[Federal Register Volume 63, Number 17 (Tuesday, January 27, 1998)]
[Proposed Rules]
[Pages 3950-4036]
From the Federal Register Online via the Government Publishing Office [www.gpo.gov]
[FR Doc No: 98-941]



[[Page 3949]]

_______________________________________________________________________

Part II





Environmental Protection Agency





_______________________________________________________________________



40 CFR Part 90



Phase 2 Emission Standards for New Nonroad Spark-Ignition Engines at or 
Below 19 Kilowatts; Proposed Rule

  Federal Register / Vol. 63, No. 17 / Tuesday, January 27, 1998 / 
Proposed Rules  

[[Page 3950]]



ENVIRONMENTAL PROTECTION AGENCY

40 CFR Part 90

[FRL-5942-9]
RIN 2060-AE29


Phase 2 Emission Standards for New Nonroad Spark-Ignition Engines 
At or Below 19 Kilowatts

AGENCY: Environmental Protection Agency (EPA).

ACTION: Notice of proposed rulemaking (NPRM).

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

SUMMARY: Today's action proposes a second phase of regulations to 
control emissions from new nonroad spark-ignition engines at or below 
19 kilowatts (25 horsepower). These engines are used principally in 
lawn and garden equipment, both in nonhandheld applications such as 
lawnmowers, and also in handheld applications such as trimmers and 
chainsaws. The proposed standards are expected to result in a 30 
percent reduction of emissions of hydrocarbons plus oxides of nitrogen 
from the current Phase 1 standards. If adopted, the standards would 
result in important reductions in emissions which contribute to 
excessively high ozone levels in many areas of the United States.

DATES: Written comments on this NPRM must be submitted on or before 
March 13, 1998. EPA will hold a public hearing on February 11, 1998 
starting at 10:00; requests to present oral testimony must be received 
on or before February 6, 1998.

ADDRESSES: Written comments should be submitted (in duplicate if 
possible) to: EPA Air and Radiation Docket, Attention Docket No. A-96-
55, Room M-1500 (mail code 6102), 401 M Street, SW, Washington, D.C. 
20460. Materials relevant to this rulemaking are contained in this 
docket and may be viewed from 8:00 a.m. until 5:30 p.m. weekdays. The 
docket may also be reached by telephone at (202) 260-7548. As provided 
in 40 CFR part 2, a reasonable fee may be charged by EPA for 
photocopying. The public hearing will be held in Ann Arbor, MI at a 
location to be determined; call (313) 668-4278 for further information.

FOR FURTHER INFORMATION CONTACT: Robert Larson, Office of Mobile 
Sources, Engine Programs and Compliance Division, (313) 668-4278, 
[email protected].

SUPPLEMENTARY INFORMATION:

Table of Contents

I. Regulated Entities
II. Legal Authority and Background
III. Overview of Proposed Provisions
    A. More Stringent Standards and a Shift to Cleaner Technology
    1. Nonhandheld Engine HC+NOX Emission Standards
    2. Handheld Engine HC+NOX Emission Standards
    3. Useful Life Categories
    B. Assuring Emission Reductions are Achieved In-use
    1. Traditional Compliance Programs for Mobile Sources
    2. Compliance Programs for the Small SI Engine Industry
    3. The Proposed Phase 2 Compliance Program
    4. Alternative Compliance Program Options
IV. Description of Proposed Program
    A. Standards and Related Provisions
    1. HC+NOX Emission Standards
    2. NMHC+NOX Emission Standards for Class I and II 
Natural Gas Fueled Nonhandheld Engines
    3. CO Emission Standards
    4. Useful Life Categories
    5. Certification Averaging, Banking and Trading Program
    6. Certification Fuel
    B. Test Procedures
    1. Test Cycle: Requirement for the Use of a Speed Governor 
Operation for Testing of Nonhandheld Engines
    2. Test Cycle: Adjustments for Weightings for 2-mode Cycle for 
Handheld Engines
    3. Measurement of NMHC Emissions From Natural Gas Fueled 
Nonhandheld Engines
    C. Field/Bench Adjustment Program
    1. Background and Principles
    2. General Methodology
    3. Practical Requirements of the Program
    4. Alternative Methodology Considered
    D. Compliance Program
    1. Certification
    2. Production Line Testing
    3. In-use Emission Testing
    4. Criteria for Evaluating Alternatives to Mandatory Recall
    E. Flexibilities
    1. Overview of Approach to Providing Compliance Flexibilities
    2. Proposed Production Volume Cutoffs
    3. General Flexibilities
    4. Phase-In Flexibilities
    5. Flexibilities for Small Volume Engine Manufacturers and Small 
Volume Engine Families
    6. Flexibilities for Small Volume Equipment Manufacturers and 
Small Volume Equipment Models
    7. Engine Availability
    F. Nonregulatory Programs
    1. Voluntary ``Green'' Labeling Program
    2. Voluntary Fuel Spillage and Evaporative Emission Reduction 
Program
    3. Particulate matter and Hazardous Air Pollutant Testing 
Program for Handheld Engines
    G. General Provisions
    1. Model Year Definition and Annual Production Period 
Flexibilities During the Transition to Phase 2
    2. Definition of Handheld Engines
    3. Small Displacement Nonhandheld Engine Class
    4. Liquefied Petroleum Gas Fueled Indoor Power Equipment
    5. Dealer Responsibility
    6. Engines Used in Recreational Vehicles
    7. Engines Used in Rescue and Emergency Equipment
    8. Replacement Engines
V. Environmental Benefit Assessment
    A. Roles of HC and NOX in Ozone Formation
    B. Health and Welfare Effects of Tropospheric Ozone
    C. Estimated Emissions Impact of Proposed Regulations
    D. Health and Welfare Effects of CO Emissions
    E. Health and Welfare Effects of Hazardous Air Pollutant 
Emissions
    F. Particulate Matter
VI. Economic Impacts
    A. Engine Technologies
    B. Engine Costs
    1. Nonhandheld Engine Costs
    2. Handheld Engine Costs
    C. Equipment Costs
    1. Nonhandheld Equipment Manufacturers
    2. Handheld Equipment Manufacturers
    D. Operating Costs
    Nonhandheld Engines
    Handheld Engines
    E. Cost per Engine and Cost-effectiveness
    1. Cost per Engine
    2. Cost-effectiveness
VII. Public Participation
    A. Comments and the Public Docket
    B. Public Hearing
    C. Obtaining Electronic Copies of Documents
VIII. Administrative Requirements
    A. Administrative Designation and Regulatory Analysis
    B. Paperwork Reduction Act
    C. Unfunded Mandates Reform Act
    D. Regulatory Flexibility

I. Regulated Entities

    Entities potentially regulated by this action are those that 
manufacture or introduce into commerce new small spark-ignition nonroad 
engines or equipment. Regulated categories and entities include:

------------------------------------------------------------------------
                                                Examples of regulated   
                 Category                             entities          
------------------------------------------------------------------------
Industry..................................  Manufacturers or importers  
                                             of new nonroad small (at or
                                             below 19 kW) spark-ignition
                                             engines and equipment.     
------------------------------------------------------------------------

    This table is not intended to be exhaustive, but rather provides a 
guide for readers regarding entities likely to be regulated by this 
action. This table lists the types of entities that EPA is now aware 
could potentially be regulated by this action. Other types of entities 
not listed in the table could also be regulated. To determine whether 
your company is regulated by this action, you

[[Page 3951]]

should carefully examine the applicability criteria in Sec. 90.1 of 
title 40 of the Code of Federal Regulations. If you have questions 
regarding the applicability of this action to a particular entity, 
consult the person listed in the preceding FOR FURTHER INFORMATION 
CONTACT section.

II. Legal Authority and Background

    Authority for the actions set forth in this rule is granted to EPA 
by sections 202, 203, 204, 205, 206, 207, 208, 209, 213, 215, 216, and 
301(a) of the Clean Air Act as amended (42 U.S.C. 7521, 7522, 7523, 
7524, 7525, 7541, 7542, 7543, 7547, 7549, 7550, and 7601(a)).
    In the summer of 1992, EPA initiated a convening process to 
determine the feasibility of a negotiated rulemaking for the 
development of the regulatory program for small nonroad spark-ignited 
(SI) engines at or below 19 kilowatts (hereafter referred to as ``small 
SI engines''). An August 1992 report recommended an ``Exploratory 
Meeting'' which was held November 1992. Following meetings in January 
and June 1993, the group decided to pursue a regulatory negotiation 
process for the development of Phase 2 regulations for these engines, 
while EPA developed a first phase of controls for small SI engines 
through the traditional rulemaking process.
    On July 3, 1995, EPA published the Phase 1 final rule, Emission 
Standards for New Nonroad Spark-ignition (SI) Engines At or Below 19 
Kilowatts, hereafter referred to as the Phase 1 small SI engine 
regulations.1 The Phase 1 small SI engine regulations 
established an effective date of model year 1997. Although the Phase 1 
regulations were the first to establish nationwide new engine emission 
standards for this industry, the federal regulations were developed to 
harmonize with the Tier I 2 standards established by 
California's Air Resources Board.3
---------------------------------------------------------------------------

    \1\ 60 FR 34582, July 3, 1995, codified at 40 CFR part 90. The 
docket for the Phase 1 small SI engine rulemaking, EPA Air Docket 
#A-93-25, is incorporated by reference.
    \2\ The California utility and lawn and garden equipment engine 
(utility engine) emission regulations are contained in Title 13, 
California Code of Regulations (CCR), Sections 2400-2407.
    \3\ Since the July 3, 1995 promulgation of the Phase 1 program, 
four changes have been made to Phase 1. First, provisions for 
allowing a streamlined certification process were promulgated May 8, 
1996, 61 FR 20738. Second, revisions to the national security 
exemption provisions were promulgated October 4, 1996, 61 FR 52088. 
Third, revisions to the carbon monoxide (CO) emission standards for 
Class I and II engines, and provisions related to crankcase 
emissions, were promulgated, November 13, 1996, 61 FR 58296. 
Finally, provisions relating to replacement engines and 2-stroke 
engines in nonhandheld applications were published August 7, 1997, 
62 FR 42637.
---------------------------------------------------------------------------

    The engines covered by the existing Phase 1 rule include 
nonhandheld engines (Class I and II) used in applications such as 
lawnmowers, generator sets and riding mowers, and handheld engines, 
(Class III, IV and V), used in applications such as trimmers, edgers, 
brush cutters, leaf blowers, leaf vacuums, chain saws, augers and 
tillers. The proposed Phase 2 rules contained in today's notice would 
apply to the same types of engines and applications covered by Phase 1.
    On September 30, 1993, the charter for the Small Nonroad Engine 
Negotiated Rulemaking Advisory Committee was filed with Congress. The 
purpose of the committee was to help EPA develop Phase 2 small SI 
engine regulations. The committee consisted of eleven members 
representing the range of stakeholders.4 The committee 
adopted protocols and formed four task groups to examine key issues and 
bring recommendations to the full committee. The task groups included: 
Test Procedure; Technology; Certification; and Public Education and 
Market Incentives.
---------------------------------------------------------------------------

    \4\ The organizations participating in the regulatory 
negotiations as members of the Committee were: the American Lung 
Association (ALA); the Auger and Power Equipment Manufacturers 
Association (APEMA); the Engine Manufacturers Association (EMA); the 
Manufacturers of Emission Controls Association (MECA); the Natural 
Resources Defense Counsel (NRDC); the North American Equipment 
Dealers Association (NAEDA); the Outdoor Power Equipment Institute 
(OPEI); the Portable Power Equipment Manufacturers Association 
(PPEMA); the State and Territorial Air Pollution Program 
Administrators/Association of Local Air Pollution Control Officials 
(STAPPA/ALAPCO); the Wisconsin Department of Natural Resources; and 
U.S. EPA.
---------------------------------------------------------------------------

    The committee and the task groups met numerous times between 
September 1993 and February 1996, with the final committee meeting on 
February 16, 1996, in Ann Arbor, Michigan. During the course of its 
work, the committee addressed many issues, including: applicability of 
the rule; engine/equipment classification; test procedures for engines; 
standards and standard structure; effective dates and lead time of the 
program; certification, enforcement and compliance strategies; in-use 
program; market-based incentive programs; public education programs; 
technologies; and dealer responsibility.
    The committee developed data and draft language to address most of 
these issues, both through the work of the task groups and the work of 
the committee as a whole. However, the committee did not reach 
consensus on an agreement in principle or draft regulatory language 
during the course of the negotiations. While the committee did not 
achieve consensus, the regulatory negotiation process produced 
substantial useful information and provided EPA with input from 
numerous key stakeholders which has helped EPA develop the Phase 2 
small SI engine regulatory program being proposed today.5 In 
addition, during the meetings there was much useful discussion which 
has helped EPA understand the perspectives of the interests represented 
at the table.6
---------------------------------------------------------------------------

    \5\ EPA initially established EPA Air Docket A-93-29 for the 
Phase 2 rulemaking; this docket contains background materials on 
this Phase 2 rulemaking, as well as materials related to the Small 
Nonroad Engine Negotiated Rulemaking process. EPA Air Docket A-93-29 
is hereby incorporated by reference.
    \6\ The final report by the facilitators to the regulatory 
negotiation process can be found in EPA Air Docket A-93-29, Item 
#II-A-10.
---------------------------------------------------------------------------

    Following the final meeting of the regulatory negotiation committee 
in February 1996, EPA proceeded to develop the Phase 2 rule. EPA and 
other interested parties continued working to find areas of agreement 
on how certain aspects of a Phase 2 program would be addressed in the 
proposed rule. As these discussions proceeded, the involved parties 
worked together to develop written documents, Statements of Principles 
(SOPs), which have partly formed the basis of today's Phase 2 NPRM (see 
62 FR 14740, March 27, 1997). A Statement of Principles (SOP) is a 
joint written statement by the U.S. EPA and supporting parties 
outlining a comprehensive plan for developing a proposed rulemaking. In 
this case, the two SOPs lay out the framework for a proposal for Phase 
2 regulations covering small handheld and nonhandheld spark-ignited 
nonroad engines, respectively.
    The ``Handheld SOP'', addressing issues affecting engines used in 
handheld equipment, was signed in May 1996 by EPA, the Auger and Power 
Equipment Manufacturers Association (APEMA), the North American 
Equipment Dealers Association (NAEDA), the Portable Power Equipment 
Manufacturers Association (PPEMA), the State and Territorial Air 
Pollution Program Administrators/Association of Local Air Pollution 
Control Officials (STAPPA/ALAPCO), and the Wisconsin Department of 
Natural Resources. The ``Nonhandheld SOP'', addressing issues affecting 
engines used in nonhandheld equipment, was signed in December 1996 by 
EPA, Briggs & Stratton Corporation, Kawasaki Motors Corporation, 
U.S.A., Kohler Company, Kubota, Mitsubishi Engine North America, Inc., 
Onan Corporation, Suzuki Motor Corporation, Tecumseh Products Company, 
The Toro Company,

[[Page 3952]]

and Wis-Con Total Power Corporation. While the two SOPs set out a 
framework for EPA's development of the proposed Phase 2 program, the 
Agency wishes to stress that they do not represent final decisions 
regarding Phase 2 or bind EPA as to how provisions in the final rule 
must be promulgated.
    EPA published an Advanced Notice of Proposed Rulemaking (ANPRM) in 
March 1997 (see 62 FR 14740, March 27, 1997) which announced the 
signing of the two SOPs and requested comments on all aspects of the 
SOPs for purposes of developing today's proposal. EPA also specifically 
requested information on small business issues in the ANPRM. 
Significant comments received on the ANPRM are discussed in the context 
of the description of the program contained in today's proposal.

III. Overview of Proposed Provisions

    EPA is proposing today a second phase of regulations for small SI 
engines 19 kW and below (hereafter referred to as small SI engines). 
Two principal goals of the proposed Phase 2 rule are to encourage a 
shift to cleaner engine technology, and to assure that the air quality 
benefits anticipated by the rule are achieved in actual use. To achieve 
these goals, the proposed Phase 2 program builds on the current Phase 1 
program in two key ways. First, today's proposal includes more 
stringent standards for hydrocarbons (HC) plus oxides of nitrogen 
(NOX) emissions, with a requirement that engines meet these 
emission standards through their useful lives.7 Second, the 
proposal adds an in-use component to the Phase 1 compliance program to 
assure that the emission benefits are achieved in actual use.
---------------------------------------------------------------------------

    \7\ EPA is proposing a set of values for the useful life of the 
engines for regulatory purposes. The term ``useful life'' refers to 
these regulatory useful life categories, which are discussed in more 
detail in Section IV.A.4 of this preamble.
---------------------------------------------------------------------------

    As is clear from the analysis supporting this proposed rule (see 
Sections V, VI and VII, and draft Regulatory Support Document), further 
emission reductions from future model year small SI engines beyond 
those achieved through the Phase 1 program can be achieved in a cost-
effective manner. Uncontrolled, small SI engines contribute 
approximately 3.4 percent of the national HC emission inventory, 9.3 
percent of the mobile source HC emission inventory, and 34.4 percent of 
the nonroad mobile source HC emission inventory.
    The Phase 1 small SI regulations are expected to reduce the HC 
emissions from these engines by 32 percent. However, even with Phase 1 
controls in place, small SI engines continue to contribute 
significantly to the emission inventory that leads to ozone 
concentrations in nonattainment areas. After Phase 1, small SI engines 
contribute approximately 3.1 percent HC nationally, 8.4 percent of 
mobile source HC, and 31.6 percent of the nonroad mobile source HC 
inventory (note that these values do not reflect changes in inventories 
from other sectors).
    In addition, further control of HC+NOX emissions from 
future model year small SI engines beyond Phase 1 levels, as proposed 
in today's notice for Phase 2 controls, is achievable through 
technology that will be available for the engines to which the 
standards would apply, considering cost, lead time noise, energy and 
safety factors. For nonhandheld engines, proposed Phase 2 emission 
levels are expected to be achieved through a combination of 
modifications to current engine technologies, and conversions to 
cleaner, more durable technology such as overhead valve engine 
technology. For handheld engines, proposed Phase 2 emission levels are 
expected to be achieved through improvements to current 2-stroke engine 
technologies (see discussion in Section IV.A of this preamble).
    If the Phase 2 program is adopted as proposed, many elements of the 
existing Phase 1 program would remain essentially the same in the Phase 
2 program. First, the types of engines covered by the proposed Phase 2 
rule would remain essentially the same as those covered in the Phase 1 
program (see discussion, Section IV.G). In addition, EPA would retain 
the five engine class categorization from Phase 1 for regulatory 
purposes as in Table 1 (see discussion, Section IV.G.3). Third, the 
Phase 1 criteria for determining whether an engine family would be 
allowed to certify to less stringent handheld standards would be 
retained (see Section IV.G.2).

                                        Table 1.--Small SI Engine Classes                                       
----------------------------------------------------------------------------------------------------------------
                     Nonhandheld                                               Handheld                         
----------------------------------------------------------------------------------------------------------------
             Class I                   Class II            Class III           Class IV             Class V     
----------------------------------------------------------------------------------------------------------------
<225 cc.........................  225 cc.  <20 cc............  20 cc    50 cc  
                                                                           and <50 cc.                          
----------------------------------------------------------------------------------------------------------------

    In addition, other elements of the existing Phase 1 program that 
would remain essentially unchanged in this proposed Phase 2 program 
include: (1) Applicability of the rule and definitions (see 40 CFR Part 
90, Subpart A), except as discussed in Section IV.G; (2) certification 
requirements (see 40 CFR Part 90, Subpart B), except for the proposed 
requirements to determine deterioration factors and to certify that 
engines meet the standards through their useful lives (see Section 
IV.D.1), and proposed flexibilities for small volume engine 
manufacturers (see Section IV.E); (3) provisions regarding test 
equipment and test procedures (see 40 CFR Part 90, Subparts D and E), 
except for minor changes addressed in Section IV.B; (4) provisions for 
selective enforcement audits (SEAs), (see 40 CFR Part 90, Subpart F), 
except that for the Phase 2 program SEA would exist primarily as a 
backstop to manufacturer-run production line testing program (see 
Section IV.D.2; and (5) provisions pertaining to importation of 
nonconforming engines, emission-related defect reporting requirements, 
voluntary emission recall program, exclusion and exemption of nonroad 
engines from regulations, prohibited acts and general enforcement 
provisions, and emission warranty and maintenance instructions (see 40 
CFR Part 90, Subparts G, I, J, K, and L), except for provisions for 
ordered recall (see proposed Sec. 90.808) and compliance flexibilities 
for small volume equipment manufacturers (see proposed Sec. 90.1003). 
EPA solicits comment on the appropriateness of retaining these elements 
of the Phase 1 program in Phase 2.
    Elements new to the regulatory requirements for small SI engines 
included in today's proposed Phase 2 program include: (1) proposed 
emission standard levels and useful life categories (see proposed 
amendments to Subpart B, and Section IV.A); (2) a certification

[[Page 3953]]

averaging, banking and trading program for nonhandheld engines (see 
proposed Subpart C, and Section IV.A.5); (3) procedures for the 
determination of deterioration factors at the time of certification 
(see proposed amendments to Subpart B, and Section IV.D.1; (4) a 
manufacturer-run production line testing program, called CumSum (see 
proposed Subpart H, and Section IV.D.2); and (5) in-use testing 
programs for nonhandheld and handheld engines, with an in-use credit 
program for handheld engines (see proposed Subparts M and N, and 
Section IV.D.3).
    In addition, this proposal contains a number of flexibilities to 
ease the transition to this more stringent Phase 2 program, some which 
would apply to all manufacturers, and others which would be targeted to 
ease the transition specifically for small production volume 
manufacturers (see discussion, Section IV.E). Finally, today's notice 
also describes EPA's intent to pursue a voluntary ``green labeling'' 
program and a voluntary fuel spillage reduction program for nonhandheld 
and handheld engines, and a particulate matter (PM) and hazardous air 
pollutant testing program for handheld engines (see Section IV.F).
    The programs proposed today for nonhandheld and handheld engines 
are similar in many respects. They also have some important 
differences. The intertwining issues of more stringent standards and 
assurance of emission reductions in use can be addressed in a number of 
ways. The remainder of this section provides an overview of the Phase 2 
program goals of encouraging a shift to cleaner technology and assuring 
that emission reductions are achieved in-use, and a description of the 
basic proposed programs for nonhandheld and handheld engines for 
achieving these goals.

A. More Stringent Standards and a Shift to Cleaner Technology

    EPA is proposing today HC+NOX emission standards for 
nonhandheld and handheld engines that are expected to achieve important 
reductions of emissions that contribute to ozone nonattainment. The 
standards for Classes II-V would be fully phased-in by the 2005 model 
year, with Class I levels effective in the 2001 model year. Engines 
would be required to meet these levels throughout their useful lives. 
For nonhandheld engines, a certification averaging, banking and trading 
program is proposed as an integral part of feasibility of the proposed 
HC+NOX emission standards (see Section IV.A.5). A more 
complete discussion of the justification of the level of the standards 
and the technologies expected to meet these levels can be found in 
Section IV.A. This section contains a brief overview of the proposed 
nonhandheld engine emission standards, the proposed handheld emission 
standards, and the proposal for useful life categories for nonhandheld 
and handheld engines.
1. Nonhandheld Engine HC+NOX Emission Standards
    The emission standards proposed today for nonhandheld engines, 
indicated in Table 2, represent an approximate 25 percent reduction in 
HC+NOX levels from Phase 1 levels. These standards are 
expected to be achieved in a cost-effective manner by modifications to 
current engine technologies and, especially in the case of Class II 
engines, by conversion of current side valve (SV) technology engines to 
cleaner, more durable technology, such as overhead valve (OHV) 
technology engines. For Class I, where engine sales are currently 
dominated by side-valve (SV) technology engines, the proposed levels 
are expected to result in cleaner and more emissions durable SV 
technology engines, but are not in themselves expected to result in 
conversion of SV engines to OHV or comparably clean and durable engine 
technology. These modifications to SV engines can be accommodated by 
2001, the proposed effective date for the Phase 2 standard for Class I 
engines. For Class II engines, the proposed levels are expected to 
result in complete conversion to clean OHV or comparable technology. To 
allow this more significant design change, the proposed Phase II 
standards are gradually decreased from 2001 through 2005.

               Table 2.--HC+NOX Emission Standards for Nonhandheld Engines in Grams/Kilowatt-Hour               
                                                   [g/kW-hr] 1                                                  
----------------------------------------------------------------------------------------------------------------
                                                  Model year   Model year   Model year   Model year   Model year
                  Engine class                       2001         2002         2003         2004         2005   
----------------------------------------------------------------------------------------------------------------
Class I........................................         25.0         25.0         25.0         25.0         25.0
Class II.......................................         18.0         16.6         15.0         13.6     \2\ 12.1
----------------------------------------------------------------------------------------------------------------
\1\ Optional non-methane hydrocarbon (NMHC) plus NOX emission standards for natural gas fueled engines only, and
  carbon monoxide (CO) emission standards, are also proposed in today's notice, and are discussed in Section    
  IV.A.                                                                                                         
\2\ The 12.1 g/kW-hr Class II standard assumes a phase-in from 50 percent in model year 2001 to 100 percent in  
  model year 2005 of OHV or comparably clean and durable technology.                                            

    A key aspect of the proposed Phase 2 program for nonhandheld 
engines is the belief that low emission standards for nonhandheld 
engines can be met through engine technology that can be low emitting 
both when the engine is new, and also when the engine has experienced 
hour accumulation to the engine's useful life. Therefore, these Phase 2 
standards are based on useful life emission performance.
    a. OHV and SV Engine Technologies. EPA believes that features 
inherent to the design of OHV technology engines are superior to those 
of SV engines and allow for lower new engine emissions as well as lower 
emission deterioration characteristics. In general, the combustion 
chamber and cylinder head design of OHV technology engines give these 
engines the potential to produce lower emissions both when new and also 
in-use. These engines have potential to exhibit lower emissions when 
new due to location of the combustion chamber directly over the piston, 
rather than partly to the side of the piston as in SV technology 
engines. This location allows a shorter combustion time, shorter flame 
propagation, better fuel combustion, and better cooling 
characteristics. In addition, OHV technology engines are designed with 
lower surface to volume ratios, which enhance fuel combustion. OHV 
technology engines also have the potential to exhibit improved in-use 
engine durability characteristics due to the location of the valves in 
the cylinder head rather than in the block, which affords more uniform 
exposure of the valves to heat sources and thus lower distortion of 
valves and valve seats. However, the Agency recognizes that the design 
of the engine is all-important,

[[Page 3954]]

and that it is possible to improve features of both SV and OHV 
technology engines to enhance new and in-use emission characteristics 
(e.g., cylinder heads, advanced carburetion, fuel injection). The 
Agency requests comment on the fundamental supposition of this rule 
that OHV technology engines have the potential to be superior to SV 
technology engines for new and in-use emissions characteristics. 
Further discussion of SV and OHV technology engines is contained in 
Section IV.A and Chapter 3 of the Draft Regulatory Support Document 
(RSD).
    b. Class I Use of OHV Technology. The nonhandheld small SI engine 
market has traditionally been dominated by SV technology engines, with 
SV technology engines accounting for as much as 90 percent of engine 
sales in Class I and 65 percent of engine sales in Class II. The 
majority of Class I SV engines are used in low cost, consumer products 
such as walk-behind mowers. Recently, the market has been moving 
towards OHV for Class II, in recognition of OHV advantages in engine 
performance, engine durability, fuel economy, and emissions 
characteristics. These advantages would be expected to be more 
important in commercial equipment which tend to make up significant 
market for Class II engines. For Class I engines, there has not been 
this same trend to OHV technology.
    One barrier to increased penetration of OHV technology engines into 
the Class I market, which is dominated by residential, low cost 
equipment, may have been the cost associated with the conversion of 
product lines from SV technology to OHV technology. These conversion 
costs to the engine manufacturer are expected to be in the range of $5 
to $14 per engine, depending on volume; cost to the consumer would 
likely be even higher (see Section VI for further discussion of these 
costs). For residential, low cost equipment, the OHV engine's 
advantages in performance and durability may not outweigh the 
associated higher purchase price when compared to equipment using less 
expensive SV equipment, at least in the near term and in light of the 
lead time EPA is proposing for the proposed Class I standard. If 
consumers of residential equipment are particularly price sensitive, 
they may choose not to purchase new equipment if priced higher due to 
the use of an OHV engine. Rather, to the extent four stroke SV engines 
tend to continue providing operable service, consumers may choose to 
spend money on equipment maintenance, extending both the life of the 
equipment and the number of hours the existing, non-Phase II SV engines 
would be used. If this happens, sales of cleaner, Phase II engines 
could be depressed and the extended use of SV engines toward the end of 
their useful life would add disproportionately to emission from small 
engines as the emission performance of these engines tends to continue 
deteriorating with use. Moreover, promulgation of a more stringent 
Class I standard, combined with the proposed Class II standard, would 
raise questions about the need for providing significantly longer lead 
time before the standards became effective. Additionally lead time 
might be necessary to allow manufacturers to invest the greater level 
of engineering and production resources necessary to convert both Class 
I and Class II engines to OHV technology for their entire product line 
as could be necessary for a nationwide program. This additional lead 
time could delay the environmental benefits of the program.
    Due to uncertainties as to consumer acceptance of OHV engines in 
typical Class I equipment applications if required nationwide and how a 
more stringent Class I standard might effect lead time for the program 
as a whole and the resulting uncertainty of emissions benefit, the 
Agency is not at this time proposing Class I standards which would 
mandate the conversion of Class I engines to OHV technology. However, 
EPA is requesting comments on the likely impacts of such a standard. 
Even if it is not appropriate to adopt more stringent Class I standards 
now, in the future, as uncertainties regarding consumer acceptance of 
OHV Class I engines and other issues are resolved, EPA will be able to 
re-evaluate the stringency of the proposed standard and pursue any 
necessary and appropriate revisions. Additionally, the experience in 
California will likely provide useful information.
    While today's proposed emission standard for Class I engines are 
not expected to require additional conversion from SV to OHV 
technology, EPA does desire to encourage the production and sale of OHV 
engines into the Class I market on a mass volume basis. In order to 
encourage this, EPA has entered into Memoranda of Understanding (MOUs) 
with two individual engine manufacturers.8-10 These two 
companies currently represent over 80 percent of all Class I engine 
sales. The two MOUs detail the specifics of Class I OHV engine 
demonstration programs which are designed as experiments to explore the 
consumer acceptance and feasibility of developing low cost OHV 
technology which can be applied to mass production Class I engines. The 
two programs include a series of reports to EPA on the level of 
success, impediments encountered, market response, costs, emission 
rates, and so forth. The two Class I OHV demonstration programs will 
begin prior to the proposed effective dates for the Phase 2 rule. While 
the MOUs are outside the scope of the regulatory process, if 
successful, this voluntary program may generate considerable emission 
benefits in addition to those anticipated to result from the proposed 
standards.
---------------------------------------------------------------------------

    \8-10\ Copies of these MOUs are in EPA Air Docket A-96-55, Items 
II-B-03 and II-B-04.
---------------------------------------------------------------------------

    In addition, the proposed voluntary ``green labeling'' program is 
designed to encourage manufacturers to produce engines that are 
substantially below the standards proposed today. In Class I in 
particular, manufacturers may decide for market reasons to convert 
current SV engines to OHV or comparably clean and durable technology 
engines, in order to qualify for the ``green label'' (see discussion of 
the program in Section IV.F.1).
    EPA requests comment on the general issue of the impact of moving 
to OHV technology for Class I engines, including the potential impact 
on sales of new equipment, the extended use of existing SV engines, the 
impact of a more stringent Class I standard on the ability of 
manufacturers to meet the proposed Class II standard under the proposed 
schedule, any options in addition to the voluntary ``green labeling'' 
program which would encourage the sale of clean OHV technology engines 
and the implications for emissions impact which would likely result 
from these actions.
    c. Class II Use of OHV Technology. The 12.1 g/kW-hr HC + 
NOX emission standard proposed to take effect in the 2005 
model year for Class II engines is expected to result in complete 
conversion to clean OHV or comparably clean and durable engine 
technology. As is discussed below in Section IV.A, this is an 
aggressive standard for Class II engines. The transition to OHV 
technology should be eased by the phase-in of the standard and the 
certification averaging, banking, and trading provisions proposed today 
for nonhandheld engines.
2. Handheld Engine HC+NOX Emission Standards
    The standards proposed today for handheld engines represent an 
approximate 35 percent reduction from Phase 1 levels, to be phased-in 
on a

[[Page 3955]]

percentage of production basis between the 2002 and 2005 model year, as 
indicated in Table 3. These standards are expected to be achieved in a 
cost-effective manner by use of improved 2-stroke technology engines 
(as discussed in more detail in Section IV.A).

                            Table 3.--HC+NOX Emission Standards for Handheld Engines                            
                                                  [In g/kW-hr]                                                  
----------------------------------------------------------------------------------------------------------------
                                                    HC+NOX                                                      
                                                   emission    Model year   Model year   Model year   Model year
                  Engine class                   standard (g/     2002         2003         2004         2005   
                                                    kW-hr)     (percent)    (percent)    (percent)    (percent) 
----------------------------------------------------------------------------------------------------------------
Class III......................................          210                                                    
Class IV.......................................          172           20           40           70        100 1
Class V........................................         116                                                     
----------------------------------------------------------------------------------------------------------------
1 The standards would be phased-in on the basis of percentage of total eligible sales. In this proposed rule,   
  ``eligible sales'' or ``U.S. sales'' is defined as Phase 2 engines sold for purposes of being used in the     
  United States, and includes any engine exported and subsequently imported in a new piece of equipment, but    
  excludes any engine introduced into commerce, by itself or in a piece of equipment, for use in a state that   
  has established its own emission requirements applicable to such engines pursuant to a waiver granted by EPA  
  under section 209(e) of the Clean Air Act.                                                                    

    Two-stroke technology engines have traditionally been the dominant 
engine design used for handheld equipment applications. These engines 
have been well suited to meet the weight, multipositional use, and 
power requirements of these applications. However, 2-stroke technology 
engines also have very high engine emissions, compared with 4-stroke 
technologies, due in large part to fuel scavenging losses.
    With the advent of emission control requirements federally and in 
California, research into other technologies to further control 
emissions from engines used in handheld applications has occurred. 
Promising technologies include light weight 4-stroke technology 
engines, and 2-stroke technology engines with aftertreatment. However, 
little is known about the in-use performance, in-use emissions 
characteristics and cost of these technologies, or how appropriate it 
is to consider these technologies across the full range of handheld 
equipment applications. Because of these uncertainties, today's 
standards would not require conversion to 4-stroke engine technology or 
the use of aftertreatment for handheld engines. However, EPA wants to 
encourage introduction of technologies into today's market which are 
cleaner than required by the proposed standards. For example, EPA 
recognizes that some engine manufacturers have recently developed and 
marketed cleaner, lightweight 4-stroke engines for use in handheld 
equipment. The Agency believes potentially cleaner 4-stroke engines, 2-
stroke engines with aftertreatment and other advanced two-stroke 
technologies may enter the market to a limited extent on a national 
level during the time frame of the Phase 2 program. EPA's goal is to 
encourage development of such technology, and EPA believes that the 
proposed ``green labeling'' program, (discussed in Section IV.F.1) 
should provide important incentives to manufacturers to introduce 
cleaner technologies on a national basis. In addition, the Agency 
intends to conduct a technology review and a possible Phase 3 
rulemaking to address the possibility that technological advances and/
or cost reductions may occur after promulgation of the Phase 2 rule 
that could make greater, but still cost-effective reductions feasible 
in handheld engine emission levels.
3. Useful Life Categories
    Today's proposal would require that engines meet the proposed 
emission standards throughout their useful lives. EPA is today 
proposing multiple useful life categories, indicated in Tables 4 and 5, 
given the numerous applications in which these engines are used, and 
wide variation in expected engine useful life in these different 
applications. In addition, the use of these engines in applications 
which experience primarily commercial rather than primarily consumer or 
residential usage can also impact the useful life of the engine.

        Table 4.--Useful Life Categories for Nonhandheld Engines        
                                 [Hours]                                
------------------------------------------------------------------------
                                          Category   Category   Category
                                             C          B          A    
------------------------------------------------------------------------
Class I................................         66        250        500
Class II...............................        250        500       1000
------------------------------------------------------------------------


          Table 5.--Useful Life Categories for Handheld Engines         
                                 [Hours]                                
------------------------------------------------------------------------
                                                Residential   Commercial
------------------------------------------------------------------------
Class III, IV and V...........................           50          300
------------------------------------------------------------------------

    EPA is proposing that at the time of certification, engine 
manufacturers would have the responsibility to select the useful life 
period which most typically represents the in-use operating periods for 
the majority of engines in the engine family, based on information 
about that engine family including design and durability information, 
as well as information about the equipment in which the engine is 
expected to be used. Manufacturers would label the engine according to 
the useful life selection. See Section IV.A.4 for further discussion of 
the proposed useful life provisions for nonhandheld and handheld 
engines.

B. Assuring Emission Reductions are Achieved In-use

    The goal of the in-use component of the proposed Phase 2 program is 
to provide assurance that the emission reduction benefits anticipated 
by the program are achieved in actual use. This section describes how 
EPA's traditional compliance programs for mobile sources achieve this 
goal, outlines various challenges in designing a compliance program for 
the small SI industry, provides an overview of the compliance program 
proposed today for nonhandheld and handheld engines, and discusses 
alternative compliance program options.

[[Page 3956]]

1. Traditional Compliance Programs for Mobile Sources
    EPA has traditionally used three-step compliance programs to 
implement and enforce mobile source emission standards. For a given 
engine family, the first of the three steps is certification, where, 
based on emission data from test engines, which are often prototype 
engines, EPA issues a license to the engine manufacturer known as a 
certificate of conformity. This license enables the manufacturer to 
introduce engines covered under the certificate into commerce in the 
United States. This step typically includes some means of projecting 
the emissions characteristics of the engine family over its useful 
life. If the manufacturer demonstrates according to the regulatory 
provisions that the engine family meets the emission standards for the 
useful life of the engines, EPA issues a certificate of conformity.
    The second step is production line testing where the engine 
manufacturer demonstrates that actual production line engines meet 
emission standards. Production line testing provides an opportunity for 
EPA and the manufacturer to verify that designs approved based on 
certification testing are translated into mass production engines that 
meet standards and to catch production problems before they become in-
use problems.
    The last step involves the testing of in-use engines to ascertain 
whether the engines continue to meet standards during their useful 
lives in the hands of typical customers. EPA has the authority under 
Section 207(c) of the Clean Air Act to require a mandatory recall of 
vehicles or engines that have been shown not to comply with standards 
for their useful life. Such recalls are instigated based on evidence of 
nonconformities discovered through a variety of means, the most common 
of which are cases in which nonconformities are found either through 
production line testing or through in-use testing programs. In EPA's 
on-highway emission control programs, EPA's recall authority and recall 
practices have provided clear incentives to manufacturers to produce 
emissions durable engines and vehicles.
2. Compliance Programs for the Small SI Engine Industry
    The Phase 1 emission control program for small SI engines does not 
follow this typical three-step compliance program. This is because, 
unlike other programs, the Phase 1 program includes ``new engine'' 
standards only, that is, standards that the engines must meet when new, 
without the requirement that they continue to meet those standards in-
use throughout their useful lives. As such, while the Phase 1 program 
contains programs for certification and production line testing (in the 
form of EPA initiated Selective Enforcement Audits), the program does 
not contain a requirement for manufacturers to project the emissions 
characteristics of the engine family over its useful life at the time 
of certification (e.g., to determine a deterioration factor, or ``df'', 
for the engine family), nor does it contain mandatory in-use testing 
provisions. EPA promulgated such a program for Phase 1 for several 
reasons, including the belief that for a first phase of emission 
controls, significant emission reductions would occur in this sector 
even with the ``new engine'' standards. Equally important was the lack 
of data available to the Agency at the time of the rulemaking on which 
to base an in-use program (e.g., information supporting appropriate 
regulatory useful life periods and engine deterioration rates). In 
addition, EPA made clear its intention to address in-use issues in a 
second Phase of regulation.
    In addition to determining appropriate useful life periods and 
engine emission deterioration characteristics for this proposed Phase 2 
program, the Agency has also faced a key challenge of how to conduct an 
effective in-use testing program for these engines, and whether or not 
a recall program modeled on the traditional on-highway recall program 
could be an effective compliance tool for this sector of the nonroad 
engine industry. As EPA has begun to regulate a wide range of nonroad 
engines pursuant to Section 213 of the Clean Air Act, it has become 
evident that a mandatory recall program, as has been traditionally 
conducted for the on-highway industry, may not be the most effective 
program for some sectors of the nonroad engine industry, as compared 
with other means of assuring compliance in-use. This is especially true 
for the small SI engine industry, in which many of the engines are 
installed in consumer products which are not registered and thus would 
be difficult to track in the event of a recall, and in which the cost 
of conducting a potential recall could be large relative to the cost of 
the actual engines being recalled.
    For certain nonroad engine industry sectors, such as the spark-
ignition marine engine sector and the small SI engine sector, EPA has 
sought to develop alternative programs designed to provide reasonable 
means to address emissions exceedances identified through production 
line testing and in-use testing programs. For example, the spark-
ignition marine engine program includes a voluntary in-use credit 
program that EPA expects will be an effective way to address 
exceedances identified through in-use testing, and the program also 
includes provisions for the use of certification credits to address 
exceedances identified through production line testing (see 40 CFR Part 
91).
    EPA believes that these alternative programs, designed to provide a 
means to address emission exceedances, should meet several criteria in 
order to be considered as effective as EPA's traditional mandatory 
recall programs. First, they should provide an incentive to 
manufacturers to build emission-durable engines. Second, they should be 
practical to implement. Third, they should provide an incentive to 
perform accurate testing. Fourth, such programs should offset 
additional emissions that occur as a result of the exceedence of the 
standards. Finally, such programs should not be unduly burdensome to 
manufacturers.
    The compliance programs proposed today for small SI nonhandheld and 
handheld engines are intended to meet these criteria. While EPA retains 
the authority to order a recall if a substantial number of engines are 
found to be in nonconformity, and while this Phase 2 proposal does 
include regulatory language governing EPA's action in ordering recalls 
(see proposed Subparts I and M), EPA anticipates considering programs 
which would be effective alternatives to ordering a mandatory recall of 
Phase 2 certified engines. Instead, EPA would expect these alternatives 
to recall would address the exceedances of the emission standards in 
ways that meet the five criteria identified above. For nonhandheld 
engines, in some cases, the use of certification credits would be 
allowed to offset exceedances of the family emission limit 
11, 12 in the event of PLT exceedances. For handheld 
engines, the use of in-use credits would be allowed as one means of 
addressing potential exceedances of standards in the event of 
exceedances determined through production line testing or in-use 
testing programs. For both nonhandheld and handheld engines, other 
possible alternatives for addressing exceedances of emissions standards 
would include voluntary recall and other possible alternative projects 
(these issues are discussed

[[Page 3957]]

further in Section IV.D of this preamble).
---------------------------------------------------------------------------

    \11, 12\ For nonhandheld engines participating in the averaging, 
banking, and trading program described in more detail in Section 
IV.A.5, compliance would be demonstrated with the family emission 
limit, or FEL, rather than the standard.
---------------------------------------------------------------------------

3. The Proposed Phase 2 Compliance Program
    Today's program proposes ``in-use'' standards for the first time 
for this industry.13 New elements of the Phase 2 compliance 
program include processes for determining deterioration factors 
(``dfs'') at the time of certification, a manufacturer-run Production 
Line Testing program, and in-use testing components.
---------------------------------------------------------------------------

    \13\ The fact that the proposed Phase 2 emissions standards are 
``in-use'' standards, compared with the Phase 1 standards which are 
``new engine'' standards, together with the fact that these engines 
do experience emissions deterioration over time, is why, when 
compared numerically with the Phase 1 levels, Phase 2 levels in fact 
are higher in the case of Class I. Despite this apparent numerical 
discrepancy, EPA still anticipates important reductions from all 
engine classes as a result of the proposed Phase 2 standards. Since 
Phase 2 designs will account for in-use deterioration, in-use 
emission levels will be lower under the proposed Phase 2 regulations 
compared to Phase 1 engines.
---------------------------------------------------------------------------

    i. Certification and In-Use Testing. Today's proposal includes 
three different approaches to certification df determination and in-use 
testing, based on engine class and engine technology, which are 
discussed briefly below. These approaches comprise the basic program 
proposed today. EPA is also proposing additional procedures for some 
engine classes and engine technologies to increase the flexibility of 
the rule.14 All the approaches are discussed in more detail 
in Section IV.D.
---------------------------------------------------------------------------

    \14\ For example, for nonhandheld OHV technology engines, 
manufacturers would have an option to use a ``calculated df'' rather 
than the ``assigned df'' described below.
---------------------------------------------------------------------------

    First, for nonhandheld OHV technology engines, manufacturers would 
be allowed to apply an assigned deterioration factor or ``assigned df'' 
to new engine test values at the time of certification to determine a 
useful life certification value. Compared to an alternative of testing 
an engine over its full useful life to determine deterioration, these 
engines would be allowed to undergo this lower burden certification 
effort, in return for participation in an industry-wide OHV field 
durability and in-use emission performance demonstration program (as 
described in Sections IV.D.1 and IV.D.3). Second, for nonhandheld side-
valve technology engines and engines with aftertreatment, manufacturers 
would certify their engines based on accumulating hours on the engines 
to the engines' full useful lives at the time of certification. This 
relatively heavier burden at the time of certification is balanced by a 
decreased in-use testing burden. Following full useful life 
certification, these engines would not be subject to further in-use 
testing requirements. Third, for all handheld engines, manufacturers 
would certify their engines to full useful life standards at the time 
of certification using new engine test values and dfs determined based 
on ``good engineering judgment.'' Handheld engine manufacturers would 
then conduct an in-use testing program, by which each manufacturer 
would age and emissions test engines to ensure compliance in-use. A 
handheld engine manufacturer would in-use test up to 25 percent of its 
engine families each year.
    Other than the addition of the requirements to demonstrate that 
engines meet the emission standards throughout their useful lives, and 
to determine a deterioration factor at the time of certification, the 
certification procedures proposed today for the Phase 2 program are 
essentially the same as those for Phase 1. In particular, EPA is 
proposing to retain a streamlined certification application form and 
process, with simple procedures for electronic submittal of 
information, as discussed further in Section IV.D.1.
    ii. Production Line Compliance. Today's proposal would add a 
manufacturer-run Production Line Testing program known as CumSum to 
replace a Selective Enforcement Audit (SEA) program as the primary 
method of determining the compliance of new production engines. SEA 
would remain an optional or backstop program depending upon the class 
of engine, as described in Section IV.D.2.
    iii. Aging Engines To Their Useful Lives. EPA believes that aging 
engines in field usage in typical representative applications would be 
the most accurate possible program for verifying in-use emissions. As 
such, the proposed OHV field durability and in-use emissions 
performance program (``Field Durability Program'') is designed to 
produce significant quantities of reliable test data from OHV engines 
aged in typical field usage, and to verify that the conclusions used in 
the certification process with respect to the durability of OHV engines 
are accurate.
    While aging engines in typical field usage would be the optimal 
program for assuring the emission reductions are being achieved in use, 
EPA recognizes that costs associated with aging engines in the field 
and administering a field aging program could be higher than, for 
example, costs of a bench aging program. It is for this reason that EPA 
is proposing that for full useful life certification for nonhandheld 
side-valve technology engines or engines with aftertreatment, and for 
in-use testing for handheld engines, manufacturers may age engines on 
bench cycles, in lieu of field aging, provided that a field/bench 
adjustment factor has previously been established, as discussed in 
Section IV.C. EPA requests comment on the proposal to allow 
manufacturers in some cases to age engines on bench cycles in lieu of 
field aging.
    In addition, for nonhandheld engine manufacturers, who could be 
field aging engines for the OHV Field Durability Program and also for 
the field/bench adjustment program, EPA is proposing a cap on the 
number of field engine tests required in a given year. EPA requests 
comments on all aspects of the compliance program proposed today for 
Phase 2 small SI engine regulation.
4. Alternative Compliance Program Options
    The program proposed today for Phase 2 regulation of small SI 
engines is essentially the same as the program described in the ANPRM 
for this rulemaking. EPA received comments on the ANPRM relating to the 
differences between the nonhandheld and handheld sides of the industry, 
and the merits of applying concepts and programs outlined for one side 
of the industry to the other. One commenter stressed that the 
nonhandheld and handheld engine industries are very different in 
composition, in marketing, in technology, as well as in application. 
This commenter suggested that the program for nonhandheld engines 
described in the ANPRM is an integrated whole, with each provision 
linked to other provisions, and that it would be a mistake to graft 
parts of the handheld program on to the nonhandheld program. Another 
commenter suggested that the Agency should take a comprehensive and 
balanced view of the program for the two sides of the industry, and 
that elements of the two proposals should be used to create a simpler 
and more effective regulation.
    EPA is concerned that any changes to the programs being proposed 
today should be considered carefully as to their impact on the program 
as a whole, given linkages between the various elements of the programs 
proposed today. For example, the compliance program proposed for 
nonhandheld OHV technology engines is designed as an integrated whole. 
The proposal to allow manufacturers to use the assigned dfs for 
certification is reasonable because it is linked to the proposal for an 
industry-wide OHV Field Durability Program designed to verify the 
assumptions with respect to stable and low dfs. In addition, EPA 
believes this

[[Page 3958]]

conversion of engines to OHV or comparably clean and durable 
technology, together with the OHV Field Durability Program, is one of 
the strongest elements of today's proposal, an element which links 
stringent standards forcing clean technology with a field testing 
program to verify that those emission reductions are being achieved in 
use.
    However, EPA believes that there are multiple ways to design 
effective programs for reducing emissions from small SI engines, and 
for ensuring that those reductions are achieved in use. EPA requests 
comment on alternative compliance options. For example, EPA requests 
comment on an option which would allow nonhandheld manufacturers to 
establish certification dfs for SV engines and engines with 
aftertreatment through good engineering judgment (instead of the 
proposed program for full useful life aging for certification), linked 
to a program for field aging SV engines and engines with aftertreatment 
to verify the dfs established through good engineering judgment. EPA 
also requests comment on applying the in-use testing program proposed 
today for handheld engines to the nonhandheld side of the industry. EPA 
requests comments on these or other ways in which programs for the two 
sides of the industry could be designed to achieve the goals of 
providing assurance of environmental benefits in-use, easing the 
implementation burden for EPA and the industry, and achieving greater 
commonality in the programs for the two sides of the industry, where 
appropriate.

IV. Description of Proposed Program

    Section IV of today's document contains a description of the 
programs proposed for nonhandheld and handheld small SI engines for 
Phase 2 regulations, including discussion of standards and related 
provisions, test procedures, a field/bench adjustment program, 
compliance programs, flexibilities, nonregulatory programs, and other 
general provisions.

A. Standards and Related Provisions

    This section provides a detailed discussion of the standards being 
proposed for the Phase 2 program, as well as related provisions 
including useful life categories, certification averaging, banking, and 
trading provisions, and certification fuel.
    The Agency is aware of the levels which the California Air 
Resources Board (CARB) is considering for their Tier 2 standards for 
their Utility, Lawn, and Garden Engine regulation. The CARB Tier 2 
levels are more stringent and occur in a shorter time frame than the 
levels being proposed by the Agency for a Federal Phase 2 program. 
Although EPA's approach is not structured identically with CARB 
regulations, EPA believes there are two valid reasons for the 
distinction. First, Congress has recognized the need for California to 
maintain its own mobile source emission control program (see section 
209 of the CAA) because it faces difficult and distinct air pollution 
problems and, as a result, may need to adopt measures more stringent 
than those that apply in the nation as a whole (see, e.g., Motor & 
Equipment Manufacturers Association v. EPA, 627 F.2d 1095, 1110-11 
(D.C. Cir. 1979)). Second, EPA's nonroad emission standards are not 
allowed to be more stringent than is achievable for this nationwide 
program after consideration of cost and lead time according to section 
213(a)(3) of the CAA. Although California is constrained by similar 
criteria per the authorization criteria of section 209(e), 
consideration of such criteria is limited to the State of California. 
The Agency must consider cost and lead time when nonroad emission 
regulations affect the nation as a whole. As discussed in the remainder 
of this section, the Agency believes the standards contained in today's 
proposal meet the section 213(a)(3) requirements to consider cost and 
lead time in setting Federal standards.
1. HC+NOX Emission Standards
    The Agency believes the level of the standards contained in today's 
proposal would achieve the greatest degree of emission reduction 
achievable through application of technology which will be available 
and considering lead time under the proposed schedule of compliance, 
noise, energy, safety, and cost factors associated with applying such 
technology to a nationwide program. The sections below discuss how EPA 
addressed and weighed these factors in developing the proposed 
standards.
    EPA is proposing in-use HC+NOX standards of 25 g/kW-hr 
effective in model year 2001 for Class I engines, and 12.1 g/kW-hr to 
be phased-in between model years 2001 and 2005 for Class II engines, as 
presented in Table 6. EPA expects that the Class II levels would result 
in a complete shift in engine technology from side-valve (SV) to 
cleaner overhead valve (OHV) or comparably clean and durable technology 
by 2005.

                           Table 6. HC+NOX Emission Standards for Nonhandheld Engines                           
                                                  [In g/kW-hr]                                                  
----------------------------------------------------------------------------------------------------------------
                                                  Model year   Model year   Model year   Model year   Model year
                  Engine class                       2001         2002         2003         2004         2005   
----------------------------------------------------------------------------------------------------------------
Class I........................................         25.0         25.0         25.0         25.0         25.0
Class II.......................................         18.0         16.6         15.0         13.6         12.1
----------------------------------------------------------------------------------------------------------------

    EPA is proposing in-use HC+NOX emissions levels for 
Class III, IV and V engines to be phased-in between model years 2002 
and 2005 based on a percentage of U.S. sales as presented in Table 7.

                            Table 7.--HC+NOX Emission Standards for Handheld Engines                            
                                                  [In g/kW-hr]                                                  
----------------------------------------------------------------------------------------------------------------
                                                    HC+NOX                                                      
                                                   emission    Model year   Model year   Model year   Model year
                  Engine class                   standard (g/     2002         2003         2004         2005   
                                                    kW-hr)     (percent)    (percent)    (percent)    (percent) 
----------------------------------------------------------------------------------------------------------------
Class III......................................          210                                                    
Class IV.......................................          172           20           40           70          100

[[Page 3959]]

                                                                                                                
Class V........................................          116                                                    
----------------------------------------------------------------------------------------------------------------

    Unlike the nonhandheld Phase 2 program, for handheld engines, the 
phase-in process of mandatory percentages would result in Phase 1 and 
Phase 2 handheld engines being produced in the same model year, i.e., 
at least 20 percent of the engines produced in model year 2002 would be 
Phase 2 engines subject to the Phase 2 program, and up to 80 percent of 
the handheld engines produced in model year 2002 would be Phase 1 
engines subject to the Phase 1 program, followed by a 40/60 split in 
model year 2003, and a 70/30 split between Phase 2/Phase 1 engines in 
model year 2004.
    The remainder of this section describes the analysis and supporting 
data for the proposed HC+NOX standards for Class I 
nonhandheld engines, Class II nonhandheld engines, and Class III, IV, 
and V handheld engines. Each of these subsections is organized into the 
following topics: (i) Historical Sales Trends by Engine Technology--
Historical trends are important to consider when assessing the range of 
field proven technologies. Historical trends assist in understanding 
what technologies have been demonstrated in actual use, what 
manufacturers' current production capabilities are, and the 
availability of new and in-use emission performance data; (ii) In-use 
HC and NOX Emission Performance of Uncontrolled Engines--The 
Agency presents this information to highlight the in-use performance 
characteristics associated with small engine technologies and the need 
for careful consideration of the in-use performance of various control 
technologies. Phase 1 new engine emission performance data is available 
from Federal certification data. However, in-use emission performance 
on engines pulled from the field is limited; therefore, a discussion of 
the in-use performance of uncontrolled engines is warranted; (iii) New 
Engine and In-use HC and NOX Performance of Phase 1 
Technology Engines--A summary of the information available on the new 
and in-use emission performance of Phase 1 engines is presented. This 
information is used to assess the current status of the small engine 
industry, which is critical for the Agency's analysis when trying to 
predict the impact of technology changes on the industry; (iv) 
Technologies Considered for Phase 2 HC+NOX Standards--
Discussion of the technologies the Agency considered when determining 
the level of the proposed standards is presented. This includes a 
discussion of new and in-use emission performance of each technology, 
and the per engine cost associated with each technology, and; (v) 
Proposed Phase 2 HC+NOX Standard--A discussion of the Phase 
2 standards the Agency is proposing, including information on why the 
proposed standards are achievable, the proposed lead time, and a 
discussion and request for comment on more stringent standards (such as 
the CARB Tier 2 levels).
    a. HC+NOX Emission Standard for Class I Nonhandheld 
Engines. This section presents information used by the Agency to 
determine the appropriate level for the proposed HC+NOX 
exhaust emission standards for nonhandheld Class I engines. A more 
detailed explanation of the engine technologies and costs described in 
this section is contained in the Draft Regulatory Support Document 
(RSD) for this proposal, a copy of which is available in the public 
docket for this rule.
i. Class I Historical Sales Trends by Engine Technology
    Class I engine (<225 cc nonhandheld engines) sales have 
historically been dominated by low cost four-stroke side-valve engines. 
Two-stroke gasoline Class I engines are currently less than 10 percent 
of annual sales and will continue to decline as a result of the Phase 1 
emission standards, which effectively calls for their phase-out by 2003 
due to their high HC emissions. Prior to 1986, OHV engines represented 
less than one percent of annual Class I engine sales. In the past 
decade OHV engines have begun to penetrate the Class I marketplace, but 
they have hovered between 10 and 15 percent of total U.S. sales for the 
past eight years.
ii. In-use HC and NOX Emission Performance of Uncontrolled 
Class I Engines
    Unregulated Class I engines have demonstrated high new engine 
emission rates for HC and CO, and low levels of NOX, as well 
as poor in-use performance (large deterioration factors) for HC and CO, 
with little deterioration of new engine NOX 
values.15 HC deterioration has been shown to be greater than 
two times the new engine value in as little as four years of engine 
use.
---------------------------------------------------------------------------

    \15\ See ``Emission Tests of In-use Small Utility Engines'' 
Southwest Research Institute, Sept. 1991, EPA Air Docket A-91-24, 
Item #II-A-8, and ``Nonroad Engine and Vehicle Emission Study'' U.S. 
EPA Report #21A-2001, Nov. 1991, EPA Air Docket A-91-24, Item #II-A-
10.
---------------------------------------------------------------------------

iii. New Engine and In-use HC and NOX Performance of Phase 1 
Class I Technology Engines
    Phase 1 engines have improved new engine emission performance over 
uncontrolled engines, and may have improved in-use performance. The 
Draft RSD for this proposal contains publicly available information on 
engine families from all engine classes certified to the Phase 1 
program. This information shows both SV and OHV technology can meet the 
Phase 1 Class I new engine standard.
    The Agency has recently examined information presented by several 
engine manufacturers concerning emissions deterioration from Phase 1 
technology Class I side-valve and over-head valve engines.16 
A more detailed discussion of this data is presented in the Draft RSD. 
This information covers over 50 Class I engines field aged by 
manufacturers, with usage varying from 20 to 300 hours. Table 8 
contains a summary of the HC+NOX deterioration factors 
resulting from an analysis of this data.
---------------------------------------------------------------------------

    \16\ See ``Tier 1 Deterioration Factors for Small Nonroad 
Engines'', Sept. 1996, a report by Air Improvement Resources, 
available in EPA Air Docket A-96-55, Item #II-D-11.

[[Page 3960]]



 Table 8.--Summary of In-use Deterioration of Phase 1 Technology Class I
                                 Engines                                
------------------------------------------------------------------------
                                                     Class I    Class I 
                                                       OHV         SV   
------------------------------------------------------------------------
Estimated HC+NOX df at 66 Hours...................       1.35       1.87
------------------------------------------------------------------------

    Analysis of this information indicates Class I SV HC+NOX 
deterioration is higher than Class I OHV engines. The lower new engine 
emission levels of Class I OHV over SVs combined with lower in-use 
deterioration results in better in-use emission performance for Class I 
OHV engines compared to Class I SV engines.
iv. Technologies Considered for Phase 2 Class I HC+NOX 
Standards
    The Agency analyzed the emission performance and cost of several 
technologies which could be applied to Class I engines, including 
improvements to existing SV engines, conversion of existing SV engines 
to OHV technology, and the application of catalytic converters to 
existing SV and OHV engines. Four-stroke SV technology utilizes an 
engine configuration in which the intake and exhaust valves are located 
to one side of the combustion chamber (also called an L-head design), 
as compared to four-stroke OHV technology in which the intake and 
exhaust valves are located directly above the combustion chamber. 
Catalytic converters are add-on after treatment devices which operate 
by chemically reducing or oxidizing exhaust gases. The Draft RSD for 
this proposal contains additional information regarding these three 
technologies.
    As discussed previously, the majority of Class I engines utilize SV 
technology. Table 8 shows that Class I SV technology have 
HC+NOX deteriorations on the order of 1.87 times new engine 
levels at 66 hours of use. Combining this with the Phase 1 
certification level of 16.1 g/kW-hr HC+NOX indicates an in-
use level of approximately 30 g/kW-hr HC+NOX. The Agency 
believes additional reductions can be achieved with improvements to 
existing Phase 1 SV engines. A more detailed discussion of these 
improvements is contained in the Draft RSD. A summary of the 
improvements are: lowering of new engine emission levels achieved 
through enleanment of intake air-fuel ratio; improvements to valve seat 
material which will lower in-use distortion, resulting in decreased 
valve leakage and deposit formation; improvements in cylinder ring 
design, which will result in better combustion chamber sealing and 
lower oil consumption and lower combustion chamber deposits; continued 
structural improvements to cylinder design to lower cylinder distortion 
inherent in side-valve configurations; and addition of valve stem seals 
to limit the creepage of oil into the combustion chamber. As presented 
in the Draft RSD, the Agency estimates the improvements to Class I SV 
engines would cost the manufacturer as much as $4 to $7 per engine, 
depending on the engine family volume. The Agency estimates changes 
would result in improvements to both new and in-use emission 
performance, combining for a 10 to 20 percent improvement in the in-use 
HC+NOX performance beyond Phase 1 designs.
    As indicated by Table 8, Phase 1 OHV engines have better in-use 
performance compared to Phase 1 SV engines. A new engine level equal to 
the Phase 1 standard of 16.1 g/kW-hr combined with a HC+NOX 
df of 1.35 at 66 hours results in an in-use emission rate of 21.7 g/kW-
hr. This level is well below the performance of Class I SV engines, 
therefore the Agency has considered the conversion of existing Class I 
SV to OHV engines in developing the proposed Phase 2 levels. Based on 
the Federal Phase 1 new engine certification data analyzed for this 
proposal, the average Class I OHV engine emits around 10.5 g/kW-hr. 
Based on the deterioration information presented in Table 8 and design 
improvements discussed elsewhere, the Agency estimates a well designed 
nonhandheld OHV engine could have an HC+NOX deterioration 
factor of 1.3. Assuming a 10 percent compliance margin, these specific 
Class I OHV engines could achieve an average in-use emission level of 
around 15 g/kW-hr. However, it should be noted that only about 10 
percent of current Class I engines are OHV designs. The performance of 
these specific engines may not be representative of what would occur if 
all Class I engines were converted to OHV technology.
    Federal certification data indicates a small number of Class I 
engines have certified to the Federal Phase 1 standards using catalyst 
technology. Though it is technologically feasible to apply catalysts to 
both SV and OHV engines, the Agency has little information regarding 
in-use durability and emission performance of engines equipped with 
catalysts. As discussed previously, the in-use emission performance of 
small engines is a critical component of the analysis EPA has 
undertaken in the development of the Phase 2 proposal. The Agency's 
experience with on-highway catalyst technology has shown considerable 
in-use deterioration of catalysts can occur. In recent years several 
technical papers have been published regarding catalyst durability on 
small engines, however, these papers have relied on laboratory 
durability programs, such as aging catalysts on dynamometers 
17. The Agency is not aware of any actual field-aged in-use 
catalyst durability information. The Agency requests comment on the 
relationship between laboratory durability data and in-use field data, 
any information on typical in-use aged catalyst performance, and all 
available data on individual catalysts aged under typical in-use 
conditions experienced by equipment using Class I engines. The Agency 
requests additional information regarding new engine emission 
performance, in-use emission performance, and cost of catalyst 
technology for Class I SV and OHV engines.
---------------------------------------------------------------------------

    \17\ See Society of Automotive Engineers Technical Papers 
930076, 932445, 941807, and 961735 for bench aged catalyst 
information.
---------------------------------------------------------------------------

v. Proposed Phase 2 Class I HC+NOX Standard
    The Agency is proposing a corporate average exhaust emission level 
of 25 g/kW-hr HC+NOX for Class I engines beginning in model 
year 2001 (for discussion of the averaging, banking, and trading 
program, see Section IV.A.5). The Agency believes this level is 
technologically achievable, and, as discussed previously, can be met by 
improvements to existing Class I SV engines. The Agency has performed 
an analysis using the existing Phase 1 certification data (which 
contains confidential sales projections) combined with reasonable 
assumptions for in-use deterioration. This analysis indicates an 
averaging standard of 25 g/kW-hr is achievable with improvements to 
existing SV engines and considering the emission performance of 
existing Phase 1 OHV engines. A standard of 25 g/kW-hr would not 
require an increase in the penetration of Class I OHV sales. 
Manufacturers would need to make improvements to existing SV engine 
families which would require improvements to several engine components. 
However, major retooling of engine production lines would not be 
required. In addition, the use of ABT provides manufacturers with 
considerable flexibility for determining the most appropriate 
expenditure of resources when deciding which engine families will need 
specific improvements to meet the proposed levels. The lead time 
between the

[[Page 3961]]

finalization of this rule and model year 2001 would be sufficient for 
manufacturers to meet the proposed HC+NOX level.
    The Agency has considered emission standard levels more stringent 
than the proposed 25 g/kW-hr HC+NOX. As discussed above, a 
level more stringent than 25 g/kW-hr could be met by the conversion of 
existing SV technology engines to OHV technology. The Agency's analysis 
of existing Phase 1 certification data combined with confidential sales 
information indicates an in-use level of around 15 g/kW-hr could be met 
by current Phase 1 Class I OHV engines with some design improvements to 
assure in-use emissions durability. However, these Class I OHV engines 
represent only about 10 percent of Class I sales; it is uncertain what 
level of emission could be achieved by complete conversion to OHV 
technology. As discussed previously, the percentage of Class I OHV 
engine sales has remained fairly constant for the past eight years, 
despite superior durability, performance, and fuel economy. Several 
Class I engine manufacturers, including the two largest which represent 
the majority of the market in terms of sales, have discussed with the 
Agency their past attempts to sell low cost OHV engines, likely in 
competition with less expensive SV engines. Manufacturers have 
indicated they have seen little success in drawing consumers away from 
the even lower cost Class I SV engines. Engine manufacturers have 
indicated that the principle reason for the failure of OHVs to 
penetrate further into the Class I market is the cost difference 
between the two engine technologies, and consumers' unwillingness to 
pay this premium. Several engine manufacturers have indicated that low 
cost Phase 1 Class I SV engines have manufacturing costs on the order 
of $60 to $70 per engine. Engine manufacturers contend that for these 
low cost engines, the cost increase to purchase an OHV engine is large 
enough to prevent a larger market penetration by OHV engine, at least 
when they would have to compete in the market with SV engines (see 62 
FR 14752, ``Class I OHV Demonstration Program''). The Agency estimates 
the manufacturer's cost for conversion to OHV to be between $5 and $14 
per engine. Engine manufacturers have indicated concern over what they 
perceive to be the potentially dramatic impacts on the Class I engine 
sales which would result from a standard which requires conversion to 
OHV technology. As discussed in the Overview Section III.A, above, EPA 
is also concerned that possible adverse impact on sales and the 
potential need for additional lead time could result in reduction in at 
least the near term emission benefits anticipated by this proposal. The 
Agency requests comment on the market concerns expressed by engine 
manufacturers, on the potential impact on lead time associated with 
more stringent Class I standards and on the potential for delay in at 
least the near term emission reduction benefits available from Class I 
engines if more stringent standards were adopted.
    The Agency is aware of the emission standards being considered by 
CARB for the CARB Tier 2 Utility, Lawn, and Garden Engine (ULGE) 
regulation. The Agency's current understanding is that CARB is 
considering Class I engine in-use standards of 16.1g/kW-hr 
NMHC+NOX to be met by model year 2000, followed by a 
standard of 12.0g/kW-hr in model year 2004. In their comments to the 
ANPRM, California recommended a nationwide level of control equivalent 
to that being considered by CARB. Further, CARB suggested these 
standards could be met with the use of available technology, 
specifically, total conversion to OHV technology to achieve compliance 
with a 16.1 g/kW-hr NMHC+NOX standard and the addition of 
catalyst control to meet a 12.0 g/kW-hr NMHC+NOX standard. 
EPA understands that CARB is still evaluating its Tier 2 ULGE program 
and may adopt regulations which differ from these specific levels or 
implementation dates or both. As discussed under Section IV.A of this 
proposal, section 209 of the CAA allows California to set its own 
standards, considering criteria as they apply to the State of 
California. However, as discussed later in this section, the Agency 
requests comment on whether application of these emission control 
technologies as being considered by CARB are appropriate for a Federal 
program at this time, the level of emission control expected from such 
application of these technologies and what adjustments to the proposed 
Federal program might be necessary to accommodate standards which would 
require such widespread application of OHV and catalyst technology.
    The Agency has considered the potential impacts associated with the 
conversion of Class I SVs to OHV technology. Due to uncertainties as to 
consumer acceptance of OHV engines in typical Class I equipment 
applications and as to how a more stringent Class I standard might 
effect lead time for the program as a whole and the resulting 
uncertainty of emissions benefits, the Agency has chosen not to propose 
Class I standards which would mandate the conversion of Class I engines 
to OHV or comparably clean technology. However, the Agency requests 
comment on such an option. EPA specifically requests additional 
supporting information regarding this issue to be made available to the 
Agency through the public comment process on this proposed rule to 
supplement that which informed EPA's analysis of CARB's proposed Tier 2 
levels and EPA's cost estimates of converting Class I engines to OHV. 
The Agency requests comment on all aspects of the proposed Class I 
standards.
    b. HC+NOX Emission Standard for Class II Nonhandheld 
Engines.This section presents information used by the Agency to 
determine the appropriate level for the proposed HC+NOX 
exhaust emission standards for nonhandheld Class II engines. A more 
detailed explanation of the engine technologies and costs described in 
this section is contained in the Draft RSD for this proposal, a copy of 
which is available in the public docket.
i. Class II Historical Sales Trends by Engine Technology
    Class II engine sales have been dominated by 4-stroke SV engines in 
the past. As described in the Draft RSD, Class II engines were 
predominantly SV technology in the 1970's and early 1980's. Beginning 
in about 1985, OHV engines have steadily increased their annual sales 
penetration into the Class II market, averaging about a 3 percent 
increase per year; by 1995 OHV engine sales represented approximately 
35 percent of the Class II market, with the remaining 65 percent being 
SV engines.
ii. In-use HC and NOX Emission Performance of Uncontrolled 
Class II Engines
    Information regarding new engine and in-use emission performance of 
uncontrolled Class II engines is limited. While some new engine data is 
available, the Agency does not have in-use emission information on 
uncontrolled Class II engines. The limited new engine information from 
uncontrolled engines comes from the CARB Technical Support Document for 
the CARB ULGE program.18 The Agency used this information to 
estimate the new engine emission factors for the 1991 Nonroad Engine 
and Vehicle Emission Report. Those estimates were between 15.2 and 15.4 
g/kW-hr for

[[Page 3962]]

typical new engine Class II HC+NOX emission factors.
---------------------------------------------------------------------------

    \18\ California Air Resources Board Mail Out #92-06, Technical 
Support Document for California Exhaust Emission Standards and Test 
Procedure for 1994 and Subsequent Model Year Utility and Lawn and 
Garden Equipment Engines, January 1992.
---------------------------------------------------------------------------

iii. New Engine and In-use HC and NOX Performance of Phase 1 
Class II Technology Engines
    Table 9 is a summary of the new engine emission values for gasoline 
fueled SV and OHV engine families certified to the Federal Phase 1 
regulations as of September 1997.

                                      Table 9.--Summary of Federal Phase 1 Class II Gasoline Fueled Engine Families                                     
--------------------------------------------------------------------------------------------------------------------------------------------------------
                                                                                         Average new  HC+NOX   Minimum new  HC+NOX   Maximum new HC+NOX 
                           Technology                              Number of families         (g/kW-hr)             (g/kW-hr)             (g/kW-hr)     
--------------------------------------------------------------------------------------------------------------------------------------------------------
Federal Phase 1 OHV.............................................                    64                   9.0                   5.3                  12.9
Federal Phase 1 SV..............................................                    14                  11.3                   9.4                  12.9
--------------------------------------------------------------------------------------------------------------------------------------------------------

    The values in Table 9 are an average of the certified new engine 
rates. EPA has access to manufacturers' confidential sales estimates 
for model year 1997. Using these projections the sales weighted new 
engine HC+NOX emission rate is 11.7g/kW-hr for Class II SV 
engines, and 8.3g/kW-hr for Class II OHV. This certification data shows 
that OHV new engine HC+NOX emissions tend to be lower than 
SV emissions.
    In 1996 the Agency received a report from several engine 
manufacturers regarding the deterioration of Phase 1 technology Class 
II SV and OHV engines.19 A more detailed discussion of this 
information is contained in the Draft RSD for this proposal. Table 10 
contains a summary of this information.
---------------------------------------------------------------------------

    \19\ ``Tier 1 Deterioration Factors for Small Nonroad Engines'' 
September 1996, a report by Air Improvement Resources, available in 
EPA Air Docket A-96-55, Item #II-D-11.

 Table 10.--Summary of In-Use Deterioration Factors for Phase 1 Class II
                                 Engines                                
------------------------------------------------------------------------
                                                       Class    Class II
                                                      II OHV       SV   
------------------------------------------------------------------------
Estimated HC+NOX df 250 hours.....................        1.4        1.6
------------------------------------------------------------------------

iv. Technologies Considered for Phase 2 Class II HC+NOX 
Standards
    The Agency analyzed the emission performance and cost of several 
technologies which could be applied to Class II engines, including 
improvements to existing SV engines, conversion of existing SV engines 
to OHV technology, improvements to existing OHV engines, and the 
application of catalytic converters to existing SV and OHV engines. The 
Draft RSD for this proposal contains additional information regarding 
these technologies.
    The Agency considered the costs and emission performance potential 
which would result from manufacturers making improvements to Phase 1 
Class II SV engines. As discussed in the Draft RSD, several areas for 
improvement potentially exist, including: improvements to carburetors 
to lower variability and maintain more precise air/fuel control; 
enhancements to the cylinder structural integrity; improvements to 
valve stems and valve seats; and changes in piston ring design. These 
improvements would lower production variability and improve both new 
engine and in-use emission performance. The Agency estimates these 
changes would cost the manufacturer as much as $7 to $20 per engine 
depending on engine family volume and the improvements required. 
However, the Agency believes the improvement in the in-use emission 
performance from Phase 1 levels would be small. All spark-ignited 
engines have a lean performance limit, i.e., an air/fuel ratio beyond 
which additional enleanment will result in unstable combustion and poor 
engine performance. The basic design of the SV combustion chamber 
results in a lean performance limit which is reached relatively soon 
(compared to OHV technology). Improvements in the in-use performance 
can be made, but the Agency believes these improvements will also be 
relatively small. The Agency estimates that the improvements to SV 
technology considered would result in an overall 10 to 20 percent 
reduction in the in-use emissions from Phase 1 SV levels. With the 
Phase 1 Class II new engine standard equal to 13.4 g/kW-hr 
HC+NOX, and a Phase 1 Class II SV df of 1.6, the Phase 1 in-
use emission rate is 20.1g/kW-hr at 250 hours. A 10 to 20 percent 
reduction translates to an in-use emission rate between 16.8 and 18.9 
g/kW-hr.
    As described above in Section IV.A.1.a, the principal difference 
between SV and OHV engines is the location of the intake and exhaust 
valves with respect to the combustion chamber; in SV engines the valves 
are located to one side of the combustion chamber, while in OHV the 
valves are located at the top of the combustion chamber directly above 
the piston. The OHV location offers many performance advantages over 
the SV engine, including lower valve seat distortion, lower combustion 
chamber surface-to-volume ratio, and the ability to run stably at 
leaner air-fuel ratios. These differences are described in more detail 
in the Draft RSD. These differences can result in better new engine and 
in-use HC+NOX emission performance for OHV over SV 
technology. Based on confidential Phase 1 Class II OHV Federally 
certified engine families sales projections, the Agency believes an 
average new engine emission rate of 9.3 g/kW-hr, which includes a 10 
percent compliance margin, is achievable from OHV technology engines. 
This would result in an in-use emission level of 12.1 g/kW-hr (1.3 * 
9.3 g/kW-hr), which is a 42 percent reduction from Phase 1 SV levels 
(Phase 1 SV = 13.4 g/kW-hr * 1.6 = 20.1 g/kW-hr). As presented in the 
Draft RSD, the Agency estimates the conversion of Class II SV to OHV 
technology would cost the manufacturer between $10 and $17 per engine, 
depending on the engine family volume. Engine manufacturers have 
indicated the higher cost associated with conversion of Class II SV to 
Class II OHV technology is reasonable because the equipment using Class 
II engines is typically more expensive than the equipment targeted 
toward the residential market, and the increased cost resulting from 
conversion to OHV design would not have a significant adverse impact on 
Class II engine sales. While EPA has no independent information on 
consumer price sensitivity for equipment using Class I engines, it is 
understandable that the higher price of this equipment and the typical 
commercial use of such equipment could allow the performance, fuel 
efficiency, and durability benefits of Class II OHV engines to outweigh 
the incremental impact on equipment price.

[[Page 3963]]

    The Agency also considered improvements to existing Phase 1 OHV 
engines in determining the appropriate level of the Class II standard. 
In many cases, engine manufacturers have already optimized new engine 
emission performance and have incorporated improvements to engine 
designs to optimize in-use emission performance. However, as discussed 
in the Draft RSD, the Agency believes that for some Class II OHV engine 
families internal engine improvements can still be made which would 
result in lower new engine and/or better in-use performance. These 
changes include leaner carburetor calibrations to lower new engine 
HC+NOX, optimization of combustion chamber design, and 
improvements to oil control. As discussed previously, the sales 
weighted new engine Phase 1 Class II OHV HC+NOX level is 
8.3g/kW-hr, and as shown in Table 10, the Class II HC+NOX df 
is estimated to be 1.4 at 250 hours. The Agency believes changes to 
existing Class II OHV engines will primarily improve in-use emission 
performance. As presented in the Draft RSD, the Agency estimates these 
changes would cost the manufacturer as much as $3 to $8 per engine, 
depending on the engine family production volume and the improvements 
required. However, the Agency believes many engine families have 
already incorporated these design improvements. Based on existing 
Federal certification data and the deterioration information contained 
in Table 10, the Agency estimates these improvements will result in an 
in-use HC+NOX deterioration rate of 1.3 at 250 hours, and 
average new engine emission rates (including a ten percent compliance 
margin) of 9.3 g/kW-hr, for an average in-use emission rate of 12.1 g/
kW-hr.
    Federal certification data indicates a small number of Class II SV 
and OHV engines families have certified to the Federal Phase 1 
standards using catalyst technology. However, the majority of these 
engines are intended for indoor use on applications such as generators 
or floor buffers, where lowering CO emissions appears to be the primary 
focus. The majority of these catalyst equipped Class II engine families 
operate on propane fuel. No catalyst equipped Class II engine families 
have certified to the Phase 1 rule for use in lawn and garden 
equipment. Though it is technologically feasible to apply catalysts to 
both SV and OHV engines, the Agency has little information regarding 
in-use emission performance of engines equipped with catalysts. The 
Agency's experience with on-highway catalyst technology has shown that 
considerable in-use deterioration can occur. As previously discussed in 
the Class I standard section, information on laboratory aged small 
engine catalysts has appeared in recent years in the technical 
journals. The Agency requests comment on the relationship between 
laboratory and field aged catalyst durability data, any information on 
typical in-use aged catalyst performance and all available data on 
individual catalysts aged under typical in-use conditions experienced 
by equipment using Class II engines. The Agency requests additional 
information regarding the new engine emission performance, in-use 
emission performance, and cost of catalyst technology for Class II 
engines, particularly Class II engines designed for lawn and garden 
type applications.
v. Proposed Phase 2 Class II HC+NOX Standard
    The Agency is proposing a corporate average HC+NOX 
emission standard of 12.1 g/kW-hr which will be phased in over five 
years, beginning in model year 2001. Based on the information presented 
in this section, the Agency believes an in-use level of 12.1g/kW-hr can 
be met by the conversion of Phase 1 SV engines to OHV technology, and 
by internal improvements to some existing Phase 1 OHV engines.
    The proposed standards would require significant production line 
changes for many Class II engine manufacturers to convert existing SV 
models to OHV designs, as well as modifications to some Phase 1 OHV 
models which may need internal improvements to meet the 12.1 g/kW-hr 
level. To accommodate a smooth transition of existing SV engine family 
production lines to the new OHV technology or other comparably clean 
technology, the Agency is proposing a five year phase-in period, 
starting with a level of 18 g/kW-hr in 2001 and ramping down to the 
final year level of 12.1 in model year 2005. The Agency expects the 
proposed standards for Class II engines would result in increased 
penetration of and virtual total conversion to clean OHV technology by 
2005. However, the proposal does not preclude other technologies from 
meeting the proposed standard.
    The Agency recognizes that there are large differences in 
technology mixes currently being produced by Class II engine 
manufacturers. Some Class II engine manufacturers have already made 
significant investments in OHV technology prior to and during the Phase 
1 program. For some of these manufacturers the standards in the early 
years of the Phase 2 phase-in (i.e., the 2001 standard of 18g/kW-hr and 
the 2002 standard is 16.6 g/kW-hr) may not require additional 
reductions in Class II engine emissions. At the same time, the Phase 1 
standards do not require a shift to clean, durable OHV technology or 
comparably clean technology, and several Class II engine manufacturers 
currently produce a significant number of SV engines. For manufacturers 
who are relying on SV technology the proposed phase-in period will 
allow them to shift their production to new, cleaner technology which 
is capable of meeting the 2005 standard of 12.1g/kW-hr. The Agency 
believes the phase-in standards will address the inequities among 
manufacturers' current technology mixes but will also require 
manufacturers to produce the clean, durable 12.1g/kW-hr engines in 
2005. Manufacturers have indicated the early banking provision will 
pull ahead clean technology and ease the transition to the 12.1 
standard. However, due to the wide discrepancy between manufacturers' 
current technology mixes, some manufacturers may generate significant 
credits during the phase-in period. The Agency has recently performed 
an analysis, based on Federal Phase 1 certification data, which 
indicates under some conditions, early banking would result in 
significant credits being generated during the phase-in period which 
may in fact undermine the Agency's assumptions that the 12.1 standard 
in model year 2005 would require a virtual 100 percent shift to OHV or 
comparably clear technology for Class II engines. To insure the EPA's 
goals are met, the Agency is proposing a declining set of caps on how 
high the sales-weighed average level of HC+NOX family 
emission limits (FELs) could be for Class II engine families beginning 
in 2005. A discussion of this proposal is contained in Section IV.A.5.
    Engine manufacturers have commented that, while 12.1 g/kW-hr 
HC+NOX can be met with engines designed for a typical 250-
hour useful life, engines designed for the longer proposed useful life 
categories of 500 and 1000 hours need a higher standard due to their 
higher expected df as measured over these longer hour 
periods.20 Specifically, they recommend a 500-hour engine 
standard of 13.0 g/kW-hr and a 1000-hour standard of 14.0 g/kW-hr 
HC+NOX. In arriving at these recommendations, the 
manufacturers

[[Page 3964]]

assumed the new engine emission levels would be the same regardless of 
useful life category; this is also assumed by the Agency in developing 
its proposal. However, while the manufacturers also predict 
improvements in in-use emission durability, they do not expect these 
improvements would allow a constant deterioration factor (full useful 
life emission level divided by new engine emission level) regardless of 
useful life category. Rather, the manufacturers expect improved 
durability would allow typical deterioration factors of around 1.4 for 
500-hour engines and 1.5 for 1000-hour engines. In making these 
recommendations, the manufacturers acknowledge that they have not 
provided any data or analyses to validate their recommendations, but 
also argue that the Agency has no full useful life data for these 
higher hour categories which substantiate the feasibility of the 
Agency's proposed standards. EPA requests any additional data and other 
pertinent information which would help the Agency reassess the 
appropriate level of standards for the 500-hour and 1000-hour engines.
---------------------------------------------------------------------------

    \20\ See the discussion in the March 27, 1997, ANPRM, 62 FR 
14740, and the Memo to the Docket regarding the October 3, 1997 
meeting between U.S. EPA and the Engine manufacturers Association, 
EPA Air Docket A-96-55, Item #II-E-11.
---------------------------------------------------------------------------

    Based on the May, 1997 CARB Workshop on their Tier 2 standards, the 
Agency believes CARB may propose a Tier 2 in-use standard of 12.0 g/kW-
hr NMHC+NOX in model year 2000, followed by a level of 9.4 
g/kW-hr NMHC+NOX in model year 2004. CARB's 12.0 level may 
be achievable with OHV technology and is very similar to the Agency's 
proposed Phase 2 level. CARB's 9.4 g/kW-hr level is more stringent than 
the Agency's 12.1 g/kW-hr proposal. CARB suggests an in-use 9.4g/kW-hr 
standard would require technology beyond conversion to OHV, such as an 
OHV engine equipped with a catalyst. The Agency believes the costs and 
lead time which could be necessary to achieve a 9.4 g/kW-hr level for a 
national program would be considerably greater than the program 
contained in today's proposal. However, as discussed under Section IV.A 
of this proposal, section 209 of the CAA allows California to set their 
own standards, considering criteria as they apply to the State of 
California. However, as discussed below, the Agency requests comment on 
whether the application of the technology anticipated by the standards 
being considered by CARB would be appropriate for a Federal program at 
this time.
    The Agency requests comment on all aspects of the proposed Class II 
standards, and especially requests data, analyses and other information 
on the expected emission performance capability of Class II engines 
designed for in-use operating lives of 500 hours and 1000 hours.
    c. HC+NOX Emission Standards for Class III, IV and V 
Handheld Engines. This section presents information used by the Agency 
to determine the appropriate level for the proposed HC+NOX 
exhaust emission standards for handheld engines (engine Class III, IV 
and V). A more detailed explanation of the engine technologies and 
costs described in this section is contained in the Draft RSD for this 
proposal, a copy of which is available in the public docket for this 
rule.
i. Class III, IV and V Historical Sales Trends by Engine Technology
    Handheld engine sales have historically been dominated by crankcase 
charge scavenged two-stroke engines (``traditional 2-strokes''). 
Historical sales data indicate that until the recent introduction by 
one manufacturer, Ryobi, of a 4-stroke trimmer, 100 percent of gasoline 
engine powered handheld equipment used traditional 2-stroke engines.
ii. In-use HC and NOX Emission Performance of Uncontrolled 
Class III, IV and V Engines
    Information on uncontrolled 2-stroke engines is limited. However, 
what information is available indicates 2-stroke technology has the 
potential to experience high rates of in-use deterioration of HC, on 
the order of two times the new engine value.21
---------------------------------------------------------------------------

    \21\ See ``Emission Tests of In-use Small Utility Engines'' 
Southwest Research Institute, September 1991, EPA Air Docket A-91-
24, Item #II-A-8, ``Nonroad Engine and Vehicle Emission Study'' U.S. 
EPA Report #21A-2001, November 1991, EPA Air Docket A-91-24, Item 
#II-A-10, ``Emission Testing of In-use Handheld Engines'' Southwest 
Research Institute, March 1994, EPA Air Docket A-93-25, Item #II-A-
06, and ``Regulatory Impact Analysis and Regulatory Support 
Document, Control of Air Pollution, Emission Standards for New 
Nonroad Spark-Ignition Engines at or Below 19 kilowatts'' U.S. EPA, 
May 1995, EPA Air Docket A-93-25, Item #V-B-01.
---------------------------------------------------------------------------

    This same information indicated that little in-use deterioration of 
NOX emissions occur from traditional 2-stroke engines.
iii. New Engine and In-use HC and NOX Performance of Class 
III, IV and V Phase 1 Technology Engines
    Federal Phase 1 certification data shows that over 150 two-stroke 
engine families have been certified for the 1997 and 1998 model years. 
A summary of the emission performance of these Phase 1 technology 
engine families is shown in Table 11.

                                         Table 11.--Summary of Federal Phase 1 Handheld 2-stroke Engine Families                                        
--------------------------------------------------------------------------------------------------------------------------------------------------------
                                                                                         Average new  HC+NOX   Minimum New  HC+NOX   Maximum New  HC+NOX
                          Engine class                             Number of families         (g/kW-hr)             (g/kW-hr)             (g/kW-hr)     
--------------------------------------------------------------------------------------------------------------------------------------------------------
Class III.......................................................                     4                   216                   177                   258
Class IV........................................................                   131                   189                    97                   236
Class V.........................................................                    19                   136                    90                   161
--------------------------------------------------------------------------------------------------------------------------------------------------------

    The average emission rates for the Phase 1 Class III, IV and V 
traditional 2-stroke engines are 28 percent, 23 percent and 18 percent 
below the combined Phase 1 HC and NOX standards. Federal 
certification data also show three Class IV four-stroke technology 
engine families and three Class IV two-stroke with catalysts engine 
families have been certified to the Federal rule. The average 
HC+NOX certification levels for these engine families are 27 
and 165 g/kW-hr respectively.
    Information on in-use emission performance of Phase 1 technology 2-
strokes is also limited. In preparation for the Phase 1 regulation, 
several members of the Portable Power Equipment Manufacturers 
Association (PPEMA) ran a test program which included manufacturer 
controlled field testing of seven Phase 1 technology 2-stroke engines, 
six aged to 50 hours, and one to 225 hours.22 This data 
shows relatively low deterioration in HC+NOX emissions, with 
dfs ranging from slightly less than 1.0 to approximately 1.2 at 50 
hours, and slightly less than 1.0 for the 225 hour engine.
---------------------------------------------------------------------------

    \22\ See Appendix C of ``Regulatory Support Document, Control of 
Air Pollution, Emission Standards for New Nonroad Spark-Ignition 
Engines at or Below 19 kilowatts'' U.S. EPA, May 1995, EPA Air 
Docket A-93-25, Item #V-B-01.

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

[[Page 3965]]

    The Agency has little information on the in-use performance of 4-
stroke handheld technology or on handheld catalyst technology.
iv. Technologies Considered for Phase 2 Class III, IV and V 
HC+NOX Standards
    The Agency analyzed the emission performance and cost of several 
technologies which could be applied to handheld engines. These include 
improvements to existing 2-stroke engines, conversion of existing 2-
stroke engines to 4-stroke technology, and the application of catalytic 
converters to existing 2-stroke engines. The Draft RSD for this 
proposal contains additional information regarding these technologies.
    For Phase 1 2-stroke technology engines, fuel lost during the 
scavenging process represents the largest fraction of exhaust HC 
emissions, and HC emissions represent greater than 95 percent of the 
exhaust HC+NOX emissions. The Agency believes several types 
of improvements can be made to Phase 1 technology 2-stroke engines. The 
following is a summary of potential areas for lowering 
HC+NOX emissions: (1) improvements in carburetors to reduce 
production variability and tighter air/fuel ratio control; (2) redesign 
of the combustion chamber to promote more complete combustion; (3) 
optimizing port shapes and timing to reduce scavenging losses; (4) 
leaner carburetor calibrations to reduce HC emissions; and (5) tighter 
manufacturing tolerances for engine components to reduce component 
variation. These improvements are discussed in more detail in the Draft 
RSD. As described in the Draft RSD, the Agency estimates the cost of 
these improvements would cost the manufacturer as much as $2 to $6 per 
engine, depending on the production volume of the engine family and the 
improvements required. The Agency would expect these changes to lower 
the new and in-use emission rates of Phase 1 two-stroke technology 
engines. PPEMA members have indicated they believe a well designed, 
properly maintained 2-stroke engine is capable of performing with no 
in-use deterioration of HC+NOX emissions. Based on the small 
amount of in-use data from Phase 1 technology engines, the Agency 
estimates the in-use performance of an improved Phase 1 technology 2-
stroke engine would deteriorate approximately 10 percent during its 
useful life. The Agency estimates that for the majority of handheld 
engines, improvements to Phase 1 2-stroke designs would result in a 30 
percent reduction in the in-use emission rates from Phase 1 designs.
    The Agency also analyzed the benefits and associated costs which 
would occur from the conversion of existing 2-stroke handheld engines 
to 4-stroke designs. Two engine manufacturers, Ryobi and Honda, have 
successfully demonstrated that 4-stroke designs are viable in at least 
some handheld equipment applications, notably a string trimmer 
application. However, the Agency is uncertain that 4-stroke technology 
would be viable in all handheld applications, particularly those 
applications which require high power and low weight, such as large, 
commercial chainsaw applications, where the lower power-to-weight ratio 
of 4-stroke engines may impede equipment performance. Four-stroke 
technology does not have the scavenging loss problem associated with 
traditional 2-strokes. Therefore 4-stroke exhaust HC emissions are 
substantially below those of a 2-stroke design. Federal Phase 1 
certification data for Class IV engines indicates a 4-stroke string 
trimmer produces new engine HC+NOX emission rates of about 
27 g/kW-hr, which is approximately 80 percent below the Phase 1 
standard. Deterioration information on small displacement 4-stroke 
engines is limited, and the Agency has no deterioration information on 
handheld 4-stroke engines. The Agency has heard from one small engine 
manufacturer that the smaller 4-stroke engines would likely have higher 
deterioration than Class I OHV 4-stroke engines, which is on the order 
of 1.4 at 66 hours.23 The Agency requests comment and 
additional information on the deterioration of smaller 4-stroke 
engines. As described in the Draft RSD, the Agency estimates the cost 
of converting an existing handheld 2-stroke to a 4-stroke engine would 
cost the manufacturer between $7 and $10 per engine, depending on the 
production volume of the engine family.
    The Agency also considered the application of catalytic convertors 
to Phase 1 2-stroke technology. One handheld engine manufacturer, 
Husquvarna, has certified three engine families to the Phase 1 rule 
which utilize a 2-stroke engine with catalyst. This engine has been 
designed for lower scavenging losses to reduce engine out emissions, 
has improved fuel metering, and also uses a catalyst to further reduce 
exhaust emissions. EPA's testing of this engine showed new engine 
emission results for HC+NOX at the nominal carburetor 
setting on the order of 90 g/kW-hr, which is 63 percent below the 
combined Phase 1 Class IV HC+NOX new engine standard. The 
Agency does not have information regarding the actual in-use 
performance of this or other catalyst equipped 2-stroke engines. The 
Agency estimates the cost of adding a catalytic convertor to an 
improved 2-stroke handheld engine would cost the manufacturer between 
$6 and $12 per engine, depending on the production volume of the 
family. This cost estimate does not include any of the additional 
improvements to the Phase 1 technology 2-stroke mentioned previously, 
such as combustion chamber improvements or scavenging design 
improvements. As previously discussed, such improvements to existing 2-
stroke designs would cost the manufacturer an additional $2 to $6 per 
engine. Therefore, the Agency estimates an improved 2-stroke design 
with a catalytic convertor would cost the manufacturer from $8 to $18 
per engine. Comments are requested on these cost estimates.
---------------------------------------------------------------------------

    \23\ See Item # II-E-08 in EPA Air Docket A-96-55 referencing a 
meeting between EPA and Honda.
---------------------------------------------------------------------------

v. Class III, IV and V Proposed Phase 2 HC+NOX Standard
    The Agency is proposing an in-use HC+NOX standard of 
210, 172 and 116 g/kW-hr for Class III, IV and V engines, respectively. 
As presented in Table 7, the proposed standards would begin in model 
year 2002, with a requirement that 20 percent of a manufacturer's U.S. 
sales meet the standards, followed by an increased percentage each year 
until model year 2005, when 100 percent of a manufacturer's U.S. sales 
would be required to meet the proposed standards.
    The Agency expects the proposed in-use standards can be met 
primarily through improvements to existing Phase 1 technology 2-stroke 
engines. As presented previously, the Agency believes improvements to 
Phase 1 technology 2-stroke engines should result in approximately a 30 
percent reduction in the in-use emissions of Phase 1 engines, which 
would be required to meet the proposed standards.
    PPEMA members have indicated the proposed standards would require 
significant research and development time as well as a large capital 
investment to change existing production capabilities. The proposed 
phase-in period plus the lead time anticipated after this rule is 
finalized will allow manufacturers at least 6 years to make the 
necessary changes to existing product lines in order to meet the 
proposed standards, which should accommodate the manufacturers' 
concerns regarding lead time.

[[Page 3966]]

    The Agency has not proposed a handheld standard which would require 
catalyst or 4-stroke technology. The Agency's experience with on-
highway technology indicates catalysts and engine technology evolved 
together to prevent significant in-use deterioration. As previously 
discussed in the section on the Class I engine standard, publicly 
available information on bench aged catalysts used on 4-stroke engines 
has become available in recent years. The Agency requests comment on 
the relationship between bench aged and typical in-use aged catalyst 
performance, and all available data on individual catalysts aged under 
typical in-use conditions experienced by handheld equipment. The Agency 
requests additional information on the new and in-use emission 
performance of catalyst-equipped handheld engines. Two engine 
manufacturers have introduced 4-stroke engines into string trimmer 
applications. There are likely some applications, such as high power 
chainsaws, where 4-stroke technology may not be feasible as a power 
unit because of weight concerns. As previously discussed, the Agency 
estimates that conversion to 4-stroke designs would cost the 
manufacturer between $7 and $10 per engine. PPEMA has reported that in 
1993 and 1994 the average retail price of a 2-stroke gasoline powered 
string trimmer or leaf blower was approximately $100, and the average 
retail price of a chainsaw was approximately $200. PPEMA members, who 
do not currently manufacture 4-stroke handheld products, have expressed 
concern regarding what they perceive to be the potential negative 
impacts on sales which would result from a large increase in engine 
costs, such as the cost of conversion to 4-stroke technology for 
handheld engines. While EPA has no independent information on consumer 
price sensitivity, it is concerned that the higher cost of equipment 
which would likely result if catalyst or 4-stroke technology were 
necessitated by a more stringent standard could result in significant 
financial burden if the industry were to absorb the cost impact or 
adverse impact on sales if the increase in cost were passed along to 
the consumer. EPA is also concerned that mandating near term conversion 
to 4-stroke technology could significantly increase the lead time 
necessary before implementing the standards and delay the emission 
benefits of the standards. The Agency requests comment on the market 
concerns expressed by these engine manufacturers as well as the 
potential impact on lead time of a more stringent standard and 
information on the cost to the consumer and in-use emissions 
performance if 2-stroke engines were required to be equipped with a 
catalyst.
    The Agency believes that during the next several years additional 
information regarding the in-use performance of new technologies, such 
as handheld 4-strokes, or traditional 2-strokes equipped with 
catalysts, may become available, perhaps in response to the CARB Tier 2 
program. In addition, EPA recognizes that technological advances and/or 
cost reductions may occur after promulgation of the Phase 2 rule that 
could make greater, but still cost-effective reductions feasible in 
handheld emission levels. The Agency proposes to conduct a technology 
review to address this possibility. In this review, EPA expects to 
examine issues including the potential for further reductions from 
existing 2-stroke engines, stratified charge 2-stroke technology, 
direct injection 2-stroke injection, the use of catalysts on handheld 
engines, and the conversion to 4-stroke technology. Following a 
technical review, the Agency intends to publish a Notice of Proposed 
Rulemaking in 2001 announcing any possible amendments to the standard 
levels or other program elements, or EPA's intention to maintain the 
existing handheld standards or program. The Agency expects that the 
final rulemaking would be completed by 2002 and, if adopted, Phase 3 
standards would be phased in on a percentage basis and over of a period 
of time similar to Phase 2, beginning no earlier than model year 2007. 
This schedule is intended to provide a minimum five year period before 
the implementation of any Phase 3 standards in order to allow 
manufacturers to recoup their investments in Phase 2 technology and 
ensure the cost-effectiveness of the Phase 2 program.
    The Agency is aware that CARB is considering a Tier 2 standard for 
all handheld engines of 72 g/kW-hr HC+NOX, which is more 
stringent than the levels being proposed for the Federal program. CARB 
has stated this level could be met by the complete conversion of 
existing 2-stroke technology to 4-stroke technology. The Agency 
believes the costs and lead time which would be necessary to achieve a 
72 g/kW-hr level for a national program could be considerably higher 
than the program contained in today's proposal. However, as discussed 
under Section IV.A of this proposal, section 209 of the CAA allows 
California to set its own standards, considering criteria as they apply 
to the State of California. However, as discussed below, the Agency 
requests comment on whether 4-stroke technology for all handheld 
applications would be appropriate for a Federal program at this time. 
The Agency requests comment on all aspects of the proposed handheld 
standards, and on what adjustments to the proposed Federal program 
might be necessary to accommodate such standards.
    d. Proposed California Standards. As mentioned previously, the 
State of California has proposed standards for both handheld and 
nonhandheld small SI engines which are considerably more stringent than 
the standards which the Agency is proposing today. In this proposal, 
the Agency has noted several reasons why the level of control being 
considered by California is not being proposed today, including 
uncertainties regarding cost, the possible impact of potential price 
increases on consumer sales, and the lead time necessary for the 
industry should they be required to adopt the required changes in 
technology nationwide. However, EPA requests comment on the feasibility 
in the Federal program of requiring such technology as anticipated by 
the standards being considered by California, the level of emission 
control which would result, the costs of such technology for a 
nationwide program, and any impact on lead time necessary to allow the 
adoption of such levels of control nationwide.
2. NMHC+NOX Emission Standards for Class I and II Natural 
Gas Fueled Nonhandheld Engines
    EPA is proposing optional separate standards for Class I and Class 
II natural gas fueled engines only, due to the fact that for these 
engines methane has very low ozone forming potential, i.e., low 
reactivity. The total hydrocarbon (THC or HC) emissions from Phase 1 
technology 4-stroke gasoline engines is between 5 and 10 percent 
methane by mass. For natural gas engines, methane is on the order of 70 
percent of total HC mass emissions. For natural gas fueled nonhandheld 
engines, the Agency is proposing an optional NMHC+NOX 
standard, as presented in Table 12.

[[Page 3967]]



               Table 12.--NMHC + NOX Emission Standards for Natural Gas Fueled Nonhandheld Engines              
                                                    [g/kW-hr]                                                   
----------------------------------------------------------------------------------------------------------------
                                                  Model year   Model year   Model year   Model year   Model year
                  Engine class                       2001         2002         2003         2004         2005   
----------------------------------------------------------------------------------------------------------------
Class I........................................         23.0         23.0         23.0         23.0         23.0
Class II.......................................         16.7         15.3         14.0         12.7         11.3
----------------------------------------------------------------------------------------------------------------

    These proposed NMHC+NOX standards have been adjusted so 
that these standards are of equivalent stringency to the 
HC+NOX standards for gasoline fueled engines, i.e., 11.3 g/
kW-hr NMHC+NOX is a deteriorated new engine 
NMHC+NOX level, assuming a new engine THC+NOX 
level of 9.3 g/kW-hr, a NMHC+NOX deterioration factor of 
1.3, and a new engine split of 54 percent NMHC, 6 percent methane and 
40 percent NOX.
    The Agency is proposing that for natural gas fueled engines, the 
standard be based on the level of NMHC+NOX reduction which a 
Phase 2 technology gasoline fueled nonhandheld engine could be expected 
to meet, not on the performance of a Phase 2 technology natural gas 
fueled engine. Natural gas fueled engines represent less than 1 percent 
of annual small engine sales and EPA recognizes that this is a 
technology that as a matter of environmental policy it may be desirable 
to encourage. The Agency believes very little environmental benefit 
would occur from basing this optional NMHC+NOX standard on 
the performance of Phase 2 technology natural gas engines. In 
consideration of the energy and safety factors associated with using 
natural gas technology rather than gasoline technology, EPA is 
proposing the NMHC+NOX standard at a level that gives 
manufacturers a greater incentive, as a result of the ABT program, to 
use natural gas technology. The Agency requests comment on this 
approach, and on whether it poses a meaningful risk of allowing over 
generation of positive credits in the ABT program.
    The NMHC+NOX standard would require an additional 
testing burden for natural gas engine manufacturers, because these 
manufacturers would need an additional emission analyzer to measure the 
methane content of the exhaust gas. However, because natural gas engine 
manufacturers have requested this optional NMHC standard, and the 
Agency does not see any adverse effects for the formation of ozone, the 
Agency believes it is appropriate for this proposal. EPA is not 
proposing NMHC + NOX standards for handheld engines. EPA is 
not aware of any natural gas fueled handheld applications. Therefore, 
no NMHC+NOX standard is needed.
    The Agency is aware that CARB may use a NMHC+NOX 
standard for all handheld and nonhandheld engine manufacturers. At this 
time, EPA does not believe an emissions benefit would occur by 
replicating this action for the Federal program. The Agency would need 
to adjust all standards downward to maintain equivalent stringency and 
require all manufacturers to begin testing for methane. If 
manufacturers of small SI engines were able to selectively target 
reductions in NMHC as compared to THC, an NMHC standard may be of some 
value to manufacturers. However, the Agency is not aware of small 
engine technologies which have this potential, other than natural gas 
fueled engines, which represent less than 1 percent of annual sales. 
Therefore, because a national NMHC standard would result in increased 
testing cost for little or no benefit, the Agency is not proposing NMHC 
standards for all small engines at this time.
3. CO Emission Standards
    In addition to HC and NOX standards, the Phase 1 final 
rulemaking (60 FR 34582) put in place a cap on the level of CO 
emissions from small SI engines. That cap was subsequently modified for 
Class I and II engines (61 FR 58296). In today's action EPA is 
proposing that the Phase 1 CO standards be adjusted to reflect in-use 
standards and to maintain the same level of stringency as afforded by 
the Phase 1 standards. Specifically, EPA proposes to take the Phase 1 
standards and multiply them by the projected CO dfs over the useful 
lives of the engines to arrive at the Phase 2 in-use CO standards. For 
Class I and II engines, available data indicates that the df ranges 
considerably between less than 1.0 and something in excess of 2.0 
depending on the engine. For Class III, IV and V engines, available 
data indicates that the df for CO ranges more narrowly and typically 
falls between 1.0 and 1.1. Consequently, EPA proposes that the 
following in-use CO standards in Table 13 apply for the Phase 2 
program:

      Table 13.--In-Use CO Emission Standards for Small SI Engines      
                              [In g/kW-hr]                              
------------------------------------------------------------------------
                                                  Engine Class          
                                      ----------------------------------
                                         I      II    III     IV     V  
------------------------------------------------------------------------
CO Standard (g/kW-hr)................    610    610    805    805    603
------------------------------------------------------------------------

    These CO standards would not be subject to the averaging, banking, 
and trading provisions of the rule available for nonhandheld engines. 
Rather, these standards would serve as caps on the CO emissions allowed 
from all engine families.
    EPA is proposing that for Class I and Class II engines, the 
proposed CO levels would be effective in the 2001 model year for a 
manufacturer's entire product line. For Class III, IV and V engines, 
those engine families complying with Phase 2 HC+NOX levels 
under the proposed phase-in for HC+NOX standards for 
handheld engines would be required to also comply with CO levels on the 
same phase-in schedule. This seemingly disparate treatment for handheld 
and nonhandheld is consistent with the other provisions of the program 
(e.g., phase-in from Phase 1 to Phase 2 for handheld but not for 
nonhandheld engines) and protects manufacturers from having to have 
engine families comply with Phase 2 CO requirements prior to those same 
engine

[[Page 3968]]

families being subject to the other Phase 2 requirements.
    EPA believes it is appropriate not to go beyond the Phase 1 
stringency for CO emissions for two main reasons. First, in most parts 
of the country CO is primarily a wintertime problem (November through 
February), while the vast majority of engines covered by this 
rulemaking are used almost exclusively during the summer months. As a 
result, most additional CO emission reductions resulting from any 
increase in the stringency of the standard would not occur at a time 
when they would provide nonattainment areas with measurable benefit 
toward meeting the National Ambient Air Quality Standard (NAAQS) for 
CO.
    Second, CO is a diminishing ambient air quality 
problem.24 There has been approximately an 80 percent 
reduction in the number of nationwide exceedances of the NAAQS for CO 
since the Clean Air Act Amendments of 1990, and this trend is expected 
to continue without further tightening of CO requirements for small SI 
engines. Many of the CO nonattainment areas in 1990 have already been 
redesignated as being in attainment, many more are in the process of 
requesting redesignation, and many of those not currently requesting 
redesignation are expected to before the time the Phase 2 standards 
would go into effect.
---------------------------------------------------------------------------

    \24\ See ``National Air Pollution Emission Trends, 1900-1995,'' 
EPA-454/R-96-007, October 1997.
---------------------------------------------------------------------------

    Taken together, these two reasons indicate that it does not make 
sense to pursue more stringent CO standards at the national level for 
small SI engines at this time. Should this situation change, EPA can 
take appropriate action at that time.
    While EPA does not believe it is appropriate at this point in time 
to pursue more stringent CO standards for small engines, we 
nevertheless do believe it is important to maintain the current level 
of stringency for CO. As discussed in the Phase 1 rulemaking, 
uncontrolled small SI engines do contribute approximately 1 percent of 
the emissions toward the national winter CO inventory.25 As 
a result, while emissions from small SI engines represent a small piece 
of the inventory, they are significant. Furthermore, many small SI 
engines are used outside in close proximity to the equipment users, 
raising possible concerns over user health effects. A recent National 
Institute of Occupational Safety and Health Alert 26 raised 
serious health concerns regarding the operation of gasoline powered 
engines inside buildings or other partially enclosed spaces due to 
potential CO poisoning. The NIOSH Alert contains a list of suggested 
practices for the proper use of equipment powered by small gasoline 
engines which should be followed. The NIOSH alert does not recommend a 
more stringent CO standard for gasoline powered small SI engines.
---------------------------------------------------------------------------

    \25\ Nonroad Engine and Vehicle Emission Study--Report, U.S. 
EPA, November 1991, EPA Air Docket A-91-24, Item #II-A-10.
    \26\ ``Preventing Carbon Monoxide Poisoning from Small Gasoline-
Powered Engines and Tools,'' Department of Health and Human Services 
Publication #96-118. Information on how to obtain this publication 
is contained in EPA Air Docket A-96-55, Item #II-B-1.
---------------------------------------------------------------------------

    Even without a more stringent CO standard for Phase 2, CO emissions 
from small engines will likely continue to decrease as manufacturers 
improve production quality (reduce tolerances and variability) and 
improve durability to meet the more stringent HC+NOX 
standards proposed for Phase 2. To the extent that this does occur, and 
Phase 2 engines are shown to clearly achieve the Phase 2 CO emission 
standards, the proposal would allow EPA the flexibility to waive the 
reporting of CO emissions in the future, thereby decreasing the 
compliance costs associated with the program as it transitions to one 
more focussed on HC+NOX emissions. EPA requests comment on 
this aspect of the proposed rule. To the extent that engines do exceed 
the Phase 2 CO emission standard, EPA could also consider in the future 
setting a more stringent CO standard, taking into account cost, lead 
time, energy and safety factors as required by the Clean Air Act.
4. Useful Life Categories.
    Section 213(a)(3) of the Clean Air Act provides that regulations 
promulgated for nonroad engines shall apply to the useful lives of the 
engines. EPA is proposing that engine families meet the proposed Phase 
2 emission standards throughout their useful lives, a requirement new 
to this Phase 2 program for small SI engines. Small SI engines can 
experience a wide range of useful lives, depending upon the 
applications and usage patterns, even within a single engine class. EPA 
believes that the three useful life categories each for Class I and 
Class II engines, and the two useful life categories each for Class 
III, IV and V engines proposed today would provide a means of sorting 
engines for regulatory purposes to reflect expected usage, without 
establishing an overly complex system of useful life categories. So 
that consumers have the best information available as to the emission 
durability of the engine being purchased, EPA is proposing that an 
indication of the useful life hours be included on the engine's 
certification label. Finally, in order to ensure that the air quality 
benefits anticipated by the proposed rule will in fact accrue, EPA is 
proposing that manufacturers select the useful life category most 
appropriate for the engine family. This section discusses the useful 
life categories proposed today for nonhandheld and handheld engines, 
proposed provisions for inclusion of the useful life hours on the 
engines' label, and proposed provisions relating to manufacturer 
selection of the appropriate useful life category.
    a. Useful Life Hours. EPA is proposing three useful life categories 
each for Class I and Class II nonhandheld engines, and two useful life 
categories each for Class III, IV and V handheld engines, as shown in 
Tables 14 and 15. These categories are based on information of the 
ranges of useful lives experienced by the engines in these Classes.

          Table 14.--Nonhandheld Engine Useful Life Categories          
                                 [Hours]                                
------------------------------------------------------------------------
                                          Category   Category   Category
                                             C          B          A    
------------------------------------------------------------------------
Class I................................         66        250        500
Class II...............................        250        500       1000
------------------------------------------------------------------------


            Table 15.--Handheld Engine Useful Life Categories           
                                 [Hours]                                
------------------------------------------------------------------------
                                       ``Residential''   ``Commercial'' 
------------------------------------------------------------------------
Class III...........................                50               300
Class IV............................                50               300
Class V.............................                50               300
------------------------------------------------------------------------

    EPA is aware that the small SI engine and equipment industry is 
comprised of a wide variety of equipment with a wide range of usage 
patterns. Handheld and nonhandheld engines are designed for many 
different types of applications, with each application having specific 
design criteria, resulting in different expected lifetimes. The most 
obvious example of these differences is the distinction between 
commercial (or professional) operators and residential (or home) 
operators. In general, commercial operators, such as commercial lawn-
care companies or rental companies, expect to accumulate high numbers 
of hours on equipment on

[[Page 3969]]

an annual basis, while a residential operator, such as a residential 
chain saw owner, expects to accumulate a relatively low number of hours 
on an annual basis. Several organizations have investigated the issues 
related to average life and annual use of equipment powered by small SI 
engines, including industry organizations, the California Air Resources 
Board, and EPA (see Chapter 3 of the Draft RSD for a summary of several 
of these reports).
    On the nonhandheld engine side, a 1992 phone survey of over 6,000 
households collected information on usage rates for consumer-owned 
walk-behind and ride-on mowers, showing that on average consumers 
accumulated 100 hours of use on walk-behind mowers (typical of Class I 
``residential'' engines) over a five year period of time, and 207 hours 
of use on ride-on mowers over a six year (five and six years being the 
estimates of when one-half of the mowers are no longer in service, or 
``B-50'' life, 27 for walk-behind and ride-on mowers, 
respectively).28 On the handheld side, a 1990 study 
demonstrated the large disparity between consumer and professional use, 
with consumer equipment expected life time estimates ranging from 53 to 
80 hours, and professional equipment expected life time estimates 
ranging from 225 to 536 hours.29 A 1990 study of both 
nonhandheld and handheld equipment in residential and commercial 
applications showed a large disparity in average lifespan between 
equipment used by residential and commercial applications, with 
residential equipment implied average lifespan estimates ranging from 
35 to 394 hours, and commercial equipment implied average lifespan 
estimates ranging from 274 to 3024 hours.30
---------------------------------------------------------------------------

    \27\ The ``B-50'' is the point at which one-half of the 
equipment are no longer in service. For regulatory purposes, EPA 
anticipates that engines would be certified to a ``useful life'' 
which most accurately reflects this ``B-50'' value. Thus, for a 
Class II engine family certified to the 250 hour useful life 
category, half of those engines would be expected to no longer be in 
service after 250 hours.
    \28\ ``Useful Life, Annual Usage, and In-Use Emissions of 
Consumer Utility Engines,'' memo from the OPEI CAAC In-Use Working 
Group to Ms. Gay MacGregor, U.S. EPA, EPA Air Docket A-96-55, Item # 
II-D-13.
    \29\ ``A 1989 California Baseline Emissions Inventory for Total 
Hydrocarbon and Carbon Monoxide Emissions from Portable Two-Stroke 
Power Equipment,'' prepared by Heiden Associates, Inc., for the 
Portable Power Equipment Manufacturers Association, July 24, 1990, 
available in EPA Air Docket A-96-55, Item #II-D-14.
    \30\ ``Utility Engine Emission report,'' prepared by Booz, Allen 
and Hamilton Inc., for the California Air Resources Board, November 
20, 1990, available in EPA Air Docket A-93-25, Item #II-I-02. These 
implied average lifespan estimates were calculated from average 
annual use and estimated ``B-50'' values.
---------------------------------------------------------------------------

    Based on these sources of information, EPA is proposing for 
regulatory purposes three useful life categories for nonhandheld 
engines, and two useful life categories for handheld engines. The 
determination of which useful life category is appropriate for a 
specific engine is largely dependent on its intended application. For 
example, Class II engines going into a consumer ride-on mower 
application may most appropriately have a regulatory useful life of 
``250 hours.'' The longer useful life categories would be appropriate 
for engines placed into ``commercial'' types of usage. For example, a 
Class II engine going into a ``commercial'' generator set application, 
may most appropriately have a regulatory useful life of 1000 hours. EPA 
believes that a number of features of engine and/or equipment design 
are reflective of the intended or expected usage of the engines. As 
discussed below, manufacturers would be expected to have information on 
the intended application of their engines which support their useful 
life category selections.
    EPA received comments on the ANPRM arguing that the Class I 
shortest useful life (66 hours) is too short, and that the minimum 
lifetime compliance period for Class I engines should be set at 120 or 
125 hours to reflect an average six year life with an average use of 20 
hours a year for mower engines. While the Agency agrees that 120 or 125 
hours may be more representative of the ``B-50'' life of residential 
Class I engines, EPA selected 66 hours as sufficient to determine the 
emission durability performance characteristic of engines in this Class 
I design category. EPA did so under the assumptions that certifying 
Class I engines to 66 hours rather than 120 or 125 hours would still 
provide adequate assurance of in-use emission performance over the life 
of the engines without the added burden which would be incurred with 
testing to the higher hours. If this proves not to be the case, EPA 
would likely have to adjust the useful life, deterioration factors and 
standards accordingly to provide such assurance. EPA requests comment 
on the tradeoff between compliance demonstration and in-use compliance 
assurance associated with the 66 hour useful life proposal.
    For handheld engines, the 50 hours category reflects 
``residential'' usage, and the 300 hour category reflects 
``commercial'' usage. For example, a trimmer in residential use may 
most appropriately be certified to a regulatory useful life of 50 
hours, while a chainsaw in commercial use may more appropriately be 
certified to a useful life of 300 hours. Again, EPA believes that a 
number of features of engine and/or equipment design are reflective of 
the intended or expected usage of the engines. As discussed below, 
manufacturers would be expected to have information in support of their 
useful life category selections for handheld engines.
    EPA received comments on the ANPRM arguing that an intermediate 
useful life category for some handheld products might be appropriate, 
for example, in the case of products with intended useful lives of 150 
hours. EPA believes that the 50 and 300 hour useful life hour 
categories are sufficient to distinguish residential and commercial 
usage, respectively. EPA has not received additional data in support of 
an intermediate useful life, and believes that it is desirable to avoid 
a proliferation of useful life categories. Thus, EPA is not proposing 
an intermediate useful life category for handheld engines. However, EPA 
requests comment and data on the issue of whether an intermediate 
category is appropriate, what would be the appropriate hours for an 
intermediate category, and what features of an engine with an 
intermediate useful life might distinguish it from engines more 
appropriately certified to a 50 or a 300 hour useful life.
    EPA also received comments on the ANPRM regarding the use of 
``residential'' and ``commercial'' to indicate the useful life for 
handheld engines. Several commenters suggested that the terms 
``residential'' and ``commercial'' are potentially misleading to 
consumers of handheld engines. One commenter was concerned that dealers 
would have the responsibility to ``qualify'' a buyer of equipment, and 
in the event of injury, the dealer would be at risk for having sold the 
wrong buyer the wrong equipment. This commenter suggested instead that 
EPA categorize engines in terms of power, size, weight, or other 
factors that clearly would not risk making dealers think they have a 
responsibility to classify the expertise of the buyer. A second 
commenter suggested EPA could base the useful life on technical 
properties of engines such as ``half crank'' and ``full crank'' rather 
than ``commercial'' and ``residential.'' A third industry commenter 
suggested that it is unnecessary and unwise for manufacturers to 
differentiate handheld engine families by the terms ``residential'' and 
``commercial,'' since these terms are not airtight, and in fact have 
substantial overlap for some models. This commenter suggested using 
useful life categories ``A'' and ``B'' instead, where a Category A 
engine (or

[[Page 3970]]

engine family) would be ``a handheld engine model or family designated 
by the manufacturer, at the time of certification, as an engine 
intended primarily for commercial use. Such an engine or family would 
be subject to testing requirements and warranty obligations for its 
regulatory useful life. The regulatory useful life of a Category A 
engine shall be 300 hours.'' A Category B engine (or engine family) 
would be ``an engine model or family designated by the manufacturer, at 
the time of certification, as an engine intended primarily for 
residential use. Such an engine or engine family would be subject to 
testing requirements and warranty obligations for its regulatory useful 
life. The regulatory useful life of a Category B engine shall be 50 
hours.''
    EPA agrees that commercial and residential are not airtight terms. 
However, EPA is proposing the following definitions for these terms and 
requests comments on these definitions. A ``residential engine'' would 
mean a handheld engine for which the engine manufacturer makes the 
statement to EPA that such engine and the equipment it is installed in 
by the engine manufacturer, where applicable, is not produced, 
advertised, marketed or intended for commercial or professional usage. 
A ``commercial engine'' would mean a handheld engine that is not a 
residential engine.
    In response to the commenter's concerns about dealer 
responsibilities, EPA believes that inclusion of the terms 
``residential'' and ``commercial'' should not pose a risk to dealers, 
and that the proposed duty of engine manufacturers to certify and label 
their engines for purposes of emissions durability would not transfer 
into a duty on the dealer's part to restrict sale of ``commercial'' 
products to ``residential'' purchasers. EPA requests comment on all 
aspects of the proposal for handheld useful life categories and the 
proposed definitions of ``commercial'' and ``residential'', or other 
alternative designations for the 50 and 300 hour useful life 
categories. In particular, EPA requests comment on eliminating the use 
of residential and commercial as regulatory terms, and simply retaining 
the ``50'' and ``300'' hour useful life categories.
    In summary, the Agency's analysis indicates there is a large 
disparity in the useful life of engines within all five engine classes. 
The Agency is interested in striking a compromise between the need for 
representative useful lives, and the reality that different engines 
within a single class are designed for vastly different usage patterns. 
For this reason the Agency believes it is appropriate to have multiple 
useful life categories, but the Agency believes there should be a limit 
on the number of categories, to prevent an overly complex 
categorization system. Based on the information presented in this 
section, the Agency believes the proposed useful life categories 
presented in Tables 14 and 15 are appropriate. The Agency requests 
comment on these proposed useful life categories.
    b. Useful Life on the Engine's Label. EPA is proposing that 
manufacturers would indicate their selection of useful life category by 
adding information concerning the engine's ``emissions compliance 
period'' to the engine's label. This information would be an important 
tool for consumers and purchasers of engines. EPA anticipates that 
manufacturers will use the useful life hours of the engine as a 
marketing tool. For example, a manufacturer might advertise that an 
engine family is emissions durable to 1000 hours, or is certified by 
EPA as a ``commercial'' engine. Thus, the requirement that 
manufacturers indicate the emissions compliance period on the engine's 
label would also have potential as a marketplace mechanism to help 
encourage manufacturers to select longer useful life categories.
    For nonhandheld engines, EPA is proposing that the manufacturer 
would add to the compliance statement on the engine's label, 
``EMISSIONS COMPLIANCE PERIOD: [useful life] HOURS.'' In addition, 
consistent with the ANPRM, EPA is proposing as an option for 
nonhandheld manufacturers, rather than indicating the useful life in 
hours, the manufacturer may add to the compliance statement on the 
engine's label ``EMISSIONS COMPLIANCE PERIOD: CATEGORY [A, B, OR C]. 
REFER TO OWNER'S MANUAL FOR FURTHER INFORMATION.'' In this case, the 
owner's manual would be required to contain the statement: ``This 
engine has been shown to meet emission standards for a period of 
[useful life] hours.'' EPA is proposing this option in light of 
concerns voiced by manufacturers that putting the useful life of the 
engine, in hours, on the engines' label, could be misleading to 
consumers in that the emissions compliance period may or may not 
represent the expected lifetime of the engine. Nevertheless, EPA 
believes that putting the engine's useful life in hours on the engine's 
label could serve as an important mechanism to educate and inform 
consumers as to the emissions durability of the product they are 
considering. EPA requests comment on whether the option to allow a 
manufacturer to instead designate the useful life by using Category [A, 
B or C] on the engine's label, with information on the emissions 
compliance period in hours in the owners manual, is an effective 
substitute to achieve this goal of educating consumers.
    In the case of handheld engines, the manufacturer would add to the 
compliance statement on the engine's label, for residential engines, 
``EMISSIONS COMPLIANCE PERIOD: 50 HOURS,'' and for commercial engines, 
``EMISSIONS COMPLIANCE PERIOD: 300 HOURS.'' Again, EPA believes that 
including the useful life, in hours, on the engine's label, is an 
important mechanism for educating consumers as to the emissions 
durability of the engine. EPA requests comment on whether requiring the 
designation ``EMISSIONS COMPLIANCE PERIOD: 50 RESIDENTIAL HOURS,'' or 
``EMISSIONS COMPLIANCE PERIOD: 300 COMMERCIAL HOURS'' would be more 
effective as the proposed requirement to only include the emissions 
compliance period, by hours, on the label. Similar to the option for 
nonhandheld engines, EPA is requesting comment on an option which would 
allow handheld engine manufacturers to use label statements which 
include a useful life category code (such as A, B, or C) and 
referencing the owner's manual to determine what the code means.
    c. Manufacturer selection of useful life category. One of EPA's 
goals in the proposed Phase 2 program is to assure that engines are 
emissions durable for their useful lives, so that the air quality 
benefits anticipated for the rule are in fact achieved. EPA believes 
that the selection of the appropriate useful life category for an 
engine family is essential to achieving this goal. An appropriate 
useful life selection is important from an emissions compliance 
durability perspective, in terms of assuring that engines meet the 
appropriate emissions standards for the period of time that they are 
expected to be in service. However, EPA is concerned that since the 
useful life of engines, in hours, would be included in certification 
credit calculations for nonhandheld engines, and in-use credit 
calculations for handheld engines, and since these credits have real 
value, a manufacturer may have an important incentive to choose a 
useful life category for a particular family to maximize the 
manufacturer's credit balance, rather than to reflect the most accurate 
useful life selection for that family.
    For example, in the case of a nonhandheld engine family whose FEL 
is significantly below the standard and is therefore generating 
substantial

[[Page 3971]]

credits, a manufacturer could generate four times as many certification 
credits if that family were certified to 1000 hours rather than 250 
hours. Similarly, for a handheld engine family whose in-use test 
results are well below the standard, that family could generate six 
times as many in-use credits if certified to 300 hours rather than 50 
hours. However, in cases where the credit generating engine is not 
expected to be used for 1000 hours (or 300 hours, in the handheld 
example), those clean air benefits may never be realized if the typical 
engine for that family is scrapped substantially before reaching 1000 
hours of use. The ``surplus'' credits might be used to make up for 
higher emissions of other engine families even though the credits were 
generated based on an overestimation of the useful life. On the other 
hand, for engines which are emitting above the standard, the 
manufacturer might have an incentive to certify to the shortest useful 
life period, to minimize the credits needed to offset that engine's 
higher emissions. This could become an even greater concern if that 
engine is in fact expected to be placed into an application which 
experiences longer hours of use than indicated by the selected useful 
life category.
    From an air quality perspective, a consumer education perspective, 
as well as from a marketing or competitive perspective, EPA believes 
that selection of an appropriate useful life is important, and 
certifying an engine to an inappropriate or inaccurate useful life 
presents serious problems. However, no one technical feature of an 
engine model would necessarily dictate that it be placed in one or 
another useful life category, and the distinctions between the useful 
life categories proposed today are not based on objective technical 
differences between engines (e.g., half crank, full crank).
    EPA also recognizes that historically engine manufacturers have not 
always tracked the sale of engines, and may not have been able to 
ascertain the type of application in which an engine is used. On the 
other hand, EPA is also aware that in many cases manufacturers are able 
to determine the end application for a particular engine, and that in 
many cases an engine is designed for a specific end use.
    Manufacturers, stressing that the nonhandheld SOP, as reflected in 
the March 1997 ANPRM, discussed useful life selection as being solely 
at the manufacturer's discretion, have maintained that marketing and 
competitive concerns would ensure that manufacturers select the most 
accurate and appropriate useful life category, and that additional 
requirements that manufacturers support their useful life selections 
are not needed. EPA understands that manufacturers have strong views 
regarding the nonhandheld SOP's discussion of useful life selection. 
However, the SOP indicates that it would be appropriate to certify 
engines to longer useful life categories when they are intended for 
longer hours of operations in-use. The signatories of the SOP further 
recognized that the greater use of an engine during the ozone season 
directly relates to its impact on air quality. In addition, since the 
signing of the SOP, EPA has become concerned that a number of various 
incentives are at play for the manufacturer when it comes to selection 
of a useful life category for an engine, including the requirement to 
demonstrate the engines' emissions durability, testing requirements and 
warranty obligations, generation or use of emissions credits, consumer 
education, and marketing and competitive issues. EPA is concerned that 
a manufacturer might inappropriately select useful life categories for 
certification so as to put itself in a position of competitive 
advantage compared to other manufacturers that fairly and accurately 
select useful life categories, and that the risk of this could cause 
other manufacturers to follow suit in order to remain competitive.
    Therefore, to assure that no individual manufacturer is unfairly 
biasing its useful life selections in order to take advantage of the 
credits programs, EPA is proposing that all manufacturers would declare 
the applicable useful life category for each engine family at the time 
of certification, and would be required to retain at their facilities 
data appropriate to support their selections of useful life categories, 
to be furnished to the Administrator upon request. The manufacturer 
would be required to select the category which most closely 
approximates the actual useful lives of the equipment into which the 
engines are expected to be installed. The rule would also require 
manufacturers to have data supporting their selections sufficient to 
show that the majority of engines or a sales weighted average of 
engines of that family are used in applications having a useful life 
best represented by the chosen category. EPA would not expect to 
request such data unless there is evidence of problems with a 
manufacturer's useful life selections. Such problems might be 
indicated, for example, if all or the major portion of a manufacturer's 
credit-generating engine families were certified to the longest useful 
life categories, or if all or the major portion of a manufacturer's 
credit-using engine families were certified to the shortest useful life 
categories.
    EPA is proposing that data in support of a useful life category 
selection could include: surveys of the life spans of the equipment in 
which the engines are installed; engineering evaluations of field aged 
engines to ascertain when engine performance deteriorates to the point 
where usefulness and/or reliability is impacted to a degree sufficient 
to necessitate overhaul or replacement; warranty statements and 
warranty periods; marketing materials regarding engine life; failure 
reports from engine customers; and engineering evaluations of the 
durability, in hours, of specific engine technologies, engine 
materials, or engine designs. EPA expects that retaining these types of 
data at their facilities would not be unduly burdensome to 
manufacturers, and that in most cases these types of data would be 
information that the manufacturer already has on hand. EPA requests 
comment on these types of data and their usefulness in helping to 
distinguish the most accurate and appropriate useful life category for 
a particular engine family.
    Finally, EPA proposes that in the event that EPA reviewed data 
provided by the manufacturer in support of the useful life selection, 
and upon review of that and such other information available and 
discussion with the manufacturer EPA believed that a different useful 
life category would be more appropriate, the Agency would work with 
that manufacturer to determine a more appropriate selection of useful 
life categories. EPA requests comment on all aspects of this proposal.
5. Certification Averaging, Banking and Trading Program
    With today's notice, EPA is proposing a certification averaging, 
banking and trading (ABT) program for nonhandheld small SI engines. The 
proposed program would be the first ABT program for nonhandheld small 
SI engines. The Phase 1 rule did not include an ABT program due to 
uncertainties regarding the in-use emission levels of engines certified 
to the Phase 1 standards. (The Phase 1 standards apply to ``new'' 
engines and do not require any determination of in-use deterioration as 
the proposed Phase 2 standards do.)
    The Agency is not proposing a certification ABT program for 
handheld engines at this time. Based on the levels of the proposed 
standards and discussion with engine manufacturers, EPA does not 
believe a certification ABT program is warranted or desired for

[[Page 3972]]

handheld engines. The Agency specifically requests comment on this 
issue. As discussed later, EPA is proposing an in-use credit program 
for handheld small SI engines that would be used to address potential 
in-use emission exceedances. The reader is directed to Section IV.D.3 
of today's notice for further details of the proposed in-use credit 
program for handheld engines.
    The nonhandheld small SI engine ABT program proposed today is a 
market-based incentive program designed to provide an incentive for 
early introduction of clean technologies, and provides engine 
manufacturers with additional flexibility for meeting the proposed 
HC+NOX standards, while protecting the environmental 
benefits of the program. Implementation of the program should also 
reduce the cost of controlling HC+NOX emissions from 
nonhandheld engines.
    EPA believes that the proposed ABT program is consistent with the 
statutory requirements of section 213 of the Clean Air Act. Although 
the language of section 213 is silent on the issue of averaging, it 
allows EPA considerable discretion in determining what regulations are 
most appropriate for implementing section 213. The statute does not 
specify that a specific standard or technology must be implemented, and 
it requires EPA to consider costs, lead time, and other factors in 
making its determination of ``the greatest degree of emissions 
reduction achievable through the application of technology which the 
Administrator determines will be available.'' As noted in the proposal 
for Tier I nonroad compression-ignition engine standards, which also 
contained a certification ABT program, section 213(a)(3) also indicates 
that EPA's regulations may apply to nonroad engine classes in the 
aggregate, and need not apply to each nonroad engine individually (see 
58 FR 28809, May 17, 1993).
    At the same time, EPA believes that any ABT program must be 
consistent with the statutory requirement that standards reflect the 
greatest degree of emission reduction achievable through the 
application of available technology. EPA believes the proposed ABT 
program is fully consistent with such a requirement. The proposed 
HC+NOX emission standard of 25.0 g/kW-hr for Class I engines 
and the series of declining HC+NOX standards for Class II 
engines were developed under the assumption that an ABT program would 
take effect at the same time as proposed standards, once adopted. In 
fact, as discussed earlier in Section IV.A.1, the conclusion that the 
proposed standards for Class I and Class II engines are feasible for 
all affected nonhandheld engines within the time available to 
manufacturers, is based in part on the availability of the proposed ABT 
program. In addition, the flexibilities provided to engine 
manufacturers via an ABT program should allow compliance with the 
proposed standard at a lower cost than may otherwise be the case. It is 
also possible that ABT allows the standard to be implemented sooner 
since, for example, not every family may need to be redesigned to meet 
the lower standard. If each engine family had to comply with the 
standards, the standards might be higher and/or the standards might 
need to be implemented later.
    As noted above, the three aspects of the proposed ABT program are 
averaging, banking, and trading. Averaging means the exchange of 
emission credits among engine families within a given engine 
manufacturer's product line. Averaging allows a manufacturer to certify 
one or more engine families at levels above the applicable emission 
standard. However, the increased emissions would have to be offset by 
one or more engine families within that manufacturer's product line 
certified below the same emission standard, such that the average 
emissions in a given model year from all the manufacturer's families 
(weighted for engine power, useful life, load factor, and sales) are at 
or below the level of the emission standard. Averaging results would be 
calculated for each specific model year and, as proposed today, would 
be calculated for each engine class. The mechanism by which this is 
accomplished would be certification of the engine family to a ``family 
emission limit'' (FEL) set by the manufacturer, which may be above or 
below the standard. An FEL that is established above the standard could 
not exceed an upper limit specified in the ABT regulations. Once an 
engine family is certified to an FEL, that FEL would become the 
enforceable emissions limit used for compliance purposes and each 
engine in the engine family would be subject to compliance with the 
FEL.
    Banking means the retention of emission credits by the engine 
manufacturer generating the credits for use in future model year 
averaging or trading. EPA believes that banking, including today's 
proposed provision which would allow early banking under certain 
conditions during the two years prior to implementation of the 
standards, would improve the feasibility of meeting standards by 
encouraging the development and early introduction of advanced emission 
control technology, allowing certain engine families to act as 
trailblazers for new technology. This can help provide valuable 
information to manufacturers on the technology prior to manufacturers 
needing to apply the technology throughout their product lines. An 
incentive for early introduction arises because the banked credits 
could subsequently be used by the manufacturer to ease the compliance 
burden of new, more stringent standards.
    Trading means the exchange of emission credits between engine 
manufacturers which then can be used for averaging purposes, banked for 
future use, or traded to another engine manufacturer. Trading can be 
advantageous to smaller manufacturers who might have limited 
opportunity to optimize their costs through the use of averaging. 
Trading can also be advantageous to larger manufacturers because 
extending the effective averaging set through trading can allow for 
overall optimization of costs across manufacturers.
    EPA is proposing that participation in the proposed ABT program for 
Phase 2 nonhandheld small SI engines would be voluntary. For those 
manufacturers who choose to utilize the program, compliance of 
individual engine families with their FELs would be determined and 
enforced in the same manner as compliance with the emission standards 
in the absence of an ABT program. In addition, except where 
specifically permitted in the case of production line testing failure 
(see section IV.D.2. of today's notice), the final number of credits 
available to the manufacturer in each engine class at the end of a 
model year after considering the manufacturer's use of credits from ABT 
would have to be greater than or equal to zero. Specific elements of 
the proposed ABT program for nonhandheld small SI engines are discussed 
below.
    a. Calculation of Credits. Credits would be calculated as a 
function of the difference between the applicable Phase 2 emission 
standard and the FEL, the power, the useful life, the load factor, and 
the number of eligible engines sold of the engine family participating 
in the program. (Since the standards are expressed in terms of grams/
kW-hour, the ``power'' and ``load factor'' variables are included to 
allow averaging across engines designed to different power.) EPA would 
expect manufacturers to follow the regulations for establishing its 
engine families and not disaggregate their families into multiple 
families or combine their existing families into fewer families to 
maximize credit generation or minimize credit usage.

[[Page 3973]]

EPA is proposing the following equation for calculating the emission 
credits from a given engine family, whether generating positive or 
negative credits.

Credits=(Standard-FEL) x (Power) x  (Useful Life) x (Load Factor) x  
(Sales)

    ``Standard'' represents the applicable Phase 2 emission standard as 
proposed by EPA. ``FEL'' is the family emission limit for the engine 
family as established by the manufacturer. ``Power'' represents the 
engine's maximum modal power produced during the certification test 
cycle. For those engine families that contain more than one 
configuration with different power ratings, EPA is proposing that the 
``Power'' term be the sales-weighted maximum modal power determined 
across all configurations within the engine family. EPA assumes 
manufacturers know the general power characteristics of each of their 
engine configurations they are producing, and therefore, determining 
the power information necessary for the ABT calculations will not place 
any additional testing burden on manufacturers. EPA requests comment on 
this assumption.
    ``Useful Life'' is the useful life category to which the engine 
family is certified, and represents the period of time for which the 
manufacturer is responsible for compliance with the emissions 
standards. ``Load Factor'' refers to the fraction of rated power at 
which the engine operates in use, on average. For the two main 
certification test cycles, referred to as cycle ``A'' and cycle ``B'', 
which EPA believes represent typical in-use operation, a load factor of 
0.47 is proposed. For alternative test cycles, as approved by EPA, the 
load factor would need to be calculated based on the characteristics of 
the test procedure as described in the proposed regulations.
    ``Sales'' represents the eligible number of Phase 2 engines sold in 
the United States in the applicable model year, excluding those engines 
subject to California regulations. Manufacturers would be allowed to 
use sales projections for initial certification. However, actual sales 
based on the location of the point of first retail sale (for example, 
retail customer or dealer) would have to be submitted at the end of the 
model year to verify end-of-year compliance. The Agency is proposing 
that manufacturers exclude engines subject to California's emission 
standards from the estimates of eligible engine sales because 
California will likely require all engines sold in California to meet 
its own tighter HC+NOX standards. If California engines were 
included, then the credits generated by California sales would allow 
more engines with higher emission rates to be sold in states outside of 
California. This would detract from the goals of the Phase 2 program, 
and possibly undermine the emissions reductions expected to be achieved 
by the program throughout the country. Engines sold outside of the 
United States, including Canada and Mexico, would also be excluded from 
the manufacturer's estimates of sales unless those engines are 
subsequently imported back into the United States in a new piece of 
nonhandheld equipment.
    Because only those engines sold in the United States, excluding 
engines subject to California's standards, would be included in the ABT 
program, manufacturers would need to determine the number of such 
engines sold each year to yield accurate estimates of credit generation 
and usage. Due to the difficulty in tracking point of first retail 
sales in the nonhandheld market compared to other markets (e.g., the 
on-highway segment where a more direct engine and vehicle distribution 
system exists), EPA is requesting comments on alternative methods 
manufacturers could use to determine their eligible sales for credit 
calculations. One possible option would be to allow engine 
manufacturers to query their customers, on an annual basis, to 
ascertain the percentage of Phase 2 engines of each family that 
constitute eligible sales. Based on the results of the query, the 
Agency could allow manufacturers to extrapolate those results, assuming 
they received responses sufficient to cover some high percentage of 
their sales, say 90 percent or more, to its total sales of engines in 
the United States. The Agency is open to considering other alternative 
methods for tracking engines for credit calculation purposes that 
provide high levels of confidence that eligible sales are accurately 
counted. EPA specifically requests comments on such alternatives and 
other information that would further address the Agency's concerns that 
eligible sales estimates be as accurate as possible. In addition, the 
Agency requests comments on appropriate methods for estimating the 
export of engines and the sales of engines subject to California's 
standards, since one method for estimating eligible sales for ABT 
purposes could be to deduct these two groups from total sales.
    As discussed in Section IV.E of today's notice, EPA is proposing 
several compliance flexibility provisions for engine manufacturers and 
equipment manufacturers that would allow the limited use of Phase 1 
engines in the Phase 2 time frame. To avoid penalizing manufacturers 
that produce engines to be used under the proposed flexibility 
provisions, EPA is proposing that manufacturers exclude such engines 
from the ABT program calculations. In other words, engine manufacturers 
would not be required to use credits to certify these Phase 1 engines 
used for the proposed flexibility provisions even though they would 
likely exceed the proposed Phase 2 standards.
    Another proposed flexibility provision described in Section IV.E of 
today's notice would allow engine manufacturers to certify beyond the 
2005 model year Class II side-valve engine families with annual sales 
of less than 1,000 units to an HC+NOX cap of 24.0 g/kW-hr. 
For such engine families, EPA is proposing that manufacturers do not 
need to include such families in the ABT program calculations for 2005 
and later model years. For the interim years, 2001 through 2004, a 
manufacturer could also exclude Class II side-valve engine families 
with annual sales of less than 1,000 units from the ABT program 
calculations as long as the deteriorated HC+NOX emission 
level of the engine is less than 24.0 g/kW-hr. Class II side-valve 
engine families with annual sales of less than 1,000 units that are 
certified above the 24.0 g/kW-hr HC+NOX level must be 
included in the manufacturers' ABT calculations during the interim 
years.
    EPA is proposing an upper limit on the level of emissions allowed 
from those engine families a manufacturer wishes to include in the ABT 
program. Under the proposal, manufacturers would not be allowed to 
certify engines that have FELs above the upper limits described below. 
Typically, when EPA adopts an ABT program, the upper limit is set at 
the level of the previous standard. However, because the Phase 1 
standards did not require manufacturers to take into account 
deterioration over the useful life of the engine as the proposed Phase 
2 standards do, EPA believes it is appropriate to use the Phase 1 
standards as the basis for calculating the upper limits and apply a 
deterioration factor to determine the equivalent deteriorated level of 
the Phase 1 emission standards. Based on the predominant side-valve 
engine technology certified under the Phase 1 program, EPA estimates 
that a typical Phase 1 engine would have emissions at the end of the 
useful life period about twice its new engine emission 
level.31

[[Page 3974]]

Therefore a deterioration factor of 2.0 is appropriate for estimating 
the equivalent useful life level of engines designed to meet the Phase 
1 standards. Based on the Phase 1 HC+NOX standards and a 
deterioration factor of 2.0, EPA is proposing HC+NOX upper 
limits of 32.2 g/kW-hr for Class I engines and 26.8 g/kW-hr for Class 
II engines. Therefore, a manufacturer would be allowed to certify an 
engine family only if the HC+NOX FEL were at or below these 
proposed levels (and only if they had the appropriate number of credits 
to offset the family's credit needs). For families not participating in 
the ABT program, each family must comply with the standard which in 
effect is an analogous upper limit. EPA requests comment on the 
appropriateness of the proposed upper limits for engine families 
included in the ABT program.
---------------------------------------------------------------------------

    \31\ See ``Summary of EPA Analysis Regarding Upper Limits for 
Phase 2 Averaging, Banking & Trading Program for Nonhandheld 
Engines'', Item #II-B-05 in EPA Air Docket A-96-55.
---------------------------------------------------------------------------

    Due to concerns over the amount of credits manufacturers could 
accumulate, as described below, EPA is proposing a declining set of 
caps on how high the sales-weighted average level of HC+NOX 
FELs could be for Class II engine families beginning in 2005. Based on 
the certification information of Phase 1 nonhandheld engines submitted 
by manufacturers to EPA and assumptions about typical deterioration 
factors and compliance margins, it appears that some engine 
manufacturers have the potential to earn significant credits from their 
Class II engines prior to the 2005 model year. (Because the proposed 
emission standard for Class I engines assumes side-valve technology and 
because most Class I engines are expected to remain side-valve 
technology, it does not appear that there would be the same potential 
for significant credit generation by Class I engine manufacturers.) 
Manufacturers who adopt OHV technology earlier than anticipated by the 
proposed Class II phase-in standards appear best positioned to 
accumulate significant credits. The ability to generate credits during 
the transition years would occur primarily because the typically lower-
emitting OHV engines could earn credits up to the proposed applicable 
model year standards (which, as noted earlier, would decline for each 
model year between 2001 and 2005 and assume an industry changeover to 
the cleaner OHV engines from the higher-emitting side-valve engines).
    The environment benefits when a manufacturer produces engines 
which, on average, are cleaner than required during the transition 
years. However, EPA is concerned that some manufacturers, because their 
current product line is predominantly made up of OHV technology, would 
be able to accumulate significant credits during the phase-in years 
without any additional effort to improve emission performance. These 
credits could be, in turn, used by such manufacturers beginning in 2005 
to, in effect, delay the need for that manufacturer to produce engines 
meeting the proposed 2005 model year standard. This action could put 
such manufacturers in a competitively advantageous position compared to 
manufacturers who did not have substantial credits and therefore needed 
to produce a product line which, on average, met the 2005 model year 
standard. Such action could similarly undermine the goal of this rule 
(and the SOP) to have 100 percent OHV technology (or similar technology 
meeting the 2005 model year standards) in place across the industry for 
Class II by 2005.
    In order to ensure that this transition to cleaner technology 
occurs by the 2005 model year and to minimize the risk of credit 
``build-up'' resulting in a delay of conversion to OHV or OHV-
comparable technology, EPA is proposing that a manufacturer's sales-
weighted average of Class II HC+NOX FELs may not exceed 13.6 
g/kW-hr in 2005, 13.1 g/kW-hr in 2006, and 12.6 g/kW-hr in 2007 or 
later. EPA believes this approach would ensure that Class II engines 
are converted to OHV or OHV-comparable technology by roughly 2005 while 
still encouraging the early introduction of cleaner, more durable 
technology and ensuring that manufacturers have the flexibility they 
need to comply with the proposed standards. EPA requests comment on the 
proposed caps and alternative approaches that would ensure the 
introduction of OHV or OHV-comparable technology by approximately 2005 
while maintaining the flexibility offered to manufacturers by ABT and 
the encouragement to pull ahead cleaner, more durable technology.
    As described earlier, EPA is proposing separate NMHC+NOX 
standards for natural gas-fueled engines which are intended to be as 
stringent as the proposed HC+NOX standards for the remaining 
nonhandheld small SI engines. All credit calculations for natural gas-
fueled engines would be calculated against those standards. In 
addition, because the proposed standards are equivalent in stringency, 
and the market for nonhandheld natural gas-fueled small SI engines is 
extremely small (i.e., less than 0.1 percent of current nonhandheld 
sales), EPA is proposing to allow manufacturers to freely exchange 
NMHC+NOX credits from nonhandheld engines fueled by natural 
gas with HC+NOX credits from nonhandheld engines fueled by 
fuels other than natural gas in the ABT program.
    b. Life of Credits. For all credits generated by Class I and Class 
II engines under the certification ABT program, EPA is proposing an 
unlimited credit life. EPA believes that unlimited life for these 
credits will promote the feasibility of the proposed Phase 2 Class I 
and Class II standards because it increases the value of these credits 
to the manufacturer by providing greater flexibility for the use of the 
credits. It is consistent with the general emission reduction goal of 
ABT programs, not only because of the increased manufacturer incentive 
but also because it reduces the incentive for manufacturers to use 
their credits as quickly as possible. As a result, unused credits, 
which are extra emission reductions beyond what the EPA regulations 
require, may remain off the market longer. It should be noted that EPA 
would expect to reconsider the appropriate life of Phase 2 emission 
credits in connection with any post-Phase 2 rulemaking for nonhandheld 
engines.
    c. Early Use of the ABT Program. EPA is proposing that 
manufacturers be allowed to use the ABT program prior to implementation 
of the Phase 2 standards to provide an incentive to accelerate 
introduction of cleaner technologies into the market. The Agency 
believes that making bankable credits available prior to 2001 would 
reward those manufacturers who take on the responsibility of complying 
with the proposed standards sooner than required and would result in 
early environmental benefits. Under the proposed provisions, 
manufacturers would be allowed to begin using portions of the ABT 
program starting two model years before the proposed standards take 
effect provided the manufacturer certifies and complies with the 
proposed 2001 model year standards of 25.0 g/kW-hr for Class I engines 
and 18.0 g/kW-hr for Class II engines for their entire product line in 
a given nonhandheld engine class. The manufacturer could show it is in 
compliance with the proposed standards for each individual engine 
family or on average using the averaging provisions of the proposed ABT 
program. If a manufacturer meets this condition, the manufacturer could 
generate early credits to be banked for use in the 2001 or later model 
years

[[Page 3975]]

from only those engines certified below 16.0 g/kW-hr HC+NOX 
for Class I engines and below 12.1 g/kW-hr for Class II engines (or 
15.0 g/kW-hr NMHC+NOX for Class I natural-gas fueled engines 
and 11.3 g/kW-hr for Class II natural-gas fueled engines). However, all 
early credits would be calculated against the initial Phase 2 standards 
of 25.0 g/kW-hr HC+NOX for Class I engines and 18.0 g/kW-hr 
HC+NOX for Class II engines (or the corresponding 
NMHC+NOX standards of 23.0 g/kW-hr and 16.7 g/kW-hr, 
respectively, for natural-gas fueled engines). If the manufacturer 
certifies its product line to the proposed Phase 2 standards early 
through the use of averaging, the manufacturer could bank credits for 
use in 2001 and later, but could only bank credits from those engines 
which were not needed to show early compliance with the proposed Phase 
2 standards. In other words, manufacturers would not be allowed to bank 
credits from engines whose credits were already used to offset other 
engines with FELs above the proposed Phase 2 standards. This would 
prevent manufacturers from ``double counting'' credits needed to show 
early compliance with the proposed standards. Manufacturers would not 
be allowed to trade their early credits to other manufacturers until 
the 2001 model year or later.
    In establishing the proposed set of declining standards for Class 
II engines, EPA assumed a certain phase-in of OHV or comparably clean 
and durable technology. As described in the March 1997 ANPRM, the 
proposed series of Class II HC+NOX standards were based on 
the assumption that 50 percent of Class II engines would employ OHV or 
comparably clean and durable technology in 2001 (i.e., could meet a 
12.1 g/kW-hr HC+NOX standard without the use of credits). 
For the remaining years, the phase-in schedule assumed for ``OHV 
emission performance'' (``OEP'') technology was 62.5 percent in 2002, 
75 percent in 2003, 87.5 percent in 2004, and 100 percent in 2005. EPA 
believes this phase-in of OHV or comparably clean and durable 
technology is important due to the inherent emission benefits 
anticipated from this technology in use. Related to the concerns 
discussed above regarding credit life for pre-2005 credits, the Agency 
is concerned that manufacturers of Class II engines could bank early 
credits and use such credits to continue certifying a line of engine 
families that do not meet the OEP production phase-in schedule assumed 
by EPA in establishing the proposed standards. Therefore, EPA is 
proposing that manufacturers only be allowed to use early banked 
credits beginning in 2001 or later if they are meeting the OEP 
production phase-in schedule estimates for that model year. EPA 
believes prohibiting the use of early banked credits unless 
manufacturers meet such conditions will encourage the manufacturers to 
meet the OEP production phase-in schedule assumed in developing the 
proposed Phase 2 standards.
    d. Cross-Class Exchange of Credits for Certification Purposes. 
Today's proposal contains limitations on the cross-class exchange of 
credits during certification. The limitations are meant to assure the 
ABT program fulfills its intended function of encouraging a transition 
to cleaner, more durable technology for both classes of nonhandheld 
engines and achieves the expected environmental benefits of the 
program. The proposed limitations are also intended to assure that the 
proposed ABT program does not affect competition between engine 
manufacturers.
    With regard to encouraging cleaner, more durable technology, the 
proposed schedule of standards for Class II engines was established 
with the assumption that engine manufacturers will phase-in OHV 
technology over roughly the five year period from 2001 to 2005 based on 
the schedule noted earlier. In order to encourage manufacturers to 
follow the assumed OEP production phase-in schedule, EPA is proposing 
that limited cross-class exchange of credits for certification 
purposes, as noted below, would be allowed only if a manufacturer's 
Class II engine production meets or exceeds the assumed OEP production 
phase-in schedule for Class II engines presented earlier.
    With regard to competition in the nonhandheld market, about two-
thirds of nonhandheld engine manufacturers currently produce both Class 
I and Class II engines. The remaining one-third of the nonhandheld 
engine manufacturers produce only Class II engines. At this time, EPA 
is not aware of any nonhandheld engine manufacturers that only produce 
Class I engines. Allowing manufacturers to exchange credits across 
engine classes could cause a competitive disadvantage for those 
manufacturers who only produce Class II engines because they would not 
have the advantage of being able to use positive credits from Class I 
engines. Therefore, with regard to the cross-class exchange of credits, 
EPA is proposing that manufacturers would be allowed to exchange 
credits from credit generating Class II engines to credit using Class I 
engines for certification purposes. However, due to the competitive 
concerns noted above, EPA is not proposing to allow the exchange of 
credits from credit generating Class I engines to credit using Class II 
engines for certification purposes.
    e. Use of Credits to Address Nonconformity Determined After 
Certification. As noted elsewhere in today's notice, EPA is proposing a 
number of provisions that address post-certification compliance aspects 
of the proposed standards. In two specific cases, EPA is proposing to 
allow manufacturers to use credits from the certification ABT program 
to address noncompliance determined after the time of certification. As 
noted in the discussion on compliance, EPA does not believe that the 
typical type of enforcement action that could be taken when a 
substantial nonconformity is identified (i.e., an engine family recall 
order) would generally be workable for nonhandheld small SI engines 
given the nature of the nonhandheld market. Whereas handheld engine 
nonconformities after certification would be addressed through the use 
of in-use credits, EPA is not proposing an in-use credit program for 
nonhandheld engines, as discussed in Section IV.D.
    Instead, EPA is proposing to allow manufacturers to use 
certification ABT credit to address two different types of 
nonconformance. First, manufacturers would be allowed to use ABT 
credits to offset limited emission shortfalls for past production of 
engines determined through the Production Line Testing (PLT) program as 
described in Section IV.D.2. of today's notice. Second, manufacturers 
would be allowed to use ABT credits to offset emission shortfalls from 
Class II OHV engines that arise as a result of an adjustment to 
deterioration factors originally determined through good engineering 
judgement, as described in Section IV.E of today's notice. Under the 
proposed provisions, manufacturers would be allowed to use all credits 
available to them to offset such emission shortfalls. EPA does not 
believe it is necessary to limit the use of cross-class credits for 
these situations. Allowing manufacturers to exchange credits from one 
class to another should not raise the same concerns with regard to new 
engine competition as noted earlier because the manufacturer is 
addressing a nonconformance problem for engines that have already been 
sold and used in the field for a significant period of time. EPA 
requests comment on the proposed provisions for using certification ABT 
credits to address nonconformance with

[[Page 3976]]

the Phase 2 emission standards determined after certification.
    EPA is not proposing to allow manufacturers to use ABT credits to 
remedy a past production nonconformance situation in the Selective 
Enforcement Audit (SEA) program. As described in today's notice, EPA is 
planning to primarily rely on the PLT program to monitor the emissions 
performance of production engines. However, in the case of nonhandheld 
engines only, manufacturers would in some cases have the option of 
traditional SEA in lieu of PLT as a production line compliance program. 
In addition, SEAs could be conducted in cases where EPA has evidence of 
improper testing procedures or nonconformities not being addressed 
through PLT. As discussed in section IV.D.3, if EPA determines that an 
engine family is not complying with the standards as the result of an 
SEA, EPA plans to work with the manufacturer on a case-by-case basis to 
determine an appropriate method for dealing with the nonconformity. The 
option(s) agreed upon by EPA and the engine manufacturer may, or may 
not, include the use of ABT credits to make up for any ``lost'' 
emission benefits uncovered by the SEA.
    As noted earlier, EPA solicits comments on all aspects of the 
proposed ABT program, including comments on the benefit of the program 
to manufacturers in meeting the proposed emission standards and any 
potential air quality impacts which might be associated with them.
6. Certification Fuel
    The program for nonhandheld engines discussed in the March 1997 
ANPRM specified that the proposed range for eligible certification 
fuels for Phase 2 would be the same as under Phase 1. The program for 
handheld engines in the ANPRM was silent on this issue. EPA received 
comment on the ANPRM that the continued use of Phase 1 certification 
fuels for Phase 2 testing is appropriate so long as the same fuel may 
be used to certify handheld engines under both EPA and CARB 
regulations.
    EPA is proposing today that certification test fuel requirements 
for the Phase 2 program would remain the same as in the Phase 1 
program, as specified at 40 CFR 90.308(b). While California ``Phase 2'' 
reformulated gasoline is not a proposed certification test fuel, EPA 
believes that continuation of the Phase 1 program for Phase 2 would 
continue to provide a means of harmonizing the Federal and California 
programs. As described in the February 1997 Draft U.S. EPA Small Engine 
Certification Guidance, Section X ``Certification Fuel'', manufacturers 
have four options for choice of certification fuel for Phase 1 
32; EPA is proposing that these options would continue for 
this rule.
---------------------------------------------------------------------------

    \32\ See ``U.S. EPA Small Engine Certification Guidance, Draft, 
February 19, 1997,'' available in EPA Air Docket A-96-55, Item #II-
C-03.
---------------------------------------------------------------------------

    The first option is to use average in-use gasoline specified at 40 
CFR Part 90, Subpart D, Appendix A, Table 3. The second option is 
federal certification fuel (e.g., Indolene), specified at 40 FR 
86.1313-94(a), Table N94-1. Third, manufacturers may use other fuels, 
such as natural gas, propane, methanol, or others, under conditions 
described at 40 CFR 90.308(b)(2) and (3). Fourth, manufacturers may 
request EPA approval for certification testing on fuels such as 
California ``Phase 2'' reformulated gasoline, which do not meet the 
requirements for ``other fuels'' under 40 CFR 90.308(b)(2) or (3). For 
this option, manufacturers would request EPA approval of an alternate 
test procedure (e.g., alternate test fuel) under 40 CFR 90.120(b)(1). 
Manufacturers may elect to use an alternative test procedure provided 
it yields results equal to the results from the specified test 
procedures (e.g., test fuels described at 40 CFR 90.308(b)), its use is 
approved by EPA, and the basis for equivalent results is fully 
described in the manufacturer's certification application (see 40 CFR 
90.120(b)(1)). EPA would work with manufacturers to assist them in 
making the required technical demonstrations to show equivalency of the 
emission results. The continuation of these Phase 1 certification fuel 
requirements would continue to provide mechanisms for manufacturers to 
use the same fuel for certification to both EPA and California Air 
Resources Board regulations, as specified above.

B. Test Procedures

    Test procedures are contained in today's proposal which would be 
used by engine manufacturers for the purpose of measuring emissions and 
determining emission rates for regulated emissions for certified 
engines. The test procedures being proposed today are in most respects 
identical to the procedures required for the certification of Phase 1 
engines. Test procedures were discussed during the Regulatory 
Negotiation process, with the key issue being the appropriateness of 
the Phase 1 test cycles for Phase 2 engines. The draft Regulatory 
Support Document for this proposal contains a summary of the test 
procedure issues addressed during the Regulatory Negotiation process.
    In general, the Agency believes the Phase 1 test procedures are 
appropriate for measuring engine emissions from Phase 2 
engines.33 In today's action, EPA is proposing the Phase 1 
test procedures with the following minor changes. First, nonhandheld 
engines sold with an engine rotational speed governor would have to use 
the governor for speed control while running the appropriate test 
cycle. Second, the mode weightings for the handheld test cycle, Cycle 
C, would be adjusted to 0.85 for Mode 1 and 0.15 for Mode 2. Finally, 
appropriate changes to the test procedure and emission calculations 
have been proposed for the measurement of methane from natural gas 
fueled engines in order to determine non-methane hydrocarbon emissions 
for natural gas fueled nonhandheld engines. These proposed changes are 
discussed below. EPA requests comment on these issues.
---------------------------------------------------------------------------

    \33\ For a discussion on the adequacy of the Phase 1 test 
procedure, see Chapter 1.1 in ``Regulatory Support Document, Control 
of Air Pollution, Emission Standards for New Nonroad Spark-Ignition 
Engines At or Below 19 kilowatts'' U.S. EPA, May 1995, EPA Air 
Docket A-93-25, Item #V-B-01.
---------------------------------------------------------------------------

1. Test Cycle: Requirement for the Use of a Speed Governor Operation 
for Testing of Nonhandheld Engines
    Many small engines manufactured today make use of a speed control 
governor (``governor'') to regulate engine rotational speed. In 
general, the governor is a mechanically or electronically controlled 
device that attempts to maintain engine rotational speed in a 
particular range as the engine experiences different loads. A typical 
example is the walk-behind mower, where the governor is designed to 
control engine throttle position in response to various loads to 
maintain the engine's rotational speed, and thus, mower blade rotating 
speed, to provide an adequate grass cut. For the Phase 1 test 
procedure, manufacturers are allowed to over-ride or disconnect the 
speed governing device and use an external piece of equipment, i.e., a 
throttle controller, for the purpose of replicating the speed and load 
conditions required by the test cycle (see 40 CFR 90.409(a)(3)). After 
the finalization of the Phase 1 rule during the regulatory negotiation 
process, the Test Procedure Task Group formed by the Regulatory 
Negotiation committee recognized that the use of the engine's designed 
governor, not an external throttle controller, may be a more accurate 
prediction of an engine's in-use performance. The Test Procedure Task 
Group members generally agreed that a

[[Page 3977]]

Phase 2 test procedure should require the use of the engine's speed 
governor for speed control during the Federal Test Procedure (FTP) for 
those engines which are equipped by the manufacturer with a speed 
governor. However, there was not general agreement or detailed 
discussion of the specific requirements of how the speed governor 
should be used during the FTP. At this time the Agency believes the 
most appropriate method to operate engines on the speed governor for an 
emissions test would be to use fixed throttle operation for the 100 
percent load mode, and then to use the engine governor for all 
subsequent power modes (75 percent, 50 percent, 25 percent and 10 
percent). For each power mode, the engine speed governor set-point 
would be adjusted to the nominal test cycle set-point, 85 percent of 
rated speed for Cycle A, and 100 percent rated speed for Cycle B. This 
test method allows for a consistent and repeatable method of 
determining the 100 percent load condition, yet would allow the 
engine's governor to regulate speed for the remaining load conditions. 
This method is also straightforward and would be relatively simple to 
implement in a laboratory. The Agency requests comment on this test 
method and on other test methods which may be more appropriate.
2. Test Cycle: Adjustments for Weightings for 2-Mode Cycle for Handheld 
Engines
    The Agency is proposing a change in the weighting factors for the 
handheld test procedure. For the Phase 1 rule, a weighting factor of 90 
percent is applied to the 100 percent power mode, and a factor of 10 
percent is applied to the idle mode, in order to combine the modal 
results for the final weighted emission value. The Agency is proposing 
for Phase 2 that a weighting factor of 85 percent is used for the 100 
percent power mode, and 15 percent be used for the idle mode. This 
proposal is based on a study performed by members of PPEMA during the 
regulatory negotiation process.34 PPEMA members collected 
real-time speed and throttle position data on several types of handheld 
equipment used during actual in-use operation. This data was analyzed 
and combined with estimates of annual use, load factors, and annual 
sales to weight the results of the field testing. EPA's summary of this 
report is contained in the Draft RSD. The Agency agrees with the 
report's conclusion that a more appropriate set of weighting factors 
for handheld engines is 85 percent for the 100 percent power mode and 
15 percent for the idle mode. Therefore this change is being proposed 
for Phase 2.
---------------------------------------------------------------------------

    \34\ See ``Hand Held Composite Duty Cycle Report'', February 
1995, prepared by members of the Portable Power Equipment 
Manufacturers Association, available in EPA Air Docket A-96-55, Item 
# II-D-18.
---------------------------------------------------------------------------

3. Measurement of NMHC Emissions From Natural Gas Fueled Nonhandheld 
Engines
    In order to accommodate the proposed optional non-methane 
hydrocarbon (NMHC) standard for natural gas fueled nonhandheld engines, 
the Agency is proposing to incorporate by reference the appropriate 
sections from 40 CFR Part 86 which relate to the measurement of methane 
emissions from spark-ignited engines. These appropriate sections were 
published as part of a final rulemaking titled ``Standards for 
Emissions From Natural Gas-Fueled, and Liquefied Petroleum Gas-Fueled 
Motor Vehicles and Motor Vehicle Engines, and Certification Procedures 
for Aftermarket Conversions'' see 59 FR 48472, published on September 
21, 1994. The specific sections being incorporated can be found in the 
proposed regulatory language contained in this proposal at 
Sec. 90.301(d) and Sec. 90.401(d).

C. Field/Bench Adjustment Program

    The ANPRM contemplates a so-called ``bench field correlation 
program'' for both handheld and nonhandheld small spark ignited 
engines.35 For handheld engines, it is part of the in-use 
testing program (ANPRM, Appendix A, Section J(2)); for nonhandheld 
engines, it is part of the certification program (ANPRM, Appendix B, 
Sections 4(a) and (b)). In either case, the basic premise for these 
programs is the same: to allow manufacturers to age engines on the 
bench to demonstrate expected compliance in-use, it is necessary to 
demonstrate the ``correlation'' between field aging and bench aging.
---------------------------------------------------------------------------

    \35\ The use of the term ``correlation'' was meant to describe 
an adjustment factor that can be applied to bench-aged engines to 
approximate field-aged conditions, and not a true statistical 
correlation.
---------------------------------------------------------------------------

    The ANPRM sets out slightly different requirements for the proposed 
handheld and nonhandheld programs. Specifically, the ANPRM stipulates 
that the handheld correlation program would be conducted under EPA 
guidance; a portion of the engines would be aged in situations in which 
the manufacturer does not exercise control over the engines' 
maintenance, or limit their usage such that the engines are no longer 
used in a way that is representative of typical in-use engines; the 
full federal test procedure would be used; all pollutants would be 
measured; residential engines would be aged to their full regulatory 
life but commercial engines could be aged to 75 percent of their full 
regulatory life; samples sizes would be determined in the NPRM process; 
and there would be periodic spot checks of the correlation (ANPRM, 
Annex A, Section J(2)).
    The ANPRM provisions for the nonhandheld engines are less 
comprehensive. For this category, the correlation program was 
specifically discussed for engines using side-valve or aftertreatment 
technologies. In addition, the ANPRM describes a simple ``correlation'' 
method (ratio of mean emission rates); would require periodic re-
calculation (every other year for the first five years of the program 
and then every five years thereafter, e.g., 2001, 2003, 2005, 2010, 
2015, etc.); and calls for changes in the correlation to apply 
prospectively only.
    In today's NPRM, EPA is proposing a unified program, to be called 
the ``field/bench adjustment program,'' 36 that would apply 
to both nonhandheld engines that use side-valve or aftertreatment 
technologies and to handheld engines. EPA believes it is appropriate to 
design one program to apply to both categories of engines both because 
it is less complicated for manufacturers that produce both kinds of 
engines and because it simplifies the compliance program for 
administrative purposes. EPA seeks comment on the application of the 
same program and methodology to both categories of engines. The 
remainder of this section will set out the background for field/bench 
adjustment and the principles of such a program, a proposed 
methodology, and various practical requirements for the application of 
the program. It will end with a brief discussion of an alternative 
methodology.
---------------------------------------------------------------------------

    \36\ This nomenclature more accurately reflects the purpose of 
the program.
---------------------------------------------------------------------------

1. Background and Principles
    There are at least three ways to demonstrate compliance with in-use 
standards such as those proposed in today's rule. In general, the most 
representative way is to demonstrate compliance on engines that have 
been aged to their full regulatory lives by actual end-users. This 
ensures that the emissions reflect actual in-use conditions, including 
the presence of dirt and other matter such as clippings, operation at 
several degrees of orientation, operation in very hot ambient 
temperatures, etc. At the same time, consumer-based field aging is

[[Page 3978]]

difficult, not the least because it is cumbersome to organize a program 
with a sufficient number of end-users. In addition, it may take some 
end-use consumers years to put an appropriate number of hours on the 
engine through normal use.
    The second method is to demonstrate compliance on engines that have 
been aged to their full regulatory lives on the bench. While this 
method can be more practical for the manufacturer, it also abstracts 
away many operational or environmental conditions that can affect 
deterioration.
    The third way, and the way being proposed in today's notice, is a 
consolidation of some elements of the other two methods. Under it, 
manufacturers could bench age engines and then adjust the emission test 
results to reflect actual in-use conditions as represented by field 
aging. This would be accomplished by developing a field/bench 
adjustment factor that would be applied to emissions from bench-aged 
emissions to simulate field aging.
    Thus, the objective of this field/bench adjustment program is to 
develop an adjustment factor based on the mathematical relationship 
between emissions from field-aged and bench-aged engines. For obvious 
reasons, it is very important to design a field/bench adjustment 
program that will yield an adjustment factor that is as closely related 
as possible to the true relationship between field and bench aging. Any 
deviation will result in an adjustment factor that either under-
corrects or over-corrects the bench results, the ultimate result being 
an impact on the stringency of the emission limits. In addition, this 
field/bench adjustment program should take advantage of statistical 
techniques, both to take into account the inherent uncertainty in 
sampling 37 and to allow EPA to impose some restrictions on 
the use of this simplified compliance method. In today's notice, EPA is 
proposing to allow manufacturers to use the simple ratio of the field 
and bench mean emission results as an adjustment factor if the width of 
a confidence interval around the bench-aged and field-aged mean 
emission rates does not exceed a certain percentage of the standard. 
This restriction would limit the emission results for each sample, 
permitting a closer fix on the true population relationship.
---------------------------------------------------------------------------

    \37\ To take full advantage of the field/bench adjustment 
program, engine manufacturers will presumably prefer to bench and 
field age only a relatively small number of engines. Thus, the 
results of the program will heavily depend on the characteristics of 
the sample (it is generally the case that a different sample would 
have different emission results and a different adjustment factor).
---------------------------------------------------------------------------

2. General Methodology
    Drawing on the elements of the ``bench field correlation program'' 
set out in the ANPRM and the criteria discussed above, EPA is proposing 
the following methodology to calculate the adjustment factor that would 
be applied to bench-aged emissions to approximate field aging. EPA 
seeks comments on all aspects of this program.
    Two samples of engines would be aged, one in the field and one on 
the bench. The aging procedures for all engines in the field sample 
would be the same, and the aging procedures for all engines in the 
bench sample would be the same. The manufacturer would develop a test 
plan which would specify the conditions under which the engines would 
be aged on the bench and in the field. EPA would reserve the right to 
review any test plan, for handheld or nonhandheld engines, and to 
require the manufacturer to revise it if it does not reflect 
appropriate testing conditions. This review would enable EPA to 
exercise some oversight of the program without requiring the entire 
program to be performed under EPA guidance, as anticipated in the 
handheld program described in the ANPRM. With regard to sample size, 
today's proposed program contains only two constraints: the bench-aged 
and field-aged samples must initially be of equal size and must contain 
at least three engines. This minimum number is necessary to perform the 
statistical tests described below.
    Next, each engine would be tested on the full federal test 
procedure after it has been run for its useful life. Then, for each 
sample, the mean HC+NOX emission rate would be calculated 
and two independent confidence intervals would be constructed, one 
around the mean of the field-aged engines, and one around the mean of 
the bench-aged engines, using the student's T distribution and a 90% 
confidence level.
    The formula for the confidence interval would be:
    [GRAPHIC] [TIFF OMITTED] TP27JA98.000
    
where

    x is the sample mean,
    t(1-/2; n-1) is the appropriate parameter from Student's t 
table,
    depending on the level of confidence chosen by EPA,
    s is the sample standard deviation, and
    n is the number of engines in the sample.

    The width of each confidence interval would then be compared to the 
``maximum allowable interval width'' proposed today. EPA is proposing 
+/-20% of the standard as the maximum allowable interval width. If the 
confidence intervals around each of the field-aged and bench-aged means 
each are no wider than the maximum allowable interval width (e.g, +/
-20% of the standard), then the adjustment factor that would be applied 
in the future to bench-aged engines to simulate field aging would be 
the ratio of the means (x.8F/
x.8B), provided this ratio is greater than or 
equal to one.
    EPA is proposing that these constraints be applied to both handheld 
and nonhandheld engines, but seeks comment as to whether the confidence 
levels and maximum allowable interval widths should be different among 
them. EPA chose 90% confidence levels for constructing the confidence 
intervals for the field-aged and bench-aged engines, and +/-20% of the 
standard maximum allowable interval widths, based on computer 
simulations 38, 39; however, manufacturers or others 
commenting on this proposal may have information that suggest other 
levels.
---------------------------------------------------------------------------

    \38, 39\ See ``Simulation to Determine Confidence Level and 
Maximum Allowable Interval Width for Field/Bench Adjustment Factor 
Program,'' EPA Air Docket A-93-29, Item #II-B-01.
---------------------------------------------------------------------------

    Under the proposed program, if either or both of the confidence 
intervals do not pass the above-described statistical test, the 
manufacturer would have the choice of three remedies. First, the 
manufacturer could increase the size of the failing sample and repeat 
the statistical tests with the increased number of engines. Often, 
increasing the size of the sample will lead to a smaller sample 
variance, although this is not always the case with small samples. A 
manufacturer could repeat this remedy as many times as desired. Note 
that it would not be necessary to increase the size of both samples; 
only the sample that failed the statistical test would need to be 
increased. Alternatively, if the statistical tests are failed, the 
manufacturer could adjust the test plan and rerun the program, subject 
to EPA approval. In the third alternative, the manufacturer could 
choose to age all engines in the field for the purposes of the 
compliance program.
3. Practical Requirements of the Program
    This section describes several practical elements of this proposed 
field/bench adjustment program and how it would work if adopted as 
proposed.

[[Page 3979]]

    a. Initial Field/Bench Adjustment Factor Calculation. The ANPRM 
does not discuss an initial date by which the first correlation would 
have to be performed, and thus the first adjustment factor calculated. 
EPA is today proposing that a manufacturer may propose a field/bench 
adjustment program test plan up to 48 months prior to certification for 
Phase 2, and if EPA did not reject the proposed test plan within 90 
days of submission of a complete test plan, the proposed test plan 
would automatically be accepted. EPA is also proposing that, at least 
90 days before beginning bench aging for certification or in-use 
testing purposes, the manufacturer would provide a report to EPA for 
approval describing the aging and testing conducted for the field/bench 
adjustment program. This timing would ensure that adjustment factors 
have been established in time for demonstrating compliance with Phase 2 
standards. EPA is also proposing that the initial field/bench 
adjustment program be performed on engines representative of Phase 2 
engines.
    b. Periodic Rechecks. The ANPRM contemplates that both the handheld 
and the nonhandheld correlation programs would require the correlation 
to be periodically rechecked, although only for the nonhandheld engines 
was a specific recheck schedule provided (every other year for the 
first five years of the program and every five years thereafter, e.g., 
2001, 2003, 2005, 2010, 2015, etc.). In today's notice, EPA is 
proposing that the recheck period be the same for both handheld and 
nonhandheld engines. However, EPA suspects that the recheck period 
described in the ANPRM's nonhandheld program may be more comprehensive 
than is necessary. Specifically, it may be the case that the field/
bench adjustment factor will not need to be checked so often, 
especially if technologies, production tolerances, and emission results 
do not change that much from year to year. As a result, EPA is 
proposing that the field/bench adjustment factor be re-estimated as 
often as every five years as determined by EPA on a case-by-case basis, 
except that EPA may require more frequent rechecks in model years prior 
to the 2006 model year. EPA seeks comment on this proposed recheck 
schedule. EPA also proposes that any new adjustment factor subsequent 
to a recheck be applied regardless of how similar it is to the 
adjustment factor from the previous correlation effort. However, the 
new adjustment factor would apply only prospectively, beginning with 
the next model year. EPA seeks comment on whether a longer lead time 
should be specified, for example, requiring the new adjustment factor 
to be applied with the engine model being certified at least six months 
after the new adjustment factor is determined. This would allow more 
time for engine manufacturers to adjust their designs, if necessary. 
Finally, EPA is not proposing any restrictions on the direction of 
modification of the field/bench adjustment factor that may results from 
future rechecks: it could be revised up or down, but not below 1.0.
    c. Hours to Age. EPA is proposing that all bench-aged engines be 
aged to their full regulatory lives. Field-aged nonhandheld engines and 
field-aged residential handheld engines would also be aged to their 
full regulatory lives. However, following the program described in the 
ANPRM, under the proposed program field-aged commercial handheld 
engines could be field-aged to a minimum of 75 percent of their full 
regulatory lives. This flexibility is proposed today to reflect 
concerns that it may be hard to age these engines in the field due to 
equipment problems not related to emissions and engine durability which 
might be experienced at the end of the useful life. At the same time, 
as described below, field aging need not be done by actual end users 
but, instead, could be done by the manufacturer using a test plan that 
mimics as closely as possible actual field use. Under these conditions, 
the equipment may be less likely to break. Field aging to a minimum of 
75 percent of regulatory useful life is being proposed as a cost 
savings measure for commercial engines which have the longest 
regulatory useful lives. Furthermore, EPA believes that test results on 
commercial engines aged to at least 75 percent of their regulatory 
useful lives can be appropriately extrapolated to the full regulatory 
useful life of the engine due to the generally more durable design of 
commercial engines which would tend to result in more predictable 
emission determination performance. Therefore, EPA seeks comment on the 
costs and benefits associated with field aging handheld commercial 
engines to their full regulatory lives. Finally, EPA is proposing that 
all engines in the same sample (bench or field) be aged to the same 
number of hours.
    d. Test Plan. EPA is proposing that the manufacturer develop a test 
plan for both field and bench aging. All such test plans would be 
required to use the federal test procedure. The handheld program 
described in the ANPRM specified that ``a portion of the field engines 
will be aged in individual usage or fleets where the manufacturer does 
not carry out or exercise control over the engines' maintenance, or 
limit their usage such that engines are no longer used in a way that is 
representative of typical in-use conditions.'' Manufacturers would have 
three ways to field-age engines: in individual usage, in an independent 
fleet, or in a fleet that may be controlled by the manufacturer but 
over which the manufacturer does not control the maintenance process or 
inappropriately limit use. EPA proposes to extend this choice to both 
handheld and nonhandheld engines. However, EPA proposes that, if the 
manufacturer chooses to field-age the engines in a non-independent 
fleet, the applicable test plan must explain how the engines will be 
used to approximate, as closely as possible, actual in-use conditions, 
and also the kind of maintenance program to be followed, which should 
approximate expected in-use maintenance by end-users. The key is to 
ensure that the engines will experience similar load demands and 
environmental factors. For example, in the case of lawn mowers, the 
test plan for a non-independent fleet would have to specify how the 
engine would be exercised in a way to be representative of typical in-
use conditions, which likely include cutting both high and low grass, 
under wet and dry conditions, etc. Alternatively, if the manufacturer 
chooses to age the engines in an independent fleet, the test plan would 
have to detail how the use of the engine will be documented and how the 
user will ensure that it is used in a variety of different conditions. 
Finally, EPA could review this test plan and could require changes if 
the plan does not adequately approximate in-use conditions.
    e. Technology Subgroups. For both individual-manufacturer and 
industry-wide programs (see f., below), the analysis could be done on 
engine technology subgroups which could be expected to have similar 
emission deterioration characteristics, that is, groups of engine 
families from one or more manufacturers having similar size, 
application, useful life and emission control equipment. It would not 
be appropriate for engines with significant differences in in-use 
emissions performance characteristics to be included in the same 
technology subgroup. Manufacturers would be required to provide a 
justification satisfactory to EPA that the engines families would be 
expected to have similar emission deterioration characteristics, and 
would thus be

[[Page 3980]]

appropriately grouped in the same technology subgroup.
    f. Individual-Manufacturer or Industry-Wide Estimation. EPA is 
proposing that the above-described field/bench adjustment program and 
estimation of the field/bench adjustment factor can be performed on 
either an individual-manufacturer basis or on an industry-wide basis. 
Any manufacturer who wants to use a field/bench adjustment factor 
instead of field aging engines would have to either conduct its own 
program, or participate in an industry-wide program. In other words, 
the engines that will benefit from the application of an adjustment 
factor would have to be included in the sample used to estimate that 
adjustment factor. This requirement would ensure that a manufacturer 
could not simply apply a field/bench adjustment factor estimated by 
another manufacturer that may not reflect the performance of the 
engines to which it is applied.
    An industry-wide analysis would be subject to several additional 
constraints. First, EPA is proposing that all manufacturers 
participating in the same sample use the same test plan, except that 
maintenance schedules could vary across manufacturers to reflect 
differences in manufacturer-specified maintenance guidance to end-
users. This is to reflect the fact that although manufacturers may pool 
their emissions results in the industry-wide program, they may want to 
test their engines separately. This uniformity is important to avoid 
biased aggregation of results. Second, the sample of engines used to 
estimate the field/bench adjustment factor would have to include at 
least one bench engine and one field engine from the same engine family 
from each participating manufacturer, but no fewer than three bench-
aged engines three field-aged engines per technology subgroup. EPA 
seeks comment on whether the emissions should be sales weighted, to 
give a better picture of emissions across the category. EPA requests 
comment on how such a sales weighting procedure could be accomplished 
and still protect the confidentiality of sales information that might 
be covered by the confidential business information provisions of 90 
CFR part 2. Third, EPA proposes to limit entries into and exits from 
the industry-wide program: a manufacturer could enter or drop out only 
before the adjustment factor goes into use for the first time. This 
will prevent constant revision of the adjustment factor. If a 
manufacturer drops out of the industry-wide adjustment program, the 
field/bench adjustment factor would have to be recalculated, both for 
that manufacturer and the industry. This is necessary to ensure that 
the field/bench adjustment factor reflects only the experience of the 
engines to which it will be applied. Presumably, a manufacturer will 
drop out only if its individual adjustment factor is more favorable 
than the industry-wide adjustment factor. Thus, if the industry-wide 
adjustment factor is not recalculated, then it will understate the 
experiences of the engines to which it will be applied. EPA seeks 
comment on whether such restrictions are necessary.
    g. Restriction on Using Test Results for Other Purposes. One 
comment on the ANPRM requested that engine manufacturers be allowed to 
combine certification, correlation, and in-use testing for a family, 
such that bench results from the bench aged engines from the field/
bench adjustment program can be used to satisfy in-use testing 
requirements. EPA proposes to allow test results from engines used for 
the field/bench adjustment program to be considered for purposes of 
determining handheld deterioration factors based on good engineering 
judgment. EPA believes this is appropriate because in the handheld 
certification program compliance is determined by applying a 
deterioration factor to new engines. Thus, the actual engines that are 
used for certification are not the field-aged engines. However, the 
test results from the field/bench adjustment program would not be 
acceptable to satisfy the in-use testing requirements for handheld 
engines, since this would create a situation in which engines that were 
used to estimate a parameter for the compliance program are also used 
to demonstrate compliance. Similarly, EPA would not allow the test 
results from the field/bench adjustment program to be used for 
demonstrating certification for the nonhandheld program. The 
nonhandheld engine compliance program relies on emission results from 
engines aged to their full regulatory lives. As in the handheld engine 
in-use testing example above, if the engines used in the field/bench 
adjustment program were also allowed to be used to demonstrate 
compliance, this would create a situation in which engines that were 
used to estimate a parameter for the compliance program are also used 
to demonstrate compliance. Finally, EPA proposes to prohibit emission 
results from engines tested to determine compliance with other parts of 
today's program from being used for purposes of calculating the field/
bench adjustment factor. This restriction is necessary because 
otherwise manufacturers could choose among all of their test results 
and submit only the best emission results from a fairly large pool of 
engines, thus biasing the field/bench adjustment calculation. EPA does 
not believe this restriction will be burdensome, since manufacturers 
will be able to estimate a field/bench adjustment factor with as few as 
two engines (one bench-aged, one field-aged) if they participate in an 
industry-wide program, or six engines (three bench-aged and three 
field-aged) if they decide to establish their own adjustment factor.
    h. Other Pollutants. The handheld program described in the ANPRM 
contemplated that all pollutants be measured. EPA is proposing that CO 
emissions be measured and adjustment factors for CO be determined for 
both the nonhandheld and handheld programs. However, EPA believes that 
the data set upon which statistical tests used to establish appropriate 
adjustment factors for HC+NOX are determined are sufficient 
to establish the relationship between CO emissions in the field and on 
the bench. Therefore, EPA proposes to allow manufacturers to use the 
same set of data to calculate a CO adjustment factor as would be used 
to establish the HC+NOX field/bench adjustment factor. EPA 
requests comment on this proposal.
4. Alternative Methodology Considered
    EPA believes that the methodology described above is most 
appropriate because it balances the desires of industry for a simple 
program with the desire of EPA to put reasonable statistical 
constraints on the program without making it too difficult to perform 
or apply. However, there are other methods that can be used. Notably, 
EPA considered a statistical methodology in which a confidence interval 
would be constructed around the ratio of the means, and the adjustment 
factor would be the upper bound of that confidence 
interval.40
---------------------------------------------------------------------------

    \40\ See ``Simulation to Determine Confidence Level and Maximum 
Allowable Interval Width for Field/Bench Adjustment Factor 
Program,'' EPA Air Docket A-93-29, Item #II-B-01. For a description 
of this alternative approach, see ``A Procedure for Adjustment of 
Emissions Results for Bench Aged Small Engines,'' located in EPA Air 
Docket A-96-55, Item #II-D-40.
---------------------------------------------------------------------------

    While both techniques attempt to apply statistical concepts, this 
alternative methodology could be considered in some ways more 
statistically sound than the one proposed above. However, it may be 
practically more difficult to use. Most importantly, the adjustment 
factor derived from this alternative methodology would be sensitive to 
the number of engines tested: a larger

[[Page 3981]]

number of engines will most often result in a smaller adjustment 
factor, although this need not always be the case. Thus, manufacturers 
will be faced with either testing a large number of engines to ensure 
the smallest adjustment factor (closest to the straight ratio of the 
sample means) or using a larger adjustment factor with concomitant 
effects on the adjusted emission rate. EPA is concerned that this 
dynamic could lead manufacturers to test a large number of both bench-
aged and field-aged engines. In addition, the adjustment factor derived 
from this alternative methodology will always be a conservative 
estimate of the relationship between bench and field-aged results, 
because it is the upper bound of the confidence interval, and it will 
always be greater than the simple ratio of the means. Yet, it is not 
clear why choosing a conservative adjustment factor is preferable to a 
simple ratio of the sample means. Nevertheless, EPA seeks comment on 
the use of this methodology and other alternative approaches as opposed 
to the proposed methodology.

D. Compliance Program

    This section discusses the three step compliance program proposed 
today for the Phase 2 regulation of small SI engines, consisting of 
certification, production line testing, and in-use emission testing. As 
discussed above in Section III, today's proposal contains three basic 
elements new to the Phase 2 program. First, manufacturers would be 
required at the time of certification to account for emissions 
deterioration throughout the useful life of the engines. Second, EPA is 
today proposing a manufacturer-run production line testing program to 
replace the existing Selective Enforcement Audit (SEA) program as the 
primary method of determining the compliance of new production engines. 
Finally, EPA is proposing in-use emission testing programs for 
nonhandheld and handheld engines. EPA is also proposing appropriate 
remedies to address noncompliance with emission standards. Such 
remedies include mandatory recall but would also consider alternatives 
to mandatory recall, in the event of nonconformities found through 
production line testing or in-use testing programs. The basic proposed 
program for nonhandheld and handheld engine compliance is described in 
this section; Section IV.E outlines certain compliance flexibilities 
which may be made available to certain manufacturers depending on a 
manufacturer's size, the class of engines, or other factors.
1. Certification
    The certification process as required in the Act is an annual 
process. The Act prohibits the sale, importation or introduction into 
commerce of regulated engines when not covered by a certificate. The 
certification process proposed in this notice differs from that 
required in Phase 1 in that it would require the manufacturer to 
demonstrate that the engines will meet standards throughout their 
useful lives. To account for emission deterioration over time, 
manufacturers would be required to either age engines out to their full 
useful lives to obtain certification, or to adjust their certification 
test results by assigned or calculated deterioration factors (dfs), as 
is currently done under other EPA mobile source rules. Where 
appropriate and with suitable justification, dfs would be allowed to be 
carried over from one model year to another and from one engine family 
to another. This section describes nonhandheld and handheld engine 
certification provisions, provisions for certification to CO standards, 
and EPA efforts to streamline the certification process.
    a. Nonhandheld Certification. This notice proposes that 
certification for Class I and Class II nonhandheld engines continue as 
in Phase 1 except for the inclusion of an estimation of in-use 
deterioration. This deterioration estimate would be used to predict 
full useful life emission performance which would then be the basis for 
certification compliance decisions. The method for estimating in-use 
deterioration for certification purposes would depend on the type of 
engine technology.
    i. Side-Valve Engines and Engines with Aftertreatment. For all 
side-valve engines and engines with aftertreatment, this notice 
proposes that one engine from each engine family would either be field 
aged in a representative application to its full useful life, or bench 
aged to its full useful life to demonstrate compliance with the 
standards.41 If a manufacturer chose the bench aging option, 
it would be required to use a bench cycle approved in advance by the 
Administrator, adjusting the results using the field/bench adjustment 
factor established through the process described above at Section IV.C. 
In either case, the manufacturer would be required to run the full test 
procedure described in this rule when the engine is stabilized, 
accumulate hours on the engine, and then run a full test procedure at 
full useful life hours to determine a test value for certification.
---------------------------------------------------------------------------

    \41\ For nonhandheld engines participating in the averaging, 
banking and trading program described in more detail above in 
Section IV.A.5, compliance would be demonstrated with the family 
emission limit (FEL) rather than the standard.
---------------------------------------------------------------------------

    The final field-aged results or the final adjusted results of the 
fully bench-aged engines would be compared against the applicable 
standard to determine compliance at the time of certification. In 
addition, a df would be calculated from the final test results compared 
against low hour stabilized test results. While not directly used in 
the certification program, this df would be used to adjust the results 
of engines tested in Production Line Testing program described below in 
Section IV.D.2.
    For Class II SV engines and Class II engines with aftertreatment 
certified to the 250 hour useful life category, the manufacturer would 
have the option to bench age the engine to less than the full useful 
life and calculate a df at the engine's full useful life using a method 
of data extrapolation acceptable to the Administrator, as described 
below in Section IV.E.
    ii. Overhead Valve Engines. As discussed elsewhere in this notice, 
EPA expects the Phase 2 rule to result in a virtually complete 
technological shift for Class II nonhandheld engines from SV to OHV or 
comparably clean and durable technology engines. In addition, EPA 
believes that OHV technology engines have the potential to show low and 
stable emissions deterioration characteristics as compared with SV 
technology engines.
    EPA is today proposing that manufacturers of OHV technology engines 
be allowed to use an industry-wide assigned df for certification 
purposes. This program should allow manufacturers to focus more of 
their efforts on transitioning to a cleaner technology, by reducing the 
certification test burden on the engine manufacturers at the beginning 
of the Phase 2 program. EPA believes that offering manufacturers the 
opportunity to use an industry-wide assigned df rather than calculated 
dfs is reasonable for OHVs. A key element of the proposal for an 
assigned df is the proposed requirement that all manufacturers of OHV 
technology engines would participate in an industry-wide OHV Field 
Durability and In-use Performance Demonstration Program (``Field 
Durability Program'') described in Section IV.D.3, below. This program 
would be designed to demonstrate the validity of the assigned df by 
producing significant amounts of data from real field-aged engines. If 
the OHV Field Durability Program data indicate that the assigned df is 
inappropriate, EPA would conduct a

[[Page 3982]]

rulemaking to modify these proposed provisions to correct the assigned 
df program. This section describes the assigned df program for OHV 
engines, as well as an option for manufacturers to calculate dfs 
through field testing engines at the time of certification.

Assigned dfs For OHV Nonhandheld Engines

    EPA is proposing that manufacturers of OHV technology engines would 
be allowed to use a multiplicative assigned df of 1.3 for OHV engines 
in all useful life categories for projecting emissions deterioration 
for compliance purposes. In the ANPRM, EPA discussed a value of 1.3 as 
the assigned df value for Class I and Class II OHV technology engines 
in the shortest useful life categories (i.e., 66 and 250 hours, 
respectively). In addition, EPA indicated that it would consider during 
the rulemaking process whether or not to propose an assigned df for all 
useful life categories, and if so, what the appropriate assigned df 
values would be. EPA indicated that the assigned df for Class II OHVs 
in the 500 and 1000 hour useful life categories would likely fall 
between 1.3 and 1.5. In addition, if an assigned df of 1.5 at 1000 
hours, for example, appeared to be the appropriate value, EPA would 
propose a standard for the 1000 hour category adjusted by ratio to the 
proposed 12.1 g/kW-hr standard proposed for the 250 hour category.
    EPA received comment on the ANPRM that the assigned df should be 
higher than 1.3 for the higher useful life categories, with a 
corresponding higher emission standard for the higher useful life 
categories. This commenter suggested that the application of a 1.3 df 
to longer useful life periods could reduce product offerings and impose 
unjustified costs on small equipment manufacturers. EPA received a 
similar recommendation for higher dfs for the 500 and 1000-hour useful 
life categories.42 Specifically, an assigned df of 1.4 and a 
HC+NOX compliance standard of 13.0 g/kW-hr were recommended 
for 500-hour engines and an assigned df of 1.5 and a HC+NOX 
compliance standard of 14.0 were recommended for 1000-hour engines. In 
making these recommendations, the represented manufacturers argued that 
EPA had no full life emission performance information for these 
categories of engines. Although acknowledging they were providing no 
data to substantiate their recommendation, these manufacturers believe 
these higher dfs and emission standards provide a better assessment of 
equivalent stringency for these categories of engines compared to 250-
hour engines certified with a 1.3 df to a 12.1 g/kW-hr standard.
---------------------------------------------------------------------------

    \42\ See Memo to the Docket regarding the October 3, 1997 
meeting between U.S. EPA and the Engine Manufacturers Association, 
EPA Air Docket A-96-55, Item #II-E-11.
---------------------------------------------------------------------------

    EPA also received comment that use of assigned dfs should be 
limited to small volume manufacturers as a cost savings measure, and 
that the use of experimentally-derived dfs is preferable to the use of 
assigned dfs. This commenter argues that if the assigned df level is 
set too high, it could penalize those manufacturers who develop 
extremely durable engines, but if an assigned df were set too low, the 
result could be an underestimation of the emissions impact associated 
with an engine family or even the entire category. A final commenter 
asserted that assigned dfs are a bad idea; that the program described 
in the ANPRM results in a program in which future standards are 
uncertain due to the possibility of another rulemaking to adjust dfs; 
and that in the interval, engines may exceed the in-use standards 
because there is little incentive for manufacturers to reduce the 
deterioration rates of their engines.
    EPA believes an industry-wide assigned df combined with the OHV 
Field Durability Program to validate assumptions as to the durability 
of OHV technology engines is a sound program. The Agency fully expects 
the assigned df to accurately reflect the industry-wide average df of 
OHV engines certified to the proposed standards at least in the near 
term. As manufacturers gain improved capabilities to produce OHV 
engines (as would be expected as an increasing proportion of small 
engines become OHVs), the industry-wide df could shift to a lower 
value. There is no expectation, however, for a shift to a higher 
average df. The OHV Field Durability Program is expected to yield 
significant quantities of in-use data designed to verify the 
assumptions as to the emissions durability characteristics of OHV 
technology engines underlying today's proposal. The future standards 
are not uncertain if the industry average assigned dfs prove to be low 
and stable, as anticipated by this proposed rule.
    EPA is today proposing a 1.3 assigned df for all useful life 
categories for Class I and Class II engines, based on EPA analysis of 
available test data on engines aged in the field, provided by engine 
manufacturers.43 While the data are limited, the data on 
Class II engines designed for longer useful life periods do not point 
to any value other than 1.3 for an assigned df for longer useful life 
hours. While no data were available on Class I engines designed for 
longer useful lives, EPA believes that a 1.3 assigned df at longer 
useful lives is a reasonable value. Longer useful life engines are 
designed for enhanced durability, and this is reflected in the 
emissions deterioration of the engines as well, with longer useful life 
engines experiencing the same emissions deterioration at longer hours 
as do short useful engines at short hours. Additional information on 
the derivation of the proposed assigned df of 1.3 is contained in the 
docket to this rulemaking.44 Commenters who suggested a 
value other than 1.3 for assigned dfs at longer useful life hours did 
not supply data in support of their recommendations. However, EPA 
recognizes that the data upon which this proposal is based are very 
limited. EPA requests additional data on which to base the analysis for 
determining values for assigned dfs for OHV engines at longer useful 
lives. In particular, EPA requests comment on and any data supporting 
the assigned df and level of standards recommended by engine 
manufacturers (that is, 1.4 df and 13.0 g/kW-hr for 500-hour engines, 
and 1.5 df and 14.0 g/kW-hr for 1000-hr engines).
---------------------------------------------------------------------------

    \43\ See ``Tier 1 Deterioration Factors for Small Nonroad 
Engines'', September 1996, a report by Air Improvement Resources, 
available in EPA Air Docket A-96-55, Item #II-D-11.
    \44\ See ``Summary of EPA Analysis of Nonhandheld Engine HC and 
NOX Exhaust Emission Deterioration Data for 500 Hour 
Useful Life Class II OHV Engines,'' EPA Memorandum, August 4, 1997, 
available in EPA Air Docket A-96-55, Item #II-B-02.
---------------------------------------------------------------------------

    Finally, EPA is concerned that an industry-wide assigned df could 
reduce the incentive for a manufacturer to improve the durability of 
its engines. If manufacturers would be able to rely on an assigned df 
for certification performance regardless of in-use emission 
performance, manufacturers could design and produce engines which 
actually had much higher in-use deterioration than the assigned df. 
Manufacturers would be motivated to do so if they receive cost or other 
advantages from such a strategy. This is a real possibility since, in 
general, less expensive designs such as those with larger production 
tolerances or no oil control rings would also be expected to have 
higher emission deterioration. To protect against this, EPA is 
proposing limits on the use of assigned dfs. Specifically, EPA is 
proposing that if it determines the manufacturer's actual in-use sales 
weighted average df for a

[[Page 3983]]

useful life category (e.g., all OHV families certified to a 500-hour 
useful life) exceeds the assigned df by more than 15 percent (i.e., 
actual in-use df is 1.5 or greater), then EPA may require the 
manufacturer to generate engine family-specific dfs for one or more 
engine families in that useful life category. Similarly, if EPA 
determines that a family has an actual in-use df greater than 1.8, then 
EPA may require the manufacturer to generate an engine-specific df for 
that family. In either case, if EPA requires such engine-specific dfs, 
they would be determined on the basis of data from three field-aged 
engines per engine family. This level of testing is the same as that 
for the program being proposed for a manufacturer which opts to not use 
the assigned dfs for certification (see discussion in the following 
section, ``Calculated dfs for OHV Nonhandheld Engines''). EPA requests 
comment on the proposed thresholds for limits on the use of the 1.3 
assigned df.
    EPA recognizes that a requirement to generate an engine-family 
specific df for certification could be especially burdensome or perhaps 
practically impossible without disrupting production if the requirement 
was placed on the manufacturer close to the anticipated start of 
production for that family. EPA would take such issues into 
consideration when making any determination to require an engine-
family-specific df to be generated.
    EPA requests comment on all aspects of today's proposal for 
assigned dfs and calculated dfs for OHV technology engines, including 
the proposals for incentives for improving deterioration 
characteristics of OHV technology engines, and protections against 
misuse of the assigned dfs. EPA also requests additional data on which 
to determine the assigned dfs for OHV engines.

Calculated dfs for OHV Nonhandheld Engines

    EPA views assigned dfs for OHV technology engines as the program 
engine manufacturers would most often select due to lower costs for 
certification. However, it is desirable to allow manufacturers of 
engines having improved durability characteristics to demonstrate and 
take credit for these lower dfs. Therefore, EPA is proposing as an 
option a procedure whereby a manufacturer could generate its own dfs 
for all engine families within a useful life category, in lieu of 
applying the assigned df for those families.
    The assigned df is based on industry average data with some actual 
dfs above 1.3 and others below 1.3. EPA anticipates that manufacturers 
would choose the option of calculating their own dfs, over the option 
of selecting the 1.3 assigned df, in cases in which their engines 
exhibit superior deterioration characteristics. EPA is concerned that, 
if only these engines with superior deterioration characteristics are 
removed from the evaluation of the industry-wide assigned df values, 
then the industry average would be influenced upwards.
    Therefore, to partially mitigate this concern, EPA is proposing 
that if a manufacturer chooses to establish its own df for one engine 
family in a useful life category, then it would be required to do so 
for all of its engine families within that useful life category. Thus 
the manufacturer would determine specific dfs for all of its families 
in that useful life category. In considering the types of data that 
would be required for manufacturer-determined dfs, EPA balanced the 
need for the program to be reasonable and practicable, yet rigorous 
enough to provide confidence in the dfs.
    EPA is today proposing that calculated dfs for the full product 
line of OHV engines in a particular useful life category could be 
generated by field aging a minimum of three engines per engine family 
in a representative application to their regulatory useful lives. Each 
engine would be emission tested at least twice for all regulated 
pollutants using the full test procedure described in this rule. The 
first test point would occur after the engine had been stabilized by 
bench or field aging. The second test point would occur after the 
engine had been field aged to its useful life. The df for that engine 
family would be determined based on test data by dividing the average 
emissions at the full useful life by the average stabilized emissions 
for that family. If the manufacturer elects to conduct more than one 
test at either test point then the average of the data would be used. 
All test data would have to be at or below the standard (FEL, if 
applicable). EPA is also proposing that calculated dfs may cover 
families and model years in addition to the one upon which they were 
generated if the manufacturer submits a justification acceptable to EPA 
at the time of certification that the affected engine families can be 
reasonably expected to have similar emission deterioration 
characteristics.
    The Agency is proposing for manufacturers who choose to develop 
their own OHV dfs by field aging three engines per engine family that 
these engines must be actual field-aged engines and not bench-aged even 
if adjusted by a field/bench adjustment factor. The proposed assigned 
dfs with df verification through the OHV Field Durability Program is 
the primary program for Class I and II OHV engines. The Agency believes 
that any alternative to the primary program for nonhandheld OHV engines 
must generate emission data of similar accuracy as that on which the 
assigned df and OHV Field Durability Program is based. Without this 
requirement, the primary program would be undermined. The Agency has 
proposed a field/bench adjustment program for handheld engines and for 
non-OHV technology Class I and II engines. In both of those programs 
the Agency has proposed a level of confidence which would have to be 
met before a field/bench adjustment factor would be allowed, and is 
therefore a compromise between data accuracy and test burden (see 
Section IV.C). The test burden associated with the assigned df and OHV 
Field Durability Program has been limited to an appropriate level 
because it is covered by a maximum number of field aged engines that a 
manufacturer would be required to test on an annual basis (see Section 
IV.D.3.c ``Maximum Rates for Field Tested Nonhandheld Engines''). 
However, the proposed OHV Field Durability Demonstration does not 
permit a compromise on the accuracy of the field test data which would 
result from a field/bench adjustment program. Therefore, the Agency 
believes it is not appropriate that an alternative (i.e., manufacturer 
calculated dfs) to this primary program should allow such a compromise. 
The Engine Manufacturers Association 45 has recommended to 
the Agency that manufacturers be allowed to determine their own OHV dfs 
by performing a field/bench adjustment program. The Agency requests 
comment on this suggestion.
---------------------------------------------------------------------------

    \45\ See Memo to the Docket regarding the October 3, 1997 
meeting between U.S. EPA and the Engine Manufacturers Association, 
EPA Air Docket A-96-55, Item #II-E-11.
---------------------------------------------------------------------------

    In the ANPRM, EPA indicated that it would consider during the 
rulemaking process the appropriateness of reserving certification 
credits pending verification of the dfs through in-use testing for 
families for which the manufacturer generates its own df. EPA believes 
that today's proposal for field aging three engines per engine family 
for calculating dfs provides adequate data up front to provide 
assurance as to the deterioration of these engines, and obviates the 
need to reserve certification credits pending in-use testing. However, 
engines for which the manufacturer calculates its own df would be 
subject to the OHV Field Durability Program. EPA requests comment on 
the proposal not to reserve certification credits

[[Page 3984]]

pending verification of the dfs through in-use testing.
    Finally, to provide flexibility during the phase-in of the 12.1 g/
kW-hr Class II standard, EPA is proposing that manufacturers choosing 
to establish their own dfs for the 500 and 1000 hour useful life 
categories for Class II OHV engine families may, with the advance 
approval of the Administrator, base their dfs on good engineering 
judgement (subject to future verification, as discussed below in 
Section IV.E).
    b. Handheld Certification. This notice proposes that the 
certification of handheld engines continue as in Phase 1, except that 
manufacturers would be required to generate and apply a df to their 
stabilized emission results. EPA is proposing that manufacturers would 
be allowed to establish a df for each engine family based on 
technically appropriate analysis of test data on that engine family (or 
engine families of sufficiently similar design to be expected to have 
the same emissions durability) to reflect the emission deterioration 
expected to occur over the useful life of the engine. Manufacturers 
would be required to retain test data and description of their analysis 
to support their choice of dfs and to furnish this information to EPA 
upon request. EPA may reject the manufacturer's choice of df if it has 
evidence that the actual df is significantly higher or if the test data 
and analysis do not support the manufacturer's determination of a df. 
Data in support of the df could include data from the field/bench 
adjustment factor program as well as data from the in-use testing 
program.
    EPA believes that the proposal to allow manufacturers flexibility 
in determining the test data necessary to establish dfs for handheld 
engine families is a reasonable program designed to assure the 
environmental benefits of the program are met without placing an undue 
burden on manufacturers at the time of certification. EPA requests 
comment on all aspects of the proposed provisions for certification of 
handheld engines and determination of emission deterioration factors 
for compliance purposes.
    c. Certification to CO Emissions Standards. EPA is proposing that 
provisions for establishing CO emission dfs for use in the 
certification and production line testing programs would be the same as 
the provisions for established HC+NOX (or 
NMHC+NOX) emission dfs, except in the case of OHV technology 
engines for which the manufacturer elected to use an assigned df. For 
these engines, the manufacturer would be allowed to establish a df for 
CO emissions using good engineering judgment.
    d. Streamlining of the Certification Process. Since the 
promulgation of the Phase 1 rule, EPA has taken great strides to reduce 
the volume of information that must be submitted to obtain 
certification. A direct final rule published on May 8, 1996 (61 FR 
20738), greatly reduced the reporting requirements necessary to obtain 
certification under the Phase 1 program. This proposal would continue 
the reduced reporting requirements, adding only information items 
related to new provisions required for the Phase 2 program.
    EPA has also made strides to facilitate the electronic submittal of 
certification materials. Certification applications can currently be 
submitted on a computer disk, and the Agency hopes soon to be able to 
receive applications through a telephone data link. Further, EPA is 
working with the California Air Resources Board (CARB) in an effort to 
develop a common application format that would reduce the certification 
burden for manufacturers. EPA anticipates that for the Phase 2 program, 
EPA and CARB would accept the same application format and would have 
the same application submittal process.
2. Production Line Testing
    This section addresses the production line testing program proposed 
today for nonhandheld and handheld engines. EPA is proposing that 
manufacturers conduct a manufacturer-run production line testing (PLT) 
program using the Cumulative Sum (CumSum) procedure, as the primary 
program for ensuring the emission performance of production 
engines.46 The Phase 1 rule relies upon a traditional 
Selective Enforcement Auditing (SEA) program for production line 
compliance. SEA is a statistical sampling and testing scheme that must 
be initiated by EPA and provides a snapshot indication of whether a 
given engine family complies with applicable standards or FELs at a 
given point in time.
---------------------------------------------------------------------------

    \46\ The CumSum procedure has been promulgated for marine 
engines in EPA's spark-ignition marine rule at 40 CFR Part 91 (61 FR 
52088, October 4, 1996). In this section, ``PLT'' refers to the 
manufacturer-run CumSum procedure, or other manufacturer-run 
production line testing procedure approved by EPA. ``PLT'' does not 
include Selective Enforcement Auditing (SEA), which is addressed 
separately in Section IV.D.2.d.
---------------------------------------------------------------------------

    In the proposed Phase 2 PLT program, manufacturers would conduct 
continuous production line testing of all engine families and feed the 
results of that testing back into their design and production 
processes. CumSum is a statistical sampling and testing procedure which 
results in random periodic sampling and testing of engines from each 
engine family. The proposed CumSum procedure is useful both as an 
assessment tool for EPA and a quality control tool for engine 
manufacturers. The CumSum procedure assures that all configurations are 
susceptible to testing proportional to their production, and provides 
for continuous testing throughout the model year (except in cases in 
which an engine family shows clear compliance with the standards, in 
which cases testing can halt early, in as few as two engines). The 
CumSum procedure also allows manufacturers to monitor their own 
production and to fit production line testing into their normal 
production quality control procedures. The procedure is capable of 
detecting significant changes in the average level of a process, while 
ignoring minor fluctuations that are simply acceptable variation in the 
process. In summary, EPA believes that the CumSum procedure provides an 
effective measure for meeting EPA's goal of assuring that production 
engines comply with the applicable standards or FEL before they leave 
the production facility.
    As testing of each engine family begins with a new model year, the 
CumSum process computes an action limit and a test statistic based on 
the deteriorated test results for each pollutant for each family. As 
new data are received, both the action limit and the test statistic are 
updated. The action limit and the test statistic are functions of the 
standard deviation of the sample. If the test results are clearly below 
the standard or FEL, and the standard deviation of the test result is 
appropriately low, the process will declare a halt to testing. With 
very low emitting engines, this can occur in as few as two tests. If 
test data are highly variable or the test results are very close to the 
standard or FEL, testing may proceed to as many as thirty tests per 
family (the proposed maximum test limit) spread equally throughout the 
model year. If the test statistic crosses the action limit for two 
sequential tests, then the process indicates a nonconformity and the 
manufacturer would be required to take corrective measures.
    EPA is proposing a manufacturer-run PLT program for both 
nonhandheld and handheld engines. However, for nonhandheld engines, 
while PLT is the preferred option, EPA also is proposing an alternative 
program under which manufacturers would have the option to elect to be 
subject to the traditional SEA program (rather than PLT), as described 
in Section IV.D.2.d, below. In addition,

[[Page 3985]]

EPA is proposing to retain SEA for ``backstop'' purposes when 
manufacturer-run PLT is being conducted for nonhandheld and handheld 
engines, as described below. Under the proposal, in some cases, some 
manufacturers or engine families may have the option not to conduct 
production line testing requirements, including manufacturers of very 
clean engine families, or manufacturers or families which qualify for 
small volume flexibilities, as described in Section IV.E. The following 
discussions outline the proposed CumSum procedure, reporting of PLT 
results, procedures in the event of PLT failures, the use of SEA, and 
other topics related to production line compliance testing.
    a. The CumSum Procedure. The proposed CumSum procedure is outlined 
in this section. At the start of each model year, manufacturers would 
begin to test each newly-certified engine family at a rate of one 
percent of production. After conducting two tests, a manufacturer would 
determine the required sample size for the rest of the model year 
according to the sample size equation.47 For carry-over 
engine families, to reduce testing burden, the manufacturer would 
determine the necessary sample size by conducting one test, then 
combining the test result with the last test result from the previous 
model year, and finally calculating the required sample size for the 
rest of the model year according the sample size equation. Tests would 
be required to be distributed evenly throughout the remainder of the 
model year. After each new test, the sample size would be recalculated 
with the updated sample mean, sample standard deviation, and 95 percent 
confidence coefficient.
---------------------------------------------------------------------------

    \47\ For more discussion of the sample size equation, see 
Proposed Procedure for Quality Audits of Marine and Small Engines: A 
Cumulative Sum Approach, Item #IV-B-03 in EPA Air Docket A-92-28.
---------------------------------------------------------------------------

    The manufacturer would be allowed to stop testing at any time 
throughout the model year if the sample mean for each pollutant is less 
than or equal to the applicable standard or FEL, and if the number of 
tests required of the manufacturer, as calculated by the sample size 
equation, is less than the number of tests conducted. However, if at 
any time throughout the model year the sample mean for any pollutant is 
greater than the applicable standard or FEL, and if the manufacturer 
has not reached a ``fail'' decision, the manufacturer would be required 
to continue testing that engine family at the appropriate sampling 
rate.
    The maximum required sample rate for an engine family, regardless 
of the result of the sample size equation, would be the lesser of three 
tests per month to a maximum of 30 per year, or one percent of 
projected annual production, distributed evenly throughout the model 
year. For example, if the sample size equation produces a value of 252 
tests for a family with annual production of 20,000 engines, a 
manufacturer could elect to test only three engines per month to a 
maximum of 30 per year, instead of either 21 per month (which would be 
required if 252 tests were distributed evenly throughout the model 
year), or 17 per month (which would be required if one percent of 
annual production were distributed evenly throughout the model year).
    Although the sample size equation may calculate sample sizes 
greater than the proposed maximum sample rates, EPA believes that above 
some sample size, the cost of testing would become unnecessarily 
burdensome for manufacturers of small SI engines. Further, EPA believes 
that the proposed maximum sample rates (e.g., 30 engines) are 
sufficiently large to adequately characterize the emission levels of 
the engine family for the purpose of making a compliance decision. 
After determining the appropriate sample size, the manufacturer would 
construct a CumSum equation for each regulated pollutant for each 
engine family. Following each emission test, manufacturers would update 
current CumSum statistics for each pollutant according to the CumSum 
equation. Manufacturers would continue to update the CumSum statistics 
throughout the model year.48
---------------------------------------------------------------------------

    \48\ For more discussion of maximum sample rates and updating 
CumSum statistics, see Proposed Procedure for Quality Audits of 
Marine and Small Engines: A Cumulative Sum Approach, Item #IV-B-03, 
in EPA Air Docket A-92-28.
---------------------------------------------------------------------------

    Manufacturers could elect to test additional engines provided that 
testing of the additional engines is performed in accordance with the 
applicable federal testing procedures for small SI engines. Such 
testing could be used, for example, to bracket a nonconformity 
determined through the CumSum procedure, and such bracketing could be 
used to reduce a manufacturer's liability for past production. If a 
manufacturer elects to perform additional testing, the results would 
not be included in the CumSum equation. However, the results of 
additional tests would be included in the quarterly reports to EPA. 
Manufacturers would be required to randomly select which engines are to 
be included in the CumSum program prior to any knowledge of the 
emission levels of CumSum engines or engines used for additional 
testing.
    In cases where the CumSum sample size equation indicates that 
testing can be halted, the CumSum process indicates that there is 95 
percent probability for each pollutant that the mean emission level for 
the engine family is below the applicable standard (or FEL). In cases 
where the test statistic exceeds the action limit for two consecutive 
tests, then EPA is highly confident, based on extensive computer 
simulations of the CumSum program, that the mean emission level of the 
engine family for that pollutant exceeds the standard (or FEL), i.e., 
that the engine family is in noncompliance for that pollutant. The risk 
that a complying engine family will incorrectly be determined to be 
noncomplying (manufacturer risk) is set at similar levels as in EPA's 
historical SEA program. The risk that a noncomplying engine family will 
incorrectly be determined to be in compliance (consumer risk) is set at 
improved (lower) levels as in EPA's SEA program. The Agency requests 
comment on all aspects of the proposed production line testing program 
and CumSum procedure. For more information on the derivation of the 
sample size and CumSum equations and some examples of the CumSum 
procedure, see the document ``Proposed Procedure for Quality Audits of 
Marine and Small Engines: A Cumulative Sum Approach'' (EPA Air Docket 
A-92-28, Item # IV-B-03).
    b. Reporting of CumSum Results. EPA proposes that production line 
emission test results, as well as sample size calculations and CumSum 
calculations, would be reported to EPA on a quarterly basis. The Agency 
would then review the test data, sample size and CumSum calculations to 
assess the validity and representativeness of each manufacturer's 
production line testing program. If the CumSum process determines that 
an engine family is in noncompliance, the manufacturer would be 
required to report the emission test results and the appropriate sample 
size and CumSum equation calculations within two working days of the 
occurrence of the noncompliance.
    EPA received comments on the ANPRM recommending that, in the event 
of a PLT failure, manufacturers should be required to report such 
exceedances within thirty days of discovering the failure, suggesting 
that thirty days provides a reasonable time for manufacturers to 
evaluate and verify test data and determine the existence of any 
production line problems. EPA

[[Page 3986]]

believes that thirty days is too long a period for the Agency to not be 
made aware of a PLT failure. Such delays would not occur, for example, 
under a traditional SEA program. In the event of a traditional SEA, EPA 
is aware immediately of the existence of an SEA failure, and can 
immediately begin working with the manufacturer to remedy the problem. 
EPA is proposing that the appropriate PLT test results be reported 
within a two working days, a time period consistent with that 
promulgated for the gasoline marine PLT program. A two-day delay in 
reporting would not unnecessarily delay EPA's ability to begin to work 
with manufacturers during that time to determine an appropriate 
response to a PLT failure. As discussed below, the manufacturer would 
have 30 days after the date of the last test before any suspension or 
revocation of a certificate for the engine family would occur. The 
manufacturer could use that time to determine the existence of 
production line problems.
    EPA also received a comment that manufacturers should not be 
required to report all resultant test data to EPA quarterly (e.g., 
extensive raw test data in addition to calculated emissions results). 
This commenter suggests that the submission of a completed CumSum 
summary data sheet, permitting EPA to confirm that an engine family is 
in PLT compliance and to see where in the CumSum process compliance was 
attained, should be sufficient for quarterly reporting, and that 
manufacturers could maintain raw PLT data for a reasonable period of 
time and make such data available to EPA upon request.
    It is not clear which raw data this commenter would prefer be 
allowed to be retained at the manufacturer's facility. EPA is proposing 
that manufacturers would submit to EPA on a quarterly basis pertinent 
engine information, individual test results, relevant CumSum 
calculations, and other information at Section 90.709(e) of the 
proposed regulations. EPA does not believe that this reporting 
requirement is overly burdensome. EPA expects that manufacturers will 
keep track of PLT data electronically, and EPA intends to develop a 
standard CumSum summary data sheet to facilitate electronic submittal 
of data for the quarterly reports. EPA requests comments on these 
proposed provisions.
    c. Production Line Testing Failures. If an engine family is 
determined to be in noncompliance, or a manufacturer's submittal to EPA 
reveals that production line tests were not performed in accordance 
with applicable federal testing procedures, under the proposal EPA 
could suspend or revoke the manufacturer's certificate of conformity in 
whole or in part for that engine family subject to a thirty day waiting 
period (discussed in more detail below in Section IV.D.2.c.iv). EPA 
could reinstate a certificate of conformity subsequent to a suspension, 
or reissue one subsequent to a revocation, after the manufacturer 
demonstrates that improvements or modifications have brought the engine 
family into compliance. The proposed regulations include provisions for 
a hearing in which a manufacturer may challenge EPA's decision to 
suspend or revoke a certificate of conformity based on the CumSum 
procedure.
    EPA is proposing procedures whereby a manufacturer could remedy the 
emissions problems from engines produced prior to the PLT failure. In 
EPA's traditional SEA program, SEA failures have typically been 
addressed by a recall of the past production engines for the failing 
family. Future production engines are expected to be brought into 
compliance by either adjustments to the certification FEL, in cases 
where the manufacturer is participating in a certification ABT program, 
or through appropriate engine and emission control system 
modifications. As discussed in Section III of this preamble, above, EPA 
is proposing alternative remedies in the event of PLT failures, given 
the likely difficulties of applying a traditional recall program to the 
small SI engine industry. For handheld engines, these procedures 
include the use of in-use credits or other alternative remedies. For 
nonhandheld engines, these procedures include the use of certification 
credits through the adjustment of a family's FEL or other alternative 
remedies. These procedures are discussed below.
i. Handheld Engines
    EPA is proposing that when handheld manufacturers experience PLT 
failures, the excess emissions from engines that have already been 
introduced into commerce could be addressed by the application of in-
use credits or another alternative remedy. In-use credits are discussed 
in detail in Section IV.D.3, below. The emission performance of future 
production would be addressed through a running change to the existing 
configuration or certification of a new configuration such that 
compliance is demonstrated.
ii. Nonhandheld Engines
    Unlike the proposed program for handheld engines, the program 
proposed today for nonhandheld engines does not include provisions for 
in-use credit generation. Since in-use credits would not be available, 
and since recall of small SI engines is not likely to be effective, for 
nonhandheld engine manufacturers who use averaging, banking and trading 
to obtain certification, this notice proposes that, in the event of a 
CumSum failure, the manufacturer would be permitted to adjust its 
certification FEL to a level for which compliance could be 
demonstrated. This adjustment would apply to both past and future 
production of that family.
    EPA has held in past programs that manufacturers should be liable 
for their FELs, and that the past production of that family is subject 
to recall if the family exceeds its FEL during an SEA. The Agency 
continues to believe that manufacturers should set FELs appropriately 
based upon adequate testing and engineering analysis. Thus, while 
proposing that nonhandheld engine manufacturers would be permitted to 
adjust FELs for past production of an engine family, EPA expects that 
the need for manufacturers to change an engine family's FEL 
retroactively in the event of CumSum failures should be rare or 
nonexistent. If there are substantial occurrences of the need to adjust 
FELs retroactively, this would suggest that manufacturers are not 
correctly setting FELs carefully and accurately for individual 
families, in which case the Agency should appropriately revisit this 
provision.
    EPA is also proposing that nonhandheld manufacturers who experience 
CumSum failures could adjust their FELs even if they did not have 
adequate credits, provided that they could obtain the necessary credits 
by the end of the model year following the model year in which the 
production line failure occurs. If sufficient credits were still not 
obtained, the manufacturer would have two more years to obtain them, 
but would then be required to use credits on a 1.2 to 1 basis (i.e., 
such credits would be discounted twenty percent). Unlike in the 
proposed handheld engine in-use credit program, in which manufacturers 
would have opportunities to generate additional credits, the 
nonhandheld certification ABT program would not afford such 
opportunities. Thus, EPA believes it is reasonable in the program for 
nonhandheld small SI engines to provide additional time for 
manufacturers to acquire certification credits necessary to offset PLT 
exceedances. Requiring future model year credits to be discounted if 
used to remedy past production on

[[Page 3987]]

noncompliance assures that the manufacturer will not benefit 
economically from delayed compliance with the standards.
    Because EPA believes manufacturers should set FELs accurately and 
carefully, and to encourage manufacturers to set FELs accurately, EPA 
is proposing that these provisions (e.g., the retroactive use of 
credits, and the ability to carry a credit ``deficit'') would only 
apply in the case of a manufacturer who fails no more than one engine 
family in a given model year, or who fails more than one engine family 
but the total production of those families is no greater than 10 
percent of the manufacturer's U.S. sales. EPA requests comment on all 
aspects of this retroactive use of certification credits and its likely 
impact on the accuracy of the FELs determined at certification.
iii. Alternative Programs and Voluntary Recall
    In the event of PLT failures, EPA prefers that handheld 
manufacturers use in-use credits for past production engines and that 
nonhandheld engines be recertified to a higher FEL which may require 
the application of certification credits, rather than some other 
alternative to recall. However, EPA is proposing that in the case of 
handheld or nonhandheld engines where the manufacturer did not have and 
could not obtain adequate in-use or certification credits, as 
appropriate, a manufacturer could conduct a voluntary recall, if it 
could show that an appropriate response rate was likely. EPA would also 
consider the appropriateness of alternative projects. These projects 
are essentially alternatives to recall and would be designed to provide 
an environmental benefit as well as an economic incentive to the 
manufacturer to produce complying engines. Guidelines for such projects 
are discussed in more detail in Section IV.D.4, below. A mandatory 
recall could be ordered by EPA for past production engines pursuant to 
proposed Sec. 90.808 in cases where the manufacturer could not obtain 
appropriate credits and was unwilling to perform an alternative project 
acceptable to EPA.
iv. Suspensions and Revocations
    EPA is proposing for engine families that fail production line 
compliance testing, that EPA would have the authority to suspend or 
revoke the certificate for that family. However, no suspension or 
revocation for a family could occur before thirty days after the date 
of the last test. During the thirty day period, EPA intends to work 
diligently with the manufacturer, as it always has in the case of SEA 
failures, to provide certification of appropriate production line 
changes. Further, this notice proposes that EPA would approve or 
disapprove a manufacturer's production line change within fifteen days 
of receipt, or the change would be considered automatically approved.
    EPA believes that these waiting periods are reasonable to afford 
manufacturers and EPA sufficient time to work together to address 
problems, without the concern that EPA would hastily suspend or revoke 
the certificate of a family determined to be in nonconformity by a 
production line testing program. EPA believes that the proposed time 
frames are reasonable, and are consistent with longstanding EPA 
practices in the SEA program of providing a waiting period following an 
audit failure. In such failures, EPA works closely with the 
manufacturer to arrive at a solution for the problem engine family. 
With on-highway engines, such solutions have typically involved a 
recall of engines that have already been produced along with the 
recertification of the family to a new FEL, or the certification of a 
replacement engine configuration. As discussed above, for small SI 
engines, such solutions could involve the use of certification or in-
use credits, voluntary recalls, or other alternative remedies. EPA has 
never caused an assembly line to shut down because of an audit failure 
and does not intend to start such a practice where other alternatives 
can be used.
    d. Selective Enforcement Audits (SEA). While EPA is proposing the 
CumSum manufacturer-run PLT program as the preferred production line 
testing program for the Phase 2 program, EPA still sees a function for 
traditional SEA and is therefore not proposing to eliminate traditional 
SEA altogether. EPA is proposing that for both nonhandheld and handheld 
manufacturers, SEA would remain as a ``backstop'' for EPA to use in 
cases where there is evidence of improper testing procedures or 
nonconformities not being addressed by the CumSum process.
    As mentioned earlier, the Agency is also proposing an alternative 
program under which nonhandheld manufacturers could choose not to 
conduct manufacturer-run PLT program, in which case all families would 
continue to be subject to an SEA program as under Phase 1. Although 
currently not preferred by the Agency, EPA is considering this option 
since it was included in the ANPRM and received support from the 
nonhandheld industry. EPA solicits comment on the appropriateness of 
providing this option, and on whether it would be better to require PLT 
for all families. Only one approach, either PLT with SEA as a 
``backstop'', or manufacturers having the choice to use either PLT or 
SEA as the primary program, will be adopted as the final rule for 
nonhandheld manufacturers.
    Under this alternative program, EPA is also proposing that 
nonhandheld engine manufacturers be limited in their ability to switch 
back and forth between PLT and SEA. Manufacturers involved in PLT would 
be required to implement that approach for a minimum of three 
consecutive model years and to provide EPA with notice one complete 
model year prior to the model year for which they were planning to opt 
out. In addition, a manufacturer would not be allowed to opt out of PLT 
while carrying a negative certification credit balance. However, a 
manufacturer would be allowed to opt in to PLT at any time.
    Finally, where small volume engine manufacturers or small volume 
engine families would be entitled to exemptions from the PLT program 
under the proposal (see Section IV.E), those families would remain 
subject to SEA, although EPA would be unlikely to issue test orders 
without evidence of nonconformity.
    In the event of an SEA failure for handheld engine manufacturers, 
EPA is proposing that the option to use in-use credits or another 
alternative to recall would be available to remedy past production 
engines. For future production, the manufacturer would be expected to 
modify the engine to come into compliance with all applicable 
standards.
    In the event of an SEA failure for nonhandheld engine 
manufacturers, the manufacturer would have the option to adjust the FEL 
for future production of the engine family. EPA would address a remedy 
for the past production in the event of an SEA failure on a case-by-
base basis, seeking to both preserve the environmental benefits of the 
program, maintain incentives to accurately set FELs in advance, and 
minimize the burden on the industry. Such a remedy might include, for 
example, a combination of measures such as mandatory PLT for 
appropriate time periods and portions of production, recertification of 
all or part of an engine family, and generation of credits to remedy 
exceedances over an appropriate period of time. However, consistent 
with past practice, EPA does not anticipate allowing the retroactive 
use of certification credits to remedy past production failures 
determined via

[[Page 3988]]

SEA, or the carryover of any credit deficits, as would be allowed if 
the manufacturer chooses to conduct manufacturer-run PLT. Since SEA 
only evaluates production line performance during a ``snap shot'' in 
time and not throughout the entire production period, it would be 
inappropriate to use credits generated on the basis of total annual 
production to correct the SEA failure. Instead, a manufacturer would 
likely be expected to recall the noncomplying family or conduct an 
alternative remedy proposed by the manufacturer and accepted by EPA. 
EPA requests comments on the proposed provisions related to remedies 
for SEA failures.
    EPA received a comment on the ANPRM that handheld manufacturers 
should be permitted to elect to be subject to routine SEA testing, as 
they currently are under Phase 1 emissions regulations, rather than 
conducting manufacturer-run PLT. This commentor suggested that 
manufacturers may desire to elect SEA for reasons of cost, confidence 
in their quality control, or familiarity with SEA, and that such an 
option could enhance the flexibility and reduce the cost of the PLT 
process, while at the same time assuring new engine compliance with 
Phase 2 emissions regulations.
    EPA is not proposing routine SEA testing for handheld 
manufacturers. EPA believes that a manufacturer-run PLT program such as 
CumSum is a superior method of assuring that both handheld and 
nonhandheld production line engines meet the standards, that testing 
occurs continuously throughout the model year, and that each 
configuration is susceptible to testing. In addition, PLT affords 
benefits to the manufacturers of identifying problems early and 
addressing them without the disruption of an EPA-initiated SEA. EPA 
believes it is most useful and appropriate that manufacturers be 
responsible for and bear the burden of continuously monitoring their 
own emissions.
    Under the production line compliance program proposed today, EPA 
expects that nonhandheld manufacturers may in some cases choose SEA as 
their primary production line compliance program, for cost reasons or 
fear of the unknown. However, EPA believes that the downsides of the 
choice of SEA as the primary production line compliance program are 
potentially great for all involved. EPA believes that in choosing SEA, 
the manufacturers would be foregoing an effective quality control tool 
for monitoring their own production, and would risk expensive and 
disruptive SEAs. In addition, EPA would not get the same coverage of 
engine families in the testing process. The regulations proposed today 
reflect the option, consistent with the program outlined in the ANPRM, 
for nonhandheld manufacturers in some cases to choose either PLT or SEA 
as the primary production line compliance program. However, EPA is also 
proposing in the alternative that the nonhandheld production line 
compliance program would be the same as the handheld program. That is, 
the manufacturer would not have the option to choose SEA as the primary 
production line compliance method. Rather, manufacturer-run PLT would 
be the primary program in all cases, with SEA existing as a backstop. 
Again, EPA requests comment on the appropriateness of the proposed 
program which allows nonhandheld manufacturers the option to elect 
routine SEA testing in lieu of PLT testing. EPA also requests comment 
on the option that nonhandheld manufacturers would use only PLT as the 
primary production line compliance program, with SEA existing as a 
backstop, and the effectiveness of this option in providing assurance 
of environmental benefits in-use, easing the implementation burden for 
EPA and the industry, and achieving greater commonality in the 
compliance programs for the handheld and nonhandheld sides of the small 
SI engine industry.
    e. Annual Limits for SEA. The Phase 1 program contains annual 
limits on the number of SEAs the Agency may perform each year on a 
manufacturer, based on their number of engine families and sales. The 
Phase 1 annual limits serve to restrict the maximum number of audits 
for most manufacturers to a quantity equal to one fifth of the number 
of engine families (see 40 CFR 90.503(f)(1)). However, under the Phase 
1 program, any test which the family fails or for which testing is not 
completed does not count against the annual limit (see 40 CFR 
90.503(f)(3)). In addition, even if the annual limit is reached, EPA 
may initiate additional SEA testing to test families for which evidence 
exists indicating noncompliance (see 40 CFR 90.503(f)(4)).
    EPA is not proposing any changes to the Phase 1 SEA annual limit 
provisions for Phase 2 except for the additional proposed provision 
that EPA may initiate additional SEA testing beyond the annual limit 
for families or configurations which the Administrator has reason to 
believe are not being appropriately represented or tested in production 
line testing (see proposed Sec. 90.503(f)(4)).
    EPA also requests comment on an option, not proposed, to raise the 
annual limit by one or two families for each failing audit in a given 
model year in cases where manufacturers choose SEA as the primary 
production line compliance program, should the regulations allow SEA as 
the primary production line compliance program. While this option is 
not included in the proposed regulatory text, EPA requests comment on 
the potential benefits or costs of this option for a higher number of 
potential routine SEAs for manufacturers who experience SEA failures. 
EPA requests comment on all aspects of the proposal for annual limits 
for SEAs under the proposed Phase 2 program.
    f. Alternate Statistical Procedures for Production Line Testing. 
Consistent with the program outlined in the March 1997 ANPRM, EPA is 
proposing that manufacturers conducting manufacturer-run PLT could 
propose test schemes for EPA approval on a case-by-case basis other 
than the CumSum procedures described in this section and proposed in 
today's notice. EPA believes that this is reasonable because there may 
be situations where a single test scheme is not appropriate for a 
specific engine family or company. However, EPA also believes that it 
is desirable to avoid a multiplicity of testing schemes, and is 
concerned about the burden this could place on the Agency if multiple 
testing schemes are analyzed and developed with individual 
manufacturers. This notice proposes that EPA would have the right to 
review any alternate procedure to determine the ability of the 
procedure to (1) produce substantially the same levels of ``producer 
risk'' and ``consumer risk'' as the CumSum Procedure, i.e., the risk to 
a manufacturer that a complying family would fail in PLT testing, or 
the risk to the public that a failing family would pass in PLT testing; 
(2) to provide for continuous rather than point-in-time sampling; and 
(3) to include an appropriate decision mechanism for determining 
noncompliance upon which the Administrator can suspend or revoke the 
certificate of conformity. Further, it would be the requesting 
manufacturer's responsibility to provide an analysis and documentation 
that demonstrated the alternative satisfied these criteria. EPA would 
expect to reject any alternate statical procedure that did not fully 
satisfy these proposed criteria.
    g. Test Procedures for PLT. EPA believes that the best way to 
determine whether new engines meet certification

[[Page 3989]]

standards is to test them under the test used at certification. 
Therefore, EPA is proposing that the manufacturer-run PLT program 
proposed in this notice would require testing based on the full federal 
test procedure as used for certification and described in Subpart E of 
the attached regulations. EPA recognizes the potential need to permit 
minor adjustments to the test procedure to accommodate production line 
testing. Consistent with other compliance test programs for mobile 
sources, the proposed regulations allow the Administrator to approve 
such test procedure adjustments.
    h. Harmonization of Production Line Testing with CARB. EPA is 
interested in finding ways to harmonize the production line testing 
requirements proposed today for Phase 2 with any production line 
testing requirements manufacturers must meet for the California small 
engine regulatory program. In particular, EPA would expect that data 
from production line testing of a 50-state family conducted for a 
California Quality Audit program could be acceptable for the CumSum 
process, if the subject engines are sold nationwide and test engines 
are appropriately selected and tested. EPA will also continue to work 
with the California Air Resources Board to harmonize reporting formats, 
and similar information needs.
3. In-use Emission Testing
    EPA believes that a critical element in the success of its small SI 
engine program is ensuring that manufacturers build engines that 
continue to meet emission standards beyond certification and production 
stages and comply with standards for their full regulatory useful 
lives. Section 213(d) of the CAA specifically subjects nonroad engines 
to the in-use compliance provision of section 207.49 EPA has 
authority to subject manufacturers to in-use testing (conducted by the 
Agency or by the manufacturer under section 208 of the Act) and to 
remedy any noncompliance (for example, by recall and repair of engines) 
for the full regulatory useful life of an engine. In-use compliance 
enforcement has proven to be an effective incentive for manufacturers 
to build emission durable motor vehicles.
---------------------------------------------------------------------------

    \49\ Section 207(c) of the Act authorizes EPA to enforce 
compliance by vehicles and engines to applicable standards in actual 
use. Manufacturers are subject to recall ``[i]f the Administrator 
determines that a substantial number of any class or category of 
vehicles or engines, although properly maintained and used, do not 
conform to the regulations * * * when in actual use * * *''.
---------------------------------------------------------------------------

    However, as discussed above in Section III, in the case of small SI 
engines, EPA does not believe that a mandatory in-use compliance 
program which relies on recall, for example, is likely to be as 
effective and practical as it has proven to be in EPA's on-highway 
programs. Small SI engines differ from motor vehicles in that they are 
not registered and are therefore difficult to track so that their 
owners can be notified. Many are not easily transported to a servicing 
dealer for repair. The in-use programs described below are therefore 
designed to provide data on in-use performance and to provide 
incentives to manufacturers to produce emission-durable engines without 
relying on the use of recall. While the Production Line Testing 
programs described previously are very similar, the in-use programs 
proposed in this notice differ significantly for the two sides of the 
industry. Again, EPA requests comment on alternative in-use testing 
programs, such as applying the in-use testing program proposed for 
handheld engines to the nonhandheld side of the industry, as well as 
applying the field durability program proposed for OHV engines to side-
valve engines, engines with aftertreatment, and/or handheld engines.
    a. Nonhandheld Side-Valve Engines and Engines with Aftertreatment. 
For nonhandheld side-valve engines and engines with aftertreatment, the 
in-use program would consist of a certification program in which the 
engines would be aged to their full useful lives during the 
certification process and no certificates would be issued unless the 
engine family can first be shown to meet standards (or FELs) for its 
useful life, as described above in Section IV.C and Section IV.D.1. EPA 
believes that a program which does not rely on in-use testing after 
certification especially makes sense for Class II SV technology engines 
which are expected to be phased out by 2005. In addition, EPA would 
have data on SV technologies aged in the field for the field/bench 
adjustment factor program; if EPA suspected serious problems with 
regard to whether the emissions reductions anticipated by this rule 
were in fact being achieved, EPA would address these concerns through 
appropriate programmatic changes. EPA requests comment on the 
appropriateness of this full useful life certification to predict the 
in-use emissions durability of SV engines and engines with 
aftertreatment.
    b. Nonhandheld OHV Field Durability and In-use Performance 
Demonstration Program. For overhead valve nonhandheld engines, the 
proposed in-use program would be one whose primary function is to 
verify that the industry-wide deterioration factors predicted for the 
OHV engines are indeed correct. The proposed OHV field durability and 
in-use performance demonstration program (``Field Durability Program'') 
would generate significant quantities of emission data from engines 
aged in real field usage in representative pieces of equipment. If 
EPA's belief that the dfs of these engines are stable and predictable 
proves to be incorrect after receiving these data, or the assigned dfs 
specified in this rulemaking are significantly different than those 
that occur in real field usage of Phase 2 engines, then EPA would 
initiate appropriate programmatic changes through the regulatory 
process.
    The proposed Field Durability Program is designed to provide data 
on the deterioration of OHV engines in actual field usage. EPA is 
proposing that engines for the program would be selected from or placed 
into service with residential or professional users. This program would 
be designed to provide a representative picture of actual in-use 
emissions, including representative age, maintenance, and sales mix of 
engines in the field. To the extent practical, engines would be 
selected from residential customers or professional users, in order to 
most accurately reflect actual usage patterns such as number of cold 
starts, typical maintenance patterns, and overwintering. However, EPA 
would also allow engines to be selected from manufacturers' fleets, 
provided the engines and their operation and maintenance are typical of 
in-use engines. Each engine in the program would be baseline tested at 
a number of hours equal to the break-in hours used in certification. 
The engine would then be field aged in an appropriate piece of 
equipment to full useful life, at which time the engine would be 
removed and retested. The df would be determined mathematically from 
the two test points from each engine.
    Data from the OHV Field Durability and In-Use Emissions Performance 
Demonstration Program would not be designed to provide a basis for EPA 
to make in-use compliance determinations as to whether a particular 
engine family complies with its standard or FEL at the end of its 
useful life. Rather, the program is primarily designed to determine 
whether, in the aggregate, the industry-average assigned dfs for OHV 
engines are valid. Given the number of manufacturers expected to 
produce OHV engines and participate in this program, the program would 
generate meaningful volumes of real in-use data which would yield 
results indicating whether assigned dfs are realistic.

[[Page 3990]]

    This notice proposes that the OHV Field Demonstration Program 
testing could be spread over multiple years. EPA proposes that 
manufacturers provide a schedule to EPA each year of the engine 
families and approximate quantities of engines they intend to produce 
for U.S. sales over the coming four year period, as well as estimates 
of the number of field aged engines that would be tested each year for 
the field/bench adjustment program (see Section IV.C) and for 
calculating dfs for OHV engines at the time of certification (see 
Section IV.D.1). In addition, manufacturers may wish to recommend a 
proposed testing plan for the Field Durability Program that, for 
example, best fits testing into their marketing, production, test 
facility and budgetary constraints. EPA would consider such information 
in determining the engine families to be field tested over that time 
period as part of the OHV Field Durability Program.
    Manufacturers have indicated their desire to perform industry-wide 
OHV Field Durability Program testing to try to reduce the number of 
engines that must be field aged. EPA is proposing that it would 
consider requests by manufacturers to work together when it reviews a 
manufacturer's plan for engine families to be field aged. EPA will 
review proposals for joint testing to evaluate how thoroughly they 
cover a portion of overhead valve engine sales, whether they will 
provide statistically useful quantities of data, and other factors to 
help EPA ascertain whether OHV dfs from certification are accurate and 
appropriate.
    c. Maximum Rates for Field Tested Nonhandheld Engines. EPA believes 
that emission data from real field-aged engines would serve a crucial 
role in validating the use of assigned dfs, calculated dfs, and the 
aging cycles used for bench-aged certification of side-valve engines. 
While recognizing the importance of and need for these data, EPA is 
also sensitive to the cost and testing burden associated with directing 
large numbers of engines to be field aged and tested in a given year.
    In today's action, EPA is proposing that in any one year the Agency 
would not require field testing for the OHV Field Durability Program 
such that, when added to the field testing a manufacturer performs for 
the optional certification df generation or for the field/bench 
adjustment program, it would require the manufacturer to emission test 
more than 24 total engines that were field aged to their full useful 
life. EPA believes that this number will provide important quantities 
of data without placing an undue burden on manufacturers. EPA is 
proposing that it would have the right to require field testing to the 
maximum amount, and expects that the maximum testing may be required in 
the initial years of the program. Manufacturers would have the option 
to field test more engines than required by EPA. EPA anticipates it 
would reduce the testing burden as appropriate, especially for smaller 
manufacturers, in subsequent years should, for example, EPA determine 
that the data being developed is quite stable from year to year.
    The discussion of the Field Durability Program in the March 1997 
ANPRM indicated EPA would provide ``appropriate delays or waivers from 
the requirement of the bench correlation program in years when a 
manufacturer also runs the field durability program'' (see 62 FR 
14754). In the development of this proposal, EPA considered the need to 
propose procedures to provide for EPA granting delays or waivers from 
the requirements of the field/bench adjustment program in years when a 
manufacturer also runs the OHV Field Durability Program. In today's 
action, EPA is proposing no formal process by which manufacturers would 
request a waiver from the requirements of the field/bench adjustment 
program. EPA believes that the need for delays or waivers is obviated 
by the cap on the number of fully field aged engines EPA would be able 
to require to be tested in any one year.
    The discussion of the Field Durability Program outlined in the 
March 1997 ANPRM also suggested that EPA would propose an appropriate 
scaling of the field engine test burden for smaller volume 
manufacturers (see 62 FR 14754). For this proposal, EPA considered 
proposing a cap on the number of field tested engines of fewer than 24 
engines per year for smaller nonhandheld manufacturers by sales volume. 
However, EPA believes that a scaling back of the test burden would not 
be appropriate. Such a scaling would most appropriately be based on the 
inability of manufacturers to sustain the costs associated with the OHV 
Field Durability program; however, the ability to sustain the costs of 
the program would not appear to differ significantly among 
manufacturers. Therefore, EPA is proposing the same cap on the field 
engine test burden for all manufacturers. EPA believes that this 24 
engine per year cap is a manageable burden on the smaller volume 
manufacturers as well as the larger volume manufacturers. The Agency 
does not anticipate identifying families certified by manufacturers who 
would qualify as small volume engine manufacturers for in-use testing, 
unless there was evidence of a nonconformity (see discussion in Section 
IV.E). EPA requests comment on all aspects of the applicability of a 
cap to the number of field aged engines that EPA could require to be 
tested in any one year.
    d. In-Use Testing Program for Handheld Engines. In today's action, 
EPA is proposing an in-use testing program for handheld engines similar 
to that promulgated in the gasoline spark-ignition marine engine rule 
(see 40 CFR Part 91, Subpart I). As in the marine rule, EPA is also 
proposing an in-use credit program, as well as a number of criteria for 
evaluating other alternatives to mandatory recall. Mandatory recall is 
the primary remedy for noncompliance. However, as in the marine 
program, EPA is interested in considering options to mandatory recall 
and, if implemented, will monitor the use of these alternatives to make 
sure they are as effective as anticipated. EPA believes that the 
successful implementation of the in-use credits program and the other 
alternatives would provide a comprehensive remedy to address in-use 
emission noncompliance, as well as incentives to manufacturers to 
produce emission-durable engines, without the use of recall. The 
program for handheld engines proposed today differs from the gasoline 
marine engine program in that the engines may be bench-aged rather than 
field-aged, at the manufacturer's option, provided the manufacturer has 
previously established an adjustment factor between the bench aging 
cycle and field aging through the program described above at Section 
IV.C. EPA requests comment on the technical requirements which would 
allow bench-aged engines to represent the emission performance of 
field-aged products.
i. In-use Testing for Handheld Engines
    EPA is today proposing an in-use testing program for handheld 
engines which would make all engine families potentially subject to 
mandatory in-use testing by the manufacturer. The manufacturer would 
age the test engines in the field to their full useful lives. 
Alternatively, the manufacturer could choose to age the engines on a 
bench cycle to their full useful lives, providing that an adjustment 
factor had previously been established between the bench-aged and 
field-aged results, through the procedures described above in Section 
IV.C. The engines would then be emission tested for all regulated 
pollutants using the full test procedure described in this proposed 
rule. The number of engines per engine family tested would vary 
depending on test results. Except for small volume and carry-over 
engine families, the

[[Page 3991]]

minimum number of test engines would be four. For each engine that 
failed any pollutant, the manufacturer would test two additional 
engines, up to a maximum of ten. Small volume engine manufacturers 
would begin by testing two engines, adding two more for each failing 
engine up to the same maximum (see discussion of provisions for small 
volume engine manufacturers and other flexibilities in Section IV.E). 
Carry-over engine families would start with one engine. In the end, the 
emissions for each pollutant would be averaged and the family average 
compared against the appropriate standard to ascertain compliance. The 
in-use testing program proposed is designed as a method to provide 
adequate data on which to make compliance decisions, while allowing the 
testing of families which are found to emit below standard to conclude 
as expeditiously as possible.
    Manufacturers would provide a schedule to EPA each year of the 
engine families and approximate quantities of engines they intend to 
produce for U.S. sale over the coming four year period. EPA would then 
select engine families to be in-use tested by the manufacturer over 
that time period or a fraction of that time period. EPA would identify 
no more than 25 percent of a manufacturer's families for in-use testing 
in any one year.
    EPA received a comment on the ANPRM that it would be equally 
effective and potentially less costly to permit engine manufacturers to 
select the engine families for in-use testing. This would allow 
manufacturers to schedule in-use testing to better conform to 
production, marketing and budgetary constraints, and to choose their 
own mixture of commercial and residential engines to test each year. 
This commenter added that manufacturers could provide a testing 
schedule in advance to enable EPA to raise any concerns it has with a 
manufacturer's test plans.
    EPA believes it is important to retain the authority to select 
engine families for in-use testing that potentially show risk of higher 
emissions in-use than predicted at the time of certification. 
Therefore, EPA is proposing to retain the authority select the engine 
families for in-use testing. However, EPA would work with manufacturers 
in an attempt to schedule testing to take into account production, 
marketing, test facility and budgetary constraints and would invite 
manufacturers to recommend a testing program which best suits their 
needs.
ii. In-Use Credit Program for Handheld Engines
    As discussed above, the proposed in-use credit program for handheld 
engines is designed to address in-use nonconformities of handheld 
engines without the need for ordering manufacturers to conduct recalls 
of nonconforming engines. A reasonable means must exist to address in-
use noncompliance that provides incentives to manufacturers to build 
emission-durable engines, that can be implemented practically, that 
encourages additional in-use testing, that offsets additional emissions 
resulting from noncompliance, and that is not unduly burdensome. EPA 
believes that the successful implementation of the proposed in-use 
credit program described below could be part of a comprehensive remedy 
to address in-use noncompliance, and that EPA would not, in practice, 
order mandatory recall of Phase 2 engines. When a manufacturer 
determines its average in-use emission levels for each pollutant, it 
would compare those numbers against the applicable standards. Emission 
levels below the standards could generate in-use credits. Emission 
levels above the standard would require the use of in-use credits. The 
credit formula as proposed here would be a function of the sales of the 
engine family, the difference between the family emission average and 
the applicable standard, the power rating of the engine, load factor, 
and the useful life of the engine.
    In-use credits could be used to remedy emission exceedances of 
previously produced engines determined to be in nonconformity by in-use 
testing, production line testing or SEA failures. They would not be 
useable in handheld certification, and they would not be transferrable 
to nonhandheld engines, due to the considerable differences between the 
handheld and nonhandheld programs. Unlike certification credits for 
nonhandheld engines, they would not be useable for offsetting the high 
emissions from prospective production of an engine family following a 
PLT or SEA failure. In such cases, the manufacturer would be required 
to make a product change to improve emission performance of future 
production.
    EPA is proposing that these in-use credits could be used at any 
time during the Phase 2 program, and that any future rulemaking 
concerning Phase 3 standards would address the use of the Phase 2 
credits in Phase 3. EPA believes this unlimited life for in-use credits 
during the Phase 2 handheld program is justified since, if an engine 
demonstrates that it can remain under standards for its full useful 
life, then an environmental benefit has occurred and the manufacturer 
is entitled to that benefit for later use. However, unlimited life is 
not being extended beyond the Phase 2 program at this point, given the 
concern that Phase 2 credits could be used to effectively delay the 
implementation date of any Phase 3 standards. EPA requests comments on 
all aspects of credit life for in-use credits in the handheld in-use 
credits program.
    A manufacturer could use in-use credits to average against in-use 
failures identified in that model year's testing. It could bank the 
credits for use in a later model year or trade the credits to another 
manufacturer. Manufacturers could test additional families and would 
generate or require additional credits according to that testing. 
However, the manufacturer would be required to report all in-use 
testing to EPA, including any test engines that were deleted from the 
aging process or testing process, and to provide to EPA a technical 
justification to support the deletion.
    No restrictions are proposed on the application of in-use credits 
from one handheld engine class to another. EPA is not aware of any 
environmental or competitive concerns with allowing unrestricted use of 
in-use credits across handheld engine classes. EPA requests comments on 
the need for cross-class averaging restrictions, and the impact of 
having or not having them.
    EPA is also proposing an adjustment factor to increase credits 
earned as the in-use testing sample size increases, similar to the 
program promulgated for the gasoline marine engine rule (see 40 CFR 
91.1307). The proposal for an adjustment factor is reasonable because 
EPA's statistical certainty of the sample mean generally will increase 
with sample size.
    In addition, EPA is proposing a provision that would require 
manufacturers to apply in-use test results to two past and one future 
model year when the engine family being tested meets the carryover 
criteria for those model years. EPA contemplates that manufacturers 
would not make frequent significant changes to engine families and that 
carryover certification would be common. Essentially, under this 
provision, the test results from one model year could apply to up to 
four model years; the one subject to testing, the two previous model 
years and the next model year. In-use credits would be generated or 
required, as appropriate. EPA requests comment on the

[[Page 3992]]

appropriateness of and the need for these provisions.
    The handheld in-use credit program is meant, in part, to obviate 
the need to resort to a traditional recall program, and the Agency 
wants to ensure that this alternative program, or any other 
alternatives considered, provide incentives to manufacturers to design 
engine configurations that will comply with standards for their entire 
useful lives. EPA believes that manufacturers should make every effort 
to prove out their designs prior to certification so that in-use 
nonconformities will not occur. Therefore, this notice proposes that 
credits be discounted by 10 percent before they are used. This would 
require a manufacturer to obtain or generate credits sufficient to 
offset 110 percent of the emissions from a family found to be in 
noncompliance. This discount is consistent with that applied to in-use 
credits in the gasoline marine rule. Comment is requested on the 
appropriateness of such discounting and on the appropriate size of the 
discount.
4. Criteria for Evaluating Alternatives to Mandatory Recall
    This proposal contemplates that for handheld engines, in-use 
credits would be the primary method of addressing emission 
nonconformities determined through in-use testing or production line 
testing, whether through the use of credits banked or averaged, or 
credits purchased through available sources. For nonhandheld engines, 
EPA is proposing that in some cases, the use of certification credits 
would be allowed as a method of addressing emission exceedances 
determined through production line testing (as discussed above in 
Section IV.D.2).
    However, EPA is also proposing that manufacturers have available 
alternatives to using in-use credits or certification credits, if they 
lack sufficient credits and are unable to obtain them, that would still 
avoid necessitating an order for mandatory recall. One such alternative 
could be for the manufacturer to conduct a voluntary recall. However, 
EPA would consider other alternatives as well. This proposal contains a 
number of criteria for evaluating alternatives to determine whether 
they meet the goals of addressing the environmental impact of the in-
use problem while providing incentives to the manufacturer to produce 
emission-durable engines. EPA intends to allow a manufacturer to 
implement a reasonable alternative that met these criteria prior to 
making a determination of substantial nonconformity under section 207 
of the Act.
    In evaluating alternatives to mandatory recall, EPA would consider 
alternatives which (1) represent a new initiative that the manufacturer 
was not otherwise planning to perform at that time and that has a nexus 
to the emission problem demonstrated by the subject engine family; (2) 
cost substantially more than foregone compliance costs and consider the 
time value of the foregone compliance costs and the foregone 
environmental benefit of the subject family; (3) offset at least 100 
percent of the exceedance of the standard; and (4) are able to be 
implemented effectively and expeditiously and completed in a reasonable 
time.
    These proposed criteria would function as ground rules for 
evaluating projects to determine whether their nature and burden is 
appropriate to remedy the environmental impact of the nonconformity 
while providing assurance to the manufacturer that EPA would not 
require excessive projects.
    In addition to being evaluated according to the above criteria, EPA 
is proposing that alternatives would be subject to a cost cap, as 
contemplated by the proposal for handheld engines in the March 1997 
ANPRM. EPA proposes a cost cap of 75 percent above and beyond the 
foregone costs adjusted to present value, provided the manufacturer can 
appropriately itemize and justify these costs. EPA believes that this 
is an appropriate value which is both ``substantial'' and sufficient to 
encourage manufacturers to produce emission durable engines and 
maintain positive in-use credit balances.
    In deciding what cost cap to propose, EPA believes a figure of 75 
percent more than the foregone costs adjusted to present value is 
consistent with and informed by the principles inherent in the criteria 
for evaluating alternatives to recall. For example, criterion (2) would 
require that the alternative must cost substantially more than the 
costs the manufacturer was able to forego by producing a nondurable 
engine, and consider the time value of those foregone costs.
    EPA believes that manufacturers should prove out the in-use 
durability of their designs carefully before certification and desires 
to set the cost cap for alternative projects high enough that 
manufacturers will take measures to carefully evaluate in-use 
durability before certification and to bank and maintain substantial 
in-use credits to handle an unforeseen problem. EPA believes that a 
cost cap which would merely measure the foregone costs, and adjust them 
to their present value would not provide the appropriate incentive, 
because the manufacturer would ``break even'' and may become 
indifferent between assuring in-use durability up front and addressing 
it only when durability problems are detected.
    EPA is proposing in this rule that in-use credits be discounted by 
10 percent when they are used. If in-use credits are marketed freely 
and their price is determined by what it costs to generate them, a 
manufacturer would pay at least 10 percent more than it cost another 
manufacturer to comply with the standards and generate the credits. 
This suggests that the minimum figure for the cap should be at least 10 
percent of the failing manufacturer's foregone costs, after those costs 
have been adjusted to the present value. Given that under the proposal 
no more than one fourth of a manufacturer's families would be subject 
to in-use testing in a given year, a manufacturer that produces a non-
durable, non-carryover family has at most a 25 percent chance that EPA 
would be aware that such a non-durable family was being produced. A 
reasonable individual might risk a 10 percent cost penalty if the risk 
of actually having to pay it was never more than 25 percent. EPA can 
not estimate the savings a manufacturer may reap by building a non-
durable engine, and therefore can not compute the expected value of the 
savings when the 25 percent risk factor is added in.
    EPA believes a figure of 75 percent more than the foregone costs 
adjusted to present value would be both ``substantial'' and sufficient 
to encourage manufacturers to produce emission durable engines and 
maintain positive in-use credit balances. EPA notes that these projects 
are alternatives to recall and that a recall with a response rate 
similar to those in the motor vehicle program would likely have a much 
higher cost than would be permitted under a 75 percent cap. EPA 
considered proposing that the cap be tied to the cost of purchasing in-
use credits on the open market, but is concerned that these 
alternatives would be needed when there are no in-use credits available 
for sale. Further, based on EPA experience with other ABT programs, 
there is no guarantee that routine sales of credits would ever occur. 
EPA requests comment on the appropriate cap and the appropriate 
methodology for determining the cap, and the difficulties that could be 
faced in trying to ascertain foregone costs.

E. Flexibilities

    This section addresses a variety of flexibilities proposed today to 
ease the transition from the Phase 1 to the Phase 2 program, to ensure 
that the Phase 2

[[Page 3993]]

standards are cost-efficient and achievable, and to reduce the 
compliance burden while maintaining the environmental benefits of the 
rule. Following an overview of the approach to providing compliance 
flexibilities, and a discussion of the proposed cutoffs for determining 
whether a manufacturer, an engine family, or an equipment model would 
qualify for the flexibilities proposed today, this section describes 
the flexibility provisions proposed today, including general 
flexibilities, phase-in flexibilities, flexibilities to address the 
concerns of small volume engine manufacturers, flexibilities to address 
the concerns of small volume equipment manufacturers, and provisions to 
encourage engine availability. While some of these flexibilities may 
overlap, EPA is proposing these flexibilities as a means to reduce the 
compliance costs of the proposed rule for those that can least afford 
them, while maintaining the environmental benefits of the proposed rule 
and adopting the most stringent emissions standards achievable. EPA 
requests comment on the proposed flexibilities individually and as a 
whole.
1. Overview of Approach to Providing Compliance Flexibilities
    In this proposal, EPA has attempted to facilitate compliance by 
creating provisions that help avoid unnecessary hardship for engine and 
equipment manufacturers but that still achieve the desired 
environmental benefits. EPA believes that these provisions will help to 
avoid disruption of supplies of engines needed by equipment 
manufacturers and will enable both engine and equipment manufacturers 
to more easily and economically make the transition from Phase 1 to 
Phase 2. These provisions will also help ensure that the stringent 
standards proposed in the rule are achievable with technology that will 
be available during the Phase 2 time frame.
    Some engine manufacturers have expressed concern that the Phase 2 
program might be too burdensome for engine families with small volume 
production or for small volume manufacturers. These manufacturers have 
stated that, without some kind of relief, these burdens will lead them 
to stop producing certain engines rather than bear the additional 
costs. The engines most likely to be affected are special engines 
designed for niche markets. For these markets, there could be 
significant consequences to equipment manufacturers and operators if 
production of special engines were to cease. To address these concerns, 
EPA is proposing several compliance flexibilities intended especially 
to reduce the compliance burden on small volume products or small 
volume engine or equipment manufacturers.
2. Proposed Production Volume Cutoffs
    EPA has developed proposed cutoffs to determine whether a 
manufacturer or engine/equipment family would qualify for the 
flexibilities proposed today. These cutoffs are described here, with a 
more detailed discussion in Chapter 9 of the Draft RSD. EPA decided not 
to propose the Small Business Administration's definition of ``small 
business'' as the criterion for a manufacturer to qualify for the 
proposed flexibilities (the SBA definition is either 500 or 1000 
employees, depending on the SIC code of the industry). This is because, 
of 15 engine manufacturers qualifying as ``small business'' by the SBA 
definition, at least three produce large volumes of engines, between 
75,000 and 700,000 units, and have very high annual income. EPA 
believes these companies will not experience significant burdens in 
complying with the proposed Phase 2 program. Instead, EPA is proposing 
the following production volume cutoffs 50 for qualifying 
for the flexibilities proposed today.
---------------------------------------------------------------------------

    \50\ Annual production volume of U.S. sales, as defined by these 
proposed regulations. Note that the vast majority of ``small'' 
manufacturers together produce a very small fraction of the engines; 
a few very large manufacturers produce the large majority of the 
engines.
---------------------------------------------------------------------------

    First, nonhandheld engine manufacturers would be considered ``small 
volume engine manufacturers'' when their total annual production is 
10,000 units or less; handheld engine manufacturers would be considered 
``small volume engine manufacturers'' when their total annual 
production is 25,000 units or less. While over 50 percent of the 
nonhandheld engine manufacturers, and up to 30 percent of the handheld 
engine manufacturers could qualify under this proposed cutoff, fewer 
than 1 percent of the engines sold in the U.S. would be covered by 
these cutoffs.
    Second, nonhandheld small volume engine families would be those 
families of 1000 units of less; handheld small volume engine families 
would be those families of 2,500 units or less. These proposed 
thresholds were selected as high enough to include approximately 30 
percent of the engine families in each category, while low enough to 
account for less than 1 percent of the engines sold. At these levels, 
EPA believes a reasonable amount of flexibility could be provided to a 
significant number of manufacturers without undue risk of loss in 
emission control. In comments to the ANPRM, PPEMA has recommended 
10,000 units or less as a definition for small volume handheld 
families. Since this definition will impact the number of engines 
families within a manufacturer that could be exempt from PLT testing, 
EPA is uncertain as to why a larger sales volume cut-off is both 
appropriate from an enforcement perspective and of particular benefit 
to the manufacturer. EPA requests information on the necessity for 
expanding its small volume engine family definition to include larger 
volume family sales such as recommended by PPEMA (and a comparable 
volume for nonhandheld engine families), especially regarding the cost 
benefit to specific individual manufacturers, and the impact such a 
higher number would have on the confidence EPA would have that its PLT 
compliance program adequately evaluates the emission performance of the 
manufacturer's production.
    Third, equipment manufacturers using nonhandheld engines would be 
considered ``small volume equipment manufacturers'' when their total 
annual output across all models is 2500 units or less; equipment 
manufacturers using handheld engines would be considered ``small volume 
equipment manufacturers'' when their total annual output across all 
models is 5000 units or less. Again, while over 80 percent of the 
nonhandheld equipment manufacturers, and up to 67 percent of the 
handheld equipment manufacturers could qualify under this proposed 
cutoff, fewer than 2 percent of the nonhandheld engines and 1 percent 
of the handheld engines sold in the U.S. would be covered under these 
thresholds.
    Finally, equipment models using nonhandheld engines would be 
considered ``small volume equipment models'' when 500 or fewer units 
are produced per year; equipment models using handheld engines would be 
considered ``small volume equipment models'' when 2500 or fewer units 
are produced per year. On the nonhandheld side up to 3 percent of the 
equipment sold in the U.S. would be considered small volume equipment 
models. On the handheld side, up to 3.5 percent of the equipment sold 
in the U.S. would be considered small volume equipment models.
3. General Flexibilities
    The program proposed today contains several general provisions 
intended to facilitate compliance for engine manufacturers. One 
proposed flexibility, available to both handheld and

[[Page 3994]]

nonhandheld engine manufacturers, is the ability to carry-over 
certification from one year to the next. This would reduce 
certification costs after the first year for those engines using 
technology that does not change significantly from year to year.
    In addition, today's proposal contains two sets of proposed 
standard structure flexibilities which differ for handheld and 
nonhandheld engine manufacturers. For handheld engine manufacturers, 
the standards proposed in today's rule would be phased in, on a 
percentage of sales basis, which would facilitate compliance by 
allowing a manufacturer to spread initial compliance costs out over 
several years. It would also provide an opportunity for engine 
manufacturers to continue to supply Phase 1 engines to various 
equipment manufacturers, including the small volume equipment 
manufacturers that would also benefit from the special flexibilities 
described below.
    For nonhandheld engine manufacturers, a declining corporate average 
standard for Class II nonhandheld engines would achieve those same 
goals. In addition, nonhandheld engine manufacturers would benefit from 
the certification averaging, banking, and trading program, which would 
help reduce compliance costs by allowing manufacturers to meet the 
standards with the most cost-effective technologies. Today's proposal 
would also allow manufacturers of nonhandheld overhead valve engines to 
use an assigned deterioration factor for nonhandheld overhead valve 
engines, further easing the compliance burden by reducing the number of 
tests needed to determine compliance.
    For equipment manufacturers, EPA is proposing that the current 
provisions of 40 CFR 90.1003(b)(4) applicable for the transition from 
uncontrolled to Phase 1 emission regulations would also apply in 
concept during the transition from Phase 1 to Phase 2. Under today's 
proposal, equipment manufacturers would be allowed to continue to use 
Phase 1 engines until their stocks of engines are depleted, provided 
they do not engage in ``stockpiling'' (i.e., build up of an inventory 
of engines outside of normal business practices).
4. Phase-In Flexibilities
    In addition to these general flexibilities, EPA is proposing two 
other provisions that would be applicable to all manufacturers of 
certain kinds of nonhandheld engines to ease compliance during the 
phase-in of the standards and ensure their achievability. First, 
because manufacturers' testing capacities may be substantially 
constrained during the transition to fully-phased-in standards, EPA is 
proposing to allow manufacturers of Class II OHV nonhandheld engines 
who elect not to use assigned dfs to use good engineering judgment to 
establish deterioration factors for the 500 and 1000 hour useful life 
categories during the phase-in of the 12.1 g/kW-hr Class II standard, 
subject to the approval of the Administrator. Recognizing the need to 
verify deterioration factors established based on good engineering 
judgment, EPA is proposing that, beginning in 2006, the Administrator 
may direct manufacturers to verify such deterioration factors using the 
same process as that for calculating deterioration factors described in 
Section IV.D.1 above (i.e, aging at least three engines in the field 
and calculating the deterioration factor based on the average of the 
test data). EPA is also proposing that the manufacturer would be 
allowed to offset any emission shortfalls resulting from a low 
deterioration factor through the use of certification credits (see 
discussion, Section IV.A.5) or other compensating measures approved by 
the Administrator.
    Second, EPA is proposing an additional flexibility for 
manufacturers of Class II nonhandheld engines that use side-valve 
technology engines or engines with aftertreatment. During the 
transition to the Phase 2 standards, for engines which the manufacturer 
commits to cease production by the end of the 2004 model year, 
manufacturers would have the option to age engines for less than their 
full useful lives and extrapolate the deterioration factor to the full 
useful life using good engineering judgment.51 Again, 
demonstration of such good engineering judgment would need to be made 
to the satisfaction of the Administrator. For the engine families which 
the manufacturer commits to phase out, engines certified to 250 hours 
could be aged for 120 hours, engines certified to 500 hours could be 
aged to 250 hours, and engines certified to 1000 hours could be aged to 
500 hours. This flexibility, like the previous one, is intended to 
reduce the testing burden during the phase-in of the 12.1 g/kW-hr 
standard. However, EPA is not proposing to extend this flexibility to 
Class II engines which the manufacturer does not commit to cease 
production. In essence, this flexibility is designed to reduce the 
compliance burden at the start of the program for engines that are to 
be phased out, and thus to allow manufacturer to focus their resources 
on transitioning to engines that will meet the 2005 standards.
---------------------------------------------------------------------------

    \51\ As described in Section IV.D.1 of this preamble, Class II 
side-valve engines and engines with aftertreatment would be able to 
certify through a bench aging certification program, provided that a 
field/bench adjustment factor had been established.
---------------------------------------------------------------------------

5. Flexibilities for Small Volume Engine Manufacturers and Small Volume 
Engine Families
    EPA is proposing five compliance flexibilities to ensure the 
achievability of the standards and reduce the compliance burden on 
small volume engine manufacturers and small volume engine families, as 
follows.
    First, small volume engine manufacturers could opt out of mandatory 
production line testing. This option would apply only to nonhandheld 
engine manufacturers with a total annual production of 10,000 engines 
or less and to handheld engine manufacturers with a total annual 
production of 25,000 engines or less. These engines would be subject to 
SEA testing. However, EPA anticipates little such testing unless it 
receives evidence of nonconformities or other problems.
    Second, manufacturers of small volume nonhandheld engine families 
(those with total annual production of 1000 engines or less) and 
manufacturers of small volume nonhandheld engine families (those with 
total annual production of 2500 engines or less) could opt out of 
mandatory production line testing for those engine families. As above, 
these engines would remain subject to SEA testing, which would likely 
only occur if EPA had evidence of nonconformity.
    Third, manufacturers of very clean engine families, that is, those 
whose HC+NOX certification levels are at least 50 percent 
below the standard (or FEL, if applicable) could also opt out of 
mandatory production line testing for those families. These engines 
would also be subject to SEA testing, although EPA sees little 
likelihood of conducting SEAs on engines certified substantially below 
the standard (or FEL). EPA seeks comment on the margin below the 
standard (or FEL) necessary to qualify for this exemption.
    Fourth, small volume Class II side-valve technology engine families 
(whose annual production is 1,000 engines or less) would be allowed to 
meet an HC+NOX standard of 24 g/kW-hr, which represents the 
Phase 1 standard adjusted for deterioration. Note that these families 
could also opt out of mandatory production line testing, consistent 
with provision 2 above. This flexibility is intended to ensure that 
manufacturers can continue to produce these small

[[Page 3995]]

volume engines, many of which are used in niche-market specialty 
equipment.
    Fifth, small volume engine manufacturers could defer compliance 
with Phase 2 handheld requirements and Class II nonhandheld standards 
until the last year of the phase in. For handheld engines, this would 
mean that the engine manufacturer could, at its option, produce Phase 1 
engines exclusively through the 2004 model year, with full Phase 2 
compliance required in 2005. For nonhandheld Class II engines, the 
engines would be subject to the Phase 2 requirements beginning in 2001, 
but would not have to comply with the actual Phase 2 corporate average 
standards until the 2005 model year. These manufacturers could certify 
Class II engines to a standard of 24 g/kW-hr through 2004. These 
engines would neither use nor generate certification credits. If a 
small volume engine manufacturer desired to generate credits prior to 
the 2005 model year, it could do so for those engines certified below 
the applicable corporate average emission standard. Note that, 
consistent with the first provision above, these families would not 
have to be tested under mandatory production line testing. This 
flexibility is intended to provide another mechanism to reduce impact 
on small volume engine manufacturers and help ensure that manufacturers 
can continue to produce engines for specialty equipment.
    EPA is not proposing to specifically exempt from in-use testing any 
group of engines to which in-use testing requirements are applicable 
based on the group's or the manufacturer's size. The Agency believes 
that all engines should meet their standards (or FELs, as applicable) 
for their full useful life and that manufacturers should design engines 
to be emission durable. It is therefore appropriate that all engines to 
which in-use testing or demonstration requirements are applicable be 
subject to in-use testing. However, under this proposal, the choice of 
engines which would require in-use testing or demonstration is EPA's. 
EPA would not be inclined to identify for mandatory in-use testing a 
very small volume engine family or a family certified by a very small 
company unless there was evidence of a nonconformity. EPA requests 
comment on the appropriateness of this position.
6. Flexibilities for Small Volume Equipment Manufacturers and Small 
Volume Equipment Models
    Several equipment manufacturers who do not make their own engines 
have expressed concern that the transition to the Phase 2 program may 
disrupt their production because engine suppliers do not always provide 
adequate lead time for equipment redesigns needed to accommodate engine 
design changes. Engine changes could affect mounting and connection 
locations, heat rejection loads, and engine compartment requirements, 
for example. In addition, some equipment manufacturers cannot implement 
equipment design changes quickly, even with timely information from 
manufacturers because of the sheer volume of redesign work needed to 
change diverse product offerings with limited engineering staffs.
    EPA believes that the engine manufacturer flexibilities described 
above will extend the availability of engines currently used by small 
volume equipment manufacturers and will help ease the transition from 
Phase 1 to Phase 2 for those entities. However, to respond more 
directly to concerns raised by equipment manufacturers, EPA is 
proposing three compliance flexibilities to help enable equipment 
manufacturers to make the transition from Phase 1 to Phase 2 engines.
    First, EPA is proposing to temporarily exempt small volume 
equipment manufacturers from the requirement to use Phase 2 engines in 
cases where no Phase 2 engines with appropriate physical and 
performance characteristics are available to fit existing equipment 
models. This exemption would apply to those equipment manufacturers 
whose annual output across all models uses 2500 or fewer nonhandheld 
engines, or 5000 or fewer handheld engines, and would last through the 
third year after the last applicable phase-in date for that class of 
engines. Thus, for example, small volume equipment manufacturers who 
use Class II nonhandheld engines in an existing piece of equipment 
could continue using Phase 1 engines through the end of the 2008 model 
year, in cases where no suitable Phase 2 engines are available to fit 
existing equipment models.
    Second, EPA is proposing to delay the impact of the Phase 2 
requirements on individual small volume equipment models in cases where 
no suitable Phase 2 engines are available to fit existing equipment 
models. A small volume model, as proposed, is one with 500 or less 
units produced per year for nonhandheld equipment, and 2500 or fewer 
units produced per year for handheld equipment. These small volume 
models could continue to use Phase 1 engines throughout Phase 2, except 
as discussed below. EPA is proposing that this exemption would be 
allowed only for those equipment models in which a certified Phase 2 
engine will not fit, and would apply only to models in production prior 
to the effective date of the Phase 2 standards. This is to avoid 
encouraging manufacturers to bring out new models designed to use Phase 
1 engines after the Phase 2 standards have gone into effect. This 
exemption would also apply only so long as the equipment is not 
significantly modified. EPA believes that if the equipment manufacturer 
takes steps to significantly redesign a particular model, the use of a 
Phase 2 engine should be included. Finally, this exemption could apply 
only through the applicability of the Phase 2 program. EPA seeks 
comments on each of these restrictions, especially with regard to how 
they would affect equipment manufacturers who might incur a significant 
change in the cost of the engine if they were required to switch to a 
Phase 2 engine as the result of a significant model redesign.
    Finally, EPA is proposing a hardship relief provision by which any 
equipment manufacturer could obtain relief to continue using Phase 1 
engines, by demonstrating to the Administrator's satisfaction that, 
despite its best efforts, the manufacturer cannot meet the 
implementation dates without incurring substantial economic hardship, 
even with the transition flexibilities described above, due to 
unforeseeable factors beyond the equipment manufacturer's control. Such 
a situation may occur if an engine supplier were to change or drop an 
engine model very late in the implementation process. The intent of 
this provision is to recognize the concerns of equipment manufacturers 
about the uncertainty of timely supply of engines that meet equipment 
requirements by providing fair, objective criteria for hardship appeal 
that minimize the potential loss in environmental benefit, minimize the 
Agency's involvement in the financial affairs of the affected equipment 
manufacturer, and avoids straining the Agency's resources.
    As proposed, this hardship relief provision would require requests 
to be made in writing, submitted before the earliest date of 
noncompliance, include evidence that failure to comply was 
unforeseeable and was not the fault of the equipment manufacturer (such 
as a supply contract broken by the engine supplier), and include 
evidence that the inability to sell the subject equipment will have a 
major impact on the company's solvency. The Agency would work with the 
applicant to ensure that all other remedies available under the 
flexibility provisions are exhausted

[[Page 3996]]

before granting further relief, and would limit the period of relief to 
no more than one year. Furthermore, the Agency proposes that 
applications for hardship relief could only be submitted through the 
first year after the last effective date of the phase-in period. EPA 
seeks comment on all aspects of this flexibility provision and on 
whether the Agency should require those who receive relief to cover 
some of the lost environmental benefit, such as purchasing lower 
emitting engines.
7. Engine Availability
    EPA recognizes that the above-described equipment manufacturer 
flexibility provisions are of little use if Phase 1 engines are not 
available. Therefore, to help ensure availability of Phase 1 engines 
necessary for the above relief provisions to have full effect, EPA is 
proposing that engine manufacturers be allowed to build and sell the 
engines needed to meet the market demand created by these 
flexibilities. Specifically, EPA is proposing to continue to apply the 
Phase 1 compliance provisions to these engines. Thus, these Phase 1 
engines would not be subject to Phase 2 useful life, production line 
testing or in-use demonstration requirements contained in today's 
program, since Phase 1 engines are not currently subject to those 
provisions. EPA desires to minimize any disincentives that engine 
manufacturers may have to producing these engines for small volume 
equipment users and is therefore proposing that these engines would be 
counted only to the extent necessary to determine the availability of 
the specific flexibility item that was being applied. These engines 
would not count in any other calculation of compliance with phase in 
requirements or against any other ceilings or limits proposed in this 
rule. These engines would not be required to use any emission credits 
nor would they be permitted to generate any such credits.
    However, to prevent abuse of the ability to continue to produce 
Phase 1 engines, EPA believes it is necessary to impose some 
restrictions on the continued manufacture and sale of those engines. 
Therefore, EPA is proposing that equipment manufacturers procuring 
engines for use under the flexibility programs described above provide 
written assurance to the supplying engine manufacturer that such 
engines are being procured for this purpose. EPA requests comment on 
the need for a requirement that engine manufacturers maintain or 
annually provide to EPA records on the engines manufactured in support 
of the equipment manufacturer flexibilities described above, or whether 
EPA should rely on equipment manufacturer records.

F. Nonregulatory Programs

    The following is a description of three nonregulatory programs 
which, though outside of the scope of the regulation, could yield 
important environmental benefits from the small SI engine sector. The 
first program is a voluntary incentive and recognition program for low-
emitting nonhandheld and handheld engines, which would take the form of 
a ``green labeling'' program to identify engines which have emissions 
significantly lower than required by the proposed standards. The second 
program is a voluntary fuel spillage reduction program for nonhandheld 
and handheld engines. The third program is a particulate matter (PM) 
and hazardous air pollutant (HAP) testing program for handheld engines. 
These programs are described in the remainder of this section.
1. Voluntary ``Green'' Labeling Program
    EPA is very interested in encouraging the design, production, and 
sale of small engines which are substantially cleaner than would be 
required by today's proposed Phase 2 programs. EPA plans to implement a 
voluntary program which would include consumer labeling of engines and 
equipment with superior emission performance as a way of providing 
public recognition and also allowing consumers to easily determine 
which engines have especially clean emission performance. At this time, 
EPA is considering a threshold of around 50 percent of the proposed 
standard (e.g., around 12.5 g/kW-hr for Class I engines) as the level 
below which engines would qualify for ``green'' labeling. To develop 
the details of such a program, the Agency requests comment on all 
aspects of the program, including the threshold for determining a 
``green'' engine, whether the sales weighted certification level after 
dfs are applied should be used to establish the eligibility of an 
engine family, the design of and information to be included on the 
label, and other matters relevant to the successful implementation of 
the program. The Agency requests comment on program recommendations as 
part of today's proposal. In particular, the Agency seeks information 
on when such a program must be in place to effectively impact the sale 
of especially clean Phase 2 engines. The Agency is interested in 
working closely with consumer groups, engine and equipment 
manufacturers and others with an interest in making this program work. 
The Agency invites comment on the interest of any of these groups in 
working with the Agency to develop and implement this program.
2. Voluntary Fuel Spillage and Evaporative Emission Reduction Program
    EPA is planning to develop a voluntary fuel spillage and 
evaporative emission reduction program specifically for the small 
engine industry and its customers. While this program would not impose 
enforceable requirements on engine manufacturers subject to this 
rulemaking, it is important to reduce fuel spillage and other sources 
of evaporative emissions. Every year, millions of gallons of gasoline 
are lost during refueling. It is estimated that if a few ounces are 
spilled during each refueling of lawn and garden equipment, they would 
total about 17 million gallons of gasoline, most of which evaporates 
into the air to contribute to the ground-level ozone problems. To 
reduce and prevent this pollution, a variety of measures will be 
needed, most involving increased public awareness and education.
    The Agency believes it is appropriate to develop and implement a 
program targeted at the small SI industry and its customers to 
encourage public awareness and act as an incentive for technology 
investments. The Agency is interested in a voluntary partnership 
program which would involve EPA, engine manufacturers and equipment 
manufacturers, regional, state, and local air pollution agencies, 
health and environmental organizations, fuel container manufacturers, 
and other interested parties who would all contribute to the successful 
development and implementation of a voluntary fuel spillage and 
evaporative emission reduction program.
    While the design of such a program will benefit from the thoughtful 
input of all partners, the program would likely encourage the 
development of technology that will assist equipment users in reducing 
spills and evaporative emissions, provide recognition for implementing 
technology developments that will assist equipment users in reducing 
spills, and provide education and training to commercial operators of 
equipment and to those persons who influence individuals doing the 
refueling (such as equipment sales staff or small engine course 
instructors), and similar target audiences.
    Initial steps in this program involve identifying interested 
partners and convening a meeting to discuss the roles and 
responsibilities of each partner. The Agency seeks comment on the 
proposed voluntary partnership program, interest

[[Page 3997]]

in participating in this partnership, appropriate strategies and target 
audiences, and other matters pertinent to establishing this program.
3. Particulate Matter and Hazardous Air Pollutant Testing Program for 
Handheld Engines
    While section 213(a)(4) of the Clean Air Act allows EPA to 
establish standards for nonroad emissions of any air pollution which 
may reasonably be anticipated to endanger public health or welfare, 
today's notice does not propose to establish emission standards in 
Phase 2 for particulate matter (PM) or non-hydrocarbon hazardous air 
pollutants (HAP) listed under section 112(b) of the Clean Air Act. 
However, EPA and other parties have agreed that a PM and HAP test 
program will be conducted (see 62 FR 14746). The Portable Power 
Equipment manufacturers Association (PPEMA), in cooperation with EPA, 
will conduct a test program to evaluate and quantify emissions of PM 
and HAP including, but not limited to, formaldehyde, acetaldehyde, 
benzene, toluene, and 1,3 butadiene. EPA anticipates that testing will 
be conducted on Phase 2 technology handheld engines, with a sufficient 
magnitude of engines tested to represent the range of new basic 
technologies used to comply with Phase 2 small engine standards. EPA 
expects that the information generated by this program will be useful 
in informing any future implementation of section 213(a)(4) regarding 
small SI engines.

G. General Provisions

    This section includes a description of certain other general 
provisions proposed in today's notice, including provisions related to 
annual production period flexibilities during the transition to Phase 
2, the definition of handheld engines, a small displacement nonhandheld 
engine class, propane fueled indoor power equipment, dealer 
responsibility, engines used in recreational vehicles, engines used in 
rescue and emergency equipment, and replacement engines.
1. Model Year Definition and Annual Production Period Flexibilities 
During Transition to Phase 2
    The programs for nonhandheld and handheld engines proposed today 
would be effective beginning with the 2001 and 2002 model years, 
respectively. EPA is not proposing to change the Phase 1 definition of 
model year for Phase 2. That is, model year (MY) would continue to mean 
the manufacturer's annual new model production period which includes 
January 1 of the calendar year, ends no later than December 31 of the 
calendar year, and does not begin earlier than January 2 of the 
previous calendar year. When a manufacturer has no annual new model 
production period, model year would mean calendar year (see 40 CFR 
90.3). Under no circumstances would the model year definition be 
allowed to be interpreted to let existing models ``skip'' annual 
certification by pulling ahead the production of every other model 
year.
    In addition, in order to provide additional lead time for the 
implementation of the program for nonhandheld engines, EPA is proposing 
to adopt similar flexibilities for the beginning of the Phase 2 program 
for nonhandheld engines as were available for the Phase 1 program (see 
40 CFR 90.106 (a) and (b)). Thus, for the start up of Phase 2, EPA is 
proposing that every manufacturer of new nonhandheld engines produced 
during or after model year 2001 would be required to certify those 
engines to the Phase 2 program requirements. Nonhandheld engines 
manufactured during an annual production period beginning prior to 
September 1, 2000, would be allowed to certify to Phase 1 standards. 
However, annual production periods beginning prior to September 1, 
2000, would not be allowed to exceed 12 months in length. In effect, 
all nonhandheld engine families would be required to be certified to 
the Phase 2 program by September 1, 2001. EPA is not proposing this 
provision for handheld engines, which have both a later effective date 
as well as a phase-in of the Phase 2 program based on percentage of 
engine sales. EPA requests comment on whether similar provisions for 
handheld engines should be adopted (except that in the case of handheld 
engines, September 1 of each year would be the date that the percentage 
of engine sales requirements for Phase 2 certification would have to be 
met). EPA requests comments on all aspects of these provisions relating 
to annual production periods in the transition from Phase 1 to Phase 2 
certified engines.
2. Definition of Handheld Engines
    EPA is not proposing any changes to the criteria listed in Phase 1 
used to determine whether engines could be classed as Class III, IV or 
V. For Phase 2, EPA would continue to make determinations of 
applicability of the Class III, IV, or V standards based on the 
criteria found at 40 CFR 90.103(a)(2). During Phase 1, the 
multipositional use criterion has been used by EPA to make handheld 
determinations for certain two-person earth augurs, breakers and 
rammers, and power shovels. In each case, the manufacturer presented 
evidence to the satisfaction of the Agency demonstrating the 
multipositional use of the equipment, and provided a discussion of any 
constraints on engine design imposed by the usage of the equipment. The 
interpretation of multipositional use by EPA has been made relative to 
the equipment category and the technology available to meet the 
constraints imposed by the usage of the equipment.
    EPA received comment on the ANPRM that EPA should revise the 
definition of handheld.52 This commenter suggests that the 
Phase 1 definition of handheld restricts the replacement of 2-strokes 
by significantly cleaner 4-stroke engines, making it difficult to 
introduce a significantly cleaner engine for a product application. 
This commenter suggests that a different handheld definition and 
interpretation would improve the environment and permit the continued 
use of necessary products.
---------------------------------------------------------------------------

    \52\ See comments from Honda, Item #II-D-07 in EPA Air Docket A-
96-55.
---------------------------------------------------------------------------

    EPA believes that the current interpretation of criteria used to 
determine applicability of Class III, IV and V standards addresses this 
concern. Provided the 4-stroke engines are capable of performing the 
same intended functions as 2-stroke engines used in similar handheld 
applications, then EPA would likely determine that the 4-stroke engine 
also meets the criteria for applicability of the Class III, IV or V 
standards.
3. Small Displacement Nonhandheld Engine Class
    EPA has considered whether there is a need for changes or additions 
to the five classes of small SI engines for regulatory purposes. In 
particular, the Agency has considered whether there is a need for 
addition of a new, small displacement class that would be considered 
``nonhandheld.'' In comments on the ANPRM, one commenter specifically 
requested EPA to consider proposing a new class, as follows: the new 
class would be nonhandheld engines with displacements less than 75cc, 
and be subject to an in-use standard of 72.4 g/kW-hr with useful life 
categories of 125 hours and 250 hours. The commenter believes a new 
class for nonhandheld is needed for several reasons. The commenter 
believes the existing Phase 1 standards did not contemplate small 
displacement nonhandheld engines, yet the Phase 1 rule left a void in 
the market

[[Page 3998]]

which could be filled by small displacement nonhandheld engines. The 
commenter believes the Phase 1 standards prevented less than 75 cc 2-
stroke engines from being certified into some nonhandheld applications 
which utilize small displacement engines, but that the proposed Phase 2 
Class I standard is too stringent for less than 75 cc 4-strokes to 
meet.
    The Agency is not proposing the addition of a new small 
displacement nonhandheld class. The Agency believes that the proposed 
Class I standard, which can be met through averaging, will allow a full 
range of small displacement nonhandheld engines to certify to the 
proposed Phase 2 standards. If the proposed Class I standard can be met 
through averaging, the creation of a new displacement class with a 
higher standard could result in a smaller environmental benefit from 
the Phase 2 program.
    The Agency understands it is possible that some nonhandheld 
applications which use small displacement engines may no longer be able 
to utilize two-stroke engines if the Phase 2 standards are adopted as 
proposed, but believes that complying engines, perhaps of larger 
displacement, can be used. EPA requests additional information on this 
issue and the extent of its occurrence. The Agency also once again 
requests comment on the need for a new small displacement class, in 
particular, whether the proposed average Class I standard is sufficient 
to cover smaller displacement engines. The Agency also requests comment 
on the displacement cutoff (75cc), standard (72.4g/kW-hr), and useful 
lives (125 hours and 250 hours) suggested by the ANPRM commenter.
4. Liquefied Petroleum Gas Fueled Indoor Power Equipment
    Manufacturers of equipment using liquefied petroleum gas (LPG) have 
argued that their situation deserves special consideration within the 
Phase 2 regulations.53 The type of equipment they produce is 
often designed specifically for indoor use including, for example, 
floor washing and buffing equipment. The relatively low sales (likely 
fewer than 10,000 annually nationwide for the industry) and the fact 
that many of these manufacturers likely sell less than one thousand 
pieces of equipment annually means that both individually and 
collectively they account for a very small portion of the small SI 
engines sold annually. LPG is a popular fuel for indoor equipment due 
to the proven ability to calibrate LPG-fueled engines to operate at 
very low carbon monoxide (CO) levels. Low CO performance is especially 
important for indoor equipment to minimize CO exposure to the operator 
and others in the building. The Occupational Health and Safety 
Administration (OSHA) has set maximum CO standards for indoor ambient 
concentrations and some states have adopted even tighter indoor CO 
standards. While these are ambient standards, not emission limits for 
individual pieces of equipment, equipment manufacturers, to 
successfully market in this area, must be assured their equipment emits 
very low levels of CO and thus can be routinely used indoors without 
causing violations of OSHA or state indoor air quality requirements.
---------------------------------------------------------------------------

    \53\ See EPA Air Docket A-96-55, Items #II-D-02, II-D-04, and 
II-D-08.
---------------------------------------------------------------------------

    Because the specialized nature of their equipment places unique 
demands on these engines and due to the typically low sales volumes of 
many of the pieces of equipment, many of these indoor equipment 
manufacturers must not only design and produce their equipment but also 
to a significant extent are responsible for the modification of engines 
to power their equipment. In a number of cases these indoor equipment 
manufacturers buy gasoline-fueled engines and convert them to operate 
on LPG.
    While manufacturers of LPG-fueled indoor power equipment must power 
their equipment with engines which meet all the requirements of the 
small engine Phase 1 rules, the manufacturers argue that the proposed 
Phase 2 rules would add significantly to their burden. While meeting 
the proposed federal HC+NOX Phase 2 standard should not be 
particularly difficult for LPG engines compared to gasoline-fueled 
engines, the combined need to also achieve very low CO emission levels 
in order to not cause violations of indoor ambient CO standards may 
present a design challenge. The necessary controls may well exceed 
those required to meet just the Phase 2 standards and may include, for 
example, the use of electronically controlled fuel systems and perhaps 
catalysts. This could add significant cost to a relatively few engines. 
Even at a higher cost, those equipment manufacturers currently being 
supplied LPG-fueled engines by an original engine manufacturer are 
concerned that their suppliers may decide it is not worth the effort to 
supply engines complying with the Phase 2 standards. For those 
equipment manufacturers modifying engines to operate on LPG at low CO 
levels, the same technical challenges are faced while their ability to 
spread the development costs across their engines is limited by the low 
number of engines modified.
    While EPA has not done a thorough cost analysis for the impact of 
Phase 2 standards on this unique segment of the industry, EPA is 
persuaded that the technical challenges faced by this segment are 
significant. Many of these manufacturers would be considered ``small 
volume engine manufacturers'', with engines produced in ``small volume 
engine families'', under the criteria proposed today, and would 
therefore qualify for proposed compliance flexibilities for small 
volume engine manufacturers and small volume engine families. These 
include both additional flexibilities in the phase-in of the Phase 2 
standard, and also an option to opt out of mandatory production line 
testing. In effect, the additional phase-in flexibilities would allow 
nonhandheld manufacturers of indoor LPG-fueled power equipment engines, 
whose annual production of small SI engines is 10,000 units or less, to 
continue producing Class II nonhandheld engines which meet a Phase 1 
equivalent standard (24 g/kW-hr) until 2005. Beginning in 2005, when 
the Phase 2 standards are proposed to be fully phased in for gasoline-
fueled engines, these LPG-fueled engines are proposed to also be 
required to meet the Phase 2 HC+NOX standards. This extra 
lead time would allow manufacturers to spread their development efforts 
over several additional years, for those manufacturers choosing or 
required to make their own fuel modifications. In addition, while these 
engine families would be certified to the Phase 2 program, the cost of 
the proposed compliance program for these manufacturers would be 
minimized, as these manufacturers and engine families would likely 
qualify for the proposed flexibilities that would allow manufacturers 
to carry-over certification from one year to the next and to opt out of 
mandatory production line testing. The provisions for small volume 
engine manufacturers and small volume engine families are discussed in 
more detail in Section IV.E.
    Comments are requested on the impact of this proposed phase-in 
flexibility and other proposed compliance program flexibilities on the 
technical and economic ability of the indoor power equipment engine 
industry segment to successfully comply with the Phase 2 standard 
beginning in 2005, and any air quality impact concerns such a delayed 
implementation might cause.
    EPA is also requesting comment on the possible deletion of the 
existing Sec. 90.1003(b)(3). EPA believes this

[[Page 3999]]

provision may be of only limited utility for this program and believes 
it could prove problematic for small SI engines. This provision 
provides that certain activities connected to conversion of engines to 
alternative fuels will not be regarded as tampering. At one point, the 
existing regulatory paragraph makes reference to ``vehicle'' standards, 
of which, of course, there are none in the small SI program. Further, 
it might be misconstrued as requiring an engine modifier to reinstall 
hardware that was removed in the conversion process after the 
conversion was complete. Under such a misreading, a modifier engaged in 
converting gasoline engines to operate on propane might be viewed as 
having to reinstall the original gasoline carburetor on an engine after 
conversion, even if that were not feasible.
    Existing converters of small SI engines are currently certifying 
their products on the alternative fuel or are operating under EPA's 
tampering enforcement Memorandum 1-A. In light of this, for small SI 
engines, EPA believes that the discussion of the tampering implications 
of alternative fuel conversions for small SI engines could be best 
handled by the application of Memorandum 1-A. EPA does not expect that 
existing engine modifiers would be harmed by the deletion of this 
paragraph.
    Text similar to existing Sec. 90.1003(b)(3) is found in other 
nonroad rules. EPA intends, at some future date, to review the 
appropriateness and usefulness of this language in those rules.
5. Dealer Responsibility
    This proposal contains no new constraints or responsibilities for 
dealers and repair facilities from the Phase 1 rule. Dealers and repair 
shops, like all other persons, would continue to be prohibited from 
tampering or causing tampering. Tampering refers to the removal or 
rendering inoperative of any device or element of design installed on 
or in an engine for purposes of emission control.
    During the Phase 2 regulatory negotiation process, the issue of 
dealer responsibility was frequently raised out of concern that 
increasingly sophisticated control technologies would result in greater 
numbers of tampered engines being brought in for service. Another 
concern was that the Phase 2 rule not require that repair parts for 
emission control systems be obtained from the engine manufacturer.
    While all persons, including dealers and repair facilities, are 
prohibited from tampering or causing tampering, they are not prohibited 
from working on tampered engines. Under EPA tampering policies, dealers 
and repair facilities are not expected to restore tampered products to 
their originally certified and functioning configuration unless the 
repair is to the tampered system or a component of the tampered system. 
In such a case, the dealer or repair facility should restore the system 
to a certified and properly functioning condition, but need not conduct 
emission testing to verify compliance with emission standards. With 
regard to the use of emission control repair parts, dealers and repair 
facilities may use parts represented by their manufacturers to be 
functionally equivalent to original equipment parts.
6. Engines Used in Recreational Vehicles
    EPA is not proposing any changes to the provision in the Phase 1 
rule that engines used in recreational vehicles would not be subject to 
the small SI engine regulations. EPA continues to believe that these 
engines are more appropriately regulated under a rulemaking separate 
from this small SI engine program. Thus, these engines would remain 
outside the scope of the program when Phase 2 takes effect. The 
Agency's rationale for excluding engines used to propel recreational 
vehicles was presented in the preamble for the Phase 1 Notice of 
Proposed Rulemaking (NPRM) (see 59 FR 25403, 25414), and the Agency 
addressed the comments received on this topic in the Phase 1 Response 
to Comments document (see Section 3.8 ``Non-Coverage of Recreational 
Propulsion Engines'', EPA Air Docket A-93-25, Docket Item V-C-01). As 
discussed in the Preamble for the Phase 1 NPRM, ``EPA's primary reason 
for this exclusion is the extremely transient operation of the products 
in which these engines are used, which limits the ability of the 
proposed steady state test procedure to adequately represent exhaust 
emissions. This exclusion is not based on a determination that these 
engines do not contribute to air pollution and therefore need not be 
controlled.'' (59 FR 25414) EPA continues to be concerned that the test 
procedures covering the Phase 1 and Phase 2 engines may not be 
appropriate for engines used to propel recreational vehicles.
    Engines used in recreational vehicles are defined at 40 CFR 
90.1(b)(5), in part, as having a rated speed greater than or equal to 
5,000 RPM and having no installed speed governor. While EPA is not 
proposing any changes to the provisions which exclude recreational 
vehicles from this rule, EPA does wish to clarify that some engines 
with installed ``speed governors'' and with ungoverned rated speed 
above 5000 rpm still qualify as recreational. For example, engines used 
in typical recreational vehicles such as snowmobiles and 4-wheel ATVs 
which, when designed for use by children have ``speed governors'' 
installed for safety purposes to limit the top speed of the vehicle, 
have been found by EPA to be ``recreational vehicles'' in 
implementation of Phase 1. These vehicles are still operated in a 
typical fashion for recreational vehicles up to that top speed. During 
the development of the Phase 1 rule, the Agency was not aware of the 
existence of snowmobiles designed for children, and therefore not aware 
of the existence of snowmobiles with ``speed governors.'' The Agency 
would like to clarify that EPA continues to believe snowmobiles should 
not be covered under this rule, including snowmobiles designed for use 
by children which may in fact have a ``speed governor'' installed for 
safety purposes.
7. Engines Used in Rescue and Emergency Equipment
    In consideration of safety factors associated with compliance with 
the Phase 2 program, today's proposal includes a provision that would 
exempt engines which are used exclusively in emergency and rescue 
equipment from compliance with any standards if the equipment 
manufacturer can demonstrate that no certified engine is available to 
power the equipment safely and practically. Although under Phase 1 EPA 
has received no reports of problems caused by the need to use certified 
engines in emergency and rescue equipment, EPA is concerned that such 
problems could arise. EPA foresees this exemption applying especially 
to handheld items used to work in tight places to perform such tasks as 
cutting metal to extricate passengers from wrecked vehicles, if the 
size, heat or other characteristics of the certified engine would 
render its use unsafe. EPA does not foresee this exemption applying to 
portable generators, compressors or hydraulic pumps that may be used to 
power rescue equipment from a distance, since such devices are not as 
subject to the size, weight and other considerations surrounding a tool 
that contains its own source of power.
    EPA proposes this exemption to avoid any possible conflict between 
emission control and public safety. EPA wishes to reduce the chance 
that a piece of rescue equipment will go out of production or become 
more cumbersome because of the need to use certified engines. EPA sees 
no significant air quality impact

[[Page 4000]]

from such an exemption, because it would apply only to engines that are 
few in number and are subject to infrequent use for very short periods 
of time. In fact, EPA is not currently aware of any engine that is used 
exclusively in emergency or rescue equipment. The exemption, as 
proposed, would apply to engines and equipment produced during the 
remainder of the Phase 1 period as well as Phase 2 engines and 
equipment.
8. Replacement Engines
    After promulgation of the Phase 1 rule, equipment manufacturers 
approached EPA with concerns that, once the rule took effect, they 
would not be able to obtain replacement engines to repair certain items 
of more expensive equipment such as commercial mowing and construction 
equipment when their engines fail. The equipment manufacturers provided 
evidence that many Phase 1 engines, especially Class II nonhandheld 
engines, would be configured differently from uncertified engines and 
would not fit in the engine compartments of some pre-regulatory 
equipment. The equipment manufacturers explained that occasional engine 
failures are often best remedied by replacing the engine. Commercial 
operators, many of whom are small businesses, may not be able to afford 
the downtime associated with waiting for an extensive engine repair. In 
effect, repairing the engines becomes more costly than replacing the 
engines, and may be less environmentally beneficial. EPA evaluated 
these concerns and gathered information from engine manufacturers, 
equipment manufacturers and their associations. EPA concluded that 
permitting the sale of uncertified replacement engines, which likely 
constitute less than one percent of annual small SI engine sales, was a 
cheaper alternative that was no worse for air quality than the repair 
or rebuilding of the failed engines, which were not prohibited by the 
Phase 1 rule. On August 7, 1997 (62 FR 42638), EPA issued a direct 
final rule amending the Phase 1 rule to allow engine manufacturers to 
sell uncertified engines for replacement purposes subject to certain 
controls designed to prevent abuse.54 These controls require 
that the engine manufacturer ascertain that there is no currently 
certified engine that will fit in the equipment, that the engines be 
labeled for replacement purposes only, and that the engine manufacturer 
or its agent take ownership and possession of the old engine.
---------------------------------------------------------------------------

    \54\ The docket for this rulemaking, EPA Air Docket #A-97-25, is 
incorporated by reference.
---------------------------------------------------------------------------

    An environmental group has recently expressed concern to EPA about 
the replacement engine provisions for small SI engines published in the 
direct final rule described above. This group recommends that 
additional constraints and controls should be placed on the sale of 
these engines to prevent abuse since these engines either will not be 
built to comply with any standards, or will be built to comply with 
Phase 1 standards after those standards have been superseded by Phase 2 
standards.
    In today's notice, EPA is proposing to continue the replacement 
engine provision with an accommodation necessary to address Phase 1 
engines after the implementation of Phase 2. EPA is also proposing 
additional requirements to address the concerns of the environmental 
group and better ensure that the ability to use replacement engines is 
not abused.
    During Phase 2, the universe of small SI engines will expand to 
include uncertified engines, Phase 1 engines and Phase 2 engines. 
Consequently, the provision as proposed would be amended to permit 
uncontrolled engines to be sold for pre-regulatory equipment, and Phase 
1 engines to be sold for equipment built with Phase 1 engines, subject 
to certain constraints. EPA has no reason to believe that this 
provision will result in significant adverse air quality impacts. In 
fact, many replacement engines for older equipment will be certified 
Phase 2 engines. This provision provides flexibility and cost savings 
for equipment operators. It affects primarily commercial equipment 
where the equipment cost is high enough to justify major engine repairs 
or replacement and the usage of the equipment is such that downtime for 
repairs is costly. Replacement engines are not typically used in 
handheld equipment, nor in lower cost nonhandheld items such as walk 
behind mowers. A more detailed discussion of the rationale for the 
replacement engine provision can be found in the preamble to the direct 
final rule cited above.
    Although EPA does not believe that replacement engines will cause 
any significant air quality impacts, it is proposing to add safeguards 
and reporting and record keeping requirements to further ensure against 
abuse. EPA is proposing to amend the existing replacement engine 
provisions to require: (1) that manufacturers follow specific 
guidelines when ascertaining that no certified engine is available 
which can suitably repower a specific item of equipment; (2) that old 
engines being replaced are destroyed; (3) that engine manufacturers 
report to EPA annually the number of uncertified engines sold under the 
replacement engine provisions; (4) that manufacturers keep records, 
accessible to EPA, of the purchasers, quantities and equipment 
applications of replacement engines; and (5) that there be a limit on 
the time period for which uncertified replacement engines are normally 
available. EPA requests comment on the need for these additional 
requirements, and the burden they may pose to industry, equipment 
operators and engine distributors.

V. Environmental Benefit Assessment

    National Ambient Air Quality Standards (NAAQS) have been set for 
criteria pollutants which adversely affect human health, vegetation, 
materials and visibility. Concentrations of ozone (O3) are 
impacted by HC and NOX emissions. Ambient concentrations of 
CO are, of course, impacted by CO emissions. EPA believes that the 
standards proposed today would reduce emissions of HC and 
NOX and help most areas of the nation in their progress 
towards compliance with the NAAQS for ozone. The following provides a 
summary of the roles of HC and NOX in ozone formation, the 
estimated emissions impact of the proposed regulations, and the health 
and welfare effects of ozone, CO, hazardous air pollutants, and 
particulate matter.
    Much of the evaluation of the health and environmental effects 
related to HC, NOX and CO found in this section is also 
discussed in the draft Regulatory Support Document (RSD), and in the 
March 1997 ANPRM. EPA encourages comments on the Agency's beliefs 
expressed in this proposal and in the RSD, a copy of which is in the 
public docket for this rulemaking.

A. Roles of HC and NOX in Ozone Formation

    Both HC and NOX contribute to the formation of 
tropospheric ozone through a complex series of reactions. In a recent 
report, researchers emphasize that both HC and NOX controls 
are needed in most areas of the United States.55 EPA's 
primary reason for controlling emissions from small SI engines is the 
role of their HC emissions in forming ozone. Of the major air 
pollutants for which NAAQS have been designated under the CAA, the most 
widespread problem continues to be ozone, which is the most prevalent 
photochemical oxidant and an

[[Page 4001]]

important component of smog. The primary ozone NAAQS represents the 
maximum level considered protective of public health by the EPA. Ozone 
is a product of the atmospheric chemical reactions involving oxides of 
nitrogen and volatile organic compounds. These reactions occur as 
atmospheric oxygen and sunlight interact with hydrocarbons and oxides 
of nitrogen from both mobile and stationary sources.
---------------------------------------------------------------------------

    \55\ National Research Council, Rethinking the Ozone Problem in 
Urban and Regional Air Pollution, National Academy Press, 1991.
---------------------------------------------------------------------------

    A critical part of this problem is the formation of ozone both in 
and downwind of large urban areas. Under certain weather conditions, 
the combination of NOX and HC has resulted in urban and 
rural areas exceeding the national ambient ozone standard by as much as 
a factor of three. Thus it is important to control HC over wider 
regional areas if these areas are to come into compliance with the 
ozone NAAQS.

B. Health and Welfare Effects of Tropospheric Ozone

    Ozone is a powerful oxidant causing lung damage and reduced 
respiratory function after relatively short periods of exposure 
(approximately one hour). The oxidizing effect of ozone can irritate 
the nose, mouth, and throat causing coughing, choking, and eye 
irritation. In addition, ozone can also impair lung function and 
subsequently reduce the respiratory system's resistance to disease, 
including bronchial infections such as pneumonia.
    Elevated ozone levels can also cause aggravation of pre-existing 
respiratory conditions such as asthma.56 Ozone can cause a 
reduction in performance during exercise even in healthy persons. In 
addition, ozone can also cause alterations in pulmonary and 
extrapulmonary (nervous system, blood, liver, endocrine) function.
---------------------------------------------------------------------------

    \56\ United States Environmental Protection Agency, Review of 
the National Ambient Air Quality Standards for Ozone--Assessment of 
Scientific and Technical Information: OAQPS Staff Paper, EPA-450/2-
92-001, June 1989, pp. VI-11 to 13.
---------------------------------------------------------------------------

    The newly revised primary NAAQS 57 for ozone based on an 
8-hour standard of 0.08 parts per million (ppm) is set at a level that, 
with an adequate margin of safety, is protective of public health. EPA 
also believes attainment of the new primary standard will substantially 
protect vegetation. Ozone effects on vegetation include reduction in 
agricultural and commercial forest yields, reduced growth and decreased 
survivability of tree seedlings, increased tree and plant 
susceptibility to disease, pests, and other environmental stresses, and 
potential long-term effects on forests and ecosystems.
---------------------------------------------------------------------------

    \57\ See 62 FR 38896, Friday, July 18, 1997.
---------------------------------------------------------------------------

    High levels of ozone have been recorded even in relatively remote 
areas, since ozone and its precursors can travel hundreds of miles and 
persist for several days in the lower atmosphere. Ozone damage to 
plants, including both natural forest ecosystems and crops, occurs at 
ozone levels between 0.06 and 0.12 ppm.58 Repeated exposure 
to ozone levels above 0.04 ppm can cause reductions in the yields of 
some crops above ten percent.59 While strains of some crops 
are relatively resistant to ozone, many crops experience a loss in 
yield of 30 percent at ozone concentrations below the pre-revised 
primary NAAQS.60 The value of crops lost to ozone damage, 
while difficult to estimate precisely, is on the order of $2 billion 
per year in the United States.61 The effect of ozone on 
complex ecosystems such as forests is even more difficult to quantify. 
However, there is evidence that some forest types are negatively 
affected by ambient levels of ozone.62 Specifically, in the 
San Bernadino Mountains of southern California, ozone is believed to be 
the agent responsible for the slow decline and death of ponderosa pine 
trees in these forests since 1962.63
---------------------------------------------------------------------------

    \58\ U.S. EPA, Review of NAAQS for Ozone, p. X-10.
    \59\ U.S. EPA, Review of NAAQS for Ozone, p. X-10.
    \60\ See 62 FR 38856, Friday, July 18, 1997.
    \61\ U.S. EPA, Review of NAAQS for Ozone, p. X-22.
    \62\ U.S. EPA, Review of NAAQS for Ozone, p. X-27.
    \63\ U.S. EPA, Review of NAAQS for Ozone, p. X-29.
---------------------------------------------------------------------------

    Finally, by trapping energy radiated from the earth, tropospheric 
ozone may contribute to heating of the earth's surface, thereby 
contributing to global warming (that is, the greenhouse 
effect),64 although tropospheric ozone is also known to 
reduce levels of UVB radiation reaching the earth's surface, the 
increase of which is expected to result from depletion of stratospheric 
ozone.65
---------------------------------------------------------------------------

    \64\ NRC, Rethinking the Ozone Problem, p. 22.
    \65\ The New York Times, September 15, 1992, p. C4.
---------------------------------------------------------------------------

C. Estimated Emissions Impact of Proposed Regulation

    The emission standards proposed in today's action should reduce 
average in-use exhaust HC+NOX emissions from small SI 
engines 30 percent beyond Phase 1 standards by year 2025, by which time 
a complete fleet turnover is realized. This translates into an annual 
nationwide reduction of roughly 134,674 tons of exhaust 
HC+NOX in year 2025 over that expected from Phase 1. 
Reductions in CO beyond Phase 1 levels, due to improved technology, is 
also to be expected by year 2025.
    Along with the control of all hydrocarbons, the proposed standards 
should be effective in reducing emissions of those hydrocarbons 
considered to be hazardous air pollutants (HAPs), including benzene and 
1,3-butadiene. However, the magnitude of reduction would depend on 
whether the control technology reduces the individual HAPs in the same 
proportion as total hydrocarbons.
    These emission reduction estimates are based on in-use population 
projections using estimates of annual engine sales, engine attrition 
(scrappage), activity indicator, and current new engine and proposed 
in-use emission factors. Data on activity indicators were based on the 
Phase 1 small SI regulation. Estimates of annual engine sales for years 
from 1973 to 1995 were based on engine data available from the PSR 
databases 66 and national shipment data provided by Outdoor 
Power Equipment Institute (OPEI), the Portable Power Equipment 
Manufacturers Association (PPEMA), and a study done for the California 
Air Resources Board by Booz Allen & Hamilton (BAH). Sales projections 
into the future were for the most part based on estimates of population 
growth for the United States. Attrition rates (survival probability 
that an engine remains in service into a specific calendar year) for 
all engines included in this analysis were developed on the assumption 
that the equipment attrition function may be represented by a two-
parameter Weibull cumulative distribution function. The in-use emission 
factors are based on a multiplicative deterioration factor which is a 
function of the square root of the hours of equipment usage.
---------------------------------------------------------------------------

    \66\ Power Systems Research, Engine Data and Parts Link data 
bases, St. Paul, Minnesota, 1992.
---------------------------------------------------------------------------

    For the analysis summarized in Table 18, the emission inventories 
were developed for the five regulated engine classes as well as for all 
pieces of equipment using engines covered by this proposed rule. Using 
estimated engine sales and attrition, EPA projected the total in-
service engine population for each year from 1973 to 2025. EPA 
projected the total annual nationwide HC, NOX and CO 
emissions from small SI engines included in the proposal under the 
baseline (that is, with Phase 1 controls applied) and controlled (Phase 
2) scenarios.
    For the controlled scenario, EPA assumed that for both handheld and

[[Page 4002]]

nonhandheld engines the standards would be phased in on a percentage of 
production basis as proposed in today's notice. Deterioration factors 
were determined using manufacturer-supplied in-use emission data and 
other relevant information.

                         Table 18.--Projected Annual Nationwide Exhaust HC+NOX Emissions                        
                                                   [tons/year]                                                  
----------------------------------------------------------------------------------------------------------------
                                          Without proposed                      Tons reduced                    
                  Year                    controls  (Phase    With proposed     from  Phase 1      Percentage   
                                                 1)             controls      revised baseline      reduction   
----------------------------------------------------------------------------------------------------------------
2000....................................           378,700           378,700                                    
2005....................................           368,195           297,873            70,322              19.1
2010....................................           389,641           279,061           110,580              28.4
2015....................................           414,626           292,829           121,797              29.4
2020....................................           439,413           309,221           130,192              29.6
2025....................................           452,973           318,299           134,674              29.7
----------------------------------------------------------------------------------------------------------------

    For simplicity in modeling the projected emission reductions, the 
Agency has assumed in the emissions inventory model that under the 
Phase 2 program, each engine would meet the proposed standard for the 
minimum useful life category: i.e., Class I engines meet the proposed 
standards at 66 hours; Class 2 engines at 250 hours; and Classes III, 
IV, and V at 50 hours. Therefore, the Agency has under estimated the 
emission benefits of the proposed standards, because some engines will 
be certifying to the longer useful life categories, and therefore a 
greater emission reduction than predicted in Table 18 will occur. The 
Agency will attempt to address this issue for a more accurate 
prediction of the emission benefits of the proposed program for the 
final rule.
    In addition to the reductions in exhaust HC+NOX 
emissions, the Agency is also estimating the proposed standards would 
result in a small reduction in HC refueling emissions (refueling 
emissions are HC emissions caused from fuel spillage and vapor 
displacement during the refueling of a small engine). As discussed in 
the RSD, refueling emissions represent approximately an additional 
89,000 tons/year of HC in 2025 without Phase 2 controls. The Agency 
estimates that refueling emissions would be reduced under Phase 2 by 
the percent reduction in fuel consumption under Phase 2. The Agency 
estimates the proposed Phase 2 program would result in approximately a 
9 percent reduction in fuel consumption by 2025. Therefore, the Agency 
estimates refueling emissions would be reduced by 9 percent. A 9 
percent reduction in refueling emissions equates to an approximate 
8,000 ton/year reduction in HC emissions in 2025.

D. Health and Welfare Effects of CO Emissions

    Carbon monoxide (CO) is a colorless, odorless gas which can be 
emitted or otherwise enter into ambient air as a result of both natural 
processes and human activity. Although CO exists as a trace element in 
the troposphere, much of human exposure resulting in elevated levels of 
carboxyhemoglobin (COHb) in the blood is due to incomplete fossil fuel 
combustion, as occurs in small SI engines.
    The concentration and direct health effect of CO exposure are 
especially important in small SI engines because the operator of a 
small SI engine application is typically near the equipment as it 
functions. In some applications, the operator must be adjacent to the 
exhaust outlet and is in the direct path of the exhaust as it leaves 
the engine. According to numbers published in the Nonroad Engine and 
Vehicle Emission Study (NEVES), a 4-stroke, 2.9 kW lawnmower engine 
emits 1051.1 g/hr CO, while a 2-stroke, 2.9 kW engine emits 1188.4 g/hr 
CO.
    A Swedish study 67-69 on occupational exposure to 2-
stroke chainsaw exhaust concludes, among other things, that a rich 
fuel-air mixture results in high levels of CO emissions (a mean 
exposure rate of 37.0 mg/m3). The work conditions that gave 
rise to the most intense problems for loggers were deep snow, thick 
forest stands and calm weather. The main discomforts experienced by 
loggers from chainsaw exhaust were cough and eye, nose and throat 
irritation. In view of the discomfort experienced by loggers and the 
complex nature of the exposure to chainsaw exhaust, it was recommended 
that action be taken to reduce exposure by making technical 
modifications to the engine or control exhaust emissions.
---------------------------------------------------------------------------

    \67-69\ Occupational Exposure to Chain Saw Exhausts in Logging 
Operations, Am. Ind. Hyg. Assoc. J48, 1987.
---------------------------------------------------------------------------

    The toxicity of CO effects on blood and tissues, and how these 
effects manifest themselves as organ function changes, have also been 
topics of substantial research efforts. Such studies provided 
information for establishing the National Ambient Air Quality Standard 
for CO. The current primary and secondary NAAQS for CO are 9 parts per 
million for the one-hour average and 35 parts per million for the 
eight-hour average.

E. Health and Welfare Effects of Hazardous Air Pollutant Emissions

    The focus of today's action is reduction of HC emissions as part of 
the solution to the ozone nonattainment problem. However, direct health 
effects are also a reason for concern due to direct human exposure to 
emissions from small SI engines during operation of equipment using 
such engines. Of specific concern is the emission of hazardous air 
pollutants (HAPs). In some applications, the operator must be adjacent 
to the exhaust outlet and is in the direct path of the exhaust as it 
leaves the engine. Today's proposed regulations should be effective in 
reducing HAPs such as benzene and 1,3-butadiene, in so far as these are 
components of the HC emissions being reduced by the Phase 2 standards.
    Benzene is a clear, colorless, aromatic hydrocarbon which is both 
volatile and flammable. Benzene is present in both exhaust and 
evaporative emissions. Health effects caused by benzene emissions 
differ based on concentration and duration of exposure. The 
International Agency for Research on Cancer (IARC), classified benzene 
as a Group I carcinogen., namely an agent carcinogenic to humans. 
Exposure to benzene has also been linked with genetic changes in humans 
and animals. 1,3-butadiene is a colorless, flammable

[[Page 4003]]

gas at room temperature. This suspected human carcinogen is insoluble 
in water and its two conjugated double bonds make it highly reactive. 
1,3-butadiene is formed in internal combustion engine exhaust by the 
incomplete combustion of the fuel and is assumed not present in 
evaporative and refueling emissions.
    Epidemiologic studies of occupationally exposed workers were 
inconclusive with respect to the carcinogenity of 1,3-butadiene in 
humans. IARC has classified 1,3-butadiene as a Group 2A, probable human 
carcinogen. Other adverse noncancer health effects due to very high 
levels of exposure include heart, blood and lung diseases.
    Since air toxic levels generally decrease in proportion to overall 
emissions once emission control technology is applied, the amount of 
benzene and 1,3-butadiene produced by new small SI engines should 
diminish after this rule becomes effective. Consequently, exposure to 
HAPs from new small SI engines would be reduced, as would associated 
health and environmental effects. Although there is little data on 
direct health effects of small SI engines, the Swedish study concludes 
benzene emissions from chain saw engines as being rather high. No study 
has been conducted involving the health effects of HAP emissions 
specifically from small SI engines. The Agency requests additional 
information on this topic.

F. Particulate Matter

    Particulate matter, a term used for a mixture of solid particles 
and liquid droplets found in the air, has been linked to a range of 
serious respiratory health problems. These fine particles are of health 
concern because they easily reach the deepest recesses of the lungs. 
Batteries of scientific studies have linked particulate matter, 
especially fine particles (alone or in combination with other air 
pollutants), with a series of significant health problems including 
premature death, aggravated asthma and chronic bronchitis and increased 
hospital admissions. EPA has recently (July 1997) announced new NAAQS 
standards for particulate matter (PM) , by adding two new primary PM2.5 
standards set at concentrations of 15 micrograms per cubic meter 
(g/m3), annual arithmetic mean, and 
65g/m3, 24-hour average, to provide increased 
protection against the PM-related health effects found in community 
studies. EPA believes that the new standards will protect and improve 
the lives of millions of Americans.
    Separate from the proposed rule, which would not establish emission 
standards for PM or toxic air contaminants listed under section 112(b) 
of the Clean Air Act, an agreement with PPEMA to conduct PM/HAP testing 
program for handheld engines in cooperation with EPA has been reached. 
Testing under the program would be conducted on Phase 2 technology 
handheld engines at EPA, industry, and/or independent facilities. The 
test program is to be designed to evaluate and quantify emissions of 
particulate matter and toxics including, but not limited to: 
formaldehyde, acetaldehyde, benzene, toluene and 1,3 butadiene.

VI. Economic Impacts

    EPA has calculated the cost effectiveness of this proposed rule by 
estimating costs and emission benefits from these engines. EPA made 
best estimates of the combination of technologies that an engine 
manufacturer might use to meet the new standards, best estimates of 
resultant changes to equipment design, engine manufacturer compliance 
program costs and engine fuel savings in order to assess the expected 
economic impact of the proposed Phase 2 emission standards. Emission 
benefits are taken from the results of the environmental benefit 
assessment (Section V, above). The cost-effectiveness result of this 
rule is $390 per ton of HC+NOX when fuel savings are not 
taken into account. When fuel savings are also considered, the cost-
effectiveness calculation results in -$700 per ton of 
HC+NOX. This section describes the background and analysis 
behind these results.
    The analysis for this proposed rulemaking is based on data from 
engine families certified to EPA's Phase 1 standards. It does not 
include any engine families or production volumes that are covered by 
CARB's Tier 1 standard. The California Air Resources Board (CARB) will 
implement emission standards for many of these engines a year or two 
prior to the proposed federal Phase 2 regulations. Therefore, this rule 
only accounts for costs for each engine sold outside California and 
those engines sold in California that are not covered by the CARB Tier 
II rulemaking, such as those used in farm and construction equipment. 
Although EPA expects that engines already designed to meet CARB's 
earlier standards would incur no additional design cost to meet federal 
standards, no effort was made to estimate which models would be sold in 
California and subject to the earlier California standards. Rather for 
the purpose of this proposal, any Phase 1 engine design that would need 
to be modified to meet Phase 2 standards was assumed to incur the full 
cost of that modification including design cost. Similarly, the cost to 
equipment manufacturers was assumed to be fully attributed to this 
federal rule even if an equipment manufacturer would have to make the 
same modifications in response to the CARB Tier 2. Therefore, in both 
of these cases, the cost to the manufacturer due to these proposed 
rules is likely over estimated. EPA requests comment on these 
assumptions. The details of EPA's cost and cost-effectiveness analyses 
can be found in Chapters 4 and 7 of the Draft RSD.

A. Engine Technologies

    Table 19 lists the changes in technology, compared to Phase 1 
engines, that have been considered in the cost estimation for this 
rulemaking. As discussed in Section IV.A of this preamble, the proposed 
standards would require different engine improvements amongst the five 
classes and engine designs within those classes.70 For 
example, several Class I SV models are expected to require some 
internal improvements to reduce new engine out emissions and several 
additional components to increase emission durability. For the purposes 
of this cost analysis, Class II standards are assumed to require that 
engines be of clean OHV design. For Classes III-V, the proposed 
standards for the handheld engines are assumed to require improved 
scavenging techniques, for the two stroke engines, to be developed to 
reduce the approximately 30 percent of the air/oil/fuel mixture that 
traditionally escapes from these engines unburned. This analysis 
assumes that engine manufacturers would not be required to adopt 
advanced technologies such as catalysts or fuel injection systems. 
Manufacturers who did adopt such technologies would choose to do so for 
other perceived benefits. Therefore, the cost of such optional 
technology is not included in this cost estimate. Additional detail 
regarding the impact of these modifications can be found in Chapter 3 
and 4 of the Draft RSD.
---------------------------------------------------------------------------

    \70\ Currently, carbureted two-stroke, four-stroke side-valve 
and four-stroke overhead valve engine designs comprise the vast 
majority of engines used in nonhandheld and handheld applications.

[[Page 4004]]



Table 19.--Potential Technology Improvements Per Class and Engine Design
------------------------------------------------------------------------
      Class               Engine design                Technologies     
------------------------------------------------------------------------
I...............  4 stroke--SV.................  Carburetor             
                                                  Improvements.         
                                                 Combustion Chamber     
                                                  Improvements and      
                                                  Intake System.        
                                                 Improved Oil           
                                                  Consumption (Piston   
                                                  oil control rings,    
                                                  valve stem seals).    
I...............  4 stroke--OHV................  None necessary.        
I...............  2 stroke.....................  None necessary.        
II..............  4 stroke--SV.................  Conversion to clean    
                                                  OHV.                  
II..............  4 stroke--OHV................  Piston and piston ring 
                                                  improvements.         
                                                 Improved combustion and
                                                  intake system.        
III-V...........  2 stroke.....................  Carburetor             
                                                  Improvements.         
                                                 Improved Scavenging and
                                                  Combustion Chamber    
                                                  Design.               
                                                 Manufacturing Tolerance
                                                  Improvements.         
IV..............  4-stroke.....................  None necessary.        
------------------------------------------------------------------------

B. Engine Costs

    The engine cost increase is based on incremental purchase prices 
for new engines and is comprised of variable costs (for hardware, 
assembly time and compliance programs), and fixed costs (for R&D and 
retooling). Variable costs were applied on a per engine basis and fixed 
costs were amortized at seven percent over five years. Engine 
technology cost estimates were based on the study by ICF and EF&EE in 
October 1996 entitled ``Cost Study for Phase Two Small Engine Emission 
Regulations''. Details of the assumed costs and analysis can be found 
in Chapters 4 and 7 of the Draft RSD.
1. Nonhandheld Engine Costs
    Based on analysis of the EPA Phase 1 certification database, and 
use of the ABT program available to nonhandheld engines, it is assumed 
that four high production Class I SV engine families will need to 
incorporate all those technologies listed in Table 19. Incorporation of 
these technologies will require the engine manufacturer to incur both 
variable and fixed costs.
    Analysis of Class II engine families, from the EPA Phase 1 
certification database and use of the ABT calculation, shows that a 
number of Class II SV engine families will be converted to OHV engine 
design and a large number of OHV engine families will need to 
incorporate emission improvements. Such technologies will require both 
variable and fixed expenditures.
    The proposed Phase 2 emission standards for this diverse industry 
would impact companies differently depending on the existing product 
offerings. Some companies currently manufacture very clean Class II OHV 
engines geared toward the commercial market and would be required to 
make very few changes in their current models. Companies that target 
the consumer market with SV and perhaps less expensive OHV engines 
would require application of the emission reduction technologies.
2. Handheld Engine Costs
    Analysis of the Phase 1 certification database for handheld engines 
shows that nearly all engine families of two stroke design will require 
technologies to reduce engine emissions. Redesign of the existing two-
stroke engine is allocated to fixed costs as companies perform R&D, 
build prototypes and perform numerous emission tests to achieve 
production-ready models.

C. Equipment Costs

    While equipment manufacturers would bear no responsibility for 
meeting emission standards, they may need to make changes in the design 
of their equipment models to accommodate the Phase 2 engines. EPA's 
treatment of the impacts of the proposal therefore includes an analysis 
of costs for equipment manufacturers. The 1996 PSR EOLINK database was 
utilized as the source of information for equipment manufacturers, 
models and sales estimates for all classes. The costs for equipment 
conversion was derived from the ICF/EF&EE cost study 71 and 
improved through the work by ICF and EPA on the small business impact 
analysis. Full details of EPA's cost analysis can be found in Chapter 4 
of the Draft RSD. EPA has assumed that capital costs would be amortized 
at seven percent over ten years.
---------------------------------------------------------------------------

    \71\ ICF and Engine, Fuel and Emissions Engineering, 
Incorporated; ``Cost Study for Phase Two Small Engine Emission 
Regulations'', Draft Final Report, October 25, 1996, in EPA Air 
Docket A-93-29, Item #II-A-04.
---------------------------------------------------------------------------

1. Nonhandheld Equipment Manufacturers
    Based on engine technologies estimated for this rulemaking, it is 
assumed that Class I engine redesign would have no impact on equipment 
manufacturers since the proposed standard would not require external 
changes or adversely impact the engine's performance.
    The Class II engine change from SV to OHV design will have the 
largest impact on equipment changes. Review of the PSR database for 
equipment manufacturers that utilize Class II SV engines reveals that 
the majority (90 percent) of small engine equipment is produced from 32 
companies with the remaining 353 companies representing only 10 percent 
of the overall production.
    EPA's work analyzing small business impacts, as summarized in the 
work with ICF Incorporated,72 indicates that many of the 
small businesses, indicated by the PSR database to use SV Class II 
engines, have already converted or are in the process of converting to 
using OHV engine design due to market forces or changes in their engine 
manufacturer's offerings. These companies tend to produce professional 
or commercial equipment and competition has driven the use of OHV 
engines. The study also revealed that at least one equipment 
manufacturer that produces a large volume of equipment, has already 
switched their lines from SV to OHV. For today's proposal, EPA assumed 
only the one large manufacturer has already incurred the costs of 
converting to the use of OHV engine. For the purpose of this proposal, 
EPA has assumed that any switch from SV to OHV engines by equipment 
manufacturers is a cost incurred due to this proposal. The cost 
estimates were based on equipment application (garden tractor, tiller, 
commercial turf, etc.) and in the case of the commercial turf 
equipment, on the power of the engine within that application. 
Flexibilities within this proposal which may lessen

[[Page 4005]]

the impact of the costs of this rulemaking to equipment manufacturers 
were also not taken into account.
---------------------------------------------------------------------------

    \72\ ``Small Business Impact Analysis of New Emission Standards 
for Small Spark-Ignition Nonroad Engines and Equipment'', ICF 
Incorporated, September 1997, located in EPA Air Docket A-96-55, 
Item#II-A-01 .
---------------------------------------------------------------------------

2. Handheld Equipment Manufacturers
    The majority of technologies assumed in this analysis for handheld 
engines, see Table 19, include only internal redesign and thereby no 
change in the external design of the handheld engine is expected. 
Therefore, it is assumed that the outer dimensions and performance 
characteristics would be similar to the existing models and therefore 
the handheld equipment would not require any changes. Equipment costs 
have been included for manufacturers of augers who will need to 
incorporate changes to the transmission boxes in order to incorporate 
different speed-torque signatures of Phase 2 compliant engines.

D. Operating Costs

    The total life-cycle operating costs for this proposed rulemaking 
include any expected decreases in fuel consumption. Life cycle costs 
have been calculated per class using the nonroad small engine emission 
model. The model calculates fuel savings from the year 2001-2026 and 
takes into account factors including equipment scrappage, projected 
yearly sales increase per equipment type and engine power. Details on 
the assumptions and calculations on fuel savings are included in 
Chapter 4 and 7 of the Draft RSD.
1. Nonhandheld Engines
    No fuel consumption savings have been assumed from Class I engines. 
The addition of oil control piston rings and valve stem seals are not 
expected to affect fuel economy or maintenance requirements and changes 
to carburetion are expected to be only slight. The Class II SV engine 
conversion to OHV design is expected to result in improved fuel economy 
since data show that OHV engines can run at leaner air to fuel ratio's 
than SV engines.
2. Handheld Engines
    Redesigned two-stroke engines are assumed to result in significant 
fuel savings as fuel/oil/air scavenging is significantly reduced.

E. Cost Per Engine and Cost-Effectiveness

1. Cost Per Engine
    Total costs for this proposed rulemaking vary per year as engine 
families are phased-in to compliance with the Phase 2 standards over 
several years, capital costs are recovered and compliance programs are 
conducted. The term ``uniform annualized cost'' is used to express the 
cost of this rulemaking over the years of this analysis.
    The methodology used for estimating the uniform annualized cost per 
engine is as follows. Cost estimates from 1996 and 1997 model years, 
for technology and compliance programs respectively, were estimated and 
increased at an inflation rate of 4 percent per year to the years in 
which they were assumed to be incurred. For engine technology costs, 
one set of technologies per class and engine design was assumed (see 
Table 19). The Phase 1 database was then analyzed to determine the 
number of engine families per class that would likely incorporate the 
emission reduction technologies. The estimated costs per year were then 
calculated by multiplying the number of engine families and 
corresponding production volume by the fixed and variable costs per 
technology grouping, respectively. Retail markups used are 16 percent 
by the engine manufacturer, 5 percent by the equipment manufacturer and 
5 percent by the mass merchandiser. All markups are based on industry 
specific information from Phase 1. For compliance program costs, each 
program was outlined and assigned costs based on the likely number of 
participants or engine families to be included in each program which 
were determined from the Phase 1 certification database. The costs per 
year were discounted seven percent to the first year of Phase 2 
regulation, 2001 for nonhandheld and 2002 for handheld engine classes, 
respectively. A uniform annualized cost was then calculated. Costs per 
engine are calculated from the uniform annualized cost for the first 
full year of implementation of the Phase 2 standard, 2005, and the last 
year of this analysis, 2026. The average cost per engine is calculated 
from these two values and the results are presented in Table 20.
    The yearly fuel savings (tons/yr) per class were calculated from 
the nonroad small engine emission model. The tons/yr were converted to 
savings ($) per year through conversion to gallons per year multiplied 
by $0.765 (a 1995 average refinery price to end user). The yearly fuel 
savings were discounted by 3 percent to the first year of Phase 2 
regulation, 2001 for nonhandheld engines and 2002 for handheld engines. 
The yearly results were totaled and then divided by an annualized 
factor to yield the uniform annualized fuel savings. The engine 
lifetime fuel savings for each engine class was calculated for the 
production years of 2005 and 2026. The average of these two values was 
utilized as the average fuel savings per engine per class is shown in 
Table 20. In particular, EPA notes that its estimate of fuel savings 
for Class II engine conversion to OHV technology is greater than the 
estimated cost of this conversion and thus would be economically 
beneficial to the consumer. EPA requests comment on its analysis of the 
fuel economy benefit for Class II conversion from SV to OHV technology 
and information as to why the market has not responded with a greater 
penetration of the more fuel efficient OHV technology.
    The average resultant cost per engine class is calculated by 
subtracting the average fuel savings from the average cost, see Table 
20. See Chapter 7 of the Draft RSD for more details of this analysis.

                     Table 20.--Engine Life Time Fuel Savings and Resultant Cost Per Engine                     
                                    [Costs based on uniform annualized costs]                                   
----------------------------------------------------------------------------------------------------------------
                                                                     Cost per       Savings per   Resultant cost
                              Class                                   engine          engine        per engine  
----------------------------------------------------------------------------------------------------------------
I...............................................................           $0.87           $0.00          $0.87 
II..............................................................           10.54           33.20        ($22.66)
III.............................................................            0.74            0.45           0.29 
IV..............................................................            1.92            0.99           0.92 
V...............................................................           16.21            4.12          12.07 
----------------------------------------------------------------------------------------------------------------


[[Page 4006]]

2. Cost Effectiveness
    EPA has estimated the cost-effectiveness (i.e., the cost per ton of 
emission reduction) of the proposed HC+NOX standard over the 
typical lifetime of the small SI equipment that would be covered by 
today's proposed rule. EPA has examined the cost-effectiveness by 
performing a nationwide cost-effectiveness in which the net present 
value of the cost of compliance per year is divided by the nationwide 
emission benefits per year over a period of 26 years. This is 
sufficient time to achieve fleet turnover. The resultant cost-
effectiveness is $390 cost/ton HC+NOX without fuel savings. 
Chapter 7 of the Draft RSD contains a more detailed discussion of the 
cost-effectiveness analysis. EPA requests comments on all aspects of 
the cost-effectiveness analysis.
    The overall cost-effectiveness of this rule on HC+NOX 
emission reductions, with fuel savings, is shown in Table 21. Table 21 
contains the cost effectiveness of other nonroad rulemakings, which 
contain fuel savings, to which the cost-effectiveness of this 
rulemaking can be compared.

    Table 21.--Cost-effectiveness of the Proposed Standards With Fuel   
                                 Savings                                
------------------------------------------------------------------------
                                   NPV cost/NPV ton                     
            Standard                  (with fuel          Pollutants    
                                       savings)                         
------------------------------------------------------------------------
Proposed Small SI Engines <19 kW              -$700  HC+NOX             
 Phase 2.                                                               
Small SI Engines <19 kW Phase 1.               $217  HC+NOX             
Spark Ignition Marine Engines...              $1000  HC                 
Proposed Nonroad CI Standards...          $180-$400  HC+NOX             
------------------------------------------------------------------------

VII. Public Participation

A. Comments and the Public Docket

    The Agency welcomes comments on all aspects of this proposed 
rulemaking. All comments (preferably in duplicate), with the exception 
of proprietary information, should be directed to the EPA Air Docket 
Section, Docket No. A-96-02 (see ADDRESSES). Commenters who wish to 
submit proprietary information for consideration should clearly 
separate such information from other comments by:
     Labeling proprietary information ``Confidential Business 
Information'' and,
     Sending proprietary information directly to the contact 
person listed (see FOR FURTHER INFORMATION CONTACT) and not to the 
public docket.
     This will help ensure that proprietary information is not 
inadvertently placed in the docket. If a commenter wants EPA to use a 
submission labeled as confidential business information as part of the 
basis for the final rule, then a nonconfidential version of the 
document, which summarizes the key data or information, should be sent 
to the docket.
    Information covered by a claim of confidentiality will be disclosed 
by EPA only to the extent allowed by and in accordance with the 
procedures set forth in 40 CFR Part 2. If no claim of confidentiality 
accompanies the submission when it is received by EPA, the submission 
may be made available to the public without notifying the commenters.

B. Public Hearing

    Anyone wishing to present testimony about this proposal at the 
public hearing (see DATES) should, if possible, notify the contact 
person (see FOR FURTHER INFORMATION CONTACT) at least two business days 
prior to the day of the hearing. The contact person should be given an 
estimate of the time required for the presentation of testimony and 
notification of any need for audio/visual equipment. A sign-up sheet 
will be available at the registration table the morning of the hearing 
for scheduling those who have not notified the contact earlier. This 
testimony will be scheduled on a first-come, first-served basis, and 
will follow the testimony that is arranged in advance.
    The Agency recommends that approximately 50 copies of the statement 
or material to be presented be brought to the hearing for distribution 
to the audience. In addition, EPA would find it helpful to receive an 
advance copy of any statement or material to be presented at the 
hearing at least two business days before the scheduled hearing date. 
This is to give EPA staff adequate time to review such material before 
the hearing. Advance copies should be submitted to the contact person 
listed.

C. Obtaining Electronic Copies of Documents

    Materials relevant to this proposed rule are contained in Docket 
No. A-96-55, located at the Air Docket, 401 M Street, S.W., Washington, 
DC 20460, and may be reviewed in Room M-1500 from 8:00 a.m. until 5:30 
p.m. Monday through Friday. As provided in 40 CFR part 2, a reasonable 
fee may be charged by EPA for photocopying docket materials.
    The preamble, regulatory language and draft Regulatory Support 
Document are also available electronically from the EPA internet Web 
site. This service is free of charge, except for any cost you already 
incur for internet connectivity. The text of the proposed rule is made 
available on the day of publication on the primary Web site listed 
below. The EPA Office of Mobile Sources also publishes these notices on 
the secondary Web site listed below.

Internet (Web)

http://www.epa.gov/docs/fedrgstr/EPA-AIR/
(Either select desired data or use search feature)

http://www.epa.gov/OMSWWW/
(Look in What's New or under the specific rulemaking topic)

    Please note that due to differences between the software used to 
develop the document and the software into which the document may be 
downloaded, changes in format, page length, etc., may occur.

VIII. Administrative Requirements

A. Administrative Designation and Regulatory Analysis

    Under Executive Order 12866, 73 the Agency must 
determine whether the regulatory action is ``significant'' and 
therefore subject to OMB review and the requirements of the Executive 
Order. The order defines ``significant regulatory action'' as one that 
is likely to result in a rule that may:
---------------------------------------------------------------------------

    \73\ 58 FR 51735 (October 4, 1993).
---------------------------------------------------------------------------

    (1) Have an annual effect on the economy of $100 million or more or 
adversely affect in a material way the economy, a sector of the 
economy, productivity, competition, jobs, the environment, public 
health or safety, or State, local or tribal governments or communities;
    (2) Create a serious inconsistency or otherwise interfere with an 
action taken or planned by another agency;

[[Page 4007]]

    (3) Materially alter the budgetary impact of entitlement, grants, 
user fees, or loan programs or the rights and obligations of recipients 
thereof;
    (4) Raise novel legal or policy issues arising out of legal 
mandates, the President's priorities, or the principles set forth in 
the Executive Order.
    A regulatory support document which presents EPA's analysis of the 
cost impacts of this proposed rule is available for review in the 
public docket. EPA estimates that the proposed standards and other 
regulatory provisions, if adopted, would not have an annual effect on 
the economy of more than $100 million, a criterion which is a major 
determinant in defining an ``economically significant regulatory 
action.'' Although not ``significant'' based on this criterion, the 
rule may adversely affect in a material way that sector of the economy 
involved with the production of small spark-ignition engines or 
equipment utilizing such engines. As such, this action was submitted to 
OMB for review. Any written comments from OMB and any EPA response to 
OMB comments are in the public docket for this proposal.

B. Paperwork Reduction Act

    The information collection requirements in this proposed rule have 
been submitted for approval to the Office of Management and Budget 
(OMB) under the Paperwork Reduction Act, 44 U.S.C. 3501 et seq. Copies 
of the ICR document may be obtained from Sandy Farmer, Regulatory 
Information Division, EPA, 401 M Street, SW (2137), Washington, DC 
20460 or by calling (202) 260-2740.
    Table 22 provides a listing of this proposed rulemaking's 
information collection requirements along with the appropriate 
information collection request (ICR) numbers. The cost of this burden 
has been incorporated into the cost estimate for this rule. The Agency 
has estimated that the public reporting burden for the collection of 
information required under this rule would average approximately 6702 
hours annually for a typical engine manufacturer. The hours spent by a 
manufacturer on information collection activities in any given year 
would be highly dependent upon manufacturer specific variables, such as 
the number of engine families, production changes, emission defects 
etc.

                   Table 22.--Public Reporting Burden                   
------------------------------------------------------------------------
                                                            OMB control 
          EPA ICR No.              Type of information          No.     
------------------------------------------------------------------------
151490.........................  Certification..........       2060-0338
23420..........................  Averaging, banking and        2060-0338
                                  trading.                              
N/A............................  Production line testing             N/A
1675.01........................  In-use testing.........       2060-0292
N/A............................  In-use credits.........             N/A
0095.07........................  Pre-certification and         2060-0007
                                  testing exemption.                    
0012...........................  Engine exclusion              2060-0124
                                  determination.                        
0282...........................  Emission defect               2060-0048
                                  information.                          
1673.01........................  Importation of                2060-0294
                                  nonconforming engines.                
------------------------------------------------------------------------

    Send comments regarding the burden estimate or any other aspect of 
this collection of information, including suggestions for reducing this 
burden to Chief, Information Policy Branch, EPA, 401 M Street, SW (PM-
223Y), Washington DC 20460; and to the Office of Information and 
Regulatory Affairs, Office of Management and Budget, Washington, DC 
20503, marked ``Attention: Desk Officer for EPA.'' The final rule will 
contain responses to OMB or public comments on the information 
collection requirements contained in this proposal.

C. Unfunded Mandates Reform Act

    Section 202 of the Unfunded Mandates Reform Act of 1995 (``Unfunded 
Mandates Act'') requires that the Agency prepare a budgetary impact 
statement before promulgating a rule that includes a Federal mandate 
that may result in expenditure by State, local, and tribal governments, 
in aggregate, or by the private sector, of $100 million or more in any 
one year. Section 203 requires the Agency to establish a plan for 
obtaining input from and informing, educating, and advising and small 
governments that may be significantly or uniquely affected by the rule.
    Under section 205 of the Unfunded Mandates Act, the Agency must 
identify and consider a reasonable number of regulatory alternatives 
before promulgating a rule for which a budgetary impact statement must 
be prepared. The Agency must select from those alternatives the least 
costly, most cost-effective, or least burdensome alternative that 
achieves the objectives of the rule, unless the Agency explains why 
this alternative is not selected or the selection of this alternative 
is inconsistent with law.
    Because this proposed rule is estimated to result in the 
expenditure by State, local and tribal governments or the private 
sector of less than $100 million in any one year, the Agency has not 
prepared a budgetary impact statement or specifically addressed the 
selection of the least costly, most cost-effective or least burdensome 
alternative. EPA has estimated the rule to cost the private sector an 
annualized cost of $90 million per year . However, the Agency has 
appropriately considered cost issues in developing this proposal as 
required by section 213(a)(3) of the Clean Air Act, and has designed 
the proposed rule such that it will in EPA's view be a cost-effective 
program. Because small governments would not be significantly or 
uniquely affected by this proposed rule, the Agency is not required to 
develop a plan with regard to small governments.

D. Regulatory Flexibility

    The Regulatory Flexibility Act (RFA) generally requires an agency 
to conduct a regulatory flexibility analysis of any rule subject to 
notice and comment rulemaking requirements unless the agency certifies 
that the rule will not have a significant economic impact on a 
substantial number of small entities. Small entities include small 
businesses, small not-for-profit enterprises, and small governmental 
jurisdictions. For the reasons set out below, this proposed rule would 
not have a significant impact on a substantial number of small 
entities.
    EPA has identified industries that would be subject to this 
proposed rule and has contacted small entities and small entity 
representatives to gain a better understanding of potential impacts of 
the proposed Phase 2 program on their businesses. This

[[Page 4008]]

information was useful in estimating potential impacts of this rule on 
affected small entities, the details of which are fully discussed in 
Chapter 8 of the Draft RSD. Small not-for-profit organizations and 
small governmental jurisdictions are not expected to be impacted by 
this proposal. Thus EPA's impact analysis focuses on small businesses. 
For purposes of the impact analysis, ``small business'' is defined by 
number of employees or dollars of annual receipts according to Small 
Business Administration (SBA) regulations. The analysis focuses 
especially on impacts to manufacturers of Class II nonhandheld and 
Classes III-V handheld engines and equipment, since Class I side-valve 
engines are only expected to need minor modifications.
    The economic impact of the proposed rule on engine and equipment 
manufacturers defined as small by the SBA was evaluated using a ``sales 
test'' approach which calculates annualized compliance costs as a 
function of sales revenue. The ratio is an indication of the severity 
of the potential impacts. The results of the analysis suggest that of 
those small entities analyzed, one small business engine manufacturer 
and two small business equipment manufacturers would experience an 
impact of greater than one percent of their sales revenue. However, 
none of these small entities would experience an impact greater than 
three percent of their sales revenue. These three companies represent 
approximately five percent of the total small business manufacturers on 
which the analysis was based. Given this, and the ratio levels at which 
these companies are projected to be impacted (i.e., less than three 
percent), EPA expects today's proposal to have a light impact on small 
business entities. The analysis assumes no passthrough of costs in 
price increases and thus can be characterized as depicting worst case 
impacts.
    While the Agency does not consider these impacts to be significant, 
the Agency desires to minimize impacts to the extent possible for those 
companies which may be adversely affected and to ensure that the 
proposed emissions standards are achievable. Thus, flexibility 
provisions for the proposed rule (discussed in Section IV.E) were 
developed based on information gained through discussions with 
potentially affected small entities. Many of the flexibilities being 
proposed in today's rule should benefit both engine and equipment 
manufacturers qualifying as small. Some, but not all, of these 
provisions were considered in the impact assessment on small entities 
(see Chapter 8 of the Draft RSD). Those flexibilities not considered, 
including a hardship relief provision described in Section IV.E, were 
developed too late in the rule development process to be included in 
the impact assessment, but as they were added in order to further 
ensure the achievability of the proposed standards it is expected that 
they would further reduce the impacts of the proposed rule. EPA 
requests comment as to whether these proposed provisions adequately 
address the needs of affected manufacturers, and small entities in 
particular.
    The results of the impact analysis show minimal impacts on small 
businesses. EPA expects impacts may be negligible if small companies 
take advantage of those additional flexibilities not considered in the 
analysis, and if companies pass through most of their costs to 
customers as was indicated as likely by most small companies contacted. 
Furthermore, EPA's outreach activities with small entities indicated 
that many engine and equipment manufacturers have already made the 
switch from side-valve engine technology to producing or using overhead 
valve engine technology for reasons other than today's proposed rule, 
and therefore may not incur substantial additional costs as a result of 
this program. Therefore, I certify that this action will not have a 
significant economic impact on a substantial number of small entities 
and therefore a regulatory flexibility analysis for this proposal has 
not been prepared. The Agency continues to be interested in the 
potential impacts of the proposed rule on small entities and welcomes 
additional comments during the rulemaking process on issues related to 
such impacts. In spite of the expected minimal impacts on small 
entities, the Agency is continuing its efforts to notify other small 
business engine and equipment manufacturers of this rule and inform 
them of their opportunities for providing feedback to the Agency.

List of Subjects in 40 CFR Part 90

    Environmental protection, Administrative practice and procedure, 
Air pollution control, Confidential business information, Imports, 
Labeling, Nonroad source pollution, Reporting and recordkeeping 
requirements, Research, Warranties.

    Dated: December 23, 1997.
Carol M. Browner,
Administrator.

    For the reasons set out in the preamble, title 40, chapter I of the 
Code of Federal Regulations is proposed to be amended as follows:

PART 90--CONTROL OF EMISSIONS FROM NONROAD SPARK-IGNITION ENGINES

    1. The authority citation for part 90 continues to read as follows:

    Authority: Sections 203, 204, 205, 206, 207, 208, 209, 213, 215, 
216, and 301(a) of the Clean Air Act, as amended (42 U.S.C. 7522, 
7523, 7524, 7525, 7541, 7542, 7543, 7547, 7549, 7550, and 7601(a)).

Subpart A--General

    2. Section 90.1 is amended by removing the period at the end of 
paragraph (b)(5)(iv) and adding a semicolon in its place, by adding 
paragraphs (b)(6) and (d) and by revising paragraph (c) to read as 
follows:


Sec. 90.1  Applicability.

* * * * *
    (b) * * *
    (6) Engines that are used exclusively in emergency and rescue 
equipment where no certified engines are available to power the 
equipment safely and practically, but not including generators, 
alternators, compressors or pumps used to provide remote power to a 
rescue tool. The equipment manufacturer bears the responsibility to 
ascertain on an annual basis and maintain documentation available to 
the Administrator that no appropriate certified engine is available 
from any source.
    (c) Engines subject to the provisions of this subpart are also 
subject to the provisions found in subparts B through N of this part, 
except that subparts C, H, M and N of this part apply only to Phase 2 
engines as defined in this subpart.
    (d) Certain text in this part is identified as pertaining to Phase 
1 or Phase 2 engines. Such text pertains only to engines of the 
specified Phase. If no indication of Phase is given, the text pertains 
to all engines, regardless of Phase.
    3. Section 90.3 is amended by adding the following definitions in 
alphabetical order to read as follows:


Sec. 90.3  Definitions.

* * * * *
    Aftertreatment means the passage of exhaust gases through a device 
or system such as a catalyst whose purpose is to chemically alter the 
gases prior to their release to the atmosphere.
* * * * *
    Commercial Engine means a handheld engine that is not a residential 
engine.
    DF or df means deterioration factor.
    Eligible sales or U.S. sales means Phase 2 engines sold for 
purposes of being used in the United States, and includes any engine 
exported and

[[Page 4009]]

subsequently imported in a new piece of equipment, but excludes any 
engine introduced into commerce, by itself or in a piece of equipment, 
for use in a state that has established its own emission requirements 
applicable to such engines pursuant to a waiver granted by EPA under 
section 209(e) of the Clean Air Act.
* * * * *
    Family Emission Limit or FEL means an emission level that is 
declared by the manufacturer to serve in lieu of an emission standard 
for certification, production line testing, Selective Enforcement 
Auditing, and in-use testing for engines participating in the 
averaging, banking and trading program. An FEL must be expressed to the 
same number of decimal places as the applicable emission standard.
* * * * *
    HC+NOX means total hydrocarbons plus oxides of nitrogen.
    In-use credit means an emission credit that represents the 
difference between the mean in-use emission results of a regulated 
pollutant, CO, HC+NOX or NMHC+NOX, and the 
applicable certification emission standard. In-use results below the 
standard lead to the calculation of positive in-use credits, while in-
use results above the standard lead to the calculation of negative in-
use credits.
* * * * *
    NMHC+NOX means nonmethane hydrocarbons plus oxides of 
nitrogen.
* * * * *
    Overhead valve engine means an otto-cycle, four-stroke engine in 
which the intake and exhaust valves are located above the combustion 
chamber within the cylinder head. Such engines are sometimes referred 
to as ``valve-in-head'' engines.
    Overhead valve emission performance or OEP engine means a Class II 
overhead valve engine, or a Class II non-overhead valve engine that 
complies with the applicable 2005 model year emission standards without 
using emission credits.
    Phase 1 engine means any handheld or nonhandheld engine, that was 
produced under a certificate of conformity issued under the regulations 
in this part and that is not a Phase 2 engine.
    Phase 2 engine means any handheld engine as defined in this subpart 
that is subject to the standards that begin to phase-in in the 2002 
model year; and any nonhandheld engine as defined in this subpart of 
the 2001 model year or later including those 1999 and 2000 model year 
engines certified under early banking provisions described in this 
part. Any engines exempted from the Phase 2 standards under this part 
are excluded from coverage under this definition.
* * * * *
    Residential engine means a handheld engine for which the engine 
manufacturer makes a written statement to EPA as part of its 
certification application that such engine and the equipment it is 
installed in by the engine manufacturer, where applicable, is not 
produced, advertised, marketed or intended for commercial or 
professional usage.
    Round, rounded or rounding means, unless otherwise specified, that 
numbers will be rounded according to ASTM-E29-93a, which is 
incorporated by reference in this part pursuant to Sec. 90.7.
* * * * *
    Side valve engine means an otto-cycle, four stroke engine in which 
the intake and exhaust valves are located to the side of the cylinder, 
not within the cylinder head. Such engines are sometimes referred to as 
``L-head'' engines.
    Small volume engine family means any handheld engine family whose 
eligible sales in a given model year are projected at the time of 
certification to be no more than 2,500 engines; or any nonhandheld 
engine family whose eligible sales in a given model year are projected 
at the time of certification to be no more than 1,000 units.
    Small volume engine manufacturer means, for handheld engines, any 
engine manufacturer whose total eligible sales of handheld engines 
subject to regulation under this part are projected at the time of 
certification of a given model year to be no more than 25,000 handheld 
engines; and, for nonhandheld engines, any engine manufacturer whose 
total eligible sales of nonhandheld engines are projected at the time 
of certification of a given model year to be no more than 10,000 
nonhandheld engines.
    Small volume equipment manufacturer means, for handheld equipment, 
any equipment manufacturer whose production of handheld equipment 
subject to regulation under this part or powered by engines regulated 
under this part, does not exceed 5000 pieces for a given model year or 
annual production period excluding that equipment intended for 
introduction into commerce for use in a state that has established its 
own emission requirements applicable to such equipment or engines in 
such equipment, pursuant to a waiver granted by EPA under section 
209(e) of the Clean Air Act. For nonhandheld equipment, the term 
``small volume equipment manufacturer'' has the same meaning except 
that it is limited to 2500 pieces rather than 5000.
    Small volume equipment model means, for handheld equipment, any 
unique model of equipment whose production subject to regulations under 
this part or powered by engines regulated under this part, does not 
exceed 2500 pieces for a given model year or annual production period 
excluding that equipment intended for introduction into commerce for 
use in a state that has established its own emission requirements 
applicable to such equipment or engines in such equipment, pursuant to 
a waiver granted by EPA under section 209(e) of the Clean Air Act. For 
nonhandheld equipment, the term ``small volume equipment model'' has 
the same meaning except that it is limited to 500 pieces rather than 
2500.
    Technology subgroup means a group of engine families from one or 
more manufacturers having similar size, application, useful life and 
emission control equipment; e.g., Class III, residential, non-catalyst, 
two stroke engine used in generator set applications.
* * * * *

Subpart B--Emission Standards and Certification Provisions

    4. Section 90.103 is amended by revising paragraph (a) introductory 
text, and paragraphs (a)(3) and (a)(5) and by adding paragraphs (a)(6) 
through (a)(9) to read as follows:


Sec. 90.103  Exhaust emission standards.

    (a) Exhaust emissions for new Phase 1 and Phase 2 nonroad spark 
ignition engines at or below 19 kilowatts (kW), shall not exceed the 
following levels. Throughout this part, NMHC+NOX standards 
are applicable only to natural gas fueled engines at the option of the 
manufacturer, in lieu of HC+NOX standards. The tables for 
Phase 1 and Phase 2 exhaust emissions levels follow:

[[Page 4010]]



                                  Table 1.--Phase 1 Exhaust Emission Standards                                  
                                            [Grams per kilowatt-hour]                                           
----------------------------------------------------------------------------------------------------------------
                                                  Hydrocarbons +                                                
                                                     oxides of                        Carbon         Oxides of  
            Engine displacement class                nitrogen      Hydrocarbons      monoxide     nitrogen (NOX)
                                                     (HC+NOX)                                                   
----------------------------------------------------------------------------------------------------------------
I...............................................            16.1  ..............             519  ..............
II..............................................            13.4  ..............             519  ..............
III.............................................  ..............             295             805            5.36
IV..............................................  ..............             241             805            5.36
V...............................................  ..............             161             603            5.36
----------------------------------------------------------------------------------------------------------------


                     Table 2.--Phase 2 Nonhandheld Exhaust Emission Standards by Model Year                     
                                            [Grams per kilowatt-hour]                                           
----------------------------------------------------------------------------------------------------------------
                                                                            Model year                          
                                                ----------------------------------------------------------------
     Engine class        Emission requirement                                                          2005 and 
                                                     2001         2002         2003         2004        later   
----------------------------------------------------------------------------------------------------------------
I....................  HC+ NOX.................         25.0         25.0         25.0         25.0         25.0
                       NMHC+NOX................         23.0         23.0         23.0         23.0         23.0
                       CO......................          610          610          610          610          610
II...................  HC+NOX..................         18.0         16.6         15.0         13.6         12.1
                       NMHC+NOX................         16.7         15.3         14.0         12.7         11.3
                       CO......................          610          610          610          610          610
                       Assumed OEP Percentage..           50         62.5           75         87.5          100
----------------------------------------------------------------------------------------------------------------


     Table 3.--Phase 2 Handheld Exhaust Emission Standards Showing Phase-In by Aggregate Percentage of Sales    
                                            [Grams per kilowatt-hour]                                           
----------------------------------------------------------------------------------------------------------------
                                        Emission standard                         Model year                    
                                   -----------------------------------------------------------------------------
           Engine class                                                                                2005 and 
                                       HC+NOX         CO          2002         2003         2004        later   
                                                               (percent)    (percent)    (percent)    (percent) 
----------------------------------------------------------------------------------------------------------------
III...............................          210          805  ...........  ...........  ...........  ...........
IV................................          172          805           20           40           70          100
V.................................          116          603  ...........  ...........  ...........  ...........
----------------------------------------------------------------------------------------------------------------

* * * * *
    (3) Notwithstanding paragraph (a)(2) of this section, two stroke 
engines used to power lawnmowers or other nonhandheld equipment may 
meet Phase 1 Class III, IV or V standards and requirements, as 
appropriate, through model year 2002 subject to the provisions of 
Sec. 90.107(e), (f) and (h). Such engines shall not be included in any 
computations of Phase 2 nonhandheld credits or sales nor in any 
computations used to ascertain compliance with Phase 2 phase-in 
requirements for handheld engines.
* * * * *
    (5) Notwithstanding paragraph (a)(2) of this section, engines used 
exclusively to power products which are used exclusively in wintertime, 
such as snowthrowers and ice augers, at the option of the engine 
manufacturer, need not certify to or comply with standards regulating 
emissions of HC, NOX, HC+NOX or 
NMHC+NOX , as applicable. If the manufacturer exercises the 
option to certify to standards regulating such emissions, such engines 
must meet such standards. If the engine is to be used in any equipment 
or vehicle other than an exclusively wintertime product such as a 
snowthrower or ice auger, it must be certified to the applicable 
standard regulating emissions of HC, NOX, HC+NOX 
or NMHC+NOX as applicable.
    (6) During the phase-in of Phase 2 emission requirements for 
handheld engines, as applicable, those engine families not certified to 
Phase 2 requirements shall be certified to and shall meet Phase 1 
requirements.
    (7) Manufacturers of Phase 2 Class II engines must comply with the 
OEP percentages shown in Table 2 of this section in each model year in 
cases where the manufacturer desires to engage in cross class averaging 
of emission credits as permitted under subpart C of this part, and in 
cases where the manufacturer desires to use credits banked by itself or 
another manufacturer in the 1999 or 2000 model year as permitted under 
subpart C of this part. Compliance with OEP percentages shall be 
determined by dividing the manufacturer's eligible sales of Class II 
engines that are overhead valve engines or are certified at or below 
the 2005 HC+NOX (NMHC+NOX) standard, by the 
manufacturer's total eligible sales of Class II engines for the subject 
model year. Side valve engine families with annual US sales of less 
than 1000 may be excluded from the calculation.
    (8) Notwithstanding the standards shown in Table 2 of this section, 
the HC+NOX (NMHC+NOX) standard for Phase 2 Class 
II sidevalve engine families with annual production of 1000 or less 
shall be 24.0 g/kW-hr (22.0 g/kW-hr) for model years 2005 and later. 
Engines produced subject to this provision may not exceed this standard 
and are excluded from the averaging,

[[Page 4011]]

banking and trading program and any related credit calculations after 
the 2004 model year. During the 2001 through 2004 model years these 
engines are subject to applicable Phase 2 standards, but shall not 
require the application of certification credits if their 
HC+NOX (NMHC+NOX) certification level is 24.0 g/
kW-hr (22.0 g/kW-hr) or less.
    (9) Notwithstanding the standards shown in Table 2 of this section, 
small volume engine manufacturers as defined in this part may, at their 
option, certify Phase 2 Class II engines to an HC+NOX 
(NMHC+NOX) standard of 24.0 g/kW-hr (22.0 g/kW-hr) through 
the 2004 model year. Such engines shall not exceed this standard and 
are excluded from the averaging, banking and trading program through 
the 2004 model year.
* * * * *
    5-6. Section 90.104 is amended by adding introductory text and 
paragraphs (d) through (i) to read as follows:


Sec. 90.104  Compliance with emission standards.

    Paragraphs (a) through (c) of this section apply to Phase 1 engines 
only. Paragraphs (d) through (i) of this section apply only to Phase 2 
engines.
* * * * *
    (d) The exhaust emission standards (FELs, where applicable) for 
Phase 2 engines set forth in this part apply to the emissions of the 
engines for their full useful lives as determined pursuant to 
Sec. 90.105.
    (e) For all Phase 2 engines:
    (1) If all test engines representing an engine family have 
emissions, when properly tested according to procedures in this part, 
less than or equal to each Phase 2 emission standard (FEL, where 
applicable) in a given engine displacement class and given model year, 
when multiplicatively adjusted by the deterioration factor determined 
in this section, that family complies with that class of emission 
standards for purposes of certification. If any test engine 
representing an engine family has emissions adjusted multiplicatively 
by the deterioration factor determined in this section, greater than 
any one emission standard (FEL, where applicable) for a given 
displacement class, that family does not comply with that class of 
emission standards.
    (2) Except as otherwise permitted under this section, each 
manufacturer of handheld engines must comply with the Phase 2 phase-in 
schedule shown in Sec. 90.103. Compliance with the Phase 2 phase-in 
schedule shall be determined each model year by dividing the 
manufacturer's total eligible sales of Phase 2 handheld engines of that 
model year by the manufacturer's total eligible sales of handheld 
engines subject to regulation under this part.
    (3) In each model year during the Phase 2 phase-in period for 
handheld engines (i.e. model years 2002, 2003, and 2004), manufacturers 
of handheld engines shall project, updating as appropriate, and make 
available to the Administrator upon request, the sales figures 
necessary to complete the calculation required in paragraph (e)(2) of 
this section. Within 270 days after the end of each model year in the 
Phase 2 phase-in period, each manufacturer shall submit a report to the 
Administrator showing its calculation of compliance with the phase-in 
schedule.
    (4) Small volume manufacturers of handheld engines as defined in 
this part are not subject to the phase-in requirements applicable to 
the 2002, 2003 or 2004 model years.
    (f) Each manufacturer of nonhandheld engines must comply with all 
provisions of the averaging, banking and trading program outlined in 
subpart C of this part for each engine family participating in that 
program.
    (g)(1) Deterioration factors for HC+NOX and 
NMHC+NOX emissions for all nonhandheld OHV Phase 2 engines 
without aftertreatment may be taken from Table 1 of this section or may 
be calculated according to the process described in paragraph (h) of 
this section. Except where the Administrator directs a nonhandheld 
engine manufacturer to calculate a df under paragraph (g)(2) or (g)(3) 
of this section, if a manufacturer elects to calculate a df for an 
engine family, it must do so for all families of that class in the same 
useful life category. Where a manufacturer elects to take an 
HC+NOX or NMHC+NOX df from the table, it may use 
good engineering judgment to determine an appropriate CO df, provided 
it maintains and makes available to the Administrator upon request, 
such rationale and supporting data used to determine the CO df.
    (2) If the Administrator has evidence for a given class and useful 
life category indicating that a sales weighted average of a 
manufacturer's actual dfs of those families for which an assigned df is 
being used, exceeds the assigned df by more than 15%, the Administrator 
may require the manufacturer to submit appropriate data to establish a 
df for some or all of the engine families. Such data may be generated 
through the process described in paragraph (h) of this section or 
through another process approved by the Administrator.
    (3) If the Administrator has evidence indicating that the actual df 
of an engine family for which a manufacturer is using an assigned df, 
exceeds 1.8, the Administrator may require the manufacturer to submit 
appropriate data to establish a df for that engine family. Such data 
may be generated through the process described in paragraph (h) of this 
section or through another process approved by the Administrator.
    (4) Table 1 follows:

   Table 1.--Assigned HC+NOx and NMHC+NOx Deterioration Factors for Nonhandheld Phase 2 Overhead Valve Engines  
                                             Without Aftertreatment                                             
----------------------------------------------------------------------------------------------------------------
                                                                                                           `    
----------------------------------------------------------------------------------------------------------------
Class I...........................  Usefule life (hours).................           66          250          500
                                    Deterioration factor.................          1.3          1.3          1.3
Class II..........................  Useful life (hours)..................          250          500         1000
                                    Deterioration factor.................          1.3          1.3          1.3
----------------------------------------------------------------------------------------------------------------

    (h) Manufacturers shall obtain an assigned df or calculate a df, as 
appropriate, for each regulated pollutant for all Phase 2 handheld and 
nonhandheld engine families. Such dfs shall be used, as applicable, for 
certification, production line testing, and Selective Enforcement 
Auditing. For handheld engines, and nonhandheld engines not using 
assigned dfs from Table 1 of this section, manufacturers shall 
calculate dfs for each pollutant through one of the following options:
    (1) For handheld engines, dfs shall be determined using good 
engineering judgment and reflect the exhaust emission deterioration 
expected over the useful life of the engine except that no df may be 
less than 1.0. EPA may reject a df if it has evidence that the df is 
not appropriate for that family. The manufacturer must retain actual 
emission test data to support its choice of df and furnish that data to 
the Administrator upon request. Acceptable

[[Page 4012]]

data sources include, but are not limited to:
    (i) In-use data from an earlier model year of this family or a 
closely related family;
    (ii) Data from engines used in the field/bench adjustment program 
described in subpart M of this part.
    (2) For nonhandheld engines:
    (i) On at least three test engines representing the configuration 
chosen to be the most likely to exceed HC+NOX 
(NMHC+NOX) emission standards, (FELs where applicable), and 
constructed to be representative of production engines pursuant to 
Sec. 90.117, conduct full Federal test procedure emission testing 
pursuant to the regulations of Subpart E of this part at the number of 
hours representing stabilized emissions pursuant to Sec. 90.118. 
Average the results and round to the same number of decimal places 
contained in the applicable standard, expressed to one additional 
significant figure. Conduct such emission testing again following field 
aging in actual usage to a number of hours equivalent to the applicable 
useful life hours, plus or minus five percent. Average the results and 
round to the same number of decimal places contained in the applicable 
standard, expressed to one additional significant figure. Divide the 
full useful life average emissions for each regulated pollutant by the 
stabilized average emission results and round to two significant 
figures. The resulting number shall be the df, unless it is less than 
1.0, in which case the df shall be 1.0; or
    (ii) On at least three test engines representing the configuration 
chosen to be the most likely to exceed HC+NOX 
(NMHC+NOX) emission standards (FELs where applicable), and 
constructed to be representative of production engines pursuant to 
Sec. 90.117, conduct full Federal test procedure emission testing 
pursuant to the regulation of Subpart E of this part at no fewer than 
three points as follows: at the number of hours representing stabilized 
emissions pursuant to Sec. 90.118; again following field aging in 
actual usage to a number of hours equivalent to the applicable useful 
life hours, plus or minus five percent; and also at no fewer than one 
point spaced approximately equally between the other two. The test 
results for each pollutant shall be rounded to the same number of 
decimal places contained in the applicable standard, expressed to one 
additional significant figure and plotted as a function of hours on the 
engine, rounded to the nearest whole hour. The best fit straight line, 
determined by the method of least squares, shall be drawn. Using this 
line, interpolate the emissions of each pollutant at 12 hours and at a 
number of hours equal to the applicable useful life. Divide the 
interpolated useful life emissions by the interpolated emissions at 12 
hours and round this figure to two significant figures. The resultant 
number shall represent the df unless it is less than 1.0, in which case 
the df shall be 1.0; or
    (iii) Perform another process, approved in advance by the 
Administrator, which will have the objective of adequately ascertaining 
the relationship of field aged emissions at full useful life with those 
tested with stabilized emissions at low hours; or
    (iv) For manufacturers of Class II overhead valve engines 
certifying to 500 or 1000 hour useful lives, such manufacturers may 
establish dfs for such engines based on good engineering judgment that 
has been proposed in advance and determined to be satisfactory to the 
Administrator, for certification of model years 2001 through 2004. The 
Administrator may, in model year 2006 or later, direct the manufacturer 
to verify, in a period of time the Administrator determines to be 
reasonable, such dfs using methods described in paragraphs (h)(2)(i), 
(ii) or (iii) of this section. If the dfs established by the 
manufacturer under this provision underestimate the dfs determined by 
the methods under paragraphs (h)(2)(i), (ii) or (iii) of this section, 
by 15% or more, the Administrator shall provide the manufacturer with a 
period of two model years in which to obtain sufficient certification 
emission credits from other nonhandheld engines to cover the credit 
shortfall calculated by substituting the df determined under this 
provision for the original df in the equation in Sec. 90.207(a).
    (3) Calculated deterioration factors may cover families and model 
years in addition to the one upon which they were generated if the 
manufacturer submits a justification acceptable to the Administrator in 
advance of certification that the affected engine families can be 
reasonably expected to have similar emission deterioration 
characteristics.
    (i)(1) Except as allowed in paragraph (i)(2) of this section, 
nonhandheld sidevalve engines or nonhandheld engines with exhaust 
aftertreatment shall be certified by field aging one engine in actual 
usage or by bench aging one engine on an aging cycle determined to 
represent field aged engines under Sec. 90.1207 and Sec. 90.1208, to 
its full useful life followed by emission testing using applicable test 
procedures under this part. Emission test results for such bench aged 
engines shall be adjusted using adjustment factors calculated under 
Sec. 90.1208 to determine the certification levels. The dfs for such 
engines shall be calculated during this bench aging process using the 
techniques described in paragraphs (h)(2)(i), (ii) or (iii) of this 
section, except that bench aging of one engine may be used in place of 
field aging. In calculating the dfs of bench aged nonhandheld sidevalve 
engines or nonhandheld engines with aftertreatment, the emission test 
data at the number of hours equal to full useful life, shall first be 
multiplied by the adjustment factor applicable to that engine family 
and determined under Sec. 90.1208.
    (2) Sidevalve Class II or aftertreatment-equipped Class II engines 
for which the manufacturer commits in writing, at the time of 
certification, to cease production by the end of the 2004 model year, 
are eligible for reduced certification testing, at the manufacturer's 
option. Bench aging or field aging for the certification of such 
engines may be stopped at 120 hours for engines having a useful life of 
250 hours as determined pursuant to regulations in this part; at 250 
hours for engines having a useful life of 500 hours; and at 500 hours 
for engines having a useful life of 1000 hours. In such cases, based on 
emission results from stabilized engines and engines aged as described 
in this paragraph (i), the manufacturer shall project emissions to 250, 
500 or 1000 hours, as applicable, using good engineering judgment 
acceptable to the Administrator. The manufacturer shall then adjust 
bench aged emissions (if applicable) with the adjustment factor 
determined pursuant to Sec. 90.1208 for purposes of certification and 
computation of credits or credit needs. The manufacturer shall compute 
dfs for bench aged engines from the adjusted emission levels using good 
engineering judgment acceptable to the Administrator. For field aged 
engines, the manufacturer shall compute dfs from the projected 250, 500 
or 1000 hour emissions, as applicable, using good engineering judgment 
acceptable to the Administrator.
    7. Section 90.105 is revised to read as follows:


Sec. 90.105  Useful life periods for Phase 2 engines.

    (a) Manufacturers shall declare the applicable useful life category 
for each engine family at the time of certification as described in 
this section. Unless otherwise approved by the Administrator, such 
category shall be that category which most closely

[[Page 4013]]

approximates the actual useful lives of the equipment into which the 
engines are expected to be installed. Manufacturers shall retain data 
appropriate to support their choice of useful life category for each 
engine family. Such data shall be sufficient to show that the majority 
of engines or a sales weighted average of engines of that family are 
used in applications having a useful life best represented by the 
chosen category. Such data shall be furnished to the Administrator upon 
request.
    (1) For handheld engines:
    (i) Engines declared by the manufacturer at the time of 
certification as residential, as defined in Sec. 90.3, shall have a 
useful life for purposes of regulation under this part of 50 hours.
    (ii) Engines declared by the manufacturer at the time of 
certification as commercial, as defined in Sec. 90.3, shall have a 
useful life for purposes of regulation under this part of 300 hours.
    (2) For nonhandheld engines: Manufacturers shall select a useful 
life category from Table 1 of this section at the time of 
certification, as follows:

    Table 1.--Useful Life Categories for Nonhandheld Engines (Hours)    
------------------------------------------------------------------------
                                          Category   Category   Category
                                             C          B          A    
------------------------------------------------------------------------
Class I................................         66        250        500
Class II...............................        250        500       1000
------------------------------------------------------------------------

    (3) Data to support a manufacturer's choice of useful life 
category, for a given engine family, may include but are not limited 
to:
    (i) Surveys of the life spans of the equipment in which the subject 
engines are installed;
    (ii) Engineering evaluations of field aged engines to ascertain 
when engine performance deteriorates to the point where usefulness and/
or reliability is impacted to a degree sufficient to necessitate 
overhaul or replacement;
    (iii) Warranty statements and warranty periods;
    (iv) Marketing materials regarding engine life;
    (v) Failure reports from engine customers; and
    (vi) Engineering evaluations of the durability, in hours, of 
specific engine technologies, engine materials or engine designs.
    (b) [Reserved]
    8. Section 90.106 is amended by revising paragraph (a) and adding 
new paragraph (b)(3) to read as follows:


Sec. 90.106  Certificate of conformity.

    (a)(1) Except as provided in Sec. 90.2(b), every manufacturer of 
new engines produced during or after model year 1997 must obtain a 
certificate of conformity covering such engines; however, engines 
manufactured during an annual production period beginning prior to 
September 1, 1996 are not required to be certified.
    (2) Except as required in paragraph (b)(3) of this section, 
nonhandheld engines manufactured during an annual production period 
beginning prior to September 1, 2000 are not required to meet Phase 2 
requirements.
    (b) * * *
    (3) Manufacturers who commence an annual production period for a 
nonhandheld engine family between January 1, 2000 and September 1, 2000 
must meet Phase 2 requirements for that family only if that production 
period will exceed 12 months in length.
* * * * *
    9. Section 90.107 is amended by adding a semicolon at the end of 
paragraph (d)(5), by removing ``and'' at the end of paragraph (d)(9), 
by removing the period at the end of paragraph (d)(10) and adding a 
semicolon in its place, and by adding new paragraphs (d)(11) and 
(d)(12) to read as follows:


Sec. 90.107  Application for certification.

* * * * *
    (d) * * *
    (11) This paragraph (d)(11) is applicable only to Phase 2 engines.
    (i) Manufacturers of nonhandheld engines participating in the 
Averaging, Banking and Trading Program as described in Subpart C of 
this part shall declare the applicable Family Emission Limit (FEL) for 
HC+NOX (NMHC+NOX).
    (ii) Provide the applicable useful life as determined under 
Sec. 90.105;
    (12) In cases where the regulations in Sec. 90.114(f) are 
applicable, a copy of the language to be included in the documents 
intended for the ultimate purchaser to describe the emission compliance 
period.
* * * * *
    10. Section 90.108 is amended by adding paragraphs (c) and (d) to 
read as follows:


Sec. 90.108  Certification.

* * * * *
    (c) For certificates issued for engine families included in the 
averaging, banking and trading program as described in subpart C of 
this part:
    (1) All certificates issued are conditional upon the manufacturer 
complying with the provisions of subpart C of this part and the 
averaging, banking and trading related provisions of other applicable 
sections, both during and after the model year of production.
    (2) Failure to comply with all applicable averaging, banking and 
trading provisions in this part will be considered to be a failure to 
comply with the terms and conditions upon which the certificate was 
issued, and the certificate may be determined to be void ab initio.
    (3) The manufacturer shall bear the burden of establishing to the 
satisfaction of the Administrator that the conditions upon which the 
certificate was granted were satisfied or waived.
    (d) The Administrator may, upon request by a manufacturer, waive 
any requirement of this part otherwise necessary for the issuance of a 
certificate. The Administrator may set such conditions in a certificate 
as he or she deems appropriate to assure that the waived requirements 
are either satisfied or are demonstrated, for the subject engines, to 
be inappropriate, irrelevant or met by the application of a different 
requirement under this chapter. The Administrator may indicate on such 
conditional certificates that failure to meet these conditions may 
result in suspension or revocation or the voiding ab initio of the 
certificate.
    11. Section 90.113 is amended by revising the section heading and 
adding two sentences to the beginning of paragraph (a) to read as 
follows:


Sec. 90.113  In-use testing program for Phase 1 engines.

    (a) This section applies only to Phase 1 engines. In-use testing 
requirements for Phase 2 engines are found in subpart M of this part.* 
* *
* * * * *
    12. Section 90.114 is amended by removing ``and'' at the end of 
paragraph (c)(9), by removing the period at the end of paragraph 
(c)(10) and adding a semicolon in its place, and by adding new 
paragraphs (c)(11), (c)(12) and (f) to read as follows:


Sec. 90.114  Requirement of certification--engine information label.

* * * * *
    (c)* * *
    (11) For nonhandheld Phase 2 engines, the useful life category as 
determined by the manufacturer pursuant to Sec. 90.105. Such useful 
life category shall be shown by one of the following statements to be 
appended to the statement required under paragraph (c)(7) of this 
section:
    (i) ``EMISSIONS COMPLIANCE PERIOD: [useful life] HOURS''; or
    (ii) ``EMISSIONS COMPLIANCE PERIOD: CATEGORY [fill in C, B or A as 
indicated and appropriate from the chart in Sec. 90.105], REFER TO 
OWNER'S

[[Page 4014]]

MANUAL FOR FURTHER INFORMATION'';
    (12) For handheld Phase 2 engines, the useful life category as 
determined by the manufacturer pursuant to Sec. 90.105. Such useful 
life category shall be shown by the following statement to be appended 
to the statement required under (c)(7) of this section: ``EMISSIONS 
COMPLIANCE PERIOD: [ 50 or 300, as applicable] HOURS''.
* * * * *
    (f)(1) Manufacturers electing to use the labeling language of 
paragraph (c)(11)(ii) of this section must provide in the documents 
intended to be conveyed to the ultimate purchaser, the statement:

    The Emissions Compliance Period referred to on the label 
entitled ``Important Engine Information'' indicates the number of 
operating hours for which the engine has been shown to meet Federal 
emission requirements. For engines less than 225 cc displacement, 
Category C= 66 hours, B= 250 hours and A = 500 hours. For engines of 
225 cc or more, Category C = 250 hours, B = 500 hours and A = 1000 
hours.

    (2) The manufacturer must provide, in the same document as the 
statement in paragraph (f)(1) of this section, a statement of the 
engine's displacement or an explanation of how to readily determine the 
engine's displacement. The Administrator may approve alternate language 
to the statement in paragraph (f)(1) of this section, provided that the 
alternate language provides the ultimate purchaser with a clear 
description of the number of hours represented by each of the three 
letter categories for the subject engine's displacement.
    13. Section 90.116 is amended by revising paragraph (d)(6) and 
(d)(7) and adding paragraphs (d)(8) through (d)(10) to read as follows:


Sec. 90.116  Certification procedure--determining engine displacement, 
engine class, and engine families.

* * * * *
    (d) * * *
    (6) The location of valves, where applicable, with respect to the 
cylinder (e.g., side valves or overhead valves);
    (7) The number of catalytic converters, location, volume and 
composition;
    (8) The thermal reactor characteristics;
    (9) The fuel required (e.g., gasoline, natural gas, LPG); and
    (10) The useful life category.
* * * * *
    14. Section 90.117 is amended by revising paragraph (a) to read as 
follows:


Sec. 90.117  Certification procedure--test engine selection.

    (a) For Phase 1 engines, the manufacturer must select, from each 
engine family, a test engine that the manufacturer determines to be 
most likely to exceed the emission standard. For Phase 2 engines, the 
manufacturer must select, from each engine family, a test engine of a 
configuration that the manufacturer determines to be most likely to 
exceed the HC+NOX [NMHC+NOX] Family Emission 
Limit (FEL), or HC+NOX [NMHC+NOX] standard if no 
FEL is applicable.
* * * * *
    15. Section 90.118 is amended by revising the section heading and 
adding a new paragraph (e) to read as follows:


Sec. 90.118  Certification procedure--service accumulation and usage of 
deterioration factors.

* * * * *
    (e) For purposes of establishing whether Phase 2 engines comply 
with applicable exhaust emission standards or FELs, the test results 
for each regulated pollutant as measured pursuant to Sec. 90.119 shall 
be multiplied by the applicable df determined under Sec. 90.104 (g), 
(h) or (i). The product of the two numbers shall be rounded to the same 
number of decimal places contained in the applicable standard, and 
compared against the applicable standard or FEL, as appropriate.
    16. Section 90.122 is amended by revising the first sentence of 
paragraph (a) and adding paragraph (d)(4) as follows:


Sec. 90.122  Amending the application and certificate of conformity.

    (a) The engine manufacturer must notify the Administrator when 
either an engine is to be added to a certificate of conformity, an FEL 
is to be changed, or changes are to be made to a product line covered 
by a certificate of conformity. * * *
* * * * *
    (d)* * *
    (4) If the Administrator determines that a revised FEL meets the 
requirements of this subpart and the Act, the appropriate certificate 
of conformity will be amended, or a new certificate will be issued to 
reflect the revised FEL. The certificate of conformity is revised 
conditional upon compliance with Sec. 90.207(b).
* * * * *
    17. Subpart C, which was formerly reserved, is added to part 90 to 
read as follows:

Subpart C--Certification Averaging, Banking, and Trading Provisions for 
Nonhandheld Engines

Sec.
90.201  Applicability.
90.202  Definitions.
90.203  General provisions.
90.204  Averaging.
90.205  Banking.
90.206  Trading.
90.207  Credit calculation and manufacturer compliance with emission 
standards.
90.208  Certification.
90.209  Maintenance of records.
90.210  End-of-year and final reports.
90.211  Request for hearing.

Subpart C--Certification Averaging, Banking, and Trading Provisions 
for Nonhandheld Engines


Sec. 90.201  Applicability.

    The requirements of this subpart C are applicable to all Phase 2 
nonhandheld spark-ignition engines subject to the provisions of subpart 
A of this part except as provided in Sec. 90.103(a). These provisions 
are not applicable to any Phase 1 engines or to any Phase 2 handheld 
engines. Participation in the averaging, banking and trading program is 
voluntary, but if a manufacturer elects to participate, it must do so 
in compliance with the regulations set forth in this subpart. The 
provisions of this subpart are applicable for HC+NOX 
(NMHC+NOX) emissions but not for CO emissions.


Sec. 90.202  Definitions.

    The definitions in subpart A of this part apply to this subpart. 
The following definitions also apply to this subpart:
    Averaging means the exchange of emission credits between engine 
families within a given manufacturer's product line.
    Banking means the retention of emission credits by the manufacturer 
generating the emission credits or obtaining such credits through 
trading, for use in future model year averaging or trading as permitted 
in this part.
    Emission credits represent the amount of emission reduction or 
exceedance, by an engine family, below or above the applicable 
HC+NOX (NMHC+NOX) emission standard, 
respectively. FELs below the standard create ``positive credits,'' 
while FELs above the standard create ``negative credits.'' In addition, 
``projected credits'' refer to emission credits based on the projected 
applicable production/sales volume of the engine family. ``Reserved 
credits'' are emission credits generated within a model year waiting to 
be reported to EPA at the end of the model year. ``Actual credits'' 
refer to emission credits based on actual applicable sales volume as 
contained in the end-of-year

[[Page 4015]]

reports submitted to EPA. Some or all of these credits may be revoked 
if EPA review of the end-of-year reports or any subsequent audit 
action(s) reveals problems or errors of any nature with credit 
computations.
    Point of first retail sale means the point at which the engine is 
first sold directly to an end user. Generally, this point is the retail 
engine or equipment dealer. If the engine is sold first to an equipment 
manufacturer for installation in a piece of equipment, the equipment 
manufacturer may be the point of first retail sale if the equipment 
manufacturer can determine with reasonable certainty whether the engine 
is or is not exported or destined for retail sale in a state that has 
adopted applicable emission standards pursuant to a waiver granted by 
EPA under section 209(e) of the Act once it has been installed in a 
piece of equipment.
    Trading means the exchange of emission credits between 
manufacturers.


Sec. 90.203  General provisions.

    (a) The certification averaging, banking, and trading provisions 
for HC+NOX and NMHC+NOX emissions from eligible 
engines are described in this subpart.
    (b) A nonhandheld engine family may use the averaging, banking and 
trading provisions for HC+NOX and NMHC+NOX 
emissions if it is subject to regulation under this part with certain 
exceptions specified in paragraph (c) of this section. 
HC+NOX and NMHC+NOX credits shall be 
interchangeable subject to the limitations on credit generation, credit 
usage, cross class averaging and other provisions described in this 
subpart.
    (c) A manufacturer shall not include in its calculation of credit 
generation and may exclude from its calculation of credit usage, any 
new engines:
    (1) Which are exported, unless the manufacturer has reason or 
should have reason to believe that such engines have been or will be 
imported in a piece of equipment; or
    (2) Which are subject to state engine emission standards pursuant 
to a waiver granted by EPA under section 209(e) of the Act, unless the 
manufacturer demonstrates to the satisfaction of the Administrator that 
inclusion of these engines in averaging, banking and trading is 
appropriate.
    (d) For an engine family using credits, a manufacturer may, at its 
option, include its entire production of that engine family in its 
calculation of credit usage for a given model year.
    (e) A manufacturer may certify engine families at Family Emission 
Limits (FELs) above or below the applicable emission standard subject 
to the limitation in paragraph (f) of this section, provided the 
summation of the manufacturer's projected balance of credits from all 
credit transactions for each engine class in a given model year is 
greater than or equal to zero, as determined under Sec. 90.207.
    (1) A manufacturer of an engine family with an FEL exceeding the 
applicable emission standard must obtain positive emission credits 
sufficient to address the associated credit shortfall via averaging, 
banking, or trading.
    (2) An engine family with an FEL below the applicable emission 
standard may generate positive emission credits for averaging, banking, 
or trading, or a combination thereof.
    (3) In the case of an SEA failure, credits may be used to cover 
subsequent production of engines for the family in question if the 
manufacturer elects to recertify to a higher FEL. Credits may not be 
used to remedy a nonconformity determined by a Selective Enforcement 
Audit (SEA) or by in-use testing, except that the Administrator may 
permit the use of credits to address a nonconformity determined by an 
SEA where the use of such credits is one component of a multi-part 
remedy for the previously produced engines and the remedy, including 
the use of credits and the quantity of credits being used, is such that 
the Administrator is satisfied that the manufacturer has strong and 
lasting incentive to accurately verify its new engine emission levels 
and will set or reset its FELs for current and future model years so 
that production line compliance is assured.
    (4) In the case of a production line testing failure pursuant to 
subpart H of this part, a manufacturer may revise the FEL based upon 
production line testing results obtained under subpart H of this part 
and upon Administrator approval pursuant to Sec. 90.122(d). The 
manufacturer may use certification credits to cover both past 
production and subsequent production of nonhandheld engines as needed.
    (f) No engine family may have an FEL that is greater than 32.2 g/
kW-hr for Class I engines or 26.8 g/kW-hr for Class II engines.
    (g)(1) All credits generated under this subpart will be designated 
as Class I or Class II credits, as appropriate. Except as described in 
Sec. 90.204(b), credits generated in a given model year by an engine 
family subject to the Phase 2 emission requirements may only be used in 
averaging, banking or trading, as appropriate, for any nonhandheld 
engine family of the same class for which the Phase 2 requirements are 
applicable. Credits generated in one model year may not be used for 
prior model years, except as allowed under Sec. 90.207(c) or 
Sec. 90.104(h)(2)(iv).
    (2) For the 2005 model year and for each subsequent model year, 
manufacturers of Class II engines must provide a demonstration that the 
sales weighted average FEL for HC+NOX (including 
NMHC+NOX FELs), for all of the manufacturer's Class II 
engines, will not exceed 13.6 g/kW-hr for the 2005 model year, 13.1 g/
kW-hr for the 2006 model year and 12.6 g/kW-hr for the 2007 and each 
subsequent Phase 2 model year. Such demonstration shall be subject to 
the review and approval of the Administrator, shall be provided at the 
time of the first Class II certification of that model year and shall 
be based on projected eligible sales for that model year.
    (h) Manufacturers must demonstrate compliance under the averaging, 
banking, and trading provisions for a particular model year by 270 days 
after the end of the model year. An engine family generating negative 
credits for which the manufacturer does not obtain or generate an 
adequate number of positive credits by that date from the same or 
previous model year engines will violate the conditions of the 
certificate of conformity. The certificate of conformity may be voided 
ab initio pursuant to Sec. 90.123 for this engine family.


Sec. 90.204  Averaging.

    (a) Negative credits from engine families with FELs above the 
applicable emission standard must be offset by positive credits from 
engine families having FELs below the applicable emission standard, as 
allowed under the provisions of this subpart. Averaging of credits in 
this manner is used to determine compliance under Sec. 90.207(b).
    (b) Cross-class averaging, i.e. the use of credits from Class I 
engines to cover Class II engines and vice versa, is permitted only for 
the two situations described in paragraphs (b)(1) and (b)(2) of this 
section and only when the affected Class II manufacturer meets the 
following minimum sales percentages for Class II overhead valve 
emission performance engines in that model year: 2001 (50%); 2002 
(62.5%); 2003 (75%); 2004 (87.5%) and 2005 and later (100%). A 
manufacturer's sales percentage of overhead valve emission performance 
engines is determined by dividing the manufacturer's eligible sales (as 
defined in this part) of Class II overhead valve emission performance

[[Page 4016]]

engines certified under this part by the manufacturer's total eligible 
sales of Class II engines certified under this part, and multiplying 
the resultant quotient by 100.
    (1) Cross class averaging is allowed for credit exchanges from 
credit generating Class II engines to credit using Class I engines.
    (2) Cross class averaging is allowed for credit exchanges from 
Class I engines to Class II engines where credits are necessary to 
address production line testing failures as permitted in Sec. 90.207 or 
to address credit shortfalls that arise due to testing pursuant to 
Sec. 90.104(h)(2)(iv) .
    (c) Subject to the limitations in Sec. 90.204(b), credits used in 
averaging for a given model year may be obtained from credits generated 
in the same model year by another engine family, credits banked in 
previous model years, or credits of the same or previous model year 
obtained through trading. The restrictions of this paragraph 
notwithstanding, credits from a given model year may be used to address 
credit needs of previous model year engines as allowed under 
Sec. 90.207(c).
    (d) The use of Class II credits from the 1999 and 2000 model years 
(early banking) is subject to regulation under this subpart and also to 
the provisions of Sec. 90.103(a)(7).


Sec. 90.205  Banking.

    (a) Beginning with the 2001 model year, a manufacturer of an engine 
family with an FEL below the applicable emission standard for a given 
model year may bank credits in that model year for use in averaging and 
trading. Negative credits may be banked only according to the 
requirements under Sec. 90.207(c). Credits may also be banked in model 
years 1999 and 2000 subject to the requirements of paragraph (b) of 
this section.
    (b) A manufacturer may bank credits for a given class of engines in 
the 1999 and 2000 model years for use in the 2001 and later model 
years, provided:
    (1) For Class I credits: the manufacturer certifies its entire 
Class I production to the applicable 2001 model year requirements. 
HC+NOX (NMHC+NOX) credits may only be banked from 
engine families certified below 16.0 g/kW-hr (15.0 g/kW-hr) where those 
credits are not needed to bring the manufacturer's total Class I sales 
into compliance with the 2001 model year standard.
    (2) For Class II credits: the manufacturer certifies its entire 
Class II product line to the applicable 2001 model year requirements. 
HC+NOX (NMHC+NOX) credits may only be banked from 
engine families certified below 12.1 (11.3 g/kw-hr) for engines where 
those credits are not needed to bring the manufacturer's total Class II 
sales into compliance with the 2001 model year standard.
    (3) Engines certified under the provisions of this paragraph are 
subject to all of the requirements of this part applicable to Phase 2 
engines.
    (c) A manufacturer may bank actual credits only after the end of 
the model year and after EPA has reviewed the manufacturer's end-of-
year reports. During the model year and before submittal of the end-of-
year report, credits originally designated in the certification process 
for banking will be considered reserved and may be redesignated for 
trading or averaging in the end-of-year report and final report.
    (d) Credits declared for banking from the previous model year that 
have not been reviewed by EPA may be used in averaging or trading 
transactions. However, such credits may be revoked at a later time 
following EPA review of the end-of-year report or any subsequent audit 
actions.


Sec. 90.206  Trading.

    (a) An engine manufacturer may exchange emission credits with other 
nonhandheld engine manufacturers in trading.
    (b) Credits for trading can be obtained from credits banked in 
previous model years or credits generated during the model year of the 
trading transaction.
    (c) Traded credits can be used for averaging, banking, or further 
trading transactions.
    (d) Traded credits are subject to the limitations on cross-class 
averaging, use for past model years, and the use of credits from early 
banking as set forth in Sec. 90.204(b), (c) and (d).
    (e) In the event of a negative credit balance resulting from a 
transaction, both the buyer and the seller are liable, except in cases 
involving fraud. Certificates of all engine families participating in a 
negative trade may be voided ab initio pursuant to Sec. 90.123.


Sec. 90.207  Credit calculation and manufacturer compliance with 
emission standards.

    (a) (1) For each engine family, HC+NOX 
[NMHC+NOX] certification emission credits (positive or 
negative) are to be calculated according to the following equation and 
rounded to the nearest gram. Consistent units are to be used throughout 
the following equation:

Credits = Sales  x  (Standard--FEL)  x  Power  x  Useful life  x  Load 
Factor

Where:

    Sales = eligible sales as defined in this part. Annual sales 
projections are used to project credit availability for initial 
certification. Eligible sales volume is used in determining actual 
credits for end-of-year compliance determination.
    Standard = the current and applicable Small SI engine 
HC+NOX (NMHC+NOX) emission standard in grams 
per kilowatt hour as determined in Sec. 90.103.
    FEL = the family emission limit for the engine family in grams 
per kilowatt hour.
    Power = the sales weighted maximum modal power, in kilowatts, as 
calculated from the applicable federal test procedure as described 
in this part. This is determined by multiplying the maximum modal 
power of each configuration within the family by its eligible sales, 
summing across all configurations and dividing by the eligible sales 
of the entire family.
    Useful Life = the useful life in hours corresponding to the 
useful life category for which the engine family was certified.
    Load Factor = For Test Cycle A and Test Cycle B, the Load Factor 
= 47% (i.e. 0.47).

    (2) For approved alternate test procedures, the load factor in 
paragraph (a)(1) of this section must be calculated according to the 
following formula:
[GRAPHIC] [TIFF OMITTED] TP27JA98.001

Where:

    %MTT modei = percent of the maximum FTP torque for 
mode i.
    %MTS modei = percent of the maximum FTP engine 
rotational speed for mode i.
    WF modei = the weighting factor for mode i.

    (b) Manufacturer compliance with the emission standard is 
determined on a corporate average basis at the end of each model year. 
A manufacturer is in compliance when the sum of positive and negative 
emission credits it holds for each class is greater than or equal to 
zero, except that the sum of positive and negative credits for a given 
class may be less than zero as allowed under paragraph (c) of this 
section.
    (c)(1) A manufacturer may use credits from a later model year to 
address dfs of model year 2001 through 2004 Class II engines certified 
to 500 or 1000 hours, when the dfs are shown to be underestimated 
pursuant to the provisions of Sec. 90.104(h)(2)(iv).
    (2) If, as a result of production line testing as required in 
subpart H of this part, a nonhandheld engine family is determined to be 
in noncompliance pursuant to Sec. 90.710, the manufacturer may raise 
its FEL for past and future production as necessary. Further, a 
manufacturer may carry a negative credit balance (known also as a 
credit deficit) for the subject class and model year and for the next 
three model years.

[[Page 4017]]

The credit deficit may be no larger than that created by the 
nonconforming family. If the credit deficit still exists after the 
model year following the model year in which the nonconformity 
occurred, the manufacturer must obtain and apply credits to offset the 
remaining credit deficit at a rate of 1.2 grams for each gram of 
deficit within the next two model years. The provisions of this 
paragraph are subject to the limitations in paragraph (d) of this 
section.
    (d) Regulations elsewhere in this part notwithstanding, if a 
nonhandheld engine manufacturer experiences two or more production line 
testing failures pursuant to the regulations in subpart H of this part 
in a given model year, the manufacturer may raise the FEL of previously 
produced engines only to the extent that such engines represent no more 
than 10% of the manufacturer's total eligible sales for that model 
year. For any additional engines determined to be in noncompliance, the 
manufacturer must conduct offsetting projects approved in advance by 
the Administrator.
    (e) If, as a result of production line testing under this subpart, 
a manufacturer desires to lower its FEL it may do so subject to 
Sec. 90.708(c).
    (f) Except as allowed at paragraph (c) of this section, when a 
manufacturer is not in compliance with the applicable emission standard 
by the date 270 days after the end of the model year, considering all 
credit calculations and transactions completed by then, the 
manufacturer will be in violation of these regulations and EPA may, 
pursuant to Sec. 90.123, void ab initio the certificates of engine 
families for which the manufacturer has not obtained sufficient 
positive emission credits.


Sec. 90.208  Certification.

    (a) In the application for certification a manufacturer must:
    (1) Submit a statement that the engines for which certification is 
requested will not, to the best of the manufacturer's belief, cause the 
manufacturer to be in noncompliance under Sec. 90.207(b) when all 
credits are calculated for all the manufacturer's engine families.
    (2) Declare an FEL for each engine family for HC+NOX 
(NMHC+NOX). The FEL must have the same number of significant 
digits as the emission standard.
    (3) Indicate the projected number of credits generated/needed for 
this family; the projected applicable eligible sales volume, by 
quarter; and the values required to calculate credits as given in 
Sec. 90.207.
    (4) Submit calculations in accordance with Sec. 90.207 of projected 
emission credits (positive or negative) based on quarterly production 
projections for each family.
    (5)(i) If the engine family is projected to have negative emission 
credits, state specifically the source (manufacturer/engine family or 
reserved) of the credits necessary to offset the credit deficit 
according to quarterly projected production.
    (ii) If the engine family is projected to generate credits, state 
specifically (manufacturer/engine family or reserved) where the 
quarterly projected credits will be applied.
    (b) All certificates issued are conditional upon manufacturer 
compliance with the provisions of this subpart both during and after 
the model year of production.
    (c) Failure to comply with all provisions of this subpart will be 
considered to be a failure to satisfy the conditions upon which the 
certificate was issued, and the certificate may be determined to be 
void ab initio pursuant to Sec. 90.123.
    (d) The manufacturer bears the burden of establishing to the 
satisfaction of the Administrator that the conditions upon which the 
certificate was issued were satisfied or waived.
    (e) Projected credits based on information supplied in the 
certification application may be used to obtain a certificate of 
conformity. However, any such credits may be revoked based on review of 
end-of-year reports, follow-up audits, and any other verification steps 
considered appropriate by the Administrator.


Sec. 90.209  Maintenance of records.

    (a) The manufacturer must establish, maintain, and retain the 
following adequately organized and indexed records for each engine 
family:
    (1) EPA engine family identification code;
    (2) Family Emission Limit (FEL) or FELs where FEL changes have been 
implemented during the model year;
    (3) Maximum modal power for each configuration sold;
    (4) Projected sales volume for the model year; and
    (5) Records appropriate to establish the quantities of engines that 
constitute eligible sales as defined in Sec. 90.202 for each power 
rating for each FEL.
    (b) Any manufacturer producing an engine family participating in 
trading reserved credits must maintain the following records on a 
quarterly basis for each such engine family:
    (1) The engine family;
    (2) The actual quarterly and cumulative applicable production/sales 
volume;
    (3) The values required to calculate credits as given in 
Sec. 90.207;
    (4) The resulting type and number of credits generated/required;
    (5) How and where credit surpluses are dispersed; and
    (6) How and through what means credit deficits are met.
    (c) The manufacturer must retain all records required to be 
maintained under this section for a period of eight years from the due 
date for the end-of-model year report. Records may be retained as hard 
copy or reduced to microfilm, ADP diskettes, and so forth, depending on 
the manufacturer's record retention procedure; provided, that in every 
case all information contained in the hard copy is retained.
    (d) Nothing in this section limits the Administrator's discretion 
in requiring the manufacturer to retain additional records or submit 
information not specifically required by this section.
    (e) Pursuant to a request made by the Administrator, the 
manufacturer must submit to the Administrator the information that the 
manufacturer is required to retain.
    (f) EPA may, pursuant to Sec. 90.123, void ab initio a certificate 
of conformity for an engine family for which the manufacturer fails to 
retain the records required in this section or to provide such 
information to the Administrator upon request.


Sec. 90.210  End-of-year and final reports.

    (a) End-of-year and final reports must indicate the engine family, 
the class (I or II), the actual sales volume, the values required to 
calculate credits as given in Sec. 90.207, and the number of credits 
generated/required. Manufacturers must also submit how and where credit 
surpluses were dispersed (or are to be banked) and/or how and through 
what means credit deficits were met. Copies of contracts related to 
credit trading must be included or supplied by the broker, if 
applicable. The report must include a calculation of credit balances to 
show that the credit summation for each class of engines is equal to or 
greater than zero (or less than zero in cases of negative credit 
balances as permitted in Sec. 90.207(c)). For engines subject to the 
provisions of Sec. 90.203(g)(2), the report must include a calculation 
of the sales weighted average HC+NOX (including 
NMHC+NOX) FEL.
    (b) The calculation of eligible sales for end-of-year and final 
reports must be based on the location of the point of first retail sale 
(for example, retail customer

[[Page 4018]]

or dealer) also called the final product purchase location. Upon 
advance written request, the Administrator will consider other methods 
to track engines for credit calculation purposes that provide high 
levels of confidence that eligible sales are accurately counted.
    (c)(1) End-of-year reports must be submitted within 90 days of the 
end of the model year to: Manager, Engine Compliance Programs Group 
(6403-J), U.S. Environmental Protection Agency, Washington, DC 20460.
    (2) Unless otherwise approved by the Administrator, final reports 
must be submitted within 270 days of the end of the model year to: 
Manager, Engine Compliance Programs Group (6403-J), U.S. Environmental 
Protection Agency, Washington, DC 20460.
    (d) Failure by a manufacturer to submit any end-of-year or final 
reports in the specified time for any engines subject to regulation 
under this part is a violation of Sec. 90.1003(a)(2) and section 213(d) 
of the Clean Air Act for each engine.
    (e) A manufacturer generating credits for banking only who fails to 
submit end-of-year reports in the applicable specified time period (90 
days after the end of the model year) may not use the credits until 
such reports are received and reviewed by EPA. Use of projected credits 
pending EPA review is not permitted in these circumstances.
    (f) Errors discovered by EPA or the manufacturer in the end-of-year 
report, including errors in credit calculation, may be corrected in the 
final report.
    (g) If EPA or the manufacturer determines that a reporting error 
occurred on an end-of-year or final report previously submitted to EPA 
under this section, the manufacturer's credits and credit calculations 
must be recalculated. Erroneous positive credits will be void except as 
provided in paragraph (h) of this section. Erroneous negative credit 
balances may be adjusted by EPA.
    (h) If within 270 days of the end of the model year, EPA review 
determines a reporting error in the manufacturer's favor (that is, 
resulting in an increased credit balance) or if the manufacturer 
discovers such an error within 270 days of the end of the model year, 
EPA shall restore the credits for use by the manufacturer.


Sec. 90.211  Request for hearing.

    An engine manufacturer may request a hearing on the Administrator's 
voiding of the certificate under Secs. 90.203(h), 90.206(e), 90.207(f), 
90.208(c), or 90.209(f), pursuant to Sec. 90.124. The procedures of 
Sec. 90.125 shall apply to any such hearing.

Subpart D--Emission Test Equipment Provisions

    18. Section 90.301 is amended by revising paragraph (a) and adding 
paragraph (d) to read as follows:


Sec. 90.301  Applicability.

    (a) This subpart describes the equipment required in order to 
perform exhaust emission tests on new nonroad spark-ignition engines 
and vehicles subject to the provisions of subpart A of this part. 
Certain text in this subpart is identified as pertaining to Phase 1 or 
Phase 2 engines. Such text pertains only to engines of the specified 
Phase. If no indication of Phase is given, the text pertains to all 
engines, regardless of Phase.
* * * * *
    (d) For Phase 2 Class I and Phase 2 Class II natural gas fueled 
engines, the following sections from 40 CFR part 86 are applicable to 
this subpart. The requirements of these sections which pertain 
specifically to the measurement and calculation of non-methane 
hydrocarbon (NMHC) exhaust emissions from otto cycle heavy-duty engines 
must be followed when determining the NMHC exhaust emissions from Phase 
2 Class I and Phase 2 Class II natural gas fueled engines. Those 
sections are: 40 CFR 86.1306-90 Equipment required and specifications; 
overview, 40 CFR 86.1309-90 Exhaust gas sampling system; otto-cycle 
engines, 40 CFR 86-1311-94 Exhaust gas analytical system; CVS bag 
sampling, 40 CFR 86.1313-94(e) Fuel Specification--Natural gas-fuel, 40 
CFR 86.1314-94 Analytical gases, 40 CFR 86.1316-94 Calibrations; 
frequency and overview, 40 CFR 86.1321-94 Hydrocarbon analyzer 
calibration, 40 CFR 86.1325-94 Methane analyzer calibration, 40 CFR 
86.1327-94 Engine dynamometer test procedures, overview, 40 CFR 
86.1340-94 Exhaust sample analysis, 40 CFR 86.1342-94 Calculations; 
exhaust emissions, 40 CFR 86.1344-94(d) Required information--Pre-test 
data, 40 CFR 86.1344-94(e) Required information--Test data.
    19. Section 90.302 is revised to read as follows:


Sec. 90.302  Definitions.

    The definitions in Sec. 90.3 apply to this subpart. The following 
definitions also apply to this subpart.
    Intermediate speed means the engine speed which is 85 percent of 
the rated speed.
    Natural gas means a fuel whose primary constituent is methane.
    Rated speed means the speed at which the manufacturer specifies the 
maximum rated power of an engine.

Subpart E--Gaseous Exhaust Test Procedures

    20. Section Sec. 90.401 is amended by adding paragraphs (c) and (d) 
to read as follows;


Sec. 90.401  Applicability.

* * * * *
    (c) Certain text in this subpart is identified as pertaining to 
Phase 1 or Phase 2 engines. Such text pertains only to engines of the 
specified Phase. If no indication of Phase is given, the text pertains 
to all engines, regardless of Phase.
    (d) For Phase 2 Class I and Phase 2 Class II natural gas fueled 
engines, the following sections from 40 CFR part 86 are applicable to 
this subpart. The requirements of these sections which pertain 
specifically to the measurement and calculation of non-methane 
hydrocarbon (NMHC) exhaust emissions from otto cycle heavy-duty engines 
must be followed when determining the NMHC exhaust emissions from Phase 
2 Class I and Phase 2 Class II natural gas fueled engines. Those 
sections are: 40 CFR 86.1327-94 Engine dynamometer test procedures, 
overview, 40 CFR 86.1340-94 Exhaust sample analysis, 40 CFR 86.1342-94 
Calculations; exhaust emissions, 40 CFR 86.1344-94(d) Required 
information--Pre-test data, and 40 CFR 86.1344-94(e) Required 
information--Test data.
    21. Section 90.404 is amended by adding a sentence after the first 
sentence of paragraph (b) to read as follows:


Sec. 90.404  Test procedure overview.

* * * * *
    (b) * * * For Phase 2 Class I and II natural gas fueled engines the 
test is also designed to determine the brake-specific emissions of non-
methane hydrocarbons. * * *
* * * * *
    22. Section 90.409 is amended by revising paragraph (a)(3) to read 
as follows:


Sec. 90.409  Engine dynamometer test run.

    (a) * * *
    (3) For Phase 1 engines, at the manufacturer's option, the engine 
can be run with the throttle in a fixed position or by using the 
engine's governor (if the engine is manufactured with a governor). In 
either case, the engine speed and load must meet the requirements 
specified in paragraph (b)(12) of this section. For Phase 2 Class I and 
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, and no external throttle control

[[Page 4019]]

may be used. For Phase 2 Class I and Class II engines equipped with an 
engine speed governor, during Mode 1 fixed throttle operation may be 
used to determine the 100% torque value.
* * * * *
    23. Section 90.410 is amended by revising paragraph (b) to read as 
follows:


Sec. 90.410  Engine test cycle.

* * * * *
    (b) For Phase 1 engines and Phase 2 Class III, IV, V, and Phase 2 
Class I and II engines not equipped with an engine speed governor, 
during each non-idle mode, hold both the specified speed and load 
within  five percent of point. During the idle mode, hold 
speed within  ten percent of the manufacturer's specified 
idle engine speed. For Phase 2 Class I and II engines equipped with an 
engine speed governor, during Mode 1 hold both the specified speed and 
load within  five percent of point, during Modes 2-5, hold 
the specified load with  five 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 to subpart E of this part for a description of test Modes).
* * * * *
    24. In Appendix A to Subpart E of Part 90, Table 2 is revised to 
read as follows:

Appendix A to Subpart E of Part 90--Tables

* * * * *

                                   Table 2.--Test Cycles for Class I-V Engines                                  
----------------------------------------------------------------------------------------------------------------
                Mode                   1      2      3      4      5      6      7      8      9      10     11 
----------------------------------------------------------------------------------------------------------------
               Speed                                                                                            
(4) Rated speed                                                                                                 
(4) Intermediate speed                Idle                                                                      
Mode Points A Cycle................  .....  .....  .....  .....  .....      1      2      3      4      5      6
Load Percent--A Cycle..............  .....  .....  .....  .....  .....    100     75     50     25     10      0
Weighting..........................  .....  .....  .....  .....  .....     9%    20%    29%    30%     7%     5%
----------------------------------------------------------------------------------------------------------------
Mode Points B Cycle................      1      2      3      4      5  .....  .....  .....  .....  .....      6
Load Percent--B Cycle..............    100     75     50     25     10  .....  .....  .....  .....  .....      0
Weighting..........................  .....  .....  .....  .....  .....     9%    20%    29%    30%     7%     5%
----------------------------------------------------------------------------------------------------------------
Mode Points C Cycle................      1  .....  .....  .....  .....  .....  .....  .....  .....  .....      2
Load Percent--C Cycle..............    100  .....  .....  .....  .....  .....  .....  .....  .....  .....      0
Weighting for Phase 1 Engines......    90%  .....  .....  .....  .....  .....  .....  .....  .....  .....    10%
Weighting for Phase 2 Engines......    85%  .....  .....  .....  .....  .....  .....  .....  .....  .....    15%
----------------------------------------------------------------------------------------------------------------

Subpart F--Selective Enforcement Auditing

    25. Section 90.503 is amended by revising paragraphs (f)(3) and 
(f)(4) to read as follows:


Sec. 90.503  Test orders.

* * * * *
    (f) * * *
    (3) Any SEA test order for which the family or configuration, as 
appropriate, fails under Sec. 90.510 or for which testing is not 
completed will not be counted against the annual limit.
    (4) When the annual limit has been met, the Administrator may issue 
additional test orders to test those families or configurations for 
which evidence exists indicating nonconformity, or for which the 
Administrator has reason to believe are not being appropriately 
represented or tested in Production Line Testing conducted under 
subpart H of this part, if applicable. An SEA test order issued 
pursuant to this provision will include a statement as to the reason 
for its issuance.
    26. Section 90.509 is amended by revising paragraph (b) to read as 
follows:


Sec. 90.509  Calculation and reporting of test results.

* * * * *
    (b)(1) Final test results are calculated by summing the initial 
test results derived in paragraph (a) of this section for each test 
engine, dividing by the number of tests conducted on the engine, and 
rounding to the same number of decimal places contained in the 
applicable standard. For Phase 2 engines only, this result shall be 
expressed to one additional significant figure.
    (2) Final deteriorated test results (for Phase 2 test engines only) 
are calculated by applying the appropriate deterioration factors, from 
the certification process for the engine family, to the final test 
results, and rounding to the same number of decimal places contained in 
the applicable standard.
* * * * *
    27. Section 90.510 is amended by revising paragraph (b) to read as 
follows:


Sec. 90.510  Compliance with acceptable quality level and passing and 
failing criteria for selective enforcement audits.

* * * * *
    (b) A failed engine is a Phase 1 engine whose final test results 
pursuant to Sec. 90.509(b), for one or more of the applicable 
pollutants exceed the emission standard. For Phase 2 engines, a failed 
engine is a Phase 2 engine whose final deteriorated test results 
pursuant to Sec. 90.509(b), for one or more of the applicable 
pollutants exceed the emission standard (FEL, if applicable).
* * * * *
    28. Section 90.512 is amended by revising paragraph (b) to read as 
follows:


Sec. 90.512  Request for public hearing.

* * * * *
    (b) The manufacturer's request shall be filed with the 
Administrator not later than 15 days after the Administrator's 
notification of his or her decision to suspend, revoke or void, unless 
otherwise specified by the Administrator. The manufacturer shall 
simultaneously serve two copies of this request upon the Director of 
the Engine Programs and Compliance Division and file two copies with 
the Hearing Clerk of the Agency. Failure of the manufacturer to request 
a hearing within the time provided constitutes a waiver of the right to 
a hearing. Subsequent to the expiration of the period for requesting a 
hearing as of right, the Administrator may, in his or her discretion 
and for good cause shown, grant the manufacturer a hearing to contest 
the suspension, revocation or voiding.
* * * * *

Subpart G--Importation of Nonconforming Engines

    29. Section 90.612 is amended by revising paragraph (g) to read as 
follows:


Sec. 90.612  Exemptions and exclusions.

* * * * *
    (g) Applications for exemptions and exclusions provided for in 
paragraphs (b), (c), and (e) of this section are to be mailed to: U.S. 
Environmental

[[Page 4020]]

Protection Agency, Office of Mobile Sources, Engine Compliance Programs 
Group (6403-J), Washington, D.C. 20460, Attention: Imports.
    30. Subpart H, which was previously ``reserved'', is added to part 
90 to read as follows:

Subpart H--Manufacturer Production Line Testing Program

Sec.

90.701  Applicability.
90.702  Definitions.
90.703  Production line testing by the manufacturer.
90.704  Maintenance of records; submittal of information.
90.705  Right of entry and access.
90.706  Engine sample selection.
90.707  Test procedures.
90.708  Cumulative Sum (CumSum) Procedure.
90.709  Calculation and reporting of test results.
90.710  Compliance with criteria for production line testing.
90.711  Suspension and revocation of certificates of conformity.
90.712  Request for public hearing.
90.713  Administrative procedures for public hearing.

Subpart H--Manufacturer Production Line Testing Program


Sec. 90.701  Applicability.

    (a) Except as described in paragraph (b) of this section, the 
requirements of this subpart are applicable to all Phase 2 nonroad 
engines subject to the provisions of subpart A of this part.
    (b) The requirements of this subpart are applicable to all handheld 
engine families described in paragraph (a) of this section unless 
otherwise exempted in this part. Manufacturers of nonhandheld engine 
families described in paragraph (a) of this section may choose between 
the Production Line Testing Program described in this subpart for all 
of their engine families and the Selective Enforcement Auditing Program 
described in Subpart F of this part for all of their engine families, 
subject to the restrictions of paragraph (d) of this section.
    (c) Nonhandheld engine manufacturers shall notify EPA of their 
selection when they begin their first Phase 2 model year's 
certification.
    (d) A manufacturer of nonhandheld Phase 2 engines may change from 
the Production Line Testing program described in this subpart to the 
Selective Enforcement Auditing program described in Subpart F of this 
part and vice versa, provided that:
    (1) It does so for all of its engine families at the same time;
    (2) When changing from Production Line Testing to Selective 
Enforcement Auditing, it has remained under Production Line Testing for 
a minimum of three model years;
    (3) It provides written notice to EPA one complete model year prior 
to the model year for which it is requesting to change from Production 
Line Testing to Selective Enforcement Auditing;
    (4) It provides written notice to EPA thirty (30) days prior to the 
date for which it is requesting to change from Selective Enforcement 
Auditing to Production Line Testing; and
    (5) It is not carrying a negative credit balance at the time it 
changes from Production Line Testing to Selective Enforcement Auditing.
    (e) The procedures described in this subpart are optional for small 
volume engine manufacturers and small volume engine families as defined 
in this part, and for engine families certified to a level at least 50% 
below the applicable HC+NOX (NMHC+NOX) standard 
(FEL if applicable). Engine families for which the manufacturer opts 
not to conduct testing under this subpart pursuant to this paragraph 
shall be subject to the Selective Enforcement Auditing procedures of 
Subpart F of this part.


Sec. 90.702  Definitions.

    The definitions in subpart A of this part apply to this subpart. 
The following definitions also apply to this subpart.
    Configuration means any subclassification of an engine family which 
can be described on the basis of gross power, emission control system, 
governed speed, injector size, engine calibration, and other parameters 
as designated by the Administrator.
    Test sample means the collection of engines selected from the 
population of an engine family for emission testing.


Sec. 90.703  Production line testing by the manufacturer.

    (a) Manufacturers of small SI engines shall test production line 
engines from each engine family according to the provisions of this 
subpart.
    (b) Production line engines must be tested using the test procedure 
specified in subpart E of this part except that the Administrator may 
approve minor variations that the Administrator deems necessary to 
facilitate efficient and economical testing where the manufacturer 
demonstrates to the satisfaction of the Administrator that such 
variations will not significantly impact the test results. Any 
adjustable engine parameter must be set to values or positions that are 
within the range recommended to the ultimate purchaser, unless 
otherwise specified by the Administrator. The Administrator may specify 
values within or without the range recommended to the ultimate 
purchaser.
    (c) The Administrator, on the basis of a written application from a 
manufacturer, may approve alternate methods to evaluate production line 
compliance, where such alternate methods are demonstrated by the 
manufacturer to:
    (1) Produce substantially the same levels of producer and consumer 
risk as the Cum Sum procedure described in this subpart that mean 
emissions of an engine family are below the appropriate standards (FEL, 
where applicable);
    (2) Provide for continuous rather than point-in-time sampling; and
    (3) Include an appropriate decision mechanism for determining 
noncompliance upon which the Administrator can suspend or revoke the 
certificate of conformity.


Sec. 90.704  Maintenance of records; submittal of information.

    (a) The manufacturer of any new small SI engine subject to any of 
the provisions of this subpart must establish, maintain, and retain the 
following adequately organized and indexed records:
    (1) General records. A description of all equipment used to test 
engines in accordance with Sec. 90.703. Subpart D of this part sets 
forth relevant equipment requirements in Secs. 90.304, 90.305, 90.306, 
90.307, 90.308, 90.309, 90.310 and 90.313.
    (2) Individual records. These records pertain to each production 
line test conducted pursuant to this subpart and include:
    (i) The date, time, and location of each test;
    (ii) The number of hours of service accumulated on the test engine 
when the test began and ended;
    (iii) The names of all supervisory personnel involved in the 
conduct of the production line test;
    (iv) A record and description of any adjustment, repair, 
preparation or modification performed prior to and/or subsequent to 
approval by the Administrator pursuant to Sec. 90.707(b)(1), giving the 
date, associated time, justification, name(s) of the authorizing 
personnel, and names of all supervisory personnel responsible for the 
conduct of the repair;
    (v) If applicable, the date the engine was shipped from the 
assembly plant, associated storage facility or port facility, and the 
date the engine was received at the testing facility;
    (vi) A complete record of all emission tests performed pursuant to 
this subpart (except tests performed directly by EPA), including all 
individual

[[Page 4021]]

worksheets and/or other documentation relating to each test, or exact 
copies thereof, in accordance with the record requirements specified in 
Secs. 90.405 and 90.406; and
    (vii) A brief description of any significant events during testing 
not otherwise described under paragraph (a)(2) of this section, 
commencing with the test engine selection process and including such 
extraordinary events as engine damage during shipment.
    (3) The manufacturer must establish, maintain and retain general 
records, pursuant to paragraph (a)(1) of this section, for each test 
cell that can be used to perform emission testing under this subpart.
    (b) The manufacturer must retain all records required to be 
maintained under this subpart for a period of one year after completion 
of all testing required for the engine family in a model year. Records 
may be retained as hard copy (i.e., on paper) or reduced to microfilm, 
floppy disk, or some other method of data storage, depending upon the 
manufacturer's record retention procedure; provided, that in every 
case, all the information contained in the hard copy is retained.
    (c) The manufacturer must, upon request by the Administrator, 
submit the following information with regard to engine production:
    (1) Projected production or actual production for each engine 
configuration within each engine family for which certification has 
been requested and/or approved;
    (2) Number of engines, by configuration and assembly plant, 
scheduled for production or actually produced.
    (d) Nothing in this section limits the Administrator's discretion 
to require a manufacturer to establish, maintain, retain or submit to 
EPA information not specified by this section.
    (e) All reports, submissions, notifications, and requests for 
approval made under this subpart must be addressed to: Manager, Engine 
Compliance Programs Group (6403J), U.S. Environmental Protection 
Agency, Washington, DC 20460.
    (f) The manufacturer must electronically submit the results of its 
production line testing using EPA's standardized format. The 
Administrator may exempt manufacturers from this requirement upon 
written request with supporting justification.


Sec. 90.705  Right of entry and access.

    (a) To allow the Administrator to determine whether a manufacturer 
is complying with the provisions of this subpart or other subparts of 
this part, one or more EPA enforcement officers may enter during 
operating hours and upon presentation of credentials any of the 
following places:
    (1) Any facility, including ports of entry, where any engine to be 
introduced into commerce or any emission-related component is 
manufactured, assembled, or stored;
    (2) Any facility where any test conducted pursuant to this or any 
other subpart or any procedure or activity connected with such test is 
or was performed;
    (3) Any facility where any test engine is present; and
    (4) Any facility where any record required under Sec. 90.704 or 
other document relating to this subpart or any other subpart of this 
part is located.
    (b) Upon admission to any facility referred to in paragraph (a) of 
this section, EPA enforcement officers are authorized to perform the 
following inspection-related activities:
    (1) To inspect and monitor any aspect of engine manufacture, 
assembly, storage, testing and other procedures, and to inspect and 
monitor the facilities in which these procedures are conducted;
    (2) To inspect and monitor any aspect of engine test procedures or 
activities, including test engine selection, preparation and service 
accumulation, emission test cycles, and maintenance and verification of 
test equipment calibration;
    (3) To inspect and make copies of any records or documents related 
to the assembly, storage, selection, and testing of an engine; and
    (4) To inspect and photograph any part or aspect of any engine and 
any component used in the assembly thereof that is reasonably related 
to the purpose of the entry.
    (c) EPA enforcement officers are authorized to obtain reasonable 
assistance without cost from those in charge of a facility to help the 
officers perform any function listed in this subpart and they are 
authorized to request the manufacturer to make arrangements with those 
in charge of a facility operated for the manufacturer's benefit to 
furnish reasonable assistance without cost to EPA.
    (1) Reasonable assistance includes, but is not limited to, 
clerical, copying, interpretation and translation services; the making 
available on an EPA enforcement officer's request of personnel of the 
facility being inspected during their working hours to inform the EPA 
enforcement officer of how the facility operates and to answer the 
officer's questions; and the performance on request of emission tests 
on any engine which is being, has been, or will be used for production 
line or other testing.
    (2) By written request, signed by the Assistant Administrator for 
Air and Radiation, and served on the manufacturer, a manufacturer may 
be compelled to cause the personal appearance of any employee at such a 
facility before an EPA enforcement officer. Any such employee who has 
been instructed by the manufacturer to appear will be entitled to be 
accompanied, represented, and advised by counsel.
    (d) EPA enforcement officers are authorized to seek a warrant or 
court order authorizing the EPA enforcement officers to conduct the 
activities authorized in this section, as appropriate, to execute the 
functions specified in this section. EPA enforcement officers may 
proceed ex parte to obtain a warrant or court order whether or not the 
EPA enforcement officers first attempted to seek permission from the 
manufacturer or the party in charge of the facility(ies) in question to 
conduct the activities authorized in this section.
    (e) A manufacturer must permit an EPA enforcement officer(s) who 
presents a warrant or court order to conduct the activities authorized 
in this section as described in the warrant or court order. The 
manufacturer must also cause those in charge of its facility or a 
facility operated for its benefit to permit entry and access as 
authorized in this section pursuant to a warrant or court order whether 
or not the manufacturer controls the facility. In the absence of a 
warrant or court order, an EPA enforcement officer(s) may conduct the 
activities authorized in this section only upon the consent of the 
manufacturer or the party in charge of the facility(ies) in question.
    (f) It is not a violation of this part or the Clean Air Act for any 
person to refuse to permit an EPA enforcement officer(s) to conduct the 
activities authorized in this section if the officer(s) appears without 
a warrant or court order.
    (g) A manufacturer is responsible for locating its foreign testing 
and manufacturing facilities in jurisdictions where local law does not 
prohibit an EPA enforcement officer(s) from conducting the entry and 
access activities specified in this section. EPA will not attempt to 
make any inspections which it has been informed local foreign law 
prohibits.


Sec. 90.706  Engine sample selection.

    (a) At the start of each model year, the small SI engine 
manufacturer will begin

[[Page 4022]]

to randomly select engines from each engine family for production line 
testing at a rate of one percent of the projected eligible sales of 
that family. Each engine will be selected from the end of the assembly 
line.
    (1) For newly certified engine families: After two engines are 
tested, the manufacturer will calculate the required sample size for 
the model year for each pollutant 
(HC+NOX(NMHC+NOX) and CO) according to the Sample 
Size Equation in paragraph (b) of this section.
    (2) For carry-over engine families: After one engine is tested, the 
manufacturer will combine the test with the last test result from the 
previous model year and then calculate the required sample size for the 
model year for each pollutant according to the Sample Size Equation in 
paragraph (b) of this section.
    (b)(1) Manufacturers will calculate the required sample size for 
the model year for each pollutant for each engine family using the 
Sample Size Equation in this paragraph. N is calculated for each 
pollutant from each test result. The higher of the two values for the 
number N indicates the number of tests required for the model year for 
an engine family. N is recalculated for each pollutant after each test. 
Test results used to calculate the variables in the following Sample 
Size Equation must be final deteriorated test results as specified in 
Sec. 90.709(c):
[GRAPHIC] [TIFF OMITTED] TP27JA98.002

    Where:

N = required sample size for the model year.
t95 = 95% confidence coefficient. It is dependent on the 
actual number of tests completed, n, as specified in the table in 
paragraph (b)(2) of this section. It defines one-tail, 95% confidence 
intervals.
o = actual test sample standard deviation calculated from the following 
equation:
[GRAPHIC] [TIFF OMITTED] TP27JA98.003

xii = emission test result for an individual engine.
x = mean of emission test results of the actual sample.
FEL = Family Emission Limit or standard if no FEL.
n = The actual number of tests completed in an engine family.

    (2) The following table specifies the actual number of tests (n) & 
1-tail confidence coefficients (t95):

------------------------------------------------------------------------
                                n                                   t95 
------------------------------------------------------------------------
2...............................................................  6.31  
3...............................................................  2.92  
4...............................................................  2.35  
5...............................................................  2.13  
6...............................................................  2.02  
7...............................................................  1.94  
8...............................................................  1.90  
9...............................................................  1.86  
10..............................................................  1.83  
11..............................................................  1.81  
12..............................................................  1.80  
13..............................................................  1.78  
14..............................................................  1.77  
15..............................................................  1.76  
16..............................................................  1.75  
17..............................................................  1.75  
18..............................................................  1.74  
19..............................................................  1.73  
20..............................................................  1.73  
21..............................................................  1.72  
22..............................................................  1.72  
23..............................................................  1.72  
24..............................................................  1.71  
25..............................................................  1.71  
26..............................................................  1.71  
27..............................................................  1.71  
28..............................................................  1.70  
29..............................................................  1.70  
30..............................................................  1.70  
................................................................  1.645 
------------------------------------------------------------------------

    (3) A manufacturer must distribute the testing of the remaining 
number of engines needed to meet the required sample size N, evenly 
throughout the remainder of the model year.
    (4) After each new test, the required sample size, N, is 
recalculated using updated sample means, sample standard deviations and 
the appropriate 95% confidence coefficient.
    (5) A manufacturer must continue testing and updating each engine 
family's sample size calculations according to paragraphs (b)(1) 
through (b)(4) of this section until a decision is made to stop testing 
as described in paragraph (b)(6) of this section or a noncompliance 
decision is made pursuant to Sec. 90.710(b).
    (6) If, at any time throughout the model year, the calculated 
required sample size, N, for an engine family is less than or equal to 
the actual sample size, n, and the sample mean, x, for HC + 
NOX (NMHC+NOX) and CO is less than or equal to 
the FEL or standard if no FEL, the manufacturer may stop testing that 
engine family.
    (7) If, at any time throughout the model year, the sample mean, x, 
for HC + NOX (NMHC+NOX) or CO is greater than the 
FEL or standard if no FEL, the manufacturer must continue testing that 
engine family at the appropriate maximum sampling rate.
    (8) The maximum required sample size for an engine family 
(regardless of the required sample size, N, as calculated in paragraph 
(b)(1) of this section) is the lesser of thirty tests per model year or 
one percent of projected annual production for that engine family for 
that model year.
    (9) Manufacturers may elect to test additional engines. Additional 
engines, whether tested in accordance with the testing procedures 
specified in Sec. 90.707 or not, may not be included in the Sample Size 
and Cumulative Sum equation calculations as defined in paragraph (b)(1) 
of this section and Sec. 90.708(a), respectively. However, such 
additional test results may be used as appropriate to ``bracket'' or 
define the boundaries of the production duration of any emission 
nonconformity determined under this subpart. Such additional test data 
must be identified and provided to EPA with the submittal of the 
official CumSum results.
    (c) The manufacturer must produce and assemble the test engines 
using its normal production and assembly process for engines to be 
distributed into commerce.
    (d) No quality control, testing, or assembly procedures shall be 
used on any test engine or any portion thereof, including parts and 
subassemblies, that have not been or will not be used during the 
production and assembly of all other engines of that family, unless the 
Administrator approves the modification in production or assembly 
procedures in advance.


Sec. 90.707  Test procedures.

    (a)(1) For small SI engines subject to the provisions of this 
subpart, the prescribed test procedures are specified in subpart E of 
this part.
    (2) The Administrator may, on the basis of a written application by 
a manufacturer, prescribe test procedures other than those specified in 
paragraph (a)(1) of this section for any small SI engine the 
Administrator determines is not susceptible to satisfactory testing 
using procedures specified in paragraph (a)(1) of this section.
    (b)(1) The manufacturer may not adjust, repair, prepare, or modify 
any test engine and may not perform any emission test on any test 
engine unless this adjustment, repair, preparation, modification and/or 
test is documented in the manufacturer's engine assembly and inspection 
procedures and is actually performed by the manufacturer on every 
production line engine or unless this adjustment, repair, preparation, 
modification and/or test is required or permitted under this subpart or 
is approved in advance by the Administrator.
    (2) The Administrator may adjust or cause to be adjusted any engine 
parameter which the Administrator has

[[Page 4023]]

determined to be subject to adjustment for certification, Production 
Line Testing and Selective Enforcement Audit testing, to any setting 
within the physically adjustable range of that parameter, as determined 
by the Administrator, prior to the performance of any test. However, if 
the idle speed parameter is one which the Administrator has determined 
to be subject to adjustment, the Administrator may not adjust it or 
require that it be adjusted to any setting which causes a lower engine 
idle speed than would have been possible within the physically 
adjustable range of the idle speed parameter if the manufacturer had 
accumulated 12 hours of service on the engine under paragraph (c) of 
this section, all other parameters being identically adjusted for the 
purpose of the comparison. The manufacturer may be requested to supply 
information necessary to establish an alternate minimum idle speed. The 
Administrator, in making or specifying these adjustments, may consider 
the effect of the deviation from the manufacturer's recommended setting 
on emission performance characteristics as well as the likelihood that 
similar settings will occur on in-use engines. In determining 
likelihood, the Administrator may consider factors such as, but not 
limited to, the effect of the adjustment on engine performance 
characteristics and information from similar in-use engines.
    (c) Service Accumulation. (1) Unless otherwise approved by the 
Administrator, prior to performing exhaust emission production line 
testing, the manufacturer may accumulate on each test engine a number 
of hours of service equal to the greater of 12 hours or the number of 
hours the manufacturer accumulated during stabilization in the 
certification process for each engine family. For catalyst-equipped 
engines, the manufacturer must accumulate a number of hours equal to 
the number of hours accumulated to represent stabilized emissions on 
the engine used to obtain certification.
    (2) Service accumulation must be performed in a manner using good 
engineering judgment to obtain emission results representative of 
production line engines.
    (d) Unless otherwise approved by the Administrator, the 
manufacturer may not perform any maintenance on test engines after 
selection for testing.
    (e) If an engine is shipped to a remote facility for production 
line testing, and an adjustment or repair is necessary because of 
shipment, the engine manufacturer must perform the necessary adjustment 
or repair only after the initial test of the engine, except in cases 
where the Administrator has determined that the test would be 
impossible or unsafe to perform or would permanently damage the engine. 
Engine manufacturers must report to the Administrator, in the quarterly 
report required by Sec. 90.709(e), all adjustments or repairs performed 
on test engines prior to each test.
    (f) If an engine cannot complete the service accumulation or an 
emission test because of a malfunction, the manufacturer may request 
that the Administrator authorize either the repair of that engine or 
its deletion from the test sequence.
    (g) Testing. A manufacturer must test engines with the test 
procedure specified in subpart E of this part to demonstrate compliance 
with the applicable FEL (or standard where there is no FEL). If 
alternate or special test procedures pursuant to regulations at 
Sec. 90.120 are used in certification, then those alternate procedures 
must be used in production line testing.
    (h) Retesting. (1) If an engine manufacturer reasonably determines 
that an emission test of an engine is invalid because of a procedural 
error, test equipment problem, or engine performance problem that 
causes the engine to be unable to safely perform a valid test, the 
engine may be retested. A test is not invalid simply because the 
emission results are high relative to other engines of the family. 
Emission results from all tests must be reported to EPA. The engine 
manufacturer must also include a detailed explanation of the reasons 
for invalidating any test in the quarterly report required in 
Sec. 90.709(e). If a test is invalidated because of an engine 
performance problem, the manufacturer must document in detail the 
nature of the problem and the repairs performed in order to use the 
after-repair test results for the original test results.
    (2) Routine retests may be conducted if the manufacturer conducts 
the same number of tests on all engines in the family. The results of 
these tests must be averaged according to procedures of Sec. 90.709.


Sec. 90.708  Cumulative Sum (CumSum) Procedure.

    (a) (1) Manufacturers must construct separate CumSum Equations for 
each regulated pollutant (HC+NOX (NMHC+NOX) and 
CO) for each engine family. Test results used to calculate the 
variables in the CumSum Equations must be final deteriorated test 
results as defined in Sec. 90.709(c). The CumSum Equation follows:

Ci=max [0 or (Ci-1+Xi-(FEL+F))]

Where:

Ci=The current CumSum statistic.
Ci-1=The previous CumSum statistic. Prior to any testing, 
the CumSum statistic=0 (i.e. C0=0).
Xi=The current emission test result for an individual 
engine.
FEL=Family Emission Limit (the standard if no FEL).
F=0.25 x .

    (2) After each test pursuant to paragraph (a)(1) of this section, 
Ci is compared to the action limit, H, the quantity which 
the CumSum statistic must exceed, in two consecutive tests, before the 
engine family may be determined to be in noncompliance for a regulated 
pollutant for purposes of Sec. 90.710.

Where:

H=The Action Limit. It is 5.0 x , and is a function of the 
standard deviation, .
=is the sample standard deviation and is recalculated after 
each test.

    (b) After each engine is tested, the CumSum statistic shall be 
promptly updated according to the CumSum Equation in paragraph (a) of 
this section.
    (c)(1) If, at any time during the model year, a manufacturer amends 
the application for certification for an engine family as specified in 
Sec. 90.122(a) by performing an engine family modification (i.e., a 
change such as a running change involving a physical modification to an 
engine, a change in specification or setting, the addition of a new 
configuration, or the use of a different deterioration factor) with no 
changes to the FEL (where applicable), all previous sample size and 
CumSum statistic calculations for the model year will remain unchanged.
    (2) If, at any time during the model year, a manufacturer amends 
the application for certification for an engine family as specified in 
Sec. 90.122(a) by modifying its FEL (where applicable) for future 
production, as a result of an engine family modification, the 
manufacturer must continue its calculations by inserting the new FEL 
into the sample size equation as specified in Sec. 90.706(b)(1) and 
into the CumSum equation in paragraph (a) of this section. All previous 
calculations remain unchanged. If the sample size calculation indicates 
that additional tests are required, then those tests must be performed. 
CumSum statistic calculations must not indicate that the family has 
exceeded the action limit for two consecutive tests. Where applicable, 
the manufacturer's final credit report as required by Sec. 90.210 must 
break out the

[[Page 4024]]

credits that result from each FEL and corresponding CumSum analysis for 
the set of engines built to each FEL.
    (3) If, at any time during the model year, a manufacturer amends 
the application for certification for an engine family as specified in 
Sec. 90.122(a) (or for an affected part of the year's production in 
cases where there were one or more mid-year engine family 
modifications), by modifying its FEL (where applicable) for past and/or 
future production, without performing an engine modification, all 
previous sample size and CumSum statistic calculations for the model 
year must be recalculated using the new FEL. If the sample size 
calculation indicates that additional tests are required, then those 
tests must be performed. The CumSum statistic recalculation must not 
indicate that the family has exceeded the action limit for two 
consecutive tests. Where applicable, the manufacturer's final credit 
report as required by Sec. 90.210 must break out the credits that 
result from each FEL and corresponding CumSum analysis for the set of 
engines built to each FEL.


Sec. 90.709  Calculation and reporting of test results.

    (a) Initial test results are calculated following the applicable 
test procedure specified in Sec. 90.707(a). The manufacturer rounds 
these results to the number of decimal places contained in the 
applicable emission standard expressed to one additional significant 
figure.
    (b) Final test results are calculated by summing the initial test 
results derived in paragraph (a) of this section for each test engine, 
dividing by the number of tests conducted on the engine, and rounding 
to the same number of decimal places contained in the applicable 
standard expressed to one additional significant figure.
    (c) The final deteriorated test results for each test engine are 
calculated by applying the appropriate deterioration factors, derived 
in the certification process for the engine to the final test results, 
and rounding to the same number of decimal places contained in the 
applicable standard.
    (d) If, at any time during the model year, the CumSum statistic 
exceeds the applicable action limit, H, in two consecutive tests for 
any regulated pollutant, (HC+NOX (NMHC+NOX) or 
CO) the engine family may be determined to be in noncompliance and the 
manufacturer must notify EPA within two working days of such exceedance 
by the Cum Sum statistic.
    (e) Within 30 calendar days of the end of each quarter, each engine 
manufacturer must submit to the Administrator a report which includes 
the following information:
    (1) The location and description of the manufacturer's or other's 
exhaust emission test facilities which were utilized to conduct testing 
reported pursuant to this section;
    (2) Total production and sample sizes, N and n, for each engine 
family;
    (3) The FEL (standard, if no FEL) against which each engine family 
was tested;
    (4) A description of the process to obtain engines on a random 
basis;
    (5) A description of the test engines;
    (6) For each test conducted:
    (i) A description of the test engine, including:
    (A) Configuration and engine family identification;
    (B) Year, make, and build date;
    (C) Engine identification number; and
    (D) Number of hours of service accumulated on engine prior to 
testing;
    (ii) Location where service accumulation was conducted and 
description of accumulation procedure and schedule;
    (iii) Test number, date, test procedure used, initial test results 
before and after rounding, final test results before and after rounding 
and final deteriorated test results for all exhaust emission tests, 
whether valid or invalid, and the reason for invalidation, if 
applicable;
    (iv) A complete description of any adjustment, modification, 
repair, preparation, maintenance, and/or testing which was performed on 
the test engine, was not reported pursuant to any other paragraph of 
this subpart, and will not be performed on all other production 
engines;
    (v) A CumSum analysis, as required in Sec. 90.708, of the 
production line test results for each engine family; and
    (vi) Any other information the Administrator may request relevant 
to the determination whether the new engines being manufactured by the 
manufacturer do in fact conform with the regulations with respect to 
which the certificate of conformity was issued;
    (7) For each failed engine as defined in Sec. 90.710(a), a 
description of the remedy and test results for all retests as required 
by Sec. 90.711(g);
    (8) The date of the end of the engine manufacturer's model year 
production for each engine family; and
    (9) The following signed statement and endorsement by an authorized 
representative of the manufacturer:
    This report is submitted pursuant to Sections 213 and 208 of the 
Clean Air Act. This production line testing program was conducted in 
complete conformance with all applicable regulations under 40 CFR Part 
90. No emission-related changes to production processes or quality 
control procedures for the engine family tested have been made during 
this production line testing program that affect engines from the 
production line. All data and information reported herein is, to the 
best of (Company Name) knowledge, true and accurate. I am aware of the 
penalties associated with violations of the Clean Air Act and the 
regulations thereunder. (Authorized Company Representative.)


Sec. 90.710  Compliance with criteria for production line testing.

    (a) A failed engine is one whose final deteriorated test results 
pursuant to Sec. 90.709(c), for HC + NOX 
(NMHC+NOX) or CO exceeds the applicable Family Emission 
Limit (FEL)or standard if no FEL.
    (b) An engine family shall be determined to be in noncompliance, if 
at any time throughout the model year, the CumSum statistic, 
Ci, for HC + NOX (NMHC+NOX) or CO, is 
greater than the action limit, H, for that pollutant, for two 
consecutive tests.


Sec. 90.711  Suspension and revocation of certificates of conformity.

    (a) The certificate of conformity is suspended with respect to any 
engine failing pursuant to Sec. 90.710 (a) effective from the time that 
testing of that engine is completed.
    (b) The Administrator may suspend the certificate of conformity for 
an engine family which is determined to be in noncompliance pursuant to 
Sec. 90.710(b). This suspension will not occur before thirty days after 
the engine family is determined to be in noncompliance and the 
Administrator has notified the manufacturer of its intent to suspend. 
During this thirty day period the Administrator will work with the 
manufacturer to achieve appropriate production line changes to avoid 
the need to halt engine production, if possible. The Administrator will 
approve or disapprove any such production line changes proposed to 
address a family that has been determined to be in noncompliance under 
this subpart within 15 days of receipt. If the Administrator does not 
approve or disapprove such a proposed change within such time period, 
the proposed change shall be considered approved.
    (c) If the results of testing pursuant to these regulations 
indicate that engines of a particular family produced at one plant of a 
manufacturer do not conform to the regulations with respect to which 
the certificate of conformity was issued,

[[Page 4025]]

the Administrator may suspend the certificate of conformity with 
respect to that family for engines manufactured by the manufacturer at 
all other plants.
    (d) Notwithstanding the fact that engines described in the 
application for certification may be covered by a certificate of 
conformity, the Administrator may suspend such certificate immediately 
in whole or in part if the Administrator finds any one of the following 
infractions to be substantial:
    (1) The manufacturer refuses to comply with any of the requirements 
of this subpart.
    (2) The manufacturer submits false or incomplete information in any 
report or information provided to the Administrator under this subpart.
    (3) The manufacturer renders inaccurate any test data submitted 
under this subpart.
    (4) An EPA enforcement officer is denied the opportunity to conduct 
activities authorized in this subpart and a warrant or court order is 
presented to the manufacturer or the party in charge of the facility in 
question.
    (5) An EPA enforcement officer is unable to conduct activities 
authorized in Sec. 90.705 because a manufacturer has located its 
facility in a foreign jurisdiction where local law prohibits those 
activities.
    (e) The Administrator shall notify the manufacturer in writing of 
any suspension or revocation of a certificate of conformity in whole or 
in part, except that the certificate is immediately suspended with 
respect to any failed engines as provided for in paragraph (a) of this 
section.
    (f) The Administrator may revoke a certificate of conformity for an 
engine family after the certificate has been suspended pursuant to 
paragraph (b) or (c) of this section if the proposed remedy for the 
nonconformity, as reported by the manufacturer to the Administrator, is 
one requiring a design change or changes to the engine and/or emission 
control system as described in the application for certification of the 
affected engine family.
    (g) Once a certificate has been suspended for a failed engine, as 
provided for in paragraph (a) of this section, the manufacturer must 
take the following actions before the certificate is reinstated for 
that failed engine:
    (1) Remedy the nonconformity;
    (2) Demonstrate that the engine conforms to the applicable 
standards (FELs, where applicable) by retesting the engine in 
accordance with these regulations; and
    (3) Submit a written report to the Administrator, after successful 
completion of testing on the failed engine, which contains a 
description of the remedy and test results for each engine in addition 
to other information that may be required by this part.
    (h) Once a certificate for a failed engine family has been 
suspended pursuant to paragraph (b) or (c) of this section, the 
manufacturer must take the following actions before the Administrator 
will consider reinstating the certificate:
    (1) Submit a written report to the Administrator which identifies 
the reason for the noncompliance of the engines, describes the proposed 
remedy, including a description of any proposed quality control and/or 
quality assurance measures to be taken by the manufacturer to prevent 
future occurrences of the problem, and states the date on which the 
remedies will be implemented; and
    (2) Demonstrate that the engine family for which the certificate of 
conformity has been suspended does in fact comply with the regulations 
of this part by testing as many engines as needed so that the CumSum 
statistic, as calculated in Sec. 90.708(a), falls below the action 
limit. Such testing must comply with the provisions of this part. If 
the manufacturer elects to continue testing individual engines after 
suspension of a certificate, the certificate is reinstated for any 
engine actually determined to be in conformance with the Family 
Emission Limits (or standards if no FEL) through testing in accordance 
with the applicable test procedures, provided that the Administrator 
has not revoked the certificate pursuant to paragraph (f) of this 
section.
    (i) Once the certificate has been revoked for an engine family, if 
the manufacturer desires to continue introduction into commerce of a 
modified version of that family, the following actions must be taken 
before the Administrator may issue a certificate for that modified 
family:
    (1) If the Administrator determines that the proposed change(s) in 
engine design may have an effect on emission performance deterioration, 
the Administrator shall notify the manufacturer within five working 
days after receipt of the report in paragraph (h)(1) of this section 
whether subsequent testing under this subpart will be sufficient to 
evaluate the proposed change or changes or whether additional testing 
will be required;
    (2) After implementing the change or changes intended to remedy the 
nonconformity, the manufacturer must demonstrate that the modified 
engine family does in fact conform with the regulations of this part by 
testing as many engines as needed from the modified engine family so 
that the CumSum statistic, as calculated in Sec. 90.708(a) using the 
newly assigned FEL if applicable, falls below the action limit; and
    (3) When the requirements of paragraphs (i)(1) and (i)(2) of this 
section are met, the Administrator shall reissue the certificate or 
issue a new certificate, as the case may be, to include that family. As 
long as the CumSum statistic remains above the action limit, the 
revocation remains in effect.
    (j) At any time subsequent to a suspension of a certificate of 
conformity for a test engine pursuant to paragraph (a) of this section, 
but not later than 15 days (or such other period as may be allowed by 
the Administrator) after notification of the Administrator's decision 
to suspend or revoke a certificate of conformity in whole or in part 
pursuant to paragraph (b), (c), or (f) of this section, a manufacturer 
may request a hearing as to whether the tests have been properly 
conducted or any sampling methods have been properly applied.
    (k) Any suspension of a certificate of conformity under paragraph 
(d) of this section shall:
    (1) Be made only after the manufacturer concerned has been offered 
an opportunity for a hearing conducted in accordance with Secs. 90.712 
and 90.713; and
    (2) Not apply to engines no longer in the possession of the 
manufacturer.
    (l) After the Administrator suspends or revokes a certificate of 
conformity pursuant to this section and prior to the commencement of a 
hearing under Sec. 90.712, if the manufacturer demonstrates to the 
Administrator's satisfaction that the decision to suspend or revoke the 
certificate was based on erroneous information, the Administrator shall 
reinstate the certificate.
    (m) To permit a manufacturer to avoid storing non-test engines 
while conducting subsequent testing of the noncomplying family, a 
manufacturer may request that the Administrator conditionally reinstate 
the certificate for that family. The Administrator may reinstate the 
certificate subject to the following condition: the manufacturer must 
commit to performing offsetting measures that remedy the nonconformity 
at no expense to the owners, and which are approved in advance by the 
Administrator for all engines of that family produced from the time the 
certificate is conditionally reinstated if the CumSum statistic does 
not fall below the action limit.

[[Page 4026]]

Sec. 90.712  Request for public hearing.

    (a) If the manufacturer disagrees with the Administrator's decision 
to suspend or revoke a certificate or disputes the basis for an 
automatic suspension pursuant to Sec. 90.711(a), the manufacturer may 
request a public hearing.
    (b) The manufacturer's request shall be filed with the 
Administrator not later than 15 days after the Administrator's 
notification of his or her decision to suspend or revoke, unless 
otherwise specified by the Administrator. The manufacturer shall 
simultaneously serve two copies of this request upon the Manager of the 
Engine Compliance Programs Group and file two copies with the Hearing 
Clerk for the Agency. Failure of the manufacturer to request a hearing 
within the time provided constitutes a waiver of the right to a 
hearing. Subsequent to the expiration of the period for requesting a 
hearing as of right, the Administrator may, in his or her discretion 
and for good cause shown, grant the manufacturer a hearing to contest 
the suspension or revocation.
    (c) A manufacturer shall include in the request for a public 
hearing:
    (1) A statement as to which engine configuration(s) within a family 
is to be the subject of the hearing; and
    (2) A concise statement of the issues to be raised by the 
manufacturer at the hearing, except that in the case of the hearing 
requested under Sec. 90.711(j), the hearing is restricted to the 
following issues:
    (i) Whether tests have been properly conducted (specifically, 
whether the tests were conducted in accordance with applicable 
regulations under this part and whether test equipment was properly 
calibrated and functioning);
    (ii) Whether sampling plans and statistical analyses have been 
properly applied (specifically, whether sampling procedures and 
statistical analyses specified in this subpart were followed and 
whether there exists a basis for distinguishing engines produced at 
plants other than the one from which engines were selected for testing 
which would invalidate the Administrator's decision under 
Sec. 90.711(c));
    (3) A statement specifying reasons why the manufacturer believes it 
will prevail on the merits of each of the issues raised; and
    (4) A summary of the evidence which supports the manufacturer's 
position on each of the issues raised.
    (d) A copy of all requests for public hearings will be kept on file 
in the Office of the Hearing Clerk and will be made available to the 
public during Agency business hours.


Sec. 90.713  Administrative procedures for public hearing.

    The administrative procedures for a public hearing requested under 
this subpart shall be those procedures set forth in the regulations 
found at Secs. 90.513 through 90.516. References in Sec. 90.513 to 
Sec. 90.511(j), Sec. 90.512(c)(2), Sec. 90.511(e), Sec. 90.512, 
Sec. 90.511(d), Sec. 90.503, Sec. 90.512(c) and Sec. 90.512(b) shall be 
deemed to refer to Sec. 90.711(j), Sec. 90.712(c)(2), Sec. 90.711(e), 
Sec. 90.712, Sec. 90.711(d), Sec. 90.703, and Sec. 90.712(c) and 
Sec. 90.712(b), respectively. References to ``test orders'' in 
Sec. 90.513 can be ignored.
    31. Subpart I is amended by revising the subpart heading to read as 
follows:

Subpart I--Emission-related Defect Reporting Requirements, 
Voluntary Emission Recall Program, Ordered Recalls

    32. Section 90.801 is amended by designating the existing text as 
paragraph (a) and adding paragraphs (b), (c), (d), (e), (f) and (g) to 
read as follows:


Sec. 90.801  Applicability.

* * * * *
    (b) Phase 2 engines subject to provisions of subpart B of this part 
are subject to recall regulations specified in 40 CFR part 85, subpart 
S, except as otherwise provided in this section.
    (c) Reference to section 214 of the Clean Air Act in 40 CFR 85.1801 
(a) is deemed to be a reference to section 216 of the Clean Air Act.
    (d) Reference to section 202 of the Act in 40 CFR 85.1802(a) is 
deemed to be a reference to section 213 of the Act.
    (e) Reference to ``family particulate emission limits as defined in 
part 86 promulgated under section 202 of the Act'' in 40 CFR 85.1803(a) 
and 85.1805(a)(1) is deemed to be a reference to ``family emission 
limits as defined in subpart C of this part 90 promulgated under 
section 213 of the Act''.
    (f) Reference to ``vehicles or engines'' throughout 40 CFR part 85, 
subpart S, is deemed to be a reference to ``Phase 2 nonroad small SI 
engines at or below 19 kw.''
    (g) In addition to the requirements in 40 CFR 85.1805(a)(9) for 
Phase 2 engines include a telephone number which may be used to report 
difficulty in obtaining recall repairs.
    33. Section 90.802 is amended by adding a sentence at the end of 
the introductory text to read as follows:


Sec. 90.802  Definitions.

    * * * The definitions of 40 CFR 85.1801 also apply to this part.
* * * * *
    34. Section 90.803 is amended by revising paragraph (c) to read as 
follows:


Sec. 90.803  Emission defect information report.

* * * * *
    (c) The manufacturer must submit defect information reports to 
EPA's Engine Compliance Programs Group not more than 15 working days 
after an emission-related defect is found to affect 25 or more engines 
manufactured in the same certificate or model year. Information 
required by paragraph (d) of this section that is either not available 
within 15 working days or is significantly revised must be submitted to 
EPA's Engine Compliance Programs Group as it becomes available.
* * * * *
    35. Section 90.805 is amended by revising paragraph (a) to read as 
follows:


Sec. 90.805  Reports, voluntary recall plan filing, record retention.

    (a) Send the defect report, voluntary recall plan, and the 
voluntary recall progress report to: Group Manager, Engine Compliance 
Programs Group, (6403-J), Environmental Protection Agency, Washington, 
D.C. 20460.
* * * * *
    36. A new Sec. 90.808 is added to subpart I to read as follows


Sec. 90.808  Ordered recall provisions.

    (a) Effective with respect to Phase 2 small SI engines:
    (1) If the Administrator determines that a substantial number of 
any class or category of engines, although properly maintained and 
used, do not conform to the regulations prescribed under section 213 of 
the Act when in actual use throughout their useful life (as defined 
under Sec. 90.105), the Administrator shall immediately notify the 
manufacturer of such nonconformity and require the manufacturer to 
submit a plan for remedying the nonconformity of the engines with 
respect to which such notification is given.
    (i) The manufacturer's plan shall provide that the nonconformity of 
any such engines which are properly used and maintained will be 
remedied at the expense of the manufacturer.
    (ii) If the manufacturer disagrees with such determination of 
nonconformity and so advises the Administrator, the Administrator shall 
afford the manufacturer and other interested persons an opportunity to 
present their views and evidence in support thereof at a public 
hearing. Unless, as a result of such hearing, the Administrator 
withdraws such determination of nonconformity, the Administrator shall, 
within 60 days after the completion of

[[Page 4027]]

such hearing, order the manufacturer to provide prompt notification of 
such nonconformity in accordance with paragraph (a)(2) of this section. 
The manufacturer shall comply in all respects with the requirements of 
this subpart.
    (2) Any notification required to be given by the manufacturer under 
paragraph (a)(1) of this section with respect to any class or category 
of engines shall be given to dealers, ultimate purchasers, and 
subsequent purchasers (if known) in such manner and containing such 
information as required in subparts I and M of this part.
    (3)(i) Prior to an EPA ordered recall, the manufacturer may perform 
a voluntary emissions recall pursuant to regulations at Sec. 90.804. 
Such manufacturer is subject to the reporting and recordkeeping 
requirements of Sec. 90.805.
    (ii) Once EPA determines that a substantial number of engines fail 
to conform with the requirements of section 213 of the Act or this 
part, the manufacturer will not have the option of a voluntary recall.
    (b) The manufacturer bears all cost obligation a dealer incurs as a 
result of a requirement imposed by paragraph (a) of this section. The 
transfer of any such cost obligation from a manufacturer to a dealer 
through franchise or other agreement is prohibited.
    (c) Any inspection of an engine for purposes of paragraph (a)(1) of 
this section, after its sale to the ultimate purchaser, is to be made 
only if the owner of such vehicle or engine voluntarily permits such 
inspection to be made, except as may be provided by any state or local 
inspection program.

Subpart J--Exclusion and Exemption of Nonroad Engines From 
Regulations

    37. Section 90.905 is amended by revising paragraph (f) to read as 
follows:


Sec. 90.905  Testing exemption.

* * * * *
    (f) A manufacturer of new nonroad engines may request a testing 
exemption to cover nonroad engines intended for use in test programs 
planned or anticipated over the course of a subsequent one-year period. 
Unless otherwise required by the Director, Engine Programs and 
Compliance Division, a manufacturer requesting such an exemption need 
only furnish the information required by paragraphs (a)(1) and (d)(2) 
of this section along with a description of the recordkeeping and 
control procedures that will be employed to assure that the engines are 
used for purposes consistent with Sec. 90.1004(b).
    38. Section 90.906 is amended by revising paragraphs (a) 
introductory text and (a)(3) introductory text to read as follows:


Sec. 90.906  Manufacturer-owned exemption and precertification 
exemption.

    (a) Any manufacturer owned nonroad engine, as defined by 
Sec. 90.902, is exempt from Sec. 90.1003, without application, if the 
manufacturer complies with the following terms and conditions:
* * * * *
    (3) Unless the requirement is waived or an alternative procedure is 
approved by the Director, Engine Programs and Compliance Division, the 
manufacturer must permanently affix a label to each nonroad engine on 
exempt status. This label should:
* * * * *
    39. Section 90.909 is amended by revising paragraph (c) to read as 
follows:


Sec. 90.909  Export exemptions.

* * * * *
    (c) EPA will maintain a list of foreign countries that have in 
force nonroad emission standards identical to U.S. EPA standards and 
have so notified EPA. This list may be obtained by writing to the 
following address: Group Manager, Engine Compliance Programs Group, 
Engine Programs and Compliance Division (6403-J), Environmental 
Protection Agency, Washington, D.C. 20460. New nonroad engines exported 
to such countries must comply with U.S. EPA certification regulations.
* * * * *
    40. Section 90.911 is revised to read as follows:


Sec. 90.911  Submission of exemption requests.

    Requests for exemption or further information concerning exemptions 
and/or the exemption request review procedure should be addressed to: 
Group Manager, Engine Compliance Programs Group, Engine Programs and 
Compliance Division (6403J), Environmental Protection Agency, 
Washington, D.C. 20460.

Subpart K--Prohibited Acts and General Enforcement Provisions

    41. Section 90.1003 is amended by revising paragraphs (a)(2), 
(a)(4)(i), (b)(4), and (b)(5) and by redesignating paragraphs 
(a)(4)(iii) and (a)(4)(iv) as paragraphs (a)(4)(iv) and (a)(4)(v) 
respectively, and by adding new paragraphs (a)(4)(iii) and (b)(6) to 
read as follows:


Sec. 90.1003  Prohibited acts.

    (a) * * *
    (2) (i) For a person to fail or refuse to permit access to or 
copying of records or to fail to make reports or provide information 
required under Sec. 90.1004.
    (ii) For a person to fail or refuse to permit entry, testing or 
inspection authorized under Secs. 90.126, 90.506, 90.705, 90.1004, or 
90.1209.
    (iii) For a person to fail or refuse to perform tests or to have 
tests performed as required under Secs. 90.119, 90.504, 90.703, 
90.1004, 90.1203, or 90.1250.
    (iv) For a person to fail to establish or maintain records as 
required under Secs. 90.209, 90.704, 90.805, 90.1004, or 90.1308.
    (v) For a person to fail to submit a remedial plan as required 
under Sec. 90.808.
* * * * *
    (4) * * *
    (i) To sell, offer for sale, or introduce or deliver into commerce, 
a nonroad engine unless the manufacturer has complied with the 
requirements of Sec. 90.1103.
* * * * *
    (iii) To fail or refuse to comply with the requirements of 
Sec. 90.808.
* * * * *
    (b) * * *
    (4) Certified nonroad engines shall be used in all equipment or 
vehicles that are self-propelled, portable, transportable, or are 
intended to be propelled while performing their function, unless the 
manufacturer of the equipment or vehicle can prove that the vehicle or 
equipment will be used in a manner consistent with paragraph (2) of the 
definition of nonroad engine in Sec. 90.3. Nonroad vehicle and 
equipment manufacturers may continue to use noncertified nonroad 
engines built prior to the applicable implementation date of the Phase 
1 regulations in this part until noncertified engine inventories are 
depleted; further after the applicable implementation date of the Phase 
2 regulations in this part, nonroad vehicle and equipment manufacturers 
may continue to use Phase 1 engines until Phase 1 engine inventories 
are depleted. Stockpiling (i.e., build up of an inventory of 
uncertified engines or Phase 1 engines beyond normal business practices 
to avoid or delay compliance with the Phase 1 or Phase 2 regulations in 
this part, respectively) will be considered a violation of this 
section.
    (5) A new nonroad engine, intended solely to replace an engine in a 
piece of nonroad equipment that was originally produced with an engine 
manufactured prior to the applicable implementation

[[Page 4028]]

date as described in Secs. 90.2, 90.103 and 90.106, or with an engine 
that was originally produced in a model year in which less stringent 
standards under this part were in effect, shall not be subject to the 
requirements of Sec. 90.106 or prohibitions and provisions of 
paragraphs (a)(1) and (b)(4) of this section provided that:
    (i) The engine manufacturer has ascertained that no engine produced 
by itself or the manufacturer of the engine that is being replaced, if 
different, and certified to the requirements of this subpart, is 
available with the appropriate physical or performance characteristics 
to repower the equipment. Certified engines may be ascertained to lack 
appropriate physical characteristics where the engine is too large for 
the engine compartment or can not be connected to existing manifolds, 
air supplies, water supplies, fuel supplies or controls without 
modifications that add substantial cost or result in reliability or 
safety concerns. Certified engines may be ascertained to lack 
appropriate performance characteristics if the horsepower or rated 
speed of the engine are significantly different from the original 
engine to reduce the ability of the equipment to perform its function 
safely and efficiently; and
    (ii) The engine manufacturer or its agent:
    (A) Accepts the old engine in exchange for the new engine and 
destroys the old engine; or
    (B) Obtains documentation from the purchaser sufficient to identify 
the old engine and prove that the purchaser has had the old engine 
destroyed by a separate party; and
    (iii) The engine manufacturer retains records of the engine 
purchasers and the makes and models of equipment for which the engines 
are sold. Such records shall be made available to the Administrator 
upon request and shall be sufficient to enable the Administrator to 
determine the quantities of engines being applied to different makes 
and models of equipment; and
    (iv) The engine manufacturer submits a written report to EPA, 
within 90 days of the end of each model year in which any uncertified 
replacement engines, or engines certified to an earlier model year's 
standards, were sold describing the numbers of such engines sold during 
the model year; and
    (v) The engine manufacturer has determined and documented that the 
engine being replaced was no older than ten (10) years old or ten (10) 
model years old; and
    (vi) The replacement engine is clearly labeled with the following 
language, or similar alternate language approved in advance by the 
Administrator: ``THIS engine does not comply with Federal nonroad or 
on-highway emission requirements. Sale or installation of this engine 
for any purpose other than as a replacement engine in a nonroad vehicle 
or piece of nonroad equipment whose original engine was not certified, 
or was certified to less stringent emission standards than those that 
apply to the year of manufacture of this engine, is a violation of 
Federal law subject to civil penalty''; and
    (vii) Where the replacement engine is intended to replace an engine 
built after the applicable implementation date of regulations under 
this part, but built to less stringent emission standards than are 
currently applicable, the replacement engine shall be identical in all 
material respects to a certified configuration of the same or later 
model year as the engine being replaced.
    (6)(i) Regulations elsewhere in this part notwithstanding, for 
three model years after the phase-in of each set of Phase 2 standards; 
i.e. through the 2004 model year for Class I nonhandheld engines and 
through model year 2008 for handheld engines and Class II nonhandheld 
engines, small volume equipment manufacturers as defined in this part 
may continue to use, and engine manufacturers may continue to supply, 
engines certified to Phase 1 standards (or identified and labeled by 
their manufacturer to be identical to engines previously certified 
under Phase 1 standards), provided the equipment manufacturer has 
demonstrated to the satisfaction of the Administrator that no certified 
Phase 2 engine is available with suitable physical or performance 
characteristics to power a piece of nonhandheld equipment in production 
prior to the 2001 model year, or handheld equipment in production prior 
to the 2002 model year. The equipment manufacturer must also certify to 
the Administrator that the equipment model has not undergone any 
redesign which could have facilitated conversion of the equipment to 
accommodate a Phase 2 engine.
    (ii) Regulations elsewhere in this part notwithstanding, for the 
duration of the Phase 2 regulations in this part, equipment 
manufacturers who certify to the Administrator that annual eligible 
sales of a particular model of equipment will not exceed 500 for a 
nonhandheld model in production prior to the 2001 model year, or 2500 
for a handheld model in production prior to the 2002 model year, may 
continue to use in that model, and engine manufacturers may continue to 
supply, engines certified to Phase 1 requirements, (or identified and 
labeled by their manufacturer to be identical to engines previously 
certified under Phase 1 standards). To be eligible for this provision, 
the equipment manufacturer must have demonstrated to the satisfaction 
of the Administrator that no certified Phase 2 engine is available with 
suitable physical or performance characteristics to power the 
equipment. The equipment manufacturer must also certify to the 
Administrator that the equipment model has not undergone any redesign 
which could have facilitated conversion of the equipment to accommodate 
a Phase 2 engine.
    (iii) An equipment manufacturer which is unable to obtain suitable 
Phase 2 engines and which can not obtain relief under any other 
provision of this part, may, prior to the date on which the 
manufacturer would become in noncompliance with the requirement to use 
Phase 2 engines, apply to the Administrator to be allowed to continue 
using Phase 1 engines, through the 2002 model year for Class I engines 
and through the 2006 model year for Class II, III, IV and V engines, 
subject to the following criteria:
    (A) The inability to obtain Phase 2 engines is despite the 
manufacturer's best efforts and is the result of an extraordinary 
action on the part of the engine manufacturer that was outside the 
control of and could not be reasonably foreseen by the equipment 
manufacturer; such as canceled production or shipment, last minute 
certification failure, unforeseen engine cancellation, plant closing, 
work stoppage or other such circumstance; and
    (B) The inability to market the particular equipment will bring 
substantial economic hardship to the equipment manufacturer resulting 
in a major impact on the equipment manufacturer's solvency.
    (iv) The written permission from the Administrator to the equipment 
manufacturer shall serve as permission for the engine manufacturer to 
provide such Phase 1 engines required by the equipment manufacturers 
under this paragraph (b)(6). Such engines will not count against an 
engine manufacturer's final (100%) handheld phase-in percentage 
requirements, and are excluded from the nonhandheld certification, 
averaging, banking and trading program. As Phase 1 engines, these 
engines are exempt from Production Line Testing requirements under 
subpart H of this part and in-use testing requirements under subpart M 
of this part.

[[Page 4029]]

Subpart L--Emission Warranty and Maintenance Instructions

    42. Section 90.1103 is amended by the revising paragraph (b) to 
read as follows:


Sec. 90.1103  Emission warranty, warranty period.

* * * * *
    (b)(1) The manufacturer of each new Phase 1 small SI engine must 
warrant to the ultimate purchaser and each subsequent purchaser that 
the engine is designed, built and equipped so as to conform at the time 
of sale with applicable regulations under section 213 of the Act, and 
the engine is free from defects in materials and workmanship which 
cause such engine to fail to conform with applicable regulations for 
its warranty period.
    (2) The manufacturer of each new Phase 2 small SI engine must 
warrant to the ultimate purchaser and each subsequent purchaser that 
the engine is designed, built, and equipped so as to conform for its 
designated useful life with applicable regulations under section 213 of 
the Act, and is free from defects in materials and workmanship which 
cause such engine to fail to conform with applicable regulations for 
its warranty period.
* * * * *
    43. Section 90.1104 is amended by adding paragraph (e) to read as 
follows:


Sec. 90.1104  Furnishing of maintenance instructions to ultimate 
purchaser.

* * * * *
    (e) If a manufacturer includes in an advertisement a statement 
respecting the cost or value of emission control devices or systems, 
the manufacturer shall set forth in the statement the cost or value 
attributed to these devices or systems by the Secretary of Labor 
(through the Bureau of Labor Statistics). The Secretary of Labor, and 
his or her representatives, has the same access for this purpose to the 
books, documents, papers, and records of a manufacturer as the 
Comptroller General has to those of a recipient of assistance for 
purposes of section 311 of the Act.
    44. A new subpart, Subpart M is added to part 90 to read:
Subpart M--In-Use Compliance Testing for Handheld Engines; Bench Aging 
Adjustment; In-Use Durability Demonstration Testing for Nonhandheld 
Engines
Sec.
90.1201  Applicability.
90.1202  Definitions.
90.1203  Manufacturer in-use testing program.
90.1204  Maintenance, procurement, aging and testing of engines.
90.1205  In-use test program reporting requirements.
90.1206  [Reserved]
90.1207  Bench aging adjustment factor testing.
90.1208  Bench aging adjustment; criterion for usage, calculation of 
adjustment factor, reporting requirements.
90.1209  Entry and access.
90.1210--90.1249  [Reserved]
90.1250  Field durability and in-use emission performance 
demonstration program for nonhandheld engines using overhead valve 
technology.

Subpart M--In-Use Compliance Testing for Handheld Engines; Bench 
Aging Adjustment; In-Use Durability Demonstration Testing for 
Nonhandheld Engines


Sec. 90.1201  Applicability.

    The requirements of this subpart from Sec. 90.1201 through 
Sec. 90.1249 are applicable to all handheld Phase 2 engines subject to 
the provisions of subpart A of this part. The requirements of this 
subpart, except for those involving in-use credits, in Secs. 90.1201, 
90.1202, 90.1207, 90.1208, 90.1209 and those from Sec. 90.1250 through 
Sec. 90.1299 are applicable to nonhandheld Phase 2 engines subject to 
the provisions of subpart A of this part.


Sec. 90.1202  Definitions.

    For the purposes of this subpart, except as otherwise provided, the 
definitions in subparts A and C of this part apply to this subpart.


Sec. 90.1203  Manufacturer in-use testing program.

    (a) Unless otherwise approved by the Administrator, at the time of 
the first certification for each model year beginning with the 2002 
model year, each manufacturer shall submit a schedule to the 
Administrator of the Phase 2 engine families, their useful lives, their 
design characteristics (two or four stroke; catalyst or noncatalyst, 
etc.), and their anticipated eligible sales, it intends to produce, by 
model year, over the subsequent four year period (the model year now 
being certified plus the next three model years).
    (b) At the time the manufacturer submits the schedule required 
under paragraph (a) of this section, the manufacturer may include a 
proposed plan for the Administrator's review and approval for the in-
use testing of the current model year and such future model years as it 
chooses to include. In such plans, the manufacturer shall propose the 
in-use testing of individual engine families and engine configurations 
subject to the requirements of this subpart. Such plans shall include a 
discussion of the rationale behind the choice of each family and 
configuration that the Administrator shall use to determine whether the 
manufacturer's plan meets the objective of generating in-use data on 
substantially all of a manufacturer's engines within a reasonable time 
period, and periodically updating that data.
    (c) Based upon the schedule required in paragraph (a) of this 
section, any plan submitted under paragraph (b) of this section, and/or 
such other information as it has available, the Administrator may 
annually identify handheld engine families and at the Administrator's 
option, configurations within families which the manufacturer must then 
subject to in-use testing as described in this section and in 
Sec. 90.1204. For each model year, the Administrator may identify a 
number of engine families that is no greater than the number of 
handheld engine families produced in that model year divided by four 
and rounded to the nearest whole number. If this calculation produces a 
value of zero, then the Administrator may identify no more than one 
engine family for in-use testing for that manufacturer. The 
Administrator may identify families and configurations under this 
paragraph by approving the manufacturer's plan described in paragraph 
(b) of this section, or by providing a written directive to the 
manufacturer.
    (d) For each engine family identified by the Administrator under 
paragraph (c) of this section, engine manufacturers shall perform 
emission testing of an appropriate sample of in-use engines from each 
engine family. Manufacturers shall submit data from this in-use testing 
to the Administrator.
    (e) Number of engines to be tested. An engine manufacturer shall 
test bench aged or field aged in-use engines from each engine family or 
family and configuration identified by the Administrator. Engines to be 
tested shall have accumulated a number of hours pursuant to paragraph 
(g) of this section. The number of engines to be tested by a 
manufacturer shall be determined by the following method:
    (1) A minimum of four (4) engines per family provided that no 
engine fails any standard. For each failing engine, two more engines 
shall be tested until the total number of engines equals ten (10).
    (2) For small volume engine families for the identified model year 
or for small volume engine manufacturers, a minimum of two (2) engines 
per family provided that no engine fails any standard. For each failing 
engine, two more engines shall be tested until the total number of 
engines equals ten (10).
    (3) If an engine family was certified using carry over emission 
data and has

[[Page 4030]]

been previously tested under paragraphs (e)(1) or (e)(2) of this 
section (and mean results did not exceed any applicable emission 
standard), then only one engine for that family must be tested. If that 
one engine fails any pollutant, testing must be conducted as outlined 
at paragraph (e)(1) or (e)(2) of this section, whichever is 
appropriate.
    (f) At the discretion of the Administrator, an engine manufacturer 
may test more engines than the minima described in paragraph (e) of 
this section or may concede failure before testing a total of ten (10) 
engines.
    (g) The Administrator may approve alternatives to manufacturer in-
use testing as described in this subpart, that are designed to 
determine whether an engine family is in compliance with applicable 
standards in use, where:
    (1) Engines, in their production form, or when removed from the 
piece of equipment in which they were installed, cannot safely or 
practically be operated and tested pursuant to subparts D and E of this 
part; or
    (2) The Administrator finds that unique or extraordinary 
circumstances exist that support the need for alternative methods.
    (h) Collection of in-use engines. The engine manufacturer shall 
bench age engines to their full certified useful life as described in 
subpart B of this part using a bench aging procedure approved by the 
Administrator under this subpart, or the engine manufacturer shall 
procure field aged engines which have been operated for at least the 
engine's useful life. Unless otherwise approved by the Administrator, 
the manufacturer shall complete emission testing of bench aged engines 
within 12 calendar months and complete emission testing of field aged 
engines within 24 calendar months after receiving notice that the 
Administrator has identified a particular engine family for testing. 
Field aged engines may be procured from sources associated with the 
engine manufacturer (i.e., manufacturer established fleet engines, 
etc.) or from sources not associated with the manufacturer (i.e., 
consumer-owned engines, independently-owned fleet engines, etc.).


Sec. 90.1204  Maintenance, procurement, aging and testing of engines.

    This section is applicable to handheld engines used for in-use 
testing pursuant to Sec. 90.1203.
    (a) An in-use field aged engine must have a maintenance and use 
history representative of actual in-use conditions.
    (1) To comply with this requirement, a manufacturer must obtain 
information from the end users regarding the accumulated usage, 
maintenance, operating conditions, and storage of the test engines.
    (2) Documents used in the procurement process must be maintained as 
required in Sec. 90.121.
    (3) Each engine of a sample to be field aged shall be assigned a 
random number. Unless otherwise approved by the Administrator, the 
engine with the lowest number shall be tested first, followed by the 
next higher number until testing is completed.
    (b)(1) For an engine family which is to be emission tested 
following bench aging, test engines shall be randomly chosen from 
normal engine production or storage; or randomly chosen from normal 
handheld equipment production or storage.
    (2) Each engine of a sample to be bench aged shall be assigned a 
random number. In emission testing of the bench aged engines, the 
engine with the lowest number shall be tested first, followed by the 
next higher number until testing is completed.
    (c)(1) Bench aged engines must be aged on a dynamometer using a 
bench aging cycle that has been shown to be capable of representing 
field aging for the appropriate technology subgroup pursuant to the 
regulations at Secs. 90.1207 and 90.1208.
    (2) Unless otherwise approved by the Administrator, once an engine 
has begun the bench aging process, it can be terminated and deleted 
only for catastrophic failure or safety concerns requiring major engine 
repair, or because testing of the engine family has been completed 
based upon lower numbered engines.
    (d) The manufacturer may perform minimal set-to-spec maintenance on 
components of a test engine that are not subject to parameter 
adjustment. Unless otherwise approved by the Administrator, maintenance 
to any test engine may include only that which is listed in the owner's 
instructions for engines with the amount of service and age of the test 
engine. Documentation of all maintenance and adjustments shall be 
maintained and retained as required by Sec. 90.121.
    (e) At least one valid emission test, according to the test 
procedure outlined in subpart E of this part, is required for each test 
engine. Unless otherwise approved by the Administrator, no other 
emission testing or performance testing may be performed on a test 
engine prior to the testing at the end of hour accumulation using the 
test procedure outlined in subpart E of this part.
    (f) The Administrator may waive portions or requirements of the 
test procedure, if any, that are not necessary to determine in-use 
compliance with applicable emission standards.
    (g) If a selected test engine fails to comply with any applicable 
emission standard, the manufacturer shall make a reasonable effort, 
including troubleshooting, repairing and retesting, to determine the 
cause of noncompliance. The manufacturer must report all such reasons 
of noncompliance with the in-use test report required pursuant to 
Sec. 90.1205.


Sec. 90.1205  In-use test program reporting requirements.

    (a) The manufacturer shall submit to the Administrator within 
ninety (90) days of completion of testing for a given model year's 
engines, all emission testing results generated from the in-use testing 
program. The following information must be reported for each test 
engine:
    (1) Engine family;
    (2) Model;
    (3) Engine serial number;
    (4) Date of manufacture;
    (5) Hours of use;
    (6) Date and time of each test attempt;
    (7) Results (if any) of each test attempt;
    (8) Schedules, descriptions and justifications of all maintenance 
and/or adjustments performed;
    (9) Schedules, descriptions and justifications of all modifications 
and/or repairs; and
    (10) Determinations of noncompliance.
    (b) The manufacturer must electronically submit the information 
required in this section using EPA's electronic information format. The 
Administrator may exempt manufacturers from this requirement upon 
written request with supporting justification as to the manufacturer's 
lack of adequate information processing technology.
    (c) The report required in paragraph (a) of this section must 
include a listing of any test engines that were deleted from the aging 
process or testing process and provide a technical justification to 
support the deletion.
    (d) All testing reports and requests for approvals made under this 
subpart shall be addressed to: Manager, Engine Compliance Programs 
Group (6403-J), U.S. Environmental Protection Agency, Washington, D.C. 
20460.
    (e) The Administrator may approve and/or require modifications to a 
manufacturer's in-use testing programs.

[[Page 4031]]

Sec. 90.1206  [Reserved]


Sec. 90.1207  Bench aging adjustment factor testing.

    (a) This section is applicable to the bench aging procedures for 
handheld engines for in-use emission testing and to the bench aging 
procedures for the full useful life certification testing of 
nonhandheld sidevalve engines and nonhandheld engines with 
aftertreatment.
    (b) The bench aging adjustment procedure described in Sec. 90.1208 
shall be used to determine whether a given bench aging cycle, approved 
for adjustment factor testing by the Administrator, can be used to 
represent field aged engines for handheld in-use testing under this 
subpart or for certification of nonhandheld sidevalve engines or 
nonhandheld engines with aftertreatment; and, if so, what the 
appropriate adjustment factor should be. If both the IWB and 
IWF as defined in Sec. 90.1208 are less than or equal to 20% 
of the appropriate HC+NOX (NMHC+NOX) standard, 
then the subject bench aging cycle can be used to generate emissions 
data for adjustment to represent field aged emissions.
    (c) (1) Nothing in this section shall be construed to prohibit 
different manufacturers from jointly demonstrating that a particular 
bench aging cycle, approved by the Administrator for adjustment factor 
testing, may be used to represent the field aged emissions of engines 
of a particular technology subgroup when they each agree to use the 
same bench aging cycle, when they each contribute field and bench aged 
test engines for testing of that technology subgroup under 
Sec. 90.1208, and when they each provide justification satisfactory to 
the Administrator that the engines can be expected to have similar 
emission deterioration characteristics and that a reasonable basis 
exists for such joint testing.
    (2) Unless otherwise approved by the Administrator, a manufacturer 
participating or desiring to participate in a joint adjustment factor 
testing program may not enter or drop out of the joint program for that 
technology subgroup after the adjustment factor derived from the 
program has been used one or more times for certification of 
nonhandheld engines or in-use testing of handheld engines. When a 
manufacturer does drop out, the adjustment factor must be recalculated 
without that manufacturer's data. When an additional manufacturer is 
allowed to join, the adjustment factor must be recalculated to reflect 
the data generated by the new manufacturer's engines.
    (d) Field aging of engines shall be performed in representative 
equipment in the hands of residential customers, or professional users 
or in manufacturers' fleets, except that a minimum of one third of the 
field aged engines but not less than one engine for a given engine 
family or technology subgroup, shall be aged in individual customer 
usage or in fleets where the engine manufacturer does not carry out or 
exercise control over the engines' maintenance or limit their usage 
such that the engines are not used in a way that is representative of 
typical in-use engines.
    (e) For each engine family or technology subgroup for which a 
manufacturer desires to use bench aging, the manufacturer or group of 
manufacturers, as applicable, shall propose to the Administrator the 
bench aging cycle and an engine aging plan it intends or they intend to 
use to demonstrate the appropriateness of such cycle to represent field 
aged engines. Such proposals may be made up to 48 months prior to the 
start of a given model year. EPA shall reject such proposed aging 
cycles and/or engine aging plans in writing, within 90 days of receipt, 
or they shall be considered approved for adjustment factor testing 
pursuant to this section and Sec. 90.1208. Such proposals shall 
include:
    (1) A detailed description of the engine families a cycle is 
intended to cover, a justification satisfactory to the Administrator 
that the engines can be expected to have similar emission deterioration 
characteristics, a justification of the appropriateness of the subject 
cycle to represent field aging of the engines the cycle is intended to 
cover and data sufficient for the Administrator to ascertain whether 
the bench aging cycle has been previously determined to represent field 
aging for any other engine family under the provisions of this section 
and Sec. 90.1208;
    (2) A detailed description of the proposed bench aging cycle 
including, but not limited to, such parameters as duration at each 
throttle setting, sequencing of throttle changes, loading and load 
changes, hot starts and cold starts, idles, acceleration times, 
presence of accessory loads, periods of shutdown and other factors as 
the Administrator may require;
    (3) A description of each engine to be aged in the field and on the 
bench, including make, model, engine family, displacement, power 
rating, rated speed and other such information as the Administrator may 
require to enable the Administrator to determine whether such engines 
are appropriate for evaluating the bench aging cycle for the engine 
families or technology subgroup described in paragraph (e)(1) of this 
section;
    (4) A description of the way in which individual engines will be 
selected, uniquely identified and tracked for both bench and field 
aging and for subsequent emission testing;
    (5) A description of the method by which each engine selected for 
field aging will be aged, the procedures for determining and carrying 
out appropriate engine maintenance during field aging and bench aging, 
a description and rationale for any maintenance the manufacturer 
proposes to perform additional to routine maintenance described in the 
maintenance schedule provided to the purchaser, and a description of 
records that will be kept of both bench and field engine operation and 
maintenance; and
    (6) The location(s) of the facilities or sites at which each bench 
and field aged engine will be aged and tested.
    (f) Upon approval by the Administrator of the bench aging cycle for 
evaluation testing and the engine aging plan, the manufacturer shall 
conduct hour accumulation to the full regulatory useful life of the 
engines according to the approved engine aging plan using the approved 
bench aging cycle. Such aging shall be followed by emission testing 
pursuant to the requirements of subpart E of this part. At its option, 
the manufacturer may age handheld commercial engines to 75% of their 
regulatory useful life for bench aging adjustment testing.
    (g) Handheld engines aged for adjustment factor testing pursuant to 
the requirements of this section may not be used in the Manufacturer 
In-use Test Program required under Sec. 90.1203.
    (h) The Administrator may require that testing under this section 
and the evaluation of the appropriateness of a bench aging cycle to 
represent field aging under Sec. 90.1208, be repeated for a particular 
engine family or technology subgroup as often as every five years; 
except that the Administrator may require that such testing be repeated 
more frequently in model years prior to the 2006 model year.
    (1) The Administrator shall notify a manufacturer or group of 
manufacturers of the requirement to conduct a bench aging adjustment 
factor program for a particular engine family or technology subgroup 
and the period for completion of the program. The time period for 
completion shall be no less than one year for engines having 500 or 
1000 hour useful lives.
    (2) Within sixty days of the date of the Administrator's notice, 
the manufacturer or group of manufacturers shall provide a plan for the

[[Page 4032]]

Administrator's review and approval meeting the requirements of 
paragraph (e) of this section including a proposed bench aging cycle 
and an engine aging plan.
    (i) Upon completion of engine aging and testing pursuant to the 
requirements of this section, engine manufacturers wishing to use bench 
aging and the adjustment factors calculated pursuant to Sec. 90.1208 
for in-use emission testing of handheld engines or for certification of 
nonhandheld sidevalve engines or nonhandheld engines with 
aftertreatment, as applicable, shall provide a report to the 
Administrator describing the aging and testing conducted under this 
section and Sec. 90.1208. Such report shall be submitted no less than 
90 days before the initiation of any such bench aging for in-use or 
certification testing on the engines and engine families covered by the 
plan approved under this section. The Administrator shall disapprove 
the report within 30 days of the date of receipt, or the report shall 
be automatically approved and the manufacturer may use the bench aging 
cycle and adjustment factors described in the report for its bench 
aging activities of the subject families. Such report shall contain the 
following information about the field/bench adjustment program 
conducted under this section and Sec. 90.1208:
    (1) An identifying description of the bench aging cycle sufficient 
for the Administrator to ascertain which cycle proposed pursuant to 
this section has been evaluated;
    (2) A description of all engines selected for bench aging and field 
aging for this engine family or technology subgroup, as applicable. 
Such description shall include the make, model, engine family, 
displacement, power rating, rated speed, unique identifying 
description, and other such information as the Administrator may 
require;
    (3) A description of all maintenance performed on each engine 
during hour accumulation, including a detailed explanation of the need 
for any maintenance not contained in the maintenance schedule for that 
model engine provided to engine owners;
    (4) A description of how each engine was aged (e.g., bench cycle, 
field aged-manufacturer fleet, or field aged-individual customer);
    (5) A description of any engine selected for aging pursuant to 
paragraph (i)(2) of this section that was deleted from aging or 
testing. Include a full explanation of the rationale for deletion;
    (6) Tabulations of all emission test results and all inputs and 
outcomes of the equations found in Sec. 90.1208; and
    (7) A statement signed by an appropriate official of the 
manufacturer responsible for compliance of engines with Federal 
emission requirements that clearly states that all engine selection, 
aging, maintenance, testing, results calculation, and data evaluation 
was performed in full accordance with the requirements under this part.


Sec. 90.1208  Bench aging adjustment; criterion for usage, calculation 
of adjustment factor, reporting requirements.

    (a) Manufacturers desiring to use bench aging prior to performing 
in-use emission tests on handheld engines or prior to performing 
certification testing on nonhandheld sidevalve engines or nonhandheld 
engines with aftertreatment, must first demonstrate that the chosen 
bench aging cycle appropriately represents field aging as determined 
under this section and Sec. 90.1207. Where a bench aging cycle is shown 
to appropriately represent field aging under this section and 
Sec. 90.1207, manufacturers shall calculate separate multiplicative 
bench aging adjustment factors as described in this section to adjust 
the HC+NOX (NMHC+NOX) and CO emissions of bench 
aged engines.
    (b) A minimum of six engines from each technology subgroup shall be 
aged and tested. Three of these engines must be aged on the bench and 
three must be aged in the field.
    (c) Separate 90% confidence intervals shall be calculated around 
the HC+NOX (NMHC+NOX) mean of the bench aged 
engines and the HC+NOX (NMHC+NOX) mean of the 
field aged engines. The confidence intervals are independent of each 
other and are calculated according to the following equations:
    (1)(i) For the 90% confidence interval about the mean of the group 
of bench aged engines, B90:

B90=x.8bIWb

Where:

B90=The 90% confidence interval about the mean of the group 
of bench aged engines.
xb=The HC+NOX (NMHC+NOX) sample mean 
of the group of bench aged engines.
    IWb=The confidence interval width for the group of bench 
aged engines as defined by the equation in paragraph (c)(1)(ii) of this 
section.

    (ii) IWb is defined by the following equation:

    [GRAPHIC] [TIFF OMITTED] TP27JA98.004
    
Where:

t90=The appropriate 90% critical point from Student's t 
table for 90% confidence and nb-1 observations; this value 
will decrease as nb increases.
Sb=The HC+NOX (NMHC+NOX) sample 
standard deviation of the group of bench aged engines, where:
[GRAPHIC] [TIFF OMITTED] TP27JA98.005

nb=The number of bench aged engines tested.

    (2)(i) For the 90% confidence interval about the mean of the group 
of field aged engines, F90:

F90=xfIWf

Where:

F90=The 90% confidence interval about the mean of the group 
of field aged engines.
xf=The HC+NOX (NMHC+NOX) sample mean 
of the group of field aged engines.
IWf=The confidence interval width for the group of field 
aged engines as defined by the equation in paragraph (c)(2)(ii) of this 
section.

    (ii) IWf is defined by the following equation:
    [GRAPHIC] [TIFF OMITTED] TP27JA98.006
    
Where:

t90=The appropriate 90% critical point from Student's t 
table for 90% confidence and nb-1 observations; this value 
will decrease as nb increases.
Sf=The HC+NOX (NMHC+NOX) sample 
standard deviation of the group of field aged engines, where:
[GRAPHIC] [TIFF OMITTED] TP27JA98.007

nf=The number of field aged engines tested.
    (d) Both IWb and IWf must be rounded to the 
same number of significant digits as contained in the appropriate 
standard.
    (e) If both IWb and IWf are less than or 
equal to 20% of the appropriate HC +NOX (NMHC+NOX) standard 
as defined by Sec. 90.103, then separate Bench Aging Adjustment 
factors, AFs, can be calculated for HC+NOX (NMHC+NOX) and CO 
as follows:

AF=the maximum of [(xf / xb) or 1.0]

    (f) If either or both confidence interval widths IWb or 
IWf is/are greater than 20% of the appropriate standard as 
defined by Sec. 90.103, then the manufacturer may elect to test 
additional engines included and described in the plan approved under 
Sec. 90.1207 and recalculate the relevant

[[Page 4033]]

statistics. Additional testing need only be done for the group that 
exceeds 20% of the appropriate standard. After each additional test, 
B90, F90, IWb and IWf shall 
be recalculated according to paragraph (c) of this section. Additional 
engines may be added until such time as the newly calculated confidence 
interval width (IWb or IWf, or both) are less 
than or equal to 20% of the appropriate HC+NOX 
(NMHC+NOX) standard as defined by Sec. 90.103. When both 
IWb or IWf are less than or equal to 20% of the 
appropriate standard as defined by Sec. 90.103, then separate Bench 
Aging Adjustment Factors, AFs, may be calculated for each regulated 
pollutant according to paragraph (e) of this section.
    (g) The adjustment factors calculated under paragraph (e) of this 
section shall be multiplicatively applied to the appropriate full 
useful life bench-aged handheld in-use test results or to the 
appropriate full useful life certification test results of nonhandheld 
sidevalve engines or nonhandheld engines with aftertreatment for that 
engine family or technology subgroup for all manufacturers whose 
engines were tested in the test program for that technology subgroup, 
until another bench aging adjustment program is conducted for that 
family or technology subgroup.


Sec. 90.1209  Entry and access.

    (a) To allow the Administrator to determine whether a manufacturer 
is complying with the provisions under this subpart, EPA enforcement 
officers or their authorized representatives, upon presentation of 
credentials, shall be permitted entry, during operating hours, into any 
of the following places:
    (1) Any facility where engines undergo or are undergoing bench 
aging, field aging, maintenance, repair, preparation for aging, 
selection for aging or emission testing.
    (2) Any facility where records or documents related to any of 
activities described in paragraph (a)(1) of this section are kept.
    (3) Any facility where any engine that is being tested or aged, was 
tested or aged or will be tested or aged is present.
    (b) Upon admission to any facility referred to in paragraph (a) of 
this section, EPA enforcement officers or EPA authorized 
representatives are authorized to perform those activities set forth in 
Sec. 90.705 (b) and also to inspect and make copies of records related 
to engine aging (service accumulation) and maintenance.
    (c) The provisions of Sec. 90.705(c), (d), (e), (f) and (g) also 
apply to entry and access under this subpart.


Secs. 90.1210--90.1249  [Reserved.]


Sec. 90.1250  Field durability and in-use emission performance 
demonstration program for nonhandheld engines using overhead valve 
technology.

    The testing required pursuant to this section shall be for the 
purpose of validating the appropriateness of assigned deterioration 
factors (dfs) or manufacturer determined dfs used pursuant to 
Sec. 90.104 to represent the field aged deterioration of overhead valve 
technology engine families. For brevity, such testing is referred to as 
df validation testing.
    (a) Unless otherwise approved by the Administrator, at the time of 
the first certification for each model year of Phase 2 engines, each 
manufacturer shall submit a schedule to the Administrator of the 
overhead valve technology engine families it intends to produce over 
the subsequent four year period (the model year now being certified 
plus the next three model years) including their useful lives, their 
design characteristics (i.e.; catalyst or noncatalyst, carbureted or 
fuel injected, etc.), and their anticipated eligible sales.
    (b) In the schedule submitted under paragraph (a) of this section, 
and for the same time period, the manufacturer shall specify the engine 
families for which it intends to conduct field/bench adjustment testing 
under Secs. 90.1207 and 90.1208 and shall also specify the engine 
families for which it intends to compute its own dfs pursuant to 
Sec. 90.104(h)(2). Such schedule shall include an estimate of the 
number of field aged engines that will be emission tested each calendar 
year for the programs referenced in this paragraph.
    (c) At the time the manufacturer submits the schedule required 
under paragraph (a) of this section, the manufacturer may include a 
proposed plan for the Administrator's review and approval of the 
overhead valve engine families, configurations and associated 
quantities of engines it plans to field age to full useful life and in-
use test during those four years to determine the field aged dfs for 
engine families for which assigned dfs were used in certification. In 
such plans, the manufacturer:
    (1) May consider the number of field aged engines it plans to test 
in each calendar year from paragraph (b) of this section and the limit 
on additional testing of field aged engines that can be assigned by EPA 
pursuant to paragraph (c) of this section.
    (2) Shall include a discussion of the rationale for the choice of 
each family and configuration sufficient to enable the Administrator to 
determine whether the manufacturer's plan meets the objective of 
generating in-use data sufficient to validate the appropriateness of 
the assigned dfs on a substantial portion of a manufacturer's engines 
within a reasonable time period, and providing for periodic 
revalidation of the assigned dfs.
    (d) If no plan submitted pursuant to paragraph (c) of this section 
is approved by the Administrator, then, based upon the schedule 
submitted pursuant to paragraph (a) of this section and other available 
information, and considering the field aging requirements of 
Secs. 90.1207, 90.1208 and 90.104(h)(2), and any requests from 
manufacturers to work jointly, the Administrator may provide a schedule 
of the overhead valve engine families and associated quantities of 
engines that must be field aged to full useful life and in-use tested 
during those four years to validate dfs.
    (e) EPA shall not require any nonhandheld engine manufacturer to 
conduct df validation emission testing such that df validation emission 
testing when added to that testing of field aged engines proposed by 
the manufacturer under paragraph (b) of this section would require the 
manufacturer to emission test more than 24 total field aged engines in 
one calendar year for bench aged field adjustment testing pursuant to 
Secs. 90.1207 and 90.1208, df generation testing pursuant to 
Sec. 90.104(h)(2), and df validation testing pursuant to this section.
    (f) The Administrator may provide a schedule for engine testing to 
validate dfs pursuant to this section by approving the plan submitted 
by the manufacturer under paragraph (c) of this section, or by a 
written directive to the manufacturer under paragraph (d) of this 
section. Unless otherwise approved by the Administrator, for each test 
engine tested to fulfill the testing schedule provided by the 
Administrator under paragraph (c) or (d) of this section, the 
manufacturer shall conduct a baseline emission test at a number of 
hours equal to that on the corresponding certification engine followed 
by field aging to the certified useful life. Each engine shall then be 
emission tested using the applicable test procedures described in this 
part measuring all regulated pollutants. Field aging shall be performed 
in representative equipment in the hands of residential customers, or 
professional users or in manufacturers' fleets, under usage and 
conditions representative of typical use.
    (1) Unless otherwise approved by the Administrator, equipment shall 
be considered to be representative if it is of the type (e.g., walk 
behind lawnmowers or concrete saws) of equipment into which at least 
one third of the engines

[[Page 4034]]

are installed. If no one application of the engine constitutes one 
third of sales, then equipment shall be representative if it is taken 
from either or both of the two types of applications having the largest 
U.S. sales volumes.
    (2) Unless otherwise approved by the Administrator, test engines 
that receive maintenance additional to that recommended to the 
purchaser in the owner's manual shall not be considered representative 
of typical use.
    (g) No later than 90 days following the end of each model year, 
each manufacturer subject to this section shall provide a tabulation, 
by engine family, of all engines undergoing hour accumulation under 
this regulation, the number of hours accumulated on each engine, the 
equipment application for each engine and the basis for that choice of 
equipment. Such tabulation shall include the engine family, the engine 
identification number assigned for tracking purposes, the type of 
application, the projected test date and the geographic location (city 
and state) where hour accumulation is occurring. Such tabulation, or a 
separate tabulation submitted at the same time, shall contain all in-
use test results that have been generated during the preceding model 
year. Such tabulation shall include the engine family, the engine 
identification number assigned for tracking purposes, the type of 
application, the applicable certification deterioration factor and the 
calculated HC+NOX deterioration factor determined from the 
testing required in this subpart.
    45. Subpart N is added to part 90 to read as follows:

Subpart N--In-Use Credit Program for New Handheld Engines

Sec.
90.1301  Applicability.
90.1302  Definitions.
90.1303  General provisions.
90.1304  Averaging.
90.1305  Banking.
90.1306  Trading.
90.1307  Credit calculation.
90.1308  Maintenance of records.
90.1309  Reporting requirements.
90.1310  Request for hearing.

Subpart N--In-Use Credit Program for New Handheld Engines


Sec. 90.1301  Applicability.

    Phase 2 handheld engines subject to the provisions of subpart A of 
this part are eligible to participate in the in-use credit program 
described in this subpart for HC +NOX (NMHC+NOX) 
and CO emissions.


Sec. 90.1302  Definitions.

    The definitions in subpart A of this part and the definition of 
``point of first retail sale'' from subpart C of this part apply to 
this subpart. The following definitions shall also apply to this 
subpart:
    Averaging means the exchange of handheld engine in-use emission 
credits between engine families within a given manufacturer's product 
line.
    Banked credits refer to positive emission credits based on actual 
applicable production/sales volume as contained in the end of model 
year in-use testing reports submitted to EPA. Some or all of these 
banked credits may be revoked if EPA review of the end of model year 
in-use testing reports or any subsequent audit action(s) uncovers 
problems or errors.
    Banking means the retention of handheld engine in-use emission 
credits by the manufacturer generating the emission credits or 
obtaining such credits through trading, for use in future model year 
averaging or trading as permitted by these regulations.
    Carry-over engine family means an engine family which undergoes 
certification using carryover test data from previous model years.
    Compliance level for an engine family is determined by averaging 
the in-use test results from each test engine of the family. The 
compliance level for an individual configuration may be determined in 
cases where the Administrator directs the testing of an individual 
configuration.
    Emission credits or in-use credits represent the amount of emission 
reduction or exceedance, for each regulated pollutant, by a handheld 
engine family below or above, respectively, the applicable 
certification standard to which the engine family is certified. 
Emission reductions below the standard are considered ``positive 
credits,'' while emission exceedences above the standard are considered 
``negative or required credits.''
    Trading means the exchange of handheld engine in-use emission 
credits between manufacturers and/or brokers.


Sec. 90.1303  General provisions.

    (a) The in-use credit program for eligible Phase 2 handheld engines 
is described in this subpart. Participation in this program is 
voluntary.
    (b) Any handheld Phase 2 engine family subject to the provisions of 
subpart A of this part is eligible to participate in the in-use credit 
program described in this subpart.
    (c) Credits generated and used in the nonhandheld engine 
certification averaging, banking, and trading program pursuant to the 
provisions of subpart C of this part are not interchangeable with 
credits generated and used in the handheld engine in-use credit 
program. In-use credits under this subpart may not be used to address 
the emissions of any nonhandheld engine. Nor may nonhandheld 
certification credits be used to address any in-use credit need 
determined under this subpart.
    (d) An engine family with a compliance level, as determined by in-
use testing pursuant to subpart M of this part and paragraph (h) of 
this section, below the applicable standard to which the engine family 
is certified may generate emission credits for averaging, banking, or 
trading in the in-use credit program.
    (e) Positive credits generated in a given model year may be used in 
that model year and/or in any subsequent model year during the Phase 2 
program.
    (f) A manufacturer of an engine family with a compliance level 
exceeding the applicable standard to which the engine family is 
certified, may, prior to the date of the report required under 
paragraph (i) of this section, use previously banked credits, purchase 
credits from another manufacturer, or perform additional testing 
pursuant to paragraph (h) of this section to address (as calculated 
elsewhere in this subpart) the associated credit deficit (negative 
credits or a need for credits).
    (g) In the case of in-use testing of engine families that were 
certified using carry-over data, and in the absence of other applicable 
test data acceptable to the Administrator, the test results from one 
model year's testing shall apply to up to four years of production of 
that family: the model year tested, the next model year (if carried 
over to that year), and one or two previous model years (if carried 
over from the previous year or the two previous years, respectively). 
In-use credits shall be generated or used, as appropriate.
    (h) A manufacturer must notify EPA of plans to test additional 
engine families beyond those identified by EPA pursuant to regulations 
in subpart M of this part for the in-use testing program. Such notice 
must be submitted 30 days prior to initiation of service accumulation. 
If the additional testing discovers an engine family to be in 
noncompliance with the applicable standard, the testing must be treated 
as if it were a failure of the normal in-use testing requirement of an 
engine family. If the additional testing shows the engine family to be 
in compliance with the applicable standard, in-use credits may be 
generated subject to the provisions of this subpart.

[[Page 4035]]

    (i) Manufacturers must demonstrate a zero or positive credit 
balance under the in-use credit program for all regulated pollutants 
for a particular model year within 90 days of the end of the in-use 
testing of that model year's engine families. At that time 
manufacturers must file a report with EPA pursuant to Sec. 90.1309.
    (j) Manufacturers shall maintain separate balances for 
HC+NOX (NMHC+NOX) and CO credits. 
HC+NOX and NMHC+NOX credits are interchangeable 
with each other but not with CO credits.


Sec. 90.1304  Averaging.

    (a) A manufacturer may use averaging across engine families to 
demonstrate a zero or positive credit balance for a model year. 
Positive credits to be used in averaging may be obtained from credits 
generated by another engine family of the same model year, credits 
banked in previous model years, or credits obtained through trading.
    (b) Credits used to demonstrate a zero or positive credit balance 
must be used at a rate of 1.1 to 1.


Sec. 90.1305  Banking.

    (a) A manufacturer of a handheld engine family with an in-use 
compliance level below the standard to which the engine family is 
certified for a given model year may bank positive in-use credits for 
that model year for use in in-use averaging and trading.
    (b) A manufacturer may consider credits to be banked, for use in 
future averaging or trading, 30 days after the submission of the report 
required by Sec. 90.1309(a). During the 30 day period EPA will work 
with the manufacturer to correct any error in calculating banked 
credits, if necessary.


Sec. 90.1306  Trading.

    (a) A handheld engine manufacturer may exchange positive in-use 
emission credits with other handheld engine manufacturers through 
trading.
    (b) In-use credits for trading can be obtained from credits banked 
for model years prior to the model year of the engine family requiring 
in-use credits.
    (c) Traded in-use credits can be used for averaging, banking, or 
further trading transactions.
    (d) Unless otherwise approved by EPA, a manufacturer that generates 
positive in-use credits must wait 30 days after it has both completed 
in-use testing for the model year for which the credits were generated 
and submitted the report required by Sec. 90.1309(a) before it may 
transfer credits to another manufacturer or broker.
    (e) In the event of a negative credit balance resulting from a 
transaction, both the buyer and the seller are liable, except in cases 
involving fraud. Engine families participating in a trade that leads to 
a negative credit balance may be subject to recall under subparts I and 
M of this part if the engine manufacturer having the negative credit 
balance is unable or unwilling to obtain sufficient credits in the time 
allowed under Sec. 90.1303(i).


Sec. 90.1307  Credit calculation.

    For each participating engine family, and for each regulated 
pollutant (HC+NOX (NMHC+NOX) and CO) emission 
credits (positive or negative) are to be calculated according to the 
following equation and rounded to the nearest gram. Consistent units 
are to be used throughout the equation:

Credits = Sales  x  (Standard--CL)  x  Power  x  Useful life  x  AF  x  
LF
Where:

Useful Life = the useful life in hours corresponding to the useful life 
category for which the engine family was certified.
Power = the sales weighted maximum modal power, in kilowatts, as 
calculated from the applicable federal test procedure as described in 
this part. This is determined by multiplying the maximum modal power of 
each configuration within the family by its eligible sales, summing 
across all configurations and dividing by the eligible sales of the 
entire family. Where testing is limited to certain configurations 
designated by the Administrator, the maximum modal power for the 
individual configuration(s) shall be used.
Sales = the number of eligible U.S. sales, as defined in subpart A of 
this part, for the engine family or configuration as applicable.
Standard = The applicable emission standard to which the engine family 
was certified under subpart B of this part.
CL = compliance level of the in-use testing for the subject pollutant 
in g/kW-hr.
AF = adjustment factor for the number of tests conducted as determined 
from the following table, except that when a manufacturer concedes 
failure before completion of testing as permitted under 
Sec. 90.1203(f), the adjustment factor shall be 1.0:

----------------------------------------------------------------------------------------------------------------
                                                                                                                
----------------------------------------------------------------------------------------------------------------
No. Engines tested.............  1-5                  6-7                 8-9                 10 or more.       
Adjustment factor..............  0.5                  0.75                0.9                 1.0               
----------------------------------------------------------------------------------------------------------------

LF = Load Factor of 0.85 for test cycle C. For manufacturers using 
alternative or special test cycles approved by the Administrator, the 
Load Factor is calculated using the Load Factor formula for nonhandheld 
engines found in Sec. 90.207.


Sec. 90.1308  Maintenance of records.

    (a) Any manufacturer that is participating in the in-use credit 
program set forth in this subpart shall establish, maintain, and retain 
the records required by Sec. 90.209 with respect to its participation 
in the in-use credit program.
    (b) EPA may void ab initio a certificate of conformity for an 
engine family for which the manufacturer fails to retain the records 
required under this section or to provide such information to the 
Administrator upon request.


Sec. 90.1309  Reporting requirements.

    (a) Any manufacturer who participates in the in-use credit program 
is required to submit an in-use credit report with the end of the model 
year in-use testing report required under Sec. 90.1205 within 90 days 
of the end of the in-use testing of a given model year's engine 
families. This report must show the calculation of credits from all the 
in-use testing conducted by the manufacturer for a given model year's 
engines. Such report shall show the applications of credits, the 
trading of credits, the discounting of credits that are used and the 
final credit balance. Such report shall calculate credit generation or 
usage for past model years and estimate credit generation or usage for 
the next model year when carry over families are tested pursuant to 
Sec. 90.1303(g). The manufacturer may submit corrections to such end of 
model year reports in a final report for a period of up to 270 days 
after the end of the in-use testing of a given model year's engine 
families.
    (b) The calculation of eligible sales for end-of-year and final 
reports must be based on the location of the point of first retail sale 
(for example, retail customer or dealer) also called the final product 
purchase location. Upon advance

[[Page 4036]]

written request, the Administrator will consider other methods to track 
engines for credit calculation purposes that provide high levels of 
confidence that eligible sales are accurately counted.
    (c) Reports shall be submitted to: Manager, Engine Compliance 
Programs Group (6403-J), U.S. Environmental Protection Agency, SW., 
Washington, DC 20460.
    (d) A manufacturer that fails to submit a timely end of year report 
as required in paragraph (a) of this section will be considered 
ineligible to have participated in the in-use credit program.
    (e) If EPA or the manufacturer determines that a reporting error 
occurred on an end of model year report previously submitted to EPA 
under this subpart, or an engine family in-use testing report submitted 
to EPA under subpart I of this part, the manufacturer's credits and 
credit calculations will be recalculated. Erroneous positive credits 
will be void. Erroneous negative credits may be adjusted by EPA. An 
update of previously submitted ``point of first retail sale'' 
information is not considered an error and no increase in the number of 
credits will be allowed unless an actual error occurred in the 
calculation of credits due to an error in the ``point of first retail 
sale'' information from the time of the original end of model year 
report.


Sec. 90.1310  Request for hearing.

    An engine manufacturer may request a hearing on the Administrator's 
voiding of an engine family's certificate of conformity under 
Sec. 90.1308(b). The administrative procedures for a public hearing 
requested under this subpart shall be those procedures set forth in 
Secs. 90.512, 90.513, 90.514 and 90.515.
[FR Doc. 98-941 Filed 1-26-98; 8:45 am]
BILLING CODE 6560-50-P