the 5,000 unit cut-off for equipment manufacturers would 
include almost all of the small business equipment manufacturers using 
SBA's employee-based definition. However to ensure all small businesses 
have access to the flexibilities described below, EPA is also proposing 
to allow equipment manufacturers which exceed the production cut-off 
level noted above but have fewer than 500 employees for equipment 
manufacturers, or 750 employees for construction equipment 
manufacturers, or 1,000 employees for generator manufacturers, to 
request treatment as a small-volume equipment manufacturer (see Sec.  
1054.635). In such a case, the manufacturer would need to provide 
information to EPA

[[Page 28163]]

demonstrating that the manufacturer has fewer employees than the 
applicable employee cut-off level.
(a) Additional Lead Time
    As described in Section V.E.3., EPA is proposing a transition 
program for all equipment manufacturers that produce Class II 
equipment. Under that program, equipment manufacturers can install 
Phase 2 engines in limited numbers of Class II equipment over the first 
four years the Phase 3 standards apply (i.e., 2011 through 2014). The 
number of equipment that can use Phase 2 engines is based on 30 percent 
of an average annual production level of Class II equipment. To 
implement this two-year extension for small-volume equipment 
manufacturers within the context of the transition program for 
equipment manufacturers, EPA is proposing that small-volume 
manufacturers may use Phase 2 engines at a level of 200 percent of an 
average annual production level of Class II equipment. Small-volume 
equipment manufacturers could use these allowances over the four year 
period of the transition program (see Sec.  1054.625). Therefore, a 
small-volume equipment manufacturer could potentially use Phase 2 
engines on all their Class II equipment for two years, consistent with 
the SBAR Panel's recommendation, or they might, for example, sell half 
their Class II equipment with Phase 2 engines for four years assuming 
sales stay constant over time.
(b) Simplified Certification Procedure
    We are proposing a simplified engine certification procedure for 
all equipment manufacturers, including small-volume equipment 
manufacturers. See Section V.E.4 for further discussion of this 
provision.
(c) Hardship Provisions
    Because nonhandheld equipment manufacturers in many cases depend on 
engine manufacturers to supply certified engines in time to produce 
complying equipment, we are also proposing a hardship provision for all 
nonhandheld equipment manufacturers, regardless of size. The proposed 
hardship would allow the manufacturer to request more time if they are 
unable to obtain a certified engine and they are not at fault and would 
face serious economic hardship without an extension (see Sec.  
1068.255). Section VIII.C.10 provides a description of the proposed 
hardship provision that would apply to nonhandheld equipment 
manufacturers.

G. Technological Feasibility

(1) Level of Standards
    We are proposing new, more stringent exhaust HC+NOX 
standards for Class I and II Small SI engines. We are also proposing a 
new CO standard for Small SI engines used in marine generator 
applications.
    In the 2005 model year manufacturers certified over 500 Class I and 
II engine families to the Phase 2 standards using a variety of engine 
designs and emission control technology. All Class I engines were 
produced using carbureted air-fuel induction systems. A small number of 
engines used catalyst-based emission control technology. Similarly, 
Class II engines were predominately carbureted. A limited number of 
these engines used catalyst technology, electronic engine controls and 
fuel injection, or were water cooled. In both classes, several engine 
families were certified at levels that would comply with the proposed 
Phase 3 standards. Also, a number of families were very close to the 
proposed emission standards. This suggests that, even accounting for 
the relative increase in stringency associated with our proposed Phase 
3 requirements, a number of families either will not need to do 
anything or will require only modest reductions in their emission 
performance to meet the proposed standards. However, many engine 
families clearly will have to do more to improve their emissions 
performance.
    Based on our own testing of advanced technology for these engines, 
our engineering assessments, and statements from the affected industry, 
we believe the proposed requirements will require many engine 
manufacturers to adopt exhaust aftertreatment technology using 
catalyst-based systems. Other likely changes include improved engine 
designs and fuel delivery systems. Finally, adding electronic controls 
or fuel injection systems may obviate the need for catalytic 
aftertreatment for some engine families, with the most likely 
candidates being multi-cylinder engine designs.
(2) Implementation Dates
    We are proposing HC+NOX exhaust emission standards of 
10.0 g/kW-hr for Class I engines starting in the 2012 model year and 
8.0 g/kW-hr for Class II engines starting in the 2011 model year. For 
both classes of nonhandheld engines, we are proposing to maintain the 
existing CO standard of 610 g/kW-hr. We expect manufacturers to meet 
these standards by improving engine combustion and adding catalysts.
    For spark-ignition engines used in marine generators, we are 
proposing a more stringent Phase 3 CO emission standard of 5.0 g/kW-hr. 
This would apply equally to all sizes of engines subject to the Class I 
and II Small SI standards, with implementation dates as described above 
relative to Class I and Class II engines.
(3) Technological Approaches
    Our feasibility assessment began by evaluating the emissions 
performance of current technology for Small SI engines and equipment. 
These initial efforts focused on developing a baseline for emissions 
and general engine performance so that we could assess the potential 
for new emission standards for engines and equipment in this category. 
This process involved laboratory and field evaluations of the current 
engines and equipment. We reviewed engineering information and data on 
existing engine designs and their emissions performance. Patents of 
existing catalyst/muffler designs for Class I engines were also 
reviewed. We engaged engine manufacturers and suppliers of emission 
control-related engine components in discussions regarding recent and 
expected advances in emissions performance beyond that required to 
comply with the current Phase 2 standards. Finally, we purchased 
catalyst/muffler units that were already in mass production by an 
original equipment manufacturer for use on European walk-behind lawn 
mowers and conducted engineering and chemical analyses on the design 
and materials of those units.
    We used the information and experience gathered in the above effort 
along with the previous catalyst design experience of our engineering 
staff, to design and build prototype catalyst-based emission control 
systems that were capable of effectively and safely achieving the 
proposed Phase 3 requirement based on dynamometer and field testing. We 
also used the information and the results of our engine testing to 
assess the potential need for improvements to engine and fuel system 
designs, and the selective use of electronic engine controls and fuel 
injection on some engine types. A great deal of this effort was 
conducted in association with our more exhaustive study regarding the 
efficacy and safety of implementing advanced exhaust emission controls 
on Small SI engines, as well as new evaporative requirements for these 
engines. In other testing, we evaluated advanced emission controls on a 
multi-cylinder Class II engine with electronic fuel injection. The 
results of that study are also discussed in Section XII.

[[Page 28164]]

    In our test program to assess the feasibility of achieving the 
proposed Phase 3 HC+NOX standard, we evaluated 15 Class I 
engines of varying displacements and valve-train designs. Each of these 
engines was equipped with a catalyst-based control system and all 
achieved the applicable standard at the end of their regulatory useful 
lives. Our work also suggests that manufacturers of Class I engines may 
also need to improve the durability of their basic engine designs, 
ignition systems, or fuel metering systems for some engines in order to 
comply with the emission regulations.
    We tested five single-cylinder, overhead-valve Class II engines 
with prototype catalyst/muffler control systems. Three of the engines 
were carbureted and two were equipped with electronic engine and fuel 
controls. This latter technology improves the management of air-fuel 
mixtures and ignition spark timing. This itself can reduce engine-out 
emissions relative to a carbureted system and also allows the use of 
larger catalyst volumes and higher precious metal loading. Each of the 
engines achieved the requisite emission limit for HC+NOX 
(e.g., 8.0 g/kW-hr). Based on this work and information from one 
manufacturer of emission controls, we believe that either a catalyst-
based system or electronic engine controls appear sufficient to meet 
the standard. Nonetheless, some applications may require the use of 
both technologies. Finally, similarly to Class I engines, we found that 
manufacturers of Class II engines may also need to improve the 
durability of their ignition systems or fuel metering systems for some 
engines in order to comply with the emission regulations.
    Multi-cylinder Class II engines are very similar to their single-
cylinder counterparts regarding engine design and combustion 
characteristics. There are no multi-cylinder Class I engines. Base on 
these attributes and our testing of two twin-cylinder engines, we 
conclude that the proposed Phase 3 HC+NOX standard is 
technically feasible.
    Nonetheless, we also found that multi-cylinder engines may present 
unique concern with the application of catalytic control technology 
under atypical operation conditions. More specifically, the concern 
relates to the potential consequences of combustion misfire or a 
complete lack of combustion in one of the two or more cylinders when a 
single catalyst/muffler design is used. A single muffler is typically 
used in Class II applications. In a single-catalyst system, the 
unburned fuel and air mixture from the malfunctioning cylinder would 
combine with hot exhaust gases from the other, properly operating 
cylinder. This condition would create high temperatures within the 
muffler system as the unburned fuel and air charge from the misfiring 
cylinder combusts within the exhaust system. This could potentially 
destroy the catalyst.
    One solution is simply to have a separate catalyst/muffler for each 
cylinder. Another solution is to employ electronic engine controls to 
monitor ignition and put the engine into ``limp-mode'' until necessary 
repairs are made. For engines using carburetors, this would effectively 
require the addition of electronic controls. For engines employing 
electronic fuel injection that may need to add a small catalyst, it 
would require that the electronic controls incorporate ignition misfire 
detection if they do not already utilize the inherent capabilities 
within the engine management system.
    As described earlier, we also expect some engine families may use 
electronic fuel injection to meet the proposed Phase 3 standard without 
employing catalytic aftertreatment. Engine families that already use 
these fuel metering systems and are reasonably close to complying with 
the proposed requirement are likely to need only additional calibration 
changes to the engine management system for compliance. In addition, we 
expect that some engine families which currently use carbureted fuel 
systems will convert directly to electronic fuel injection. 
Manufacturers may adopt this strategy to couple achieving the standard 
without a catalyst and realizing other advantages of using fuel 
injection such as easier starting, more stable and reliable engine 
operation, and reduced fuel consumption.
    Our evaluation of electronic fuel injection systems that could be 
used to attain the proposed standard found that a rather simple, low-
cost system should be sufficient. We demonstrated this proof of concept 
as part of the engine test program we conducted for our safety study. 
In that program, we fitted two single-cylinder Class II engines with an 
electronic control unit and fuel system components developed for Asian 
motor-scooters and small-displacement motorcycles. The sensors for the 
system were minimized to include a throttle position sensor, air charge 
temperature sensor, oil temperature sensor, manifold absolute pressure 
sensor, and a crankshaft position sensor. This is in contrast to the 
original equipment manufacturer fuel injection systems currently used 
in some equipment with two-cylinder Class II engine applications that 
employ more sophisticated and expensive automotive-based components.
    Finally, there are a number of Class II engines that use gaseous 
fuels (i.e., liquid propane gas or compressed natural gas). Based on 
our engineering evaluation of current and likely emission control 
technology for these engines, we conclude that there are no special 
concerns relative to achieving the proposed Phase 3 HC+NOX 
standard.
    Turning to the proposed Phase 3 CO standard for Class I and II 
Small SI engines used in marine generator applications, these engines 
have several rather unique design considerations that are relevant to 
achieving the proposed CO standard. Marine generator engines are 
designed to operate for very long periods. Manufacturers generally 
design the engines to operate at lower loads to accommodate continuous 
operation. Manufacturers also design them to take advantage of the 
cooling available from the water in the lake or river where the boat is 
operating (seawater). By routing seawater through the engine block, or 
using a heat exchanger that transfers heat from the engine coolant to 
the seawater, manufacturers are able to maintain engine temperatures as 
well or better than automotive engines. Stable temperatures in the 
engine block make a very significant difference in engine operation, 
enabling much less distortion of the cylinders and a much more 
consistent combustion event. These operating characteristics make it 
possible to introduce advanced technology for controlling emissions. 
Manufacturers also use this cooling water in a jacketing system around 
the exhaust in order to minimize surface temperatures and reduce the 
risk of fires on boats.
    The vast majority of gasoline marine generators are produced by two 
engine manufacturers. Recently, these two manufacturers have announced 
that they are converting their marine generator product lines to new 
designs which can achieve more than a 99 percent reduction in CO 
emissions. These manufacturers stated that this action is to reduce the 
risk of CO poisoning and is a result of boat builder demand. These low 
CO emission designs used closed-loop electronic fuel injection and 
catalytic control. Both of these manufacturers have certified some low 
CO engines and have expressed their intent to convert their full 
product lines in the near future. These manufacturers also make use of 
electronic controls to monitor catalyst function.

[[Page 28165]]

(4) Consideration of Regulatory Alternatives
    In developing the proposed emission standards, we considered what 
was achievable with catalyst technology. Our technology assessment work 
indicated that the proposed emission standards are feasible in the 
context of provisions for establishing emission standards prescribed in 
section 213 of the Clean Air Act. We also considered what could be 
achieved with larger, more efficient catalysts and improved fuel 
induction systems. In particular, Chapter 4 of the Draft RIA presents 
data on Class I engines with more active catalysts and on Class II 
engines with closed-loop control fuel injection systems in addition to 
a catalyst. In both cases larger emission reductions were achieved.
    Based on this work we considered HC+NOX standards which 
would have involved a 50 percent reduction for Class I engines and a 
65-70 percent reduction for Class II engines. Chapter 11 of the Draft 
RIA evaluates these alternatives, including an assessment of the 
overall technology and costs of meeting more stringent standards. For 
Class I engines a 50 percent reduction standard would require base 
engine changes not necessarily involved with the standards we are 
proposing and the use of a more active catalyst. For Class II engines 
this would require the widespread use of closed loop control fuel 
injection systems rather than carburetors, some additional engine 
upgrades, and the use three-way catalysts. We believe it is not 
appropriate at this time to propose more stringent exhaust emission 
standards for Small SI engines. Our key concern is lead time. More 
stringent standards would require three to five years of lead time 
beyond the 2011 model year start date we are proposing for the program. 
We believe it would be more effective to implement the proposed Phase 3 
standards to achieve near-term emission reductions needed to reduce 
ozone precursor emissions and to minimize growth in the Small SI 
exhaust emissions inventory in the post 2010 time frame. More efficient 
catalysts, engine improvements, and closed loop electronic fuel 
injection could be the basis for more stringent Phase 4 emission 
standards at some point in the future.
(5) Our Conclusions
    We believe the proposed Phase 3 exhaust emission standards for 
nonhandheld Small SI engines will achieve significant emission 
reductions. Manufacturers will likely meet the proposed standards with 
a mix of three-way catalysts packaged in the mufflers and fuel-
injection systems. Test data using readily available technologies have 
demonstrated the feasibility of achieving the proposed emission levels.
    As discussed in Section X, we do not believe the proposed standards 
would have negative effects on energy, noise, or safety and may lead to 
some positive effects.

VI. Evaporative Emissions

A. Overview

    Evaporative emissions refer to hydrocarbons released into the 
atmosphere when gasoline or other volatile fuels escape from a fuel 
system. The primary source of evaporative emissions from nonroad 
gasoline engines and equipment is known as permeation, which occurs 
when fuel penetrates the material used in the fuel system and reaches 
the ambient air. This is especially common through rubber and plastic 
fuel-system components such as fuel lines and fuel tanks. Diurnal 
emissions are another important source of evaporative emissions. 
Diurnal emissions occur as the fuel heats up due to increases in 
ambient temperature. As the fuel heats, liquid fuel evaporates into the 
vapor space inside the tank. In a sealed tank, these vapors would 
increase the pressure inside the tank; however, most tanks are vented 
to prevent this pressure buildup. The evaporating fuel therefore drives 
vapors out of the tank into the atmosphere. Diffusion emissions occur 
when vapor escapes the fuel tank through an opening as a result of 
random molecular motion, independent of changing temperature. Running 
loss emissions are similar to diurnal emissions except that vapors 
escape the fuel tank as a result of heating from the engine or some 
other source of heat during operation rather than from normal daily 
temperature changes. Refueling losses are vapors that are displaced 
from the fuel tank to the atmosphere when someone fills a fuel tank. 
Refueling spitback is the spattering of liquid fuel droplets coming out 
of the filler neck during a refueling event. Spillage is fuel that is 
spilled while refueling. Regulatory provisions to set standards for 
several of these types of evaporative emissions effectively define the 
terms for establishing the specific test procedures for measuring 
emissions. See the proposed regulatory text for more information.
    This proposal is part of a larger effort to control evaporative 
emissions from all mobile sources. Motor vehicles have stringent 
evaporative emission controls based on SHED testing of complete 
vehicles.\82\ As a result, motor vehicle manufacturers must control 
diurnal emissions, permeation through all fuel-system components, 
running loss emissions, refueling vapor displacement, refueling 
spitback, and to some extent, spillage. We recently established 
evaporative emission standards for recreational vehicles and Large SI 
engines (67 FR 68242, November 8, 2002). These standards include 
permeation requirements for fuel tanks and fuel lines. In addition, 
equipment using Large SI engines must control diurnal emissions and 
running losses. Fuel systems used with Small SI engines and Marine SI 
engines are not yet subject to evaporative emission standards.
---------------------------------------------------------------------------

    \82\ An entire vehicle is placed in a SHED (Sealed Housing for 
Evaporative Determination) and total evaporative emissions are 
measured over prescribed test cycles.
---------------------------------------------------------------------------

    In August 2002, we proposed permeation and diurnal emission 
standards for fuel systems related to Marine SI engines (67 FR 53050, 
August 14, 2002). We finalized other portions of that proposal but 
chose to delay promulgation of Marine SI evaporative standards. At the 
time of the earlier proposal there were still open issues regarding 
emission control technologies for rotational-molded fuel tanks and for 
pressurizing fuel tanks as a diurnal emission control strategy. Since 
then, EPA has continued gathering information and performing tests on 
new technologies that could be used to address these issues. In this 
notice we are updating the proposed evaporative emission standards for 
Marine SI fuel systems. The standards in this proposal incorporate this 
new information.
    We are also proposing standards for controlling evaporative 
emissions from fuel systems used with Small SI engines. These proposed 
standards include requirements for controlling permeation, diffusion, 
and running loss emissions.

B. Fuel Systems Covered by This Rule

    The proposed evaporative emission standards would apply to fuel 
systems for both Small SI engines and Marine SI engines. The marine 
standards apply to fuel systems related to both propulsion and 
auxiliary engines. In some cases, specific standards are proposed only 
for certain types of equipment, as described below. These standards 
would apply only to new products, as described in Section VII.A.
    We are proposing to write the regulations related to evaporative 
emission standards in 40 CFR part 1060,

[[Page 28166]]

which is devoted to evaporative emission controls from nonroad engines 
and equipment. The exhaust standard-setting part (part 1045 for Marine 
SI and part 1054 for Small SI) defines the emission standards, but 
references part 1060 for certification and testing procedures, in 
addition to definitions, compliance-related issues, and other special 
provisions. Section VII describes further how the different parts work 
together in the certification process. Also, as described in Section 
XI, we are proposing to allow component manufacturers and some 
equipment manufacturers to certify products under the provisions of 
part 1060 with respect to recreational vehicles. We also plan to 
clarify in a separate action that marine and land-based compression-
ignition engines that operate on volatile liquid fuels (such as 
methanol or ethanol) are subject to evaporative requirements related to 
part 1060. The draft regulations in part 1060 describe how those 
provisions would apply for compression-ignition engines, but these 
regulations impose no obligations until we adopt those as requirements 
in a separate rulemaking.
    The following definitions are important in establishing which 
components would be covered by the proposed standards: ``evaporative,'' 
``fuel system,'' ``fuel line,'' ``portable nonroad fuel tank,'' and 
``installed marine fuel tank.'' See the full text of these definitions 
in the proposed regulations at Sec.  1060.801.
    Note in particular that the proposed standards would apply to fuel 
lines, including hose or tubing that contains liquid fuel. This would 
include fuel supply lines but not vapor lines or vent lines not 
normally exposed to liquid fuel. We consider fuel return lines for 
handheld engines to be vapor lines, not fuel lines. Data in Chapter 5 
of the Draft RIA suggest that permeation rates through vapor lines and 
vent lines are already lower than the proposed standard; this is due to 
the low vapor concentration in the vapor line. In contrast, permeation 
rates for materials that are consistently exposed to saturated fuel 
vapor are generally considered to be about the same as that for liquid 
fuel. The standards also do not apply to primer bulbs exposed to liquid 
fuel only for priming. This standard would apply to marine filler necks 
that are filled or partially filled with liquid fuel after a refueling 
event where the operator fills the tank as full as possible. In the 
case where the fuel system is designed to prevent liquid fuel from 
standing in the fill neck, the fill neck would be considered a vapor 
line and not subject to the proposed fuel line permeation standard. We 
request comment on the appropriateness of applying permeation standards 
to filler necks, vapor lines and vent lines for Small SI engines and 
Marine SI engines.
    One special note applies to fuel systems for auxiliary marine 
engines. These engines must meet exhaust emission standards that apply 
to land-based engines. This is appropriate because these engines, 
typically used to power generators, operate more like land-based 
engines than like marine propulsion engines. For evaporative emissions, 
however, it is important that the fuel systems for propulsion and 
auxiliary engines be subject to the same standards because these 
engines typically draw fuel from a common fuel tank and share other 
fuel-system components. We are therefore proposing to apply the Marine 
SI evaporative emission standards and certification requirements to the 
fuel systems for both auxiliary and propulsion marine engines on marine 
vessels.
    Our evaporative emission standards for automotive applications are 
based on a comprehensive measurement from the whole vehicle. However, 
the evaporative standards in this proposal are generally based on 
individual fuel-system components. For instance, we are proposing 
permeation standards for fuel lines and fuel tanks rather than for the 
equipment as a whole.\83\ We are taking this approach for several 
reasons. First, most production of Small SI equipment and Marine SI 
vessels is not vertically integrated. In other words, the fuel line 
manufacturer, the engine manufacturer, the fuel tank manufacturer, and 
the equipment manufacturer are often separate companies. In addition, 
there are several hundred equipment manufacturers and boat builders, 
many of which are small businesses. Testing the systems as a whole 
would place the entire certification burden on the equipment 
manufacturers and boat builders. Specifying emission standards and 
testing for individual components allows for measurements that are 
narrowly focused on the source of emissions and on the technology 
changes for controlling emissions. This correspondingly allows for 
component manufacturers to certify that their products meet applicable 
standards. We believe it would be most appropriate for component 
manufacturers to certify their products since they are best positioned 
to apply emission control technologies and demonstrate compliance. 
Equipment manufacturers and boat builders would then be able to 
purchase certified fuel-system components rather than doing all their 
own testing on individual components or whole systems to demonstrate 
compliance with every requirement. In contrast, controlling running 
loss emissions cannot be done on a component basis so we are proposing 
to require engine or equipment manufacturers to certify that they meet 
the running loss standard. We would otherwise expect most equipment 
manufacturers to simply identify a range of certified components and 
install the components as directed by the component manufacturer to 
demonstrate compliance with the proposed emission standards.
---------------------------------------------------------------------------

    \83\ An exception to component certification is the design 
standard for contolling running loss emissions.
---------------------------------------------------------------------------

    Second, a great deal of diversity exists in fuel-system designs 
(hose lengths, tank sizes/shapes, number of connections, etc.). In most 
cases, the specific equipment types are low-volume production runs so 
sales would not be large enough to cover the expense of SHED-type 
testing. Third, there are similarities in fuel lines and tanks that 
allow for component data to be used broadly across products in spite of 
extensive variety in the geometry and design of fuel systems. Fourth, 
many equipment types, primarily boats, would not fit in standard-size 
SHEDs and would require the development of very large, very expensive 
test facilities if the entire vessel were tested.
    Finally, by proposing separate standards for fuel line permeation, 
fuel tank permeation, diurnal emissions, and diffusion emissions, we 
are able to include simplified certification requirements without 
affecting the level of the standards. Specifying a comprehensive test 
with a single standard for all types of evaporative emissions would 
make it difficult or impossible to rely on design-based certification. 
Requiring emission tests to cover the wide range of equipment models 
would greatly increase the cost of compliance with little or no 
increase in the effectiveness of the certification program. We believe 
the proposed approach allows substantial opportunity for market forces 
to appropriately divide compliance responsibilities among affected 
manufacturers and accordingly results in an effective compliance 
program at the lowest possible cost to society.
    The proposed emission standards generally apply to the particular 
engines and their associated fuel systems. However, for ease of 
reference, we may refer to evaporative standards as being related to 
Small SI equipment or Marine SI vessels, meaning the relevant

[[Page 28167]]

evaporative standards for engines and fuel systems used in such 
equipment or vessels.\84\ See Section VI.F for a more detailed 
description of certification responsibilities for all the proposed 
evaporative standards.
---------------------------------------------------------------------------

    \84\ ``Small SI equipment'' includes all nonroad equipment 
powered by Small SI engines. ``Marine SI vessels'' includes all 
vessels powered by engines that run on volatile liquid fuels. In 
almost all cases these engines are powered by gasoline. Note also 
that volatile liquid fuels include methanol or ethanol, which could 
be used in a compression-ignition engine. While we are aware of no 
such equipment or vessels today, they would be covered by the 
proposed regulations. In this preamble, we nevertheless refer to all 
the vessels that fall within the scope of the proposed regulations 
as marine SI vessels. Throughout this section, we generally refer to 
Small SI equipment and Marine SI vessels as ``equipment,'' 
consistent with the proposed regulatory text.
---------------------------------------------------------------------------

C. Proposed Evaporative Emission Standards

    We are proposing permeation standards for Small SI equipment and 
Marine SI vessels, covering permeation from fuel tanks and fuel lines. 
We are also proposing diurnal emission standards for Marine SI vessels. 
We are proposing diffusion emission standards but not diurnal emission 
standards for nonhandheld Small SI equipment. In addition, we are 
proposing a running loss standard for nonhandheld Small SI equipment 
(except wintertime engines), with a variety of specified options for 
manufacturers to demonstrate compliance. Based on the current state of 
technology, we believe the proposed standards are a logical extension 
of the standards proposed for marine vessels in August 2002 and the 
standards finalized for recreational vehicles in November 2002.
    All the proposed evaporative emission standards would apply to new 
equipment for a useful life period in years that matches the useful 
life of the corresponding engine. We propose to specify a five-year 
useful life for evaporative requirements for Small SI equipment (we are 
not proposing a year-based useful life requirement related to exhaust 
emissions for Small SI engines). Manufacturers have expressed concern 
that they will not have time to gain five years of in-use experience on 
low-permeation fuel tanks by the proposed dates of the tank permeation 
standards. Unlike barrier fuel line, which is well established 
technology, some fuel tanks may use barrier technologies that have not 
been used extensively in other applications. An example of this 
technology would be barrier surface treatments that must be properly 
matched to the fuel tank material. Therefore, we are proposing a 
shorter useful life of two years for Marine SI and Small SI fuel tanks 
through the 2013 model year to allow manufacturers to gain experience 
in use (see Sec. Sec.  1045.145 and 1054.145). We do not expect this 
interim provision to affect manufacturer designs or in-use compliance 
efforts. We do not believe this interim provision to specify a shorter 
useful life period is necessary for other fuel-system components, 
either because there is adequate durability experience in other sectors 
or because the control inherently does not involve a concern over in-
use deterioration.
    The rest of this section summarizes the proposed standards, 
additional requirements, and implementation dates. Unless otherwise 
stated, implementation dates specified below refer to the model year. 
Section VI.D describes how manufacturers may use emission credits to 
meet fuel tank permeation standards. Section VI.E describes the test 
procedures corresponding to each standard. Section VI.F describes how 
component and equipment manufacturers certify their products and how 
their responsibilities overlap in some cases. Section VI.F also 
describes the simplified process of design-based certification for 
meeting many of the proposed standards.
(1) Fuel Line Permeation Standards and Dates
    The proposed fuel line permeation standard applies to fuel lines 
intended for use in new Small SI equipment and Marine SI vessels is 15 
g/m\2\/day at 23 [deg]C on a test fuel containing 10 percent ethanol 
(see Sec.  1060.102 and Sec.  1060.515). The form of the standard 
refers to grams of permeation over a 24-hour period divided by the 
inside surface area of the fuel line. This proposed standard is 
consistent with that adopted for fuel lines in recreational vehicles. 
The move toward low-permeation fuel lines in recreational vehicles--and 
further development work in this area since the first proposed rule for 
marine evaporative emissions--demonstrates that low-permeation fuel 
lines are available on the market today for Small SI equipment and 
Marine SI vessels. In addition, many manufacturers are already using 
low-permeation technologies in response to permeation standards in 
California. We are therefore proposing that this standard apply 
beginning with 2008 for nonhandheld Small SI equipment and 2009 for 
Marine SI vessels. For handheld equipment, we are proposing a fuel line 
permeation implementation date of 2012, except that small-volume 
families as defined in Sec.  1054.801 would have until 2013. Although 
low-permeation fuel line technology is available, handheld equipment is 
not currently subject to fuel line permeation requirements in 
California and does not typically use low-permeation fuel lines today. 
In addition, much of the fuel line used on handheld equipment is not 
straight-run fuel line for which low-permeation replacements are 
readily available; thus, more lead time is required. We request comment 
on the proposed standard and implementation dates.
    Component manufacturers would be required to certify to the 
proposed emission standard for fuel lines (this may involve 
certification to a family emission limit above the emission standard 
for handheld engines, as described in Section VI.D), except in certain 
circumstances. Equipment manufacturers may need to certify that their 
fuel lines meet the proposed emission standards if they use any 
sections or pieces of fuel line that are not already certified by the 
fuel line manufacturer, or if they comply using emission credits, as 
described in Section VI.F.
    To address the short lead time associated with the 2008 
requirements for Small SI equipment, we are proposing an interim 
arrangement in which engine manufacturers would include compliant fuel 
lines under their existing certification (see Sec.  90.127). This would 
prevent the need for other companies to submit new applications for 
certification that would need to be processed immediately. This 
arrangement would allow for engine manufacturers to start complying 
well ahead of the time that the fuel line standards become mandatory. 
The certification requirements described above for component 
manufacturers would start once Small SI engines and equipment would be 
subject to Phase 3 standards.
    By specifying standards for fuel-system components rather than the 
entire fuel system, we must separately address appropriate requirements 
for connecting pieces, such as valves, O-rings, seals, plugs, and 
grommets that are exposed to liquid fuel but are not part of the fuel 
line. We are proposing to require that these ancillary pieces meet the 
broad specifications described in Sec.  1060.101(f), which generally 
requires that fittings and connections be designed to prevent leaks. As 
described in Section VI.E.1, we are also proposing to allow testing of 
fuel line assemblies that include connecting pieces, primer bulbs, and 
other fuel line components as a single item (see Sec.  1060.102). For 
example, manufacturers may certify fuel lines for portable marine fuel 
tanks as

[[Page 28168]]

assemblies of fuel line, primer bulbs, and self-sealing end 
connections. Finally, we are proposing to require that detachable fuel 
lines be self-sealing when they are removed from the fuel tank or the 
engine because this would otherwise result in high evaporative 
emissions (see Sec.  1060.101). To the extent that equipment 
manufacturers and boat builders certify their products, they would need 
to describe how they meet the equipment-based requirements proposed in 
Sec.  1060.101(e) and (f) in their application for certification. If 
boat builders rely on certified components instead of certifying, they 
would need to keep records describing how they meet the equipment-based 
requirements proposed in Sec.  1060.101(e) and (f).
    Handheld equipment manufacturers have raised concerns that fuel 
lines constructed of available low-permeation materials may not perform 
well in some handheld applications under extreme cold weather 
conditions such as below -30 [deg]C. These products often use injected 
molded fuel lines with complex shapes and designs needed to address the 
unique equipment packaging issues and the high vibration and random 
movement of the fuel lines within the overall equipment when in use. 
Industry has expressed concern and the data in Chapter 5 of the Draft 
RIA suggest that durability issues may occur from using certain low-
permeation materials in these applications when the weather is 
extremely cold and that these could lead to unexpected fuel line leaks. 
Handheld equipment types that could be considered as cold-weather 
products include cut-off saws, clearing saws, brush cutters over 40cc, 
commercial earth and wood drills, ice augers, and chainsaws.
    The extreme cold temperatures needed to induce the potential fuel 
line failures are very rare but do occur each year in Alaska and the 
continental United States. EPA considered a number of different options 
aimed at developing special provisions for equipment most likely to be 
used in these extreme cold weather situations without providing relief 
to all of the equipment sold in the broad categories identified by 
industry as cold weather products. These included focusing the 
provisions on products used by professionals (longer useful life 
equipment or Class V equipment only), geographic-based retrofit kits, 
product segregation, and special labeling. While each of the options 
has some merit, none could provide the full assurance that handheld 
equipment using low-permeation fuel lines not compatible with extreme 
cold weather would not be used in such weather conditions. While very 
low temperature materials are available that can achieve the fuel line 
permeation standards discussed above, these materials come at a 
substantially higher cost than that for fuel lines used in non cold 
weather products and none have been evaluated in fuel lines on the 
handheld equipment at issue.
    If we consider a less stringent standard, we believe there are 
lower cost materials available that could be used to achieve permeation 
reductions in equipment designed for cold weather applications without 
creating potential safety concerns related to fuel leaks. As discussed 
in the Draft RIA, rubbers with high acrylonitrile (ACN) content are 
used in some handheld applications. These materials have about half the 
permeation of lower ACN-content rubbers also used in handheld 
applications. To capture the capability of these materials to reduce 
permeation emissions without creating other issues for cold weather 
products, we are proposing a fuel line permeation standard of 175 g/
m\2\/day in 2013 for cold-weather products. We request comment on 
appropriateness of this standard and whether there are materials that 
could be used to achieve larger fuel line permeation reductions from 
cold-weather products.
    We request comment on what products should be considered to be 
cold-weather products and if it would be possible to distinguish 
between products used in warm versus cold climates. We also request 
comment regarding whether the proposed ABT program discussed below for 
handheld equipment would provide enough flexibility to manufacturers to 
address cold weather issues through credit trading rather than through 
a differentiated standard.
    Outboard engine manufacturers have expressed concern that it would 
be difficult for them to meet proposed 2009 date for the sections of 
fuel lines that are mounted on their engines under the engine cowl. 
While some sections of straight-run fuel line are used on the 
outboards, many of the smaller sections between engine mounted fuel-
system components and connectors are preformed or even injection-molded 
parts. Outboard engine manufacturers stated that they would need 
additional time to redesign and perform testing on low-permeation fuel 
lines under the cowl. PWC and SD/I manufacturers have indicated that 
this is not an issue on their engines because they are dominantly 
straight-run pieces. Outboard engine manufacturers have also stated 
that, in contrast to under cowl fuel line, they would be able to 
facilitate the introduction of low-permeation fuel line, from the fuel 
tank to the engine, in 2008.
    We request comment on implementing an optional program where the 
implementation dates for fuel line under the cowl can be delayed beyond 
2009, provided low-permeation fuel line from the fuel tank to the 
engine is used beginning on January 1, 2008. Under this approach, 
permeation standards for primer bulbs on fuel lines from the tank to 
the engine would still begin in 2009. One specific approach would be to 
phase in the use of low-permeation fuel lines on outboards based on the 
total inside surface area of the under cowl fuel lines. For instance 
the following phase-in could be implemented: 30 percent in 2010, 60 
percent in 2011, and 90 percent in 2012. This would allow manufacturers 
to transition to the use of low-permeation fuel lines in an orderly 
fashion. Also, it would give them some flexibility to continue to use 
short sections of uncontrolled fuel lines, in the longer term, that are 
more difficult or costly to replace with low-permeation fuel lines. At 
some point in the future, such as 2015, we could require the use of 100 
percent low-permeation fuel lines. Manufacturers would be expected to 
target 100 percent use of low-permeation fuel lines in new engine 
designs. If the surface area percentages were weighted across a 
manufacturers entire product line of outboard engines (rather than on a 
per-engine basis), it would allow manufacturers to use 100 percent low-
permeation fuel lines on new engine designs, while making less changes 
to engines that are planned to be phased out of production.
    We also request comment on how the above program could be 
implemented given that the fuel line from the tank to the engine is 
typically installed by the boat builder while the under-cowl fuel line 
is installed by the engine manufacturer. One approach that has been 
considered is requiring the engine manufacturer to specify low-
permeation fuel line in its installation instructions beginning in 
2008. The engines would not be made available to boat builders who do 
not begin using low-permeation fuel lines in 2008.
(2) Fuel Tank Permeation Standards and Dates
    Except as noted below, we are proposing a fuel tank permeation 
standard of 1.5 g/m\2\/day for tanks intended for use in new Small SI 
equipment and Marine SI vessels based on the permeation rate of 
gasoline containing 10 percent ethanol at a test temperature of 28 
[deg]C (see Sec.  1060.103 and Sec.  1060.520). The emission standard 
is

[[Page 28169]]

based on the inside surface area of the fuel tank rather than the 
volumetric capacity because permeation is a function of surface area 
exposed to fuel. This proposed standard is consistent with that adopted 
for fuel tanks in recreational vehicles.
    We are proposing a fuel tank permeation standard of 2.5 g/m\2\/day 
for handheld equipment with structurally integrated nylon fuel tanks 
(see Sec.  1060.801 for the proposed definition of structurally 
integrated nylon fuel tanks). These fuel tanks are molded as part of 
the general structure of the equipment. In most cases, these fuel tanks 
are made of glass-reinforced nylon for strength and temperature 
resistance. These nylon constructions typically have significantly 
lower permeation rates than other plastics used for fuel tanks, such as 
high-density polyethylene; however, based on data in Chapter 5 of the 
Draft RIA the nylon constructions may not be able meet a standard of 
1.5 g/m\2\/day. Therefore, we believe a higher standard is necessary 
for these fuel tank constructions. We request comment on this separate 
permeation standards for structurally integrated fuel tanks.
    Many Small SI equipment manufacturers are currently using low-
permeation fuel tanks for products certified in California. The 
California tank permeation test procedures use a nominal test 
temperature of 40 [deg]C with California certification gasoline while 
we are proposing to require testing at 28 [deg]C with gasoline 
containing 10 percent ethanol. We are proposing to allow manufacturers 
the alternative of testing their fuel tanks at 40 [deg]C with our test 
fuel. Because permeation increases as a function of temperature, we are 
proposing an alternative standard of 2.5 g/m2/day for fuel tanks tested 
at 40 [deg]C. For structurally integrated nylon fuel tanks, the 
alternative standard at 40 [deg]C would be 4.0 g/m\2\/day.
    We consider three distinct classes of marine fuel tanks: (1) 
Portable marine fuel tanks (generally used with small outboards); (2) 
personal watercraft (PWC) fuel tanks; and (3) other installed marine 
fuel tanks (generally used with SD/I and larger outboards). The fuel 
tank permeation standards are proposed to start in 2011 for all Small 
SI equipment using Class II engines and for personal watercraft and 
portable marine fuel tanks. For Small SI equipment using Class I 
engines and for other installed marine fuel tanks, we propose to apply 
the same standard starting in 2012. Most of the marine fuel tanks with 
the later standards are produced in low volumes using rotational-molded 
cross-link polyethylene or fiberglass construction, both of which 
generally present a greater design challenge. We believe the additional 
lead time will be necessary for these fuel tanks to allow for a smooth 
transition to low-permeation designs. For Small SI equipment, these 
dates also align with the schedule for introducing the proposed Phase 3 
exhaust emission standards.
    Component manufacturers would be required to certify to the 
proposed permeation emission standard for fuel tanks (this may involve 
certification to a family emission limit above the emission standard, 
as described in Section VI.D), except in certain circumstances. 
Equipment manufacturers would need to certify that their fuel tanks 
meet the proposed emission standards if they are not already certified 
by the fuel tank manufacturer, or if they comply using emission 
credits, as described in Section VI.F. However, we are proposing that 
manufacturers of portable marine fuel tanks be required to certify that 
their products meet the new permeation standard. This is necessary 
because portable fuel tanks are not sold to boat builders for 
installation in a vessel. There is therefore no other manufacturer who 
could be treated as the manufacturer and responsible for meeting 
emission standards that apply to portable marine fuel tanks.
    For handheld equipment, we are proposing a phased-in implementation 
of the fuel tank permeation standards. Manufacturers would be required 
to meet the proposed fuel tank permeation standards in 2009 for 
products that they already certify in California (see Sec.  90.129). 
The remaining equipment, except for structurally integrated nylon fuel 
tanks and small-volume families, would be subject to the proposed tank 
permeation standards in 2010 (see Sec.  1054.110). Structurally 
integrated nylon fuel tanks would be subject to the proposed standards 
in 2011 and small-volume families would have to meet the proposed tank 
permeation standards beginning in 2013. Manufacturers would need to 
start using EPA-specified procedures starting in 2010, except that 
equipment certified using carryover data would be allowed to use data 
collected using procedures specified for compliance in California for 
model years 2010 and 2011 (see Sec.  1054.145).
    For the purpose of the proposed fuel tank permeation standards, a 
fuel cap mounted on the fuel tank is considered to be part of the fuel 
tank. We consider a fuel cap to be mounted on the fuel tank unless the 
fuel tank is designed to have a filler neck at least 12 inches long 
with the opening at least six inches above the top of the fuel tank. 
The fuel cap would therefore be included in the tank permeation 
standard and test. The cap may optionally be tested separately from the 
tank and the results combined to determine the total tank permeation 
rate (see Sec.  1060.521). Cap manufacturers could also test their caps 
and certify them separately to a separate 1.5 g/m\2\/day cap permeation 
standard. The permeation requirements apply independently of the 
diffusion standards described below, which address venting of fuel 
vapors. We are concerned that allowing certification of fuel caps could 
add complexity to the certification process. It would also add a 
measure of uncertainty in our efforts to ensure compliance with 
emission standards--for fuel tanks certified to permeation standards 
alone, it would be hard ensure that the fuel tanks in the final 
installation would be in a certified configuration with respect to 
diffusion emissions. We therefore request comment on the value to 
manufacturers of allowing fuel caps to be certified independently from 
the fuel tank. Note that a single certification fee would apply to fuel 
tanks that are certified to permeation and diffusion emission 
standards, but only if there is no optional fuel cap certification. 
With the option of fuel cap certification, a separate certification fee 
would apply to diffusion and permeation families, even if a single fuel 
tank manufacturer certifies to both standards.
(3) Diurnal Emission Standards and Dates
    We are proposing diurnal emission standards for fuel tanks intended 
for use in new Marine SI vessels (see Sec.  1045.107). We consider 
three distinct classes of marine fuel tanks: (1) Portable marine fuel 
tanks (used with small outboards); (2) personal watercraft (PWC) fuel 
tanks; and (3) other installed fuel tanks (used with SD/I and larger 
outboards). For diurnal emissions from portable fuel tanks, we are 
proposing a design requirement that the tank remain sealed up to a 
pressure of 5.0 psi, starting in the 2009 model year (see Sec.  
1060.105). We are also proposing that portable fuel tanks must continue 
to be self-sealing when disconnected from an engine.
    We are proposing a general emission standard of 0.40 g/gal/day 
based on a 25.6-32.2 [deg]C temperature profile for installed tanks. 
The applicable test procedures are described in Section VI.E.3. 
Manufacturers have expressed concerns that some very large boats stay 
in the water throughout the boating season and therefore will see a 
much smaller daily swing in fuel

[[Page 28170]]

temperatures, which corresponds with a smaller degree of diurnal 
emissions. We are proposing to address this concern with an alternative 
standard and test procedure that would apply only for nontrailerable 
boats. Using available measurements related to fuel temperatures and 
emission models to relate temperatures to projected diurnal emission 
levels, we are proposing an alternative standard of 0.16 g/gal/day 
based on a 27.6-30.2 [deg]C temperature cycle for fuel tanks installed 
in nontrailerable boats. For the purposes of this rule, we are 
proposing to define a nontrailerable boat as 26 feet or more in length, 
which is consistent with the U.S. Fish and Wildlife Service definition 
for ``nontrailerable recreational vessels'' in 50 CFR 86.12. The 
diurnal emission standards would apply starting in 2009 for PWC fuel 
tanks and in 2010 for other installed fuel tanks.
    Component manufacturers would be required to certify to the 
proposed diurnal emission standard for fuel tanks, except in certain 
circumstances. Equipment manufacturers would need to certify that their 
fuel tanks meet the proposed emission standards if they are not already 
certified by the fuel tank manufacturer, as described in Section VI.F. 
As described above for permeation standards, we are proposing to 
require manufacturers of portable marine fuel tanks to certify that 
they meet the proposed diurnal emission standards since there is no 
``equipment manufacturer'' to assume certification responsibility for 
those tanks.
    We believe the proposed requirements would achieve at least a 50 
percent reduction in diurnal emissions from PWC and other installed 
marine fuel tanks and nearly a 100 percent reduction from portable 
marine tanks. We request comment on the proposed diurnal emission 
standards for Marine SI vessels.
    It is common today for portable marine fuel tanks to maintain an 
airtight seal when the engine is not operating. These tanks typically 
have caps that are fitted with a valve that can be manually opened 
during engine operation and closed when the fuel tank is stored. 
Although this technology could be used to control diurnal emissions 
effectively, it depends on user intervention. We are proposing that 
portable fuel tanks be required to be fitted with a self-sealing vent 
rather than a manually-controlled vent. For instance, a one-way 
diaphragm valve could be used to allow air in when fuel is drawn from 
the tank (to prevent vacuum conditions), but otherwise seal the fuel 
tank. Current portable marine fuel tanks are small and designed to hold 
pressure when the manual valve is closed. We are proposing to require 
that portable marine fuel tanks be designed to maintain a seal to allow 
for pressure buildup resulting from normal temperature swings. These 
tanks should include valves that prevent a vacuum in the tank during 
engine operation which could restrict fuel flow to the engine and 
potentially stall the engine. We believe portable marine fuel tanks 
with valves that seal automatically will control diurnal emissions 
without relying on user operation. We are proposing to implement this 
design standard beginning with the 2009 model year. We request comment 
on this approach.
    Manufacturers will likely control emissions from installed marine 
fuel tanks either by sealing the fuel system up to 1.0 psi or by using 
a carbon canister in the vent line. As discussed below, we believe PWC 
manufacturers will likely seal the fuel tank with a pressure-relief 
valve while manufacturers of other boats with installed fuel tanks are 
more likely to use carbon canisters. However, either technology would 
be acceptable for either kind of installed marine fuel tank as long as 
every system meets the numerical standard applicable to the specific 
tank.
    Personal watercraft currently use sealed fuel systems for 
preventing fuel from exiting, or water from entering, the fuel tank 
during typical operation. These vessels use pressure-relief valves for 
preventing excessive positive pressure in the fuel system; the pressure 
to trigger the valve may range from 0.5 to 4.0 psi. Such fuel systems 
would also need a low-pressure vacuum relief valve to allow the engine 
to draw fuel from the tank during operation. In the 2002 proposal, we 
discussed a diurnal emission standard largely based on the use of a 
sealed system with a 1.0 psi pressure-relief valve. The Personal 
Watercraft Industry Association (PWIA) expressed support in their 
comments for this proposal. We estimate that diurnal emissions from a 
sealed system with a 1.0 psi pressure-relief valve would be about half 
that of the same system on a PWC with an open vent. For personal 
watercraft, we are proposing an implementation date of 2009 because the 
anticipated technology is widely used today.
    The National Marine Manufacturers Association (NMMA) expressed 
concern in their comments on the 2002 proposal that pressurized fuel 
tanks could lead to safety issues for larger installed fuel tanks. NMMA 
commented that these tanks would deform under pressure and that 
pressure could lead to fuel leaks. Manufacturers also commented that 
bladder fuel systems, which would not be pressurized, would be too 
expensive. At the time of the 2002 proposal, we considered the use of 
carbon canisters to control diurnal emissions, but were concerned that 
active purging would occur infrequently due to the low hours of 
operation per year seen by many boats. However, we have since collected 
data on carbon canisters showing that canisters can reduce emissions by 
more than 50 percent with passive purge that occurs during the normal 
breathing process without creating any significant pressure in the fuel 
tank. For installed marine fuel tanks, other than PWC, we are therefore 
proposing an implementation date of 2010 to allow additional lead time 
for designing and producing canisters for marine vessels.
    During the SBREFA process described in Section VI.I, NMMA expressed 
general support of the feasibility of using carbon canisters on boats. 
However, they commented that there are many small boat builders that 
may need additional time to become familiar with and install carbon 
canisters in their boats. We request comment on either a three-year 
phase-in (say 33/66/100 percent over the 2010 through 2012 model years) 
or an extra year of lead time for small businesses to comply with the 
proposed diurnal emission standards. We also request comment on which 
small business companies would be eligible for this flexibility. One 
option would be to use the SBA definition of a small boat builder which 
is based on having fewer than 500 employees. Another option would be to 
base the flexibility on the annual boat sales of the company. One issue 
with the latter approach would be the wide range of boat sizes and 
sales prices in the marine industry. With a given number of employees, 
many more small than large boats can be manufactured in a year.
    If a manufacturer uses a canister-based system to comply with the 
standard applicable to the specific tank, we are also proposing to 
require that manufacturers design their systems not to allow liquid 
gasoline to reach the canister during refueling or from fuel sloshing 
(see Sec.  1060.105). Liquid gasoline would significantly degrade the 
carbon's ability to capture hydrocarbon vapors. One example of an 
approach to protect the canister from exposure to liquid gasoline is a 
design in which the canister is mounted higher than the fuel level and 
a small orifice or a float valve is installed in the vent line to stop 
the flow of liquid gasoline to the canister.
    Several manufacturers have stated that it is common for users to 
fill their fuel tank until they see fuel coming out

[[Page 28171]]

of the vent line. In addition to being a source of hydrocarbon 
emissions, if liquid fuel were to reach a carbon canister, it would 
significantly reduce the effectiveness of the canister. Solutions for 
this problem are relatively straightforward and have been used in 
automotive applications for many years. We are therefore proposing to 
require that boat builders use good engineering judgment in designing 
fuel systems that address diurnal emission control in a way that does 
not increase the occurrence of fuel spitback or spillage during 
refueling beginning in the years specified in Table VI-1. While this 
provision is not detailed or prescriptive, it communicates a 
requirement that manufacturers appropriately take refueling design into 
account, and it allows EPA to make enforcement decisions as the 
industry establishes sound practices in this area. In addition, we are 
proposing that manufacturers would have to meet certain specifications 
with their fuel tank caps, including requirements to tether the cap to 
the equipment and designing the cap to provide physical or audible 
feedback when the vapor seal is established. Also, adding vents to a 
fuel tank would generally not be allowed. To the extent that boat 
builders certify their vessels to meet emission standards, they would 
need to describe how they meet these refueling-related requirements in 
their application for certification. If boat builders rely on certified 
components instead of applying for certification, they would need to 
keep records describing how they meet these refueling-related 
requirements; Section VI.F describes how such companies can meet 
certification requirements without applying for a certificate.
    Any increase in fuel temperature resulting from engine operation 
would cause a potential for emissions that is very similar to diurnal 
emissions. We are therefore proposing to disallow manufacturers from 
disabling their approaches for controlling diurnal emissions during 
engine operation (see Sec.  1060.105). This would ensure that any 
running loss emissions that would otherwise occur will be controlled to 
a comparable degree as diurnal emissions.
    We are not proposing diurnal emission standards for Small SI 
equipment. However, we request comment on such a requirement. We 
believe passively purging carbon canisters could reduce diurnal 
emissions by 50 to 60 percent from Small SI equipment. Active purging 
would result in even greater reductions. However, we believe some 
important issues would need to be resolved, such as cost, packaging, 
and vibration. The cost sensitivity is especially noteworthy given the 
relatively low emissions levels (on a per-equipment basis) from such 
small fuel tanks. We request comment on the appropriate level of such a 
standard and when it could be implemented.
    There are some small outboard marine engines that have fuel tanks 
directly mounted on the engine. In these cases, the fuel tank could be 
considered to be more similar to those on Small SI equipment than other 
marine fuel tanks. Typically, these outboard engines have fuel tanks on 
the order of 1-2 liters in size. Manufacturers have expressed concern 
about the practicality of using carbon canisters for these applications 
due to space constraints and durability impacts of engine handling. We 
request comment on excluding fuel tanks less than 2 liters in size that 
are mounted on outboard engines from the proposed diurnal emission 
requirements. Since it may be a viable alternative, comments should 
address the feasibility of using sealed fuel tanks with pressure relief 
in these applications. Similar to Small SI equipment, marine fuel tanks 
mounted on the engine are directly exposed to heat from the engine 
during operation. In the case where diurnal standards were not applied 
to these fuel tanks, we request comment on applying the proposed 
diffusion and running loss standards, described below, to these fuel 
tanks.
(4) Diffusion Standards and Dates
    As described above, diffusion emissions occur when vapor escapes 
the fuel tank through an opening as a result of random molecular 
motion, independent of changing temperature. Diffusion emissions can be 
easily controlled by venting fuel tanks in a way that forces fuel 
vapors to go through a long, narrow path to escape. We are proposing 
that manufacturers may choose between certifying to a performance 
standard or a design standard. Under a performance standard, we specify 
a test procedure and a maximum emission rate. Under a design standard, 
we specify certain designs that a manufacturer may use to comply with 
the standard. This standard would take effect at the same time as the 
exhaust emission standards--2011 for Class II engines and 2012 for 
Class I engines.
    We are proposing a performance standard of 0.80 g/day for diffusion 
emissions for fuel tanks intended for use in new nonhandheld Small SI 
equipment (Sec.  1060.105). This standard would not apply to a 
manufacturer who certifies using one of the four alternative design 
standards described below.
    1. We are proposing a design standard for diffusion in which the 
tank must be sealed except for a single vent line. This vent line would 
need to be at least 180 mm long and have a ratio of length to the 
square of the diameter of at least 5.0 mm-1 (127 
inches-1). For example, a vent line with 6 mm inside 
diameter would have to be at least 180 mm long to meet this design 
standard.
    2. We are proposing a second alternative design standard for 
diffusion in which vapors from a fuel tank are vented solely through a 
tortuous path through the fuel cap. Many fuel cap manufacturers use 
this cap design today to prevent fuel from splashing out through the 
vent during operation. As described in Chapter 5 of the Draft RIA, we 
tested three low-diffusion fuel caps used on Class I equipment with 
high annual sales. Based on these designs, we proposing to define a 
tortuous path fuel cap as one that is vented through a small path in 
the gasket and then around the threads where the cap screws onto the 
fuel tank. Specifically, we are proposing an average path length to 
total cross sectional area in the gasket pathways of greater than 1 
mm-1 and a vent path through at least 360[deg] of the 
threads.
    3. We are proposing a third alternative design standard for 
diffusion in which the fuel tank is sealed except for a vent through a 
carbon canister. Carbon canisters are one technology that manufacturers 
may use to meet diurnal emission standards in California.
    4. We are proposing a fourth alternative design standard for 
diffusion in which a fuel tank is sealed so that vapors may not exit 
the fuel tank. Under this design standard, it would be acceptable to 
have a pressure relief valve with an opening pressure of at least 0.5 
psi.
    We request comment on the appropriateness of setting a design 
standard for diffusion and on the designs described above. We also 
request comment on any additional diffusion data from fuel caps that 
are capable of meeting the proposed performance-based diffusion 
standard and on the design of these fuel caps. Even without the 
alternative of a design standard, we anticipate that fuel cap 
manufacturers, with a small number of designs covering a large number 
of equipment models, would be able to perform the necessary testing for 
a performance-standard without being unreasonably burdened.
    Fuel tank manufacturers would be required to certify that their 
products limit venting sufficiently to meet the proposed diffusion 
emission standard, except in certain circumstances. Fuel

[[Page 28172]]

cap manufacturers may optionally certify their fuel caps to the 
diffusion emission standard, in which case they would become subject to 
all the compliance requirements related to the standards, including 
certification. Equipment manufacturers would need to certify that their 
fuel tanks meet the proposed emission standards if they are not already 
certified by the fuel tank manufacturer, as described in Section VI.F.
    We are also proposing that equipment manufacturers subject to 
diffusion emission standards must ensure that the fuel cap is tethered 
to the fuel tank or the equipment to prevent it from being accidentally 
misplaced (see Sec.  1060.101). A missing fuel tank cap would bypass 
any design intended to control these losses and could lead to very high 
emission rates. Fuel cap or fuel tank manufacturers could address this 
as part of their component certification. If this is not part of the 
component certification, an equipment manufacturer would need to 
describe how it meets the tethering requirement in its application for 
certification.
    We are not proposing diffusion standards for handheld equipment. 
Handheld equipment use fuel caps that are either sealed or have 
tortuous venting pathways to prevent fuel from spilling during 
operation. We believe these fuel cap designs limit diffusion emissions 
sufficiently that handheld equipment already meet the proposed 
standard. In addition, we are not proposing diffusion standards for 
Marine SI vessels. The diurnal emission standard for Marine SI vessels 
will lead manufacturers to adopt technologies that automatically limit 
diffusion losses, so there is no need to propose a separate diffusion 
standard for those systems. Similarly, we would not finalize the 
proposed diffusion standard if we adopt a diurnal emission standard for 
Small SI equipment. We request comment on the proposed diffusion 
standard for nonhandheld equipment and whether it should apply to 
handheld equipment and marine vessels as well.
(5) Running Loss Emission Standards and Dates
    We are proposing standards to control running loss emissions from 
nonhandheld Small SI equipment beginning in the same year as the 
proposed Phase 3 exhaust emission standards--2012 for Class I engines 
and 2011 for Class II engines (see Sec.  1060.104). Equipment 
manufacturers would need to certify that their equipment models meet 
the proposed running loss requirements since component certification is 
not practical.
    We have measured fuel temperatures and found that some types of 
equipment experience significant fuel heating during engine operation. 
This was especially true for fuel tanks mounted on or near the engine. 
This occurs in many types of Small SI equipment.
    It would be very difficult to define a measurement procedure to 
consistently and accurately quantify running losses. Also, a 
performance standard with such a procedure would introduce a 
challenging testing requirement for hundreds of small-volume equipment 
manufacturers. Moreover, we believe there are several different design 
approaches that will reliably and effectively control running losses. 
We are therefore not proposing to control running losses using the 
conventional approach of establishing a procedure to measure running 
losses and adopting a corresponding emission standard. Manufacturers 
could choose from one of the following approaches to meet this 
requirement:
     Vent running loss fuel vapors from the fuel tank to the 
engine's intake manifold in a way that burns the fuel vapors in the 
engine instead of venting them to the atmosphere. The use of an 
actively purged carbon canister would qualify under this approach.
     Use a bladder to minimize fuel vapor volume in a sealed 
fuel tank.
     Design the equipment so that fuel temperature does not 
rise more than 8 [deg]C during normal operation. Such a design may use 
insulation or forced cooling to minimize temperature increases. This 
would require measuring fuel temperatures to show that each covered 
equipment configuration does not exceed the temperature threshold (see 
Sec.  1060.535).
     Show that the equipment qualifies as wintertime equipment.
    We believe any of these approaches will ensure that manufacturers 
will be substantially controlling running losses, either by preventing 
or managing running loss vapors. While none of these approaches are 
expected to require extensive design changes or lead time, any 
manufacturer choosing the option to vent running loss fuel vapors into 
the engine's intake manifold would need to make this change in 
coordination with the engine design. As a result, we believe it is 
appropriate to align the timing of the running loss standards with the 
introduction of the proposed Phase 3 standards.
    We request comment on the proposed running loss requirement for 
nonhandheld Small SI equipment. We also request comment on any other 
design approaches that will reliably and effectively control running 
losses. Examples of other approaches may be to seal the fuel tank for 
pressures up to 3.5 psi or, for equipment that does not include fuel 
recirculation, locate the fuel tank at least 12 inches away from the 
engine and other heat sources (such as exhaust pipes, hydraulic lines, 
etc.).
    We are not proposing to apply the running loss requirements to 
handheld Small SI engines. We believe running loss emission standards 
should not apply to handheld engines at this time because the likely 
approach to controlling running losses could require that manufacturers 
revisit their design for controlling exhaust emissions. As described 
above, we are not proposing to change the exhaust emission standards 
for handheld engines in this rulemaking. In addition, there are some 
technical challenges that would require further investigation. For 
example, the compact nature of the equipment makes it harder to isolate 
the fuel tank from the engine and the multi-positional nature of the 
operation may prevent a reliable means of venting fuel vapors into the 
intake manifold while the engine is running. We request comment on the 
appropriateness of requiring manufacturers to address running loss 
emissions from handheld engines.
    Furthermore, we are not proposing to apply running loss 
requirements to Marine SI engines. Installed marine fuel tanks are 
generally not mounted near the engine or other heat sources so running 
losses should be very low. A possible exception to this is personal 
watercraft since they are designed with the fuel tank closer to the 
engine. However, under the proposed standard for controlling diurnal 
emissions, we expect that manufacturers will design their fuel tanks to 
stay pressurized up to 1 psi. This would also help control running loss 
emissions. We request comment on applying running loss controls to 
Marine SI engines. In particular, we request comment on the possibility 
that other design configurations would have higher running loss 
emissions. One example may be outboard applications in which a fuel 
tank is mounted directly on the engine.
(6) Requirements Related to Refueling
    Refueling spitback and spillage emissions represent a substantial 
additional amount of fuel evaporation that contributes to overall 
emissions from equipment with gasoline-fueled engines. We are not 
proposing measurement procedures with corresponding emission standards 
to address these emission sources. However, we believe equipment 
manufacturers can take significant steps

[[Page 28173]]

to address these refueling issues by incorporating sound practices into 
their equipment designs. For example, designing a marine filler neck 
with a horizontal segment near the fuel inlet will almost inevitably 
lead to high levels of spillage since fuel flow will invariably reach 
the nozzle, leading to substantial fuel flow out of the fuel system. In 
contrast, designing for automatic shutoff would prevent this. Also, 
maintaining a vertical orientation of the filler neck would allow the 
fuel to flow back into the filler neck and into the tank after the 
nozzle shuts off.
    For Small SI equipment, designing fuel inlets that are readily 
accessible and large enough to see the rising fuel level (either 
through the tank wall or the fuel inlet) will substantially reduce 
accidental spillage during refueling. We are therefore proposing to 
require that equipment manufacturers design and build their equipment 
such that operators could reasonably be expected to fill the fuel tank 
without spitback or spillage during the refueling event (see Sec.  
1060.101). This proposed requirement mirrors the following requirement 
recently adopted with respect to portable fuel containers (72 FR 8428, 
February 26, 2007):

    You are required to design your portable fuel containers to 
minimize spillage during refueling to the extent practical. This 
requires that you use good engineering judgment to avoid designs 
that will make it difficult to refuel typical vehicle and equipment 
designs without spillage. (40 CFR 59.611(c)(3))

    While the proposed requirement is not as objective and quantifiable 
as the other standards and requirements we are proposing, we believe 
this is important, both to set a requirement for manufacturers in 
designing their products and to give EPA the ability to require 
manufacturers to select designs that are consistent with good 
engineering practice regarding effective refueling strategies. To the 
extent that equipment manufacturers and boat builders certify their 
products to emission standards, they would need to describe how they 
meet this refueling-related requirement in their application for 
certification. If boat builders rely on certified components instead of 
applying for certification, they would need to keep records describing 
how they meet this refueling-related requirement; Section VI.F 
describes how such companies can meet certification requirements 
without applying for a certificate. We request comment on this approach 
to addressing refueling emissions from nonroad spark-ignition engines. 
We also request comment on the possibility of relying on current or 
future published industry standards to establish designs for equipment 
and fueling containers that minimize refueling emissions under normal 
in-use conditions.
    Spitback and spillage are a particular concern for gasoline-fueled 
boats. Marine operators have reported that relatively large quantities 
of gasoline are released into the marina environment during refueling 
events. The American Boat and Yacht Council (ABYC) has a procedure in 
place to define a standard practice to address refueling. However, this 
procedure calls for testing by refueling up to a 75 percent fill level 
at a nominal flow rate of 5 gallons per minute. This procedure is 
clearly not consistent with prevailing practices and is not effective 
in preventing spills. We believe the most effective means of addressing 
this problem is for ABYC to revise their test procedure to reflect 
current practices. Specifically, we would recommend a procedure in 
which the marine fuel tank is filled at flow rates between 5 and 20 
gallons per minute until automatic shutoff occurs.
    A variety of technological solutions are available to address 
spitback and spillage from marine vessels. The simplest would be a 
system much like is used on cars. A small-diameter tube could run along 
the filler neck from the top of the tank to a point near the top of the 
filler neck. Once liquid fuel would reach the opening of the filler 
neck and the extra tube, the fuel would go faster up the small-diameter 
tube and trigger automatic shutoff before the fuel climbs up the filler 
neck. This design would depend on the user to use the equipment 
properly and may not be fully effective, for example, with long filler 
necks and low refueling rates. An alternative design would involve a 
snug fit between the nozzle's spout and the filler neck, which would 
allow for a tube to run from a point inside the tank (at any 
predetermined level) directly to the shutoff venturi on the spout. The 
pressure change from the liquid fuel in the tank reaching the tube's 
opening would trigger automatic shutoff of the nozzle. This system 
would prevent overflowing fuel without depending on the user. These are 
just two of several possible configurations that would address fuel 
spillage from marine vessels.
    We request comment on the degree of fuel spillage with current 
technologies and practices with marine vessels. We request comment on 
the potential for ABYC standards to address fuel spillage or on the 
need for EPA to adopt such procedures and standards. We request comment 
on the specific procedures that would be appropriate for measuring 
spitback and spillage. Finally, we request comment on adopting 
provisions such as those in 40 CFR 80.22 to regulate the dimensions of 
refueling nozzles for marine applications, including a specification of 
a nominal nozzle diameter of 1.1870.010 inches and nominal 
venturi placement \5/8\ inch from the terminal end of the nozzle.
(7) Summary Table of Proposed Evaporative Emission Standards
    Table VI-1 summarizes the proposed standards and implementation 
dates discussed above for evaporative emissions from Small SI equipment 
and Marine SI vessels. Where a standard does not apply to a given class 
of equipment, ``NA'' is used in the table to indicate ``not 
applicable.''

                                       Table VI.-1.--Proposed Evaporative Emission Standards and Model Year Dates
--------------------------------------------------------------------------------------------------------------------------------------------------------
         Standard/ category             Hose  permeation        Tank  permeation            Diurnal               Diffusion             Running loss
--------------------------------------------------------------------------------------------------------------------------------------------------------
                                                                   Proposed Standards
--------------------------------------------------------------------------------------------------------------------------------------------------------
Standard level.....................  15 g/m2 /day..........  1.5 g/m2 /day.........  0.40 g/gal/day.......  0.80 g/day...........  Design standard.
--------------------------------------------------------------------------------------------------------------------------------------------------------
                                                        Implementation Dates: Small SI Equipment
--------------------------------------------------------------------------------------------------------------------------------------------------------
Handheld...........................  2012 a b..............  2009-2013 c d.........  NA...................  NA...................  NA.
--------------------------------------------------------------------------------------------------------------------------------------------------------
Class I............................  2008..................  2012..................  NA...................  2012 g...............  2012.
Class II...........................  2008..................  2011..................  NA...................  2011 g...............  2011.

[[Page 28174]]

 
                                                          Implementation Dates: Marine Vessels
--------------------------------------------------------------------------------------------------------------------------------------------------------
Portable tanks.....................  2009..................  2011..................  2009 e...............  NA...................  NA.
PWC................................  2009..................  2011..................  2009.................  NA...................  NA.
Other installed tanks..............  2009..................  2012..................  2010 f...............  NA...................  NA.
--------------------------------------------------------------------------------------------------------------------------------------------------------
\a\ 2013 for small-volume families and cold weather equipment.
\b\ Fuel line permeation standard of 175 g/m2 /day for cold-weather equipment.
\c\ 2.5 g/m2 /day for structurally integrated nylon fuel tanks.
\d\ 2009 for families certified in California, 2013 for small-volume families, 2011 for structurally integrated nylon fuel tanks, and 2010 for remaining
  families.
\e\ Design standard.
\f\ Fuel tanks installed in nontrailerable boats (>=26 ft. in length) may meet a standard of 0.16 g/gal/day over an alternative test cycle.
\g\ Alternatively, may meet a design standard.

D. Emission Credit Programs

    A common feature of mobile source emission requirements is an 
emission credit program that allows manufacturers to generate emission 
credits based on certified emission levels for engine families that are 
more stringent than the standard. See Section VII for background 
information and general provisions related to emission credit programs.
    We believe it is appropriate to consider compliance based on 
emission credits relative to permeation standards for fuel lines used 
with handheld engines and for fuel tanks used in all applications. As 
described above, the emission standards apply to the fuel tanks and 
fuel lines directly, such that we would generally expect component 
manufacturers to certify their products. However, we believe it is best 
to avoid placing the responsibility for demonstrating a proper emission 
credit balance on component manufacturers for three main reasons. 
First, it is in many cases not clear whether these components will be 
produced for one type of application or another. Component 
manufacturers might therefore be selling similar products into 
different applications that are subject to different standards--or no 
standards at all. Component manufacturers may or may not know in which 
application their products will be used. Second, there will be 
situations in which equipment manufacturers and boat builders take on 
the responsibility for certifying components. This may be the result of 
an arrangement with the component manufacturer, or equipment 
manufacturers and boat builders might build their own fuel tanks. We 
believe it would be much more difficult to manage an emission credit 
program in which manufacturers at different places in the manufacturing 
chain would be keeping credit balances. There would also be a 
significant risk of double-counting of emission credits. Third, most 
component manufacturers would be in a position to use credits or 
generate credits, but not both. Equipment manufacturers and boat 
builders are more likely to be in a position where they would keep an 
internal balance of generating and using credits to meet applicable 
requirements. Our experience with other programs leads us to believe 
that an emission credit program that depends on trading is not likely 
to be successful.
    We are therefore proposing emission credit provisions in which 
equipment manufacturers and boat builders keep a balance of credits for 
their product line. Equipment manufacturers and boat builders choosing 
to comply based on emission credits would need to certify all their 
products that either generate or use emission credits. Component 
manufacturers would be able to produce their products with emission 
levels above or below applicable emission standards but would not be 
able to generate emission credits and would not need to maintain an 
accounting to demonstrate a balance of emission credits.
    We are aware that some component manufacturers would be making 
products that generate emission credits that would belong to equipment 
manufacturers or boat builders. Equipment manufacturers or boat 
builders could in turn use those emission credits to enable them to buy 
components from different competing component manufacturers. This would 
potentially put fuel tank manufacturers producing low-FEL products at a 
competitive disadvantage with other manufacturers producing high-FEL 
fuel tanks. We request comment on the best approach to setting up an 
ABT program. We specifically request comment on special provisions that 
may be appropriate to address these competitiveness issues for 
component manufacturers.
(1) Averaging, Banking, and Trading for Nonhandheld Equipment and 
Marine Vessels
    We are proposing averaging, banking, and trading (ABT) provisions 
for fuel tank permeation from nonhandheld Small SI equipment and Marine 
SI vessels (see subpart H in parts 1045 and 1054). See the following 
section for similar provisions for handheld Small SI equipment.
    We are aware of certain control technologies that would allow 
manufacturers to produce fuel tanks that reduce emissions more 
effectively than we would require. These technologies may not be 
feasible or practical in all applications, but we are proposing to 
allow equipment manufacturers using such low-emission technologies to 
generate emission credits. In other cases, an equipment manufacturer 
may want to or need to use emission credits that would allow for fuel 
tanks with permeation rates above the applicable standards. Equipment 
manufacturers would quantify positive or negative emission credits by 
establishing a Family Emission Limit (FEL) to define the applicable 
emission level, then factoring in sales volumes and useful life to 
calculate a credit total. This FEL could be based on testing done 
either by the component manufacturer or the equipment manufacturer. 
Through averaging, these emission credits could be used by the same 
equipment manufacturer to offset other fuel tanks in the same model 
year that do not have control technologies that control emissions to 
the level of the standard. Through banking, such an equipment 
manufacturer could use the emission credits in later model years to 
offset high-emitting fuel tanks. The emission credits could also be 
traded to another equipment manufacturer to offset that company's high-
emitting fuel tanks.

[[Page 28175]]

    We believe an ABT program is potentially very advantageous for fuel 
tanks because of the wide variety of tank designs. The geometry, 
materials, production volumes, and market dynamics for some fuel tanks 
are well suited to applying emission controls but other fuel tanks pose 
a bigger challenge. The proposed emission credit program allows us to 
set a single standard that applies broadly without dictating that all 
fuel tanks be converted to use low-permeation technology at the same 
time.
    We are requesting comment on one particular issue. We are not 
proposing to limit the life of evaporative emission credits under the 
proposed banking program. However, we are concerned that this could 
result in a situation where credits generated by a fuel tank sold in a 
model year are not used until many years later when the fuel tanks 
generating the credits have been scrapped and are no longer part of the 
fleet. EPA believes there may be value to limiting the use of credits 
to the period that the credit-generating fuel tanks exist in the fleet. 
For this reason, EPA requests comment on limiting the lifetime of the 
credits generated under the proposed evaporative emission ABT program 
to five years. The five-year period is consistent with the proposed 
useful life for fuel tank evaporative emissions.
    We are proposing not to allow manufacturers to generate emission 
credits by using metal fuel tanks. These tanks would have permeation 
rates well below the standard, but there is extensive use of metal 
tanks today, so it would be difficult to allow these emission credits 
without undercutting the stringency of the standard and the expected 
emission reductions from the standard.
    Emission control technologies and marketing related to portable 
marine fuel tanks are quite different than for installed tanks. Since 
these fuel tanks are not installed in vessels that are subject to 
emission standards, the fuel tank manufacturer would need to take on 
the responsibility for certification. As a result, we would treat these 
companies as both component manufacturer and equipment manufacturer 
with respect to their portable fuel tanks. As described above, we are 
proposing that component manufacturers not be responsible for 
compliance as part of an emission credit program. We would expect all 
portable fuel tank manufacturers to also make nonportable fuel tanks, 
which would again lead to a confusing combination of manufacturers 
maintaining credit balances to demonstrate compliance. In addition, 
most if not all portable fuel tanks are made using high-density 
polyethylene in a blow-molding process. The control technologies for 
these tanks are relatively straightforward and readily available so we 
do not anticipate that these companies will need emission credits to 
meet the proposed standards. We are therefore proposing to require 
portable marine fuel tanks to meet emission standards without an 
emission credit program.
    We are proposing not to allow cross-trading of emission credits 
between Small SI equipment and Marine SI vessels. The proposed 
standards are intended to be technology-forcing for each equipment 
category. We are concerned that cross-trading may allow marginal 
credits in one area to hamper technological advances in another area. 
We are also proposing not to allow credit exchanges with Small SI 
equipment certified in California because California has its own 
emission standards for these products. Similarly, if California ARB 
adopts different evaporative requirements or separate ABT provisions 
for Marine SI vessels, we would not allow credit exchanges with marine 
vessels certified in California. These restrictions are consistent with 
our existing ABT programs. We also would not allow credit exchanges 
between handheld and nonhandheld equipment or between Class I and Class 
II equipment. We are concerned that cross trading between these 
equipment types could give an unfair competitive advantage to equipment 
manufacturers with broader product lines. We request comment regarding 
whether the competitive nature of the market warrants such a 
restriction in cross-trading between Class I and Class II equipment.
    In the early years of the ABT program we are proposing not to have 
an FEL cap. This would give manufacturers additional time to use 
uncontrolled fuel tanks, primarily in small-volume applications, until 
they could convert their full product lines to having fuel tanks with 
permeation control. After an initial period of three years after the 
implementation date of the fuel tank standards, we are proposing an FEL 
cap of 5.0 g/m2 /day (8.3 g/m2 /day if tested at 
40 [deg]C). For Class II equipment, portable marine fuel tanks, and 
personal watercraft, the FEL cap would begin in 2014. For Class I 
equipment, handheld equipment, and other installed marine fuel tanks, 
the FEL cap would begin in 2015. See Sec.  1045.107 and Sec.  1054.110. 
For small volume, Small SI equipment families, we are proposing an FEL 
cap of 8.0 g/m2 /day (13.3 g/m2 /day if tested at 
40 [deg]C). The purpose of the FEL cap would be to prevent the long-
term production of fuel tanks without permeation control, while still 
providing regulatory flexibility. We request comment on the level of 
the FEL that would be necessary to achieve this goal.
    While the FEL cap is intended to require manufacturers to move 
toward widespread use of emission control technologies, we are aware of 
technologies that have measured emission levels between the proposed 
standard and the proposed FEL cap. As a result, the effect of an FEL 
cap may be that there will be little or no use of emission credits as a 
compliance strategy once the FEL cap applies. We request comment on the 
usefulness of maintaining an ABT program after we implement an FEL cap.
    We are proposing that emission credits under the tank permeation 
standards would be calculated using the following equation: Credits 
[grams] = (Standard - FEL) x useful life [years] x 365 days/year x 
inside surface area [m2]. Both the standard and the FEL are 
in units of g/m2 /day based on testing at 28 [deg]C.
    As discussed earlier, we are proposing an alternative standard for 
tank permeation testing performed at 40 [deg]C. Because permeation is 
higher at this temperature than the primary test temperature, emissions 
credits and debits calculated at this test temperature would be 
expected to be higher as well. An FEL 10 percent below the standard 
would generate 0.15 grams of credit for the primary standard and 0.25 
grams of credit for the alternative standard. Therefore, we are 
proposing that credits and debits that are calculated based on the 
alternative standard be adjusted using a multiplicative factor of 0.6 
(1.5/2.5 = 0.6).
    We request comment on the need for averaging, banking and trading 
for fuel tanks and on the specific provisions proposed above.
(2) Averaging, Banking, and Trading Program for Handheld Equipment
    We are proposing an ABT program for handheld equipment that would 
include fuel tanks and fuel lines. Under this program, a manufacturer 
would be able to use credits from fuel tanks to offset debits from fuel 
lines, or vice versa. This category of equipment generally involves 
very short sections of fuel lines, which are often made using complex, 
injection-molded designs. We believe an ABT program would help handheld 
equipment manufacturers meet fuel line permeation standards sooner than 
would otherwise be possible.

[[Page 28176]]

    As discussed earlier, we are proposing a higher standard level of 
2.5 g/m2 /day for structurally integrated handheld fuel 
tanks. This standard is intended to reflect the measured permeation 
rates and characteristics of materials used in these fuel tanks and 
manufacturer concerns regarding uncertainty about the permeation rates 
from tanks used in the wider range of products and the lack of 
definitive control strategies to reduce emissions while meeting other 
product requirements. A similar issue exists for cold-weather fuel 
lines, for which we are proposing a less stringent permeation standard 
of 175 g/m2 /day to address uncertainty associated with the 
availability of appropriate low-permeation cold-weather materials in 
the time frame of the new standards. We are concerned that windfall 
credits that may be generated for these applications if products are 
produced that are below the adjusted standards, but do not meet the 
primary standards for fuel tanks and fuel lines. To address this issue, 
we are proposing that credits would only be earned below 1.5 g/
m2 /day for fuel tanks and below 15 g/m2 /day for 
fuel lines on handheld equipment. To promote early introduction of low-
permeation products, we are proposing to allow manufacturers to be able 
to earn credits on this basis even before the permeation standards go 
into effect. Credit use would be calculated based on the applicable 
standards. Emission credits would otherwise be calculated using the 
same equation described in Section VI.D.1 above.
    Both the fuel line and fuel tank standards are in units of g/
m2 /day. However, fuel line testing is performed at 23 
[deg]C while tank testing is performed at 28 [deg]C. Because permeation 
tends to increase with increases in temperature, we request comment 
regarding whether the credits should be adjusted to account for 
temperature. This adjustment would be smaller than the adjustment 
described above for a 28 [deg]C versus 40 [deg]C test.
    For non-structurally integrated fuel tanks, we are proposing to 
apply an FEL cap of 5.0 g/m2 /day (8.3 g/m2 /day 
if tested at 40[deg]C) beginning in 2015. For structurally integrated 
fuel tanks we are proposing an FEL cap of 3.0 g/m2 /day (5.0 
g/m2 /day if tested at 40 [deg]C) in 2015. We believe this 
cap gives adequate flexibility for manufacturers to address variability 
in the permeation rates of these fuel tanks. For small volume, Small SI 
equipment families (including handheld and nonhandheld equipment), we 
are proposing a long term FEL cap of 8.0 g/m2 /day (13.3 g/
m2 /day if tested at 40[deg]C) to provide additional 
regulatory flexibility where costs cannot be spread over high 
production volumes. We request comment on the need for continuing an 
ABT program once there is an FEL cap, as described for nonhandheld 
equipment above.
(3) Other Evaporative Sources
    We are not proposing an emission credit program for other 
evaporative sources. We believe technologies are readily available to 
meet the applicable standards for fuel line permeation, diurnal 
emissions and diffusion emissions (see Section VI.H.). The exception to 
this is for fuel lines on handheld equipment as discussed above. In 
addition, the diurnal emission standards for portable marine fuel tanks 
and PWC fuel tanks are largely based on existing technology so any 
meaningful emission credit program with the proposed standards would 
result in windfall credits. The running loss standard is not based on 
emission measurements and refueling-related requirements are based on 
design specifications only, so it is not appropriate or even possible 
to calculate emission credits.
(4) Early-Allowance Programs
    Manufacturers may in some cases be able to meet the proposed 
emission standards earlier than we would require. We are proposing 
provisions for equipment manufacturers using low-emission evaporative 
systems early to generate allowances before the standards apply. These 
early allowances could be used, for a limited time, after the 
implementation date of the standards to sell equipment or fuel tanks 
that have emissions above the standards. We are proposing two types of 
allowances. The first is for Small SI equipment as a whole where for 
every year that a piece of equipment is certified early, another piece 
of equipment could delay complying with the proposed standards by an 
equal time period beyond the proposed implementation date. The second 
is similar but would be just for the fuel tank rather than the whole 
equipment (Small SI or Marine SI). Equipment or fuel tanks certified 
for the purposes of generating early allowances would be subject to all 
applicable requirements. These allowances are similar to the emission 
credit program elements described above but they are based on counting 
compliant products rather than calculating emission credits. 
Establishing appropriate credit calculations would be difficult because 
the early compliance is in some cases based on products meeting 
different standards using different procedures.
(a) Nonhandheld Small SI Equipment
    Many Small SI equipment manufacturers are currently certifying 
products to evaporative emission standards in California. The purpose 
of the proposed early-allowance program is to provide an incentive for 
manufacturers to begin selling low-emission products nationwide. We are 
proposing to give allowances to manufacturers for equipment meeting the 
California evaporative emission standards that are sold in the United 
States outside of California and are therefore not subject to 
California's emission standards. Manufacturers would need to have 
California certificates for these equipment types. See Sec.  1054.145.
    Allowances could be earned in any year before 2012 for Class I 
equipment and before 2011 for Class II equipment. We are proposing that 
the allowances may be used through the 2014 model year for Class I and 
through the 2013 model year for Class II equipment. We are proposing 
not to allow trading of allowances between Class I and Class II. To 
keep this program simpler, we are not proposing to adjust the 
allowances based on the anticipated emission rates from the equipment. 
Therefore, we believe it is necessary to at least distinguish between 
Class I and Class II equipment. We request comment on the early 
allowance program described above for nonhandheld Small SI equipment.
(b) Fuel Tanks
    We are also proposing an early-allowance program for nonhandheld 
Small SI equipment for fuel tanks (see Sec.  1054.145). This program 
would be similar to the program described above for equipment 
allowances, except that it would be for fuel tanks only. We would 
accept California-certified configurations. Allowances could be earned 
prior to 2011 for Class II equipment and prior to 2012 for Class II 
equipment; allowances could be used through 2013 for Class II equipment 
and through 2014 for Class II equipment. Allowances would not be 
exchangeable between Class I and Class II equipment. See Section V.E.3 
for a description of how this provision would interact with the 
proposed transition program for equipment manufacturers.
    The proposed early-allowance program for marine fuel tanks would be 
similar except that there are no California standards for these tanks 
(see Sec.  1045.145). Manufacturers certifying early to the proposed 
fuel tank permeation standards would be able to earn allowances that 
they could use to

[[Page 28177]]

offset high-emitting fuel tanks after the proposed standards go into 
place. We are proposing not to allow cross-trading of allowances 
between portable fuel tanks, personal watercraft, and other installed 
fuel tanks. Each of these categories includes significantly different 
tank sizes and installed tanks have different implementation dates and 
are expected to use different permeation control technology. For 
portable fuel tanks and personal watercraft, allowances could be earned 
prior to 2011 and used through the 2013 model year. For other installed 
tanks, allowances could be earned prior to 2012 and used through the 
2014 model year.

E. Testing Requirements

    Compliance with the emission standards is determined by following 
specific testing procedures. This section describes the proposed test 
procedures for measuring fuel line permeation, fuel tank permeation, 
diurnal emissions, and diffusion emissions. We also describe 
measurement procedures related to running loss emissions. As discussed 
in Section VI.H, we are proposing design-based certification as an 
alternative to testing for certain standards.
(1) Fuel Line Permeation Testing Procedures
    We are proposing that fuel line permeation be measured at a 
temperature of 23  2 [deg]C using a weight-loss method 
similar to that specified in SAE J30 \85\ and J1527 \86\ recommended 
practices (see Sec.  1060.515). We are proposing two modifications to 
the SAE recommended practice. The first modification is for the test 
fuel to contain ethanol; the second modification is to require 
preconditioning of the fuel line through a fuel soak. These 
modifications are described below and are consistent with our current 
requirements for recreational vehicles.
---------------------------------------------------------------------------

    \85\ Society of Automotive Engineers Surface Vehicle Standard, 
``Fuel and Oil Hoses,'' SAE J30, June 1998 (Docket EPA-HQ-OAR-2004-
0008-0176).
    \86\ SAE Recommended Practice J1527, ``Marine Fuel Hoses,'' 
1993, (Docket EPA-HQ-OAR-2004-0008-0195-0177).
---------------------------------------------------------------------------

(a) Test Fuel
    The recommended practice in SAE J30 and J1527 is to use ASTM Fuel C 
(defined in ASTM D471-98) as a test fuel. We are proposing to use a 
test fuel containing 10 percent ethanol. We believe the test fuel must 
contain ethanol because it is commonly blended into in-use gasoline and 
because ethanol substantially increases the permeation rates for many 
materials.
    Specifically, we are proposing to use a test fuel of ASTM Fuel C 
blended with 10 percent ethanol by volume (CE10).\87\ Manufacturers 
have expressed support of this test fuel because it is a consistent 
test fluid compared to gasoline and because it is widely used today by 
industry for permeation testing. In addition, most of the data used to 
develop the proposed fuel line permeation standards were collected on 
this test fuel. This fuel is allowed today as one of two test fuels for 
measuring permeation from fuel lines under the recreational vehicle 
standards.
---------------------------------------------------------------------------

    \87\ ASTM Fuel C is a mix of equal parts toluene and isooctane. 
We refer to gasoline blended with ethanol as E10.
---------------------------------------------------------------------------

    We request comment on allowing permeation testing using EPA 
certification gasoline (known as indolene and specified in 40 CFR 
1065.710) blended with 10 percent ethanol as the test fuel (IE10). This 
test fuel is also specified in the recreational vehicle standards and 
has the advantage of being more similar to in-use fuel than CE10. Based 
on data contained in Chapter 5 of the Draft RIA, most materials used in 
fuel line constructions have lower permeation rates on IE10 than CE10. 
Because the proposed standards are based primarily on data collected 
using CE10 as a test fuel, we also request comment on how the level of 
the standard would need to be adjusted for testing performed on IE10.
(b) Preconditioning Soak
    The second difference from weight-loss procedures in SAE practices 
is in fuel line preconditioning. We believe the fuel line should be 
preconditioned with an initial fuel fill followed by a long enough soak 
to ensure that the permeation rate has stabilized. We are proposing a 
soak period of four to eight weeks at 23  5 [deg]C. 
Manufacturers should use the longer soak period as necessary to achieve 
a stabilized permeation rate for a given fuel line design, consistent 
with good engineering judgment. For instance, thick-walled marine fuel 
line may take longer to reach a stable permeation rate than the fuel 
line used in Small SI equipment. After this fuel soak, the fuel 
reservoir and fuel line would be drained and immediately refilled with 
fresh test fuel prior to the weight-loss test. We request comment on 
the need to require a longer fuel soak, especially for marine lines.
(c) Alternative Approaches
    We also propose to allow permeation measurements using alternative 
equipment and procedures that provide equivalent results (see Sec.  
1060.505). To use these alternative methods, manufacturers would first 
need to get our approval. Examples of alternative approaches that we 
anticipate manufacturers may use are the recirculation technique 
described in SAE J1737 or enclosure-type testing such as in 40 CFR part 
86.\88\ Note that the proposed test fuel, test temperatures, and 
preconditioning soak described above would still apply. Because 
permeation increases with temperature we would accept data collected at 
higher temperatures (greater than 23 [deg]C) for a demonstration of 
compliance.
---------------------------------------------------------------------------

    \88\ SAE Recommended Practice J1737, ``Test Procedure to 
Determine the Hydrocarbon Losses from Fuel Tubes, Hoses, Fittings, 
and Fuel Line Assemblies by Recirculation,'' 1997, (Docket EPA-HQ-
OAR-2004-0008-0178).
---------------------------------------------------------------------------

    For portable marine fuel tanks, the fuel line assembly from the 
engine to the fuel tank typically includes two sections of fuel line 
with a primer bulb in-between and quick-connect assemblies on either 
end. We are proposing a provision to allow manufacturers to test the 
full assembly as a single fuel line to simplify testing for these fuel 
line assemblies (see Sec.  1060.102). This gives the manufacturer the 
flexibility to use a variety of materials as needed for performance 
reasons while meeting the fuel line permeation standard for the fully 
assembled product. Measured values would be based on the total measured 
permeation divided by the total internal surface area of the fuel line 
assembly. However, where it is impractical to calculate the internal 
surface area of individual parts of the assembly, such as a primer 
bulb, we would allow a simplified calculation that treats the full 
assembly as a straight fuel line. This small inaccuracy would cause 
reported emission levels (in g/m\2\/day) to be slightly higher so it 
would not jeopardize a manufacturer's effort to demonstrate compliance 
with the applicable standard.
    We request comment on the above approaches for fuel line permeation 
testing and on the proposed test fuel.
(2) Fuel Tank Permeation Testing Procedures
    The proposed test procedure for fuel tank permeation includes 
preconditioning, durability simulation, and a weight-loss permeation 
test (see Sec.  1060.520). The preconditioning and the durability 
testing may be conducted

[[Page 28178]]

simultaneously; manufacturers would put the tank through durability 
testing while the tank is undergoing its preconditioning fuel soak to 
reach a stabilized permeation level. We request comment on the proposed 
tank permeation test procedures and options.
(a) Test Fuel
    Similar to the proposed fuel line testing procedures, we are 
proposing to use a test fuel containing 10 percent ethanol to help 
ensure in-use emission reductions with the full range of in-use fuels. 
We are proposing to specify IE10 as the test fuel; this is made up of 
90 percent certification gasoline and 10 percent ethanol (see 40 CFR 
1065.710). This is the same test fuel specified for testing fuel tanks 
for recreational vehicles. In addition, IE10 is representative of in-
use test fuels. We are proposing that Fuel CE10 may be used as an 
alternative test fuel. Data in Chapter 5 of the Draft RIA suggest that 
permeation tends to be somewhat higher on CE10 than IE10, so testing on 
CE10 should be an acceptable demonstration of compliance. We request 
comment on the proposed test fuels.
    We included a provision allowing recreational vehicle manufacturers 
to perform emission measurements after preconditioning using IE10. This 
allowance has created substantial confusion and necessitated including 
additional provisions to prevent manufacturers from exercising the test 
option in a way that undermines the objective of maintaining a 
procedure that accounts for the effect of ethanol. As a result, we 
believe it is appropriate to propose a test procedure for Small SI 
equipment and Marine SI vessels that maintains a consistent approach by 
including ethanol in the test fuel for both preconditioning and 
emission measurements. We request comment on this approach.
(b) Preconditioning Fuel Soak
    Before testing fuel tanks for permeation, the fuel tank must be 
preconditioned by allowing it to sit with fuel inside until the 
hydrocarbon permeation rate has stabilized. Under this step, we are 
proposing that the fuel tank be filled with test fuel and soaked--
either for 20 weeks at 28  5 [deg]C or for 10 weeks at 43 
 5 [deg]C. The manufacturer may need to use a longer soak 
period if necessary to achieve a stabilized permeation rate for a given 
fuel tank, consistent with good engineering judgment.
    The tank would have to be sealed during this fuel soak and we are 
proposing that any components that are directly mounted to the fuel 
tank, such as a fuel cap, must be attached. Other openings, such as 
fittings for fuel lines or petcocks, would be sealed with impermeable 
plugs. In addition, if there is a vent path through the fuel cap, that 
vent path may be sealed. Alternatively, we are proposing that the 
opening could be sealed for testing and the fuel cap tested separately 
for permeation (discussed below). If the fuel tank is designed to have 
a separate fill neck between the fuel cap and the tank that is at least 
12 inches long and at least 6 inches above the top of the fuel tank, 
the tank may be sealed with something other than a production fuel cap.
    Manufacturers may do the durability testing described below during 
the time period specified for preconditioning. The time spent in 
durability testing may count as preconditioning time as long as the 
fuel tank has fuel inside the entire time. During the slosh testing, a 
fuel fill level of 40 percent would be considered acceptable for the 
fuel soak. Otherwise, we are proposing to require that the fuel tank be 
filled to nominal capacity during the fuel soak.
(c) Durability Tests
    We are proposing three tests to evaluate the durability of fuel 
tank permeation controls: (1) Fuel sloshing; (2) pressure-vacuum 
cycling; and (3) ultraviolet exposure. The purpose of these 
deterioration tests would be to help ensure that the technology is 
durable under the wide range of in-use operating conditions. For 
sloshing, the fuel tank would be filled to 40 percent capacity with E10 
fuel and rocked for one million cycles. The pressure-vacuum testing 
would consist of 10,000 cycles from -0.5 to 2.0 psi. These two proposed 
durability tests are based on draft recommended SAE practice.\89\ The 
third durability test would be intended to assess potential impacts of 
ultraviolet sunlight (i.e., light with wavelength ranging from 300 to 
400 nanometers) on the durability of surface treatment. In this test, 
the tank would be exposed to ultraviolet light with an intensity of at 
least 0.40 W-hr/m2/min on the tank surface for 450 hours. 
Alternatively, we are proposing the tank could be exposed to direct 
natural sunlight for an equivalent period of time.
---------------------------------------------------------------------------

    \89\ Draft SAE Information Report J1769, ``Test Protocol for 
Evaluation of Long Term Permeation Barrier Durability on Non-
Metallic Fuel Tanks,'' (Docket EPA-HQ-OAR-2004-0008-0195).
---------------------------------------------------------------------------

    We are proposing to include a provision that would allow 
manufacturers to omit one or more of the durability tests if it is not 
appropriate for a certain tank design. For example, coextruded plastic 
tanks rely on a thin layer of material within the wall of the tank. 
This material is never exposed to sunlight or liquid fuel so the 
sloshing, pressure, and ultraviolet-exposure tests would not be 
necessary. At the same time, we request comment on whether other 
durability tests would be necessary to ensure that the fuel tank would 
not be compromised for safety due to changes to address permeation. 
Examples may be temperature cycling or impact testing.
(d) Weight-Loss Test
    Following the fuel soak, we are proposing that the fuel tank must 
be drained and refilled with fresh fuel immediately after to prevent 
the fuel tank from drying out. The tank would have to be sealed within 
eight hours after refreshing the fuel at the end of the soak period. 
The permeation rate from fuel tanks would be measured by comparing mass 
measurements of the tank before and after a soaking period of at least 
two weeks at a temperature of 28  2 [deg]C. In the case of 
fuel tanks with very low permeation, the weight loss of the fuel tank 
over two week period could be too small to obtain an accurate 
measurement. We are proposing that manufacturers may extend the test 
period by two weeks to obtain an accurate measurement for fuel tanks 
with low permeation rates, consistent with good engineering judgment.
    A change in atmospheric pressure over the weeks of testing can 
affect the accuracy of measured weights for testing due to the buoyancy 
of the fuel tank. The buoyancy effect on emission measurements is 
proportional to the volume of the fuel tank, so this procedure is 
appropriate even for testing very small fuel tanks. To address this we 
are proposing a procedure in which a reference fuel tank filled with 
sand or some other inert material to the approximate total weight of 
the test tank be used to zero the scale used for measuring the test 
tank. This would result in measured and reported values representing 
the change in mass from permeation losses rather than a comparison of 
absolute masses. This is similar to an approach in which weighing would 
determine absolute masses with a mathematical correction to account for 
the effects of buoyancy. We believe the proposed approach is better 
because it minimizes the possibility of introducing or propagating 
error.
    We propose to allow permeation measurements for certification using 
alternative equipment and procedures that provide equivalent results. 
To use these alternative methods, manufacturers would first need to get

[[Page 28179]]

our approval. An example of an alternative weight-loss measurement 
procedure would be to test the fuel tank in a SHED and determine the 
permeation by measuring the concentration of hydrocarbons in the 
enclosure.
(e) Fuel Cap Permeation Testing
    As discussed above, we are proposing that manufacturers would have 
the option to test the fuel cap separately from the tank and combine 
the results to determine the total tank permeation rate. In this case, 
the permeation test would be performed as described above except that 
the fuel cap would be mounted on an impermeable reservoir such as a 
metal or glass tank. The volume of the test reservoir would have to be 
at least one liter to ensure sufficient fuel vapor exposure. We are 
proposing that the ``tank'' surface area for calculating the results 
would be the smallest inside cross sectional area of the opening on 
which the cap is mounted. The fuel cap would need to be tested in 
conjunction with a representative gasket. In the case where the vent 
path is through grooves in the gasket, another gasket of the same 
material and dimensions, without the vent grooves, may be used. In the 
case where the vent is through the cap, that vent would be sealed for 
testing.
(3) Diurnal Emission Testing Procedures
    The proposed test procedure for diurnal emissions from installed 
marine fuel tanks involves placing the fuel tank in a SHED, varying the 
temperature over a prescribed profile, and measuring the hydrocarbons 
escaping from the fuel tank (see Sec.  1060.525). The final result 
would be reported in grams per gallon where the grams are the mass of 
hydrocarbons escaping from the fuel tank over 24 hours and the gallons 
are the nominal fuel tank capacity. The proposed test procedure is 
derived from the automotive evaporative emission test with 
modifications specific to marine applications.\90\ We request comment 
on the proposed diurnal test procedures described below.
---------------------------------------------------------------------------

    \90\ See 40 CFR part 86, subpart B, for the automotive 
evaporative emission test procedures.
---------------------------------------------------------------------------

(a) Temperature Profile
    We believe it is appropriate to base diurnal measurements on a 
summer day with ambient temperatures ranging from 72 to 96 [deg]F (22.2 
to 35.6 [deg]C). This temperature profile, which is also used for 
automotive testing, represents a hot summer day when ground-level ozone 
formation is most likely. Due to the thermal mass of the fuel and, in 
some cases, the inherent insulation provided by the boat hull, the fuel 
temperatures would cover a narrower range. Data presented in Chapter 5 
of the Draft RIA suggest that the fuel temperature in an installed 
marine fuel tank would see a total change of about half the ambient 
temperature swing. We are therefore proposing a test temperature range 
of 78 to 90 [deg]F (25.6 to 32.2 [deg]C) for installed marine fuel 
tanks. This testing would be based on fuel temperature instead of 
ambient temperature.
    We are proposing an alternative, narrower temperature range for 
fuel tanks installed in nontrailerable boats (>=26 ft.). Data presented 
in Chapter 5 of the Draft RIA suggest that the fuel temperature swing 
in a boat stored in the water would be about 20 percent of the ambient 
temperature swing. Based on this relationship, we are proposing an 
alternative temperature cycle for tanks installed in nontrailerable 
boats of 81.6 to 86.4 [deg]F (27.6 to 30.2 [deg]C). This alternative 
temperature cycle would be associated with an alternative standard as 
discussed earlier. See the proposed regulations at Sec.  1060.525 for 
further detail. We request comment on the proposed test temperatures, 
especially on the appropriateness of the alternative test procedure and 
standard for tanks installed in nontrailerable boats.
    The automotive diurnal test procedure includes a three-day 
temperature cycle to ensure that the carbon canister can hold at least 
three days of diurnal emissions without vapors breaking through to the 
atmosphere. For marine vessels using carbon canisters as a strategy for 
controlling evaporative emissions, we are proposing a three-day cycle 
here for the same reason. In the automotive test, the canister is 
loaded and then purged by the engine during a warm-up drive before the 
first day of testing. Here, we are proposing a different approach 
because we anticipate that canisters on marine applications will be 
passively purged. Before the first day of testing, the canister would 
be loaded to full working capacity and then run over the diurnal test 
temperature cycle, starting and ending at the lowest temperature, to 
allow one day of passive purging. The test result would then be based 
on the highest recorded value during the following three days.
    For fuel systems using a sealed system (including those that rely 
on pressure-relief valves with no canister), we believe a three-day 
test would not be necessary. Before the first day of testing, the fuel 
would be stabilized at the initial test temperature. Following this 
stabilization, the SHED would be purged, followed by a single run 
through the diurnal temperature cycle. Because this technology does not 
depend on purging or storage capacity of a canister, multiple days of 
testing should not be necessary. We are therefore proposing a one-day 
test for the following technologies: Sealed systems, sealed systems 
with a pressure-relief valve, bladder fuel tanks, and sealed fuel tanks 
with a volume-compensating air bag. We request comment on this 
simplified approach.
(b) Test Fuel
    Consistent with the automotive test procedures, we are proposing to 
specify a gasoline test fuel with a volatility of 9 psi.\91\ We are not 
proposing that the fuel used in diurnal emission testing include 
ethanol for two reasons. First, we do not believe that ethanol in the 
fuel affects the diurnal emissions or control effectiveness other than 
the effect that ethanol in the fuel may have on fuel volatility. 
Second, in-use fuels containing ethanol are generally blended in such a 
way as to control for ethanol effects in order to meet fuel volatility 
requirements. We request comment on the proposed test fuel and whether 
it would be appropriate to specify a test fuel blended with ethanol 
either as the primary test fuel or as an optional test fuel. If so, we 
request comment regarding whether the volatility of the test fuel 
should be controlled to 9 psi or if ethanol should be blended into 
certification gasoline. We also request comment on the effect of 
ethanol in the fuel on controlled diurnal emissions and if the standard 
would need to be adjusted to account for ethanol in the test fuel.
---------------------------------------------------------------------------

    \91\ Volatility is specified based on a procedure known as Reid 
Vapor Pressure (see ASTM D 323-99a).
---------------------------------------------------------------------------

    Diurnal emissions are not only a function of temperature and fuel 
volatility, but of the size of the vapor space in the fuel tank. 
Consistent with the automotive procedures, we are proposing that the 
fill level at the start of the test be 40 percent of the nominal 
capacity of the fuel tank. Nominal capacity of the fuel tank would be 
defined as the a fuel tank's volume as specified by the fuel tank 
manufacturer, using at least two significant figures, based on the 
maximum volume of fuel the tank can hold with standard refueling 
techniques. The ``permanent'' vapor space above a fuel tank that has 
been filled to capacity would not be considered in the nominal capacity 
of the fuel tank.

[[Page 28180]]

(c) Fuel Tank Configuration
    The majority of marine fuel tanks are made of plastic. Even plastic 
fuel tanks designed to meet our proposed standards would be expected to 
have some amount of permeation. However, over the length of the diurnal 
test, if it were performed on a new tank that had not been previously 
exposed to fuel, the effect of permeation on the test results should be 
insignificant. For fuel tanks that have reached their stabilized 
permeation rate (such as testing on in-use tanks), we believe it would 
be appropriate to correct for permeation. In such a case, we propose 
that the permeation rate be measured from the fuel tank and subtracted 
from the final diurnal test result. The fuel tank permeation rate would 
be measured with the established procedure for measuring permeation 
emissions, except that the test fuel would be the same as that used for 
diurnal emission testing. This test measurement would have to be made 
just before the diurnal emission test to ensure that the permeation 
rate does not change when measuring diurnal emissions. In no case would 
we allow a permeation correction higher than that corresponding to the 
applicable permeation standard for a tank with a given inside surface 
area. Because not correcting for permeation represents the worst-case 
test result, we would accept data from manufacturers in which no 
permeation correction was applied. We request comment on this approach.
(4) Diffusion Testing Procedures
    The proposed procedure for measuring diffusion emissions is very 
similar to that for diurnal emissions, with three primary differences 
(see Sec.  1060.530). First, the fuel tank should be filled to 90 
percent of its nominal capacity. Second, the fuel tank is held in a 
controlled environment to stabilize at test temperatures. Third, the 
test run is proposed to be six hours in length. Testing has shown that 
diffusion occurs at a steady rate, so we would want manufacturers to be 
able to run a full test in a single day's shift rather than running a 
test for a full 24 hours. Measured emissions are then adjusted 
mathematically for comparison to the gram-per-day standard.
    There is some concern that fluctuating temperatures during this 
test could cause small diurnal effects that would result in higher 
measured emissions. Filling the fuel tank to 90 percent would help 
minimize the potential for diurnal effects by increasing the thermal 
mass of the fuel and by reducing the volume of the vapor space. In 
addition, the proposed diffusion standard is based on data collected 
from testing in this manner.
    As described above, we are proposing to allow fuel cap 
manufacturers to voluntarily certify their fuel caps to diffusion 
standard. This would require a separate test with a fuel cap mounted on 
a test tank with a representative sealing configuration of production 
tanks.
    As described for diurnal measurements, we are proposing that 
manufacturers would be able to separately quantify permeation emissions 
occurring during the diffusion test and subtract the permeation 
contribution so the reported result isolates the test to quantifying 
diffusion emissions.
(5) Measurement Procedures Related to Running Loss Emissions
    We do not specify a procedure for measuring running loss emissions, 
but we are proposing to allow manufacturers to demonstrate control of 
running losses by showing that fuel temperatures will not increase by 
more than 8 [deg]C during normal operation (see Sec.  1060.104 and 
Sec.  1060.535). This requires testing to measure fuel temperatures on 
each equipment configuration. We are proposing a fuel temperature test 
that includes filling the fuel tank with commercially available 
gasoline and operating the equipment for one hour over a normal in-use 
duty cycle with a load factor approximately the same as the specified 
test cycle. If the equipment consumes 80 percent of the fuel capacity 
in one hour of operation, a shorter period may be used based on time 
until the fuel tank is drained to 20 percent capacity. We are proposing 
that manufacturers would be required to document a description of the 
operation and include grass height or equivalent variables affecting 
load.
    We are proposing that the testing must occur outdoors with a 
beginning ambient temperature ranging from 20 to 30 [deg]C with no 
precipitation and with average wind speeds below fifteen miles per 
hour. The ambient temperature would have to be steady or increasing 
during the test and it must be during a mostly sunny time period with a 
maximum cloud cover of 25 percent as reported by the nearest local 
airport making hourly meteorological observations.
    We are proposing that the temperature of the fuel in the tank must 
be within 2 [deg]C of (but not exceeding) the ambient temperature at 
the beginning of the test. Fuel temperature would be measured with a 
thermocouple positioned in the fuel but not touching the inside walls 
or bottom of the tank. Ambient temperature would be measured on-site in 
the shade. The equipment configuration meets the requirement to control 
running losses if measured minimum and maximum fuel temperatures 
throughout the period of operation do not differ by more than 8 [deg]C. 
In the case were the equipment has multiple fuel tanks, the temperature 
would have to be measured on each fuel tank. We request comment on this 
procedure for measuring fuel temperatures.
    We are also proposing to allow manufacturers to use an alternative 
procedure in a laboratory with prior EPA approval. The alternative test 
procedure would need to simulate outdoor conditions and consider engine 
operation, solar load, temperature, and wind speed. The manufacturer 
would be required to make a demonstration of equivalency.

F. Certification and Compliance Provisions

    Sections VII and VIII describe several general provisions related 
to certifying emission families and meeting other regulatory 
requirements. This section notes several particulars related to 
applying these general provisions to evaporative emissions.
    Marine vessels do not always include installed fuel systems. 
Manufacturers of vessels without installed fuel systems do not have the 
ability to control engine or fuel system design parameters. We are 
therefore proposing that vessels without an installed fuel system would 
not be subject to the proposed standards (see Sec.  1045.5). As a 
result, it is necessary for us to treat manufacturers of uninstalled 
fuel-system components as the equipment manufacturer with respect to 
evaporative emission standards. This includes manufacturers of outboard 
engines (including any fuel lines or fuel tanks produced with the 
engine), portable fuel tanks, and the fuel line system (including fuel 
line, primer bulb, and connectors).
    For ease of reference, Small SI equipment manufacturers, Marine SI 
boat builders, and manufacturers of portable marine fuel tanks (and 
associated fuel-system components) are all referred to as equipment 
manufacturers in this section.
(1) Liability for Certification and Compliance
    The proposed standards for fuel lines and fuel tanks apply to any 
such components that are used with or intended to be used with Small SI 
engines or Marine SI engines (see Sec.  1060.1 and Sec.  1060.601). 
Section VI.C

[[Page 28181]]

describes for each standard which manufacturer is expected to certify. 
Engine manufacturers would describe these fuel-system components in the 
same certification application in which they document their compliance 
with exhaust emission standards (see Sec.  1045.201 and Sec.  
1054.201).
    In most cases, nonroad standards apply to the manufacturer of the 
engine or the manufacturer of the nonroad equipment. Here, the products 
subject to the standards (fuel lines and fuel tanks) are typically 
manufactured by a different manufacturer. In most cases the engine 
manufacturers do not produce complete fuel systems and would therefore 
not be in a position to do all the testing and certification work 
necessary to cover the whole range of products that will be used. We 
are therefore proposing an arrangement in which manufacturers of fuel-
system components are in most cases subject to the standards and are 
subject to certification and other compliance requirements associated 
with the applicable standards. We are proposing to prohibit the 
introduction into commerce of noncompliant fuel-system components that 
are intended for installation in Small SI equipment or Marine SI 
vessels unless the component manufacturer either certifies the 
component or has a contractual arrangement for each equipment 
manufacturers using their products to certify those components. As a 
matter of good practice, any components not intended for installation 
in Small SI equipment or Marine SI vessels should be labeled 
accordingly to prevent the possibility of improper installation to 
prevent confusion in this regard.
    As described in Section VI.D, component manufacturers may certify 
with measured emission levels showing that the components meet the 
emission standard, or they may certify to an FEL above or below the 
standard. If any component manufacturer certifies using an FEL, the FEL 
becomes the emission standard for that emission family for all 
practical purposes. The component manufacturer however would not be 
required to meet any overall average for their products, but would have 
the option to certify to an FEL above or below the standard. This is to 
facilitate the use of ABT by equipment manufacturers, as discussed 
below.
    Equipment manufacturers would be subject to all the proposed 
evaporative standards. This applies for the general standards described 
above with respect to fuel caps, miscellaneous fuel-system components, 
and refueling. These standards generally depend on design 
specifications rather than emission measurements, so we believe it is 
appropriate to simply deem these products to be certified if they are 
designed and produced to meet the standards we specify. The vessel 
manufacturer would also need to keep records of the components used 
(see Sec.  1060.210). This would allow us, by operation of the 
regulation, to have certified products without requiring the paperwork 
burden associated with demonstrating compliance with these relatively 
straightforward specifications. Manufacturers could optionally apply 
for and receive a certificate of conformity with respect to these 
general standards, but this would not be necessary and we would expect 
this to be a rare occurrence.
    Equipment manufacturers would also be subject to all the proposed 
emission standards. Equipment manufacturers may comply with 
requirements related to evaporative emission standards in three 
different situations. First, equipment manufacturers might install only 
components certified by the component manufacturer, without using 
emission credits. In this case all the components must meet the 
proposed emission standard or have an FEL below the standard. Such an 
equipment manufacturer would be subject to the fuel line and fuel tank 
standards, but would be able to satisfy their requirements by using 
certified components. They would need to apply for certification only 
with respect to the remaining emission standards they are subject to, 
such as running loss emissions (if applicable). Equipment manufacturers 
must also design and produce their equipment to meet the requirements 
specified in Sec.  1060.101(f), though this would not necessarily 
involve an application for certification. Such an equipment 
manufacturer would generally need only to use certified components, add 
an emission label, and follow any applicable emission-related 
installation instructions to ensure that certified components are 
properly installed. This is similar to an equipment manufacturer that 
is required to properly install certified engines in its equipment, 
except that the equipment manufacturer must meet general design 
standards and shares the liability for meeting emission standards.
    Second, equipment manufacturers may be required to certify certain 
components based on contractual arrangements with the manufacturer of 
those components. In this case, the equipment manufacturer's 
certification causes the component manufacturer to no longer be subject 
to the standard. This approach might involve the equipment manufacturer 
relying on test data from the component manufacturer. The equipment 
manufacturer might also be producing its own fuel tanks for 
installation in its equipment, in which case it would be subject to the 
standards and all requirements related to certification and compliance. 
In either case, the equipment manufacturer would take on all the 
responsibilities associated with certification and compliance with 
respect to those components.
    Third, equipment manufacturers may comply with evaporative emission 
requirements by using certified components, some of which are certified 
to an FEL above the standard. The equipment manufacturer would then 
comply based on emission credits. In this case, the equipment 
manufacturer would take on all the certification and compliance 
responsibilities with respect to any components that are part of the 
equipment manufacturer's emission credit calculations. Equipment 
manufacturers would generally use only certified components for meeting 
evaporative emission requirements, but they might also hold the 
certificate for such components. For purposes of certification, 
equipment manufacturers would not need to submit new test data if they 
use certified components. Equipment manufacturers would make an annual 
accounting to demonstrate a net balance of credits for the model year. 
Under this approach, the component manufacturer would continue to be 
subject to the standards for its products and be required to meet the 
certification and compliance responsibilities related to the standard. 
However, as in the first option, the component manufacturer would not 
be required to meet any averaging requirements or be required to use 
emissions credits. Where equipment manufacturers use ABT with 
components that have already been certified by the component 
manufacturer, there will be overlapping certifications between the two 
parties. We propose to address this by specifying that all parties are 
responsible for meeting applicable requirements associated with the 
standards to which they have certified, but if any specific requirement 
is met by one company, we will consider the requirement to be met for 
all companies (see Sec.  1060.5). For example, either the component 
manufacturer or the equipment manufacturer could honor warranty claims, 
but we may hold both companies responsible for the violation if there 
is a failure to meet warranty obligations.
    Similarly, if we find that new equipment is sold without a valid

[[Page 28182]]

certificate of conformity for the fuel lines or fuel tanks, then the 
equipment manufacturer and all the affected fuel-system manufacturers 
subject to the standards would be liable for the noncompliance (see 
Sec.  1060.601).
    Liability for recall of noncompliant products would similarly fall 
to any manufacturer whose product is subject to the standard, as 
described above. If more than one manufacturer is subject to the 
standards for a noncompliant product, we would have the discretion to 
assign recall liability to any one of those manufacturers. In assigning 
this liability, we would generally consider factors such as which 
manufacturer has substantial manufacturing responsibility and which 
manufacturer holds the certificate (see Sec.  1060.5). However, we may 
hold equipment manufacturers liable for recall even if they don't 
manufacture or certify the defective product. This would generally be 
limited to cases where the component manufacturer is unavailable to 
execute any remedial action. For example, if a foreign component 
manufacturer discontinues their participation in the U.S. market or a 
component manufacturer goes out of business, we would turn to the 
equipment manufacturer.
    The proposed running loss standards for nonhandheld Small SI 
engines are not geared toward component certification, which 
necessitates some special provisions. If engine manufacturers sell 
their engines with a complete fuel system, which is typical for Class I 
engines, they would also be subject to and need to comply with running 
loss standards as part of their overall certification. Of the available 
alternatives for demonstrating compliance with the running loss 
standard, we would expect the only practical approach for these 
companies would be to route vapors from the fuel tank into the engine's 
air intake system for combustion. Any engine manufacturer certifying 
its engines this way would need to test for exhaust emissions with an 
installed running loss vent (see Sec.  1054.501). If equipment 
manufacturers use only fuel-system components that have been certified 
by component manufacturers (without using emission credits) and engines 
that are certified by the engine manufacturer to meet both exhaust and 
running loss standards, they would have no responsibility to certify. 
However, if the engine manufacturer does not sell its engine with a 
complete fuel system that has been certified for running loss control, 
the equipment manufacturer would need to certify with respect to the 
running loss standard.
    The running loss standard is not a typical standard based on 
emission measurements using established procedures. Some of the 
available compliance demonstrations involve straightforward design 
specifications that involve no measurement at all. The approach of 
keeping fuel temperatures from increasing above a specified threshold 
involves a test procedure with a performance standard, but does not 
involve emission measurements. As described above, it may be possible 
to identify design specifications that would replace the need for the 
proposed temperature measurements. In this case running loss control 
would be a straightforward design standard that we could treat like the 
general standards above, in which equipment manufacturers are deemed to 
be certified by operation of the regulations, rather than submitting an 
application for certification. The regulations would prohibit the sale 
of equipment without the specified running loss controls.
(2) Regulatory Requirements Related to Certification
    The established provisions for implementing exhaust emission 
standards apply similarly for evaporative emission standards; however, 
because the control technologies are very different, these requirements 
require further clarification. For example, scheduled maintenance is an 
important part of certifying engines to exhaust emission standards. 
There is little or no maintenance involved for the expected 
technologies for controlling evaporative emissions. The regulations 
still require manufacturers to identify specified maintenance 
procedures, if there are any, but there are no specific limitations on 
the maintenance intervals and no distinction for emission-related 
maintenance. Manufacturers may not do any maintenance during testing 
for certification. (See Sec.  1060.125 and Sec.  1060.235.) We also do 
not expect that emission-related warranty claims would be common, but 
we are proposing a two-year period for emission-related warranties with 
respect to evaporative emission controls.
    Similarly, we do not expect manufacturers to use evaporative 
emission control technologies that involve adjustable parameters or 
auxiliary emission control devices. Technologies that control 
evaporative emissions are generally passive designs that prevent vapors 
from escaping, in contrast to the active systems engines use to control 
exhaust emissions. The regulations state the basic expectation that 
systems must comply with standards throughout any adjustable range 
without auxiliary emission control devices, but it is clear that these 
provisions will not apply to most evaporative systems. We also do not 
allow emission control strategies that cause or contribute to an 
unreasonable risk to public health or welfare or that involve defeat 
devices. While these are additional statutory provisions that are 
meaningful primarily in the context of controlling exhaust emissions, 
we are proposing to include them for addressing evaporative emissions 
(see Sec.  1045.101). This also addresses the possibility that future 
technologies may be different in a way that makes these provisions more 
meaningful. We request comment on this approach. In particular we 
request comment on best way of adapting these provisions to evaporative 
emission controls.
    The testing specified for certifying fuel systems to the 
evaporative emission standards includes measurements for evaluating the 
durability of emission control technologies where appropriate. While we 
adopted evaporative requirements for recreational vehicles relying on a 
testing approach that used deterioration factors, we believe it is more 
appropriate to incorporate the durability testing for each family 
directly. Therefore, no requirement exists for generating deterioration 
factors for any evaporative emission standard. We request comment on 
the best approach to incorporate durability testing for evaporative 
emission standards
    We are proposing to require that Small SI engine or equipment 
manufacturers add an emission control information label if they certify 
with respect to running losses or if they certify based on the use of 
emission credits. We are proposing to require that Marine SI engine or 
vessel manufacturers add an emission control information label for 
evaporative emission only if they certify based on the use of emission 
credits. (See Sec.  1060.135.) If engine, equipment, or vessel 
manufacturers also certify fuel-system components separately, they may 
include that additional information in a combined label. If the 
equipment is produced by the same company that certifies the engine for 
exhaust standards, the emission control information label for the 
engine may include all the appropriate information related to 
evaporative emissions.
    In addition, we are proposing a simplified labeling requirement for 
fuel lines (see Sec.  1060.136). This would involve only the fuel line 
manufacturer's name, EPA's standardized designation for an

[[Page 28183]]

emission family, and the family emission limit (FEL), if applicable. 
This labeling information would need to be repeated continuously, with 
not more than 12 inches before repeating. There is some concern that if 
short sections of fuel lines are used, that sections of the fuel line 
may be found on equipment without sufficient markings on them. We 
request comment regarding whether the length of the repeated labeling 
information should be shorter than 12 inches. We are proposing 
simplified labeling requirements for fuel filters, primer bulbs, or 
short preformed fuel lines (less than 12 inches long) (see Sec.  
1060.138).
    Fuel tanks that are certified separately would need to include an 
emission control information label (see Sec.  1060.137). This would 
involve fuel tank manufacturer's name, EPA's standardized designation 
for an emission family, the FEL (if applicable), a simple compliance 
statement, and a description of the method of controlling emissions. 
For example, a label on a certified marine fuel tank would need to 
describe how it meets permeation emission standards and identify the 
part numbers of any associated components for meeting diurnal emission 
standards.
    Including the fuel tank's family emission limit is important, not 
only for EPA oversight, but also to communicate this information to 
equipment manufacturers and end users. Unlike the situation for exhaust 
emissions, the certifying manufacturer establishes the FEL, but does 
not maintain a balance of emission credits. Equipment manufacturers may 
buy fuel tanks and fuel lines that have an FEL, which would be the 
basis for calculating emission credits for the equipment manufacturer. 
Any other approach would require equipment manufacturers to be vigilant 
about verifying FEL values with EPA or the component manufacturer, or 
both. Also, as described in Section VI.F.6, we are proposing to require 
that owners find replacement fuel tanks and fuel lines with FELs that 
match or exceed the emission control performance represented by the 
original parts. This is an unrealistic expectation unless the FEL is 
readily available on the original equipment.
    Other fuel-system components would need to be labeled with the 
manufacturer's name and part number, if space allows, and EPA's 
standardized designation for an emission family (see Sec.  1060.138). 
This would apply for carbon canisters, fuel tanks that are not 
certified separately, and any other fuel-system components (such as 
fuel caps) that are certified separately. Equipment manufacturers could 
meet the requirement to label fuel tanks by placing the overall 
equipment label on the fuel tank, as long as the fuel tank and label 
are positioned such that the label can be read easily.
    Manufacturers have expressed concern that it would be very 
difficult to properly label very small fuel tanks and fuel lines. To 
the extent that engine manufacturers are certifying their products with 
respect to evaporative emissions, this problem can be addressed in part 
by putting the information related to evaporative emissions on the 
engine label already required for exhaust emissions. This is most 
likely to be the case for the smallest products. We request comment on 
any additional provisions we would need to specify to address space 
limitations on very small fuel-system components.
    While we are proposing no requirement for manufacturers to test 
production-line or in-use products, we may pursue testing of certified 
products to evaluate compliance with evaporative emission standards 
(see Sec.  1060.301).
(3) Emission Families
    To certify equipment or components, manufacturers would first 
define their emission families. This is generally based on selecting 
groups of products that have similar emission characteristics 
throughout the useful life (see Sec.  1060.230). For example, fuel 
tanks could be grouped together if they were made of the same material 
(including consideration of additives such as pigments, plasticizers, 
and UV inhibitors that may affect emissions) and the same control 
technology. For running loss control for nonhandheld Small SI engines 
and equipment, emission families are based on the selected compliance 
demonstration. For example, certifying manufacturers would have one 
emission family for all their products that vent fuel vapors to the 
engine's air intake system, and another emission family for all their 
products that comply based on keeping fuel temperatures below the 
specified threshold.
    The manufacturer would then select a single product from the 
emission family for certification testing. This product would be the 
one that is most likely to exceed the applicable emission standard. For 
instance, the ``worst-case'' fuel tank in a family of monolayer tanks 
would likely be the tank with the thinnest average wall thickness. For 
fuel lines or co-extruded fuel tanks with a permeation barrier layer, 
the worst-case configuration may be the thinnest barrier thickness.
    Testing with those products, as specified above, would need to show 
compliance with emission standards. The manufacturers would then send 
us an application for certification. After reviewing the information in 
the application, we would issue a certificate of conformity allowing 
equipment manufacturers to introduce into commerce certified equipment 
from the covered emission family, or alternatively, equipment with the 
components from certified emission families.
(4) Compliance Provisions From 40 CFR Part 1068
    As described in Section VIII, we are proposing to apply the 
provisions of 40 CFR part 1068 to Small SI and Marine SI engines, 
equipment, and vessels. This section describes how some of the 
provisions of part 1068 apply specifically with respect to evaporative 
emissions.
    The provisions of Sec.  1068.101 prohibit introducing into commerce 
new nonroad engines and equipment unless they are covered by a 
certificate of conformity and labeled appropriately. Section VI.F.1 
describes the responsibilities for engine manufacturers, equipment 
manufacturers, and manufacturers of fuel-system components with respect 
to the prohibition against introducing uncertified products into 
commerce. In the case of portable marine fuel tanks and outboard 
engines, there is no equipment manufacturer so we are proposing to 
treat manufacturers of these items as equipment manufacturers relative 
to this prohibition.
    While engine rebuilding or extensive engine maintenance is 
commonplace in the context of exhaust emission controls, there is very 
little analogous servicing related to evaporative emission controls. 
Nevertheless, it can be expected that individual components, such as 
fuel lines, fuel tanks, or other fuel-system components, may be 
replaced periodically. While the detailed rebuilding provisions of 
Sec.  1068.120 have no meaning for evaporative emission controls, the 
underlying requirement applies generally. Specifically, if someone is 
servicing a certified system, there must be a reasonable basis to 
believe that the modified emission control system will perform at least 
as well as the original system. We are not proposing any recordkeeping 
requirements related to maintenance of evaporative emission control 
systems.
    There are many instances where we specify in 40 CFR part 1068, 
subparts C and D, that engines (and the associated

[[Page 28184]]

equipment) are exempt from emission standards under certain 
circumstances, such as for testing, national security, or export. Our 
principle objective in applying these provisions to evaporative 
emission standards is to avoid confusion. We are therefore proposing 
that an exemption from exhaust emission standards, automatically 
triggers a corresponding exemption from evaporative emission standards 
for the same products. We believe it is unlikely that an equipment 
manufacturer will need a separate exemption from evaporative emission 
standards, but the exemptions related to national security, testing, 
and economic hardship would apply if such a situation were to occur. We 
believe these are the only three reasons that would ever call for 
evaporative systems to be exempt when the engines have not already been 
exempted for some reason. We request comment on this approach to 
addressing exemptions and importation provisions for evaporative 
requirements.
    Given the extended times required to precondition fuel-system 
components, we have no plans to require evaporative testing of units 
from the production line. This means that evaporative measurements are 
not part of the production-line testing program or selective 
enforcement audits. On the other hand, we may require certifying 
manufacturers to supply us with production equipment or components as 
needed for our own testing or we may find our own source of products 
for testing.
    The defect-reporting requirements of Sec.  1068.501 apply to 
certified evaporative systems. This requires the certifying 
manufacturer to maintain information, such as warranty claims, that may 
indicate an emission-related defect. The regulations describe when 
manufacturers must pursue an investigation of apparent defects and when 
to report defects to EPA. These provisions apply to every certifying 
manufacturer and their certified products, including component 
manufacturers.
(5) Interim Compliance Flexibility for Small SI Equipment
    Most Small SI equipment manufacturers are currently certifying 
products to evaporative emission requirements in California. However, 
these standards and their associated test procedures differ somewhat 
from those proposed in this document. Although the standards are 
different, we believe evaporative emission control technologies are 
available to meet the California ARB's standards and our proposed 
emission standards. To help manufacturers transition to selling low-
emission equipment nationwide, we are proposing to accept California 
ARB certification of equipment and components in the early years of the 
proposed federal program.
    As discussed above, we are proposing to accept California ARB 
certification for nonhandheld equipment and fuel tanks for the purposes 
of the proposed early-allowance program (see Sec. Sec.  1045.145 and 
1054.145). We are also proposing to accept California ARB certification 
of handheld fuel tanks through the 2011 model year (see Sec.  90.129).
    We are proposing to accept Class I/Class II fuel lines meeting 
California ARB certification or certain SAE specifications through the 
2011/2010 model years (see Sec.  90.127). These SAE specifications 
include SAE J30 R11A, SAE J30 R12, and SAE J2260 Category 1. Such fuel 
lines would need to be labeled accordingly. As described in Section 
VI.C.1, we are proposing to require that engine manufacturers certify 
fuel lines used with their engines until the proposed Phase 3 standards 
are in place. The purpose of this provision is to give Small SI 
equipment manufacturers additional lead time before they have to 
certify to the proposed standards. For any fuel lines installed on the 
equipment, but not supplied with the engine, we are proposing that the 
engine manufacturer would be required to supply low-permeation fuel 
line specifications in its installation instructions (see Sec.  
90.128). Equipment manufacturers would be required, under the 
prohibited acts specified in the regulations, to use the fuel line 
specified by the engine manufacturer.
    We are proposing to allow certification of walk-behind mowers under 
Sec.  90.127 as an alternative to the proposed fuel line permeation 
standards if manufacturers rely on SHED-based certification to meet the 
California standards that apply to the overall equipment (diurnal, tank 
permeation, and fuel line permeation). While this might allow for use 
of fuel lines that exceed the proposed standards, we believe the 
overall emission control will be at least as great from systems that 
have been tested and certified using SHED-based procedures. The Phase 3 
standards described above do not rely on diurnal emission control, so 
we do not intend to continue the provision for SHED-based testing and 
certification. However, we request comment on the possible 
administrative advantages or emission control advantages of continuing 
this alternative approach in the Phase 3 time frame.
(6) Replacement Parts
    We are proposing to apply the tampering prohibition in Sec.  
1068.101(b)(1) for evaporative systems. This means that it would be a 
violation to replace compliant fuel tanks or fuel lines with 
noncompliant products. This would effectively disable the applicable 
emission controls. To address the concern that low-cost replacement 
products will be easy to make available and difficult to prevent, we 
are proposing several new noncompliance-related provisions. In Sec.  
1060.610 we clarify the meaning of tampering for evaporative systems 
and propose two requirements. First, for the period from January 1, 
2012 to December 31, 2019, we propose to require that manufacturers, 
distributors, retailers, and importers of these replacement parts 
clearly label their products with respect to the applicable 
requirements. For example, a package might be labeled as compliant with 
the requirements in 40 CFR part 1060 or it might be labeled as 
noncompliant and appropriate only for use in applications not covered 
by EPA standards. Unless the packaging clearly states otherwise, the 
product is presumed to be intended for applications that are subject to 
EPA standards. Second, starting in 2020 we are proposing a provision 
stating that it is presumed that all replacement parts intended for 
applications covered by EPA standards will be installed in such 
equipment. This presumption significantly enhances our ability to 
enforce the tampering prohibition because the replacement part is then 
noncompliant before it is installed in a vessel or a piece of 
equipment. We believe shifting to a blanket presumption in 2020 is 
appropriate since in-use vessels and equipment will be almost 
universally subject to EPA's evaporative emission standards by that 
time.
    We are aware that producing low-permeation fuel tanks in very low 
production volumes can be costly. In particular, some equipment owners 
may need to replace a fuel tank that has been certified to a Family 
Emission Limit (FEL) that is lower than the emission standard. The 
owner would need to find and install a replacement fuel tank that is 
certified with an FEL that is the same as or lower than that of the 
replaced fuel tank. However, we are concerned that such replacement 
fuel tanks may in some cases not be available. We are proposing to 
allow equipment owners to ask for an exemption from the tampering 
prohibition if there is no low-FEL tank available. The replacement tank 
would still need to meet applicable

[[Page 28185]]

standards, but would not need to meet the more stringent emission 
levels reflected by the old tank's FEL. We request comment on the need 
for this provision. In particular, we request comment on the likelihood 
that owners would be unable to find replacement tanks that match the 
emission level of the fuel tanks being replaced.
(7) Certification Fees
    Under our current certification program, manufacturers pay a fee to 
cover the costs associated with various certification and other 
compliance activities associated with an EPA issued certificate of 
conformity. These fees are based on the projected costs to EPA per 
emission family. For the fees rule published May 11, 2004, we conducted 
a cost study to assess EPA's costs associated with conducting programs 
for the industries that we certify (69 FR 26222). A copy of the cost 
study worksheets that were used to assess the fees per category may be 
found on EPA's fees Web site at http://www.epa.gov/otaq/proprule.htm. 
We are proposing to establish a new fees category for certification 
related to the proposed evaporative emission standards. The costs for 
this category will be determined using the same method used in 
conducting the previous cost study.
    As under the current program, this depends on an assessment of the 
anticipated number of emission families and the corresponding EPA 
staffing necessary to perform this work. At this time, EPA plans to 
perform a basic level of certification review of information and data 
submitted to issue certificates of conformity for the evaporative 
emission standards, as well as conducting some testing to measure 
evaporative emissions. This is especially the case for equipment 
manufacturers that use only certified components for meeting applicable 
emission standards. We are proposing a fee of $241 based on Agency 
costs for half of a federal employee's time and three employees hired 
through the National Senior Citizens Education and Research Center 
dedicated to the administration of the evaporative certification 
program, including the administrative, testing, and overhead costs 
associated with these people. The total cost to administer the program 
is estimated to be $362,225. We divided this cost by the estimated 
number of certificates, 1503, to calculate the proposed fee.
    We will update the fees related to evaporative emission 
certificates each year when we update the fees for all categories. The 
actual fee in 2015 and later model years will depend on these annual 
calculations. The fees update will be based upon EPA's costs of 
implementing the evaporative category multiplied by the consumer price 
index (CPI), then divided by the average of the number of certificates 
received in the two years prior to the update. The CPI will be applied 
to all of EPA's costs except overhead. This is a departure from EPA's 
current fees program wherein the CPI is applied only to EPA's labor 
costs. In the most recent fees rulemaking, commenters objected to 
applying the CPI to EPA's fixed costs. In the proposed fee program for 
the evaporative category, however, there are no fixed costs. EPA 
expects all its costs to increase with inflation and we therefore think 
it is appropriate to apply the inflation adjustment to all of the 
program costs.
    Where a manufacturer holds the certificates for compliance with 
exhaust emission standards and includes certification for evaporative 
emissions in that same certificate, we would assess an additional 
charge related to compliance with evaporative emission standards to 
that for the exhaust emission certification.
    EPA believes it appropriate to charge less for a certificate 
related to evaporative emissions relative to the existing charge for 
certificates of conformity for exhaust emissions from the engines in 
these same vessels and equipment. The amount of time and level of 
effort associated with reviewing the latter certificates is higher than 
that projected for the certificates for evaporative emissions.
(8) Engineering Design-Based Certification
    Certification of equipment or components that are subject to 
performance-based emission standards depends on test data showing that 
products meet the applicable standards. We are proposing a variety of 
approaches that reduce the level of testing needed to show compliance. 
As described above, we allow manufacturers to group their products into 
emission families so that a test on a single worst-case configuration 
can be used to show that all products in the emission family are 
compliant. Also, test data from a given year could be ``carried over'' 
for later years for a given emission control design (see Sec.  
1060.235). These steps help reduce the overall cost of testing.
    Design-based certification is an additional step that may be 
available to reduce testing requirements (see Sec.  1060.240). To 
certify their products using design-based certification, certifying 
manufacturers would describe, from an engineering perspective, how 
their fuel systems meet the applicable design specifications. These 
manufacturers could then forego the testing described in Section VI.E. 
We believe there are several emission control designs that use 
established technologies that are well understood to have certain 
emission characteristics. At the same time, while engineering design-
based certification is a useful tool for reducing the test burden 
associated with certification, this does not remove a manufacturer's 
liability for meeting the emission standard throughout the useful life.
    The following sections describe how we propose to implement 
engineering design-based certification for each of the different 
performance standards. We are proposing that we may establish 
additional engineering design-based certification options where we find 
that new test data demonstrate that the use of other technology designs 
will ensure compliance with the applicable emission standards. These 
designs would need to produce emission levels comfortably below the 
proposed emission standards when variability in the emission control 
performance is considered.
(a) Fuel Line Permeation
    In our program for recreational vehicles, we specified that fuel 
lines meeting certain SAE specifications could be certified by design. 
However, we are not proposing to allow this for Small SI equipment or 
marine vessels. That decision was appropriate for recreational 
vehicles, because that program did not include provisions for component 
certification. Fuel line manufacturers will need to conduct testing 
anyway to qualify their fuel lines as meeting the various industry 
ratings so any testing burden to demonstrate compliance with EPA 
standards should be minimal. We would allow test data used to meet 
industry standards to be used to certify to the proposed standards 
provided that the data were collected in a manner consistent with this 
proposal and that the data were made available to EPA if required.
(b) Fuel Tank Permeation
    We are proposing to consider that a metal fuel tank meets the 
design criteria for a design-based certification as a low-permeation 
fuel tank. There is also a body of existing test data showing that co-
extruded fuel tanks from automotive applications have permeation rates 
that are well below the proposed standard. We are proposing to allow 
design-based certification for co-extruded high-

[[Page 28186]]

density polyethylene fuel tanks with a continuous ethylene vinyl 
alcohol barrier layer. The EVOH barrier layer would be required to be 
at least 2 percent of the wall thickness of the fuel tank.
    To address the permeability of the fuel cap, seals, and gaskets 
used on metal and co-extruded tanks, we are proposing that the design 
criteria include a specification that seals and gaskets that are not 
made of low-permeation materials must have a total exposed surface area 
smaller than 1000 mm\2\. A metal or co-extruded fuel tank with seals 
that meet this design criterion would reliably pass the standard. 
However, we believe it is not appropriate to assign an emission level 
to fuel tanks using a design-based certification option that would 
allow them to generate emission credits. Given the uncertainty of 
emission rates from the seals and gaskets, we would not consider these 
tanks to be any more effective than other fuel tanks meeting emission 
standards.
(c) Diurnal Emissions
    For portable marine fuel tanks, we are proposing a design standard 
based on automatically sealing the tank to prevent fuel venting while 
fuel temperatures are rising. The options described below for design-
based certification therefore deal only with installed marine fuel 
tanks (including personal watercraft).
    We are proposing that fuel systems sealed to 1.0 psi would meet the 
criteria for engineering design-based certification to the proposed 
diurnal emission standards. Systems that remain sealed up to positive 
pressures of 1.0 psi have a predictable relationship to changing fuel 
temperatures that ensure that total diurnal emissions over the 
specified test procedure will be below the proposed standard. This type 
of system would allow venting of fuel vapors only when pressures exceed 
1.0 psi or when the fuel cap is removed for refueling. Note that 
systems with anti-siphon valves would have to be designed to prevent 
fuel releases when the system is under pressure to meet Coast Guard 
requirements.
    Bladder fuel tanks and tanks with a volume-compensating air bag are 
specialized versions of tanks that may meet the specifications for 
systems that remain sealed up to positive pressures of 1.0 psi. In each 
of these designs, volume changes within a sealed system prevent 
pressure buildup. As long as these designs meet basic specifications 
for system integrity they would also qualify for design-based 
certification.
    We are proposing that fuel tanks equipped with a passively purged 
carbon canister to control diurnal emissions may be certified by 
design, subject to several technical specifications. To ensure that 
there is enough carbon to collect a sufficient mass of hydrocarbon 
vapors, we propose to specify a minimum butane working capacity of 9 g/
dL based on the test procedures specified in ASTM D5228-92. The carbon 
canister would need a minimum carbon volume of 0.040 liters per gallon 
of fuel tank capacity. For fuel tanks certified to the optional 
standards for tanks in nontrailerable boats ( 26 ft. in length), we are 
proposing a minimum carbon volume of 0.016 liters per gallon of fuel 
tank capacity.
    We are proposing two additional specifications for the quality of 
the carbon. We believe these specifications are necessary to ensure 
that the canister will continue to function effectively over the full 
useful life of a marine vessel. First, the carbon would need to meet a 
moisture adsorption capacity maximum of 0.5 grams of water per gram of 
carbon at 90 percent relative humidity and a temperature of 25  5 [deg]C. Second, the carbon would need to pass a dust attrition 
test similar to that in ASTM D3802-79. The moisture adsorption and dust 
attrition tests are described in more detail in Chapter 5 of the Draft 
RIA. We are also proposing that the carbon canister must be properly 
designed to ensure the in-use effectiveness of the carbon.
    The canisters would need to be designed using good engineering 
judgment to ensure structural integrity. They must include a volume 
compensator or other device to hold the carbon pellets in place under 
vibration and changing temperatures and the vapor flow would need to be 
directed so that it reaches the whole carbon bed rather than just 
passing through part of the carbon. We are proposing that the geometry 
of the carbon canister must have a length to diameter ratio of at least 
3.5.
    The emission data we used to develop these proposed engineering 
design-based certification options are presented in Chapter 5 of the 
Draft RIA. Manufacturers wanting to use designs other than those we 
discuss here would have to perform the applicable testing. However, 
once an additional technology is proven, we may consider adding it to 
the list as one that qualifies for engineering design-based 
certification. For example, if several manufacturers were to pool 
resources to test a diurnal emission control strategy and submit this 
data to EPA, we could consider this particular technology, with any 
appropriate design specifications, as one that qualifies to be 
considered compliant under engineering design-based certification. We 
would intend to revise the regulations to include any additional 
technologies we decide are suitable for design-based certification, but 
we would be able to approve the use of additional engineering design-
based certification with these technologies before changing the 
regulations. We request comment on this approach to design-based 
certification for diurnal emission control technologies and on the 
specific technologies discussed above. Section IV.H presents a more 
detailed description of these technologies and how they can be used to 
reduce evaporative emissions.

G. Small-Business Provisions

(1) Small Business Advocacy Review Panel
    On May 3, 2001, we convened a Small Business Advocacy Review Panel 
under section 609(b) of the Regulatory Flexibility Act as amended by 
the Small Business Regulatory Enforcement Fairness Act of 1996. The 
purpose of the Panel was to collect the advice and recommendations of 
representatives of small entities that could be affected by this 
proposed rule and to report on those comments and the Panel's findings 
and recommendations as to issues related to the key elements of the 
Initial Regulatory Flexibility Analysis under section 603 of the 
Regulatory Flexibility Act. We convened a Panel again on August 17, 
2006 to update our findings for this new proposal. The Panel reports 
have been placed in the rulemaking record for this proposal. Section 
609(b) of the Regulatory Flexibility Act directs the review Panel to 
report on the comments of small entity representatives and make 
findings as to issues related to identified elements of an initial 
regulatory flexibility analysis (IRFA) under RFA section 603. Those 
elements of an IRFA are:
     A description of, and where feasible, an estimate of the 
number of small entities to which the proposed rule will apply;
     A description of projected reporting, recordkeeping, and 
other compliance requirements of the proposed rule, including an 
estimate of the classes of small entities that will be subject to the 
requirements and the type of professional skills necessary for 
preparation of the report or record;
     An identification, to the extent practicable, of all 
relevant Federal rules

[[Page 28187]]

that may duplicate, overlap, or conflict with the proposed rule; and
     A description of any significant alternative to the 
proposed rule that accomplishes the stated objectives of applicable 
statutes and that minimizes any significant economic impact of the 
proposed rule on small entities.
    In addition to the EPA's Small Business Advocacy Chairperson, the 
Panel consisted of the Director of the Assessment and Standards 
Division of the Office of Transportation and Air Quality, the 
Administrator of the Office of Information and Regulatory Affairs 
within the Office of Management and Budget, and the Chief Counsel for 
Advocacy of the Small Business Administration.
    Using definitions provided by the Small Business Administration 
(SBA), companies that manufacture internal-combustion engines and that 
employ fewer than 1000 people are considered small businesses for a 
Small Business Advocacy Review (SBAR) Panel. Equipment manufacturers, 
boat builders, and fuel-system component manufacturers that employ 
fewer than 500 people are considered small businesses for the SBAR 
Panel. Based on this information, we asked 25 companies that met the 
SBA small business thresholds to serve as small entity representatives 
for the duration of the Panel process. These companies represented a 
cross-section of engine manufacturers, equipment manufacturers, and 
fuel-system component manufacturers.
    With input from small-entity representatives, the Panel drafted a 
report providing findings and recommendations to us on how to reduce 
potential burden on small businesses that may occur as a result of this 
proposed rule. The Panel Report is included in the rulemaking record 
for this proposal. We are proposing all of the recommendations as 
presented in the Panel Report. The proposed flexibility options 
recommended to us by the Panel, and any updated assessments, are 
described below.
(2) Proposed Burden Reduction Approaches for Small Businesses Subject 
to the Proposed Evaporative Emission Standards
    The SBAR Panel Report includes six general recommendations for 
regulatory flexibility for small businesses affected by the proposed 
evaporative emission standards. This section discusses the provisions 
being proposed based on each of these recommendations. In this industry 
sector, we believe the burden reduction approaches presented in the 
Panel Report should be applied to all businesses with the exception of 
one general economic hardship provision described below which is 
designed specifically for small businesses. The majority of fuel tanks 
produced for the Small SI equipment and Marine SI vessel market are 
made by small businesses or by companies producing small volumes of 
these products. The purpose of these options is to reduce the potential 
burden on companies for which fixed costs cannot be distributed over a 
large product line. For this reason, we often also consider the 
production volume when making decisions regarding burden reduction 
options.
(a) Consideration of Appropriate Lead Time
    Small businesses commented that they would need to make significant 
changes to their plastic fuel tank designs and molding practices to 
meet the proposed fuel tank permeation standards. For blow-molded tank 
designs with a molded-in permeation barrier, new blow-molding machines 
would be needed that could produce multi-layer fuel tanks. One small 
business commented that, due to the lead time needed to install a new 
machine and to perform quality checks on the tanks, they would not be 
ready to sell multi-layer blow-molded fuel tanks until 2011 for the 
Small SI and Marine SI markets.
    Small businesses that rotational-mold fuel tanks were divided in 
their opinion of when they would be ready to produce low-permeation 
fuel tanks. One manufacturer stated that it is already producing fuel 
tanks with a low-permeation inner layer that are used in Small SI 
applications. This company also sells marine fuel tanks, but not with 
the low-permeation characteristics. However, they have performed Coast 
Guard durability testing on a prototype 40 gallon marine tank using 
their technology which passed the tests. Two other small businesses, 
that rotationally mold fuel tanks, stated that they have not been able 
to identify and demonstrate a low-permeation technology that would meet 
their cost and performance needs. They commented that developing and 
demonstrating low-permeation technology is especially an issue for the 
marine industry because of the many different tank designs and Coast 
Guard durability requirements.
    Consistent with the Panel recommendations in response to the above 
comments, we are proposing to provide sufficient lead time for blow-
molded and marine rotational molded fuel tanks. We are proposing tank 
permeation implementation dates of 2011 for Class II equipment and 2012 
for Class I equipment. For marine fuel tanks, we are proposing to 
implement the tank permeation standards in 2011 with an additional year 
(2012) for installed fuel tanks which are typically rotational-molded 
marine fuel tanks (see Sec.  1054.110 and Sec.  1045.107).
    There was no disagreement on the technological feasibility of the 
Marine SI diurnal emission standard EPA is considering. Small 
businesses commented that they would like additional time to install 
carbon canisters in their vessels. They stated that some boat designs 
would require deck and hull changes to assist in packaging the 
canisters and they would like to make these changes in the normal 
turnover cycle of their boat molds. Small businesses commented that 
they would consider asking EPA to allow the use of low-permeation fuel 
line prior to 2009 as a method of creating an emission neutral option 
for providing extra time for canisters. We are requesting comment on 
phase-in schemes or other burden reduction approaches which would 
provide small businesses additional lead time to meet these 
requirements without losing overall emission reductions.
    The majority of large equipment manufacturers have indicated that 
they will be using low-permeation fuel lines in the near term as part 
of their current product plans. As a result, we are proposing an 
implementation date of 2008 for Small SI fuel line permeation standards 
for nonhandheld equipment (see Sec.  90.127). The Panel expressed 
concern that small equipment manufacturers who do not sell products in 
California may not necessarily be planning on using low-permeation fuel 
line in 2008. Therefore, we are proposing a 2009 implementation date 
for low-permeation fuel line for small businesses producing Small SI 
nonhandheld equipment.
(b) Fuel Tank ABT and Early-Incentive Program
    The Panel recommended that we propose an ABT program for fuel tank 
permeation and an early-allowance program for fuel tank permeation. Our 
proposed ABT and early-allowance programs are described above. We are 
requesting comment on including service tanks in the ABT program. These 
are tanks that are sold as replacement parts for in-use equipment.
(c) Broad Definition of Emission Family
    The Panel recommended that we propose broad emission families for 
fuel tank emission families similar to the

[[Page 28188]]

existing provisions for recreational vehicles. As described above, we 
are proposing permeation emission families be based on type of material 
(including additives such as pigments, plasticizers, and UV 
inhibitors), emission control strategy, and production methods. Fuel 
tanks of different sizes, shapes, and wall thicknesses would be grouped 
into the same emission family (see Sec.  1045.230 and Sec.  1054.230). 
Manufacturers therefore would be able to broadly group similar fuel 
tanks into the same emission family and then only test the 
configuration most likely to exceed the emission standard. Although 
Small SI and Marine SI fuel tanks would not be allowed in the same 
emission family, it could be possible to carry-across certification 
test data from one category to another.
(d) Compliance Progress Review for Marine Fuel Tanks
    One manufacturer of rotational-molded fuel tanks has stated that 
they are already selling low-permeation tanks into the Small SI market 
and they have plans to sell them into marine applications. However, 
other manufacturers of rotational-molded marine fuel tanks have 
expressed concern that they do not have significant in-use experience 
to demonstrate the durability of low-permeation rotational-molded fuel 
tanks in boats. To address this uncertainty, EPA intends to continue to 
engage on a technical level with rotational-molded marine fuel tank 
manufacturers and material suppliers to assess the progress of low-
permeation fuel tank development and compliance. If systematic problems 
are identified across the industry, this would give EPA the opportunity 
to address the problem. If problems were identified only for individual 
businesses, this would give EPA early notice of the issues that may 
need to be addressed through the proposed hardship relief provisions.
(e) Engineering Design-Based Certification
    In the existing evaporative emission program for recreational 
vehicles, manufacturers using metal fuel tanks may certify by design to 
the tank permeation standards. Tanks using design-based certification 
provisions are not included in the ABT program because they are 
assigned a certification emission level equal to the standard. The 
Panel recommended that we propose to allow design-based certification 
for metal tanks and plastic fuel tanks with a continuous EVOH barrier. 
The Panel also recommended that we propose design-based certification 
for carbon canisters. A detailed description of the proposed design-
based certification options that are consistent with the Panel 
recommendations is presented earlier in this document.
    The National Marine Manufacturers Association (NMMA) the American 
Boat and Yacht Council (ABYC) and the Society of Automotive Engineers 
(SAE) have industry recommended practices for boat designs that must be 
met as a condition of NMMA membership. NMMA stated that they are 
working to update these recommended practices to include carbon 
canister installation instructions and low-permeation fuel line design. 
The Panel recommended that EPA accept data used for meeting the 
voluntary requirements as part of the EPA certification. We are 
proposing that this data could be used as part of EPA certification as 
long as it is collected consistent with the test procedures and other 
requirements described in this proposal.
(f) Hardship Provisions
    We are proposing two types of hardship provisions consistent with 
the Panel recommendations. The first type of hardship is an unusual 
circumstances hardship which would be available to all businesses, 
regardless of size. The second type of hardship is an economic hardship 
provision which would be available to small businesses only. Sections 
VIII.C.8 and VIII.C.9 provide a description of the proposed hardship 
provisions that would apply to the range of manufacturers subject to 
the proposed Marine SI and Small SI evaporative emission requirements. 
This would include Marine SI engine manufacturers, nonhandheld engine 
manufacturers, nonhandheld equipment manufacturers, handheld equipment 
manufacturers, boat builders, and fuel-system component manufacturers.
    The proposed criteria for small businesses are presented earlier in 
Sections III.F.2 and IV.G for Marine SI engine manufacturers, Section 
V.F.2 for nonhandheld engine manufacturers, and Section V.F.3 for 
nonhandheld equipment manufacturers. For handheld equipment 
manufacturers, EPA is proposing to use the existing small-volume 
manufacturer criteria which relies on a production cut-off of 25,000 
pieces of handheld equipment per year. For boat builders and fuel-
system component manufacturers, EPA is proposing to base the 
determination of whether a company is a small business based on the SBA 
definition. The SBA small business definition for companies 
manufacturing boats subject to the proposed standards is fewer than 500 
employees. Likewise, the SBA small business definition for companies 
manufacturing fuel-system components such as fuel tanks and fuel lines 
is fewer than 500 employees.
    Because many boat builders, nonhandheld equipment manufacturers, 
and handheld equipment manufacturers will depend on fuel tank 
manufacturers and fuel line manufacturers to supply certified products 
in time to produce complying vessels and equipment, we are also 
proposing a hardship provision for all boat builders and Small SI 
equipment manufacturers, regardless of size. The proposed hardship 
would allow the boat builder or equipment manufacturer to request more 
time if they are unable to obtain a certified fuel system component and 
they are not at fault and would face serious economic hardship without 
an extension (see Sec.  1068.255). Section VIII.C.10 provides a 
description of the proposed hardship provisions that would apply to 
boat builders and Small SI equipment manufacturers.

H. Technological Feasibility

    We believe there are several strategies that manufacturers can use 
to meet the proposed evaporative emission standards. We have collected 
and will continue to collect emission test data on a wide range of 
technologies for controlling evaporative emissions. The design-based 
certification levels discussed above are based on this test data and we 
may amend the list of approved designs and emission levels as more data 
become available.
    In the following sections we briefly describe how we decided to 
propose specific emission standards and implementation dates, followed 
by a more extensive discussion of the expected emission control 
technologies. A more detailed discussion of the feasibility of the 
proposed evaporative requirements, including all the underlying test 
data, is included in Chapter 5 of the Draft RIA. See Table VI-1 for a 
summary of the proposed evaporative emission standards.
(1) Level of Standards
    The proposed fuel line and fuel tank permeation standards for Small 
SI equipment and Marine SI vessels are based on the standards already 
adopted for recreational vehicles. These applications use similar 
technology in their fuel systems. In cases where the fuel systems 
differ we have identified technological approaches that could be used 
to meet these same emission levels. The control strategies are 
discussed below. For structurally integrated nylon fuel tanks and for 
fuel

[[Page 28189]]

lines used with cold-weather equipment, we are proposing slightly 
relaxed standards based on available permeation data. In addition, we 
have proposed higher numerical standards for fuel tank permeation for 
tests performed at higher temperature (40 [deg]C vs. 28 [deg]C). These 
higher numerical standards are based on data described in Chapter 5 of 
the Draft RIA.
    For fuel tanks installed in personal watercraft and for portable 
marine fuel tanks, we are proposing diurnal emission standards based on 
the current capabilities of these systems. We are basing the proposed 
standard for other installed marine fuel tanks on the capabilities of 
passive systems that store emitted vapors in a carbon canister. The 
Draft RIA describes the test results on passively purged canisters, and 
other technologies, that led us to the proposed level of the diurnal 
emission standard.
    Control of diffusion emissions from Small SI equipment requires 
application of a simple technological approach that is widely used 
today. The Draft RIA describes the testing we conducted on fuel caps 
with tortuous vent paths and short vent lines on which we based the 
diffusion emission standard.
    We have measured running loss emissions and found that some Small 
SI products have very high emission levels. The large variety of 
manufacturers and equipment types makes it impractical to design a 
measurement procedure, which means that we are unable to specify a 
performance standard. We are proposing a design standard for running 
losses from Small SI equipment by specifying that manufacturers may use 
any of a variety of specified design solutions, as described in Section 
VI.C.6. Several of these design options are already in common use 
today.
    We are proposing to require that equipment and vessel manufacturers 
use good engineering judgment in their designs to minimize refueling 
spitback and spillage. In general, it would simply require 
manufacturers to use system designs that are commonly used today. 
Several refueling spitback and spillage control strategies are 
discussed in Chapter 5 of the Draft RIA.
(2) Implementation Dates
    Low-permeation fuel line is available today. Many Small SI 
equipment manufacturers certifying to permeation standards in 
California are selling products with low-permeation fuel line 
nationwide. In addition, many boat builders have begun using low-
permeation marine fuel lines to feed fuel from the fuel tank to the 
engine. For this reason, we are proposing to implement the fuel line 
permeation standards in 2008 for nonhandheld Small SI equipment and in 
2009 for Marine SI vessels. This date is the same as for recreational 
vehicles and is two years later than the California requirements for 
Small SI equipment. For handheld equipment, there are no fuel line 
permeation requirements in California. In addition, injection molded 
fuel lines are common in many applications rather than straight-run 
extruded fuel line. For this reason we are proposing to delay 
implementation of fuel line permeation standards for handheld equipment 
until 2012 (or 2013 for small volume emission families). We request 
comment on the proposed implementation dates for fuel line permeation 
standards.
    Similar to fuel line technology, low-permeation fuel tank 
constructions are used today in automotive and portable fuel tank 
applications. This technology is also being developed for use in 
recreational vehicles and for Small SI equipment sold in California. 
The available technology options include surface treatment and multi-
layer constructions, though rotational molding presents some unique 
design challenges. Based on discussions with fuel tank manufacturers, 
and on our own assessment of the lead time necessary to change current 
industry practices, we believe low-permeation fuel tank technology can 
be applied in the 2011-2012 model years for Small SI and Marine SI fuel 
tanks. We are proposing to implement the fuel tank permeation standards 
in 2011 for Class II equipment and portable and PWC marine fuel tanks. 
For Class I equipment and installed marine fuel tanks, we are proposing 
an implementation date of 2012. We are proposing to phase-in the 
handheld fuel tank standards on the following schedule: 2009 for 
equipment models certifying in California, 2013 for small-volume 
families, and 2010 for the remaining fuel tanks on handheld equipment. 
We believe this will facilitate an orderly transition from current fuel 
tank designs to low-permeation fuel tanks.
    We are proposing the additional year of lead time for the large 
fuel tanks installed in marine vessels largely due to concerns raised 
over the application of low-permeation rotational-molded fuel tank 
technology to marine applications. The majority of these fuel tanks are 
typically rotational-molded by small businesses. Although low-
permeation technology has emerged for these applications, we believe 
additional lead time will be necessary for all manufacturers to be 
ready to implement this technology. This will give these manufacturers 
additional time to make changes to their production processes to comply 
with the standards and to make any tooling changes that may be 
necessary. We are similarly proposing the implementation of fuel tank 
permeation standards for Class I fuel tanks installed in Small SI 
equipment in 2012, mostly to align with the implementation date for the 
Phase 3 exhaust emission standards. This is especially important for 
Class I engines where most of the engine manufacturers will also be 
responsible for meeting all evaporative emission standards. We request 
comment on the proposed implementation dates for the proposed fuel tank 
permeation standards.
    We are proposing to implement the running loss standards for 
nonhandheld Small SI equipment in the same year as the exhaust emission 
standards. We believe this is appropriate because the running loss 
vapor will in some cases be routed to the intake manifold for 
combustion in the engine. Manufacturers would need to account for the 
effect of the additional running loss vapor in their engine 
calibrations. We request comment on this approach.
    We are proposing to implement the proposed diurnal standards for 
portable marine fuel tanks and personal watercraft in 2009. We believe 
these requirements will not result in a significant change from current 
practice so this date will provide sufficient lead time for 
manufacturers to comply with standards. For other installed fuel tanks, 
however, we are proposing a later implementation date of 2010. The 
development of canisters as an approach to control diurnal emissions 
without pressurizing the tanks has substantially reduced the expected 
level of effort to redesign and retool for making fuel tanks. However, 
canister technology has not yet been applied commercially to marine 
applications and additional lead time may be necessary to work out 
various technical parameters, such as design standards and installation 
procedures to ensure component durability and system integrity. We 
request comment on the proposed diurnal implementation dates.
(3) Technological Approaches
    We believe several emission control technologies can be used to 
reduce evaporative emissions from Small SI equipment and Marine SI 
vessels. These emission control strategies are discussed below. Chapter 
5 of the Draft RIA presents more detail on these technologies and 
Chapter 6 provides information on the estimated costs. We request 
comment on these or other technological approaches for reducing

[[Page 28190]]

evaporative emissions from these engines and equipment.
(a) Fuel Line Permeation
    Fuel lines produced for use in Small SI equipment and Marine SI 
applications are generally extruded nitrile rubber with a cover for 
abrasion resistance. Fuel lines used in Small SI applications often 
meet SAE J30 R7 recommendations, including a permeation limit of 550 g/
m2/day at 23 [deg]C on ASTM Fuel C. Fuel lines for personal 
watercraft are typically designed to meet SAE J2046, which includes a 
permeation limit of 300 g/m2/day at 23 [deg]C on ASTM Fuel 
C.\92\ Marine fuel lines subject to Coast Guard requirements under 33 
CFR part 183 are designated as either Type A or Type B and either Class 
1 or Class 2. SAE J1527 provides detail on these fuel line designs. 
Type A fuel lines pass the U.S. Coast Guard fire test while Type B 
designates fuel lines that have not passed this test. Class 1 fuel 
lines are intended for fuel-feed lines where the fuel line is normally 
in contact with liquid fuel and has a permeation limit of 100 g/
m2/day at 23 [deg]C. Class 2 fuel lines are intended for 
vent lines and fuel fill necks where liquid fuel is not continuously in 
contact with the fuel line; it has a permeation limit of 300 g/
m2/day at 23 [deg]C. In general practice, most boat builders 
use Class 1 fuel lines for both vent lines and fuel-feed lines to avoid 
carrying two types of fuel lines. Most fuel fill necks, which have a 
much larger diameter and are constructed differently, use materials 
meeting specifications for Class 2 fuel lines. The marine industry is 
currently in the process of revising SAE J1527 to include a permeation 
rating of 15 g/m2/day at 23 [deg]C on fuel CE10 for marine 
fuel lines.
---------------------------------------------------------------------------

    \92\ Society of Automotive Engineers Surface Vehicle Standard, 
``Personal Watercraft Fuel Systems,'' SAE J2046, Issues 1993-01-19 
(Docket EPA-HQ-OAR-2004-0008-0179).
---------------------------------------------------------------------------

    Low-permeability fuel lines are in production today. One fuel line 
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 acetate. 
Barrier approaches in automotive applications include fuel lines with 
fluoroelastomers such as FKM and fluoroplastics such as Teflon and THV. 
In addition to presenting data on low-permeation fuel lines, Chapter 5 
of the Draft RIA lists several fuel-system materials and their 
permeation rates. Molded rubber fuel line components, such as primer 
bulbs and some handheld fuel lines, could meet the standard by using a 
fluoroelastomer such as FKM. The Draft RIA also discusses low-
permeation materials that retain their flexibility at very low 
temperatures.
    Automotive fuel lines made of low-permeation plastic tubing are 
generally made from fluoroplastics. An added benefit of these low-
permeability fuel lines is that some fluoropolymers can be made to 
conduct electricity and therefore prevent the buildup of static 
charges. This type of fuel line can reduce permeation by more than an 
order of magnitude below the level associated with barrier-type fuel 
lines, but it is relatively inflexible and would need to be molded in 
specific shapes for each equipment or vessel design. Manufacturers have 
commented that they need flexible fuel lines to fit their many designs, 
resist vibration, prevent kinking, and simplify connections and 
fittings. An alternative to custom molding is to manufacture fuel lines 
with a corrugated profile (like a vacuum hose). Producing flexible 
fluoropolymer fuel lines is somewhat more expensive but the result is a 
product that meets emission standards without compromising in-use 
performance or ease of installation.
(b) Fuel Tank Permeation
    Blow-molding is widely used for the manufacture of Small SI, 
portable marine, and PWC fuel tanks. Typically, blow-molding is 
performed by creating a hollow tube, known as a parison, by pushing 
high-density polyethylene (HDPE) through an extruder with a screw. The 
parison is then pinched in a mold and inflated with an inert gas. In 
highway applications, nonpermeable plastic fuel tanks are produced by 
blow molding a layer of ethylene vinyl alcohol (EVOH) or nylon between 
two layers of polyethylene. This process is called coextrusion and 
requires at least five layers: the barrier layer, adhesive layers on 
either side of the barrier layer, and two outside layers of HDPE that 
make up most of the thickness of the fuel tank walls. However, multi-
layer construction requires additional extruder screws, which 
significantly increases the cost of the blow-molding process. One 
manufacturer has developed a two-layer barrier approach using a 
polyarylamide inner liner. This technology is not in production yet but 
appears to be capable of permeation levels similar to the traditional 
EVOH barrier designs. This approach would enable blow-molding of low-
permeation fuel tanks with only one additional extruder screw.
    Multi-layer fuel tanks can also be formed using injection molding. 
In this method a low-viscosity polymer is forced into a thin mold to 
create the two sides of the fuel tank (e.g., top and bottom), which are 
then fused together. To add a barrier layer, a thin sheet of the 
barrier material is placed inside the mold before injecting the 
poleythylene. The polyethylene, which generally has a much lower 
melting point than the barrier material, bonds with the barrier 
material to create a shell with an inner liner.
    A less expensive alternative to coextrusion is to blend a low-
permeable resin with the HDPE and extrude it with a single screw to 
create barrier platelets. The trade name typically used for this 
permeation control strategy is Selar. The low-permeability resin, 
typically EVOH or nylon, creates noncontinuous platelets in the HDPE 
fuel tank to reduce permeation by creating long, tortuous pathways that 
the hydrocarbon molecules must navigate to escape through the fuel tank 
walls. Although the barrier is not continuous, this strategy can still 
achieve greater than a 90 percent reduction in permeation of gasoline. 
EVOH has much higher permeation resistance to alcohol than nylon so it 
would likely be the preferred material for meeting the proposed 
standard based on testing with a 10 percent ethanol fuel.
    Many fuel tanks for Small SI equipment are injection-molded out of 
either HDPE or nylon. Injection-molding can be used with lower 
production volumes than blow-molding due to lower tooling costs. In 
this method, a low-viscosity polymer is forced into a thin mold to 
create the two sides of the fuel tank; these are then fused together 
using vibration, hot plate or sonic welding. A strategy such as Selar 
has not been demonstrated to work with injection-molding due to high 
shear forces.
    An alternative to injection-molding is thermoforming which is also 
cost-effective for lower production volumes. In this process, sheet 
material is heated and then drawn into two vacuum dies. The two halves 
are then fused while the plastic is still molten to form the fuel tank. 
Low-permeation fuel tanks can be constructed using this process by 
using multi-layer sheet material. This multi-layer sheet material can 
be extruded using similar materials to multi-layer blow-molded fuel 
tank designs. A typical barrier construction would include a thin EVOH 
barrier, adhesion layers on both sides, a layer of HDPE regrind, and 
outside layers of pure virgin HDPE.
    Regardless of the molding process, another type of low-permeation 
technology for HDPE fuel tanks would

[[Page 28191]]

be to treat the surfaces with a barrier layer. Two ways of achieving 
this are known as fluorination and sulfonation. The fluorination 
process causes a chemical reaction where exposed hydrogen atoms are 
replaced by larger fluorine atoms, which creates a barrier on the 
surface of the fuel tank. In this process, batches of fuel tanks are 
generally processed post-production by stacking them in a steel 
container. The container is then 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. Fluorinating with this 
process would treat both the inside and outside surfaces of the fuel 
tank, thereby improving the reliability and durability of the 
permeation-resistance. As an alternative, fuel tanks can be fluorinated 
during production by exposing the inside surface of the fuel tank to 
fluorine during the blow-molding process. However, this method may not 
prove as effective as post-production fluorination.
    Sulfonation is another surface treatment technology where sulfur 
trioxide is used to create the barrier by reacting with the exposed 
polyethylene to form sulfonic acid groups on the surface. Current 
practices for sulfonation are to place fuel tanks on a small assembly 
line and expose the inner surfaces to sulfur trioxide, then rinse with 
a neutralizing agent. However, sulfonation can also be performed using 
a batch method. Either of these sulfonation processes can be used to 
reduce gasoline permeation by more than 95 percent.
    Over the first month or so of use, polyethylene fuel tanks can 
experience a material expansion of as much as three percent due to 
saturation of the plastic with fuel. Manufacturers have raised the 
concern that this hydrocarbon expansion could degrade the effectiveness 
of surface treatments like fluorination or sulfonation. However, we 
believe this will not significantly affect these surface treatments. 
California ARB has performed extensive permeation testing on portable 
fuel containers with and without these surface treatments. Prior to the 
permeation testing, the tanks were prepared by performing a durability 
procedure where the fuel container cycled a minimum of 1,000 times 
between--1 psi and 5 psi. In addition, the fuel containers were soaked 
with fuel for a minimum of four weeks before testing. Their test data, 
presented in Chapter 5 of the Draft RIA, show that fluorination and 
sulfonation are still effective after this durability testing. We have 
conducted our own permeation testing on fluorinated fuel tanks that 
have been exposed to fuel for more than a year with excellent results. 
These results are presented in the Draft RIA.
    Manufacturers have also commented that fuel sloshing in the tank 
under normal in-use operation could wear off the surface treatments. 
However, we believe this is unlikely to occur. These surface treatments 
actually result in an atomic change in the structure of the surface of 
the fuel tank. To wear off the treatment, the plastic itself would need 
to be worn away. In addition, testing by California ARB shows that the 
fuel tank permeation standard can be met by fuel tanks that have 
undergone 1.2 million slosh cycles. Test data on a sulfonated 
automotive HDPE fuel tank after five years of use showed no 
deterioration in the permeation barrier. These data are presented in 
Chapter 5 of the Draft RIA.
    A fourth method for molding plastic fuel tanks is called 
rotational-molding. Rotational-molding is a lower-cost alternative for 
smaller production volumes. In this method, a mold is filled with a 
powder form of polyethylene with a catalyst material. While the mold is 
rotated in an oven, the heat melts the plastic. When cross-link 
polyethylene (XLPE) is used, this heat activates a catalyst in the 
plastic, which causes a strong cross-link material structure to form. 
This method is often used for relatively large fuel tanks in Small SI 
equipment and for installed marine fuel tanks. The advantages of this 
method are low tooling costs, which allow for smaller production 
volumes, and increased strength and flame resistance. Flame resistance 
is especially important for installed marine fuel tanks subject to 33 
CFR part 183. At this time, the barrier treatment approaches discussed 
above for HDPE have not been demonstrated to be effective for XLPE.
    We have evaluated two permeation control approaches for rotational-
molded fuel tanks. The first is to form an inner layer during the 
molding process. Historically, the primary approach for this is to use 
a drop-box that opens after the XLPE tank begins to form. However, 
processes have been developed that eliminate the need for a drop box. 
With this construction a low-permeation inner liner can be molded into 
the fuel tank. Manufacturers are currently developing acetyl copolymer, 
nylon, and polybutylene terephthalate inner liners for this 
application. In fact, one fuel tank manufacturer is already selling 
tanks with a nylon inner liner into Class II Small SI equipment 
applications. Initial testing suggests that these barrier layers could 
be used to achieve the proposed standards.
    The second approach to creating a barrier layer on XLPE rotational-
molded fuel tanks is to use an epoxy barrier coating. One manufacturer 
has demonstrated that a low-permeation barrier coating can be adhered 
to an XLPE fuel tank that results in a permeation rate below the 
proposed standard. In this case, the manufacturer used a low level of 
fluorination to increase the surface energy of the XLPE so the epoxy 
would adhere properly.
    Marine fuel tanks are also fabricated out of either metal or 
fiberglass. Metal does not permeate so tanks that are constructed and 
installed properly to prevent corrosion should meet the proposed 
standards throughout their full service life. For fiberglass fuel 
tanks, one manufacturer has developed a composite that has been 
demonstrated to meet the proposed fuel tank permeation standard. 
Permeation control is achieved by incorporating fillers into a resin 
system and coating the assembled tank interior and exterior. This 
filler is made up of nanocomposites (very small particles of treated 
volcanic ash) which are dispersed into a carrier matrix. These 
particles act like the barrier platelets discussed above by creating a 
tortuous pathway for hydrocarbon migration through the walls of the 
fuel tank.
(c) Diurnal
    Portable marine fuel tanks are currently equipped with a valve that 
can be closed by the user when the tank is stored to hold vapor in the 
fuel tank. These fuel tanks are designed to hold the pressure that 
builds up when a sealed fuel tank undergoes normal daily warming. This 
valve must be opened when the engine is operating to prevent a vacuum 
from forming in the fuel tank as the fuel level in the tank decreases. 
A vacuum in the fuel tank could prevent fuel from being drawn into the 
engine. Because the valve is user-controlled, any emission control is 
dependent on user behavior. This can be corrected by replacing the 
user-controlled valve with a simple one-way valve in the fuel cap. For 
instance, a diaphragm valve that is common in many automotive 
applications seals when under pressure but opens at low-vacuum 
conditions.
    Personal watercraft currently use sealed systems with pressure-
relief valves that start venting vapors when pressures reach a 
threshold that ranges from 0.5 to 4.0 psi. We believe the proposed 
standard can be met through the use of a sealed fuel system with a 1.0 
psi pressure-relief valve. Personal watercraft should therefore be able 
to meet the proposed standard with little or no change to current 
designs.

[[Page 28192]]

    For other vessels with installed fuel tanks, manufacturers have 
commented that even 1.0 psi of pressure would be too high for their 
applications. They expressed concern that their fuel tanks had large, 
flat surfaces that would deform or leak at pressures of 0.5 psi or 
higher. This concern led us to consider several technologies for 
controlling diurnal emissions without pressurizing the tank, including 
carbon canisters, volume-compensating air bags, and bladder fuel tanks.
    The primary evaporative emission control device used in automotive 
applications is a carbon canister. With this technology, vapor 
generated in the tank is vented to a canister containing activated 
carbon. The fuel tank must be sealed such that the only venting that 
occurs is through the carbon canister. This prevents more than a 
minimal amount of positive or negative pressure in the tank. The 
activated carbon collects and stores the hydrocarbons. The activated 
carbon bed in the canister is refreshed by purging.
    In a marine application, an engine purge is not practical; 
therefore, canisters were not originally considered to be a practical 
technology for controlling diurnal vapor from boats. Since that time, 
however, we have collected information showing that the canister is 
purged sufficiently during cooling periods to reduce diurnal emissions 
effectively. When the fuel in the tank cools, fresh air is drawn back 
through the canister into the fuel tank. This fresh air partially 
purges the canister and returns hydrocarbons to the fuel tank. This 
creates open sites in the carbon so the canister can again collect 
vapor during the next heating event. Test data presented in Chapter 5 
of the Draft RIA show that a canister starting from empty is more than 
90 percent effective until it reaches the point of saturation. Once it 
reaches saturation, a canister is still capable of reducing diurnal 
emissions by more than 60 percent due to the normal airflow across the 
canister bed during cooling periods. Adding active purging during 
engine operation would improve the level of control somewhat depending 
on how often the engine is operated.
    Manufacturers have raised the concern that it is common for fuel to 
pass out the vent line during refueling. If there were a canister in 
the vent line it would become saturated with fuel. While this would not 
likely cause permanent damage to the canister, we believe marine fuel 
systems should prevent liquid fuel from exiting the vent line for both 
environmental and safety reasons. A float valve or small orifice in the 
entrance to the vent line from the fuel tank would prevent liquid fuel 
from reaching the canister or escaping from the tank. Any pressure 
build-up from such a valve would cause fuel to back up the fill neck 
and shut off the fuel dispensing nozzle. Manufacturers have also 
expressed concerns for canister durability in marine applications due 
to vibration, shock, and humidity. However, there are now marine grades 
of activated carbon that are harder and more moisture-resistant than 
typical automotive carbon. Industry installed canisters equipped with 
the marine grade carbon on 14 boats in a pilot program and no problems 
were encountered. This is discussed in more detail in Chapter 5 of the 
Draft RIA.
    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 above the liquid fuel. By minimizing 
the vapor space, the equilibrium concentration of fuel vapors occupies 
a smaller volume, resulting in a smaller mass of vapors. As the 
equilibrium vapor concentration increases with increasing temperature, 
the vapor space expands, which forces air out of the bag through the 
vent to atmosphere. Because the bag volume decreases to compensate for 
the expanding vapor space, total pressure inside the fuel tank stays 
very close to atmospheric pressure. Once the fuel tank cools in 
response to cooling ambient temperatures the resulting vacuum in the 
fuel tank will make the bag expand again by drawing air from the 
surrounding environment. Our test results show that pressure could be 
kept below 0.8 psi using a bag with a capacity equal to 25 percent of 
the fuel tank capacity. The use of a volume-compensating air bag, in 
conjunction with a pressure-relief valve, would be very effective in 
controlling diurnal emissions.
    Probably the most effective technology for reducing diurnal 
emissions from marine fuel tanks is through the use of a collapsible 
fuel bladder. In this concept, a low-permeation bladder is installed in 
the fuel tank to hold the fuel. As fuel is drawn from the bladder the 
vacuum created collapses the bladder. There is, therefore, no vapor 
space and no pressure build-up from fuel heating. No vapors would be 
vented to the atmosphere since the bladder is sealed. This option could 
also significantly reduce emissions during refueling that would 
normally result from dispensed fuel displacing vapor in the fuel tank. 
We have received comments that this would be cost-prohibitive because 
it could increase costs from 30 to 100 percent, depending on tank size. 
However, bladder fuel tanks have safety advantages and they are already 
sold by at least one manufacturer to meet market demand in niche 
applications.
(d) Running Loss
    Running loss emissions can be controlled by sealing the fuel cap 
and routing vapors from the fuel tank to the engine intake. In doing 
so, vapors generated by heat from the engine will be burned in the 
engine's combustion chamber. It may be necessary to use a valve or 
limited-flow orifice in the purge line to prevent too much fuel vapor 
from reaching the engine and to prevent liquid fuel from entering the 
line if the equipment flips over. Depending on the configuration of the 
fuel system and purge line, a one-way valve in the fuel cap may be 
desired to prevent a vacuum in the fuel tank during engine operation. 
We anticipate that a system like this would eliminate running loss 
emissions. However, higher temperatures during operation and the 
additional length of vapor line would slightly increase permeation. 
Considering these effects, we still believe that the system described 
here would reduce running losses from Small SI equipment by more than 
90 percent. Other approaches would be to move the fuel tank away from 
heat sources or to use heat protection such as a shield or directed air 
flow.
    We are not considering running loss controls for marine vessels. 
For portable fuel tanks and installed fuel tanks on larger vessels we 
would expect the significant distance from the engine and the cooling 
effect of operating the vessel in water to prevent significant heating 
of the fuel tanks during engine operation. For personal watercraft, 
fuel tanks have a sealed system with pressure relief that should help 
contain running loss emissions. For other installed fuel tanks, we 
would expect the system for controlling diurnal emissions would capture 
about half of any running losses that would occur.
(e) Diffusion
    Many manufacturers today use fuel caps that effectively limit the 
diffusion of gasoline from fuel tanks. In fact, the proposed diffusion 
emission standard for Small SI equipment is based to a large degree on 
the diffusion control capabilities of these fuel caps. As discussed in 
Chapter 5 of the Draft RIA, venting a fuel tank through a tube (rather 
than through an open orifice) also greatly reduces diffusion. We have 
conducted additional testing with short, narrow-diameter vent lines 
that provide

[[Page 28193]]

enough resistance to diffusion to meet the proposed emission standards.
    A secondary benefit of the running loss control described above for 
Small SI equipment relates to diffusion emissions. In a system that 
vents running loss vapors to the engine, venting losses would occur 
through the vapor line to the engine intake, rather than through open 
vents in the fuel cap. This approach should therefore eliminate 
diffusion emissions.
(4) Regulatory Alternatives
    We considered both less and more stringent evaporative emission 
control alternatives for fuel systems used in Small SI equipment and 
Marine SI vessels. Chapter 11 of the Draft RIA presents details on this 
analysis of regulatory alternatives. The results of this analysis are 
summarized below. We believe the proposed permeation standards are 
reflective of available technology and represent a step change in 
emissions performance. Therefore, we consider the same permeation 
control scenario in the less stringent and more stringent regulatory 
alternatives.
    For Small SI equipment, we considered a less stringent alternative 
without running loss emission standards Small SI engines. However, we 
believe controlling running loss and diffusion emissions from 
nonhandheld equipment is feasible at a relatively low cost. Running 
loss emissions can be controlled by sealing the fuel cap and routing 
vapors from the fuel tank to the engine intake. Other approaches would 
be to move the fuel tank away from heat sources or to use heat 
protection such as a shield or directed air flow. Diffusion can be 
controlled by simply using a tortuous tank vent path, which is commonly 
used today on Small SI equipment to prevent fuel splashing or spilling. 
These emission control technologies are relatively straight-forward, 
inexpensive, and achievable in the near term. Not requiring these 
controls would be inconsistent with section 213 of the Clean Air Act. 
For a more stringent alternative, we considered applying a diurnal 
emission standard for all Small SI equipment. We believe passively 
purging carbon canisters could reduce diurnal emissions by 50 to 60 
percent from Small SI equipment. However, we believe some important 
issues would need to be resolved for diurnal emission control, such as 
cost, packaging, and vibration. The cost sensitivity is especially 
noteworthy given the relatively low emissions levels (on a per-
equipment basis) from such small fuel tanks.
    For marine vessels, we considered a less stringent alternative, 
where there would be no diurnal emission standard for vessels with 
installed fuel tanks. However, installed fuel tanks on marine vessels 
are much larger in capacity than those used in Small SI applications. 
Our analysis indicates that traditional carbon canisters are feasible 
for boats at relatively low cost. While packaging and vibration are 
also issues with marine applications, we believe these issues have been 
addressed. Carbon canisters were installed on fourteen boats by 
industry in a pilot program. The results demonstrated the feasibility 
of this technology. The proposed standards would be achievable through 
engineering design-based certification with canisters that are very 
much smaller than the fuel tanks. In addition, sealed systems, with 
pressure control strategies would be accepted under the proposed 
engineering design-based certification. For a more stringent scenario, 
we consider a standard that would require boat builders to use an 
actively purged carbon canister. This means that, when the engine is 
operating, it would draw air through the canister to purge the canister 
of stored hydrocarbons. However, we rejected this option because active 
purge occurs infrequently due to the low hours of operation per year 
seen by many boats. The gain in overall efficiency would be quite small 
relative to the complexity active purge adds into the system in that 
the engine must be integrated into a vessel-based control strategy. The 
additional benefit of an actively purged diurnal control system is 
small in comparison to the cost and complexity of such a system.
(5) Our Conclusions
    We believe the proposed evaporative emission standards reflect what 
manufacturers can achieve through the application of available 
technology. We believe the proposed lead time is necessary and adequate 
for fuel tank manufacturers, equipment manufacturers, and boat builders 
to select, design, and produce evaporative emission control strategies 
that will work best for their product lines. We expect that meeting 
these requirements will pose a challenge, but one that is feasible when 
taking into consideration the availability and cost of technology, lead 
time, noise, energy, and safety. The role of these factors is presented 
in detail in Chapters 5 and 6 of the Draft RIA. As discussed in Section 
X, we do not believe the proposed standards would have negative effects 
on energy, noise, or safety and may lead to some positive effects.

VII. General Concepts Related to Certification and Other Requirements

    This section describes general concepts concerning the proposed 
emission standards and various requirements related to these standards. 
There is a variety of proposed requirements that serve to ensure 
effective implementation of the emission standards, such as applying 
for certification, labeling engines, and meeting warranty requirements. 
The following discussion reviews these requirements for Small SI 
engines and outboard and personal-watercraft engines that have already 
been subject to exhaust emission standards, explains a variety of 
changes, and describes how these provisions apply to evaporative 
emissions. Sterndrive and inboard marine engines will be subject to 
emission standards for the first time so all these requirements are new 
for those engines.
    Rather than making changes to existing regulations, we have drafted 
new regulatory text describing the new emission standards and related 
requirements and included that text in this proposal. The proposed 
regulations are written in plain-language format. In addition to the 
improved clarity of the regulatory text, this allows us to harmonize 
the regulations with our other programs requiring control of engine 
emissions.\93\
---------------------------------------------------------------------------

    \93\ For additional background related to plans for migrating 
regulations, see ``Plain Language Format of Emission Regulations for 
Nonroad Engines,'' EPA420-F-02-046, September 2002 (http://www.epa.gov/otaq/regs/nonroad/2002/f02046.pdf).
---------------------------------------------------------------------------

    The proposed regulatory text migrates the existing requirements for 
Small SI engines, including all the emission standards and other 
requirements related to getting and keeping a valid certificate of 
conformity, from 40 CFR part 90 to 40 CFR part 1054. For nonhandheld 
engines, manufacturers must comply with all the provisions in part 1054 
once the Phase 3 standards begin to apply in 2011 or 2012. For handheld 
engines, manufacturers must comply with the provisions in part 1054 
starting in 2010. Similarly, we are proposing to migrate the existing 
requirements for Marine SI engines from 40 CFR part 91 to 40 CFR part 
1045. Manufacturers must comply with the provisions in part 1045 for an 
engine once the proposed exhaust emission standards begin to apply in 
2009.
    The proposed requirements for evaporative emissions are described 
in 40 CFR part 1060, with some category-specific provisions in 40 CFR 
parts 1045 and 1054, which are referred to as the exhaust standard-
setting parts for each

[[Page 28194]]

type of engine. Adopting the provisions related to evaporative 
emissions in a broadly applicable part has two main advantages. First, 
we anticipate that in many cases boat builders, equipment 
manufacturers, and manufacturers of fuel-system components will need to 
certify their products only to the standards for evaporative emissions, 
with no corresponding responsibility for exhaust emissions. These 
companies will not need to focus on the exhaust standard-setting part 
except to read the short section defining the evaporative emission 
standards and requirements. Second, manufacturers of fuel-system 
components make products that are not necessarily unique to a specific 
category of engines. The regulations in 40 CFR parts 1045 and 1054 will 
highlight the standards that apply and provide any specific directions 
in applying the general provisions in part 1060. The standards, test 
procedures, and certification provisions are almost completely uniform 
across our programs so this combined set of evaporative-related 
provisions will make it much easier for companies to certify their 
products if they are not subject to the exhaust emission standards. In 
Section XI we describe how we might apply the provisions of part 1060 
to recreational vehicles regulated under 40 CFR part 1051.
    Other provisions describing general testing procedures, including 
detailed laboratory and equipment specifications and procedures for 
equipment calibration and emission measurements, are written in 40 CFR 
part 1065. The exhaust standard-setting parts also include testing 
specifications that are specific to each type of engine, including duty 
cycles, test-fuel specifications, and procedures to establish 
deterioration factors. See Section IX for further discussion of these 
test procedures. Engines, equipment, and vessels subject to the new 
standard-setting parts (parts 1045, 1054, and 1060) will also be 
subject to the general compliance provisions in 40 CFR part 1068. These 
include prohibited acts and penalties, exemptions and importation 
provisions, selective enforcement audits, defect reporting and recall, 
and hearing procedures. See Section VIII for further discussion of 
these general compliance provisions. Both part 1065 and part 1068 
already apply to various other engine categories. We are therefore 
publishing in this proposal only the changes needed to apply the 
existing regulations to the engines, equipment, and vessels covered by 
this rulemaking.

A. Scope of Application

    This proposal covers spark-ignition propulsion marine engines and 
vessels powered by those engines introduced into commerce in the United 
States. The proposal also covers other nonroad spark-ignition engines 
rated at or below 19 kW and the corresponding equipment. The following 
sections describe generally when emission standards apply to these 
products. Refer to the specific program discussion in Sections III 
through VI for more information about the scope of application and 
timing of the proposed standards.
(1) Do the standards apply to all engines, equipment, and vessels or 
only to new products?
    The scope of this proposal is broadly set by Clean Air Act section 
213(a)(3), which instructs us to set emission standards for new nonroad 
engines and new nonroad vehicles. Generally speaking, the proposed rule 
is intended to cover all new engines and vehicles in the identified 
categories (including any associated vehicles, vessels, or other 
equipment). Once the emission standards apply to an engine, piece of 
equipment, or fuel-system component manufacturers must get a 
certificate of conformity from us before selling them or otherwise 
introducing them into commerce in the United States. Note that the term 
``manufacturer'' includes any individual or company introducing into 
commerce in the United States engines, equipment, vessels, or 
components that are subject to emission standards. These Clean Air Act 
requirements relate to importation and any other means of introducing 
covered products into commerce. In addition to any applicable 
evaporative requirements, we also require equipment manufacturers that 
install engines from other companies to install only certified engines 
once emission standards apply. The certificate of conformity (and 
corresponding emission control information label) provides assurance 
that manufacturers have met their obligation to make engines, 
equipment, and vessels that meet emission standards over the useful 
life we specify in the regulations.
(2) How do I know if my engine or equipment is new?
    We are proposing to define ``new'' consistent with previous 
rulemakings. Under the proposed definition, a nonroad engine (or 
nonroad equipment) is considered new until its title has been 
transferred to the ultimate purchaser or the engine has been placed 
into service. This proposed definition would apply to engines, 
equipment, and vessels so the nonroad equipment using these engines 
would be considered new until their title has been transferred to an 
ultimate buyer. In Section VII.B.1 we describe how to determine the 
model year of individual engines, equipment, and vessels.
    To further clarify the proposed definition of new nonroad engine, 
we are proposing to specify that a nonroad engine, equipment, or vessel 
is placed into service when it is used for its intended purpose. We are 
therefore proposing that an engine subject to the proposed standards is 
used for its intended purpose when it is installed in a vessel or other 
piece of nonroad equipment. We need to make this clarification because 
some engines are made by modifying a highway or land-based nonroad 
engine that has already been installed on a vessel or other piece of 
equipment, so without this clarification, these engines may escape 
regulation. For example, an engine installed in a marine vessel after 
it has been used for its intended purpose as a land-based highway or 
nonroad engine is considered ``new'' under this definition. We believe 
this is a reasonable approach because the practice of adapting used 
highway or land-based nonroad engines may become more common if these 
engines are not subject to the standards in this proposal.
    In summary, an engine would be subject to the proposed standards if 
it is:
     Freshly manufactured, whether domestic or imported; this 
may include engines produced from engine block cores;
     Installed for the first time in nonroad equipment after 
having powered a car, a truck, or a category of nonroad equipment 
subject to different emission standards;
     Installed in new nonroad equipment, regardless of the age 
of the engine; or
     Imported--whether new or used, as long as the engine was 
not built before the initial emission standards started to apply.
(3) When do imported engines, equipment, and vessels need to meet 
emission standards?
    The proposed emission standards would apply to all new engines, 
equipment, and vessels that are used in the United States. According to 
Clean Air Act section 216 ``new'' includes engines or equipment that 
are imported by any person, whether freshly manufactured or used. Thus, 
the proposed program would include

[[Page 28195]]

engines that are imported for use in the United States whether they are 
imported as loose engines or are already installed on a vessel or other 
piece of nonroad equipment built elsewhere. All imported engines would 
need an EPA-issued certificate of conformity to clear customs, with 
limited exemptions (as described in Section VIII).
    If an engine or piece of nonroad equipment that was built after 
emission standards take effect is imported without a currently valid 
certificate of conformity, we would still consider it to be a new 
engine, equipment, or vessel. This means it would need to comply with 
the emission standards that apply based on its model year. Thus, for 
example, a marine vessel manufactured in a foreign country in 2009, 
then imported into the United States in 2010, would be considered 
``new.'' The engines on that piece of equipment would have to comply 
with the requirements for the 2009 model year, assuming that the engine 
has not been modified and no other exemptions apply. This provision is 
important to prevent manufacturers from avoiding emission standards by 
building products abroad, transferring their title, and then importing 
them as used products. Note that if an imported engine has been 
modified it must meet emission standards based on the year of 
modification rather than the year of manufacture. See Section V.E.6 and 
Section XI.C for proposed and contemplated restrictions related model 
years for importation of new engines and equipment.
(4) Do the standards apply to exported engines, equipment, or vessels?
    Engines, equipment, or vessels intended for export would generally 
not be subject to the requirements of the proposed emission control 
program, except that we would not exempt engines exported to countries 
having standards identical to the United States. However, engines, 
equipment, or vessels that are exported and subsequently re-imported 
into the United States must be certified. For example, this would be 
the case when a foreign company purchases engines manufactured in the 
United States for installation in nonroad equipment for export back to 
the United States. Those engines would be subject to the emission 
standards that apply on the date the engine was originally 
manufactured. If the engine is later modified and certified (or 
recertified), the engine is subject to emission standards that apply on 
the date of the modification. So, for example, foreign equipment 
manufacturers buying U.S.-made engines without recertifying the engines 
will need to make sure they purchase complying engines for the products 
they sell in the United States.
(5) Are there any new products that would be exempt from the emission 
standards?
    We are proposing to extend our basic nonroad exemptions to the 
engines, equipment, and vessels covered by this proposal. These include 
the testing exemption, the manufacturer-owned exemption, the display 
exemption, and the national security exemption. These exemptions are 
described in more detail in Section VIII.C.
    In addition, the Clean Air Act does not consider engines used 
solely for competition to be nonroad engines so the proposed emission 
standards do not apply to them. The Clean Air Act similarly does not 
consider engines used in stationary applications to be nonroad engines; 
however, EPA has proposed to apply emission standards for stationary 
spark-ignition engines that are comparable to the standards that apply 
to nonroad engines (71 FR 33804, June 12, 2006). As described in 
Section V, we are proposing in this notice to apply the Phase 3 
standards for Small SI engines equally to stationary spark-ignition 
engines at or below 19 kW. Refer to the program discussions in Sections 
III through VI for a discussion of how the various exclusions apply for 
different categories of engines.

B. Emission Standards and Testing

(1) How is the model year determined?
    The proposed emission standards are effective on a model-year 
basis. We are proposing to define model year much like we do for 
passenger cars. It would generally mean either the calendar year or 
some other annual production period based on the manufacturer's 
production practices. For example, manufacturers could start selling 
2006 model year engines as early as January 2, 2005 as long as the 
production period extends until at least January 1, 2006. All of a 
manufacturer's engines from a given model year would have to meet 
emission standards for that model year. For example, manufacturers 
producing new engines in the 2006 model year would need to comply with 
the 2006 standards.
(2) How do adjustable engine parameters affect emission testing?
    Many engines are designed with components that can be adjusted for 
optimum performance under changing conditions, such as varying fuel 
quality, high altitude, or engine wear. Examples of adjustable 
parameters include spark timing, idle speed setting, and fuel injection 
timing. While we recognize the need for this practice, we are also 
concerned that engines maintain a consistent level of emission control 
for the whole range of adjustability. We are therefore proposing to 
require that engines meet emission standards over the full adjustment 
range.
    Manufacturers would have to provide a physical stop to prevent 
adjustment outside the established range. Operators would then be 
prohibited from adjusting engines outside this range. Refer to the 
proposed regulatory text for more information about adjustable engine 
parameters. See especially the proposed sections 40 CFR 1045.115 for 
Marine SI engines and 40 CFR 1054.115 for Small SI engines.
(3) Alternate Fuels
    The emission standards apply to all spark-ignition engines 
regardless of the fuel they use. Almost all Marine SI engines and Small 
SI engines operate on gasoline, but these engines may also operate on 
other fuels, such as natural gas, liquefied petroleum gas, ethanol, or 
methanol. The test procedures in 40 CFR part 1065 describe adjustments 
needed for operating test engines with oxygenated fuels.
    In some special cases, a single engine is designed to alternately 
run on different fuels. For example, some engines can switch back and 
forth between natural gas and LPG. We request comment on the best way 
of certifying such engines so they can be in a single engine family, 
even though we would normally require engines operating on different 
fuels to be in separate engine families. We could require such 
manufacturers to conduct emission testing with emission-data engines 
operating on both fuels to establish the worst-case configuration. In 
particular, we request comment on the appropriate data for 
demonstrating compliance at the end of the service-accumulation period 
for durability testing.
    Once an engine is placed into service, someone might want to 
convert it to operate on a different fuel. This would take the engine 
out of its certified configuration, so we are proposing to require that 
someone performing such a fuel conversion go through a certification 
process. We would expect to allow certification of the complete engine 
using normal certification procedures, or the aftermarket conversion 
kit could be certified using the provisions of 40 CFR part 85, subpart 
V. This contrasts with the existing provisions that allow for fuel 
conversions that can be demonstrated

[[Page 28196]]

not to increase emission levels above the applicable standard. We 
propose to apply this requirement starting January 1, 2010. (See Sec.  
90.1003 and Sec.  1054.635.)

C. Demonstrating Compliance

    We are proposing a compliance program to accompany emission 
standards. This consists first of a process for certifying engine 
models and fuel systems (either as a part of or independently from the 
vessel or equipment). In addition to certification, we are proposing 
several provisions to ensure that emission control systems continue to 
function over long-term operation in the field. Most of these 
certification and durability provisions are consistent with previous 
rulemakings for these and other nonroad engines, equipment, and 
vessels. Refer to the discussion of the specific programs in Sections 
III through VI for additional information about these requirements for 
each engine category.
(1) How would I certify my engines, equipment, or vessels?
    Sections III through VI describe the proposed emission standards 
for new engines, equipment, and vessels. Section VI in particular 
describes which companies are responsible for certifying to the new 
standards. This section describes the general certification process.
    We are proposing a certification process similar to that already 
adopted for these and other engines and equipment. Certifying 
manufacturers generally test representative prototype engines or fuel 
system components and submit the emission data along with other 
information to EPA in an application for a Certificate of Conformity. 
If we approve the application, then the manufacturer's Certificate of 
Conformity allows the manufacturer to sell the engines, equipment, or 
vessels described in the application in the United States. We are 
proposing to include clarifying language to specify that the 
certificate is valid starting with the indicated effective date, but 
that it is not valid for any production after December 31 of the model 
year for which it is issued. We are also proposing a provision to 
preclude issuance of certificates after December 31 of a given model 
year. This would avoid a situation in which a manufacturer receives 
certification after it is no longer valid for further production.
    We are proposing that manufacturers certify their engine models by 
grouping them into emission families. Under this approach, engines 
expected to have similar emission characteristics would be classified 
in the same emission family. The emission family definition is 
fundamental to the certification process and to a large degree 
determines the amount of testing required for certification. The 
proposed regulations include specific engine characteristics for 
grouping emission families for each category of products. To address a 
manufacturer's unique product mix, we may approve using broader or 
narrower emission families as long as the manufacturer can show that 
all the engines in an engine family will have similar emission control 
characteristics over the engines' useful life.
    The useful life period specified in the regulations defines the 
period over which manufacturers are responsible for meeting emission 
standards. The useful life values included in our regulations are 
intended to reflect the period during which engines are designed to 
properly function without being remanufactured. Useful life values are 
unique for each category of engines. As proposed, for purposes of 
certification, manufacturers would be required to use test data to 
estimate the rate of deterioration for each emission family over its 
useful life. Manufacturers would show that each emission family meets 
the emission standards after incorporating the estimated deterioration 
in emission control.
    The emission-data engine is the engine from an emission family that 
will be used for certification testing. To ensure that all engines in 
the family meet the standards, we are proposing that manufacturers 
select for certification testing the engine from the family that is 
most likely to exceed emission standards. In selecting this ``worst-
case'' engine, the manufacturer uses good engineering judgment. 
Manufacturers would consider, for example, all engine configurations 
and power ratings within the emission family and the range of allowed 
options. Requiring the worst-case engine to be tested ensures that all 
engines within the emission family are complying with emission 
standards. A similar approach would be used for evaporative emission 
control systems in emission families.
    We are proposing to require manufacturers to include in their 
application for certification the results of all emission tests from 
their emission-data units (engines, fuel tanks, etc.), including any 
diagnostic-type measurements (such as ppm testing) and invalidated 
tests. This complete set of test data ensures that the valid tests 
forming the basis of the manufacturer's application are a robust 
indicator of emission control performance rather than a spurious or 
incidental test result.
    Clean Air Act section 206(h) specifies that test procedures for 
certification (including the test fuel) should adequately represent in-
use operation. We are proposing test fuel specifications intended to 
represent in-use fuels. Engines would have to meet the standards on 
fuels with properties anywhere in the range of proposed test fuel 
specifications. The test fuel is generally to be used for all testing 
associated with the regulations proposed in this document, including 
certification, production-line testing, and in-use testing.
    We are proposing to require that engine manufacturers give engine 
operators instructions for properly maintaining their engines. We are 
including limitations on the frequency of scheduled maintenance that a 
manufacturer may specify for emission-related components to help ensure 
that emission control systems do not depend on an unreasonable 
expectation of maintenance in the field. These maintenance limits would 
also apply during any service accumulation that a manufacturer may do 
to establish deterioration factors. This approach is common to all our 
engine programs. We are proposing new regulatory language to clarify 
that engine manufacturers may perform emission-related maintenance 
during service accumulation only to the extent that they can 
demonstrate that such maintenance will be done with in-use engines. It 
is important to note, however, that these provisions would not limit 
the maintenance an operator could perform. It would merely limit the 
maintenance that operators would be expected to perform on a regularly 
scheduled basis. Some of these requirements are new for engines that 
are already subject to standards. We believe it is important to define 
limits to these maintenance parameters, especially with the expectation 
that engines will begin to incorporate aftertreatment technologies. See 
Sec.  1045.125 and Sec.  1054.125 of the proposed regulations for more 
information.
(2) What emission labels are required?
    Once an emission family is certified every product a manufacturer 
produces from that emission family would need an emission label with 
basic identifying information. We request comment on the proposed 
requirements for the design and content of engine labels, which are 
detailed in Sec.  1045.135 and Sec.  1054.135 of the proposed 
regulation text.
    The current regulations require equipment manufacturers to put a 
duplicate label on the equipment if the

[[Page 28197]]

engine is installed in a way that obscures the label on the engine. We 
are proposing to clarify this requirement for duplicate labels to 
ensure that labels are accessible without creating a supply of 
duplicate labels that are not authentic or are not used appropriately. 
Specifically, we are proposing to require engine manufacturers to 
supply duplicate labels to equipment manufacturers that request them 
and keep records to show how many labels they supply. Similarly, we are 
proposing that equipment manufacturers must request from engine 
manufacturers a specific number of duplicate labels, with a description 
of which engine and equipment models are involved and why the duplicate 
labels are necessary. Equipment manufacturers would need to destroy any 
excess labels and keep records to show the disposition of all the 
labels they receive. This would make it easier for us to verify that 
engines are meeting requirements and it would be easier for U.S. 
Customs to clear imported equipment with certified engines.
(3) What requirements apply to auxiliary emission control devices?
    Clean Air Act section 203(a) and existing regulations prohibit the 
use of a defeat device (see 40 CFR 90.111 and 91.111). The defeat 
device prohibition is intended to ensure that engine manufacturers do 
not use auxiliary emission control devices (AECD) in a regulatory test 
procedure that reduce the effectiveness of the emission control system 
during operation that is not substantially included in the regulatory 
test procedure.\94\ We are proposing to require manufacturers to 
describe their AECDs and explain why these are not defeat devices.
---------------------------------------------------------------------------

    \94\ Auxiliary emission control device is defined at 40 CFR 90.2 
and 91.2 as `` any element of design that senses temperature, 
vehicle speed, engine RPM, transmission gear, or any other parameter 
for the purpose of activating, modulating, delaying or deactivating 
the operation of any part of the emission control system.''
---------------------------------------------------------------------------

    Under the current regulations, there has been limited use of AECDs. 
However, with the proposed new emission standards and the corresponding 
engine technologies, we expect manufacturers to increase their use of 
engine designs that rely on AECDs. Disclosure of the presence and 
purpose of an AECD is essential in allowing us to evaluate the AECD and 
determine whether it represents a defeat device.
(4) What warranty requirements apply to engines or other products that 
are subject to emission standards?
    Consistent with our current emission control programs, we are 
proposing that manufacturers provide a design and defect warranty 
covering emission-related components. If the manufacturer offers a 
longer mechanical warranty for the engine or any of its components 
without an additional charge, the proposed regulations would require 
that the emission-related warranty period must be at least as long as 
the commercial warranty for the engine or the applicable components. 
Extended warranties that are available for an extra price would not 
trigger a need for a longer emission-related warranty. See the proposed 
regulation language for a description of which components are emission-
related.
    If an operator makes a valid warranty claim for an emission-related 
component during the warranty period, the engine manufacturer is 
generally obligated to replace the component at no charge to the 
operator. The engine manufacturer may deny warranty claims if the 
operator failed to do prescribed maintenance that contributed to the 
warranty claim.
    We are also proposing a defect reporting requirement that applies 
separately from the emission-related warranty (see Section VIII.F). In 
general, defect reporting applies when a manufacturer discovers a 
pattern of component failures whether that information comes from 
warranty claims, voluntary investigation of product quality, or other 
sources.
(5) Can I meet standards with emission credits?
    We are proposing a new emission-credit program for sterndrive and 
inboard marine engines and for evaporative emissions. We are also 
proposing to revise the existing emission-credit provisions for 
outboard and personal-watercraft engines and for Small SI engines. An 
emission-credit program is an important factor we take into 
consideration in setting emission standards that are appropriate under 
Clean Air Act section 213. An emission-credit program can reduce the 
cost and improve the technological feasibility of achieving standards, 
helping to ensure the standards achieve the greatest achievable 
reductions, considering cost and other relevant factors, in a time 
frame that is earlier than might otherwise be possible. Manufacturers 
gain flexibility in product planning and the opportunity for a more 
cost-effective introduction of product lines meeting a new standard. 
Emission-credit programs also create an incentive for the early 
introduction of new technology, which allows certain emission families 
to act as trailblazers for new technology. This can help provide 
valuable information to manufacturers on the technology before they 
apply the technology throughout their product line. This early 
introduction of clean technology improves the feasibility of achieving 
the standards and can provide valuable information for use in other 
regulatory programs that may benefit from similar technologies.
    Emission-credit programs generally involve averaging, banking, or 
trading. Averaging would allow a manufacturer to certify one or more 
emission families at emission levels above the applicable emission 
standards as long as the increased emissions are offset by one or more 
emission families certified below the applicable standards. The over-
complying families generate credits that are used by the under-
complying families. Compliance is determined on a total mass emissions 
basis to account for differences in production volume, power, and 
useful life among emission families. The average of all emissions for a 
particular manufacturer's production must be at or below the level of 
the applicable emission standards. This calculation generally factors 
in sales-weighted average power, production volume, useful life, and 
load factor. Banking and trading would allow a manufacturer to generate 
emission credits and bank them for future use in its own averaging 
program in later years or sell them to another company.
    A manufacturer choosing to participate in an emission-credit 
program would certify each participating emission family to a Family 
Emission Limit (FEL). In its certification application, a manufacturer 
would determine a separate FEL for each pollutant included in the 
emission-credit program. The FEL selected by the manufacturer becomes 
the emission standard for that emission family. Emission credits are 
based on the difference between the emission standard that applies and 
the FEL. The engines have to meet the FEL for all emission testing. At 
the end of the model year, manufacturers would generally need to show 
that the net effect of all their emission families participating in the 
emission-credit program is a zero balance or a net positive balance of 
credits. A manufacturer could generally choose to include only a single 
pollutant from an emission family in the emission-credit program or, 
alternatively, to establish an FEL for each of the regulated 
pollutants.

[[Page 28198]]

    Refer to the program discussions in Sections III through VI for 
more information about emission-credit provisions for individual engine 
or equipment categories. We request comment on all aspects of the 
emission-credit programs discussed in this proposal. In particular, we 
request comment on the structure of the proposed emission-credit 
programs and how the various provisions may affect manufacturers' 
ability to utilize averaging, banking, or trading to achieve the 
desired emission-reductions in the most efficient and economical way.
(6) How does EPA define maximum engine power?
    Maximum engine power is used to calculate the value of emission 
credits. For Small SI engines, it is also used to determine whether the 
standards apply; for example engines above 1000 cc are subject to Small 
SI standards only if maximum engine power is at or below 19 kW. For 
Marine SI engines, maximum engine power is also used to determine the 
emission standard that applies to a particular engine and to calculate 
emission credits. The regulations give no specific direction for 
defining maximum power for determining whether part 90 applies. Marine 
SI engine manufactures declare a rated power based on a procedure 
specified in a voluntary consensus standard, while credit calculations 
are based on sales-weighted average power for an engine family. We are 
concerned that these terms and specifications are not objective enough 
to ensure consistent application of regulatory requirements to all 
manufacturers. To the extent that manufacturers can determine different 
values of rated power or maximum engine power, they could be subject to 
different emission standards and calculate emission credits differently 
for otherwise identical engines. We believe it is important that a 
single power value be determined objectively according to a specific 
regulatory definition. Note that maximum engine power is not used 
during engine testing.
    We are proposing to standardize the determination of maximum engine 
power by relying primarily on the manufacturer's design specifications 
and the maximum torque curve that the manufacturer expects will 
represent the actual production engines. Under this approach the 
manufacturer would take the torque curve that is projected for an 
engine configuration, based on the manufacturer's design and production 
specifications, and convert it into a ``nominal power curve'' that 
would relate the maximum expected power to engine speed when a 
production engine is mapped according to our specified mapping 
procedures. The maximum engine power is the maximum power point on that 
nominal power curve. This has become the standard approach for all our 
emission control programs.
    Manufacturers would report the maximum engine power of each 
configuration in the application for certification. As with other 
engine parameters, manufacturers would ensure that the engines they 
produce under the certificate have maximum engine power consistent with 
those described in their applications. However, since we recognize that 
variability is a normal part of engine production, we allow a tolerance 
around the nominal value. We would instead require only that the power 
specified in the application be within the normal power range for 
production engines (see Sec.  1045.140 and Sec.  1054.140). We would 
typically expect the specified power to be within one standard 
deviation of the mean power of the production engines. If a 
manufacturer determines that the specified power is outside of the 
normal range for production engines, we may require the manufacturer to 
amend the application for certification. Manufacturer could 
alternatively change their engines to conform to the parameters 
detailed in the application for certification. In deciding whether to 
require a change to the application for certification, we would 
consider the degree to which the specified power differed from that of 
the production engines, the normal power variability for those engines, 
whether the engine used or generated emission credits, and whether the 
error affects which standards apply to the engine.
(7) What are the proposed production-line testing requirements?
    We are proposing to modify production-line testing requirements for 
engines already subject to exhaust emission standards and to extend 
these requirements to sterndrive and inboard marine engines. According 
to these requirements, manufacturers would routinely test production-
line engines to help ensure that newly assembled engines control 
emissions at least as well as the emission-data engines tested for 
certification. Production-line testing serves as a quality-control 
step, providing information to allow early detection of any problems 
with the design or assembly of freshly manufactured engines. This is 
different than selective enforcement auditing where we would give a 
test order for more rigorous testing for production-line engines in a 
particular emission family (see Section VIII.E).
    If an engine fails to meet an emission standard, the manufacturer 
must modify it to bring that specific engine into compliance. If too 
many engines exceed emission standards, the manufacturer will need to 
correct the problem for the engine family. This correction may involve 
changes to assembly procedures or engine design, but the manufacturer 
must, in any case, do sufficient testing to show that the emission 
family complies with emission standards.
    The proposed production-line testing programs would depend on the 
Cumulative Sum (CumSum) statistical process for determining the number 
of engines a manufacturer needs to test. We have used CumSum procedures 
for production-line testing with several other engine categories. Each 
manufacturer selects engines randomly at the beginning of a new 
sampling period. If engines must be tested at a facility where final 
assembly is not yet completed, manufacturers must randomly select 
engine components and assemble the test engine according to their 
established assembly instructions. The sampling period is a calendar 
quarter for engine families over 1,600 units. The minimum testing rate 
for these families is five engines per year. For engine families with 
projected sales at or below 1,600 units, the sampling period is a 
calendar year and the minimum testing rate is two engines. We may waive 
testing requirements for Marine SI engine families with projected sales 
below 150 units per year and for Small SI engine families with 
projected sales below 5,000 units per year. The CumSum program uses the 
emission results to calculate the number of tests required for the 
remainder of the sampling period to reach a pass or fail determination. 
If tested engines have relatively high emissions, the statistical 
sampling method calls for an increased number of tests to show that the 
emission family meets emission standards. The remaining number of tests 
is recalculated after the manufacturer tests each engine. Engines 
selected should cover the broadest range of production configurations 
possible. Tests should also be distributed evenly throughout the 
sampling period to the extent possible.
    Under the CumSum approach, a limited number of individual engines 
can exceed the emission standards before the Action Limit is met and 
the engine family itself fails under the production-line testing 
program. If an engine family fails, we may suspend the certificate. The 
manufacturer would then need to take steps to address the

[[Page 28199]]

nonconformity, which may involve amending the application for 
certification. This could involve corrected production procedures, a 
modified engine design. This may also involve changing the Family 
Emission Limit if there is no defect and the original Family Emission 
Limit was established using good engineering judgment. Note, however, 
that we propose to require manufacturers to adjust or repair every 
failing engine and retest it to show that it meets the emission 
standards. Note also that all production-line emission measurements 
must be included in the periodic reports to us. This includes any type 
of screening or surveillance tests (including ppm measurements), all 
data points for evaluating whether an engine controls emissions ``off-
cycle,'' and any engine tests that exceed the minimum required level of 
testing.
    While the proposed requirements may involve somewhat more testing 
than is currently required under 40 CFR part 90 or 91, there are 
several factors that limit the additional burden. First, the testing 
regulations in 40 CFR part 1065 specify that manufacturers may use 
field-testing equipment and procedures to measure emissions from 
production-line engines. This may substantially reduce the cost of 
testing individual engines by allowing much lower-cost equipment for 
measuring engines following assembly.
    Second, we are proposing to reduce the testing requirements for 
emission families that consistently meet emission standards. The 
manufacturer may request a reduced testing rate for emission families 
with no production-line tests exceeding emission standards for two 
consecutive years. The minimum testing rate is one test per emission 
family for one year. Our approval for a reduced testing rate would 
apply for a single model year.
    Third, as we have concluded in other engine programs, some 
manufacturers may have unique circumstances that call for different 
methods to show that production engines comply with emission standards. 
We therefore propose to allow a manufacturer to suggest an alternate 
plan for testing production-line engines as long as the alternate 
program is as effective at ensuring that the engines will comply. A 
manufacturer's petition to use an alternate plan should address the 
need for the alternative and should justify any changes from the 
regular testing program. The petition must also describe in detail the 
equivalent thresholds and failure rates for the alternate plan. If we 
approve the plan, we would use these criteria to determine when an 
emission family would become noncompliant. It is important to note that 
this allowance is intended only to provide flexibility and is not 
intended to affect the stringency of the standards or the production-
line testing program.
    Refer to the specific program discussions in Sections III, IV, and 
V for additional information about production-line testing for 
different types of engines.

D. Other Concepts

(1) What are the proposed emission-related installation instructions?
    For manufacturers selling loose engines to equipment manufacturers, 
we are proposing to require that the engine manufacturer develop a set 
of emission-related installation instructions. This would include 
anything that the installer would need to know to ensure that the 
engine operates within its certified design configuration. For example, 
the installation instructions could specify a total capacity needed 
from the engine cooling system, placement of catalysts after final 
assembly, or specification of parts needed to control evaporative 
emissions. If equipment manufacturers fail to follow the established 
emission-related installation instructions, we would consider this 
tampering, which could subject them to significant civil penalties. 
Refer to the proposed regulations for more information about specific 
provisions related to installation instructions (see Sec.  1045.130 and 
Sec.  1054.130).
(2) What is an agent for service?
    We are proposing to require that manufacturers identify an agent 
for service in the United States in their application for 
certification. The named person should generally be available within a 
reasonable time to respond to our attempts to make contact, either by 
telephone, e-mail, or in person. The person should also be capable of 
communicating about matters related to emission program requirements in 
English. (See Sec.  1045.205 and Sec.  1054.205).
(3) Are there special provisions for small manufacturers of these 
engines, equipment, and vessels?
    The scope of this proposal includes many engine, equipment, and 
vessel manufacturers that have not been subject to our regulations or 
certification process. Many of these manufacturers are small 
businesses. The sections describing the proposed emission control 
program include discussion of proposed special compliance provisions 
designed to address small business issues for the different types of 
engines and other products covered by the rule. Section XIV.B gives an 
overview of the inter-agency process in which we developed these small-
volume provisions.

VIII. General Nonroad Compliance Provisions

    This section describes a wide range of compliance provisions that 
apply generally to all the engines and equipment that would be subject 
to the proposed standards. Several of these provisions apply not only 
to engine manufacturers but also to equipment manufacturers installing 
certified engines, remanufacturing facilities, operators, and others.
    For standards that apply to equipment or fuel-system components, 
the provisions generally applicable to engine manufacturers would also 
apply to the equipment or component manufacturers. While this preamble 
section is written as if it would apply to engine exhaust standards, 
the same provisions would apply for equipment or component evaporative 
standards. We are proposing extensive revisions to the regulations to 
more carefully make these distinctions.
    As described in Section VII, we are proposing to migrate these 
general compliance provisions from 40 CFR parts 90 and 91 to the 
established regulatory text in 40 CFR part 1068. The provisions in part 
1068 already apply to other engine categories and we believe they can 
be applied to Small SI engines and Marine SI engines with minimal 
modification. Note that Section XI.C describes a variety of proposed 
changes and updates to the regulatory provisions in part 1068. We 
request comment on all aspects of part 1068 for these engines. The 
following discussion follows the sequence of the existing regulatory 
text in part 1068.\95\
---------------------------------------------------------------------------

    \95\ The regulatory text in the proposal does not republish the 
provisions of part 1068 that we are not proposing to change. For the 
latest full-text version of part 1068, see http://www.gpoaccess.gov/ecfr/index.html. Note that part 1068 is in Title 40, Protection of 
Environment.
---------------------------------------------------------------------------

A. Miscellaneous Provisions (Part 1068, Subpart A)

    This regulation contains some general provisions, including general 
applicability and the definitions that apply to part 1068. Other 
provisions concern good engineering judgment, how we would handle 
confidential information, how the EPA Administrator delegates decision-

[[Page 28200]]

making authority, and when we may inspect facilities, engines, or 
records.
    The process of testing engines and preparing an application for 
certification requires the manufacturer to make a variety of judgments. 
This includes, for example, selecting test engines, operating engines 
between tests, and developing deterioration factors. EPA has the 
authority to evaluate whether a manufacturer's use of engineering 
judgment is reasonable. The regulations describe the methodology we use 
to address any concerns related to a manufacturer's use of good 
engineering judgment in cases where the manufacturer has such 
discretion (see 40 CFR 1068.5). We will take into account the degree to 
which any error in judgment was deliberate or in bad faith. This 
subpart is consistent with provisions already adopted for light-duty 
highway vehicles and various other nonroad engines.

B. Prohibited Acts and Related Requirements (Part 1068, Subpart B)

    The proposed provisions in this subpart lay out a set of 
prohibitions for engine manufacturers, equipment manufacturers, 
operators, and engine rebuilders to ensure that engines comply with the 
emission standards. These provisions are summarized below but readers 
are encouraged to review the regulatory text. These provisions are 
intended to help ensure that each new engine sold or otherwise entered 
into commerce in the United States is certified to the relevant 
standards, that it remains in its certified configuration throughout 
its lifetime, and that only certified engines are used in the 
appropriate nonroad equipment.
(1) General Prohibitions (Sec.  1068.101)
    This proposed regulation contains several prohibitions consistent 
with the Clean Air Act. We generally prohibit selling a new engine in 
the United States without a valid certificate of conformity issued by 
EPA, deny us access to relevant records, or keep us from entering a 
facility to test or inspect engines. In addition, no one may 
manufacture any device that will make emission controls ineffective or 
remove or disable a device or design element that may affect an 
engine's emission levels, which we would consider tampering. We have 
generally applied the existing policies developed for tampering with 
highway engines and vehicles to nonroad engines.\96\ Other prohibitions 
reinforce manufacturers' obligations to meet various certification 
requirements. We also prohibit selling engine parts that prevent 
emission control systems from working properly. Finally, for engines 
that are excluded from regulation based on their use in certain 
applications, we generally prohibit using these engines in applications 
for which emission standards apply.
    Each prohibited act has a corresponding maximum penalty as 
specified in Clean Air Act section 205. As provided for in the Federal 
Civil Penalties Inflation Adjustment Act of 1990, Pub. L. 10-410, these 
maximum penalties are in 1970 dollars and should be periodically 
adjusted by regulation to account for inflation. The current penalty 
amount for most violations is $32,500.\97\
---------------------------------------------------------------------------

    \96\ ``Interim Tampering Enforcement Policy,'' EPA memorandum 
from Norman D. Shutler, Office of General Counsel, June 25, 1974 
(Docket A-2000-01; document II-B-20).
    \97\ EPA acted to adjust the maximum penalty amount in 1996 (61 
FR 69364, December 31, 1996). See also 40 CFR part 19.
---------------------------------------------------------------------------

(2) Equipment Manufacturer Provisions (Sec.  1068.105)
    The provisions of Sec.  1068.105 require equipment manufacturers to 
use certified engines in their new equipment once the emission 
standards begin to apply. We would allow a grace period for equipment 
manufacturers to deplete their supply of uncertified engines if they 
follow their normal inventory practices for buying engines, rather than 
stockpiling noncompliant (or previous-tier) engines to circumvent the 
new standards.
    We require equipment manufacturers to observe the engine 
manufacturers' emission-related installation instructions to ensure 
that the engines remain consistent with the application for 
certification. This may include such things as radiator specifications, 
diagnostic signals and interfaces, and placement of catalytic 
converters.
    If equipment manufacturers install a certified engine in a way that 
obscures the engine label, we propose to require that they add a 
duplicate label on the equipment. The equipment manufacturer would need 
to request from the engine manufacturer a specific number of duplicate 
labels, describe which engine and equipment models are involved, and 
explain why the duplicate labels are necessary. Equipment manufacturers 
would need to destroy any excess labels and keep records to show the 
disposition of all the labels they receive. This would make it easier 
for us to verify that engines are meeting requirements and it would be 
easier for U.S. Customs to clear imported equipment with certified 
engines.
    Equipment manufacturers not fulfilling the responsibilities we 
describe in this section would be in violation of one or more of the 
prohibited acts described above.
(3) In-Service Engines (Sec.  1068.110)
    The regulations generally prevent manufacturers from requiring 
owners to use any certain brand of aftermarket parts as well as give 
the manufacturers responsibility for engine servicing for emission-
related warranty issues, leaving the responsibility for all other 
maintenance with the owner. This proposed regulation would also reserve 
our right to do testing (or require testing), for example, to 
investigate potential defeat devices or in-use noncompliance, as 
authorized by the Clean Air Act.
(4) Engine Rebuilding (Sec.  1068.120)
    We are proposing to apply rebuild provisions for all the nonroad 
engines subject to the proposed emission standards. This approach is 
similar to what applies to heavy-duty highway engines and most other 
nonroad engines. This is necessary to prevent an engine rebuilder from 
rebuilding engines in a way that disables the engine's emission 
controls or compromises the effectiveness of the emission control 
system. We are proposing minimal recordkeeping requirements for 
businesses involved in commercial engine rebuilding to show that they 
comply with the regulations.
    In general, anyone who rebuilds a certified engine must restore it 
to its original (or a lower-emitting) configuration. Rebuilders must 
also replace some critical emission control components such as fuel 
injectors and oxygen sensors in all rebuilds for engines that use those 
technologies. Rebuilders must replace an existing catalyst if there is 
evidence that it is not functional; for example, if rattling pieces 
inside a catalyst show that it has lost its physical integrity, it 
would need to be replaced. See Sec.  1068.120 for more detailed 
information.
    These rebuilding provisions define good maintenance and rebuilding 
practices to help someone avoid violating the prohibition on ``removing 
or disabling'' emission control systems. These provisions therefore 
apply also to individuals who rebuild their own engines. However, we do 
not require such individuals to keep records to document compliance.
    We request comment on applying these proposed requirements for 
engine rebuilding and maintenance to the engines and vehicles subject 
to this rulemaking. In addition, we request

[[Page 28201]]

comment on the associated recordkeeping requirements.

C. Exemptions (Part 1068, Subpart C)

    We are proposing to apply several exemptions for certain specific 
situations, consistent with previous rulemakings. In general, exempted 
engines would need to comply with the requirements only in the sections 
related to the exemption. Note that additional restrictions could apply 
to importing exempted engines (see Section VIII.D). We may also require 
manufacturers (or importers) to add a permanent label describing that 
the engine is exempt from emission standards for a specific purpose. In 
addition to helping us enforce emission standards, this would help 
ensure that imported engines clear Customs without difficulty.
(1) Testing
    Anyone would be allowed to request an exemption for engines used 
only for research or other investigative purposes.
(2) Manufacturer-Owned Engines
    Engines that are used by engine manufacturers for development or 
marketing purposes could be exempted from regulation if they are 
maintained in the manufacturers' possession and are not used for any 
revenue-generating service. In contrast with the testing exemption, 
only certificate holders would be able to use this exemption.
(3) Display Engines
    Anyone may request an exemption for an engine if it is for display 
only.
(4) National Security
    Engine manufacturers could receive an exemption for engines they 
can show are needed by an agency of the federal government responsible 
for national defense. For cases where the engines will not be used on 
combat applications, the manufacturer would have to request the 
exemption with the endorsement of the procuring government agency.
(5) Exported Engines
    Engines that will be exported to countries that do not have the 
same emission standards as those that apply in the United States would 
be exempted without need for a request. This exemption would not be 
available if the destination country has the same emission standards as 
those in the United States.
(6) Competition Engines
    New engines that are used solely for competition are excluded from 
regulations applicable to nonroad engines. For purposes of our 
certification requirements, a manufacturer would receive an exemption 
if it can show that it produces the engine specifically for use solely 
in competition (see Sections III through V for specific provisions). In 
addition, engines that have been modified for use in competition would 
be exempt from the prohibition against tampering described above 
(without need for request). The literal meaning of the term ``used 
solely for competition'' would apply for these modifications. We would 
therefore not allow the engine to be used for anything other than 
competition once it has been modified. This also applies to someone who 
would later buy the engine, so we would require the person modifying 
the engine to remove or deface the original engine label and inform a 
subsequent buyer in writing of the conditions of the exemption.
(7) Replacement Engines
    An exemption would be available to engine manufacturers without 
request if that is the only way to replace an engine from the field 
that was produced before the current emission standards took effect. If 
less stringent standards applied to the old engine when it was new, the 
replacement engine would also have to meet those standards.
(8) Unusual Circumstance Hardship Provision
    Under the unusual circumstances hardship provision, any 
manufacturer subject to the proposed standards would be able to apply 
for hardship relief if circumstances outside their control cause the 
failure to comply and if failure to sell the subject engines or 
equipment or fuel system component would have a major impact on the 
company's solvency (see Sec.  1068.245). An example of an unusual 
circumstance outside a manufacturer's control may be an ``Act of God,'' 
a fire at the manufacturing plant, or the unforeseen shutdown of a 
supplier with no alternative available. The terms and time frame of the 
relief would depend on the specific circumstances of the company and 
the situation involved. As part of its application for hardship, a 
company would be required to provide a compliance plan detailing when 
and how it would achieve compliance with the standards. This hardship 
provision would be available to all manufacturers of engines, 
equipment, boats, and fuel system components subject to the proposed 
standards, regardless of business size.
(9) Economic Hardship Provision for Small Businesses
    An economic hardship provision would allow small businesses subject 
to the proposed standards to petition EPA for limited additional lead 
time to comply with the standards (see Sec.  1068.250). A small 
business would have to make the case that it has taken all possible 
business, technical, and economic steps to comply, but the burden of 
compliance costs would have a significant impact on the company's 
solvency. Hardship relief could include requirements for interim 
emission reductions and/or the purchase and use of emission credits. 
The length of the hardship relief decided during review of the hardship 
application would be up to one year, with the potential to extend the 
relief as needed. We anticipate that one to two years would normally be 
sufficient. As part of its application for hardship, a company would be 
required to provide a compliance plan detailing when and how it would 
achieve compliance with the standards. This hardship provision would be 
available only to small manufacturers of engines, equipment, boats, and 
fuel system components subject to the standards. For the purpose of 
determining which manufacturers qualify as a small business, EPA is 
proposing criteria based on either a production cut-off or the number 
of employees. The proposed criteria for determining which companies 
qualify as a small business are contained in Section III.F.2 for SD/I 
engines, Section IV.G for OB/PWC engines, Sections V.F.2 for 
nonhandheld engines, V.F.3 for nonhandheld equipment, and Section 
VI.G.2.f for handheld equipment, boats, and fuel system components.
(10) Hardship for Equipment Manufacturers, Vessel Manufacturers, and 
Secondary Engine Manufacturers
    Equipment manufacturers and boat builders in many cases will depend 
on engine manufacturers and fuel system component manufacturers to 
supply certified engines and fuel system components in time to produce 
complying equipment or boats by the date emission standards begin to 
apply. We are aware of other regulatory control programs where 
certified engines have been available too late for equipment 
manufacturers to adequately accommodate changing engine size or 
performance characteristics. To address this concern, we are proposing 
to allow Small SI equipment manufacturers and Marine SI boat builders 
to request up to one extra year before using certified engines or fuel 
system components if

[[Page 28202]]

they are unable to obtain certified product and they are not at fault 
and would face serious economic hardship without an extension. See 
Sec.  1068.255 for the proposed regulatory text related to this 
hardship.
    In addition, we are aware that some manufacturers of nonroad 
engines are dependent on another engine manufacturer to supply base 
engines that are then modified for the final application. Similar to 
equipment or vessel manufacturers, these ``secondary engine 
manufacturers'' may face difficulty in producing certified engines if 
the manufacturer selling the base engine makes an engine model 
unavailable with short notice. These secondary engine manufacturers 
generally each buy a relatively small number of engines and would 
therefore not necessarily be able to influence the marketing or sales 
practices of the engine manufacturer selling the base engine. As a 
result, we are proposing that secondary engine manufacturers could 
apply for this hardship as well. However, because these secondary 
engine manufacturers control the final design of their modified engine 
and could benefit in the market if they are allowed to produce a 
product certified to less stringent standards than their competitors, 
we would generally not approve an exemption unless the secondary engine 
manufacturer committed to a plan to make for any calculated loss in 
environmental benefit. Provisions similar to this hardship were already 
adopted for Large SI engines and recreational vehicles. See the 
existing regulatory text in Sec.  1068.255(c).
(11) Delegated Final Assembly
    The regulations in 40 CFR 1068.260 allow for flexible manufacturing 
for companies that produce engines that rely on aftertreatment. These 
regulations allow for equipment manufacturers to receive separate 
shipment of aftertreatment devices with the obligation resting on the 
equipment manufacturer to correctly install the aftertreatment on the 
engine when installing the engine in the equipment. Allowing for this 
practice requires an exemption from provisions which prohibit an engine 
from being introduced into commerce in its uncertified configuration. 
The provisions in Sec.  1068.260 to prevent improper use of this 
exemption include requirements to (1) Have contractual arrangements 
with equipment manufacturers; (2) submit affidavits to EPA regarding 
the use of the exemption; (3) include the price of the aftertreatment 
in the cost of the engine (to avoid giving equipment manufacturers an 
incentive to reduce costs inappropriately); and (4) periodically audit 
the affected equipment manufacturers.
    These provisions are not likely to be necessary for most Marine SI 
engine manufacturers. We do not expect outboard or personal watercraft 
engine manufacturers to use aftertreatment technology. For sterndrive/
inboard engines, we expect catalyst designs generally to be so integral 
to the exhaust manifold that engine manufacturers will include them 
with their engines. However, their may be some less common designs, 
such as engines on large vessels or airboats, where engine 
manufacturers may want to use the provisions allowing for separate 
shipment of aftertreatment. We are therefore proposing to adopt the 
provisions of Sec.  1068.260 without change for Marine SI engines.
    Manufacturers of handheld Small SI engines typically build both the 
engine and the equipment so we are proposing not to allow for delegated 
assembly with these engines.
    In contrast, nonhandheld engines (especially Class II) are built by 
engine manufacturers and sold to equipment manufacturers, often without 
complete fuel or exhaust systems. Ensuring that consumers get only 
engines that are in a certified configuration therefore requires a 
carefully crafted program. As described in Section V.E.2, we are 
proposing special provisions to accommodate the unique circumstances 
related to nonhandheld Small SI engines.
(12) Uncertified Engines Subject to Emission Standards
    In some cases we require manufacturers to meet certain emission 
standards without requiring certification, most commonly for 
replacement engines. In 40 CFR 1068.265 we spell out manufacturers 
obligations for these compliant but uncertified engines. Manufacturers 
must have test data showing that their engines meet the applicable 
emission standards and are liable for the emission performance of their 
engines, much like for certified engines, but are not required to 
submit an application for certification and get EPA approval before 
selling the engine. We propose to apply these provisions without 
modification for Small SI engines and Marine SI engines.

D. Imports (Part 1068, Subpart D)

    In general, the same certification requirements would apply to 
engines and equipment whether they are produced in the United States or 
are imported. The regulations in part 1068 also include some additional 
provisions that would apply if someone wants to import an exempted or 
excluded engine.
    All the proposed exemptions described above for new engines would 
also apply to importation, though some of these exemptions apply only 
on a temporary basis. An approved temporary exemption would be 
available only for a defined period. We could require the importer to 
post bond while the engine is in the United States. There are several 
additional proposed exemptions that would apply only to imported 
engines.
     Identical configuration: This is a permanent exemption to 
allow individuals to import engines that were designed and produced to 
meet applicable emission standards. These engines may be different than 
certified engines only in the fact that the emission label is missing 
because they were not intended for sale in the United States.
     Ancient engines: We would generally treat used engines as 
new if they are imported without a certificate of conformity. However, 
this permanent exemption would allow for importation of uncertified 
engines if they are more than 20 years old and remain in their original 
configuration.
     Repairs or alterations: This is a temporary exemption to 
allow companies to repair or modify engines. This exemption does not 
allow for operating the engine except as needed to do the intended 
work. This exemption would also apply for the practice for retiring 
bigger engines; noncompliant engines may be imported under this 
exemption for the purpose of recovering the engine block.
     Diplomatic or military: This is a temporary exemption to 
allow diplomatic or military personnel to use uncertified engines 
during their term of service in the U.S.
    We request comment on all these exemptions for domestically 
produced and imported engines and vehicles.

E. Selective Enforcement Audit (Part 1068, Subpart E)

    Clean Air Act section 206(b) gives us the discretion in any program 
with vehicle or engine emission standards to do selective enforcement 
auditing of production engines. We would do a selective enforcement 
audit by choosing an engine family and giving the manufacturer a test 
order that details a testing program to show that production-line 
engines meet emission standards. The regulation text describes the 
audit procedures in greater detail.

[[Page 28203]]

    We intend generally to rely on manufacturers' testing of 
production-line engines to show that they are consistently building 
products that conform to the standards. However, we reserve our right 
to do selective enforcement auditing if we have reason to question the 
emission testing conducted and reported by the manufacturer or for 
other reasons.

F. Defect Reporting and Recall (Part 1068, Subpart F)

    We are proposing to apply the defect reporting requirements of 
Sec.  1068.501 to replace the provisions of 40 CFR part 85 for nonroad 
engines. The requirements obligate manufacturers to tell us when they 
learn that emission control components or systems are defective and to 
conduct investigations under certain circumstances to determine if an 
emission-related defect is present. We are also proposing a requirement 
that manufacturers initiate these investigations when warranty claims 
and other available information indicate that a defect investigation 
may be fruitful. For this purpose, we consider defective any part or 
system that does not function as originally designed for the regulatory 
useful life of the engine or the scheduled replacement interval 
specified in the manufacturer's maintenance instructions.
    We believe the investigation requirement proposed in this rule will 
allow both EPA and the engine manufacturers to fully understand the 
significance of any unusually high rates of warranty claims that may 
have an impact on emissions. We believe prudent engine manufacturers 
already conduct a thorough investigation when available data indicate 
recurring parts failures as part of their normal practice to ensure 
product quality. Such data are valuable and readily available to most 
manufacturers and, under this proposal, must be considered to determine 
whether or not there is a possible defect of an emission-related part.
    Defect reports submitted in compliance with the current regulations 
are based on a single threshold applicable to engine families of all 
production volumes. No affirmative requirement for gathering 
information about the full extent of the problem applies. Many Small SI 
engine families have very high sales volumes. The proposed approach may 
therefore result in fewer total defect reports that should be submitted 
compared with the traditional approach because the number of defects 
triggering the submission requirement generally rises in proportion to 
the engine family size. Under the existing regulations, very small 
engine families would likely never report even a prominent defect 
because a relatively high proportion of such engines would have to be 
known to be defective before reporting is required under a scheme with 
fixed thresholds. The proposed threshold for reporting for the smallest 
engine families is therefore lower than under the current regulations.
    We are aware that accumulation of warranty claims will likely 
include many claims and parts that do not represent defects, so we are 
establishing a relatively high threshold for triggering the 
manufacturer's responsibility to investigate whether there is, in fact, 
a real occurrence of an emission-related defect.
    This proposal is intended to require manufacturers to use 
information we would expect them to keep in the normal course of 
business. We believe in most cases manufacturers would not be required 
to institute new programs or activities to monitor product quality or 
performance. A manufacturer that does not keep warranty information may 
ask for our approval to use an alternate defect-reporting methodology 
that is at least as effective in identifying and tracking potential 
emission-related defects as the proposed requirements. However, until 
we approve such a request, the proposed thresholds and procedures 
continue to apply.
    The proposed investigation thresholds are ten percent of total 
production to date up to a total production of 50,000 engines, but 
never fewer than 50 for any single engine family in one model year. For 
production between 50,000 and 550,000 units, the investigation 
threshold would increase at a marginal rate of four percent. For all 
production above 550,000 an investigation threshold of 25,000 engines 
would apply. For example, for an engine family with a sales volume of 
20,000 units in a given model year, the manufacturer would have to 
investigate potential emission-related defects after identifying 2,000 
possible defects. For an engine family with a sales volume of 450,000 
units in a given model year, the manufacturer would have to investigate 
potential emission-related defects after identifying 21,000 possible 
defects. These thresholds reflect the relevant characteristics of 
nonroad engines, such as the varying sales volumes, engine 
technologies, and warranty and maintenance practices.
    To carry out an investigation to determine if there is an emission-
related defect, manufacturers would have to use available information 
such as preexisting assessments of warranted parts. Manufacturers would 
also have to gather information by assessing previously unexamined 
parts submitted with warranty claims and replacement parts which are 
available or become available for examination and analysis. If 
available parts are deemed too voluminous to conduct a timely 
investigation, manufacturers would be permitted to employ appropriate 
statistical analyses of representative data to help draw timely 
conclusions regarding the existence of a defect. These investigative 
activities should be summarized in the periodic reports of recently 
opened or closed investigations, as discussed below. It is important to 
note that EPA does not regard having reached the investigation 
thresholds as conclusive proof of the existence of a defect, only that 
initiation of an appropriate investigation is merited to determine 
whether a defect exists.
    The second threshold in this proposal specifies when a manufacturer 
must report that an emission-related defect exists. This threshold 
involves a smaller number of engines because each potential defect has 
been screened to confirm that it is an emission-related defect. In 
counting engines to compare with the defect-reporting threshold, the 
manufacturer would consider a single engine family and model year. 
However, when a defect report is required, the manufacturer would 
report all occurrences of the same defect in all engine families and 
all model years that use the same part. The threshold for reporting a 
defect is two percent of total production for any single engine family 
for production up to 50,000 units, but never fewer than 20 for any 
single engine family in one model year. For production between 50,000 
and 550,000 units, the investigation threshold would increase at a 
marginal rate of one percent. For all production above 550,000 an 
investigation threshold of 6,000 engines would apply.
    It is important to note that while EPA regards occurrence of the 
defect threshold as proof of the existence of a reportable defect, it 
does not regard that occurrence as conclusive proof that recall or 
other action is merited.
    If the number of engines with a specific defect is found to be less 
than the threshold for submitting a defect report, but warranty claims 
or other information later indicate additional potentially defective 
engines, under this proposal the information must be aggregated for the 
purpose of determining whether the threshold for submitting a defect 
report has been met. If a manufacturer has knowledge from any source 
that the threshold for submitting a defect report has been met,

[[Page 28204]]

a defect report would have to be submitted even if the trigger for 
investigating has not yet been met. For example, if manufacturers 
receive information from their dealers, technical staff, or other field 
personnel showing conclusively that a recurring emission-related defect 
exists, they would have to submit a defect report if the submission 
threshold is reached.
    At specified times, the manufacturer would have to report open 
investigations as well as recently closed investigations that did not 
require a defect report. We are not proposing a fixed time limit for 
manufacturers to complete their investigations. However, the periodic 
reports required by the regulations will allow us to monitor these 
investigations and determine if it is necessary or appropriate for us 
to take further action.
    We request comment on all aspects of this approach to defect 
reporting. We also request comment on whether these reporting 
requirements should also apply to the current Phase 2 compliance 
program and if so, when these provisions should be applied.
    Under Clean Air Act section 207, if we determine that a substantial 
number of engines within an engine family, although properly used and 
maintained, do not conform to the appropriate emission standards, the 
manufacturer must remedy the problem and conduct a recall of the 
noncomplying engine family. However, we recognize that in some cases 
recalling noncomplying nonroad engines may not achieve sufficient 
environmental protection, so instead of making a determination of a 
substantial number of nonconforming engines (and thereby triggering a 
recall responsibility), we may allow manufacturers in some cases to 
nominate alternative remedial measures to address most potential 
noncompliance situations.

G. Hearings (Part 1068, Subpart G)

    According to this regulation, manufacturers would have the 
opportunity to challenge our decision to deny an application for 
certification or to suspend, revoke, or void an engine family's 
certificate. This also applies to our decision to reject the 
manufacturer's use of good engineering judgment (see Sec.  1068.5), and 
to our decisions related to emission-credit programs. Part 1068, 
subpart G, references the proposed procedures for a hearing to resolve 
such disputes.

IX. General Test Procedures

    The regulatory text in part 1065 is written with the intent to 
apply broadly to EPA engine programs. Part 1065 was originally adopted 
on November 8, 2002 (67 FR 68242) and currently applies for nonroad 
diesel engines, large nonroad spark-ignition engines and recreational 
vehicles under 40 CFR parts 1039, 1048 and 1051, respectively. The 
regulatory text was substantially revised in a recent rulemaking to 
make a variety of corrections and improvements (70 FR 40420, July 13, 
2005).
    This proposal applies to anyone who tests engines to show that they 
meet the emission standards for Small SI engines or Marine SI engines. 
This includes certification testing as well as all production-line and 
in-use testing. See the program descriptions above for testing 
provisions that are unique to each category of engines.
    We are proposing to apply the existing test provisions in part 1065 
for all Small SI engines and Marine SI engines. See Sections III 
through V for testing issues that are specific to the particular engine 
categories. In addition, we are proposing to allow manufacturers to use 
the provisions of part 1065 even before the proposed new standards take 
effect. This would allow manufacturers to migrate to the new test 
procedures sooner. This may involve upgrading to different types of 
analyzers that are specified in part 1065 but not in part 90 or part 
91. It may also involve recoding computers to do modal calculations 
specified in part 1065 instead of the weight-based calculations in part 
90 or part 91. At the same time, this would allow EPA to do 
confirmatory testing using the upgraded procedures without waiting for 
the proposed new standards to apply. This is important because EPA 
testing facilities are used for many different programs and the 
conversion to testing according to part 1065 specifications is well 
underway. We are aware that the new test specifications regarding 
engine mapping, generating duty cycles, and applying cycle-validation 
criteria would affect the emission measurements so we would follow the 
manufacturers' methods for these parameters in any case. For any other 
parameters, we would understand any differences between test procedures 
specified in parts 90, 91, and 1065 either to have no effect on 
emission measurements or to improve the accuracy of the measurement.
    We have identified various provisions in part 90 and part 91 that 
may need correction or adjustment. We request comment on the following 
possible changes:
     Changing the standard temperature condition for volume-
related calculations in Sec.  90.311(a)(2) and Sec.  91.311(a)(2) from 
25 [deg]C to 20 [deg]C. This would be consistent with EPA's test 
regulations, including the specifications in Sec.  1065.640.
     Removing the requirement to derive calibration and span 
gas concentrations from NIST Standard Reference Materials in Sec.  
90.312(c) and Sec.  91.312(c). This goes beyond the traceability 
requirements of other EPA test regulations and standard lab practices. 
We could instead refer to Sec.  1065.750 for calibration and span gas 
concentrations.
     Changing the direction for specifying gas concentrations 
in Sec.  90.312(c)(3) and Sec.  91.312(c)(3) from a volumetric basis to 
a molar basis.
     Correcting inconsistent requirements related to gas 
dividers. The regulations at Sec.  90.312(c)(4) and Sec.  91.312(c)(4) 
specify an accuracy of 2 percent, while Sec.  90.314(c) and 
Sec.  91.314(c) specify an accuracy of 1.5 percent. We 
could select one of these values, or we could refer to the gas divider 
specifications in Sec.  1065.248 and Sec.  1065.307.
     Correcting inconsistent specifications related to the 
timing of CO interference checks. The regulations at Sec.  90.317(b) 
and Sec.  91.317(b) specify that interference checks occur as part of 
annual maintenance, Sec.  90.325(a) and Sec.  91.325(a) specify that 
interference checks occur after any major repairs that could affect 
analyzer performance. We believe it would be most appropriate to make 
these consistent based on the specification in Sec.  1065.303, which 
calls for interference checks to occur after major maintenance.
    As we have done in previous programs, we are proposing specific 
test procedures to define how measurements are to be made but would 
allow the use of alternate procedures if they are shown to be 
equivalent to our specified procedures.\98\ The test procedures 
proposed in part 1065 are derived from our test procedures in 40 CFR 
part 86 for highway heavy-duty gasoline engines and light-duty 
vehicles. The procedures have been simplified (and to some extent 
generalized) to better fit nonroad engines. The procedures in part 1065 
currently apply to recreational vehicles and to nonroad spark-ignition 
engines above 19 kW. We request comment on all aspects of these 
proposed test procedures. We also request comment regarding whether any 
additional parts of the test procedures contained in 40 CFR part 86 
(for highway vehicles and engines), in other parts that apply to 
nonroad engines, or

[[Page 28205]]

in ISO 8178 should be incorporated into the final test procedures.
---------------------------------------------------------------------------

    \98\ Note that the published procedures still apply if we 
approve a manufacturer's use of an alternative procedure. EPA 
testing may be done using the published procedures or the alternate 
procedures approved for a given engine family.
---------------------------------------------------------------------------

A. Overview

    Part 1065 is organized by subparts as shown below:
     Subpart A: General provisions; global information on 
applicability, alternate procedures, units of measure, etc.
     Subpart B: Equipment specifications; required hardware for 
testing
     Subpart C: Measurement instruments
     Subpart D: Calibration and verifications; for measurement 
systems
     Subpart E: Engine selection, preparation, and maintenance
     Subpart F: Test protocols; step-by-step sequences for 
laboratory testing and test validation
     Subpart G: Calculations and required information
     Subpart H: Fuels, fluids, and analytical gases
     Subpart I: Oxygenated fuels; special test procedures
     Subpart J: Field testing and portable emissions 
measurement systems
     Subpart K: Definitions, references, and symbols
    The regulations prescribe scaled specifications for test equipment 
and measurement instruments by parameters such as engine power, engine 
speed and the emission standards to which an engine must comply. That 
way this single set of specifications will cover the full range of 
engine sizes and our full range of emission standards. Manufacturers 
will be able to use these specifications to determine what range of 
engines and emission standards may be tested using a given laboratory 
or field testing system.
    The content already adopted in part 1065 is mostly a combination of 
material from our most recent updates to other test procedures and from 
test procedures specified by the International Organization for 
Standardization (ISO). There are also some provisions we created 
specifically for part 1065, generally to address very recent advances 
such as measuring very low concentrations of emissions, using new 
measurement technology, using portable emissions measurement systems, 
and performing field testing.
    The content in part 1065 also reflects a shift in our approach for 
specifying measurement performance. In the past we specified numerous 
calibration accuracies for individual measurement instruments, and we 
specified some verifications for individual components such as 
NO2-to-NO converters. We have shifted our focus away from 
individual instruments and toward the overall performance of complete 
measurement systems. We did this for several reasons. First, some of 
what we specified in the past precluded the implementation of new 
measurement technologies. These new technologies, sometimes called 
``smart analyzers,'' combine signals from multiple instruments to 
compensate for interferences that were previously tolerable at higher 
emissions levels. These analyzers are useful for detecting low 
concentrations of emissions. They are also useful for detecting 
emissions from raw exhaust, which can contain high concentrations of 
interferences, such as water vapor. This is particularly important for 
field testing, which will most likely rely upon raw exhaust 
measurements. Second, this new ``systems approach'' requires periodic 
verifications for complete measurement systems, which we feel will 
provide a more robust assurance that a measurement system as a whole is 
operating properly. Third, the systems approach provides a direct 
pathway to demonstrate that a field test system performs similarly to a 
laboratory system. Finally, we feel that our systems approach will lead 
to a more efficient way of ensuring measurement performance in the 
laboratory and in the field. We believe this efficiency will stem from 
less frequent calibrations of individual instruments and higher 
confidence that a complete measurement system is operating properly.
    Below is a brief description of the content of each subpart. The 
discussion highlights some recent changes to part 1065. We are not 
proposing any changes to part 1065 as part of this proposal, but we 
intend to make various changes to part 1065 as part of a concurrent 
rulemaking to set new emission standards for marine diesel and 
locomotive engines. Manufacturers of engines that are the subject of 
this proposal are encouraged to stay abreast of testing changes that we 
propose in this other rulemaking.
(1) Subpart A General Provisions
    In Subpart A we identify the applicability of part 1065 and 
describe how procedures other than those in part 1065 may be used to 
comply with a standard-setting part. In Sec.  1065.10(c)(1) we specify 
that testing must be conducted in a way that represents in-use engine 
operation, such that in the rare case where provisions in part 1065 
result in unrepresentative testing, we may cooperate with manufacturers 
to work out alternative testing approaches for demonstrating compliance 
with emission standards. Another aspect of representative testing 
relates to the desire to maintain consistency between certification 
testing and in-use testing. If we or manufacturers test in-use engines, 
we would expect the engine to be removed from the equipment and 
installed on an engine dynamometer for testing with no changes to the 
engine (including the governor, fuel system, exhaust system and other 
components).
    In Sec.  1065.10(c)(7) and Sec.  1065.12 we describe a process by 
which we may approve alternative test procedures that we determine to 
be equivalent to (or more accurate than) the specified procedures. 
Given the new testing specifications in part 1065 and the standard-
setting parts, and this more detailed approach to approving alternative 
test procedures, we will not allow manufacturers to continue testing 
based on any earlier approvals for alternative testing under part 90 or 
part 91. Any manufacturer wishing to continue testing with any method, 
device, or specification that departs from that included in this 
proposal would need to request approval for such testing under Sec.  
1065.10(c)(7).
    Other information in this subpart includes a description of the 
conventions we use regarding units and certain measurements and we 
discuss recordkeeping. We also provide an overview of how emissions and 
other information are used for determining final emission results. The 
regulations in Sec.  1065.15 include a figure illustrating the 
different ways we allow brake-specific emissions to be calculated.
    In this same subpart, we describe how continuous and batch sampling 
may be used to determine total emissions. We also describe the two ways 
of determining total work that we approve. Note that the figure 
indicates our default procedures and those procedures that require 
additional approval before we will allow them.
(2) Subpart B Equipment Specifications
    Subpart B first describes engine and dynamometer related systems. 
Many of these specifications are scaled to an engine's size, speed, 
torque, exhaust flow rate, etc. We specify the use of in-use engine 
subsystems such as air intake systems wherever possible to best 
represent in-use operation when an engine is tested in a laboratory.
    Subpart B also describes sampling dilution systems. These include 
specifications for the allowable components, materials, pressures, and 
temperatures. We describe how to sample crankcase emissions.
    The regulations in Sec.  1065.101 include a diagram illustrating 
all the available equipment for measuring emissions.

[[Page 28206]]

(3) Subpart C Measurement Instruments
    Subpart C specifies the requirements for the measurement 
instruments used for testing. These specifications apply to both 
laboratory and field testing. In subpart C we recommend accuracy, 
repeatability, noise, and response time specifications for individual 
measurement instruments, but note that we require that overall 
measurement systems meet the calibrations and verifications in Subpart 
D.
    In some cases we allow new instrument types to be used where we 
previously did not allow them. For example, we now allow the use of a 
nonmethane cutter for NMHC measurement, a nondispersive ultraviolet 
analyzers for NOX measurement, zirconia sensors for 
O2 measurement, various raw-exhaust flow meters for 
laboratory and field testing measurement, and an ultrasonic flow meter 
for CVS systems.
(4) Subpart D Calibrations and Verifications
Subpart D describes what we mean when we specify accuracy, 
repeatability and other parameters in Subpart C. These specifications 
apply to both laboratory and field testing. We are adopting 
calibrations and verifications that scale with engine size and with the 
emission standards to which an engine is certified. We are replacing 
some of what we have called ``calibrations'' in the past with a series 
of verifications, such as a linearity verification, which essentially 
verifies the calibration of an instrument without specifying how the 
instrument must be initially calibrated. Because new instruments have 
built-in routines that linearize signals and compensate for various 
interferences, our existing calibration specifications sometimes 
conflicted with an instrument manufacturer's instructions. In addition, 
there are new verifications in subpart D to ensure that the new 
instruments we specify in Subpart C are used correctly.
(5) Subpart E Engine Selection, Preparation, and Maintenance
    Subpart E describes how to select, prepare, and maintain a test 
engine. We updated these provisions to include both gasoline and diesel 
engines.
(6) Subpart F Test Protocols
    Subpart F describes the step-by-step protocols for engine mapping, 
test cycle generation, test cycle validation, pre-test preconditioning, 
engine starting, emission sampling, and post-test validations. We 
adopted an improved way to map and generate cycles for constant-speed 
engines that would better represent in-use engine operation. We adopted 
a more streamlined set of test cycle and validation criteria. We allow 
modest corrections for drift of emission analyzer signals within a 
certain range.
(7) Subpart G Calculations and Required Information
    Subpart G includes all the calculations required in part 1065. We 
adopted definitions of statistical quantities such as mean, standard 
deviation, slope, intercept, t-test, F-test, etc. By defining these 
quantities mathematically we intend to resolve any potential ambiguity 
when we discuss these quantities in other subparts. We have written all 
calculations for calibrations and emission calculations in 
international units to comply with 15 CFR part 1170, which removes the 
voluntary aspect of the conversion to international units for federal 
agencies. Furthermore, Executive Order 12770 (56 FR 35801, July 29, 
1991) reinforces this policy by providing Presidential authority and 
direction for the use of the metric system of measurement by Federal 
agencies and departments. For our standards that are not completely in 
international units (i.e., grams/horsepower-hour, grams/mile), we 
specify in part 1065 the correct use of internationally recognized 
conversion factors.
    We also specify emission calculations based on molar quantities for 
flow rates instead of volume or mass. This change eliminates the 
frequent confusion caused by using different reference points for 
standard pressure and standard temperature. Instead of declaring 
standard densities at standard pressure and standard temperature to 
convert volumetric concentration measurements to mass-based units, we 
declare molar masses for individual elements and compounds. Since these 
values are independent of all other parameters, they are known to be 
universally constant.
(8) Subpart H Fuels, Fluids, and Analytical Gases
    Subpart H specifies test fuels, lubricating oils and coolants, and 
analytical gases for testing. We are not identifying any detailed 
specification for service accumulation fuel. Instead, we specify that 
service accumulation fuel must be either a test fuel or a commercially 
available in-use fuel. This helps ensure that testing is representative 
of in-use engine operation. We are adding a list of ASTM specifications 
for in-use fuels as examples of appropriate service accumulation fuels. 
Compared to the proposed regulatory language, we have clarified that 
Sec.  1065.10(c)(1) does not require test fuels to be more 
representative than the specified test fuels. We have added an 
allowance to use similar test fuels that do not meet all of the 
specifications provided they do not compromise the manufacturer's 
ability to demonstrate compliance. We also now allow the use of ASTM 
test methods specified in 40 CFR part 80 in lieu of those specified in 
part 1065. We did this because we may more frequently review and update 
the ASTM methods in part 80 versus those in part 1065.
    Proper testing requires the use of good engineering judgment to 
maintain the stability of analytical gases.
(9) Subpart I Oxygenated Fuels
    Subpart I describes special procedures for measuring certain 
hydrocarbons whenever oxygenated fuels are used. We updated the 
calculations for these procedures in Subpart G. We have made some 
revisions to the proposed text to make it consistent with the original 
content of the comparable provisions in part 86. We have also added an 
allowance to use the California NMOG test procedures to measure 
alcohols and carbonyls.
(10) Subpart J Field Testing and Portable Emissions Measurement Systems
    Portable Emissions Measurement Systems (PEMS) for field testing for 
marine spark-ignition engines must generally meet the same 
specifications and verifications that laboratory instruments must meet 
according to subparts B, C, and D. However, we allow some deviations 
from laboratory specifications. In addition to meeting many of the 
laboratory system requirements, a PEMS must meet an overall 
verification relative to laboratory measurements. This verification 
involves repeating a duty cycle several times. The duty cycle itself 
must have several individual field-test intervals (e.g., NTE events) 
against which a PEMS is compared to the laboratory system. This is a 
comprehensive verification of a PEMS. We also adopted a procedure for 
preparing and conducting a field test and adopted drift corrections for 
emission analyzers. Given the evolving state of PEMS technology, the 
field-testing procedures provide for a number of known measurement 
techniques. We have added provisions and conditions for using PEMS in 
an engine dynamometer laboratory to conduct laboratory testing.

[[Page 28207]]

(11) Subpart K Definitions, References, and Symbols
    Subpart K includes all the defined terms, identification of 
reference materials, and lists of acronyms and abbreviations used 
throughout part 1065.

B. Special Provisions for Nonroad Spark-Ignition Engines

    While part 1065 defines a wide range of specifications to define 
appropriate test procedures, several parameters are unique to each 
program. For example, each category of engines has one or more duty 
cycles that describe exactly how to operate each engine during the 
test. These category-specific provisions are described in part 1045, 
subpart F, for Marine SI engines and in part 1054, subpart F, for Small 
SI engines.
    Manufacturers may run the specified steady-state duty cycle either 
as a series of discrete modes or as a ramped-modal cycle. The ramped-
modal cycle specifies the same engine speeds and loads as in 
conventional discrete-mode testing, but the modes are connected by 
gradual ramps in engine speed and torque for a single, continuous 
emission-sampling period. The different modes are connected with 
twenty-second linear speed and torque transitions during which 
emissions are measured. Emission sampling therefore starts at the 
beginning of a ramped-modal cycle and does not stop until its last mode 
is completed.
    Ramped-modal cycles involve a different sequence of modes than is 
specified for discrete-mode testing. For example, the first mode, which 
is engine idle, is split so that half the idle mode occurs at the 
beginning of the test and half occurs at the end of the test. This 
helps facilitate certain technical aspects of emission sampling. 
Instead of using weighting factors for each steady-state mode, a 
ramped-modal cycle specifies different time durations for each mode. 
Time durations of the modes and transitions are proportioned to the 
established modal weighting factors for the specified cycle.
    There are several advantages to ramped-modal testing. Using 
discrete-mode testing, manufacturers sample emissions for an 
unspecified time duration near the end of each individual mode. The 
result is several separate measurements that must be combined 
mathematically to yield an overall emission result in g/kW-hr. The 
ramped-modal cycle has a single emission-sampling period. This 
decreases testing variability and reduces the overall cost of running 
tests. Ramped-modal testing also enables the use of batch sampling 
systems such as bag samplers.

X. Energy, Noise, and Safety

    Section 213 of the Clean Air Act directs us to consider the 
potential impacts on safety, noise, and energy when establishing the 
feasibility of emission standards for nonroad engines. Furthermore, 
section 205 of EPA's 2006 Appropriations Act requires us to assess 
potential safety issues, including the risk of fire and burn to 
consumers in use, associated with the proposed emission standards for 
nonroad spark-ignition engines below 50 horsepower.\99\ As further 
detailed in the following sections, we expect that the proposed exhaust 
and evaporative emission standards will either have no adverse affect 
on safety, noise, and energy or will improve certain aspects of these 
important characteristics. A more in depth discussion of these topics 
relative to the proposed exhaust and evaporative emission standards is 
contained in Chapters 4 and 5 of the Draft RIA, respectively. Also, our 
conclusions relative to safety are fully documented in our 
comprehensive safety study which is discussed in the next section.
---------------------------------------------------------------------------

    \99\ Department of the Interior, Environment, and Related 
Agencies Appropriations Act, 2006, Pub. L. 109-54, Title II, sec. 
205, 119 Stat. 499, 532 (August 2, 2005).
---------------------------------------------------------------------------

A. Safety

    We conducted a comprehensive, multi-year safety study of spark-
ignition engines that focused on the four areas where we are proposing 
new emission standards.\100\ These areas are:
---------------------------------------------------------------------------

    \100\ ``EPA Technical Study on the Safety of Emission Controls 
for Nonroad Spark-Ignition Engines < 50 Horsepower,'' Office of 
Transportation and Air Quality, U.S. Environmental Protection 
Agency, Washington, DC, EPA420-R-06-006, March 2006. This document 
is available in Docket EPA-HQ-OAR-2004-0008. This report was also 
subject to peer review, as described in a peer review report that is 
also available in the docket.
---------------------------------------------------------------------------

     New catalyst-based HC+NOX exhaust emission 
standards for Class I and Class II nonhandheld spark-ignition engines;
     New fuel evaporative emission standards for nonhandheld 
and handheld equipment;
     New HC+NOX exhaust emission standards for 
outboard and personal watercraft engines and vessels, and a new CO 
exhaust emission standard for nonhandheld engines used in marine 
auxiliary applications; and
     New fuel evaporative emission standards for outboard and 
personal watercraft engines and vessels.
    Each of these four areas is discussed in greater detail in the next 
sections.
(1) Exhaust Emission Standards for Small Spark-Ignition Engines
    The technology approaches that we assessed for achieving the 
proposed Small SI engine standards included exhaust catalyst 
aftertreatment and improvements to engine and fuel system designs. In 
addition to our own testing and development effort, we also met with 
engine and equipment manufacturers to better understand their designs 
and technology and to determine the state of technological progress 
beyond EPA's Phase 2 emission standards.
    The scope of our safety study included Class I and Class II engine 
systems that are used in residential walk-behind and ride-on lawn mower 
applications, respectively. Residential lawn mower equipment was chosen 
for the following reasons.
     Lawn mowers and the closely-related category of lawn 
tractors overwhelmingly represent the largest categories of equipment 
using Class I and Class II engines.
     Consumer Product Safety Commission (CPSC) data indicate 
that more thermal burn injuries are associated with lawn mowers than 
occur with other nonhandheld equipment; lawn mowers therefore represent 
the largest thermal burn risk for these classes of engines.
     General findings regarding advanced emission control 
technologies for residential lawn and garden equipment carry over to 
commercial lawn and turf care equipment as well as to other nonhandheld 
equipment using Class I and Class II engines.
    We conducted the technical study of the incremental risk on several 
fronts. First, working with CPSC, we evaluated their reports and 
databases and other outside sources to identify those in-use situations 
which create fire and burn risk for consumers. The outside sources 
included meetings, workshops, and discussions with engine and equipment 
manufacturers. From this information, we identified ten scenarios for 
evaluation that covered a comprehensive variety of in-use conditions or 
circumstances which potentially could lead to an increased risk in 
burns or fires.
    Second, we conducted extensive laboratory and field testing of both 
current technology (Phase 2) and prototype catalyst-equipped advanced-
technology engines and equipment (Phase 3) to assess the emission 
control performance and thermal characteristics of the engines and 
equipment. This testing included a comparison of exhaust system, 
engine, and equipment

[[Page 28208]]

surface temperatures using still and full motion video thermal imaging 
equipment.
    Third, we conducted a design and process Failure Mode and Effects 
Analyses (FMEA) comparing current Phase 2 and Phase 3 compliant engines 
and equipment to evaluate incremental changes in risk probability as a 
way of evaluating the incremental risk of upgrading Phase 2 engines to 
meet Phase 3 emission standards.\101\ This is an engineering analysis 
tool to help engineers and other professional staff to identify and 
manage risk. In an FMEA, potential failure modes, causes of failure, 
and failure effects are identified and a resulting risk probability is 
calculated from these results. This risk probability is used by the 
FMEA team to rank problems for potential action to reduce or eliminate 
the causal factors. Identifying these causal factors is important 
because they are the elements that a manufacturer can consider to 
reduce the adverse effects that might result from a particular failure 
mode.
---------------------------------------------------------------------------

    \101\ ``EPA Technical Study on the Safety of Emission Controls 
for Nonroad Spark-Ignition Engines < 50 Horsepower,'' Office of 
Transportation and Air Quality, U.S. Environmental Protection 
Agency, Washington, DC, EPA420-R-06-006, March 2006. This document 
is available in Docket EPA-HQ-OAR-2004-0008.
---------------------------------------------------------------------------

    Our technical work and subsequent analysis of all of the data and 
information strongly indicate that effective catalyst-based standards 
can be implemented without an incremental increase in the risk of fire 
or burn to the consumer either during or after using the equipment. 
Similarly, we did not find any increase in the risk of fire during 
refueling or in storage near typical combustible materials. For 
example, our testing program demonstrated that properly designed 
catalyst-mufflers could, in some cases, actually result in systems that 
were significantly cooler than many current original equipment 
mufflers. A number of design elements appear useful to properly 
managing heat loads including: (1) The use of catalyst designs that 
minimize CO oxidation through careful selection of catalyst size, 
washcoat composition, and precious metal loading; (2) positioning the 
catalyst within the cooling air flow of the engine fan or redirecting 
some cooling air over the catalyst area with a steel shroud; (3) 
redirecting exhaust flow through multiple chambers or baffles within 
the catalyst-muffler; and (4) larger catalyst-muffler volumes than the 
original equipment muffler.
(2) Fuel Evaporative Emission Standards for Nonhandheld and Handheld 
Engines and Equipment
    We reviewed the fuel line and fuel tank characteristics for 
nonhandheld and handheld equipment and evaluated control technology 
which could be used to reduce evaporative emissions from these two 
subcategories. The available technology is capable of achieving 
reductions in fuel tank and fuel line permeation without an adverse 
incremental impact on safety. For fuel lines and fuel tanks, the 
applicable consensus safety standards, manufacturer specific test 
procedures and EPA requirements are sufficient to ensure that there 
will be no increase in the types of fuel leaks that lead to fire and 
burn risk during in-use operation. Instead, these standards will reduce 
vapor emissions both during operation and in storage. That reduction, 
coupled with some expected equipment redesign, is expected to lead to 
reductions in the risk of fire or burn without affecting component 
durability.
    The Failure Mode and Effects Analyses, which was described in the 
previous section, also evaluated permeation and running loss controls 
on nonhandheld engines. We found that these controls would not increase 
the probability of fire and burn risk from those expected with current 
fuel systems, but could in fact lead to directionally improved systems 
from a safety perspective. Finally, the running loss control program 
being proposed for nonhandheld equipment will lead to changes that are 
expected to reduce risk of fire during in-use operation. Moving fuel 
tanks away from heat sources, improving cap designs to limit leakage on 
tip over, and requiring a tethered cap will all help to eliminate 
conditions which lead to in-use problems related to fuel leaks and 
spillage. Therefore, we believe the application of emission control 
technology to reduce evaporative emissions from these fuel lines and 
fuel tanks will not lead to an increase in incremental risk of fires or 
burns and in some cases is likely to at least directionally reduce such 
risks.
(3) Exhaust Emission Standards for Outboard and Personal Watercraft 
Marine Engines and Vessels and Marine Auxiliary Engines
    Our analysis of exhaust emission standards for OB/PWC engines and 
marine auxiliary engines found that the U. S. Coast Guard (USCG) has 
comprehensive safety standards that apply to engines and fuel systems 
used in these vessels. Additionally, organizations such as the Society 
of Automotive Engineers, Underwriters Laboratories, and the American 
Boat and Yacht Council (ABYC) also have safety standards that apply in 
this area. We also found that the four-stroke and two-stroke direct 
injection engine technologies which are likely to be used to meet the 
exhaust emission standards contemplated for OB/PWC engines are in 
widespread use in the vessel fleet today. These more sophisticated 
engine technologies are replacing the traditional two-stroke carbureted 
engines. The four-stroke and two-stroke direct injection engines meet 
applicable USCG and ABYC safety standards and future products will do 
so as well. The proposed emission standards must be complementary to 
existing safety standards and our analysis indicates that this will be 
the case. There are no known safety issues with the advanced 
technologies compared with two-stroke carbureted engines. The newer-
technology engines arguably provide safety benefits due to improved 
engine reliability and range in-use. Based on the applicability of USCG 
and ABYC safety standards and the good in-use experience with advanced-
technology engines in the current vessel fleet, we believe new emission 
standards would not create an incremental increase in the risk of fire 
or burn to the consumer.
(4) Fuel Evaporative Emission Standards for Outboard and Personal 
Watercraft Engines and Vessels
    We reviewed the fuel line and fuel tank characteristics for marine 
vessels and evaluated control technology which could be used to reduce 
evaporative emissions from boats. With regard to fuel lines, fuel 
tanks, and diurnal controls, there are rigorous USCG, ABYC, United 
Laboratories, and Society of Automotive Engineers standards which 
manufacturers will continue to meet for fuel system components. All of 
these standards are designed to address the in-use performance of fuel 
systems, with the goal of eliminating fuel leaks. The low-permeation 
fuel lines and tanks needed to meet the Phase 3 requirements would need 
to pass these standards and every indication is that they would 
pass.\102\
---------------------------------------------------------------------------

    \102\ ``EPA Technical Study on the Safety of Emission Controls 
for Nonroad Spark-Ignition Engines < 50 Horsepower,'' Office of 
Transportation and Air Quality, U.S. Environmental Protection 
Agency, Washington, DC, EPA420-R-06-006, March 2006. This document 
is available in Docket EPA-HQ-OAR-2004-0008.
---------------------------------------------------------------------------

    Furthermore, the EPA permeation certification requirements related 
to emissions durability will add an additional layer of assurance. Low-
permeation fuel lines are used safely

[[Page 28209]]

today in many marine vessels. Low-permeation fuel tanks and diurnal 
emission controls have been demonstrated in various applications for 
many years without an increase in safety risk. Furthermore, a properly 
designed fuel system with fuel tank and fuel line permeation controls 
and diurnal emission controls would reduce the fuel vapor in the boat, 
thereby reducing the opportunities for fuel related fires. In addition, 
using improved low-permeation materials coupled with designs meeting 
USCG and ABYC requirements should reduce the risk of fuel leaks into 
the vessel. We believe the application of emission control technologies 
on marine engines and vessels for meeting the proposed fuel evaporative 
emission standards would not lead to an increase in incremental risk of 
fires or burns, and in many cases may incrementally decrease safety 
risk in certain situations.

B. Noise

    As automotive technology demonstrates, achieving low emissions from 
spark-ignition engines can correspond with greatly reduced noise 
levels. Direct-injection two-stroke and four-stroke OB/PWC have been 
reported to be much quieter than traditional carbureted two-stroke 
engines. Catalysts in the exhaust act as mufflers which can reduce 
noise. Additionally, adding a properly designed catalyst to the 
existing muffler found on all Small SI engines can offer the 
opportunity to incrementally reduce noise.

C. Energy

(1) Exhaust Emission Standards
    Adopting new technologies for controlling fuel metering and air-
fuel mixing, particularly the conversion of some carbureted engines to 
advanced fuel injection technologies, will lead to improvements in fuel 
consumption. This is especially true for OB/PWC engines where we expect 
the proposed standards to result in the replacement of old technology 
carbureted two-stroke engines with more fuel-efficient technologies 
such as two-stroke direct injection or four-stroke engines. Carbureted 
crankcase-scavenged two-stroke engines are inefficient in that 25 
percent or more of the fuel entering the engine may leave the engine 
unburned. EPA estimates that conversion to more fuel efficient 
recreational marine engines would save 61 million gallons of gasoline 
per year in 2030. The conversion of some carbureted Small SI engines to 
fuel injection technologies is also expected to improve fuel economy. 
We estimate approximately 18 percent of the Class II engines will be 
converted to fuel injection and that this will result in a fuel savings 
of about 10 percent for each converted engine. This translates to a 
fuel savings of about 56 million gallons of gasoline in 2030 when all 
of the Class II engines used in the U.S. will comply with the proposed 
Phase 3 standards. By contrast, the use of catalyst-based control 
systems on Small SI engines is not expected to change their fuel 
consumption characteristics.
(2) Fuel Evaporative Emission Standards
    We anticipate that the proposed fuel evaporative emission standards 
will have a positive impact on energy. By capturing or preventing the 
loss of fuel due to evaporation, we estimate that the lifetime average 
fuel savings would be about 1.6 gallons for an average piece of Small 
SI equipment and 32 gallons for an average boat. This translates to a 
fuel savings of about 41 million gallons for Small SI equipment and 30 
million gallons for Marine SI vessels in 2030 when most of the affected 
equipment used in the U.S. would be expected to have evaporative 
emission controls.

XI. Proposals Affecting Other Engine and Vehicle Categories

    We are proposing to make several regulatory changes that would 
affect engines, equipment, and vessels other than Small SI and Marine 
SI. These changes are described in the following sections. We request 
comment on all aspects of these proposed changes.

A. State Preemption

    Section 209(e) of the Clean Air Act prohibits states and their 
political subdivisions from adopting or enforcing standards and other 
requirements relating to the control of emissions from nonroad engines 
or vehicles. Section 209(e) authorizes EPA to waive this preemption for 
California for standards and other requirements for nonroad engines and 
vehicles, excluding new engines that are smaller than 175 horsepower 
used in farm or construction equipment or vehicles and new locomotives 
or new engines used in locomotives. States other than California may 
adopt and enforce standards identical to California standards 
authorized by EPA.
    EPA promulgated regulations implementing section 209(e) on July 20, 
1994 (59 FR 36987). EPA subsequently promulgated revised regulations 
implementing section 209(e) on December 30, 1997 (62 FR 67733). See 40 
CFR part 85, subpart Q. We are proposing to create a new part 1074 that 
would describe the federal preemption of state and local emission 
requirements. This is being done as part of EPA's ongoing effort to 
write its regulations in plain language format in subchapter U of title 
40 of the CFR. The proposed regulations are based directly on the 
existing regulations in 40 CFR part 85, subpart Q. With the exception 
of the simplification of the language and specific changes described in 
this section, we are not changing the meaning of these regulations.
    Pursuant to section 428 of the 2004 Consolidated Appropriations 
Act, we are proposing to add regulatory language to implement the 
legislative restriction on states other than California adopting, after 
September 1, 2003, standards or other requirements applicable to spark-
ignition engines smaller than 50 horsepower. We are also proposing to 
add, pursuant to that legislation, criteria for EPA's consideration in 
authorizing California to adopt and enforce standards applicable to 
such engines.\103\
---------------------------------------------------------------------------

    \103\ See section 428 of the Appropriations Act for 2004.
---------------------------------------------------------------------------

    On July 12, 2002, the American Road and Transportation Builders 
Association (ARTBA) petitioned EPA to amend EPA's rules implementing 
section 209(e) of the Act.\104\ In particular, ARTBA petitioned EPA to 
amend its regulations and interpretive rule regarding preemption of 
state and local requirements ``that impose in-use and operational 
controls or fleet-wide purchase, sale or use standards on nonroad 
engines.''\105\
---------------------------------------------------------------------------

    \104\ ``Petition to Amend Rules Implementing Clean Air Act 
section 209(e),'' American Road and Transportation Builders 
Association (ARTBA), July 12, 2002. Also, EPA received an additional 
communication from ARTBA urging EPA to grant the petition after the 
decision of the U.S. Supreme Court in EMA v. SCAQMD, 541 U.S. 246 
(2004). See ``ARTBA Petition,'' L. Joseph, ARTBA, to D. Dickinson & 
R. Doyle, EPA, April 30, 2004. These documents are available in 
Docket EPA-HQ-OAR-2004-0008.
    \105\ In 1994, EPA promulgated an interpretive rule at Appendix 
A to subpart A of 40 CFR part 89. The appendix provides that state 
restrictions on the use and operation of nonroad engines are not 
preempted under section 209.
---------------------------------------------------------------------------

    ARTBA believes such controls should be preempted. As we are already 
revising the preemption provisions to a certain extent in this rule, we 
believe it is appropriate to respond to ARTBA's petition in the context 
of this rule, while giving the public the ability to respond to provide 
comments regarding ARTBA's petition. EPA is not proposing to adopt the 
explicit changes requested by ARTBA in its petition; however, EPA will 
continue to review the arguments raised by ARTBA's petition, as well as 
all further arguments provided by ARTBA and other commenters during the 
period for notice and comment on

[[Page 28210]]

this issue. We will respond to the petition, and if appropriate, make 
any changes to the regulations to conform our response to ARTBA and 
other commenters in the final rule. We request comment from the public 
regarding issues related to ARTBA's petition and how we should respond.

B. Certification Fees

    Under our current certification program, manufacturers pay a fee to 
cover the costs associated with various certification and other 
compliance activities associated with an EPA issued certificate of 
conformity. These fees are based on the actual and/or projected cost to 
EPA per emission family. We are proposing to establish a new fees 
category for certification related to the proposed evaporative emission 
standards. Sections III and VI describe how these fees would apply to 
sterndrive/inboard marine engines and equipment and vessels subject to 
evaporative emission standards since these products are not currently 
required to pay certification fees.
    In addition, we are proposing to create a new part 1027 in title 40 
that would incorporate the new and existing fee requirements under a 
single part in the regulations. This is being done as part of EPA's 
ongoing effort to write its regulations in plain language format in 
subchapter U of title 40 of the CFR. The proposed regulations are based 
directly on the existing regulations in 40 CFR part 85, subpart Y. 
Aside from a variety of specific changes, moving this language to part 
1027 is not intended to affect the substance of the existing fee 
provisions. We are proposing the following adjustments and 
clarifications to the existing regulations:
     Establishing a new fees category for new evaporative 
emission standards.
     Eliminating one of the paths for applying for a reduced 
fee. The existing regulations specify that applications covering fewer 
than six vehicles or engines, each with an estimated retail sales price 
below $75,000, shall receive a certificate for five vehicles or 
engines. Holders of these certificates are required to submit an annual 
model year reduced fee payment report adjusting the fees paid. We are 
proposing to eliminate this pathway and the associated report, as they 
are complex and have been rarely used.
     Clarifying the obligation to make additional payment on a 
reduced fee certificate if the actual final sales price is more than 
the projected retail sales price for a reduced fee vehicle or engine. 
As before, the final fee payment must also reflect the actual number of 
vehicles.
     Applying the calculated fee changes for later years, which 
are based on the Consumer Price Index and the total number of 
certificates, only after the change in the fee's value since the last 
reported change has reached $50. The fee change for the ``Other'' 
category for calendar year 2005 to 2006 changed from $826 to $839 and 
for non-road compression-ignition engines from $1822 to $1831. Under 
the proposal, the fee would not change until such time as the fee 
increase would be $50.00 or greater. This might not occur after one 
year, but after two or more years the calculated increase in a fee 
based on the change in the Consumer Price Index might be more than 
$50.00. The same applies if the price goes both up and down. For 
example, if the fee published in EPA guidance for a category of engine 
was $1,000 in 2011 and the calculated fee for 2012 is $990 and in 2013 
is $1040, the fee in 2013 would remain at $1,000 since the change from 
the 2011 fee is only $40. This would minimize confusion related to 
changing fees where the calculated fee is very close to that already 
established for the previous year. It will also lessen paperwork and 
administrative burdens for manufacturers and EPA in making adjustments 
for small fees changes for applications that are completed around the 
change in a calendar year. The number of certificates may go up or down 
in any given year, while the Consumer Price Index would generally 
increase annually. As a result, this change would be revenue-neutral or 
would perhaps slightly decrease overall revenues.
     Clarifying that all fee-related records need to be kept, 
not just those related to the ``final reduced fee calculation and 
adjustment.''
     Adding www.Pay.gov or other methods specified in guidance 
as acceptable alternative methods for payment and filing of fee forms. 
We anticipate several changes in administration of the fees program in 
coming months. It is likely that future payment of fees by electronic 
funds transfers (other than wire payments through the Federal Reserve) 
will be available only through online payments via www.Pay.gov. We are 
also receiving an increasing number of fee forms through e-mail 
submissions, which has proved to be a reliable and convenient method. 
We will be establishing a specific e-mail address for these 
submissions.
     Establishing a single deadline for all types of refunds: 
total, partial for reduced fees, and partial for corrections. In all 
cases, refund requests must be received within six months of the end of 
the model year. A common type of request is due to an error in the fee 
amount paid as a result of changed fees for a new calendar year. We 
frequently apply these overpayments to other pending certification 
applications. This is less burdensome than applying for a simple 
refund, both for EPA and for most manufacturers. Applications to apply 
such refunds to other certification applications must also be received 
within six months of the end of the model year of the original engine 
family or test group.
     Emphasizing with additional cross references that the same 
reduced fee provisions that apply to Independent Commercial Importers 
also apply to modification and test vehicle certificates under 40 CFR 
85.1509 and 89.609: the number of vehicles covered is listed on the 
certificate, a revision of the certificate must be applied for and 
additional reduced fee payments made if additional vehicles are to be 
covered, and the certificate must be revised to show the new total 
number of vehicles to be covered.

C. Amendments to General Compliance Provisions in 40 CFR Part 1068

    The provisions of part 1068 currently apply for nonroad diesel 
engines regulated under 40 CFR part 1039, Large SI engines regulated 
under 40 CFR part 1048, and recreational vehicles regulated under 40 
CFR part 1051. We are proposing to apply these provisions also for 
Small SI and Marine SI engines, equipment, and vessels. Any changes we 
make to part 1068 will apply equally for these other types of engines 
and vehicles. We therefore encourage comment from any affected 
companies for any of these proposed changes.
    The most significant change we are proposing for part 1068 is to 
clarify the language throughout to make necessary distinctions between 
engines, equipment, and fuel-system components--and particularly 
between equipment using certified engines and equipment that has been 
certified to meet equipment-based standards. This becomes necessary 
because the evaporative emission standards proposed in this document 
apply in some cases to equipment manufacturers and boat builders, while 
the exhaust emission standards apply only to engine manufacturers. Some 
provisions in part 1068 apply to equipment manufacturers differently if 
they hold a certificate of conformity rather than merely installing 
certified engines (or certified fuel-system components). The proposed 
changes in regulatory language are intended to help make those 
distinctions. See Sec.  1068.2 for a

[[Page 28211]]

description of the proposed terminology that we intend to use 
throughout part 1068.
    We are aware that in some cases manufacturers produce nonroad 
engines by starting with a complete or partially complete engine from 
another manufacturer and modifying it as needed for the particular 
application. This is especially common for Marine SI and Large SI 
engines and equipment, but it may also occur for other types of nonroad 
engines and equipment. We are concerned that an interpretation of the 
prohibited acts in Sec.  1068.101 would disallow this practice because 
the original engine manufacturer is arguably selling an engine that is 
not covered by a certificate of conformity even though emission 
standards apply. We are addressing this first by proposing to define 
``engine'' for the purposes of the regulations (see Sec.  1068.30). To 
do this, we differentiate between complete engines and partially 
complete engines, both of which need to be covered by a certificate. 
Partially complete engines would include any engine, consisting of the 
engine block plus at least one attached component such that the engine 
is not yet in its final, certified configuration. We are also proposing 
to allow for a path by which the original engine manufacturer would not 
need to certify partially complete engines or request approval for an 
exemption (see Sec.  1068.262). To do this though, the original engine 
manufacturer would need a written request from a secondary engine 
manufacturer who already holds a valid certificate of conformity for 
the engine based on its final configuration and application. These 
proposed provisions are intended generally to be clarifications of the 
existing regulatory provisions, particularly those in Sec.  1068.330 
for imported engines.
    One situation involving partially complete engines involves the 
engine block as a replacement part where the original engine had major 
structural damage. In this case the engine manufacturer will typically 
sell an engine block with piston, crankshaft, and other internal 
components to allow the user to repower with many of the components 
from the original engine. Under the proposed definitions, these short 
blocks or three-quarter blocks would be new engines subject to emission 
standards. We believe it would be appropriate to address this situation 
in the regulations with the replacement engine provisions in Sec.  
1068.240, which provides a path for making new engines that are exempt 
from current emission standards. We request comment on applying these 
replacement-engine provisions to engine blocks as replacement parts.
    We are proposing to further clarify the requirement for engine 
manufacturers to sell engines in their certified configuration. The 
existing provisions in Sec.  1068.260 describe how manufacturers may 
use delegated assembly to arrange for equipment manufacturers to 
separately source aftertreatment components for engines that depend on 
aftertreatment to meet emission standards. We are proposing to include 
language to clarify that we will consider an engine to be in its 
certified configuration in certain circumstances even if emission-
related components are not assembled to the engine. This is intended to 
reflect common practice that has developed over the years. We are also 
proposing to clarify that engines may be shipped without radiators or 
other components that are unrelated to emission controls, and that we 
may approve requests to ship engines without emission-related 
components in some circumstances. This would generally be limited to 
equipment-related components such as vehicle-speed sensors. We could 
specify conditions that we determine are needed to ensure that shipping 
the engine without such components will not result in the engine being 
operated outside of its certified configuration.
    We adopted a definition of ``nonroad engine'' that continues to 
apply today (see Sec.  1068.30). This definition distinguishes between 
portable or transportable engines that may be considered either nonroad 
or stationary, depending on the way they will be used. The distinction 
between nonroad and stationary engines is most often relevant for new 
engines in determining which emission standards apply. However, we have 
received numerous questions related to equipment whose usage has 
changed so that the original designation no longer applies. The 
definition does not address these situations. We are therefore 
proposing to adopt provisions that would apply when an engine 
previously used in a nonroad application is subsequently used in an 
application other than a nonroad application, or when an engine 
previous used in a stationary application is moved (see Sec.  1068.31).
    In addition, we are proposing several amendments to part 1068 to 
clarify various items. These include:
     Sec.  1068.101(a)(1): Revising the prohibited act to 
specify that engines must be ``covered by'' a certificate rather than 
``having'' a certificate. The revised language is more descriptive and 
consistent with the Clean Air Act.
     Sec.  1068.101(a)(1)(i): Clarifying that engines or 
equipment are considered to be uncertified if they are not in a 
configuration that is included in the applicable certificate of 
conformity. This would apply even if the product had an emission label 
stating that it complies with emission standards.
     Sec.  1068.101(a)(2): Clarifying the prohibition on 
recordkeeping to apply also to submission of records to the Agency.
     Sec.  1068.101(b)(2): Adding a prohibition against using 
engines in a way that renders emission controls inoperative, such as 
misfueling or failing to use additives that the manufacturer specifies 
as part of the engine's certified configuration. This is more likely to 
apply for compression-ignition engines than spark-ignition engines.
     Sec.  1068.101(b)(7): Clarifying the prohibitions related 
to warranty to require the submission of specified information in the 
application for certification; adding language to identify obligations 
related to recall; and preventing the manufacturer from communicating 
to users that warranty coverage is conditioned on using authorized 
parts or service facilities. These provisions are consistent with 
requirements that apply in other EPA programs.
     Sec.  1068.105(a): Revising the regulation to allow 
equipment manufacturers to use up normal inventories of previous model 
year engines only if it is a continuation of ongoing production with 
existing inventories. These provisions would not apply for an equipment 
manufacturer starting to produce a new equipment model.
     Sec.  1068.105: Eliminating paragraph (b) related to using 
highway certification for nonroad engines or equipment, since these 
provisions are spelled out specifically for each nonroad program where 
appropriate.
     Sec.  1068.105(b): Clarifying the requirement to follow 
emission-related installation instructions to include installation 
instructions from manufacturers that certify components to evaporative 
emission standards.
     Sec.  1068.120: Clarifying the rebuilding provisions to 
apply to maintenance related to evaporative emissions.
     Sec.  1068.240: Clarifying that the scope of the exemption 
for new replacement engines is limited to certain engines; also 
clarifying that the replacement engine provisions apply for replacing 
engines that meet alternate emission standards (such as those produced

[[Page 28212]]

under the Transition Program for Equipment Manufacturers).
     Sec.  1068.250: Revising the applicability of the hardship 
provisions to small businesses more broadly by referring to a term that 
is defined in Sec.  1068.30; this would include small businesses as 
identified in the standard-setting part, or any companies that meet the 
criteria established by the Small Business Administration.
     Sec.  1068.250: Clarifying the timing related to hardship 
approvals, and the ability to get extensions under appropriate 
circumstances.
     Sec.  1068.260: Revising the provisions related to 
delegated assembly as described in Section XI.F and clarifying that 
reduced auditing rates as specified in paragraph (a)(6) should be based 
on the number of equipment manufacturers involved rather than the 
number of engines; also specifying that manufacturers may itemize 
invoices to ensure that the Customs valuation for assessment of import 
duties is based on the price of the imported engine without the 
aftertreatment components that are being shipped separately. We request 
comment on adding a provision allowing for a separate invoice for 
aftertreatment components that are shipped separately.
     Sec.  1068.305: Clarifying that the requirement to submit 
importation forms applies to all engines, not just nonconforming 
engines; also adding a requirement to keep these records for five 
years. Both of these changes are consistent with the Customs 
regulations at 19 CFR 12.74.
     Part 1068, Appendix I: Clarifying that the fuel system 
includes evaporative-related components and that the parts comprising 
the engine's combustion chamber are emission-related components.
    Manufacturers have also expressed a concern that the engine 
rebuilding provisions in Sec.  1068.120 do not clearly address the 
situation in which rebuilt engines are used to repower equipment where 
the engine being replaced meets alternate emission standards (such as 
those produced under the Transition Program for Equipment 
Manufacturers). These engines are not certified to the emission 
standards that would otherwise apply for the given model year, so there 
may be some confusion regarding the appropriate way of applying these 
regulatory requirements.
    In Section V.E.6 we describe several proposed special compliance 
provisions that are intended to improve our ability to oversee our 
emission control program for Small SI engines. For example, we are 
proposing that manufacturers take steps to ensure that they will be 
able to honor emission-related warranty claims, meet any compliance- or 
enforcement-related obligations that may arise, and import new engines 
and equipment in a timely manner after we adopt new standards. We 
request comment on the appropriateness of adopting any or all of those 
provisions under part 1068 such that they would apply to all engines 
and equipment subject to part 1068. We also request comment on any 
adjustments to those provisions that would be appropriate for other 
categories of engines and equipment, whether we choose to adopt these 
provisions in this proposal or in a separate rulemaking.
    In addition, we request comment on early application of the 
provisions of part 1068 before the standards proposed in this notice 
take effect. For example, for any provisions not directly related to 
the emission standards, we could revise the regulations in part 90 and 
part 91 to reference the corresponding provisions in part 1068. We 
similarly request comment on making these changes for diesel engines 
regulated under part 89 (land-based) and part 94 (marine). This would 
allow us to accelerate the transition to plain-language regulations and 
prevent confusion from maintaining multiple versions of similar 
provisions for several years. We would also be able to substantially 
decrease printing costs. The provisions most appropriately considered 
for early transition to part 1068 include: (1) Selective enforcement 
audits, (2) exemptions, (3) importation provisions, (4) defect 
reporting and recall, (5) hearing procedures, and (6) treatment of 
confidential information.
    We are also seeking comment on revisions to 40 CFR 1068.101. 
Section 203 of the Act (42 U.S.C. 7522) states that performing certain 
acts, ``and causing thereof,'' constitutes a prohibited act. We are 
interested in revising the regulations to specifically include this 
prohibition on the ``causing'' of any of the prohibited acts listed in 
the statute and the regulations. Adding this clarification would help 
people who are subject to the regulations to more fully understand what 
actions are prohibited and may potentially subject them to enforcement 
proceedings under the Act. The revisions themselves would not be 
intended to add new enforcement authorities beyond what is already 
specified in the statute.
    If we consider it a violation to cause someone to commit a 
prohibited act, then persons causing any prohibited act would also be 
subject to the full administrative and judicial enforcement actions 
allowable under the Act and the regulations. The prohibition on 
``causing'' a prohibited act would apply to all persons and would not 
be limited to manufacturers or importers of regulated engines or 
equipment.
    If this provision is adopted, EPA would interpret the ``causation'' 
aspect of section 203 broadly. In assessing whether a person has caused 
a prohibited act, EPA would evaluate the totality of circumstances. For 
example, in certain circumstances EPA believes a retailer may be 
responsible for causing the importation of engines or equipment not 
covered by a valid certificate of conformity or otherwise in violation 
of our regulations, such as the labeling requirements. In addition to 
the prohibitions that apply to manufacturers and importers generally 
under section 203, EPA will also consider many factors in assessing 
whether a manufacturer, importer, retailer, distributor or other person 
has caused a prohibited act, including, but not limited to, the 
following: (1) The contractual or otherwise established business 
relationship of those persons involved in producing and/or selling new 
engines and equipment; (2) the particular efforts or influence of the 
alleged violator contributing to, leading to or resulting in the 
prohibited act; and (3) the efforts, or lack thereof, of the person to 
prevent such a violation. EPA will evaluate the entire circumstances in 
determining whether a person caused another person to commit a 
prohibited act such as importing engines or equipment in violation of 
our regulations.

D. Amendments Related to Large SI Engines (40 CFR Part 1048)

    Manufacturers of Large SI engines are encouraged to review the 
proposed changes described in Section XI.C related to 40 CFR part 1068.
    Some of the issues related to Marine SI engines described in 
Section III relate to Large SI engines. In particular, the uncertain 
availability of certain base engine models from General Motors for use 
in nonroad applications poses a challenge for efforts to certify the 
engines to the Large SI standards. In particular, the uncertain lead 
time associated with getting the new engines and the level of effort 
expected for certifying the existing engine models that are planned for 
obsolescence make it difficult for companies, especially small 
businesses, to go through the certification process and recover costs 
for repeated testing. Of greatest concern are requirements related to 
developing deterioration factors for these engines. The existing 
regulations allow for assigned deterioration factors for small 
businesses, but these apply only to companies with fewer than 200

[[Page 28213]]

employees. We are therefore proposing to expand the definition of 
small-volume engine manufacturer to also include companies with annual 
U.S. sales of no more than 2000 Large SI engines. This would align with 
the provisions already adopted by California ARB. Similarly, we are 
proposing a provision allowing for assigned deterioration factors for 
small-volume engine families for Small SI engines (see Section V). A 
similar dynamic applies for Large SI engines. Any such allowance would 
apply to engine families with projected sales up to 300 or 500 units to 
reflect to different production volumes. We request comment on allowing 
assigned deterioration factors for small-volume engine families for 
Large SI engines, and on the appropriate threshold for this provision.
    We are also proposing to revise the provisions related to 
competition engines to align with the proposal for Small SI engines. 
Any Small SI engine that is produced under the competition exemption 
will very likely exceed 19 kW. As a result, we believe it is 
appropriate to make these provisions identical to avoid confusion.
    Manufacturers have notified us that the transient test for 
constant-speed engines does not represent in-use operation in a way 
that significantly affects measured emission levels. This notification 
is required by Sec.  1065.10(c)(1). In particular, manufacturers have 
pointed out that the specified operation involves light engine loads 
such that combustion and exhaust temperatures do not rise enough to 
reach catalyst light-off temperatures. As a result, meeting the 
standard using the constant-speed transient test would require the use 
of significantly oversized catalysts, which would add significant costs 
without a commensurate improvement for in-use emission control. We 
faced a similar dilemma in the effort to adopt transient standards for 
nonroad diesel engines, concluding that the transient standards should 
not apply until we develop a more suitable duty cycle that more 
appropriately reflects in-use operation. We are proposing to take this 
same approach for Large SI engines, waiving the requirement constant 
speed engines to meet the transient standards until we are able to 
develop a more appropriate duty cycle. Manufacturers must continue to 
meet the standards for steady-state testing and the field-testing 
standards continue to apply. We are also proposing to clarify that 
manufacturers certifying constant-speed engines should describe their 
approach to controlling emissions during transient operation in their 
application for certification.
    Manufacturers have also pointed out that a multiplicative 
deterioration factor is problematic for engines with very low emission 
levels. While the HC+NOX emissions may be as high as 2.7 g/
kW-hr, manufacturers are certifying some engine families with 
deteriorated emission levels below 0.1 g/kW-hr. These very low emission 
levels are well below the standard, but the measurement systems are 
challenged to produce a precisely repeatable emission level at that 
point. As a result, measurement variability and minor engine-to-engine 
variability can lead to small absolute differences in emission levels 
that become magnified by a deterioration factor that reflects the 
extremely small low-hour measurement. We are therefore proposing to 
specify that manufacturers use an additive deterioration factor if 
their low-hour emission levels are below 0.3 g/kW-hr. This change would 
accommodate the mathematical and analyzer effects of very low emission 
levels without changing the current practice for the majority of 
engines that are certified with emission levels closer to the standard. 
This change would remove the incentive for manufacturers to increase 
their engine's emission levels to avoid an artificially large 
deterioration factor. The only exception would be for cases in which 
good engineering judgment dictates that a multiplicative deterioration 
factor would nevertheless be appropriate for engines with very low 
emissions. This may be the case if an engine's deterioration can be 
attributed, even at very low emission levels, to proportionally 
decreased catalyst conversion of emissions from an aged engine. It is 
important to note that Large SI engine manufacturers are subject to in-
use testing to demonstrate that they meet emission standards throughout 
the useful life. Should such testing indicate that an additive 
deterioration factor does not appropriately reflect actual performance, 
we would require manufacturers to revise their deterioration factors 
appropriately, as required under the current regulations. If such 
discrepancies appear for multiple manufacturers, we would revise the 
regulation to again require multiplicative deterioration factors for 
all aftertreatment-based systems. We also request comment on a further 
refinement of the form of the deterioration factor to more closely 
reflect the degradation in catalyst conversion efficiency. For example, 
measuring engine-out emissions would allow for calculating catalyst 
conversion efficiency, such that changes in this parameter over an 
engine's useful life could be factored into a calculation to 
characterize an engine's actual rate of deterioration.
    Most Large SI engines are installed in equipment that has metal 
fuel tanks. This formed the basis of the regulatory approach to set 
evaporative emission standards and certification requirements. 
Manufacturers have raised questions about the appropriate steps to take 
for systems that rely on plastic fuel tanks. These tanks are able to 
meet standards, but questions have been raised about the engine 
manufacturer's role in certifying a range of fuel tanks with their 
engines. We request comment on the extent to which the current 
regulatory requirements might limit the range of fuel tank designs.
    The current permeation standards for Large SI equipment references 
Category 1 fuel lines as defined in the version of SAE J2260 that was 
issued in November, 1996. In 2004, the Society of Automotive Engineers 
(SAE) updated SAE J2260. Manufacturers have asked whether we will 
approve fuel lines based on the updated procedures. The new procedures 
have two primary differences related to fuel line permeation. First, 
the test fuel was changed from CM15 to CE10.\106\ Second, the 
associated limits for the different categories of fuel line permeation 
were revised. Data presented in Chapter 5 of the Draft RIA suggest that 
permeation from low-permeation fuel line materials can be less than 
half on CE10 than on CM15. The permeation specification for Category 1 
fuel line was revised by SAE from 0-25 g/m\2\/day to 3-10 g/m\2\/day. 
(A new Category 0 was added at 0-3 g/m\2\/day.) Directionally, the new 
Category 1 permeation limits seem to account for the change in the test 
fuel. In addition, ethanol fuel blends are commonly used in-use while 
methanol fuel blends are less common. We request comment on updating 
the regulations for Large SI equipment to reference the Category 1 fuel 
line specifications in the updated version of SAE J2260 (revised 
November 2004). We also request comment on whether this new 
specification would affect the stringency of the standard or the choice 
of fuel line constructions for this equipment.
---------------------------------------------------------------------------

    \106\ ``C'' refers to fuel C as specified in ASTM D 412, E10 
refers to 10 percent ethanol, and M15 refers to 15 percent methanol.
---------------------------------------------------------------------------

    We are also proposing several technical amendments to part 1048. 
Many of these simply correct

[[Page 28214]]

typographical errors or add references to the proposed regulatory cites 
in part 1054. Several changes are intended merely to align regulatory 
language with that of other programs, including those that would be 
subject to the standards proposed in this notice. In addition, we are 
proposing the following changes:
     Sec.  1048.5: Clarifying that locomotive propulsion 
engines are not subject to Large SI emission standards, even if they 
use spark-ignition engines. This is based on the separate provisions 
that apply to locomotives in Clean Air Act section 213.
     Sec.  1048.101: Clarifying manufacturer's responsibility 
to meet emission standards for different types of testing, especially 
to differentiate between field-testing standards and duty-cycle 
standards.
     Sec.  1048.105: Clarifying that only the permeation 
standards of SAE J2260 apply to fuel lines used with Large SI engines.
     Sec.  1048.105: Clarifying that the requirement to prevent 
fuel boiling is affected by the pressure in the fuel tank. The 
regulation currently characterizes the boiling point of fuel only at 
atmospheric pressure. Pressurizing the fuel tank increases the boiling 
point of the fuel.
     Sec.  1048.105: Reorganizing the regulatory provisions to 
align with the new language in 40 CFR part 1060. This is not intended 
to change any of the applicable requirements.
     Sec.  1048.110: Clarifying that ``malfunctions'' relate to 
engines failing to maintain emission control and not to diagnostic 
systems that fail to report signals; and clarifying that the 
malfunction indicator light needs to stay illuminated for malfunctions 
or for system errors.
     Sec.  1048.120: Clarifying that the emission-related 
warranty covers only those components from 40 CFR part 1068, Appendix 
I, whose failure will increase emissions.
     Sec.  1048.125: Clarifying the provisions related to 
noncritical emission-related maintenance.
     Sec.  1048.135: Revising the engine labeling requirements 
to allow omission of the manufacturing date only if the date is stamped 
or engraved on the engine, rather than allowing manufacturers to keep 
records of engine build dates. This is important for verifying that 
engines comply with standards based on their build date.
     Sec.  1048.205: Removing detailed specifications for 
describing auxiliary emission control devices in the application for 
certification. This responds to the concern expressed by manufacturers 
that the existing, very prescriptive approach requires much more 
information than is needed to adequately describe emission control 
systems. We are proposing to leave in place a broad requirement to 
describe emission control systems and parameters in sufficient detail 
to allow EPA to confirm that no defeat devices are employed. 
Manufacturers should be motivated to include substantial information to 
make such determinations in the certification process, rather than 
being subject to this type of investigation for emission control 
approaches that are found to be outside of the scope of the application 
for certification.
     Sec.  1048.205: Adding requirement to align projected 
sales volumes with actual sales from previous years. This does not 
imply additional reporting or recordkeeping requirements. It is 
intended simply to avoid situations where manufacturers intentionally 
mis-state their projected sales volume to gain some advantage under the 
regulations.
     Sec.  1048.205: Specifying that manufacturers must submit 
modal emission results rather than just submitting a weighted average. 
Since this information is already part of the demonstration related to 
the field-testing standards, this should already be common practice.
     Sec.  1048.220: Clarifying that if manufacturers change 
their maintenance instructions after starting production for an engine 
family, they may not disqualify engines for in-use testing or warranty 
claims based on the fact that operators did not follow the revised 
maintenance instructions.
     Sec.  1048.225: Clarifying the terminology to refer to 
``new or modified engine configurations'' rather than ``new or modified 
nonroad engines.'' This is necessary to avoid using the term ``new 
nonroad engine'' in a way that differs from the definitions in Sec.  
1048.801.
     Sec.  1048.230: Clarifying that engine families relate 
fundamentally to emission certification and that we would expect 
manufacturers to suggest a tailored approach to specifying engine 
families under Sec.  1048.230(d) to occur only in unusual 
circumstances.
     1048.240: Adding a requirement for design-based 
certification for the diurnal standards that fuel tanks need to use 
low-permeation materials.
     1048.245: Adding the provision to allow for component 
certification for plastic fuel tanks. The revised language clarifies 
the requirement related to allowing pressure relief for vacuum 
pressures and for controlling permeation rates from plastic fuel tanks.
     Sec.  1048.250: Adding a requirement for manufacturers to 
report their sales volumes for an engine family if they are using a 
provision that depends on production volumes.
     Sec.  1048.301: Clarifying that engine families with 
projected sales volumes below 150 units may have reduced testing rates 
for production-line testing. This level of production does not allow 
for adequate testing to use the statistical techniques before exceeding 
specified maximum testing rates.
     Sec.  1048.305: Clarifying that (1) Tested engines should 
be built in a way that represents production engines; (2) the field-
testing standards apply for any testing conducted (this may involve 
simply comparing modal results to the field-testing standards); and (3) 
we may review a decision to use emission results from a retested engine 
instead of the original results.
     Sec.  1048.310: Clarifying the relationship between 
quarterly testing and compliance with the annual testing requirements.
     Sec.  1048.315: Correcting the equation for the CumSum 
statistic to prevent negative values.
     Sec.  1048.410: Clarifying that repeat tests with an in-
use test engine are acceptable, as long as the same number of repeat 
tests are performed for all engines.
     Sec.  1048.415: Clarifying that the provisions related to 
defect reporting in 40 CFR 1068.501 apply for in-use testing.
     Sec.  1048.501: Removing specified mapping procedures, 
since these are addressed in 40 CFR part 1065.
     Sec.  1048.505: Removing redundant text and removing 
sampling times specified in Table 1, since these are addressed in Sec.  
1048.505(a)(1).
     Sec.  1048.505: Correcting the mode sequence listed in the 
table for the ramped-modal testing.
     Sec.  1048.505: Clarifying that cycle statistics for 
discrete-mode testing must be calculated separately for each mode.
     Sec. Sec.  1048.605 and 1048.610: Requiring some 
demonstration that the sales restrictions that apply for these sections 
are met, and clarifying the provisions related to emission credits for 
vehicles that generate or use emission credits under 40 CFR part 86.
     Sec.  1048.801: Revising several definitions to align with 
updated definitions adopted (or proposed) for other programs.
    We request comment on changing Sec.  1048.220 to prevent 
manufacturers from distributing revised emission-related maintenance 
instructions until we have approved them. We are taking this approach 
for Small SI and Marine

[[Page 28215]]

SI engines in this proposal (see Sec. Sec.  1045.220 and 1054.220) 
because we believe it would be inappropriate for manufacturers to 
specify increased or decreased emission-related maintenance without EPA 
approval of those changes. The same concern applies equally to all 
nonroad spark-ignition engines and vehicles, so we would expect to 
apply the same policy to all these engines.
    For Small SI and Marine SI engines we are proposing to require 
manufacturers of imported engines to include basic information in the 
application for certification, including identification of associated 
importers, specific ports intended for importation, and testing 
facilities where testing could be done in the United States. We request 
comment on extending these provisions to Large SI engines. See Sec.  
1054.205.

E. Amendments Related To Recreational Vehicles (40 CFR Part 1051)

    Manufacturers of recreational vehicles are encouraged to review the 
proposed changes described in Section XI.C related to 40 CFR part 1068.
    We are proposing in this notice to establish a process by which 
manufacturers of fuel system components certify that their products 
meet emission standards. For recreational vehicles we adopted a program 
in which the exhaust and evaporative emission standards apply to the 
vehicle so we did not set up a process for certifying fuel-system 
components. We continue to believe that evaporative emission standards 
should apply to the vehicle. However, we are proposing to allow 
manufacturers of fuel-system components to opt in to this program by 
certifying their fuel tanks or fuel lines to the applicable standards. 
While this would be a voluntary step, any manufacturer opting into the 
program in this way would be subject to all the requirements that apply 
to certificate holders. While manufacturers of recreational vehicles 
would continue to be responsible for meeting standards and certifying 
their vehicles, it may be appropriate to simplify their compliance 
effort by allowing them to rely on the certification of the fuel-line 
manufacturer or fuel-tank manufacturer.
    We also request comment on specifying that vehicle manufacturers 
use the certification and testing procedures proposed in 40 CFR part 
1060 to meet the evaporative emission standards included in part 1051. 
This would not be intended to affect the stringency of current 
requirements. This would simply allow us to maintain consistent 
requirements across programs and avoid publishing redundant 
specifications.
    We are also proposing several technical amendments to part 1051. 
Many of these simply correct typographical errors or add references to 
the proposed regulatory cites in part 1054. Several changes are 
intended merely to align regulatory language with that of other 
programs, including those that would be subject to the standards 
proposed in this notice.
    In addition, we are proposing the following changes:
     Sec.  1051.1: Revising the speed threshold for offroad 
utility vehicles to be subject to part 1051. Changing from ``25 miles 
per hour or higher'' to ``higher than 25 miles per hour'' aligns this 
provision with the similar threshold for qualifying as a motor vehicle 
in 40 CFR 85.1703.
     Sec.  1051.5: Clarifying the status of very small 
recreational vehicles to reflect the provisions in the current 
regulations in 40 CFR part 90 to treat such vehicles with a dry weight 
under 20 kilograms as Small SI engines.
     Sec.  1051.25: Clarifying that manufacturers of 
recreational vehicles that use engines certified to meet exhaust 
emission standards must still certify the vehicle with respect to the 
evaporative emission standards.
     Sec.  1051.120: Clarifying that the emission-related 
warranty covers only those components from 40 CFR part 1068, Appendix 
I, whose failure will increase emissions.
     Sec.  1051.125: Clarifying the provisions related to 
noncritical emission-related maintenance.
     Sec.  1051.135: Revising the labeling requirements to 
allow omission of the manufacturing date only if the date is stamped or 
engraved on the vehicle, rather than allowing manufacturers to keep 
records of vehicle build dates. This is important for verifying that 
vehicles comply with standards based on their build date.
     Sec.  1051.135: Adding a requirement to include family 
emission limits related to evaporative emissions to the emission 
control information label. Since this change may involve some time for 
manufacturers to comply, we are proposing to apply this starting with 
the 2009 model year.
     Sec.  1051.137: Clarifying how the labeling requirements 
apply with respect to the averaging program and selected family 
emission limits.
     Sec.  1051.205: Removing detailed specifications for 
describing auxiliary emission control devices in the application for 
certification. This responds to the concern expressed by manufacturers 
that the existing, very prescriptive approach requires much more 
information that is needed to adequately describe emission control 
systems. We are proposing to leave in place a broad requirement to 
describe emission control systems and parameters in sufficient detail 
to allow EPA to confirm that no defeat devices are employed. 
Manufacturers should be motivated to include substantial information to 
make such determinations in the certification process, rather than 
being subject to this type of investigation for emission control 
approaches that are found to be outside of the scope of the application 
for certification.
     Sec.  1051.205: Requirements to align projected sales 
volumes with actual sales from previous years. This does not imply 
additional reporting or recordkeeping requirements. It is intended 
simply to avoid situations where manufacturers intentionally mis-state 
their projected sales volume to gain some advantage under the 
regulations.
     Sec.  1051.220: Clarifying that if manufacturers change 
their maintenance instructions after starting production for an engine 
family, they may not disqualify vehicles for warranty claims based on 
the fact that operators did not follow the revised maintenance 
instructions.
     Sec.  1051.225: Clarifying the terminology to refer to 
``new or modified vehicle configurations'' rather than ``new or 
modified vehicles.'' This is necessary to avoid confusion with the term 
``new vehicle'' as it relates to introduction into commerce.
     Sec.  1051.225: Clarifying the provisions related to 
changing an engine family's Family Emission Limit after the start of 
production.
     Sec.  1051.255: Adopting a different SAE standard for 
specifying low-permeability materials to allow for design-based 
certification of metal fuel tanks with gaskets made of polymer 
materials. The existing language does not adequately characterize the 
necessary testing and material specifications.
     Sec.  1051.230: Clarifying that engine families relate 
fundamentally to emission certification and that we would expect 
manufacturers to suggest a tailored approach to specifying engine 
families under Sec.  1051.230(e) to occur only in unusual 
circumstances.
     Sec.  1051.250: Adding a requirement for manufacturers to 
report their sales volumes for an engine family if they are using a 
provision that depends on production volumes.
     Sec.  1051.301: Clarifying that engine families with 
projected sales volumes

[[Page 28216]]

below 150 units may be exempted from production-line testing. This 
level of production does not allow for adequate testing to use the 
statistical techniques before exceeding specified maximum testing 
rates.
     Sec.  1051.305: Clarifying that tested vehicles should be 
built in a way that represents production vehicles.
     Sec.  1051.310: Clarifying the relationship between 
quarterly testing and compliance with the annual testing requirements; 
and clarifying the testing provisions that apply for engine families 
where the production period is substantially less than a full year.
     Sec.  1051.315: Correcting the equation for the CumSum 
statistic to prevent negative values.
     Sec.  1051.325: Clarifying the basis on which we would 
approve retroactive changes to the Family Emission Limit for an engine 
family that has failed under production-line testing.
     Sec.  1051.505: Clarifying that cycle statistics for 
discrete-mode testing must be calculated separately for each mode.
     Sec. Sec.  1051.605 and 1051.610: Requiring some 
demonstration that the sales restrictions that apply for these sections 
are met.
     Sec.  1051.650: Add a requirement to certify vehicles that 
are converted to run on a different fuel. We expect this is a rare 
occurrence, but one that we should make subject to certification 
requirements (see Section VII.B.3).
     Sec.  1051.701: Clarifying that manufacturers using 
emission credits to meet emission standards must base their credit 
calculations on their full product line-up, rather than considering 
only those engine families with Family Emission Limits above or below 
the emission standard. We are also clarifying that a single family may 
not generate emission credits for one pollutant while using emission 
credits for another pollutant, which is common to all our emission 
control programs.
     Sec.  1051.735: Adding a requirement to keep records 
related to banked emission credits for as long as a manufacturer 
intends for those credits to be valid. This is necessary for us to 
verify the appropriateness of credits used for demonstrating compliance 
with emission standards in later model years.
     Sec.  1051.801: Revising several definitions to align with 
updated definitions adopted (or proposed) for other programs.
    We request comment on changing Sec.  1051.220 to prevent 
manufacturers from distributing revised emission-related maintenance 
instructions until we have approved them. We are taking this approach 
for Small SI and Marine SI engines in this proposal (see Sec. Sec.  
1045.220 and 1054.220) because we believe it would be inappropriate for 
manufacturers to specify increased or decreased emission-related 
maintenance without EPA approval of those changes. The same concern 
applies equally to all nonroad spark-ignition engines and vehicles, so 
we would expect to apply the same policy to all these engines.
    For Small SI and Marine SI engines we are proposing to require 
manufacturers of imported engines to include basic information in the 
application for certification, including identification of associated 
importers, specific ports intended for importation, and testing 
facilities where testing could be done in the United States. We request 
comment on extending these provisions to recreational vehicles. See 
Sec.  1054.205.

F. Amendments Related to Heavy-Duty Highway Engines (40 CFR Part 85)

    We are proposing to make several adjustments to the provisions 
related to delegated assembly specified in Sec.  85.1713. These 
adjustments include:
     Removing the provision related to auditing outside the 
United States since equipment manufactured in other countries would not 
be subject to these provisions
     Clarifying that the exemption expires when the equipment 
manufacturer takes possession of the engine, but not before it reaches 
the point of final assembly
     Clarifying the prohibition related to following 
installation instructions to ensure that engines will be in their 
certified configuration when installed in a piece of equipment.
    We believe all these amendments are straightforward adjustments 
that are appropriate for maintaining a program that allows for 
appropriate oversight and implementation.

G. Amendments Related to Stationary Spark-Ignition Engines (40 CFR Part 
60)

    On June 12, 2006 we proposed emission standards for stationary 
spark-ignition engines (71 FR 33804). The June 2006 proposal specified 
that stationary spark-ignition engines at or below 19 kW would be 
subject to all the same emission standards and certification 
requirements that apply to Small SI engines. If we would include the 
new Phase 3 standards for Small SI engines in 40 CFR part 90, these 
requirements would apply automatically to those stationary engines. 
However, since the Phase 3 standards will be in 40 CFR part 1054, as 
described in Section V, we are proposing to revise the regulatory 
language for stationary spark-ignition engines in 40 CFR part 60, 
subpart JJJJ, to directly reference the Phase 3 standards part 1054.

XII. Projected Impacts

A. Emissions from Small Nonroad and Marine Spark-Ignition Engines

    As discussed in previous sections, this proposal will reduce 
exhaust emissions from specific sizes of nonhandheld Small SI and 
Marine SI engines. It will also reduce evaporative emissions from the 
fuel systems used on nonhandheld and handheld Small SI equipment and 
Marine SI vessels (for simplicity we collectively include the 
evaporative emission requirements from equipment or vessels when 
referring to Small SI or Marine SI engines in the remainder of this 
section). The proposed exhaust and evaporative emission standards will 
directly affect volatile organic hydrocarbon compounds (VOC), oxides of 
nitrogen (NOX), and to a lesser extent carbon monoxide (CO). 
Also, we anticipate that the emission control technology which is 
likely to be used to meet the exhaust emission standards will affect 
directly emitted particulate matter, most importantly particles with 
diameters of 2.5 micrometers or less (PM2.5). It will also 
incrementally reduce air toxic emissions. A detailed analysis of the 
effects of this proposal on emissions and emission inventories can be 
found in Chapter 3 of the Draft RIA.
    The contribution of exhaust and evaporative emissions from Small SI 
and Marine SI engines to total 50-state emission inventories is 
significant and will remain so into the future. Table XII-1 presents 
the nationwide inventory for these engines for both 2001 and 2020. (The 
inventories cover all Small SI and Marine SI engines including the 
portion of Small SI engines regulated by the California ARB.) Table 
XII-1 shows that for the primary pollutants affected by this proposal, 
these engines contribute about 25 to 30 percent of the nationwide VOC 
emissions from all mobile sources. The nationwide contribution to the 
total mobile source NOX inventory is about 5 percent or 
less. Finally, for PM2.5, the contribution ranges from about 
25 to 30 percent.

[[Page 28217]]



 Table XII-1.--Contribution of Small Nonroad and Marine SI Engines to National (50-State) Mobile Source Emission
                                                   Inventories
----------------------------------------------------------------------------------------------------------------
                                                               2001                            2020
                                                 ---------------------------------------------------------------
                                                     Small SI/                       Small SI/
                    Pollutant                        marine SI      Percent of       marine SI      Percent of
                                                    inventory,     mobile source    inventory,     mobile source
                                                       tons          inventory         tons          inventory
----------------------------------------------------------------------------------------------------------------
VOC.............................................       2,239,056              28       1,351,739              27
NOX.............................................         159,051               1         201,789               4
PM2.5...........................................          42,294               9          39,271              16
CO..............................................      20,867,436              24      16,373,518              31
----------------------------------------------------------------------------------------------------------------

(1) VOC
    Table XII-2 shows the VOC emissions and emission reductions we 
expect both with and without the proposed standards for engines, 
equipment, and vessels affected by the proposal. In 2001, Small SI and 
Marine SI emitted approximately 1,081,000 and 961,000 tons of VOC, 
respectively. Without the proposed standards, these emissions will 
decrease because of the effect of the existing emission control 
requirements to about 1,005,000 and 490,000 tons by 2040, respectively. 
With the proposed controls, this pollutant will be further reduced by 
34 percent for Small SI engines and 74 percent for Marine SI engines by 
2040. The VOC emission inventory trends over time for both categories 
of engines that are subject to the proposal are shown in Figure XII-1.

   Table XII-2.--National (50-State) VOC Emissions and Emission Reductions for Small SI and Marine SI Engines
----------------------------------------------------------------------------------------------------------------
                                                      Without      With proposed                      Percent
          Year                   Category          proposed rule       rule          Reduction       reduction
----------------------------------------------------------------------------------------------------------------
2001....................  Small Engine..........       1,080,898       1,080,898
                          Marine................         961,240         961,240
                          Both..................       2,042,138       2,042,138
2015....................  Small Engine..........         708,331         510,617         197,714              28
                          Marine................         513,105         372,020         141,086              27
                          Both..................       1,221,436         882,637         338,799              28
2020....................  Small Engine..........         764,453         508,677         255,776              33
                          Marine................         466,624         232,697         233,927              50
                          Both..................       1,231,078         741,375         489,703              40
2030....................  Small Engine..........         884,188         581,766         302,422              34
                          Marine................         464,490         135,956         328,533              71
                          Both..................       1,348,678         717,723         630,955              47
2040....................  Small Engine..........       1,005,403         659,976         345,427              34
                          Marine................         490,052         127,158         362,893              74
                          Both..................       1,495,455         787,135         708,320              47
----------------------------------------------------------------------------------------------------------------


[[Page 28218]]

[GRAPHIC] [TIFF OMITTED] TP18MY07.001

(2) NOX
    Table XII-3 shows the NOX emissions and emission 
reductions we expect both with and without the proposed standards for 
engines affected by the proposal. In 2001, Small SI and Marine SI 
emitted approximately 102,000 and 41,500 tons of NOX, 
respectively. Without the proposed standards, these emissions will 
increase to about 135,000, and 95,400 tons by 2040, respectively. With 
the proposed controls, this pollutant will be reduced by 47 percent for 
Small SI engines and 51 percent for Marine SI engines by 2040. The 
NOX emission inventory trends over time for both categories 
of engines that are subject to the proposal are shown in Figure XII-2.

   Table XII-3.--National (50-State) NOX Emissions and Emission Reductions for Small SI and Marine SI Engines
----------------------------------------------------------------------------------------------------------------
                                                      Without      With proposed                      Percent
          Year                   Category          proposed rule       rule          Reduction       reduction
----------------------------------------------------------------------------------------------------------------
2001....................  Small Engine..........         101,928         101,928
                          Marine................          41,514          41,514
                          Both..................         143,442         143,442
2015....................  Small Engine..........          94,432          58,117          36,315              38
                          Marine................          73,583          59,024          14,558              20
                          Both..................         168,015         117,141          50,874              30
2020....................  Small Engine..........         102,310          55,241          47,069              46
                          Marine................          80,655          55,656          24,999              31
                          Both..................         182,965         110,896          72,069              39
2030....................  Small Engine..........         118,615          62,778          55,837              47
                          Marine................          89,225          46,859          42,366              47
                          Both..................         207,840         109,637          98,203              47
2040....................  Small Engine..........         135,136          71,361          63,775              47
                          Marine................          95,440          46,874          48,567              51
                          Both..................         230,577         118,235         112,342              49
----------------------------------------------------------------------------------------------------------------


[[Page 28219]]

[GRAPHIC] [TIFF OMITTED] TP18MY07.002

(3) PM2.5
    Table XII-4 shows the PM2.5 emissions and emission reductions we 
expect both with and without the proposed standards for engines 
affected by the proposal. In 2001, Small SI and Marine SI emitted 
23,200 and 15,600 tons of PM2.5, respectively. Without the proposed 
standards, the PM2.5 emissions from Small SI engines will increase to 
39,100 by 2040, while those from Marine SI will decrease to about 6,000 
tons in that year due to the effects of the existing emission control 
requirements for certain types of recreational marine engines, e.g, 
outboards. With the proposed controls, this pollutant will be reduced 
by 5 percent for Small SI engines and a further 84 percent for Marine 
SI engines by 2040. The PM2.5 emission inventory trends over time for 
both categories of engines that are subject to the proposal are shown 
in Figure XII-3.

  Table XII-4.--National (50-State) PM2.5 Emissions and Emission Reductions for Small SI and Marine SI Engines
----------------------------------------------------------------------------------------------------------------
                                                      Without      With proposed                      Percent
          Year                   Category          proposed rule       rule          Reduction       reduction
----------------------------------------------------------------------------------------------------------------
2001....................  Small Engine..........          23,163          23,163
                          Marine................          15,625          15,625
                          Both..................          38,789          38,789
2015....................  Small Engine..........          27,747          26,647           1,100               4
                          Marine................           6,823           4,666           2,157              32
                          Both..................          34,570          31,313           3,256               9
2020....................  Small Engine..........          30,009          28,574           1,435               5
                          Marine................           5,908           2,448           3,461              59
                          Both..................          35,917          31,022           4,896              14
2030....................  Small Engine..........          34,535          32,849           1,686               5
                          Marine................           5,719           1,107           4,613              81
                          Both..................          40,255          33,956           6,299              16
2040....................  Small Engine..........          39,079          37,153           1,926               5
                          Marine................           6,016             985           5,031              84
                          Both..................          45,095          38,138           6,957              15
----------------------------------------------------------------------------------------------------------------


[[Page 28220]]

[GRAPHIC] [TIFF OMITTED] TP18MY07.003

(4) CO
    Table XII.-5 shows the CO emissions and emission reductions we 
expect both with and without the proposed standards for engines 
affected by the proposal. In 2001, Small SI and Marine SI emitted 
16,108,000 and 2,585,000 tons of PM2.5, respectively. Without the 
proposed standards, these emissions will increase slightly for Small SI 
engines to 16,727,000 and decrease slightly for Marine SI engines to 
2,122,000 tons by 2040, respectively. With the proposed controls, this 
pollutant will be reduced by 16 percent for Small SI engines and a 
further 22 percent for Marine SI engines by 2040. The CO emission 
inventory trends over time for both categories of engines that are 
subject to the proposal are shown in Figure XII-4.

    Table XII-5.--National (50-State) CO Emissions and Emission Reductions for Small SI and Marine SI Engines
----------------------------------------------------------------------------------------------------------------
                                                      Without      With proposed                      Percent
          Year                   Category          proposed rule       rule          Reduction       reduction
----------------------------------------------------------------------------------------------------------------
2001....................  Small Engine..........      16,108,103      16,108,103  ..............  ..............
                          Marine................       2,584,786       2,584,786  ..............  ..............
                          Both..................      18,692,890      18,692,890  ..............  ..............
2015....................  Small Engine..........      11,797,078      10,317,051       1,480,027              13
                          Marine................       2,031,684       1,883,241         148,443               7
                          Both..................      13,828,762      12,200,291       1,628,471              12
2020....................  Small Engine..........      12,712,775      10,782,258       1,930,518              15
                          Marine................       1,968,663       1,718,956         249,707              13
                          Both..................      14,681,439      12,501,214       2,180,225              15
2030....................  Small Engine..........      14,700,521      12,411,661       2,288,860              16
                          Marine................       2,009,248       1,607,678         401,570              20
                          Both..................      16,709,768      14,019,339       2,690,429              16
2040....................  Small Engine..........      16,726,708      14,113,517       2,613,191              16
                          Marine................       2,122,336       1,665,392         456,943              22
                          Both..................      18,849,044      15,778,910       3,070,134              16
----------------------------------------------------------------------------------------------------------------


[[Page 28221]]

[GRAPHIC] [TIFF OMITTED] TP18MY07.004

B. Estimated Costs

    In assessing the economic impact of setting emission standards, we 
have made a best estimate of the costs associated with the technologies 
we anticipate manufacturers will use in meeting the standards. In 
making our estimates for the proposed rule, we have relied on our own 
technology assessment, which includes information developed by EPA's 
National Vehicle and Fuel Emissions Laboratory (NVFEL). Estimated costs 
include variable costs (e.g. hardware and assembly time) and fixed 
costs (e.g. research and development, retooling, engine certification 
and test cell upgrades to 40 CFR 1065 requirements). We projected that 
manufacturers will recover the fixed costs over five years of 
production and used an amortization rate of 7 percent in our analysis. 
The analysis also considers total operating costs, including 
maintenance and fuel consumption. Cost estimates based on the projected 
technologies represent an expected change in the cost of engines as 
they begin to comply with new emission standards. All costs are 
presented in 2005 dollars. Full details of our cost analysis can be 
found in Chapter 6 of the Draft RIA. Estimated costs related to exhaust 
emissions were also subject to peer review, as described in a set of 
peer review reports that are available in the docket for this 
rulemaking.
    Cost estimates based on the current projected costs for our 
estimated technology packages represent an expected incremental cost of 
equipment in the near term. For the longer term we have identified 
factors that would cause cost impacts to decrease over time. First, as 
noted above, we project that manufacturers will spread their fixed 
costs over the first five years of production. After the fifth year of 
production, we project that the fixed costs would be retired and the 
unit costs could be reduced as a result.
    The cost analysis considers both long-term and short-term costs. We 
expect that over time, manufacturers will undergo a learning process 
that will lead to lower variable costs. For instance, the analysis 
incorporates the expectation that Small SI engine manufacturers will 
optimize the catalyst muffler offerings available and thereby 
streamline their production and reduce costs. The cost analysis 
generally incorporates this learning effect by decreasing estimated 
variable costs by 20 percent starting in the sixth year of production. 
Long-term impacts on costs are expected to decrease as manufacturers 
fully amortize their fixed costs and learn to optimize their designs 
and production processes to meet the standards more efficiently. The 
learning curve has not been applied to Small SI EFI systems due to the 
fact that the technologies are currently well established on similar 
sized engines in other applications.
    We project average costs to comply with the proposed exhaust 
emission standards for Small SI engines and equipment to range from $9-
$15 per Class I equipment to meet the Phase 3 standards. We anticipate 
the manufacturers will meet the emission standard with several 
technologies including engine improvements and catalysts. For Class II 
equipment, we project average costs to range from $22-$47 per equipment 
to meet the proposed emission standards. We anticipate the 
manufacturers of Class II engines would meet the proposed exhaust 
emission standards by engine improvements and adding catalysts and/or 
electronic fuel injection to their engines.
    For Small SI equipment, we have also estimated a per-unit cost for 
the proposed evaporative emission standards. The average short-term 
costs without fuel savings are projected to be $0.82 for handheld 
equipment, $3.16 for Class I equipment, and $6.90 for Class II 
equipment. These costs are based on fuel tank and fuel line permeation 
control, and for non-handheld equipment, running loss and diffusion

[[Page 28222]]

control. Because evaporative emissions are composed of otherwise usable 
fuel that is lost to the atmosphere, measures that reduce evaporative 
emissions will result in fuel savings. We estimate that the average 
fuel savings, due to permeation control, be about 1.2 gallons over the 
5-year average operating lifetime. This translates to a discounted 
lifetime savings of more than $2 at an average fuel price of $1.81 per 
gallon.
    For marine engines, we estimated per-engine costs for OB, PWC, and 
SD/I engines for meeting the proposed exhaust emission standards. The 
short-term cost estimates without fuel savings are $280 for OB, $360 
for PWC, and $360 for SD/I engines. For OB/PWC engines, we anticipate 
that manufacturers would meet the standards through the expanded 
production of existing low-emission technologies such as four-stroke 
and direct-injection two-stroke engines. For SD/I engines, we 
anticipate that manufacturers would use catalytic control to meet the 
proposed standards.
    For marine vessels, we have also estimated a per-unit cost for the 
proposed evaporative emission standards. The average short-term costs 
without fuel savings are projected to be $12 for boats with portable 
fuel tanks, $17 for PWC, and $74 for boats with installed fuel tanks. 
These costs are based on fuel tank and fuel line permeation control and 
diurnal emission control. For portable fuel tanks, diurnal emission 
control is based on an automatic sealing vent, for PWC we estimate that 
changes will not be necessary from current designs, and for other boats 
with installed fuel tanks, the estimated costs are based on the use of 
a passively-purged carbon canister. Because evaporative emissions are 
composed of otherwise usable fuel that is lost to the atmosphere, 
measures that reduce evaporative emissions will result in fuel savings. 
We estimate that the average fuel savings, due to permeation control, 
be about 31 gallons over the 15-year average operating lifetime. This 
translates to a discounted lifetime savings of about $36 at an average 
fuel price of $1.81 per gallon.

C. Cost per Ton

    We have calculated the cost per ton of the Phase 3 standards 
contained in this proposal by estimating costs and emission benefits 
for these engines. We made our best estimates of the combination of 
technologies that engine manufacturers might use to meet the new 
standards, best estimates of resultant changes to equipment design, 
engine manufacturer compliance program costs, and fuel savings in order 
to assess the expected economic impact of the proposed Phase 3 emission 
standards for Small SI engines and Marine SI engines. Emission 
reduction benefits are taken from the results of the Inventory chapter 
of the RIA (Chapter 3).
    A summary of the annualized costs to Small SI and Marine SI engine 
manufacturers is presented in Table XII-6. These annualized costs are 
over a 30-year period and presented both with a 3-percent and a 7-
percent discount rate. The annualized fuel savings for Small SI engines 
are due to reduced fuel costs from the use of electronic fuel injection 
on Class II engines as well as fuel savings from evaporative measures 
on all Small SI engines. The annualized fuel savings for Marine SI 
engines are due to reduced fuel costs from the expected elimination of 
2-stroke outboard motors from the new engine fleet as well as fuel 
savings from evaporative emission controls on all vessels.

  Table XII-6.--Estimated Annualized Cost to Manufacturers and Annualized Fuel Savings over 30 Years Due to the
                                 Phase 3 Small SI and Marine SI Engine Standards
                                     [2005$, 3 and 7 percent discount rates]
----------------------------------------------------------------------------------------------------------------
                                                                 Annualized cost to      Annualized fuel savings
                                                               manufactuers (millions/        (millions/yr)
           Engine category              Emissions category               yr)           -------------------------
                                                             --------------------------
                                                                   3%           7%           3%           7%
----------------------------------------------------------------------------------------------------------------
Small SI Engines....................  Exhaust...............         $281         $267          $71          $63
                                      Evaporative...........           70           67           58           52
                                      Aggregate.............          350          334          129          114
Marine SI Engines...................  Exhaust...............          134          141           76           67
                                      Evaporative...........           26           26           29           25
                                      Aggregate.............          160          167          105           92
----------------------------------------------------------------------------------------------------------------

    We have estimated the Small SI and Marine SI engine cost per ton of 
the Phase 3 HC+NOX standards over the typical lifetime of 
the equipment that are covered by this proposal. We have examined the 
cost per ton by performing a nationwide cost per ton analysis in which 
the net present value of the cost of compliance per year is divided by 
the net present value of the HC+NOX benefits over 30 years. 
The resultant discounted cost per ton is presented in Table XII-7. The 
total (exhaust and evaporative) cost per ton, using a 7 percent 
discount rate, with fuel savings is $950 for Small SI equipment and 
$350 for marine vessels. For the proposal as a whole, the cost per ton 
of HC+NOX reduction is $660. Reduced operating costs offset 
a portion of the increased cost of producing the cleaner Small SI and 
Marine SI engines. Reduced fuel consumption also offsets the costs of 
permeation control. Chapter 7 of the RIA contains a more detailed 
discussion of the cost per ton analysis.

                      Table XII-7.--Estimated Cost Per Ton of the HC+NOX Emission Standards
                                     [2005$, 3 and 7 percent discount rates]
----------------------------------------------------------------------------------------------------------------
                                                                                      Discounted cost per ton
                                                                                 -------------------------------
                            Category                              Implementation   Without fuel      With fuel
                                                                       dates       savings  (3%/   savings  (3%/
                                                                                        7%)             7%)
----------------------------------------------------------------------------------------------------------------
Small SI Exhaust................................................       2011-2012     $1700/$1860     $1270/$1420

[[Page 28223]]

 
Small SI Evaporative............................................       2008-2013         720/770         120/170
Marine SI Exhaust...............................................       2009-2013         690/820         300/430
Marine SI Evaporative...........................................       2009-2012         530/630         (70)/35
Aggregate.......................................................       2008-2013        660/1120         226/660
----------------------------------------------------------------------------------------------------------------

    As is discussed above, we are also expecting some reduction in 
direct PM emissions and carbon monoxide. These reductions will come 
primarily as product of the technology being used to meet HC and 
NOX standards and not directly as a result of the 
implementation of specific technology to achieve these gains. Thus, we 
have elected to focus our cost per ton analysis on HC+NOX.
    One useful purpose of cost per ton analysis is to compare this 
program to other programs designed to achieve similar air quality 
objectives. Toward that end, we made a comparison between the 
HC+NOX cost per ton values presented in Table C-2 and the 
HC+NOX cost per ton of other recent mobile source programs. 
Table XII-8 summarizes the HC+NOX cost per ton of several 
recent EPA actions for controlled emissions from mobile sources. While 
the analyses for each rule were not completely identical, it is clear 
that the Small SI and Marine SI values compare favorably with the other 
recent actions.

   Table XII-8.--Cost Per Ton of Previously Implemented HC+NOX Mobile
                             Source Programs
              [2005$, 7 percent discount with fuel savings]
------------------------------------------------------------------------
                                                            Discounted
                         Program                           cost per ton
------------------------------------------------------------------------
2002 HH engines Phase 2.................................             840
2001 NHH engines Phase 2................................           * neg
1998 Marine SI engines..................................            1900
2004 Comm Marine CI.....................................             200
2007 Large SI exhaust...................................              80
2006 ATV exhaust........................................             300
2006 Off-highway motorcycle.............................             290
2006 Recreational marine CI.............................             700
2010 Snowmobile.........................................            1430
2006 <50cc highway motorcycle...........................            1860
2010 Class 3 highway motorcycle.........................            1650
------------------------------------------------------------------------
* Fuel savings outweigh engineering/hardware costs.

D. Air Quality Impact

    Information on the air quality impacts of this proposed action can 
be found in Section II of this preamble. Section II includes health 
effect information on ozone, PM, CO and air toxics. It also includes 
modeled projections of future ozone concentrations with and without the 
controls detailed in this proposal. The proposed emission reductions 
would lead to reductions in ambient concentrations of ozone, PM, CO and 
air toxics.

E. Benefits

    This section presents our analysis of the health and environmental 
benefits that can be expected to occur as a result of the proposed 
Small SI and Marine SI engine standards throughout the period from 
initial implementation through 2030. Nationwide, the engines that are 
subject to the proposed emission standards in this rule are a 
significant source of mobile source air pollution. The proposed 
standards would reduce exposure to hydrocarbon, CO and NOX 
emissions and help avoid a range of adverse health effects associated 
with ambient ozone and PM2.5 levels. In addition, the 
proposed standards would help reduce exposure to CO, air toxics, and 
PM2.5 for persons who operate or who work with or are 
otherwise active in close proximity to these engines.
    EPA typically quantifies PM- and ozone-related benefits in its 
regulatory impact analyses (RIAs) when possible. In the analysis of 
past air quality regulations, ozone-related benefits have included 
morbidity endpoints and welfare effects such as damage to commercial 
crops. EPA has not recently included a separate and additive mortality 
effect for ozone, independent of the effect associated with fine 
particulate matter. For a number of reasons, including (1) Advice from 
the Science Advisory Board (SAB) Health and Ecological Effects 
Subcommittee (HEES) that EPA consider the plausibility and viability of 
including an estimate of premature mortality associated with short-term 
ozone exposure in its benefits analyses and (2) conclusions regarding 
the scientific support for such relationships in EPA's 2006 Air Quality 
Criteria for Ozone and Related Photochemical Oxidants (the CD), EPA is 
in the process of determining how to appropriately characterize ozone-
related mortality benefits within the context of benefits analyses for 
air quality regulations. As part of this process, we are seeking advice 
from the National Academy of Sciences (NAS) regarding how the ozone-
mortality literature should be used to quantify the reduction in 
premature mortality due to diminished exposure to ozone, the amount of 
life expectancy to be added and the

[[Page 28224]]

monetary value of this increased life expectancy in the context of 
health benefits analyses associated with regulatory assessments. In 
addition, the agency has sought advice on characterizing and 
communicating the uncertainty associated with each of these aspects in 
health benefit analyses.
    Since the NAS effort is not expected to conclude until 2008, the 
agency is currently deliberating how best to characterize ozone-related 
mortality benefits in its rulemaking analyses in the interim. For the 
analysis of the proposed standards, we do not quantify an ozone 
mortality benefit. So that we do not provide an incomplete picture of 
all of the benefits associated with reductions in emissions of ozone 
precursors, we have chosen not to include an estimate of total ozone 
benefits in the proposed RIA. By omitting ozone benefits in this 
proposal, we acknowledge that this analysis underestimates the benefits 
associated with the proposed standards. Our analysis, however, 
indicates that the rule's monetized PM2.5 benefits alone 
substantially exceed our estimate of the costs.
    The PM2.5 benefits are scaled based on relative changes 
in PM2.5 precursor emissions (direct PM and NOX) 
between this rule and the proposed Clean Air Nonroad Diesel (CAND) 
rule. As explained in Section 8.2.1 of the RIA for this rule, the 
PM2.5 benefits scaling approach is limited to those studies, 
health impacts, and assumptions that were used in the proposed CAND 
analysis. As a result, PM-related premature mortality is based on the 
updated analysis of the American Cancer Society cohort (ACS; Pope et 
al., 2002).\107\ However, it is important to note that since the CAND 
rule, EPA's Office of Air and Radiation (OAR) has adopted a different 
format for its benefits analyses in which characterization of the 
uncertainty in the concentration-response function is integrated into 
the main benefits analysis. This new approach follows the 
recommendation of NRC's 2002 report ``Estimating the Public Health 
Benefits of Proposed Air Pollution Regulations'' to begin moving the 
assessment of uncertainties from its ancillary analyses into its main 
benefits presentation through the conduct of probabilistic 
analyses.\108\ Within this context, additional data sources are 
available, including a recent expert elicitation and updated analysis 
of the Six-Cities Study cohort (Laden et al., 2006).\109\ Please see 
the PM NAAQS RIA for an indication of the sensitivity of our results to 
use of alternative concentration-response functions. The 
PM2.5-related benefits associated with the proposed 
standards are presented in table XII-9.
---------------------------------------------------------------------------

    \107\ Pope, C.A., III, R.T. Burnett, M.J. Thun, E.E. Calle, D. 
Krewski, K. Ito, and G.D. Thurston. 2002. ``Lung Cancer, 
Cardiopulmonary Mortality, and Long-term Exposure to Fine 
Particulate Air Pollution.'' Journal of the American Medical 
Association 287:1132-1141.
    \108\ National Research Council (NRC). 2002. Estimating the 
Public Health Benefits of Proposed Air Pollution Regulations. 
Washington, DC: The National Academies Press.
    \109\ Laden, F., J. Schwartz, F.E. Speizer, and D.W. Dockery. 
2006. Reduction in Fine Particulate Air Pollution and Mortality. 
American Journal of Respiratory and Critical Care Medicine. 173: 
667-672.
---------------------------------------------------------------------------

    It should be noted that since the CAND rule, EPA's Office of Air 
and Radiation (OAR) has adopted a different format for its benefits 
analysis in which characterization of uncertainty is integrated into 
the main benefits analysis. The benefits scaling approach used in the 
analysis of the proposed standards limits our ability to integrate 
uncertainty into the main analysis. For the benefits analysis of the 
final standards, we will adopt this integrated uncertainty approach. 
Please see the PM NAAQS RIA for an indication of the uncertainty 
present in the base estimate of benefits and the sensitivity of our 
results to the use of alternative concentration-response functions.

   Table XII-9.--Estimated Monetized PM-Related Health Benefits of the
                           Proposed Standards
------------------------------------------------------------------------
                                              Total Benefits a, b, c
                                                 (billions 2005$)
                                         -------------------------------
                                               2020            2030
------------------------------------------------------------------------
Using a 3% discount rate................        $2.1 + B        $3.4 + B
Using a 7% discount rate................        $1.9 + B       $3.1 + B
------------------------------------------------------------------------
\a\ Benefits include avoided cases of mortality, chronic illness, and
  other morbidity health endpoints. PM-related mortality benefits
  estimated using an assumed PM threshold at background levels (3 [mu]g/
  m3). There is uncertainty about which assumed threshold to use and
  this may impact the magnitude of the total benefits estimate. For a
  more detailed discussion of this issue, please refer to Section
  8.6.2.2 of the RIA.
\b\ For notational purposes, unquantified benefits are indicated with a
  ``B'' to represent the sum of additional monetary benefits and
  disbenefits. A detailed listing of unquantified health and welfare
  effects is provided in Table XII-12.
\c\ Results reflect the use of two different discount rates: 3 and 7
  percent, which are recommended by EPA's Guidelines for Preparing
  Economic Analyses\110\ and OMB Circular A-4.\111\ Results are rounded
  to two significant digits for ease of presentation and computation.

(1) Quantified Human Health and Environmental Effects of the Proposed 
Standards
---------------------------------------------------------------------------

    \110\ U.S. Environmental Protection Agency. September 2000. 
Guidelines for Preparing Economic Analyses. EPA 240-R-00-003.
    \111\ U.S. Office of Management and Budget (OMB). 2003. Circular 
A-4 Guidance for Federal Agencies Preparing Regulatory Analyses, 
Available at: http://www/whitehouse.gov/omb/inforeg/iraguide.html. 
Accessed December 15, 2005.
---------------------------------------------------------------------------

    In this section we discuss the PM2.5 benefits of the 
proposed standards. To estimate PM2.5 benefits, we rely on a 
benefits transfer technique. The benefits transfer approach uses as its 
foundation the relationship between reductions in precursors to 
PM2.5 (NOX and direct PM2.5 emissions) 
and ambient PM2.5 concentrations modeled across the 
contiguous 48 states (and DC) for the Clean Air Nonroad Diesel (CAND) 
proposal.\112\ For a given future year, we first calculate the ratio 
between CAND direct PM2.5 emission reductions and direct 
PM2.5 emission reductions associated with the proposed 
control standards (proposed emission reductions/CAND emission 
reductions). We calculate a similar ratio for NOX. We then 
multiply these ratios by the percent that direct PM2.5 and 
NOX emissions, respectively, contribute towards population-
weighted reductions in ambient PM2.5 due to the CAND 
standards. This calculation results in a ``benefits apportionment 
factor'' for the relationship between direct PM emissions and ambient 
PM2.5 and NOX emissions and ambient 
PM2.5, which are then applied to the incidence and monetized 
benefits from the CAND proposal. In this way, we apportion the results 
of the proposed CAND analysis to its underlying PM precursor emission 
reductions and scale the apportioned

[[Page 28225]]

benefits to reflect differences in emission reductions between the two 
rules.\113\ This benefits transfer method is consistent with the 
approach used in other recent mobile and stationary source rules.\114\
---------------------------------------------------------------------------

    \112\ See 68 FR 28327, May 23, 2003.
    \113\ Note that while the proposed regulations control 
hydrocarbons (VOCs), which contribute to PM formation, the benefits 
transfer scaling approach only scales benefits based on 
NOX, SO2, and direct PM emission reductions. 
PM benefits will likely be underestimated as a result, though we are 
unable to estimate the magnitude of the underestimation. Note also 
that PM-related mortality benefits estimated for the CAND analysis 
used an assumed PM threshold at background levels (3 [mu]g/
m3). There is uncertainty about which threshold to use 
and this may impact the magnitude of the total benefits estimate. 
For a more detailed discussion of this issue, please refer to 
Chapter 8.2 of the RIA.
    \114\ See: Mobile Source Air Toxics proposed rule (71 FR 15803, 
March 29, 2006); Clean Air Nonroad Diesel final rule (69 FR 38958, 
June 29, 2004); Nonroad Large Spark-Ignition Engines and 
Recreational Engines standards (67 FR 68241, November 8, 2002); 
Final Industrial Boilers and Process Heaters NESHAP (69 FR 55217, 
September 13, 2004); Final Reciprocating Internal Combustion Engines 
NESHAP (69 FR 33473, June 15, 2004); Final Clean Air Visibility Rule 
(EPA-452/R-05-004, June 15, 2005); Ozone Implementation Rule 
(documentation forthcoming).
---------------------------------------------------------------------------

    Table XII-10 presents the primary estimates of reduced incidence of 
PM-related health effects for the years 2020 and 2030 for the proposed 
emission control strategy.\115\ In 2030, we estimate that PM-related 
annual benefits include approximately 450 fewer premature fatalities, 
290 fewer cases of chronic bronchitis, 800 fewer non-fatal heart 
attacks, 460 fewer hospitalizations (for respiratory and cardiovascular 
disease combined), 310,000 days of restricted activity due to 
respiratory illness and approximately 52,000 fewer work-loss days. We 
also estimate substantial health improvements for children from reduced 
upper and lower respiratory illness, acute bronchitis, and asthma 
attacks.
---------------------------------------------------------------------------

    \115\ The ``primary estimate'' refers to the estimate of 
benefits that reflects the suite of endpoints and assumptions that 
EPA believes yields the expected value of air quality improvements 
related to the proposed standards. The impact that alternative 
endpoints and assumptions have on the benefit estimates are explored 
in appendixes to the RIA.

    Table XII-10.--Estimated Annual Reductions in Incidence of Health
                                Effects a
------------------------------------------------------------------------
                                            2020 annual     2030 annual
              Health effect                  incidence       incidence
                                             reduction       reduction
------------------------------------------------------------------------
PM-Related Endpoints:
    Premature Mortality \b\--
    Adult, age 30 and over plus Infant,              290             450
     age < 1 year.......................
    Chronic bronchitis (adult, age 26                200             290
     and over)..........................
    Non-fatal myocardial infarction                  490             800
     (adult, age 18 and over)...........
    Hospital admissions--respiratory                 160             270
     (all ages) \c\.....................
    Hospital admissions--cardiovascular              130             200
     (adults, age > 18) \d\.............
    Emergency room visits for asthma                 210             310
     (age 18 years and younger).........
    Acute bronchitis, (children, age 8-              470             700
     12)................................
    Lower respiratory symptoms                     5,600           8,300
     (children, age 7-14)...............
    Upper respiratory symptoms                     4,300           6,300
     (asthmatic children, age 9-18).....
    Asthma exacerbation (asthmatic                 7,000          10,000
     children, age 6-18)................
    Work loss days......................          38,000          52,000
    Minor restricted activity days               220,000         310,000
     (adults age 18-65).................
------------------------------------------------------------------------
\a\ Incidence is rounded to two significant digits. The PM estimates
  represent benefits from the proposed rule nationwide. The ozone
  estimates only represent benefits from the Eastern 37 states and DC,
  though the program is national in scope.
\b\ PM-related adult mortality based upon studies by Pope, et al
  2002.\116\ PM-related infant mortality based upon studies by Woodruff,
  Grillo, and Schoendorf,1997.\117\
\c\ Respiratory hospital admissions for PM include admissions for
  chronic obstructive pulmonary disease (COPD), pneumonia and asthma.
\d\ Cardiovascular hospital admissions for PM include total
  cardiovascular and subcategories for ischemic heart disease,
  dysrhythmias, and heart failure.

(2) Monetized Benefits
    Table XII-11 presents the estimated monetary value of reductions in 
the incidence of health and welfare effects. Annual PM-related health 
benefits are approximately $3.4 billion in 2030, assuming a 3 percent 
discount rate (or $3.1 billion assuming a 7 percent discount rate). All 
monetized estimates are stated in 2005 dollars. These estimates account 
for growth in real gross domestic product (GDP) per capita between the 
present and the years 2020 and 2030. As the table indicates, total 
benefits are driven primarily by the reduction in premature fatalities 
each year, which accounts for well over 90 percent of total benefits.
---------------------------------------------------------------------------

    \116\ Pope, C.A., III, R.T. Burnett, M.J. Thun, E.E. Calle, D. 
Krewski, K. Ito, and G.D. Thurston. 2002. ``Lung Cancer, 
Cardiopulmonary Mortality, and Long-term Exposure to Fine 
Particulate Air Pollution.'' Journal of American Medical Association 
287:1132-1141.
    \117\ Woodruff, T.J., J. Grillo, and K.C. Schoendorf. 1997. 
``The Relationship Between Selected Causes of Postneonatal Infant 
Mortality and Particulate Infant Mortality and Particulate Air 
Pollution in the United States.'' Environmental Health Perspectives 
105(6):608-612.
---------------------------------------------------------------------------

    Table XII-11 indicates with a ``B'' those additional health and 
environmental benefits of the rule that we were unable to quantify or 
monetize. These effects are additive to the estimate of total benefits, 
and are related to the following sources:
     There are many human health and welfare effects associated 
with ozone, PM, and toxic air pollutant reductions that remain 
unquantified because of current limitations in the methods or available 
data. A full appreciation of the overall economic consequences of the 
proposed standards requires consideration of all benefits and costs 
expected to result from the new standards, not just those benefits and 
costs which could be expressed here in dollar terms. A listing of the 
benefit categories that could not be quantified or monetized in our 
benefit estimates are provided in Table XII-12.
     The PM air quality model only captures the benefits of air 
quality improvements in the 48 states and DC; PM benefits for Alaska 
and Hawaii are not reflected in the estimate of benefits.
---------------------------------------------------------------------------

    \118\ U.S. Environmental Protection Agency, 2000. Guidelines for 
Preparing Economic Analyses. www.yosemite1.epa.gov/ee/epa/eed/hsf/pages/Guideline.html.
    \119\ Office of Management and Budget, The Executive Office of 
the President, 2003. Circular A-4. http://www.whitehouse.gov/omb/circulars.

[[Page 28226]]



 Table XII-11.--Estimated Annual Monetary Value of Reductions in Incidence of Health and Welfare Effects (2005$)
                                                      a, b
----------------------------------------------------------------------------------------------------------------
                                                                                  2020 estimated  2030 estimated
                                                                                     value of        value of
               Health effect                              Pollutant                 reductions      reductions
                                                                                    (millions)      (millions)
----------------------------------------------------------------------------------------------------------------
PM-Related Premature mortality c, d
    Adult >30 years........................  PM2.5..............................
        3 percent discount rate............  ...................................          $2,000          $3,100
        7 percent discount rate............  ...................................           1,800           2,800
    Child <1 year..........................  ...................................               5               6
Chronic bronchitis (adults, 26 and over)...  PM2.5..............................              90             140
Non-fatal acute myocardial infarctions
    3 percent discount rate................  ...................................              50              77
    7 percent discount rate................  PM2.5..............................              48              75
Hospital admissions for respiratory causes.  PM2.5..............................             2.9             5.0
Hospital admissions for cardiovascular       PM2.5..............................             3.1             4.7
 causes.
Emergency room visits for asthma...........  PM2.5..............................            0.07            0.11
Acute bronchitis (children, age 8-12)......  PM2.5..............................            0.20            0.30
Lower respiratory symptoms (children, age 7- PM2.5..............................            0.11            0.16
 14).
Upper respiratory symptoms (asthma, age 9-   PM2.5..............................            0.13            0.19
 11).
Asthma exacerbations.......................  PM2.5..............................            0.36            0.54
Work loss days.............................  PM2.5..............................             5.8             7.0
Minor restricted activity days (MRADs).....  PM2.5..............................              14              19
Monetized Total \e\
    Base estimate:
        3 percent discount rate............  PM2.5..............................       2,100 + B       3,400 + B
        7 percent discount rate............  ...................................       1,900 + B      3,100 + B
----------------------------------------------------------------------------------------------------------------
\a\ Incidence is rounded to two significant digits. The PM estimates represent benefits from the proposed rule
  nationwide.
\b\ Monetary benefits adjusted to account for growth in real GDP per capita between 1990 and the analysis year
  (2020 or 2030).
\c\ Valuation of premature mortality based on long-term PM exposure assumes discounting over the SAB recommended
  20 year segmented lag structure described in the Regulatory Impact Analysis for the Final Clean Air Interstate
  Rule (March 2005). Results show 3 percent and 7 percent discount rates consistent with EPA and OMB guidelines
  for preparing economic analyses (US EPA, 2000 and OMB, 2003).118, 119
\d\ Adult mortality based upon the ACS cohort study (Pope et al., 2002). Infant mortality based upon studies by
  Woodruff, Grillo, and Schoendorf, 1997.
\e\ B represents the monetary value of health and welfare benefits not monetized. A detailed listing is provided
  in Table XII-12.


  Table XII-12.--Unquantified and Non-Monetized Effects of the Proposed
          Small Spark Ignition/Recreational Marine Engine Rule
------------------------------------------------------------------------
                                       Effects not included in primary
         Pollutant/effects                 estimates--changes in:
------------------------------------------------------------------------
Ozone Health a....................  Premature mortality: short-term
                                     exposures b.
                                    Hospital admissions: respiratory.
                                    Emergency room visits for asthma.
                                    Minor restricted-activity days.
                                    School loss days.
                                    Asthma attacks.
                                    Cardiovascular emergency room
                                     visits.
                                    Acute respiratory symptoms.
                                    Chronic respiratory damage.
                                    Premature aging of the lungs.
                                    Non-asthma respiratory emergency
                                     room visits.
                                    Increased exposure to UVb.
Ozone Welfare.....................  Yields for
                                       --commercial forests.
                                       --some fruits and vegetables.
                                       --non-commercial crops.
                                    Damage to urban ornamental plants.
                                    Impacts on recreational demand from
                                     damaged forest aesthetics.
                                    Ecosystem functions.
                                    Increased exposure to UVb.
PM Health c.......................  Premature mortality--short term
                                     exposures d.
                                    Low birth weight.
                                    Pulmonary function.
                                    Chronic respiratory diseases other
                                     than chronic bronchitis.
                                    Non-asthma respiratory emergency
                                     room visits.
                                    Exposure to UVb ()e.
PM Welfare........................  Visibility in Class I areas.
                                    Residential and recreational
                                     visibility in non-Class I areas.
                                    Soiling and materials damage.
                                    Damage to ecosystem functions.
                                    Exposure to UVb () e.

[[Page 28227]]

 
Nitrogen and Sulfate Deposition     Commercial forests due to acidic
 Welfare.                            sulfate and nitrate deposition.
                                    Commercial freshwater fishing due to
                                     acidic deposition.
                                    Recreation in terrestrial ecosystems
                                     due to acidic deposition.
                                    Existence values for currently
                                     healthy ecosystems.
                                    Commercial fishing, agriculture, and
                                     forests due to nitrogen deposition.
                                    Recreation in estuarine ecosystems
                                     due to nitrogen deposition.
                                    Ecosystem functions.
                                    Passive fertilization.
CO Health.........................  Behavioral effects.
HC Health f.......................  Cancer (benzene, 1,3-butadiene,
                                     formaldehyde, acetaldehyde).
                                    Anemia (benzene).
                                    Disruption of production of blood
                                     components (benzene).
                                    Reduction in the number of blood
                                     platelets (benzene).
                                    Excessive bone marrow formation
                                     (benzene).
                                    Depression of lymphocyte counts
                                     (benzene).
                                    Reproductive and developmental
                                     effects (1,3-butadiene).
                                    Irritation of eyes and mucus
                                     membranes (formaldehyde).
                                    Respiratory irritation
                                     (formaldehyde).
                                    Asthma attacks in asthmatics
                                     (formaldehyde).
                                    Asthma-like symptoms in non-
                                     asthmatics (formaldehyde).
                                    Irritation of the eyes, skin, and
                                     respiratory tract (acetaldehyde).
                                    Upper respiratory tract irritation
                                     and congestion (acrolein).
HC Welfare........................  Direct toxic effects to animals.
                                    Bioaccumulation in the food chain.
                                    Damage to ecosystem function.
                                    Odor.
------------------------------------------------------------------------
a In addition to primary economic endpoints, there are a number of
  biological responses that have been associated with ozone health
  effects including increased airway responsiveness to stimuli,
  inflammation in the lung, acute inflammation and respiratory cell
  damage, and increased susceptibility to respiratory infection. The
  public health impact of these biological responses may be partly
  represented by our quantified endpoints.
b Recent analyses provide evidence that short-term ozone exposure is
  associated with increased premature mortality. As a result, EPA is
  considering how to incorporate ozone mortality benefits into its
  benefits analyses as a separate estimate of the number of premature
  deaths that would be avoided due to reductions in ozone levels.
c In addition to primary economic endpoints, there are a number of
  biological responses that have been associated with PM health effects
  including morphological changes and altered host defense mechanisms.
  The public health impact of these biological responses may be partly
  represented by our quantified endpoints.
d While some of the effects of short-term exposures are likely to be
  captured in the estimates, there may be premature mortality due to
  short-term exposure to PM not captured in the cohort study upon which
  the primary analysis is based.
e May result in benefits or disbenefits.
f Many of the key hydrocarbons related to this rule are also hazardous
  air pollutants listed in the Clean Air Act.

(3) What Are the Significant Limitations of the Benefits Analysis?
    Every benefit-cost analysis examining the potential effects of a 
change in environmental protection requirements is limited to some 
extent by data gaps, limitations in model capabilities (such as 
geographic coverage), and uncertainties in the underlying scientific 
and economic studies used to configure the benefit and cost models. 
Deficiencies in the scientific literature often result in the inability 
to estimate quantitative changes in health and environmental effects, 
such as potential increases in premature mortality associated with 
increased exposure to carbon monoxide. Deficiencies in the economics 
literature often result in the inability to assign economic values even 
to those health and environmental outcomes which can be quantified. 
These general uncertainties in the underlying scientific and economics 
literature, which can cause the valuations to be higher or lower, are 
discussed in detail in the RIA and its supporting references. Key 
uncertainties that have a bearing on the results of the benefit-cost 
analysis of the proposed standards include the following:
     The exclusion of potentially significant and unquantified 
benefit categories (such as health, odor, and ecological benefits of 
reduction in ozone, air toxics, and PM);
     Errors in measurement and projection for variables such as 
population growth;
     Uncertainties in the estimation of future year emissions 
inventories and air quality, especially regarding the discrepancy 
between the modeled and proposed suite of standards and their impact on 
emissions inventories;
     Uncertainties associated with the scaling of the PM 
results of the modeled benefits analysis to the proposed standards, 
especially regarding the assumption of similarity in geographic 
distribution between emissions and human populations and years of 
analysis;
     Uncertainty in the estimated relationships of health and 
welfare effects to changes in pollutant concentrations including the 
shape of the concentration-response function, the size of the effect 
estimates, and the relative toxicity of the many components of the PM 
mixture;
     Uncertainties in exposure estimation; and
     Uncertainties associated with the effect of potential 
future actions to limit emissions.
    As Table XII-11 indicates, total benefits are driven primarily by 
the reduction in premature fatalities each year. Elaborating on the 
list of uncertainties above, some key assumptions underlying the 
primary estimate for the premature mortality category include the 
following:

[[Page 28228]]

     Inhalation of fine particles is causally associated with 
premature death at concentrations near those experienced by most 
Americans on a daily basis. Although biological mechanisms for this 
effect have not yet been completely established, the weight of the 
available epidemiological, toxicological, and experimental evidence 
supports an assumption of causality. The impacts of including a 
probabilistic representation of causality were explored in the expert 
elicitation-based results of the recently published PM NAAQS RIA. 
Because the analysis of the proposed standards is constrained to the 
studies included in the CAND PM benefits scaling approach, we are 
unable to conduct the same analysis of expert elicitation-based 
mortality incidence for the proposed standards.\120\ However, we 
qualitatively describe the expert elicitation-based mortality results 
associated with the final PM NAAQS to provide an indication of the 
sensitivity of our PM-related premature mortality results to use of 
alternative concentration-response functions. We present this 
discussion in the RIA.
---------------------------------------------------------------------------

    \120\ The scaling approach relies on the incidence and valuation 
estimates derived from the studies available at the time of the CAND 
analysis. Incidence estimates and monetized benefits derived from 
new information, including mortality derived from the full expert 
elicitation, are not available for scaling. Please refer to section 
2 of this preamble and Chapter 12 of the RIA for more information 
about the benefits scaling approach.
---------------------------------------------------------------------------

     Since the publication of CAIR, a follow up to the Harvard 
six-city study on premature mortality was published (Laden et al., 2006 
based on Dockery et al., 1993),121 122 which both confirmed 
the effect size from the first study and provided additional evidence 
that reductions in PM2.5 directly result in reductions in 
the risk of premature death. The impacts of including this study in the 
primary analysis were explored in the results of the recently published 
PM NAAQS RIA. Because the analysis of the proposed standards is 
constrained to the studies included in the CAND PM benefits scaling 
approach, we are unable to characterize PM-related mortality based on 
Laden et al. However, we discuss the implications of these results in 
the RIA for the proposed standards.
---------------------------------------------------------------------------

    \121\ Laden, F., J. Schwartz, F.E. Speizer, and D.W. Dockery. 
2006. Reduction in Fine Particulate Air Pollution and Mortality. 
American Journal of Respiratory and Critical Care Medicine. 173: 
667-672.
    \122\ Dockery, D.W., C.A. Pope, X.P. Xu, J.D. Spengler, J.H. 
Ware, M.E. Fay, B.G. Ferris, and F.E. Speizer. 1993. ``An 
Association between Air Pollution and Mortality in Six U.S. 
Cities.'' New England Journal of Medicine 329(24):1753-1759.
---------------------------------------------------------------------------

     All fine particles, regardless of their chemical 
composition, are equally potent in causing premature mortality. This is 
an important assumption, because PM produced via transported precursors 
emitted from Small SI and Marine SI engines may differ significantly 
from PM precursors released from electric generating units and other 
industrial sources. However, no clear scientific grounds exist for 
supporting differential effects estimates by particle type.
     The concentration-response function for fine particles is 
approximately linear within the range of ambient concentrations under 
consideration. Thus, the estimates include health benefits from 
reducing fine particles in areas with varied concentrations of PM, 
including both regions that may be in attainment with PM2.5 
standards and those that are at risk of not meeting the standards.
    Taking into account these uncertainties, we believe this benefit-
cost analysis provides a conservative estimate of the expected economic 
benefits of the proposed standards in future years because of the 
exclusion of potentially significant benefit categories. Acknowledging 
benefits omissions and uncertainties, we present a best estimate of the 
total benefits based on our interpretation of the best available 
scientific literature and methods. Furthermore, our analysis reflects 
many methodological improvements that were incorporated into the 
analysis of the final Clean Air Interstate Rule (CAIR), including a 
revised value of a statistical life, a revised baseline rate of future 
mortality, and a revised mortality lag assumption. Details of these 
improvements can be found in the RIA for this rule and in the final 
CAIR rule RIA.\123\ Once again, however, it should be noted that since 
the CAIR rule, EPA's Office of Air and Radiation (OAR) has adopted a 
different format for its benefits analysis in which characterization of 
uncertainty is integrated into the main benefits analysis. Please see 
the PM NAAQS RIA for an indication of the uncertainty present in the 
base estimate of benefits and the sensitivity of our results to the use 
of alternative concentration-response functions.
---------------------------------------------------------------------------

    \123\ See Chapter 4 of the Final Clean Air Interstate Rule RIA 
(http://www.epa.gov/cair) for a discussion of EPA's ongoing efforts 
to address the NAS recommendations in its regulatory analyses.
---------------------------------------------------------------------------

(4) How Do the Benefits Compare to the Costs of the Proposed Standards?
    The proposed rule establishes separate standards that reduce the 
evaporative and exhaust emissions from Small SI and Marine SI engines. 
A full appreciation of the overall economic consequences of these 
provisions requires consideration of the benefits and costs expected to 
result from each standard. Due to limitations in data availability and 
analytical methods, however, we are only able to present the benefits 
of the entire proposed rule in the aggregate for both PM2.5 
and ozone. There are also a number of health and environmental effects 
associated with the proposed standards that we were unable to quantify 
or monetize (see Table XII-12).
    Table XII-13 contains the estimates of monetized PM2.5-
related benefits of the proposed standards and estimated social welfare 
costs for each of the proposed control programs. The annual social 
welfare costs of all provisions of this proposed rule are described 
more fully in the next section. The results in Table XII-13 suggest 
that the 2020 and 2030 monetized benefits of the proposed standards are 
much greater than the expected social welfare costs. Specifically, the 
annual benefits of the program would be approximately $2.1 + B billion 
annually in 2020 using a three percent discount rate (or $1.9 + B 
billion using a seven percent discount rate), compared to estimated 
social welfare costs of approximately $252 million in that same year. 
The net benefits are expected to increase to $3.4 + B billion annually 
in 2030 using a three percent discount rate (or $3.1 + B billion using 
a seven percent discount rate), even as the social welfare costs of 
that program fall to $241 million.
    In Table XII-13, we present the costs and PM-related benefits 
related to each of the two broad engine classes regulated by the 
proposed standards: Small SI and Marine SI engines. Table XII-13 also 
presents the costs and PM-related benefits related to the specific 
engine classes regulated by the proposed standards: Small SI--Class I, 
Class II, and Handheld (HH); Marine SI--Sterndrive/Inboard (SD/I), and 
Outboard/Personal Water Craft (OB/PWC). Using the same PM scaling 
approach described in Chapter 8.2 of the RIA, we are able to split out 
the estimated PM benefits related to the different Small SI and Marine 
SI engine classes. One can see that in all cases, the PM benefits 
accrued by the engine classes are greater than the costs, even when 
fuel savings is not factored into the cost estimate. The benefit-to-
cost ratio would be even greater if we

[[Page 28229]]

estimated the ozone benefits related to the proposed standards.

  Table XII-13.--Summary of Annual Benefits, Costs, and Net Benefits of
            the Proposed Small SI and Marine SI Engine Rule a
------------------------------------------------------------------------
                                       2020  (Millions   2030  (Millions
             Description                  of  2005          of  2005
                                          dollars)          dollars)
------------------------------------------------------------------------
Estimated Social Welfare Costs b c
    Small SI........................             $351              $404
        Class I.....................              145               167
        Class II....................              199               229
        HH d........................                7                 8
    Marine SI.......................              154               164
        SD/I........................               41                44
        OB/PWC......................              113               120
            Total...................              505               569
            Fuel Savings............             (253)             (327)
Total Social Welfare Costs..........              252               241
Estimated Benefits e f
    PM-Only Small SI Benefits
        3 percent discount rate.....              861             1,280
        7 percent discount rate.....              782             1,160
            Class I
                3 percent discount                478               647
                 rate...............
                7 percent discount                434               587
                 rate...............
            Class II
                3 percent discount                383               627
                 rate...............
                7 percent discount                348               570
                 rate...............
    PM-Only Marine SI Benefits
        3 percent discount rate.....            1,280             2,110
        7 percent discount rate.....            1,160             1,190
            SD/I
        3 percent discount rate.....              209               487
        7 percent discount rate.....              190               442
            OB/PWC
        3 percent discount rate.....            1,070             1,620
        7 percent discount rate.....              969             1,470
Total PM-Only Benefits g
        3 percent discount rate.....          2,140+B           3,380+B
        7 percent discount rate.....          1,940+B           3,070+B
Annual Net PM-Only Benefits (Total
 Benefits-Total Costs) g
        3 percent discount rate.....          1,890+B           3,140+B
        7 percent discount rate.....          1,690+B          2,830+B
------------------------------------------------------------------------
\a\ All estimates are rounded to three significant digits and represent
  annualized benefits and costs anticipated for the years 2020 and 2030.
  Columnar totals may not sum due to rounding.
\b\ Note that costs are the annual total costs of reducing all
  pollutants associated with each provision of the proposed control
  package, while the benefits reflect the value of reductions in PM2.5
  only.
\c\ To calculate annual fixed costs, we use a 7 percent average before-
  tax rate of return on private capital (see Chapter 9). We do not
  present annual costs using an alternative rate of return. In Chapter
  9, however, we use both a 3 percent and 7 percent social discount rate
  to calculate the net present value of total social costs consistent
  with EPA and OMB guidelines for preparing economic analyses (US EPA,
  2000 and OMB, 2003).124 125
\d\ Handheld emission reductions associated with the proposed standards,
  volatile organic hydrocarbons, are not accounted for in the PM
  benefits scaling approach. The PM benefit scaling approach is based
  upon changes in NOX and direct PM2.5 (see section 8.2 of the RIA). We
  therefore do not estimate any PM-related benefits associated with
  emission reductions in the handheld engine class.
\e\ PM-related benefits in this table are nationwide.
\f\ Valuation of premature mortality based on long-term PM exposure
  assumes discounting over the SAB recommended 20-year segmented lag
  structure described in section 8.3 of the RIA. Valuation of non-fatal
  myocardial infarctions is based on the cost-of-illness over a 5-year
  period after the incident. The valuation of both endpoints therefore
  requires the use of a discount rate. We present the PM-related
  benefits results using a 3 percent and 7 percent social discount rate
  consistent with EPA and OMB guidelines for preparing economic analyses
  (US EPA, 2000 and OMB, 2003).
\g\ Not all possible benefits or disbenefits are quantified and
  monetized in this analysis. B is the sum of all unquantified benefits
  and disbenefits. Potential benefit categories that have not been
  quantified and monetized are listed in Table XII-12.

F. Economic Impact Analysis
---------------------------------------------------------------------------

    \124\ U.S. Environmental Protection Agency, 2000. Guidelines for 
Preparing Economic Analyses. http:// www.yosemite1.epa.gov/ee/epa/eed/hsf/pages/Guideline.html.
    \125\ Office of Management and Budget, The Executive Office of 
the President, 2003. Circular A-4. http://www.whitehouse.gov/omb/circulars.
---------------------------------------------------------------------------

    We prepared an Economic Impact Analysis (EIA) to estimate the 
economic impacts of the proposed emission control program on the Small 
SI and Marine SI engine and equipment markets. In this section we 
briefly describe the Economic Impact Model (EIM) we developed to 
estimate the market-level changes in price and outputs for affected 
markets, the social costs of the program, and the expected distribution 
of those costs across affected stakeholders. We also present the 
results of our analysis. We request comment on all aspects of the 
analysis,

[[Page 28230]]

including the model and the model inputs.
    We estimate the net social costs of the proposed program to be 
about $241 million in 2030.126, 127 This estimate reflects 
the estimated compliance costs associated with the Small SI and Marine 
SI engine standards and the expected fuel savings from improved 
evaporative controls. When the fuel savings are not taken into account, 
the results of the economic impact modeling suggest that the social 
costs of these programs are expected to be about $569 million in 2030. 
Consumers of Small SI and Marine products are expected to bear about 75 
percent of these costs. Small SI engine and equipment manufacturers are 
expected to bear 6 percent and 19 percent, respectively. We estimate 
fuel savings of about $327 million in 2030, which will accrue to 
consumers.
---------------------------------------------------------------------------

    \126\ All estimates presented in this section are in 2005$.
    \127\ This analysis is based on an earlier version of the 
engineering compliance developed for this rule. The net present 
value of the engineering costs used in this analysis (without taking 
the fuel savings into account, at a 3 percent discount rate over the 
period of the analysis) is $10.0 billion, which is about $100 
million less than the net present value of the final estimated 
engineering costs, $10.1 billion. We do not expect that a difference 
of this magnitude would change the overall results of this economic 
impact analysis, in terms of market impacts and how the costs are 
expected to be shared among stakeholders.
---------------------------------------------------------------------------

    With regard to market-level impacts in 2030, the average price 
increase for Small SI engines is expected to be about 9.1 percent ($17 
per unit). The average price increase for Marine SI engines is expected 
to be about 1.7 percent ($195 per unit). The largest average price 
increase for Small SI equipment is expected to be about 5.6 percent 
($15 per unit) for Class I equipment. The largest average price 
increase for Marine SI vessels is expected to be about 2.1 percent 
($178 per unit) for Personal Watercraft.
(1) What is an Economic Impact Analysis?
    An Economic Impact Analysis (EIA) is prepared to inform decision 
makers about the potential economic consequences of a regulatory 
action. The analysis consists of estimating the social costs of a 
regulatory program and the distribution of these costs across 
stakeholders. These estimated social costs can then be compared with 
estimated social benefits (as presented in Section XII.E). As defined 
in EPA's Guidelines for Preparing Economic Analyses, social costs are 
the value of the goods and services lost by society resulting from (a) 
The use of resources to comply with and implement a regulation and (b) 
reductions in output.\128\ In this analysis, social costs are explored 
in two steps. In the market analysis, we estimate how prices and 
quantities of goods affected by the proposed emission control program 
can be expected to change once the program goes into effect. In the 
economic welfare analysis, we look at the total social costs associated 
with the program and their distribution across stakeholders.
---------------------------------------------------------------------------

    \128\ EPA Guidelines for Preparing Economic Analyses, EPA 240-R-
00-003, September 2000, p 113. A copy of this document can be found 
at http://yosemite.epa.gov/ee/epa/eed.nsf/webpages/Guidelines.html.
---------------------------------------------------------------------------

(2) What Is the Economic Impact Model?
    The EIM is a behavioral model developed for this proposal to 
estimate price and quantity changes and total social costs associated 
with the emission controls under consideration. The EIM simulates how 
producers and consumers of affected products can be expected to respond 
to an increase in production costs as a result of the proposed emission 
control program. In this EIM, compliance costs are directly borne by 
producers of affected goods. Depending on the producers' and consumers' 
sensitivity to price changes, producers of affected products will try 
to pass some or all of the increased production costs on to the 
consumers of these goods through price increases. In response to the 
price increases, consumers will decrease their demand for the affected 
good. Producers will react to the decrease in quantity demanded by 
decreasing the quantity they produce; the market will react by setting 
a higher price for those fewer units. These interactions continue until 
a new market equilibrium quantity and price combination is achieved. 
The amount of the compliance costs that can be passed on to the 
consumers is ultimately limited by the price sensitivity of consumers 
and producers in the relevant market (represented by the price 
elasticity of demand or supply). The EIM explicitly models these 
behavioral responses and estimates the new equilibrium prices and 
output and the resulting distribution of social costs across these 
stakeholders (producers and consumers).
(3) What Economic Sectors Are Included in This Economic Impact 
Analysis?
    There are two broad economic sectors affected by the emission 
control program described in this proposal: (1) Small SI engines and 
equipment, and (2) Marine SI engines and equipment. For Small SI 
engines and equipment we distinguish between handheld and nonhandheld 
sectors. For handheld, we model one integrated handheld engine and 
equipment category. On the nonhandheld side, we model 6 engine 
categories, depending on engine class and useful life (Class I: UL125, 
UL250, and UL500; Class II: UL250, UL500, UL1000), and 8 equipment 
categories (agriculture/construction/general industrial; utility and 
recreational vehicles; lawn mowers; tractors; other lawn and garden; 
generator sets/welders; pumps/compressors/pressure washers; and 
snowblowers). For Marine SI engines and equipment, we distinguish 
between sterndrives and inboards (SD/I), outboards (OB), and personal 
watercraft (PWC). SD/I and OB are further categorized by whether they 
are luxury or not. All of these markets are described in more detail in 
Chapter 9 of the RIA and in the industry characterizations prepared for 
this proposal.
    This analysis assumes that all of these products are purchased and 
used by residential households. This means that to model the behavior 
change associated with the proposed standards we model all uses as 
residential lawn and garden care or power generation (Small SI) or 
personal recreation (Marine SI). We do not explicitly model commercial 
uses (how the costs of complying with the proposed programs may affect 
the production of goods and services that use Small SI or Marine SI 
engines or equipment as production inputs); we treat all commercial 
uses as if they were residential uses. We believe this approach is 
reasonable because the commercial share of the end use markets for both 
Small SI and Marine SI equipment is very small.\129\ In addition, for 
any commercial uses of these products the share of the cost of these 
products to total production costs is also small (e.g., the cost of a 
Small SI generator is only a very small part of the total production 
costs for a construction firm). Therefore, a price increase of the 
magnitude anticipated for this control program is not expected to have 
a noticeable impact on prices or quantities of goods or services 
produced using Small SI or Marine SI equipment as inputs (e.g., 
commercial turf care, construction, or fishing).
---------------------------------------------------------------------------

    \129\ The Outdoor Power Equipment Institute (OPEI) provides 
annual estimates of Small SI shipments (unit volumes) broken out 
into commercial and residential markets. For 2003 and 2004, the 
commercial share for NHH products is estimated to be 3.3 percent and 
2.8 percent, respectively; for all Small SI products is estimated to 
be 1.4 percent and 1.2 percent. Similarly, commercial uses of Marine 
SI vessels are limited. See the industry characterizations prepared 
for this proposal for more information (RTI, 2006).

---------------------------------------------------------------------------

[[Page 28231]]

    In the EIM the Small SI and Marine SI markets are not linked (there 
is no feedback mechanism between the Small SI and Marine SI market 
segments). This is appropriate because the affected equipment is not 
interchangeable and because there is very little overlap between the 
engine producers in each market. These two sectors represent different 
aspects of economic activity (lawn and garden care and power generation 
as opposed to recreational marine) and production and consumption of 
one product is not affected by the other. In other words, an increase 
in the price of lawnmowers is not expected to have an impact on the 
production and supply of personal watercraft, and vice versa. 
Production and consumption of each of these products are the results of 
other factors that have little cross-over impacts (the need for 
residential garden upkeep or power generation; the desire for personal 
recreation).
(4) What Are the Key Features of the Economic Impact Model?
    A detailed description of the features of the EIM and the data used 
in this analysis is provided in Chapter 9 of the RIA prepared for this 
rule. The model methodology is firmly rooted in applied microeconomic 
theory and was developed following the methodology set out in OAQPS's 
Economic Analysis Resource Document.\130\
---------------------------------------------------------------------------

    \130\ U.S. Environmental Protection Agency, Office of Air 
Quality Planning and Standards, Innovative Strategies and Economics 
Group, OAQPS Economic Analysis Resource Document, April 1999. A copy 
of this document can be found at http://www.epa.gov/ttn/ecas/econdata/Rmanual2.
---------------------------------------------------------------------------

    The EIM is a computer model comprised of a series of spreadsheet 
modules that simulate the supply and demand characteristics of the 
markets under consideration. The initial market equilibrium conditions 
are shocked by applying the compliance costs for the control program to 
the supply side of the markets (this is done by shifting the relevant 
supply curves by the amount of the compliance costs). The EIM uses the 
model equations, model inputs, and a solution algorithm to estimate 
equilibrium prices and quantities for the markets with the regulatory 
program. These new prices and quantities are used to estimate the 
social costs of the model and how those costs are shared among affected 
markets.
    The EIM uses a multi-market partial equilibrium approach to track 
changes in price and quantity for the modeled markets. As explained in 
EPA's Guidelines for Preparing Economic Analyses, ``partial 
equilibrium'' means that the model considers markets in isolation and 
that conditions in other markets are assumed either to be unaffected by 
a policy or unimportant for social cost estimation. Multi-market 
analysis models go beyond partial equilibrium by extending the inquiry 
to more than just single markets and attempt to capture at least some 
of the interaction between markets--in this case, between selected 
engine and equipment markets sectors.\131\
---------------------------------------------------------------------------

    \131\ EPA Guidelines for Preparing Economic Analyses, EPA 240-R-
00-003, September 2000, p. 125-6.
---------------------------------------------------------------------------

    The EIM uses an intermediate run time frame. This means that some 
factors of production are fixed and some are variable. In very short 
analyses, all factors of production would be assumed to be fixed, 
leaving the producers with no means to respond to the increased 
production costs associated with the regulation (e.g., they cannot 
adjust labor or capital inputs). Under this time horizon, the costs of 
the regulation fall entirely on the producer. In the long run, all 
factors of production are variable and producers can adjust production 
in response to cost changes imposed by the regulation (e.g., using a 
different labor/capital mix). In the intermediate run there is some 
resource immobility which may cause producers to suffer producer 
surplus losses, but they can also pass some of the compliance costs to 
consumers.
    The EIM assumes a perfectly competitive market structure. The 
perfect competition assumption is a widely accepted economic practice 
for this type of analysis, and only in rare cases are other approaches 
used.\132\ It should be noted that the perfect competition assumption 
is not about the number of firms in a market, it is about how the 
market operates. The markets included in this analysis do not exhibit 
evidence of noncompetitive behavior: there are no indications of 
barriers to entry, the firms in these markets are not price setters, 
and there is no evidence of high levels of strategic behavior in the 
price and quantity decisions of the firms. These markets are also 
mature markets as evidenced by unit sales growing at the rate of 
population increases. Pricing power in such markets is typically 
limited. In addition, the products produced within each market are 
somewhat homogeneous in that engines and equipment from one firm can be 
purchased instead of engines and equipment from another firm. Finally, 
according to contestable market theory, oligopolies and even monopolies 
will behave very much like firms in a competitive market if it is 
possible to enter particular markets without cost (i.e., there are no 
sunk costs associated with market entry or exit). This is the case with 
these markets, as there is significant excess production capacity in 
both the Small SI and Marine SI industries, in part due to improved 
productivity and efficiency in current plants. Idle production capacity 
also limits the ability of firms to raise prices, since competitors can 
easily capture market share by increasing their production at the 
expense of a producer that increases its prices. For all of these 
reasons it is appropriate to use a perfect competition model to 
estimate the economic impacts of this proposal.
---------------------------------------------------------------------------

    \132\ See, for example, EPA Guidelines for Preparing Economic 
Analyses, EPA 240-R-00-003, September 2000, p 126.
---------------------------------------------------------------------------

    The perfect competition assumption has an impact on the way the EIM 
is structured. In a competitive market the supply curve is based on the 
industry marginal cost curve; fixed costs do not influence production 
decisions at the margin. Therefore, in the market analysis the model is 
shocked by variable costs only. However, the nature of the Small SI and 
Marine SI markets suggests the market supply curve shifts in the model 
should include fixed and variable compliance costs. This is because 
Small SI and Marine SI engine and equipment manufacturers produce a 
product that changes very little over time. These manufacturers may not 
engage in research and development to improve their products on a 
continuous basis (as opposed to highway vehicles or nonroad engines and 
equipment). If this is the case, then the product changes that would be 
required to comply with the proposed standards would require these 
manufacturers to devote new funds and resources to product redesign and 
facilities changes. In this situation, Small SI and Marine SI engine 
and equipment manufacturers would be expected to increase their prices 
by the full amount of the compliance costs (both fixed and variable) to 
attempt to recover those costs. To reflect these conditions, the supply 
shift in this EIM is based on both fixed and variable costs, even 
though the model assumes perfect competition. A sensitivity analysis 
was performed to investigate the impacts under the alternative 
scenarios of shifting the supply curve by the variable costs only. The 
results of that analysis can be found in the RIA prepared for this 
proposal. We request comment on the extent to which manufacturers can 
be expected to devote additional funds to cover the fixed costs 
associated with the standards, or whether they in fact do provide for 
product development resources on a continuous basis and can

[[Page 28232]]

be expected to use those funds to cover the fixed costs. We also 
request comment on whether companies would attempt to pass fixed costs 
to consumers as an additional price increase and, if so, how much of 
the fixed costs would be based on and for how long.
    The market interactions modeled in the EIM are those between 
producers and consumers of the specified engines and the equipment that 
use those engines. The EIM does not consider sales distribution 
networks or how the regulated goods are sold to final consumers through 
wholesalers and/or retailers. This is appropriate because the proposed 
regulatory program does not impose additional costs on the distribution 
networks and those relationships are not expected to change as a result 
of the standards. In the case of Small SI equipment, however, concerns 
have been raised about the potential for dominant retailers (big box 
stores such as Wal-Mart, Sears and K-Mart) to affect the ability of 
manufacturers to pass along cost increases associated with new emission 
control requirements, forcing them to absorb the compliance costs 
associated with the proposed standards. As described in greater detail 
in Chapter 9 of the RIA, dominant retailers are not expected to affect 
market interactions in ways that would offset the assumption of perfect 
competition by preventing firms from passing on increases in costs 
associated with the control program. This is because all firms in the 
market are expected to comply with the control program, and all will 
experience an increase in marginal costs. Profit-maximizing 
manufacturers will continue to follow a marginal cost pricing rule 
regardless of the distribution arrangements. If large retail 
distributors attempted to prevent efficient manufacturers from raising 
prices in response to the standards, manufacturers would likely respond 
to a retailer's price pressure by reducing output. This would result in 
large excess demand in the equipment market which would ultimately have 
to be satisfied through a new higher equilibrium price, which in turn 
would result in greater supply, thus bidding the price down to a new 
market equilibrium after the application of the control program.
    The relationships modeled in the EIM do not include substitution 
away from Small SI and Marine SI engines and equipment to diesel or 
electric alternatives. This is appropriate because consumers are not 
likely to make these substitutions. Substitution to diesel Small SI 
equipment is not a viable option for most residential consumers, either 
because diesel equipment does not exist (e.g., diesel string trimmers) 
or because there would be a large price premium that would discourage 
the use of diesel equipment (e.g., diesel lawnmowers and diesel 
recreational marine vessels). In addition, most households are not 
equipped to handle the additional fuel type and misfueling would carry 
a high cost. Finally, the lack of a large infrastructure system already 
in place like the one supporting the use of gasoline equipment for 
residential and recreational purposes, including refueling and 
maintenance, represents a large barrier to substitution from gasoline 
to diesel equipment. On the electric side, the impact of substitution 
to electric for Small SI equipment (there are no comparable options for 
Marine SI) is also expected to be negligible. Gasoline is the power 
source of choice for small and inexpensive equipment due to its low 
initial cost. Gasoline equipment is also inherently portable, which 
make them more attractive to competing electric equipment that must be 
connected with a power grid or use batteries that require frequent 
recharging.
    The EIM is a market-level analysis that estimates the aggregate 
economic impacts of the control program on the relevant market. It is 
not a firm-level analysis and therefore the supply elasticity or 
individual compliance costs facing any particular manufacturer may be 
different from the market average. This difference can be important, 
particular where the rule affects different firms' costs over different 
volumes of production. However, to the extent there are differential 
effects on individual firms, EPA believes that the wide array of 
compliance flexibilities provided in this proposal are adequate to 
address any cost inequities that are likely to arise.
    Finally, consistent with the proposed emission controls, this EIA 
covers engines sold in 49 states. California engines are not included 
because California has its own state-level controls for Small SI and 
Marine SI engines. The sole exceptions are Small SI engines used in 
agriculture and construction applications in California. These engines 
are included in the control program and in this analysis because the 
Clean Air Act pre-empts California from setting standards for these 
engines.
(5) What Are the Key Model Inputs?
    Key model inputs for the EIM are the behavioral parameters, the 
market equilibrium quantities and prices, and the compliance cost 
estimates.
    The model's behavioral paramaters are the price elasticities of 
supply and demand. These parameters reflect how producers and consumers 
of the engines and equipment affected by the standards can be expected 
to change their behavior in response to the costs incurred in complying 
with the standards. More specifically, the price elasticity of supply 
and demand (reflected in the slope of the supply and demand curves) 
measure the price sensitivity of consumers and producers. The price 
elasticities used in this analysis are summarized in Table XII.F-1 and 
are described in more detail in Chapter 9 of the RIA. An ``inelastic'' 
price elasticity (less than one) means that supply or demand is not 
very responsive to price changes (a one percent change in price leads 
to less than one percent change in demand). An ``elastic'' price 
elasticity (more than one) means that supply or demand is sensitive to 
price changes (a one percent change in price leads to more than one 
percent change in demand). A price elasticity of one is unit elastic, 
meaning there is a one-to-one correspondence between a change in price 
and change in demand.

                                 Table XII. F-1.--Behavioral Parameters Used in Small SI/Marine SI Economic Impact Model
--------------------------------------------------------------------------------------------------------------------------------------------------------
               Sector                        Market             Demand elasticity            Source           Supply elasticity            Source
--------------------------------------------------------------------------------------------------------------------------------------------------------
Engine.............................  Small SI and Marine SI  Derived...............  N/A..................  3.8 (elastic)........  EPA Econometric
                                      Engine Market.                                                                                Estimate.
Small SI Equipment.................  All handheld..........  -1.9 (elastic)........  EPA Econometric        3.4 (elastic)........  EPA Econometric
                                                                                      Estimate.                                     Estimate.
                                     Lawn Mowers...........  -0.2 (inelastic)......  EPA Econometric        Same as above........
                                                                                      Estimate.
                                     Other lawn & garden...  -0.9 (inelastic)......  EPA Econometric        Same as above........
                                                                                      Estimate.

[[Page 28233]]

 
                                     Gensets/welders (class  -1.4 (elastic)........  EPA Econometric        3.3 (elastic)........  EPA Econometric
                                      I).                                             Estimate.                                     Estimate.
                                     Gensets/welders (class  -1.1 (elastic)........  EPA Econometric        Same as above........
                                      II).                                            Estimate.
                                     All other non-handheld  -1.0 (unit elastic)...  EPA Econometric        3.4 (elastic) Same as
                                                                                      Estimate.              above.
Marine SI Equipment................  PWC...................  -2.0 (elastic)........  EPA Econometric        3.4 (elastic)........  EPA Econometric
                                                                                      Estimate.                                     Estimate.
                                     All other vessels       Same as above.........  .....................  2.3 (elastic)........  EPA Econometric
                                      types.                                                                                        Estimate.
--------------------------------------------------------------------------------------------------------------------------------------------------------

    The estimated supply and demand elasticities were based on best 
data we could find. We used (1) The industry-level data published by 
the National Bureau of Economic Research (NBER)-Center for Economic 
Studies (Bartlesman, Becker, and Gray, 2000); (2) Current Industrial 
Reports (CIR) series from the U.S. Census Bureau; (3) several data 
series reported in a study by Air Improvement Resource Inc. and 
National Economic Research Associates (AIR/NERA, 2003) for the walk-
behind lawnmowers; (4) the U.S. Census Bureau historical data on 
household income and housing starts (U.S. Census Bureau, 2002; 2004); 
(5) price, wage, and material cost indexes from the Bureau of Labor 
Statistics (BLS) (BLS, 2004a,b,c,d,e); (6) the implicit gross domestic 
product (GDP) price deflator reported by the U.S. Bureau of Economic 
Analysis (BEA, 2004). It should be noted that the aggregate data we 
used to estimate elasticities include data on other markets as well as 
the Small SI or Marine SI markets. If we had been able to obtain 
market-specific data for Small SI or Marine SI only, the estimated 
price elasticities may have been different.
    The estimated supply elasticities for all of the equipment and 
engine markets are elastic, ranging from 2.3 for all recreational 
marine except PWC, to 3.3 for generators, 3.4 for PWCs and all Small SI 
except generators, and 3.8 for engines. This means that quantities 
supplied are expected to be fairly sensitive to price changes (e.g., a 
1% change in price yields a 3.3 percent change in quantity of 
generators produced).
    On the demand side, the Marine SI equipment market estimated demand 
elasticity is elastic, at -2.0. This is consistent with the 
discretionary nature of purchases of recreational marine vessels 
(consumers can easily decide to spend their recreational budget on 
other alternatives).
    The estimated demand elasticity for handheld equipment is elastic, 
at -1.9. This suggests that consumers are more sensitive to price 
changes for handheld equipment than for other Small SI equipment. In 
other words, they are more likely to change their purchase decision for 
a small change in the price of a string trimmer, perhaps opting for 
trimmer shears or deciding to forego trimming altogether.
    The estimated demand elasticity for lawnmowers is very inelastic at 
-0.2. This suggests that consumers of this equipment are not very 
sensitive to price changes. Most of this equipment is sold to 
individual homeowners, who are often required by local authorities to 
keep their lawns trimmed. Household ownership of a gasoline lawnmower 
is often their least expensive option. Lawncare services are more 
expensive since the price for these services includes labor and other 
factors of production. Purchasing other equipment may also not be 
attractive, since electric and diesel mowers are generally more 
expensive and often less convenient. Finally, the option of using 
landscape alternatives (e.g., prairie, wildflower, or rock gardens) may 
not be attractive for homeowners who may also use their yards for 
recreational purposes. For all these reasons, the price sensitivity of 
homeowners to lawnmower prices would be expected to be inelastic.
    All the other demand elasticities, for gensets, welders, 
compressors, and ag/construction equipment, are about unit elastic, at 
-1.0 meaning a 1 percent change in price is expected to result in a 1 
percent change in demand.
    The demand elasticities for the engine markets are internally 
derived as part of the process of running the model. This is an 
important feature of the EIM, which allows it to link the engine and 
equipment components of each model and simulate how compliance costs 
can be expected to ripple through the affected market. In actual 
markets, for example, the quantity of lawnmowers produced in a 
particular period depends on the price of engines (the Small SI engine 
market) and the demand for equipment by residential consumers. 
Similarly, the number of engines produced depends on the demand for 
engines (the lawnmower market) which depends on consumer demand for 
equipment. Changes in conditions in one of these markets will affect 
the others. By designing the model to derive the engine demand 
elasticities, the EIM simulates these connections between supply and 
demand among the product markets and replicates the economic 
interactions between producers and consumers.
    Initial market equilibrium quantities for these markets are 
simulated using the same current year sales quantities used in the 
engineering cost analysis. The initial market equilibrium prices for 
Small SI and Marine SI engines and equipment were derived from industry 
sources and published data and are described in Chapter 9 of the Draft 
RIA.
    The compliance costs used to shock the model, to simulate the 
application of the control program, are the same as the engineering 
costs described in Chapter 6. However, the EIM uses an earlier version 
of the engineering compliance developed for this rule. The net present 
value of the engineering costs used in this analysis (without taking 
the fuel savings into account, at a 3 percent discount rate over the 
period of the analysis) is $10.0 billion, which is about $100 million 
less than the net present value of the final estimated engineering 
costs, $10.1 billion. We do not expect that a difference of this 
magnitude would change the overall results of this economic impact 
analysis, in terms of market impacts and how the costs are expected to 
be shared among stakeholders.
    As explained in Section XII.F.4, the EIM uses both fixed and 
variable engineering costs to shock the initial equilibrium conditions. 
The fixed costs are amortized over the first 5 years of the standards 
and include a 7 percent cost of capital. For some elements of the 
program (i.e., evaporative emission controls), fixed costs are incurred

[[Page 28234]]

throughout the period of analysis due to the need to replace tooling.
    Additional costs that need to be considered in the EIM are the 
operating costs (fuel savings) associated with the evaporative emission 
controls. These fuel savings are not included in the market analysis 
for this economic impact analysis. This is because all available 
evidence suggests that fuel savings do not affect consumer decisions 
with respect to the purchase of this equipment. Unlike motor vehicles 
or other consumer goods, neither Small SI nor Marine SI equipment is 
labeled with expected fuel consumption or expected annual operating 
costs. Therefore, there is no information available for the consumer to 
use to make this decision. Instead consumers base their purchase 
decision on other attributes of the product for which the manufacturer 
provides information. For lawn mowers this may be the horsepower of the 
engine, whether the machine has a bag or has a mulching feature, its 
blade size, etc. For PWC it may be how many people it can carry, its 
maximum speed, its horsepower, etc. In many cases, especially for Small 
SI equipment, the consumer may not even be aware of the fuel savings 
when operating the equipment, especially if he or she uses the same 
portable fuel storage container to fuel several different pieces of 
equipment.
    These fuel savings are included in the social cost analysis. This 
is because they are savings that accrue to society. These savings are 
attributed to consumers of the relevant equipment. As explained in more 
detail in Section 9.3.5 of the Draft RIA, the social cost analysis is 
based on the equivalent of the pre-tax price of gasoline in that 
analysis. Although the consumer will realize a savings equal to the 
pump price of gasoline (post-tax), part of that savings is offset by a 
tax loss to governmental agencies and is thus a loss to consumers of 
the services supported by those taxes. This tax revenue loss, 
considered a transfer payment in this analysis, does not affect the 
benefit-cost analysis results.
(6) What Are the Results of the Economic Impact Modeling?
    Using the model and data described above, we estimated the economic 
impacts of the proposed emission control program. We performed a market 
analysis for all years and all engine and equipment types. In this 
section we present summarized results for selected markets and years. 
More detail can be found in the appendices to Chapter 9 of the RIA and 
in the docket for this rule.\133\ Also included in Appendix 9H to that 
chapter are sensitivity analyses for several key inputs.
---------------------------------------------------------------------------

    \133\ Li, Chi. 2007. Memorandum to Docket EPA-HQ-OAR-2004-0008. 
Detailed Results From Economic Impact Model.
---------------------------------------------------------------------------

    The EIA consists of two parts: a market analysis and a welfare 
analysis. The market analysis looks at expected changes in prices and 
quantities for affected products. The welfare analysis looks at 
economic impacts in terms of annual and present value changes in social 
costs.
    As explained in Section XII.F.4, the EIM is shocked by the sum of 
fixed and variable costs. For the market analysis, this leads to a 
small increase in estimated price impacts for the years 2011 through 
2016, the period during which fixed costs are recovered. The increase 
is small because, for many elements of the program, annual per unit 
fixed costs are smaller than annual per unit variable costs. For the 
welfare analysis, applying both fixed and variable costs means that the 
burden of the social costs attributable to producers and consumers 
remains fixed throughout the period of analysis. This is because 
producers pass the fixed costs to consumers at the same rate as the 
variable costs instead of having to absorb them internally.
(a) Market Impact Analysis
    In the market analysis, we estimate how prices and quantities of 
goods affected by the proposed emission control program can be expected 
to change once the program goes into effect. The analysis relies on the 
initial market equilibrium prices and quantities for each type of 
equipment and the price elasticity of supply and demand. It predicts 
market reactions to the increase in production costs due to the new 
compliance costs (variable and fixed). It should be noted that this 
analysis does not allow any other factors of production to vary. In 
other words, it does not consider that manufacturers may adjust their 
production processes or marketing strategies in response to the control 
program. Also, as explained above, while the markets are shocked by 
both fixed and variable costs, the market shock is not offset by fuel 
savings.
    A summary of the estimated market impacts is presented in Table 
XII.F-2 for 2013, 2018, and 2030. These years were chosen because 2013 
is the year of highest compliance; after 2018, the fixed costs are 
recovered and the market impacts reflect variable costs as well as 
growth in equipment population; and 2030 illustrates the long-term 
impacts of the program.
    Market level impacts are reported for the engine and equipment 
markets separately. This is because the EIM is a two-level model that 
treats these markets separately. However, changes in equipment prices 
and quantities are due to impacts of both direct equipment compliance 
costs and indirect engine compliance costs that are passed through to 
the equipment market from the engine market through higher engine 
prices.
    The average market-level impacts presented in this section are 
designed to provide a broad overview of the expected market impacts 
that is useful when considering the impacts of the rule on the economy 
as a whole. The average price impacts are product-weighted averages of 
the results for the individual engine and equipment categories included 
in that sub-sector (e.g., the estimated Marine SI engine price and 
quantity changes are weighted averages of the estimated results for all 
of the Marine SI engine markets). The average quantity impacts are the 
sum of the decrease in units produced units across sub-markets. Price 
increases and quantity decreases for specific types of engines and 
equipment are likely to be different.
    Although each of the affected equipment in this analysis generally 
requires one engine (the exception being Marine SI sterndrive/
inboards), the estimated decrease in the number of engines produced in 
Table XII.F-2 is less than the estimated decrease in the number of 
equipment produced. At first glance, this result seems counterintuitive 
because it does not reflect the approximate one-to-one correspondence 
between engines and equipment. This discrepancy occurs because the 
engine market-level analysis examines only output changes for engines 
that are produced by independent engine manufacturers and subsequently 
sold to independent equipment manufacturers. Engines produced and 
consumed by vertically integrated equipment/engine manufacturers are 
not explicitly modeled. Therefore, the market-level analysis only 
reflects engines sold on the ``open market,'' and estimates of output 
changes for engines consumed internally are not reflected in this 
number.\134\ Despite the fact that changes

[[Page 28235]]

in consumption of internally consumed engines are not directly reported 
in the market-level analysis results, the costs associated with these 
engines are included in the market-level analysis (as supply shift for 
the equipment markets). In addition, the cost and welfare analyses 
include the compliance costs associated with internally consumed 
engines.
---------------------------------------------------------------------------

    \134\ For example, PWC and handheld equipment producers 
generally integrate equipment and engine manufacturing processes and 
are included in the EIM as one-level equipment markets. Since there 
is no engine market for these engines, the EIM does not include PWC 
and handheld engine consumption changes in engine market-level 
results.
---------------------------------------------------------------------------

Marine SI Market Analysis

    The average price increase for Marine SI engines in 2013, the high 
cost year, is estimated to be about 2.3 percent, or $257. By 2018, this 
average price increase is expected to decline to about 1.7 percent, or 
$196, and remain at that level for later years. The market impact 
analysis predicts that with these increases in engine prices the 
expected average decrease in total sales in 2013 is about 2.0 percent, 
or 8,800 engines. This decreases to about 1.6 percent in 2018, or about 
7,000 engines.
    On the vessel side, the average price change reflects the direct 
equipment compliance costs plus the portion of the engine costs that 
are passed on to the equipment purchaser (via higher engine prices). 
The average price increase in 2013 is expected to be about 1.3 percent, 
or $232. By 2018, this average price increase is expected to decline to 
about 1 percent, or $178. These price increases are expected to vary 
across vessel categories. The category with the largest price increase 
in 2013 is expected to be personal watercraft engines, with an 
estimated price increase of about 2.8 percent in 2013; this is expected 
to decrease to 2.1 percent in 2018. The smallest expected change in 
2013 is expected to be for sterndrive/inboards and outboard 
recreational vessels, which are expected to see price increases of 
about 0.7 percent. The market impact analysis predicts that with these 
increases in vessel prices the expected average decrease in quantity 
produced in 2013 is about 2.7 percent, or 11,000 vessels.\135\ This is 
expected to decrease to about 2.0 percent in 2018, or about 8,600 
vessels. The personal watercraft category is expected to experience the 
largest decline in 2013, about 5.6 percent (4,800 vessels). The 
smallest percentage decrease in production is expected for sterndrive/
inboards at 1.4 percent (1,300 vessels); the smallest absolute decrease 
in quantity is expected for outboard recreational vessels, at 113 
vessels (1.5 percent).
---------------------------------------------------------------------------

    \135\ It should be noted that the absolute change in the number 
of engines and equipment does not match. This is because the 
absolute change in the quantity of engines represents only engines 
sold on the open market. Reductions in engines consumed internally 
by integrated engine/equipment manufacturers are not reflected in 
this number but are captured in the social cost analysis.
---------------------------------------------------------------------------

Small SI Market Analysis

    The average price increase for Small SI engines in 2013, the high 
cost year, is estimated to be about 11.7 percent, or $22. By 2018, this 
average price increase is expected to decline to about 9.1 percent, or 
$17, and remain at that level for later years. The market impact 
analysis predicts that with these increases in engine prices the 
expected average decrease in total sales in 2013 is expected to be 
about 2.3 percent, or 371,000 engines. This is expected to decrease to 
about 1.7 percent in 2018, or about 299,000 engines.
    On the equipment side, the average price change reflects the direct 
equipment compliance costs plus the portion of the engine costs that 
are passed on to the equipment purchaser (via higher engine prices). 
The average price increase for all Small SI equipment in 2013 is 
expected to be about 3.1 percent, or $14. By 2018, this average price 
increase is expected to decline to about 2.4 percent, or $10. The 
average price increase and quantity decrease differs by category of 
equipment. As shown in Table XII.F-2, the price increase for Class I 
equipment is estimated to be about 6.9 percent ($19) in 2013, 
decreasing to 5.5 percent ($15) in 2018. The market impact analysis 
predicts that with these increases in equipment prices the expected 
average decrease in the quantity of Class I equipment produced in 2013 
is about 2.2 percent, or 219,400 units.\136\ This is expected to 
decrease to about 1.8 percent in 2018, or about 189,700 units. For 
Class II equipment, a higher price increase is expected, about 3.9 
percent ($41) in 2013, decreasing to 2.6 percent ($25) in 2018. The 
expected average decrease in the quantity of Class II equipment 
produced in 2013 is about 4.3 percent, or 157,300 units, decreasing to 
2.8 percent, or about 114,000 units, in 2018.
---------------------------------------------------------------------------

    \136\ See previous note.
---------------------------------------------------------------------------

    For the handheld equipment market, prices are expected to increase 
about 0.3 percent for all years, and quantities are expected to 
decrease about 0.6 percent.

                          Table XII.F-2.--Estimated Market Impacts for 2013, 2018, 2030
                                                     [2005$]
----------------------------------------------------------------------------------------------------------------
                                                                   Change in price         Change in quantity
                           Market                            ---------------------------------------------------
                                                                Absolute     Percent      Absolute     Percent
----------------------------------------------------------------------------------------------------------------
                                                      2013
----------------------------------------------------------------------------------------------------------------
Marine:
    Engines.................................................         $257          2.3       -8,846         -2.0
    Equipment...............................................          232          1.3      -10,847         -2.7
        SD/I................................................          252          0.7       -1,336         -1.4
        OB Recreational.....................................          638          0.7         -113         -1.5
        OB Luxury...........................................          206          1.1       -4,579         -2.1
        PWC.................................................          237          2.8       -4,819         -5.6
Small SI:
    Engines.................................................           22         11.7     -371,097         -2.3
    Equipment...............................................           14          3.1     -482,942         -1.9
        Class I.............................................           19          6.9     -219,400         -2.2
        Class II............................................           41          3.9     -157,306         -4.3
        HH..................................................          0.3          0.3     -106,236         -0.6
----------------------------------------------------------------------------------------------------------------

[[Page 28236]]

 
                                                      2018
----------------------------------------------------------------------------------------------------------------
Marine:
    Engines.................................................          196          1.7       -7,002         -1.6
    Equipment...............................................          178          1.0       -8,563         -2.0
        SD/I................................................          195          0.5       -1,072         -1.1
        OB Recreational.....................................          496          0.6          -91         -1.1
        OB Luxury...........................................          160          0.8       -3,634         -1.6
        PWC.................................................          178          2.1       -3,766         -4.2
Small SI:
    Engines.................................................           17          9.1     -298,988         -1.7
    Equipment...............................................           10          2.4     -401,025         -1.4
        Class I.............................................           15          5.5     -189,771         -1.8
        Class II............................................           25          2.6     -113,999         -2.8
        HH..................................................          0.2          0.3      -97,255         -0.5
----------------------------------------------------------------------------------------------------------------
                                                      2030
----------------------------------------------------------------------------------------------------------------
Marine:
    Engines.................................................          195          1.7       -7,728         -1.6
    Equipment...............................................          179          1.0       -9,333         -2.0
        SD/I................................................          195          0.5       -1,161         -1.1
        OB Recreational.....................................          496          0.6          -98         -1.1
        OB Luxury...........................................          160          0.8       -3,998         -1.7
        PWC.................................................          178          2.1       -4,076         -4.2
Small SI:
    Engines.................................................           17          9.1     -354,915         -1.7
    Equipment...............................................           10          2.4     -475,825         -1.4
        Class I.............................................           15          5.6     -225,168         -1.8
        Class II............................................           25          2.6     -135,400         -2.8
        HH..................................................          0.2          0.3     -115,257         -0.5
----------------------------------------------------------------------------------------------------------------

(b) Economic Welfare Analysis
    In the economic welfare analysis we look at the costs to society of 
the proposed program in terms of losses to consumer and producer 
surplus. These surplus losses are combined with the fuel savings to 
estimate the net economic welfare impacts of the proposed program. 
Estimated annual net social costs for selected years are presented in 
Table XII-F-3. This table shows that total social costs for each year 
are slightly less than the total engineering costs. This is because the 
total engineering costs do not reflect the decreased sales of engines 
and equipment that are incorporated in the total social costs.

                   Table XII.F-3.--Estimated Annual Engineering and Social Costs, Through 2038
                                                [2005$, $million]
----------------------------------------------------------------------------------------------------------------
                                                                                        Net
                                       Total                                        engineering     Net social
              Year                  engineering    Total social    Fuel savings        costs           costs
                                       costs           costs                        (including      (including
                                                                                   fuel savings)   fuel savings)
----------------------------------------------------------------------------------------------------------------
2008............................            $9.5            $9.5            $3.1            $6.4            $6.4
2009............................           171.7           168.8            13.7           157.9           155.1
2010............................           191.1           188.0            25.4           165.7           162.6
2011............................           470.5           463.4            64.9           405.7           398.5
2012............................           647.3           638.2           103.5           543.8           534.7
2013............................           652.5           643.4           136.5           516.0           506.9
2014............................           621.1           613.1           161.2           459.9           451.9
2015............................           627.0           619.0           182.3           444.7           436.7
2016............................           520.9           515.2           200.9           320.0           314.2
2017............................           492.6           487.5           216.2           276.4           271.3
2018............................           497.2           492.0           229.9           267.3           262.1
2019............................           503.6           498.4           242.1           261.5           256.2
2020............................           510.0           504.7           253.1           256.9           251.6
2021............................           516.4           511.0           263.3           253.1           247.8
2022............................           522.7           517.3           272.9           249.8           244.4
2023............................           529.1           523.7           281.4           247.7           242.3
2024............................           535.8           530.3           289.3           246.5           241.0
2025............................           542.3           536.7           296.6           245.6           240.0
2026............................           548.7           543.1           303.6           245.1           239.5

[[Page 28237]]

 
2027............................           555.2           549.4           310.1           245.1           239.3
2028............................           561.6           555.8           316.3           245.3           239.5
2029............................           568.0           562.2           322.0           246.1           240.2
2030............................           574.5           568.6           327.3           247.2           241.3
2031............................           580.9           575.0           332.3           248.6           242.6
2032............................           587.4           581.3           337.1           250.3           244.2
2033............................           593.8           587.7           341.7           252.1           246.0
2034............................           600.3           594.1           346.1           254.2           248.0
2035............................           606.7           600.5           350.4           256.3           250.1
2036............................           613.1           606.9           354.5           258.6           252.3
2037............................           619.6           613.2           358.5           261.1           254.7
2038............................           626.0           619.6           362.5           263.6           257.1
NPV at 3% a.....................         9,996.2         9,882.2         4,356.2         5,640.1         5,526.0
NPV at 7% a.....................         5,863.6         5,794.1         2,291.5         3,572.1        3,502.6
----------------------------------------------------------------------------------------------------------------
\a\ EPA EPA presents the present value of cost and benefits estimates using both a three percent and a seven
  percent social discount rate. According to OMB Circular A-4, ``the 3 percent discount rate represents the
  ``social rate of time preference'* * * [which] means the rate at which `society' discounts future consumption
  flows to their present value''; ``the seven percent rate is an estimate of the average before-tax rate of
  return to private capital in the U.S. economy* * * [that] approximates the opportunity cost of capital.''

    Table XII.F-4 shows how total social costs are expected to be 
shared across stakeholders, for selected years. According to these 
results, consumers in the Marine SI market are expected to bear 
approximately 66 percent of the cost of the Marine SI program. This is 
expected to be offset by the fuel savings. Vessel manufacturers are 
expected to bear about 22 percent of that program, and engine 
manufacturers the remaining 11 percent. In the Small SI market, 
consumers are expected to bear 79 percent of the cost of the Small SI 
program. This will also be offset by the fuel savings. Equipment 
manufacturers are expected to bear about 17 percent of that program, 
and engine manufacturers the remaining 4 percent. The estimated 
percentage changes in surplus are the same for all years because the 
initial equilibrium conditions are shocked by both fixed and variable 
costs; producers would pass the fixed costs to consumers at the same 
rate as the variable costs.

                      Table XII.F-4: Summary of Estimated Social Costs for 2013, 2018, 2030
                                               [2005 $, $ million]
----------------------------------------------------------------------------------------------------------------
                                                     Absolute
                     Market                          change in    Percent change   Fuel savings    Total change
                                                      surplus       in  surplus                     in surplus
----------------------------------------------------------------------------------------------------------------
                                                      2013
----------------------------------------------------------------------------------------------------------------
Marine SI:
    Engine Manufacturers........................         -$21.54              11  ..............         -$21.54
    Equipment Manufacturers.....................          -42.23              22  ..............          -42.23
    End User (Households).......................         -125.14              66          $42.27          -82.87
                                                 ---------------------------------------------------------------
        Subtotal................................         -188.91  ..............  ..............         -146.64
                                                 ---------------------------------------------------------------
Small SI:
    Engine Manufacturers........................          -18.36               4  ..............          -18.36
    Equipment Manufacturers.....................          -80.16              18  ..............          -80.16
    End User (Households).......................         -355.95              78           94.26         -261.69
                                                 ---------------------------------------------------------------
        Subtotal................................         -454.47  ..............  ..............         -360.21
                                                 ---------------------------------------------------------------
            Total...............................         -643.38  ..............          136.53         -506.85
----------------------------------------------------------------------------------------------------------------
                                                      2018
----------------------------------------------------------------------------------------------------------------
Marine SI:
    Engine Manufacturers........................          -17.29              11  ..............          -17.29
    Equipment Manufacturers.....................          -34.02              22  ..............          -34.02
    End User (Households).......................         -100.19              66           87.12          -13.07
                                                 ---------------------------------------------------------------
        Subtotal................................         -151.50  ..............  ..............          -64.38
                                                 ---------------------------------------------------------------
Small SI:
    Engine Manufacturers........................          -13.89               4  ..............          -13.89
    Equipment Manufacturers.....................          -57.65              17  ..............          -57.65

[[Page 28238]]

 
    End User (Households).......................         -268.95              79          142.78         -126.17
                                                 ---------------------------------------------------------------
        Subtotal................................         -340.49  ..............  ..............         -197.71
                                                 ---------------------------------------------------------------
            Total...............................         -491.99  ..............          229.90         -262.09
----------------------------------------------------------------------------------------------------------------
                                                      2030
----------------------------------------------------------------------------------------------------------------
Marine SI:
    Engine Manufacturers........................          -18.81              11  ..............          -18.81
    Equipment Manufacturers.....................          -36.97              23  ..............          -36.97
    End User (Households).......................         -108.52              66          149.36           40.84
                                                 ---------------------------------------------------------------
        Subtotal................................         -164.30  ..............  ..............          -14.94
                                                 ---------------------------------------------------------------
Small SI:
    Engine Manufacturers........................          -16.49               4  ..............          -16.49
    Equipment Manufacturers.....................          -68.45              17  ..............          -68.45
    End User (Households).......................         -319.31              79          177.89         -141.42
                                                 ---------------------------------------------------------------
        Subtotal................................         -404.25  ..............  ..............         -226.36
                                                 ---------------------------------------------------------------
            Total...............................         -568.55  ..............          327.25         -241.30
----------------------------------------------------------------------------------------------------------------

    Table XII.F-5 contains more detailed information on the sources of 
the social costs for 2013. This table shows that vessel and equipment 
manufacturers are expected to bear more of the burden of the program 
than engine manufacturers. On the marine side, the loss of producer 
surplus for the vessel manufacturers has two sources. First, they would 
bear part of the burden of the equipment costs. Second, they would also 
bear part of the engine costs, which are passed on to vessel 
manufacturers in the form of higher engine prices. Vessel manufacturers 
would not be able to pass along a greater share of the engine and 
vessel compliance costs to end consumers due to the elastic price 
elasticity of demand for consumers of these vessels. On the Small SI 
side, equipment manufacturers can pass on more of the compliance costs 
to end consumers because the price elasticity of demand in these 
markets is less elastic.

          Table XII.F-5.--Distribution of Estimated Surplus Changes by Market and Stakeholder for 2013
                                                [2005$, million$]
----------------------------------------------------------------------------------------------------------------
                                    Engineering
             Scenario                compliance    Producer     Consumer      Total         Fuel         Net
                                        costs      surplus      surplus      surplus      savings      surplus
----------------------------------------------------------------------------------------------------------------
                                                    Marine SI
----------------------------------------------------------------------------------------------------------------
Engine Manufacturers..............       $133.2       -$21.5  ...........       -$21.5  ...........       -$21.5
---------------------------------------------