[Federal Register Volume 61, Number 26 (Wednesday, February 7, 1996)]
[Proposed Rules]
[Pages 4600-4621]
From the Federal Register Online via the Government Publishing Office [www.gpo.gov]
[FR Doc No: 96-2230]



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


ENVIRONMENTAL PROTECTION AGENCY
40 CFR Parts 89, 90, and 91

[FRL-5412-3]
RIN 2060-AE54


Control of Air Pollution; Emission Standards for New Gasoline 
Spark-Ignition and Diesel Compression-Ignition Marine Engines; 
Exemptions for New Nonroad Compression-Ignition Engines at or Above 37 
Kilowatts and New Nonroad Spark-Ignition Engines at or Below 19 
Kilowatts

AGENCY: Environmental Protection Agency (EPA).

ACTION: Supplemental notice of proposed rulemaking; proposed revisions.

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

SUMMARY: Pursuant to section 213(a)(3) of the Clean Air Act as amended, 
EPA published a Notice of Proposed Rulemaking (NPRM) on November 9, 
1994 (59 FR 55930) for emission standards for new gasoline spark-
ignition and diesel compression-ignition marine engines. EPA believes 
that the proposed standards will help nonattainment areas come into 
compliance with the ozone National Ambient Air Quality Standards.
    The Agency is now publishing this Supplemental Notice of Proposed 
Rulemaking (SNPRM) because EPA wishes to refine its proposals regarding 
compliance programs, and because EPA wishes to address some of the 
comments received on the NPRM. Many of the provisions of this SNPRM 
seek to minimize regulatory burdens proposed in the NPRM without 
reducing environmental benefits. The proposals include, for example, 
modified compliance requirements for small manufacturers and 
manufacturers of sterndrive/inboard engines or old technology two-
stroke outboard/personal watercraft engines. Also, this Notice proposes 
an in-use averaging, banking, and trading program, and addresses 
comments regarding consistency with the regulations on land-based 
nonroad compression-ignition engines rated at or above 37 kilowatts 
(kW). The Agency is proposing adjustments to the form of the proposed 
standards for gasoline spark-ignition marine engines, and is proposing 
changes to the level of the standards for sterndrive and inboard 
engines. Finally, this Notice proposes to revise the criteria for a 
national security exemption in the regulations regarding marine 
engines, land-based nonroad compression-ignition engines 
(37kW), and land-based nonroad spark-ignition engines 
(19kW).
DATES: The comment period for this rulemaking will reopen on February 
7, 1996, for purposes of taking comment on issues raised in this SNPRM 
and will remain open until March 8, 1996, or 30 days after the date of 
a public hearing, if one is held.
    The Agency will hold a public hearing regarding the content of this 
SNPRM on February 22, 1996, if it receives the request to testify at a 
hearing by February 20, 1996. The Agency will cancel this hearing if no 
one requests to testify. Members of the public should call the contact 
persons indicated below to notify EPA of their interest in testifying 
at the hearing; they may call the contact persons after February 20, 
1996, to determine whether the hearing will be held.

ADDRESSES: Interested parties may submit written comments (in 
duplicate, if possible) for EPA consideration by addressing them as 
follows: EPA Air Docket (LE-131), Attention: Docket Number A-92-28, 
room M-1500, 401 M Street, SW., Washington, D.C. 20460.
    The public hearing will be held at the National Vehicle and Fuel 
Emission Laboratory, 2565 Plymouth Road, Ann Arbor, Michigan, at 9 a.m.
    Materials relevant to this rulemaking are contained in this docket 
and may be reviewed at this location 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.

FOR FURTHER INFORMATION CONTACT: Deanne R. North, Office of Mobile 
Sources, Engine Programs and Compliance Division, (313) 668-4283, or 
James A. Blubaugh, Office of Mobile Sources, Engine Programs and 
Compliance Division, (202) 233-9244.

SUPPLEMENTARY INFORMATION:

I. Obtaining Copies of the Regulatory Language

    The Agency has not included in this document the proposed 
regulatory language. Electronic copies (on 3.5''diskettes) of the 
proposed regulatory language may be obtained free of charge by 
visiting, writing, or calling the Environmental Protection Agency, 
Engine Programs and Compliance Division, 2565 Plymouth Road, Ann Arbor, 
MI 48105, (313) 668-4288. Refer to Docket A-92-28. A copy is also 
available for inspection in the docket (see ADDRESSES).
    The preamble and regulatory language are also available 
electronically on the Technology Transfer Network (TTN), which is an 
electronic bulletin board system (BBS) operated by EPA's Office of Air 
Quality Planning and Standards. The service is free of charge, except 
for the cost of the phone call. Users are able to access and download 
TTN files on their first call using a personal computer and modem per 
the following information.
    TTN BBS: 919-541-5742 (1200-14400 bps, no parity, 8 data bits, 1 
stop bit) Voice Helpline: 919-541-5384. Also accessible via Internet: 
TELNET ttnbbs.rtpnc.epa.gov Off-line: Mondays from 8:00 a.m. to 12:00 
Noon ET
    A user who has not called TTN previously will first be required to 
answer some basic informational questions for registration purposes. 
After completing the registration process, proceed through the 
following menu choices from the Top Menu to access information on this 
rulemaking.

 GATEWAY TO TTN TECHNICAL AREAS (Bulletin Boards)
 OMS--Mobile Sources Information
 Rulemaking and Reporting
<6> Non-Road
<1> File area #1. Non-Road Marine Engines

    At this point, the system will list all available files in the 
chosen category in chronological order with brief descriptions. To 
download a file, select a transfer protocol that is supported by the 
terminal software on your own computer, then set your own software to 
receive the file using that same protocol.
    If unfamiliar with handling compressed (that is, ZIP'ed) files, go 
to the TTN top menu, System Utilities (Command: 1) for information and 
the necessary program to download in order to unZIP the files of 
interest after downloading to your computer. After getting the files 
you want onto your computer, you can quit the TTN BBS with the 
oodbye command.
    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.

II. Contents

III. Statutory Authority and Background
    A. Statutory Authority
    B. Background
IV. Proposed Changes; Discussion of Issues
    A. Emission Standards for Spark-Ignition Engines
    1. HC+NOX Emission Standard
    2. Proposed Emission Standard Levels 
    
[[Page 4601]]

    B. Administrative Program Flexibility
    1. Sterndrive and Inboard Engine Manufacturers
    2. Small Marine Compression-Ignition Engine Manufacturers
    3. Old Technology Two-stroke Outboard Engine and Personal 
Watercraft Manufacturers
    4. Other Potential Administrative Program Changes
    C. Small Manufacturer Criteria
    1. Sterndrive and Inboard Engine Manufacturers
    2. Marine Compression-Ignition Engine Manufacturers
    3. Outboard Engine and Personal Watercraft Manufacturers
    D. Relative Use By Age Function
    E. Manufacturer Production Line Testing Program
    F. In-use Credit Program
    G. Labeling Requirements
    H. Addition of Combat Exclusion and Competition Exclusion
    I. Engine Family Definition
    J. Harmonization with the International Maritime Organization
V. Public Participation
    A. Comments and the Public Docket
    B. Public Hearing
VI. Administrative Requirements
    A. Reporting and Recordkeeping Requirements
    B. Impact on Small Entities
    C. Executive Order 12866
    D. Unfunded Mandates Reform Act of 1995

III. Statutory Authority and Background

A. Statutory Authority

    Authority for the actions proposed in this notice is granted to EPA 
by 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)].

B. Background

    Pursuant to section 213(a) of the Clean Air Act as amended 
(hereafter, ``CAA''), EPA undertook a study of emissions from nonroad 
engines and vehicles to determine whether such emissions are 
significant contributors to ozone or carbon monoxide (CO) 
concentrations in more than one nonattainment area. A nonattainment 
area is a specified area that has failed to attain the applicable 
National Ambient Air Quality Standard (NAAQS) for a given pollutant. 
Based on the 1991 Nonroad Engine and Vehicle Emission Study (available 
in the docket),1 EPA determined that nonroad emissions do, in 
fact, contribute significantly to ozone and CO concentrations in more 
than one NAAQS nonattainment area.2

    \1\ EPA Publication Number 211A-2001 (November, 1991).
    \2\ 59 FR 31306 (June 17, 1994).
---------------------------------------------------------------------------

    Under section 213(a)(3) of the CAA, EPA is required to regulate 
those categories or classes of new nonroad engines and vehicles that 
contribute to ozone and CO air pollution. On November 9, 1994, EPA 
published a Notice of Proposed Rulemaking (NPRM) for emission standards 
for new gasoline spark-ignition (SI) and diesel compression-ignition 
(CI) marine engines pursuant to section 213(a) of the CAA.3 The 
primary pollutants affected by this rule include ozone and hydrocarbons 
(HC) for gasoline SI engines and oxides of nitrogen (NOX) for CI 
engines. In addition, EPA proposed that this rule have some impact on 
particles smaller than 10 microns (PM10) and carbon monoxide (CO). 
EPA believes the standards proposed in this rule will reduce HC 
emissions from SI engines and reduce NOX from CI engines and help 
areas come into compliance with the ozone NAAQS.

    \3\ 59 FR 55930 (November 9, 1994).
---------------------------------------------------------------------------

    The proposed gasoline SI marine engine HC emission standards should 
decrease HC emissions from marine engines by approximately 75% from 
projected baseline emission levels by the year 2025. HC emission levels 
are estimated to be stabilized at this percentage reduction through 
complete fleet turnover by the year 2051. Emission reductions due to 
this regulation for diesel CI marine engines are expected to be 
equivalent on a per-engine basis to the reductions achieved from land-
based CI engines. Land-based CI engines were estimated to achieve a 
reduction in NOX of approximately 37% per year on a per-engine 
basis (see 59 FR 31306).
    In the course of the comment period for the NPRM, some commenters 
suggested that EPA consider new approaches to some of the items 
addressed in the proposal; also, it became apparent that some aspects 
of the proposed regulation were not addressed in sufficient detail in 
the NPRM and needed additional development for further comment. This 
SNPRM seeks to address these matters. Today's notice modifies only 
those aspects of the November 9, 1994, NPRM that are identified herein; 
the remainder of the proposals set forth in the NPRM remain unchanged 
except to the extent necessary to make them consistent with the 
proposals set forth in this SNPRM.
    EPA has received an extension of the court ordered deadline for the 
final rulemaking. The final rulemaking must now be signed by the 
Administrator on May 31, 1996. The court ordered deadline for this 
rulemaking is set forth in a consent decree resulting from consolidated 
lawsuits brought by Sierra Club and the Natural Resources Defense 
Council against the Administrator.4

    \4\  Sierra Club v. Browner, Civil no. 93-0124 NHJ (D.D.C.).
---------------------------------------------------------------------------

IV. Proposed Changes; Discussion of Issues

A. Emission Standards for Spark-Ignition Engines

    EPA has received comment suggesting that a more appropriate form of 
average emission standard for spark-ignition engines is of the type 
``HC+NOX.'' Comment has indicated that an HC+NOX average 
emission standard more appropriately recognizes the inherent SI engine 
technology trade-off between reductions in HC and necessary increases 
in NOX. EPA is proposing a HC+NOX average emission standard 
structure for spark-ignition marine engines. Additionally, comment was 
received indicating that the SD/I emission standards as proposed were 
unnecessarily stringent and counterproductive. EPA is proposing 
different SD/I emission standards for HC and NOX (now proposed as 
an average HC+NOX standard) that will not require any physical 
changes to SD/I engines.
1. HC+NOX Emission Standard
    From an engineering perspective, it is clear that exhaust or engine 
out HC reductions from charge crankcase scavenged 2-stroke engines 
(e.g., old technology 2-stroke) of the magnitude proposed in the NPRM 
lead to a small NOX increase for all spark-ignition internal 
combustion engines that do not utilize catalyst or exhaust gas 
recirculation technology. The HC and NOX balance can be adjusted 
to some extent through other means, but some NOX increase is 
inevitable if HC reductions are finalized on the order of magnitude 
proposed. EPA recognized this fact in the NPRM by setting a HC average 
emission standard for outboards and personal watercraft (OB/PWC) that 
achieved dramatic reduction (i.e., a 75% reduction) and setting a 
NOX standard that was targeted at the average of the necessary 
increase in NOX (i.e., 6.0 g/kw-hr) across the fleet.
    Comment received in response to the NPRM from some in industry 
indicated that the NOx emission standard proposed was too stringent and 
that a HC+NOX average emission standard structure would be more 
appropriate. Commentors indicated that a HC+NOX average emission 
standard would provide them with needed flexibility when attempting to 
appropriately calibrate the OB/PWC four-stroke and direct-injection two 
stroke technology. 

[[Page 4602]]
Further, the NPRM standards would likely cause a low HC engine that 
generated positive emission credits according to the HC average 
standard to at the same time generate negative NOX credits. 
Therefore, the low HC engine would have to cover the negative NOx 
credits with positive NOX credits from other engines. These other 
engines would by nature have higher HC. Commentors suggested that the 
way to address this perverse effect would be to set a HC+NOX 
average emission standard. Although the perverse effect exists under 
combined or separate HC and NOX standards, a combined standard 
gives manufacturers more flexibility to calibrate engines, while still 
achieving overall targets. This calibration flexibility may be 
appropriate because OB/PWC four stroke technology and direct injection 
two stroke technology have similar overall levels of HC+NOX, while 
four stroke technology is cleaner on HC but would be more likely to 
have emissions above the separate average NOX emission standard. 
An emission standard which allows HC and NOX to be averaged 
together may treat these two control technologies more equitably.
    EPA did not propose a HC+NOX average emission standard 
structure in the NPRM and requested comment surrounding the relative 
valuation of HC versus NOX in terms of air quality. Air quality is 
determined according to a variety of local and regional conditions, 
including the relative background concentrations of volatileSince the 
NPRM, EPA has moved forward with two rulemakings that contain 
HC+NOX emission standards based on a 1 to 1 weighting of the two 
pollutants. This type of emission standard, HC+NOX, was finalized 
for small gasoline engines under 19 kilowatt,5 was discussed in an 
ANPRM for on-highway heavy-duty engines,6 and has been promulgated 
for on-highway heavy-duty engine emission standards in the past. The 
issue of weighting other than 1 to 1 did not appear to be a concern in 
public comment to these prior rulemakings. EPA requests further comment 
on the issue of weighting.

    \5\ 60 FR 34582 (July 3, 1995).
    \6\ 60 FR 45580 (August 31, 1995).
---------------------------------------------------------------------------

    Further a HC+NOX average ard structure is inherently inter-
pollutant averaging. The Agency is not opposed to considering inter-
pollutant averaging as a form of emission standard structure.
    With respect to this particular marine regulation, EPA believes 
this combined HC+NOX average emission standard may be less of a 
potential concern from the perspective of air quality and HC/NOX 
weighting given the magnitude of the large HC inventory reductions 
proposed and the comparatively tiny increase in NOX inventories, 
which are small to begin with, resulting from the separate HC and 
NOX emission standards proposed in the NPRM. Further, the NOX 
emission standard is proposed to be phased into a combined HC+NOX 
emission standard over the 9 year phase-in period at a gradual rate, 
rather than allowing the final year NOX increase in the first year 
of implementation (see detailed discussion of proposed NOX phase-
in in section IV.A.1.a below). Thus, it doesn't appear reasonable to 
say that a HC+NOX average emission standard structure would have a 
significant negative environmental impact. However, EPA requests 
comment should anyone think there may be a negative environmental 
impact.
    EPA requests comment on its proposal to finalize a HC+NOX 
average emission standard for spark-ignition gasoline engines. 
Commenters are encouraged to comment on the appropriateness of an 
HC+NOX average emission standard, as well as any variation on the 
proposal. EPA is particularly interested in any data that may further 
characterize the relative value of HC versus NOX with respect to 
air quality. Among other possibilities, should EPA determine that the 
combined standard would have a negative environmental impact, EPA may 
finalize separate HC and NOX average standards for SI engines. 
However, the flexibilities afforded by a HC+NOX emission standard 
may encourage manufacturers greater flexibility to bring clean HC 
technology into the marketplace earlier than if the standards were 
separate.
2. Proposed Emission Standard Levels
    a. OB/PWC. EPA proposes to retain the NPRM average emission 
standard levels for OB/PWC of 6.0 g/kw-hr NOX and the associated 
HC average emission standards which result in a 75% reduction in HC by 
model year (MY) 2006. The HC+NOX average emission standard for OB/
PWC is proposed to be the sum of these NPRM proposed average emission 
standard levels, although NOX is proposed to be phased-in 
gradually over the 9 year phase-in period. Therefore, the following 
formulas and tables summarize the HC+NOX average emission standard 
proposed today for OB/PWC.7

    \7\  The level of the OB/PWC emission standard for CO proposed 
in the NPRM remains unchanged.

HCbase=(151+557/P0.9) or 300 g/kW-hr, whichever is lower, 
where:
HCbase=hydrocarbon base emission standard in g/kW-hr
P=rated power of the engine family in kilowatt (kW).

    This HCbase is reduced over a 9 year phase-in period beginning 
in MY 1998 and ending in MY 2006. The average HC standard curve for a 
given MY is determined by the product of the HCbase curve function 
and the MY factor as shown in Table 1. The MY factor reflects equal 
percentage reductions per year from the baseline over the nine year 
phase-in period, resulting in a 75 percent decrease when fully 
implemented. For example, the average HC emission standard in 2004 is 
the product of the 2004 HC MY factor, 0.417, and the HCbase 
function. The resulting HC average emission standard function for MY 
2004 is as follows:

    Also, given the limitation on HCbase of 300 g/kW-hr maximum, 
the 2004 emission standard may not be greater than 0.417-300=125.1 g/
kW-hr.

         Table 1--Gasoline Spark-Ignition OB/PWC Marine Engines         
                     [HC Average Emission Standards]                    
------------------------------------------------------------------------
                                                                  HC MY 
                          Model year                             factor 
------------------------------------------------------------------------
1998..........................................................     0.917
1999..........................................................     0.833
2000..........................................................     0.750
2001..........................................................     0.667
2002..........................................................     0.583
2003..........................................................     0.500
2004..........................................................     0.417
2005..........................................................     0.333
2006 and after................................................     0.250
------------------------------------------------------------------------

    Table 2 contains the HC+NOX average emission standards 
proposed today. These average emission standards represent the 
summation of the average emission standards proposed in the NPRM.

         Table 2.--Gasoline Spark-Ignition OB/PWC Marine Engines        
                   [HC+NOX Average Emission Standards]                  
------------------------------------------------------------------------
                                        HC+NOX average emission standard
              Model year                              by MY             
------------------------------------------------------------------------
1998..................................  (0.917 x (151+557/P0.9))+ (1/9 x
                                         (6.0-2.0))+2.0.                
1999..................................  (0.833 x (151+557/P0.9))+ (2/9 x
                                         (6.0-2.0))+2.0.                
2000..................................  (0.750 x (151+557/P0.9))+ (3/9 x
                                         (6.0-2.0))+2.0.                
2001..................................  (0.667 x (151+557/P0.9))+ (4/9 x
                                         (6.0-2.0))+2.0.                
2002..................................  (0.583 x (151+557/P0.9))+ (5/9 x
                                         (6.0-2.0))+2.0.                
2003..................................  (0.500 x (151+557/P0.9))+ (6/9 x
                                         (6.0-2.0))+2.0.                
2004..................................  (0.417 x (151+557/P0.9))+ (7/9 x
                                         (6.0-2.0))+2.0.                

[[Page 4603]]
                                                                        
2005..................................  (0.333 x (151+557/P0.9))+ (8/9 x
                                         (6.0-2.0))+2.0.                
2006 and after........................  (0.250 x (151+557/P0.9))+6.0.   
------------------------------------------------------------------------
    The proposed HC+NOX average emission standards in Table 2 are 
derived by adding the average HC emission standards that were proposed 
in the NPRM to phased-in NOX levels. The NOX baseline is 2.0 
g/kw-hr and is gradually increased over the phase-in to 6.0 g/kw-hr. 
EPA chose this phase-in approach for the NOX part of the average 
HC+NOX emission standard because it encourages manufacturers to 
avoid increases in NOX all at once.
    Further, this approach eliminates any windfall credits from the 
point of view of the market as a whole. If the standard were phased-in 
differently, such as setting the 6.0 NOX level earlier than MY 
2006, it could be said that windfall NOX credits would be 
generated because the overall fleet NOX average would be less than 
6.0 g/kw-hr. Currently, the overall fleet average is at approximately 
2.0 g/kw-hr. To allow a 6.0 g/kw-hr NOX average in 1998 allows 
windfall credits equal to nearly 4.0 g/kw-hr of NOX from the 
emission rate perspective. EPA mitigated this problem somewhat in the 
NPRM proposal by disallowing NOX credit banking. However, it would 
be cumbersome to disallow NOX credit banking under today's 
proposal for a combined HC+NOX average standard in which a 
NOX credit is the same as a HC credit. Therefore, EPA thinks the 
most equitable way to phase-in the targeted NOX level is to 
gradually phase it in from the current levels to the targeted level in 
equal percentages over the 9 year phase-in, which is what is proposed 
in Table 2.
    At the same time, because of the inherent flexibility with a 
combined HC+NOX emission standard, the gradual phase-in should not 
inhibit the introduction of clean technology early. Further, the 
HC+NOX emission standard clearly achieves the same overall control 
as the proposal, if not better control for NOX. EPA requests 
comment on the way in which the NOX average emission standard 
should be combined with the HC emission standard over the phase in 
period. Comment should address the specific NOX numbers that are 
recommended for each phase-in year and the rationale supporting the 
recommendation, including whether windfall credits are associated with 
the recommendation.
    b. SD/I Engines. Comment received on the NPRM from some in industry 
indicated that the proposed emission standards for sterndrive and 
inboard (SD/I) engines are inappropriate because they would increase 
costs and thereby depress sales of SD/I engines, the cleaner engines in 
the marketplace. As stated in the NPRM, uncontrolled SD/I's are cleaner 
than OB/PWC's would be in the MY 2006 at the end of the phase-in. When 
EPA proposed emission standards for SD/I engines in the NPRM, EPA 
thought the standards would incur very little, if any, additional costs 
because they would simply require tighter calibration of SD/I engines.
    Now, comments suggest that the necessary engines changes would be 
more costly than EPA expected and would adversely affect SD/I operation 
and sales. The emission standards proposed in the NPRM would require 
the manufacturers to spend money recalibrating the engines. The 
recalibration would cause the engines to have poor operating 
characteristics. Alternatively, because manufacturers may not meet the 
corporate average by recalibration alone, exhaust gas recirculation may 
need to be applied. Exhaust gas recirculation is costly and would not 
provide much environmental benefit. Hence, EPA now believes it would be 
counterproductive for EPA to finalize the emission standards for SD/I 
engines proposed in the NPRM because that action would introduce 
negative market forces which would discourage manufacturers from 
expanding the market with new models of cleaner SD/I engines and 
discourage people from buying the cleaner engines.
    For these reasons, EPA is now proposing to apply two-thirds of the 
final phase-in MY 2006 OB/PWC HC+NOX emission standard to SD/I 
engines as an emission cap beginning in SD/I MY 1998.8 Thus, SD/I 
engines would not be allowed to exceed two-thirds of the MY 2006 OB/PWC 
average emission standard of (0.250 x (151+557/P0.9))+6) in the 1998 
MY. Therefore, the 1998 MY emission standard for SD/I engines is shown 
in the following equation.

    \8\ The level of the SD/I emission standard for CO proposed in 
the NPRM would remain unchanged.
---------------------------------------------------------------------------

    EPA believes that SD/I engines are much cleaner than this proposed 
emission level. All data available to EPA clearly shows that all SD/I 
engines have emission levels that are much below this level. Therefore, 
manufacturers will not need to make any changes to SD/I engines to 
achieve two-thirds of the MY 2006 OB/PWC average emission standard as a 
cap type standard.
    EPA requests comment on this emission standard proposal for SD/I 
engines, particularly any comment indicating that any particular type 
of SD/I engine would be likely to exceed the proposed level. Refer to 
the docket for further discussion of the emission levels associated 
with SD/I engines.9 EPA does not think backsliding is a concern at 
the proposed emission standard level, primarily because if backsliding 
were to occur, it seems that it would have occurred already, since 
these engines are currently unregulated and future technology is more 
likely to result in lower emissions, not backsliding.

    \9\ Refer to EPA Air Docket A-92-28, docket submission IV-H-01.
---------------------------------------------------------------------------

    EPA is considering whether a report should be submitted by the SD/I 
industry or by SD/I manufacturers that indicates the emission levels of 
the engines based on the voluntary testing that is performed by 
manufacturers. For example, manufacturers already do testing of the SD/
I engines. Requiring a biennial report of this data (e.g., test results 
on specific test procedures, engine family identification, test fuel, 
type of engine: prototype, development, production, in-use or field 
engine) would seem to adequately identify if backsliding is or is not 
occurring. EPA requests comment whether EPA should finalize such a 
requirement or whether EPA should ask the SD/I manufacturers to submit 
these reports voluntarily.
    In the alternative, EPA proposes not to apply emission standards to 
SD/I engines. EPA believes Section 213(a)(3) of the CAA offers the 
Agency the flexibility either to finalize the emission standards for 
SD/Is proposed above or not to impose emissions standards for SD/I 
engines, given the unique circumstances presented by SD/Is.
    Section 213(a)(3) directs EPA to establish emission standards for 
``classes or categories'' of new nonroad engines which achieve ``the 
greatest degree of emission reduction achievable through the 
application of technology * * *., giving appropriate consideration to 
the cost of applying such technology within the period of time 
available to manufacturers'' and other factors. 42 U.S.C. 7545(a)(3). 
EPA is proposing in this rule to treat all marine spark-ignition 
engines as one ``class or category'' of new nonroad engines for which 
EPA is establishing emission standards. SD/Is would constitute a 
subclass or subcategory of the marine SI class or category. Given this 
approach, 

[[Page 4604]]
EPA believes that, depending on circumstances on which it seeks comment 
below, the HC+NOX emission standards proposed for OB/PWC plus 
either (1) an SD/I HC+NOX standard set at two-thirds the MY 2006 
OB/PWC HC+NOX standard, or (2) no SD/I HC or NOX standard at 
all, satisfy the criteria set forth in Section 213(a)(3). In the unique 
circumstances presented by the marine SI industry, HC+NOX 
emissions from unregulated SD/I engines will be dramatically cleaner 
than HC+NOX emissions from MY 2006 and later OB/PWC regulated at 
the levels proposed in this rule. Given this fact, and the opportunity 
for some substitution of SD/I for OB/PWC in the market place, it is 
appropriate for EPA to consider what emissions standards for SD/Is, if 
any, will most appropriately satisfy the criteria of Section 213(a)(3) 
viewing marine SI engines as a whole.
    EPA believes that requiring SD/Is to comply with two-thirds of the 
MY 2006 OB/PWC HC+NOX emission standard would achieve greater 
emission reductions from marine SI engines as a class or category than 
would the SD/I emission standard levels proposed in the NPRM, at less 
cost to the SI industry. If EPA were to finalize the SD/I HC+NOX 
standard proposed in this Notice, EPA would encourage the cleaner SD/I 
engine in the market by virtually eliminating any regulatory costs on 
SD/I engines (see discussion of minimal administrative burdens for SD/I 
engines, in Section IV.B.1 below). The minimal regulatory burden and 
consequent lower cost to SD/I engines compared to OB/PWC should 
encourage manufacturers to offer a greater range of SD/I engines, 
including smaller SD/Is that could compete with OB/PWC. Public health 
and the environment in turn would benefit from the emissions reductions 
achieved through any substitution of SD/Is for OB/PWC, and at lesser 
cost to the SI engine industry than the more stringent SD/I emission 
standards proposed in the NPRM.
    EPA is proposing in the alternative to finalize no HC or NOX 
emission standard for SD/Is because this alternative may achieve 
greater emission reductions from SI engines as a class or category than 
would the SD/I HC+NOX standard proposed above, at lesser cost to 
the SI industry. While the option of applying two-thirds of the MY 2006 
OB/PWC HC+NOX standard level to SD/I would not require 
manufacturers to physically change the engines in any way, to the 
extent that an emission standard causes any costs on SD/I such 
standards may tend to limit SD/I substitution for OB/PWC and therefore 
offer somewhat less emission reductions than would no SD/I standard at 
all, while imposing more cost on the SI industry.
    On the other hand, finalizing no HC or NOX emission standard 
for SD/I raises a potential concern. There is the issue whether SD/I 
standards at the level proposed here would offer a useful backstop 
against emissions backsliding by SD/I. EPA believes that emissions 
backsliding, i.e. worsening emissions performance, may not be a 
realistic concern with SD/I because of several reasons. First, engines 
have been moving to electronic fuel injection which will provide better 
control over engine and emission performance. Customer demands for both 
low odor and low smoke discourage manufacturers from selling engines 
that have higher emissions. Also, the engine block manufacturers are 
improving the emission performance of the engine blocks because of 
carryover of on-highway emission performance through engine block 
design.
    If EPA finalizes the alternative of no HC or NOX emission 
standard for the SD/I subcategory, EPA is considering a suboption of 
issuing guidance to states that provides information of the relative 
emissions form the class or category of SI engines. This guidance would 
explain that EPA did not finalize emission standards on the subclass or 
subcategory of SD/I engines because they are relatively clean. EPA 
requests comment on whether industry could or should provide data 
either voluntarily or by requirement that exemplifies the emission 
characteristices of the fleet of SD/I engines in support of this 
guidance based on the current testing of SD/I engines which industry 
already performs as noted above. EPA also requests comment on the 
frequency with which manufacturers should be asked to require to 
provide such data.
    EPA requests comment on both proposals regarding HC and NOX 
emission standards for SD/I. Commenters should specify whether they 
prefer some level of HC and NOX emission standards for SD/Is or 
none at all, and why they prefer one approach versus the other. If the 
commenter favors some level of HC and NOX emission standards, EPA 
requests comment on the proposal to combine the standards into one 
HC+NOX standard and on the proposal to set the HC+NOX 
standard for SD/I at two-thirds of the proposed MY 2006 HC+NOX 
standard for OB/PWC. Should a commenter prefer a different HC+NOX 
standard level, EPA encourages the commenter to identify the standard 
level that it prefers and offer an explanation for this preference. EPA 
also seeks comment on which approach toward SD/I emissions best 
satisfies the criteria set forth in Section 213(a)(3), and why.

B. Administrative Program Flexibility

    The Agency is proposing several modifications to the proposed rules 
in order to offer administrative program flexibility to certain types 
of engine technology and certain categories of small manufacturers, as 
described below in section C., ``Small Manufacturer Criteria.''
    Specifically, EPA proposes to allow the Administrator to certify 
all sterndrive and inboard (SD/I) engine families on the basis of much 
less information than that proposed in the NPRM. As explained in more 
detail below, EPA proposes to find as part of this rulemaking that EPA 
currently has enough testing and other information regarding engines 
which meet EPA's proposed regulatory definition for SD/I such that 
additional emissions testing is not needed to determine if an SD/I 
engine family should be certified as conforming to the HC+NOX and 
CO standards proposed in this rule. This finding would make it 
unnecessary for manufacturers to submit test results in order to 
receive a certificate of conformity. To apply for a certificate for an 
engine family, the manufacturer need only submit a simple affirmation 
that the engine family meets the SD/I definition and related 
affirmations. Upon receipt of the affirmations, EPA would typically 
issue a certificate of conformity. In addition, EPA proposes to exempt 
all SD/I engine manufacturers from the proposed regulatory provisions 
concerning manufacturer-conducted production line and in-use testing 
requirements, related test equipment and test procedure provisions, and 
selective enforcement auditing.
    The Agency received comments urging EPA to drop manufacturer-
conducted production line testing and in-use testing requirements for 
all marine CI engines proposed in the NPRM. The Agency now proposes to 
apply the EPA-directed production line and in-use testing provisions to 
marine CI engines that already apply to similar land-based CI engines, 
as set forth in 40 CFR part 89 and discussed in more detail below. 
Also, the Agency is proposing some administrative program flexibilities 
for old technology two-stroke outboard and personal watercraft (OB/PWC) 
engines, for the reasons set forth below.
    Finally, EPA proposes that the administrative programs for small 
marine CI engine manufacturers focus 

[[Page 4605]]
on simplified certification and reduced enforcement requirements.
    EPA believes it has authority under the CAA to offer the 
administrative program flexibility that is described in more detail 
below. The CAA states that the marine engine emission standards, when 
finalized, shall be subject to Sections 206, 207, and 208 of the Act, 
``with such modifications of the applicable regulations * * * as the 
Administrator deems appropriate.'' 42 U.S.C. 7547(d). This statutory 
language grants EPA substantial discretion to offer flexibility in the 
compliance provisions of the marine engine final rule. The paragraphs 
below describe each of the administrative program flexibility 
provisions proposed in this SNPRM and explains EPA's rationale for 
offering such flexibility.
1. Sterndrive and Inboard Engine Manufacturers
    The Agency believes that any regulations it issues for marine 
engines should offer substantial compliance flexibility to 
manufacturers of gasoline-fueled SD/I engines, in part because the 
market is comprised mostly of small manufacturers, but principally 
because the engines are inherently low-emitting compared to the OB/PWC 
alternative. In fact, the market is moving towards even cleaner 
technology (e.g., electronic fuel control) in the future without 
regulation. In the absence of compliance flexibility, small SD/I engine 
manufacturers would be particularly at risk, because their receipts 
would not bear the cost of compliance as proposed in the NPRM. The 
Agency does not wish to drive out of business manufacturers of engines 
that are already relatively clean.
    The Agency recognizes that four-stroke SD/I engines are currently 
cleaner than OB/PWC engines with respect to hydrocarbon (HC) emissions. 
Even at the 75 percent HC reduction level proposed in the NPRM for OB/
PWC engines, SD/I engines will still be much cleaner on average than 
controlled OB/PWC. Because EPA wants its regulations to encourage 
purchasers to substitute SD/I engines for OB/PWC engines rather than 
hinder that substitution, it is proposing certification flexibility for 
all manufacturers of SD/I engines as a means of keeping the costs of 
SD/I engines low.
    In the NPRM, EPA discussed the issue of averaging between OB/PWC 
engines and SD/I engines as a way to encourage the purchase of the 
inherently cleaner SD/I engines. The Agency stated at that time that 
substitution of SD/I engines for OB/PWC engines was possible in some 
horsepower ranges and was environmentally desirable. In developing the 
NPRM, EPA considered averaging systems and other mechanisms (such as 
relative standard stringency) to encourage this substitution.
    Comments on the proposal stated that many SD/I engine manufacturers 
were in fact very small operations that marinized engine blocks 
purchased from automobile manufacturers. Some of these companies only 
employ two people. Additionally, EPA received comment that the 
certification and testing burden was very onerous for such entities. 
The standards originally proposed for SD/I engines were set at a level 
that EPA believed would keep prices low and encourage growth in the SD/
I market relative to the market for OB/PWC. However, these commenters 
believed that, rather than encouraging the growth of the cleaner SD/I 
market, EPA's proposed administrative program would have the unintended 
effect of forcing small SD/I manufacturers out of business, shrinking 
competition, and raising SD/I prices.
    The Agency remains interested in encouraging the relatively clean 
SD/I technology and is concerned that burdens of certification and 
other administrative programs would decrease the substitution of SD/I 
engines for the higher-polluting OB and PWC. Therefore, EPA is 
proposing a very minimal certification process for all manufacturers of 
SD/I engines.
    The certification process is proposed to simply include 
manufacturer submittal of an affirmation that the engine family meets 
the regulatory definition of a sterndrive or inboard engine, an 
affirmation that the manufacturer has no test data or other engine 
information indicating that the engine family would not comply with the 
emission standard, and an affirmation that the engine meets applicable 
safety requirements. Upon receipt, the Administrator would issue a 
certificate of conformity, unless, based on all available information, 
the Administrator has reason to believe that the engine family may not 
comply with the applicable emission standards and safety requirements 
and therefore is not able to determine that the engine family conforms 
and should be issued a certificate.
    While EPA believes current SD/Is meet the proposed emission 
standards,10 circumstances could arise in the future where EPA may 
have reason to doubt that a particular engine family meets the 
applicable emission standards. Therefore, the Agency proposes that in 
such circumstances EPA may require, at its discretion, other 
information on the engine family in addition to the affirmations 
specified above. For example, the Administrator may require the 
manufacturer to demonstrate (on the basis of previous emission tests, 
development tests, or other information) that the engine will conform 
with applicable emission standards.

    \10\ Refer to EPA Air Docket A-92-28, docket submission IV-H-01
---------------------------------------------------------------------------

    EPA expects this proposed certification process for SD/I engine 
families to be straightforward, involving no more than a one-page 
submittal to EPA and an expedient issuance of a Certificate of 
Conformity. The submitter would not need to include test results in its 
submission if its engine meets the regulatory definition of an SD/I 
because EPA would determine as part of this rule that EPA currently has 
enough emission testing and other information regarding engines meeting 
the SD/I definition such that additional testing is not needed to 
determine whether an SD/I engine family should be certified. EPA is 
exploring electronic data submission alternatives that may make this 
process virtually burden free for the manufacturers.
    EPA believes that the engines currently are well below the proposed 
emission standards level. EPA has test results on the regulated test 
procedure and confidential test result information submitted by 
manufacturers. All data is presented in the docket, with any 
confidential data masked so as to not reveal its origin (Refer to 
Docket A-92-28 submission IV-H-01). EPA encourages comment on this data 
and the submission of further data that either supports or refutes the 
data presented.
    The Agency believes Section 206 of the CAA offers it the 
flexibility to determine through rulemaking that EPA currently has 
enough testing and other information such that additional emissions 
testing is not needed to determine whether an SD/I engine family should 
be certified as conforming to the applicable emission regulations 
(i.e., a cap of two-thirds of the proposed MY 2006 OB/PWC HC+NOX 
emission standard and a CO cap of 400 g/kW-hr). While the language of 
Section 206 contemplates an individualized, adjudicatory procedure, 
Supreme Court precedent allows EPA to establish issues common to many 
adjudications through rulemaking. See American Hospital Assn. v NLRB, 
499 U.S. 606, 612 (1991) (``[E]ven if a statutory scheme requires 
individualized determinations, the decisionmaker has the authority to 
rely on rulemaking to resolve certain issues of general applicability 
unless Congress 

[[Page 4606]]
clearly expresses an intent to withhold that authority.''); Heckler v. 
Campbell, 461 U.S. 458, 467 (1983). Therefore, EPA believes that, 
should information available to EPA support a finding that additional 
emissions testing is not needed to determine whether SD/I engine 
families should be certified, it is appropriate for EPA to make this 
finding through rulemaking and offer for SD/I engine families the 
minimally burdensome individualized determination of conformity 
described above. The individualized determination would basically 
address whether a specific engine family fits the definitions for 
sterndrive or inboard engines proposed in this rulemaking, whether 
there is any information specific to this engine family that would 
indicate it in fact would not be expected to conform to the standards, 
and whether the specific engine family meets the safety criteria of 
Section 206(a)(3).
    In addition, the Agency is proposing to exempt SD/I manufacturers 
from manufacturer production-line testing, manufacturer in-use testing, 
and related requirements. EPA is proposing that Part 91 subparts A, B, 
H, I (recall portions only), J, K, L, and M apply to SD/I engine 
manufacturers, in order to minimize compliance burdens on these 
manufacturers. While EPA is proposing to exempt these manufacturers 
from subpart G, Selective Enforcement Auditing, EPA retains authority 
under Sections 206(b)(1) and 208 to test newly manufactured engines and 
to inspect production facilities and processes to determine whether the 
manufacturer is complying with the information submitted for 
certification. Further, EPA would retain authority under Section 
206(b)(2) of the CAA to suspend or revoke the certificate for engines 
that do not conform with applicable emission standards.
    The Agency requests comment as to which subparts of Part 91 should 
apply to SD/I engine manufacturers and comment on the proposed 
certification process as outlined above. Further, EPA requests comment 
on its proposal to find through rulemaking that EPA currently has 
enough testing or other information regarding engines which meet the 
regulatory definition of SD/I such that additional emissions testing is 
not needed to determine whether an SD/I engine family conforms to the 
HC+NOx and CO emission standards proposed in this rule. Commenters 
are encouraged to evaluate the data on which EPA proposes to rely and 
to submit to EPA any additional relevant data, together with the 
commenter's evaluation of the submitted data. EPA requests comment on 
whether it is appropriate to treat small SD/I manufacturers (as defined 
in more detail below) separately from other SD/I manufacturers and if 
so, which subparts of Part 91 should apply to small SD/I manufacturers.
    Finally, EPA requests comment on the following definitions. A 
sterndrive engine is defined as a 4-stroke engine (unless otherwise 
designated by the Administrator (e.g., a personal watercraft engine)) 
that is designed such that the drive unit is external to the hull of 
the marine vessel, while the engine is internal to the hull of the 
marine vessel. An inboard engine is defined as a 4-stroke engine 
(unless otherwise designated by the Administrator (e.g., a personal 
watercraft engine)) that is designed such that the propeller shaft 
penetrates the hull of the marine vessel while the engine and the 
remainder of the drive unit is internal to the hull of the marine 
vessel. Commenters are encouraged to offer any changes to the 
definitions which are needed to allow the commenter to concur with 
EPA's proposal to offer non-testing certification to qualifying SD/I 
engine families. Commenters should explain the reasons for any proposed 
alterations to the definitions.
2. Marine Compression-Ignition Engine Manufacturers
    In comments responding to the NPRM, several small CI engine 
marinizers characterized the proposed administrative program 
requirements as very burdensome for the small proportion of marine 
emissions attributable to such engines. Marinizers informed EPA that 
they are truly small manufacturers, in terms of market share, 
production, and number of employees, compared to all other 
manufacturers of nonroad diesel engines. Furthermore, they believe that 
their marinized engines may already meet or nearly meet the proposed 
standards.
    The Agency now proposes certification program flexibility for small 
marine CI engine manufacturers; that is, small marine CI engine 
manufacturers will have reduced certification data requirements. EPA 
proposes that part 89 subparts A, B, C, G, H, I, J, and K be applicable 
to manufacturers of small marine CI engines. This means that the 
manufacturers must undergo simplified certification, while receiving 
reduced enforcement requirements. While EPA is not proposing to apply 
subpart F, Selective Enforcement Auditing, to small marine CI 
manufacturers, EPA retains authority under Sections 206(b)(1) and 208 
to test newly manufactured engines and to inspect production facilities 
and processes to determine whether the manufacturer is complying with 
the information submitted for certification. The Agency requests 
comment as to which subparts of part 89 should apply to the small 
marine CI engine manufacturers. Further, EPA would retain authority 
under Sec. 206(b)(2) of the CAA to suspend or revoke the certificate 
for engines that do not conform with applicable emission standards.
    The Agency has also received comments in response to the NPRM 
regarding the proposed production line testing and in-use testing 
requirements for marine CI engines. The comments did not support 
finalizing these two types of testing because EPA does not require them 
for land-based CI engines in 40 CFR Part 89, and many of the land-based 
CI engine blocks are used for marine CI engines. If EPA were to 
finalize its requirements as proposed, it would impose different 
requirements for essentially some of the same engines. The Agency has 
considered these comments and now proposes not to require marine CI 
engines to comply with the production line and in-use testing 
provisions that EPA proposed in the NPRM; instead, it proposes that 
marine CI engines be subject to the SEA testing and recall provisions 
that apply to land-based CI engines greater than 37 kilowatts (kW), as 
set forth in 40 CFR part 89.
3. Old Technology Two-Stroke Outboard Engine and Personal Watercraft 
Manufacturers
    The Agency received significant comment in response to the NPRM 
regarding the cost of the CO cap and the administrative programs 
relative to old technology two-stroke OB/PWC engines. Manufacturers 
argued that imposing the costs of certification testing and enforcement 
program testing on old technology OB/PWC engines is not a cost-
effective requirement, because these engines will be phased out of 
production anyway. Also, many old technology OB/PWC exceed the CO cap, 
and money would have to be spent to redesign product and production 
lines to meet the CO standard even though the old technology will be 
phased-out. Manufacturers would rather spend their limited resources on 
developing and producing cleaner, new technology OB/PWC engines.
    The new, cleaner technology will require extensive changes in 
production lines and engine design. The engine changes do not simply 
consist of using a different fuel system, but involve designing whole 
new engines. Therefore, millions of dollars will be 

[[Page 4607]]
needed for capital costs over the phase-in period. Manufacturers have 
also stated that they would potentially have to build twice as many 
test facilities, in order to accommodate testing of old technology OB/
PWC engines as well as new technology OB/PWC engines, and half of those 
facilities would become idle as soon as manufacturers phased out the 
old technology.
    In view of these comments, EPA believes it is appropriate to 
mitigate compliance costs for old technology OB/PWC engines through 
compliance flexibility, waivers, and an exemption from the CO cap, if 
one if finalized,11 until late in the phase-in in order to free up 
the manufacturers' limited resources for new technology development.

    \11\ EPA is currently considering whether to include a CO cap in 
the final rule in light of the comments it received on the NPRM and 
its authority under the CAA.
---------------------------------------------------------------------------

    Because adequate test data on old technology engines currently 
exists, and that data shows engine emissions are so high that 
production variance and in-use deterioration are almost negligible 
effects on the emission rate, imposing compliance costs on the old 
technology engines would appear to yield little environmental benefit. 
On the other hand, allowing flexibility in the administrative programs 
for old technology, which will be phased out of production, will free 
up money for the manufacturers to develop, produce, and market clean 
technology OB/PWC engines.
    The Agency proposes to define old technology OB/PWC engine families 
to be those that were in production for the 1997 and previous model 
years and that did not utilize newer technologies, such as four-stroke 
technology, direct-injection two-stroke technology, or catalyst 
technology. It is important to restrict this definition to engines that 
were in production both in 1997 and in previous years so an old 
technology engine family that would first be brought into production in 
1997, just before the regulations take effect, could not take advantage 
of flexibilities proposed here. EPA does not want to allow the 
flexibilities to be gamed in this way. For example, the jetboat market 
segment is expanding rapidly. A manufacturer electing to bring in a new 
jetboat engine family in 1997 that uses old technology (e.g., simple 
two-stroke engine design) should not enjoy the flexibilities proposed 
here. The reason is that the manufacturer is choosing to spend money to 
bring more old, dirty technology into the market rather than spending 
the money on new, cleaner technology. To allow flexibility in this case 
on the eve of the implementation of the proposed 1998 MY emission 
standards is contrary to EPA's intent in offering this flexibility. EPA 
is offering this flexibility to allow the manufacturers to spend money 
on new, cleaner technology rather than old, dirty technology that will 
need to be phased-out of production.
    EPA is proposing to allow manufacturers of old technology OB/PWC 
engines for a limited time to comply with reduced data submission 
requirements for certification,\12\ to be exempted from Part 91 
subparts D, E, F, G, I (non-recall portions only), J, and M containing 
compliance monitoring programs, and to be exempted from any CO cap if 
it should be included in the final rule. However, EPA does not intend 
to allow such provisions in the latter years of the nine year phase-in 
of emission standards. The Agency believes it is important to the 
integrity of the averaging standard that at the end of the nine year 
phase-in, all OB/PWC engine families submit the certification and 
testing data normally required for obtaining certificates of conformity 
and that all OB/PWC engine families comply with enforcement provisions 
and any CO cap. Further, because the phase-in period is nine years, EPA 
believes that allowing this flexibility for the first six years will be 
adequate to mitigate costs and will also encourage the introduction of 
the cleanest technology sooner.

    \12\ EPA proposes that it may accept, at its discretion, summary 
information on an old technology OB/PWC engine family in lieu of the 
full Application for Certification. The Agency plans to explain its 
policy on accepting summary information via guidance to be generated 
and distributed to manufacturers subsequent to the publication of 
the final rule. The Agency welcomes comments on whether guidance is 
the appropriate mechanism for defining the summary certification 
procedures, or whether EPA should engage in notice and comment 
rulemaking (at some later date) to define the procedures.
---------------------------------------------------------------------------

    Therefore, EPA intends to implement such flexibility through the 
sixth year of the phase-in. The Agency requests comment on allowing 
this flexibility for old technology OB/PWC engines and on EPA's 
intended implementation strategy for this flexibility. Commenters are 
encouraged to express separately their views on each aspect (i.e., 
certification, enforcement, and the CO cap) of the flexibility proposed 
here as EPA may finalize all, some, or none of them. EPA also requests 
comment on the proposed definition of old technology OB/PWC and 
alternative approaches.
4. Other Potential Administrative Program Changes
    a. Recall. EPA is considering omitting from the final spark-
ignition marine engine regulations (Part 91) those portions of proposed 
Subpart I regarding recall.
    Section 213(d) of the CAA provides that new nonroad engine emission 
standards ``shall be subject'' to the provisions of sections 206 
through 209. EPA believes that this statutory provision is self-
executing, so that the marine engine standards proposed in this 
rulemaking would be subject to the remedial provisions of section 
207(c)(1) as well as, for example, the certification provisions of 
section 206. Further, EPA believes that the remaining language in 
section 213(d) does not require promulgation of regulations to 
implement section 207(c) for marine engines, except where they are 
``necessary to determine compliance with, and enforce,'' such new 
nonroad engine standards.\13\

    \13\ See also discussion of section 207(c) and recall in the 
NPRM 59 FR 55943-46.
---------------------------------------------------------------------------

    In this rulemaking, EPA believes that it may not be necessary to 
promulgate regulations specifying procedures to implement Sec. 207(c) 
in order to determine compliance with and enforce the proposed marine 
spark-ignition emission standards. The in-use averaging, banking, and 
trading (ABT) program proposed in this SNPRM makes it highly improbable 
that EPA would ever find that a substantial number of marine engines do 
not conform to the applicable emissions standard or FEL, since any 
noncompliance may be offset through ABT. Moreover, EPA expects that 
remedial action under section 207(c) would be largely ineffective, both 
because industry structure and engine owner turnover make it difficult 
for a manufacturer to identify the owners of a nonconforming engine, 
and because safety-related recalls of marine engines have generated 
little consumer response in the past. For these reasons, EPA would not 
expect to make a determination of nonconformity under section 207(c) 
regarding marine engines. Furthermore, in the unlikely event EPA should 
find it appropriate to take such action, EPA retains authority either 
to take action directly under section 207(c) or to promulgate 
appropriate regulations at that time.
    EPA seeks comment whether EPA should omit the recall portions of 
subpart I (Part 91) from the final rule. Commenters are encouraged to 
explain the basis for their opinion, including all policy reasons and 
all circumstances regarding the marine engine industry which favor one 
approach versus another, as well as the statutory basis for the 
preferred approach. 

[[Page 4608]]

    EPA is also considering not applying Part 89, Subpart H (recall 
regulations) to small marine CI engine manufacturers (as proposed to be 
defined in this Notice). As with marine SI engines, EPA expects that 
remedial action under Sec. 207(c) for marine CI engines would be 
largely ineffective, both because industry structure and engine 
turnover make it difficult for a manufacturer to identify the owners of 
a nonconforming engine and because safety-related recalls of marine 
engines have generated little consumer response in the past.
    Unlike marine SI engine manufacturers, however, marine CI engine 
manufacturers would not have available credits generated through an in-
use credit program to offset any in-use noncompliance. Instead, it may 
be appropriate not to apply Part 89, subpart H, to small marine CI 
engine manufacturers because of the relative burden of Sec. 207(c)(1) 
remediation on smaller marine CI engine manufacturers compared to 
larger marine CI engine manufacturers. Even if the recall regulations 
were to apply, EPA would expect that any Agency decision on whether to 
take formal action under Sec. 207(c) would take into consideration the 
circumstances involved, including the nature of the industry and the 
specific manufacturer involved.
    EPA seeks comment, including any available data, on the relative 
potential burden of recall on marine CI engine manufacturers depending 
on their size. EPA also seeks comment whether EPA should apply Part 89, 
subpart H to small marine CI engine manufacturers (as proposed to be 
defined in this Notice) or whether any differences in potential 
manufacturer burden should be addressed through EPA's discretion in 
implementing the recall authority. Commenters are encouraged to explain 
the basis for their opinion, as well as any variation in the definition 
of small marine CI engine manufacturer which would affect their 
opinion.
    b. 1998 MY Corporate Average Compliance. In the NPRM, EPA proposed 
that manufacturers begin the phase-in of the spark-ignition gasoline 
engine regulations in the 1998 MY. Production of the 1998 MY will begin 
in May, 1997, and this final rulemaking is scheduled to be published in 
May, 1996. Manufacturers have informed EPA that 1998 MY compliance will 
be challenging because they must have their product line certified and 
in compliance on a corporate average basis within one year.
    EPA is concerned with this short amount of time between the final 
rule and the first year of compliance. However, EPA strongly wishes to 
implement the rulemaking for the 1998 MY so as to achieve emission 
reductions in 1998 from this source. As the emission reductions in MY 
1998 are very important, EPA is reluctant to forego the 1998 MY. 
Therefore, EPA is not revising its proposal to begin implementation in 
1998 MY.
    On the other hand, the timing is clearly tight. There may be 
circumstances where, despite the manufacturer's best efforts, 
compliance in the 1998 MY is unattainable. When compliance is 
unattainable due to circumstances that are clearly beyond the control 
of a manufacturer, it may be reasonable to allow the manufacturer some 
flexibility in compliance.
    Compliance, for this rulemaking, means having a zero or positive 
emission credit balance for the manufacturer's product line at the end 
of the MY. Therefore, a manufacturer would potentially be in 
noncompliance if it could not generate or buy sufficient positive 
emission credits to offset the amount of negative emission credits 
reflected in its product line.
    When the manufacturer is in such a situation for the 1998 MY due to 
circumstances that are clearly beyond its control, EPA is considering 
allowing the manufacturer to combine its 1998 MY ending credit balance 
(i.e., a negative balance) with the 1999 MY credit balance. In this 
special circumstance, the compliance period would be a 2 year averaging 
period across the 1998 and 1999 MYs. The advantage of this approach is 
that it inherently requires remediation of the lack of reduction in 
1998 MY.
    Circumstances beyond the control of a manufacturer would be those 
types of circumstances where the manufacturer had taken clearly laid 
out steps to make sure its product plans would be met, yet could not 
produce its appropriate complying product plans due to factors 
associated with suppliers not providing appropriate inputs. For 
example, test facilities might not be operational due to delays in 
construction that are beyond the control of the manufacturer or its 
designated contractor (e.g., the test equipment supplier did not 
deliver the equipment in time to be installed).
    EPA requests comment on the need for flexibility for 1998 MY 
compliance due to circumstances outside the control of the 
manufacturer. Specifically, EPA requests comment on allowing a 2 year 
averaging period for a manufacturer which EPA determines is in such a 
situation, including comment on a different averaging period (e.g., 18 
months, 30 months). EPA requests comment on specific circumstances 
which would clearly be outside the control of a manufacturer leading to 
the inability to comply in 1998 MY. Finally, EPA requests comment on 
any other alternatives.

C. Small Manufacturer Criteria

    In their comments to the NPRM, small marine manufacturers provided 
new information to EPA. This information has heightened EPA's awareness 
that the proposal would impact different segments of the marine market 
in different ways.
    One example is the engine marinizer: rather than manufacturing the 
engines themselves, smaller marine engine companies (i.e., 
``marinizers'') will often modify engine blocks originally produced for 
other nonroad or on-highway applications for marine applications. 
Marinizers may have as few as two employees and typically do not have 
the resources to comply with certification and enforcement provisions 
proposed in the NPRM. The cost of one test for one engine can be 
$5,000-$10,000 at a contract laboratory, while the cost of test 
equipment can run up to one half million dollars. Many marinizers have 
indicated that the cost of certification reporting burdens alone will 
dramatically increase their costs and may force them out of business.
    Based on the written comments to the NPRM described above, EPA 
proposes to allow some flexibility in the certification and enforcement 
provisions proposed in the NPRM for small manufacturers. The proposed 
flexibility is described in IV.B. above. In the following sections, EPA 
proposes small manufacturer criteria for the marine CI engine and SD/I 
engine market segments. Each market segment has a different, unique 
aspect from a regulatory and market structure perspective, as explained 
below. Therefore, different small manufacturer criteria are proposed 
for each market segment.
1. Sterndrive and Inboard Engine Manufacturers
    Although EPA is proposing an emission standard for SD/I engine 
families that will only necessitate a marginal compliance burden to 
certify the engine family and no other administrative program burdens, 
EPA is proposing a small manufacturer criteria in the event that the 
rulemaking should be finalized with the need for such a criteria.
    The market for SD/I engines is composed of one very large market 
leader, several medium-sized market players, and a number of very small 
(in terms of both receipts and production volumes) manufacturers. For 
the 

[[Page 4609]]
purposes of the SD/I engine category only, EPA proposes to define small 
manufacturers as those which have less than 15 percent of the United 
States market share of SD/I engines on a unit volume basis, to be 
determined by averaging engine unit volume for the past three model 
years. The average total SD/I market unit volume for the same model 
years would be used to determine whether a manufacturer's market share 
was less than 15 percent. Manufacturers with greater than 15% market 
share are clearly the largest manufacturers.
    The Agency requests comment on this market share percentage 
criterion and on alternatives for defining a small SD/I manufacturer.
2. Marine Compression-Ignition Engine Manufacturers
    The Agency proposes that a small marine CI engine manufacturer be 
defined as one for which the business concern together with all its 
domestic and foreign affiliates (e.g., the parent company and all the 
subsidiaries): (1) Have total annual receipts under $100 million, and 
(2) have less than a 4 percent United States market share on a unit 
volume basis for all nonroad diesel engines. For example, 4 percent 
market share is approximately 12,000 units based on a total volume of 
all nonroad diesel engines of 300,000. According to the proposed 
criteria, the average annual receipts per engine from 12,000 units 
could not exceed approximately $8,300 per engine ($100 million/12,000 
engines).
    The Agency proposes to accept the definitions of ``affiliation,'' 
``annual receipts,'' and ``business concern'' that are contained in 13 
CFR Part 121 of the Small Business Administration (SBA) regulations. 
Compliance with the market share criterion will be determined on the 
basis of data averaged over the past three fiscal years, in a manner 
similar to that defined in the SBA regulations for annual receipts.
    The definition proposed herein would give flexibility in the rule 
for manufacturers of different sizes of marine CI engines and different 
production volumes. If a manufacturer meets the definition's criteria, 
EPA proposes that its engine families be eligible automatically for the 
certification program flexibility described in section IV.B.2. above.
    The Agency requests comment on the advisability of two additional 
small entity criteria that it has not proposed. The first of these 
would be to limit small manufacturer flexibility to engine families 
under 1500 kW only. This is because engines over 1500 kW are expensive, 
and therefore, certification and enforcement costs have a small effect 
on engine price and should be easily recovered. The second would be to 
limit small manufacturer flexibility using an engine speed designation 
for high-speed marine CI engines, instead of a maximum power criterion. 
The Agency requests comment on designating high-speed engines as those 
over 1000 revolutions per minute (rpm). Comment submitted in response 
to the NPRM on the issue of harmonization with the proposed emission 
standards by the International Maritime Organization included a 
recommendation on a 1500 kW cutpoint for EPA's proposal.
    The Agency also requests comment regarding specific alternative 
criteria for designating small manufacturers and on equity issues 
associated with the proposed criteria. In addition, EPA is considering 
whether to propose applying the small marine CI engine manufacturer 
definition to all aspects of the CI engine industry. EPA is not aware 
of any nonmarine CI engine manufacturers that meet this criteria. 
However, if any exist or enter the market, it seems appropriate that 
the same provisions apply. EPA welcomes comments on this issue.
3. Outboard Engine and Personal Watercraft Manufacturers
    a. Competitive Issues. Manufacturers of outboards or personal 
watercraft that commented on the proposal appear to fall into at least 
one of two categories: (1) Those which do not meet the SBA's 
definitions of ``small'' and (2) manufacturers that purchase engines 
and market them as their own, rather than being actual engine 
manufacturers. Thus, EPA is not aware of any manufacturers of OB/PWC 
engines that it believes would need compliance flexibility as small 
volume manufacturers. Moreover, this category of marine engines 
produces the highest HC emissions per unit power output, and is 
therefore the category of marine engines targeted for the largest HC 
reductions. The Agency is hesitant to offer a permanent waiver of more 
stringent testing requirements for the engines of most concern to it. 
EPA wants to be certain that manufacturers are developing, producing, 
and achieving the targeted HC emission reductions for OB/PWC. For these 
reasons, EPA does not propose to offer small entity regulatory relief 
to manufacturers of OB/PWC engines.
    However, while in the absolute sense there are no ``small'' 
manufacturers, in the relative sense there are smaller manufacturers 
relative to the larger manufacturers. Furthermore, there are 
differences in the product lines of the manufacturers. Some 
manufacturers are dominant in the personal watercraft market while at 
the same time being less dominant in the outboard market (e.g., 
Yamaha). Most of the PWC manufacturers specialize only in personal 
watercraft and do not produce outboards.
    EPA proposed in the NPRM a combined averaging set for outboards and 
personal watercraft, even though there are differences in product lines 
between manufacturers with some producing both types or only one type. 
EPA thinks this strategy best for many reasons. First, this strategy 
allows manufacturers to take advantage of the most cost-effective means 
of achieving emission reduction targets amongst engines with similar 
emission problems. Both outboards and personal watercraft currently 
utilize old technology 2-stroke engines and have similar options 
available to reduce those emissions. Second, achieving the most cost-
effective emission reductions means that the market achieves the lowest 
price increase to the consumer. Third, EPA is not interested in 
protecting manufacturer market share at the expense of higher consumer 
prices for control technology. EPA thinks that broader averaging sets 
encourage a more competitive market environment which in turn limits 
non-competitive (e.g., oligopolistic) market forces and acts to keep 
consumer prices low. Fourth, a combined OB/PWC averaging set gives more 
flexibility to manufacturers, particularly the smaller PWC 
manufacturers, to buy credits from other manufacturers (including those 
they do not directly compete with) instead of putting on control 
technology that is not cost-effective. Therefore, in effect, a combined 
OB/PWC set inherently improves small manufacturer flexibility. For 
these reasons, EPA is very hesitant to consider splitting up the 
combined OB/PWC averaging set.
    In response to the NPRM, EPA received significant comment from some 
manufacturers that only produces PWC indicating concern with the 
appropriateness of a combined OB/PWC averaging set.\14\ These 
manufacturers seemed to be concerned that manufacturers that produce 
both outboards and personal watercraft (only one such manufacturer 
currently exists) can take competitive advantage of their ability to 
average their OB engine 

[[Page 4610]]
families with their PWC engine families. The PWC-only manufacturer 
seemed to be concerned that their competitor has more flexibility to 
meet the emission standards due to the possibilities of generating 
internal to the company positive credits from the OB product line that 
can potentially delay control on PWC or provide cheaper credits to 
cover lesser control on PWC. One PWC-only manufacturer stated their 
belief that the other manufacturer will convert its OB products into 4-
stroke and that will harm the competitive position of manufacturers who 
only produce PWC and market share will be eroded. There is concern that 
the end result will be that the OB/PWC manufacturer will become a much 
more dominant manufacturer. It is feared by the commenter that no 
credits would be available in the market that would allow the same 
flexibility for the PWC manufacturers that this one manufacturer will 
inherently have under the combined OB/PWC set.

    \14\ Refer to the Kawasaki docket comments IV-D-58, statement or 
Artco, Kawasaki, and Polaris IV-D-66 and Polaris statement IV-51 for 
EPA Air Docket A-92-28.
---------------------------------------------------------------------------

    EPA would be concerned if a single manufacturer gained control over 
the PWC market simply because of the combined OB/PWC averaging set.
    On the other hand, EPA is concerned that splitting the averaging 
sets will give significant competitive advantage to the currenta 
dominant PWC manufacturers, particularly against the PWC manufacturers 
with smaller market share. The marginal cost-effectiveness analysis by 
which EPA set the 75% reduction in HC requirement allows small engine 
families, such as those produced by PWC manufacturers with smaller 
market share, to avoid manufacturing changes that are not cost-
effective in a relative sense and purchase sufficient credits in the 
market. This is because the marginal cost-effectiveness for each engine 
family was ranked and the standard was set at the point where it became 
less cost-effective to gain further emission reductions. Splitting the 
averaging set restricts the potential credit supply and the result may 
be that the most cost-effective credits are not available. Therefore, 
even if it is more marginally cost-effective to achieve emission 
reductions from OB, for example, restricting the averaging between OB 
and PWC means that some of the most cost-effective reductions may not 
be taken advantage of. Further, the smaller PWC manufacturers who would 
most need to buy credits would be restricted to purchasing credits from 
their direct competitors, instead of the OB manufacturers they do not 
directly compete with. The effect would be that the dominant PWC 
manufacturers would be able to gain competitive advantage because they 
produce larger unit volumes and can take advantage of economies of 
scale, thereby generating positive credits in a more cost-effective 
manner than lower volume manufacturers. Thus, protecting the market 
share of current dominant manufacturers by splitting the averaging set 
may have the effect of strengthening the dominant manufacturers' market 
positions against the smaller PWC manufacturers. EPA is seeking an 
emission standard structure that promotes a competitive market and 
promotes the cleanest technology. EPA thinks it essential to allow all 
smaller manufacturers the flexibility intended with the combined OB/PWC 
averaging set.
    Comments are requested on separating the averaging sets as an 
approach to address the specific problem raised with respect to the 
competitive impact of a combined OB/PWC averaging set. EPA requests 
comment on requiring separate averaging sets for a short while during 
the phase-in period or a portion of it. From an environmental point of 
view, this will likely ensure that the manufacturer who produces both 
OB and PWC invest in control technology for PWCs in the early years of 
the phase-in. EPA is hesitant to consider this option and would only 
consider it for a short while (e.g., a portion of the phase-in).
    EPA requests comment on why EPA should consider requiring separate 
sets, even for a portion of the phase-in period, if this limits the 
ability of the market to generate the most cost-effective controls 
overall. Further, EPA requests comment on the docket comments submitted 
by Kawasaki (IV-D-58 for EPA Air Docket A-92-28).
    Also, EPA notes that it would have to re-evaluate the 
appropriateness of the proposed OB/PWC emission standards if the 
averaging sets were to change, because of the potential effect of the 
separate sets on such factors as technological achievability and cost 
(see section 213(a)(3) of the CAA). The Agency seeks comment on what 
changes, if any, should be made to the proposed emission standards if 
separate averaging sets are finalized, with an explanation of the 
reasons for the commenter's preferred approach.
    Additionally, EPA requests comments on the need for any change from 
the combined OB/PWC averaging set that EPA proposed. Change is 
questionable since already manufacturers may purchase credits in the 
market rather than apply control technology in the early years of the 
phase-in thereby giving a manufacturer extra leadtime for whatever 
reason. Further, if the credit market is economically efficient (i.e., 
a manufacturer does not act in a predatory manner to gain market share) 
then the OB manufacturer would make positive credits available to PWC 
manufacturers because this would lower the cost of OB compliance, 
either raising profit margin or increasing sales, or both.
    EPA requests comment on systems that would encourage the credit 
market to function efficiently. Any comments that present ways to make 
the market function prospectively are especially encouraged. Comment is 
requested on the need to have a formalized credit market. EPA would 
prefer that such a market system not be run by EPA and requests 
comments on making this market run by an independent third party if a 
formalized market is advocated. EPA is seeking comment in order to 
determine whether it should propose action through a subsequent 
rulemaking.
    b. Market Entrants. The Agency requests comment on the issue of 
flexibility for small OB/PWC engine manufacturers that may enter the 
market in the future. EPA would be concerned should administrative 
program burdens add an additional production cost that discourages 
market entrants and limits additional competition in the marketplace, 
particularly for clean technology. The Agency would consider allowing 
administrative program flexibility for a short period of time for new 
OB/PWC market entrants similar to that which it is proposing for small 
marine CI engine manufacturers, such as reduced certification 
requirements, as described above in section IV.B. As EPA would like to 
encourage clean technology, such market entrant flexibility would only 
be considered for engines with emissions falling below the MY 2006 
average HC emission standard level.
    The Agency requests comment on defining a market entrant as a 
manufacturer that has not produced OB/PWC engines before one MY prior 
to the current MY. This suggested definition would allow flexibility 
for market entrants for the first two model years. In addition, EPA 
requests comment on whether flexibility should be limited to small 
market entrants and, if so, requests suggestions for alternative 
definitions of a small market entrant.

D. Relative Use by Age Function

    The Agency proposes to include a statistical function in the credit 
calculation formula in Sec. 91.207 of the regulations proposed for 40 
CFR Part 91, representing relative usage of engines by engine age and 
power output. EPA did 

[[Page 4611]]
not propose the use of such a function in the NPRM for the generation 
of new engine family credits. However, EPA is inclined to believe that 
for OB engines usage does vary by age of the engine and by power 
output. The relationship between age of engine and relative usage was 
assumed to be linear according to the following function, which is 
based on an assumption of 30% deviation (i.e., 1.3 and 0.7).

where

t=age of the engine in years
use=mean use in hours per year
life=mean life

    The average annual use derived for the new engine credit generation 
methodology proposed in the NPRM still appears to be an adequate 
representation.
    For outboard engines, the probability that an engine will survive 
into the future depends upon the power output (in terms of rated kW or 
rated horsepower) of the engine. Smaller engines typically last longer 
than larger engines.15 Therefore, the relative use by age function 
uses mean life as in input. In turn, the mean life is dependent upon 
power output. Power output identifies the size of the engine.

    \15\ Price Waterhouse, National Recreational Boating Survey: 
Final Report, June 30, 1992.
---------------------------------------------------------------------------

    The Agency is aware that the State of Wisconsin performed a survey 
of the 1995 summer season to obtain better information on relative use 
of engines by age. If the Wisconsin data becomes available before the 
final rule is promulgated, EPA will publish a notice of data 
availability regarding the survey results. EPA may consider the survey 
results when deciding how to finalize the rule with respect to the 
relative use by age function.

E. Manufacturer Production Line Testing Program

    The NPRM described a proposal for marine SI and CI engine 
manufacturers to perform self-audits of new marine engines. The 
proposed self-audit program would be an emissions compliance program 
for new production marine engines in which manufacturers would be 
required to test engines as they leave the production line, without EPA 
oversight.
    The Agency believes that a post-production compliance program may 
be necessary for OB/PWC only to verify that production engines comply 
with the applicable family emission limit (FEL), particularly during 
the early years of the program. The NPRM noted that the need for such a 
program is particularly vital in a regulatory situation in which 
manufacturers participate in an averaging, banking, and trading program 
and receive usable or salable credits for declaring FELs more stringent 
than the emission standard. The NPRM proposed a self-audit program 
comparable to the California Air Resources Board's (CARB's) current 
Quality Audit Program for new utility and lawn and garden engines. As 
the NPRM described, this program would assure that engines from each 
engine family will be tested periodically and their compliance 
evaluated on a quarterly basis.
    In this supplemental notice EPA proposes to modify the self-audit 
program set forth in part 91 subpart F of the NPRM's proposed 
regulatory text. First, EPA is proposing to change the name of the 
proposed Manufacturer Self-Audit Program to the Manufacturer Production 
Line Testing Program, because this title more clearly indicates that 
this proposal is applicable for emission testing engines from the 
manufacturer's production line. Second, EPA proposes to limit the 
production line testing program provisions to SI OB/PWC engines. As 
described in more detail in section IV.B.2. above, EPA now proposes 
that all marine CI engines be subject to the Selective Enforcement 
Auditing and recall provisions that have been promulgated for land-
based CI engines. Third, EPA proposes to adopt the Cumulative Sum 
(CumSum) procedure described below, rather than CARB's Quality Audit 
Program procedure, because EPA has noticed a potential problem with the 
provisions of subpart F as proposed. CARB's Quality Audit Program is 
based on a fixed sample size approach. An essential problem with this 
approach is that to keep the sample size small, the manufacturer risk 
and the consumer risk must increase 16. The only way to lower 
manufacturer and consumer risk is to increase the sample size to 
possibly burdensome levels. This results in an inherent conflict for 
the design of a quality audit procedure which requires a fixed sample 
size.

    \16\ Manufacturer risk is the risk that the quality audit 
program will detect that an engine family is in noncompliance, when 
the family is actually in compliance. Consumer risk is the risk that 
the quality audit program will fail to detect that an engine family 
is in noncompliance, when the family is actually in noncompliance.
---------------------------------------------------------------------------

    The annual sample size required by CARB's Quality Audit Program is 
set at one percent of engine family production, at least until ten 
engines are tested in an engine family. A major effort by both CARB and 
the affected manufacturers has been to find ways to reduce the 
necessary sample size, resulting in a confusing array of statistically 
ad hoc modifications to the program. Upon recognizing the limitations 
of CARB's Quality Audit Program as a model for the NPRM's marine engine 
self-audit program, EPA initiated development of another approach.
    In today's SNPRM, EPA is proposing to modify the proposed subpart F 
regulations to include a statistical procedure known as the CumSum 
procedure that will enable manufacturers to select engines at 
appropriate sampling rates for emission testing and will determine 
whether production line engines are complying with emission standards. 
CumSum procedures are used for the detection of changes in the average 
level of a process; the proposed procedure is useful both as an 
assessment tool for EPA and a quality control tool for engine 
manufacturers. 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.
    Under the procedure, described in more detail below, manufacturers 
would select engines from each engine family at appropriate sampling 
rates for emissions testing. Testing would be required to be conducted 
in accordance with the applicable federal testing procedures for marine 
engines. The test results would be input to the appropriate CumSum 
equations, and the results of the procedure would indicate whether the 
engine family is in noncompliance.
1. Sampling Rates Required for the CumSum Procedure
    Sample Size Calculation. At the start of each MY, manufacturers 
would begin to test each engine family at a rate of one percent, and 
then modify the testing rate according to a sample size equation. A 
manufacturer would determine the sample size necessary for newly-
certified engine families by conducting two tests and then calculating 
the required sample size for the rest of the MY according to the Sample 
Size Equation below. For carry-over engine families, 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 
MY, and finally calculating the required sample size for the rest of 
the MY according to the Sample Size Equation below.

Sample Size Equation

where:


[[Page 4612]]

N=Calculated sample size. NHC, NCO, and NNOX are all 
calculated from each test result. The largest of the three becomes the 
official N which becomes the number of tests required for the remainder 
of the MY. NHC, NCO, and NNOX are all recalculated after 
each test.
t95=95 percent confidence coefficient. It is dependent on the 
actual sample size, n, and is defined in the table below. It defines 
one-tail, 95 percent confidence intervals.

            Sample Size and One-Tail Confidence Coefficients            
------------------------------------------------------------------------
     n           t95           n          t95          n          t95   
------------------------------------------------------------------------
 2.........       6.3           12         1.8          22        1.72  
                    1                        0                          
3..........       2.9           13         1.7          23        1.72  
                    2                        8                          
4..........       2.3           14         1.7          24        1.71  
                    5                        7                          
5..........       2.1           15         1.7          25        1.71  
                    3                        6                          
6..........       2.0           16         1.7          26        1.71  
                    2                        5                          
7..........       1.9           17         1.7          27        1.71  
                    4                        5                          
8..........       1.9           18         1.7          28        1.70  
                    0                        4                          
9..........       1.8           19         1.7          29        1.70  
                    6                        3                          
1..........       1.8           20         1.7          30        1.70  
0                   3                        3                          
1..........       1.8           21         1.7         1.645  
1                   1                        2                          
------------------------------------------------------------------------

=sample standard deviation of the actual sample, where:
Xi=emission test result for an individual engine
x=mean of the actual sample
STD=emission standard or, if applicable, family emission limit (FEL)
n=The actual number of tests completed in an engine family
    The calculated sample size, N, determines the number of tests 
required for the rest of the MY. Tests must be distributed evenly 
throughout the remainder of the MY. After each new test, the sample 
size is recalculated with the updated sample mean, sample standard 
deviation, and 95 percent confidence coefficient.
    If at any time throughout the MY the calculated sample size for an 
engine family, N, is less than or equal to the actual sample size, n, 
and the sample mean, x, for each pollutant, is less than or equal to 
the applicable standard or FEL, the manufacturer may stop testing that 
engine family. But, if at any time throughout the MY the sample mean, 
x, for any pollutant, is greater than the applicable standard or FEL, 
the manufacturer must continue testing that engine family at the 
appropriate maximum sampling rate.
    Manufacturers may elect to test additional engines for input into 
the Sample Size Equation, provided that testing of the additional 
engines is performed in accordance with the applicable federal testing 
procedures for marine engines.
    Maximum Sample Rates. The maximum required sample size for an 
engine family (regardless of the result of the Sample Size Equation) is 
the lesser of three tests per month or one percent of projected annual 
production (distributed evenly throughout the model year). For example, 
if the Sample Size Equation produces a value of N = 252 for a family 
with annual production of 20,000 engines, a manufacturer may elect to 
test only three engines per month instead of:
    (1) 21 per month, which would be required if 252 tests were 
distributed evenly throughout the MY, or
    (2) 17 per month, if one percent of annual production were 
distributed evenly throughout the MY.
    Although the Sample Size Equation may calculate sample sizes 
greater than the proposed maximum sample rates, EPA believes sample 
sizes greater than these maximum rates would be unnecessarily 
burdensome for manufacturers of marine engines. The proposed maximum 
sample rates adequately characterize the emission levels of the engine 
family.
2. Construction of the CumSum Equation
    After determining the appropriate sample size using the Sample Size 
Equation, the manufacturer would construct the following CumSum 
Equation for each regulated pollutant for each engine family:

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
STD=The applicable standard or, if applicable, the FEL
F=0.25  x   and is the reference value

    After each test, Ci is compared to the action limit, H.

H=5.0  x   and is the action limit, the quantity which the 
CumSum statistic must exceed, in two consecutive tests, before the 
engine family is determined to be in noncompliance. (it is a function 
of the standard deviation, )
=is the sample standard deviation and is recalculated after 
each test.

    Following each emission test, manufacturers would update current 
CumSum statistics for each pollutant according to the CumSum Equation 
described above. Manufacturers would continue to update the CumSum 
statistics throughout the MY. (At no time throughout the MY are CumSum 
statistics reset to zero.)
    Manufacturers may elect to test additional engines for input into 
the CumSum Equation, provided that testing of the additional engines is 
performed in accordance with the applicable federal testing procedures 
for marine engines.
3. Criteria for Determining Noncompliance
    An engine family is determined to be in noncompliance if at any 
time throughout the MY, the CumSum statistic, Ci, exceeds the 
applicable action limit in two consecutive tests for the same 
pollutant.
    Production line emission test results, as well as sample size 
calculations and CumSum calculations, would be electronically 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 a manufacturer were to determine 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 such a determination.
    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, 
EPA may suspend or revoke the manufacturer's certificate of conformity 
in whole or in part for that engine family. The suspension or 
revocation will not occur before fifteen days after a noncompliance 
determination is made. During this fifteen day period, EPA will 
coordinate with the manufacturer to facilitate the approval of the 
required production line remedy in order to eliminate the need to halt 
production, if possible. The manufacturer must then address the engines 
produced prior to the suspension or revocation of the certificate of 
conformity. EPA may reinstate a certificate of conformity subsequent to 
a suspension, or reissue one subsequent to a revocation, after the 
manufacturer demonstrates that improvements, modifications, or 

[[Page 4613]]
replacement 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.
    The Manufacturer Production Line Testing Program would be the main 
production line emission test program for marine engines. The Selective 
Enforcement Auditing (SEA) 17 program that was proposed in the 
NPRM will serve a spot-check function and enable EPA to evaluate 
testing practices used by the manufacturer, follow up on concerns 
reported to EPA, and address any configurations not covered by 
manufacturers in their production line testing program.

    \17\ SEA is a program in which EPA selects engines from one 
engine family configuration, directly from the production line, for 
emissions testing.
---------------------------------------------------------------------------

    EPA realizes that the standard deviation, , of an engine 
family is an important aspect of the production line testing program. 
EPA intends to employ accurate engine family standard deviation in the 
CumSum Equations. The Agency requests comment on all aspects of the 
proposed production line testing program and specifically the 
appropriateness of the values chosen for the variables in the Sample 
Size and CumSum Equations. For more information on the derivation of 
the Sample Size and CumSum Equations, the selection of appropriate 
variables, and some examples of the CumSum Procedure, see ``The 
Cumulative Sum Procedure'' document in the docket.
4. Changes in FELs and Other Running Changes
    During the course of a MY, manufacturers may change certification 
FELs up or down depending on comfort level or engineering decisions. 
Manufacturers may also make changes to the engines to increase 
performance or reduce emissions. The Agency proposes to handle these 
changes in production with respect to the CumSum procedure as described 
below.
    Changing an FEL (Actual Engine Not Changed). All data accumulated 
during that MY but prior to the FEL change would be recalculated with 
the new FEL. New sample sizes would be calculated, and testing would be 
continued or halted as required. The CumSum statistic would also be 
recalculated with the new FEL and would be evaluated with respect to a 
new action limit. Testing and updating of the sample size and CumSum 
statistic would continue until testing could be halted as a result of 
the sample size calculation, a noncompliance decision, or the end of 
the MY.
    Changing an FEL (Actual Engine Changed). All data accumulated 
during that MY but prior to the FEL/engine change would be left as is. 
Sample sizes would now be calculated by inserting the new FEL into the 
Sample Size Equation. The CumSum Equation and action limit would be 
updated to reflect the new FEL. The CumSum statistic would then be 
calculated by the new equation and would be evaluated with respect to 
the new action limit. Testing and updating of the sample size and 
CumSum statistic would continue until testing could be halted as a 
result of the sample size calculation, a noncompliance decision, or the 
end of the MY.
    No Change to an FEL (Actual Engine Changed). No changes would be 
made to any of the equations or any of the accumulated data. This type 
of action is considered a typical day-to-day change on the production 
line that should be evaluated by the Sample Size and CumSum Equations.
5. Old Technology Engines
    EPA proposes to waive production line testing requirements for any 
old technology OB/PWC engine family through MY 2003. In MY 2004 and MY 
2005, any manufacturer of an old technology OB/PWC engine family may 
request, in writing, an exemption from the requirements to perform 
production line testing. EPA will have the discretion to grant a waiver 
if the Administrator determines that the engine family will be phased 
out of production by MY 2005. EPA will review requests for exemptions 
and upon granting appropriate requests will prepare and submit to the 
manufacturer a memorandum of exemption, which will set forth the terms 
and conditions of the exemption. The Agency requests comment on the 
appropriateness of exempting old technology engine families being 
phased out within six years of the effective date of the rulemaking, 
and of offering a discretionary waiver to such engine families in MY 
2004 and MY 2005. Refer to section IV.B.3. for additional discussion of 
this flexibility and the definition of ``old technology OB/PWC.''
6. Effective Date of the CumSum Procedure
    Since publishing the NPRM, it has come to EPA's attention that OB/
PWC engine manufacturers may need significant time to prepare their 
production facilities with all of the necessary equipment and resources 
to comply with the production line testing requirements. EPA is 
proposing that the requirements for the production line testing program 
become effective one year later than proposed in the NPRM. Under this 
proposal, marine engine emission standards, certification requirements, 
and in-use testing provisions would still go into effect beginning with 
MY 1998. Production line testing requirements would go into effect 
beginning in MY 1999. This proposal offers some relief to manufacturers 
while making sure that emission standards and in-use compliance are not 
delayed. Manufacturers could voluntarily submit production line testing 
data to EPA during MY 1998. The Agency requests comment on the 
appropriateness of this one-year delay in the imposition of production 
line testing requirements.
7. Request for Comment
    Although EPA is proposing modifications to the proposed Production 
Line Testing Program to greatly reduce its burden as outlined above, 
EPA also requests comment on the appropriateness of omitting such a 
program from the final rule. EPA believes that the Production Line 
Testing Program may be the best testing activity which can detect 
whether a manufacturer has failed to translate an engine design 
successfully into mass production while the manufacturer still is 
producing that design.
    This Program has the ability to catch and offer a manufacturer the 
opportunity to correct emission related problems early in an engine's 
life, thus reducing a manufacturer's in-use liability. EPA believes 
that the proposed Production Line Testing Program would also serve the 
following additional purposes: (1) ensure that manufacturers follow 
precisely the emissions test procedures listed in the CFR, (2) ensure 
that the manufacturers' test equipment accurately measure emissions, 
and (3) ensure that production engines are in conformity with 
applicable Federal emission requirements as they come off the assembly 
line and that individual engines tested conform to applicable family 
emission limits.
    EPA believes that production line testing is especially important 
for a rule where certification is built around an averaging, banking, 
and trading program. Manufacturers will be producing engines which 
generate emission credits that can be bought or sold or used to offset 
other families produced by the same manufacturer. EPA believes it is 
important to ascertain that actual production engines achieve proper 
certification family emission 

[[Page 4614]]
limits to ensure that credits are bona fide and real.
    However, EPA is considering whether the information obtained from 
this program is redundant with the information obtained from the 
proposed In-Use Testing Program. The government is generally attempting 
to reduce regulatory burden by eliminating all programs that generate 
redundant information and information that is not cost-effective. EPA 
requests comment on the relative importance of the information gleaned 
from the Production Line Testing Program and compliance measures 
associated with the In-Use Testing and In-use Credit Programs. EPA is 
considering the option of not finalizing the Production Line Testing 
Program provided that the In-Use Testing and In-Use Credit Programs are 
finalized. However, because EPA thinks production line testing 
generates relevant data and is important, EPA also requests comment on 
other options such as having production line testing in the early years 
of the program and then relaxing or eliminating production line testing 
as the in-use program generates more data.
    Should EPA opt not to finalize a Production Line Testing Program, 
EPA requests comment as to whether SEA should become a more important 
programmatic emphasis. EPA requests comment on whether SEA regulations 
(i.e., Part 91, Subpart G) should be finalized for OB/PWC if commenters 
do not think SEA should become a more important programmatic emphasis. 
Even if the SEA regulations proposed in the NPRM were not finalized, 
EPA would retain authority under Sections 206(b)(1) and 208 of the CAA 
to test or require testing of newly manufactured engines and to inspect 
production facilities and processes to determine whether a manufacturer 
is complying with the information submitted for certification. Further, 
EPA would retain authority under Section 206(b)(2) of the CAA to 
suspend or revoke the certificate for engines that do not conform with 
applicable emission standards. However, without SEA regulations, the 
SEA process could become more cumbersome. EPA seeks comments on both 
the advantages and disadvantages of finalizing Subpart G, Part 91.

F. In-Use Credit Program

    The Agency is proposing an in-use credit program for marine OB/PWC 
engines. This program would not be a substitute for the proposed 
averaging, banking, and trading (ABT) provisions used for certification 
purposes, but would be offered as a separate program that may be used 
in conjunction with the certification ABT provisions. The in-use credit 
program is designed to reduce cost without reducing environmental 
benefits by providing manufacturers with flexibility in meeting the 
proposed standards for each pollutant in-use. Participation in this 
proposed program would be voluntary.
    The flexibility that EPA proposes to provide in the in-use credit 
program is necessary for a number of reasons. In the event that engine 
families fail in-use testing, EPA believes that recalling the 
nonconforming engines would be particularly burdensome and impractical 
for this industry, mainly due to the difficulty of tracking the 
nonconforming engines. If registration with a government entity occurs, 
it is the vessel that is registered, not the vessel's engine; 
manufacturers of marine engines do not typically know in what vessels 
their engines are installed. Tracking the engines would thus be 
cumbersome and difficult, especially because manufacturers estimate 
that the owner moves or the vessel is typically sold about four years 
after the initial purchase. Therefore, recalling the engines would 
likely require substantial resources, yet not be highly effective in 
actually remedying the excess emissions.
    The Agency believes it has the authority to promulgate this in-use 
credit program under the circumstances. The CAA provides that the 
marine engine emission standards, when finalized, shall be subject to 
Section 207 of the Act, ``with such modifications of the applicable 
regulations * * * as the Administrator deems appropriate.'' 42 U.S.C. 
7547(d). Section 213 requires engines to comply with emission standards 
when in actual use throughout their regulatory useful lives, and 
Section 207 requires a manufacturer to remedy in-use nonconformity when 
EPA determines that a substantial number of properly maintained and 
used engines fail to conform with the applicable emission standards. 42 
U.S.C. 7541. Once EPA makes this determination, recall would be 
necessary to remedy the nonconformity. However, EPA believes that, 
under the circumstances here, where it has been proposed that OB/PWC 
marine engines could use ABT to comply with the emission standards at 
certification (see 59 FR 55930), it is appropriate not to make a 
determination of substantial nonconformity where a manufacturer uses 
ABT to offset in-use noncompliance. Doing so is also appropriate 
because it is expected that recall would be impractical and largely 
ineffective. Thus, the CAA offers EPA the discretion to not make a 
Section 207(c) determination of substantial nonconformity where a 
marine engine manufacturer uses ABT to offset any noncompliance with 
the statute's in-use performance requirements. Though the language of 
Section 213(d) is silent on the issue of averaging, it allows EPA 
considerable discretion in determining what modifications to the on-
highway regulatory scheme are appropriate for nonroad engines.
    In this current proposal, in-use credits would be based upon in-use 
testing conducted by the manufacturer as discussed previously in the 
NPRM. For a given engine family, the in-use compliance level (CL) would 
be determined by averaging the results from in-use testing performed 
for that engine family. If the in-use CL is below the applicable FEL to 
which the engine family is certified, the manufacturer could generate 
in-use credits for that engine family. If the in-use CL is above the 
applicable FEL, the engine family would experience a credit deficit. In 
any given year, a manufacturer may use in-use credits to average 
against excess in-use emissions of another engine family from the same 
MY, to bank for use in future model years, or to trade to other 
manufacturers. If a manufacturer completes testing for a given MY and 
is in a deficit situation, it will not be allowed to carry the deficit 
over to the next MY. To remedy a deficit situation, a manufacturer 
could purchase credits from another manufacturer or, upon EPA approval, 
test additional engine families of that MY beyond the 25 percent 
proposed in the NPRM for the in-use testing program to generate 
additional credits.18

    \18\  However, if the additional testing discovers an engine 
family that was in noncompliance with its FEL, the result would be 
handled as if it were a failure of the mandated in-use testing 
requirement of up to 25 percent of a manufacturer's engine families.
---------------------------------------------------------------------------

    However, EPA is considering allowing a manufacturer to carry-over a 
deficit to the next MY in the beginning of the phase-in period. 
Specifically, EPA is considering allowing carry-over during the first 
three years of the phase-in if no credits are available for purchase to 
remedy the deficit. EPA requests comment on the appropriateness of 
allowing a deficit carry-over, on whether it should allow this carry-
over only when no credits are available for purchase or if other 
circumstances are appropriate for carry-over, and on whether the first 
three years of the phase-in period or some other time period is an 
appropriate time period for such a deficit carry-over.
    The Agency is designing the in-use credit program around three 
principles. 

[[Page 4615]]
First, the in-use testing program will assess whether each manufacturer 
is achieving the environmental benefits intended by the standards when 
the engines are in-use. Second, manufacturers will be provided with 
strong incentive to maintain the standards in-use which will further 
encourage in-use compliance. Finally, the in-use credit program will 
provide flexibility and reduce the burden on manufacturers by allowing 
them an option to address in-use noncompliance in a way that EPA agrees 
would avoid a determination of nonconformity under Sec. 207(c) of the 
Act, and thereby avoid a recall.
    Credits associated with the certification ABT program would not be 
interchangeable with credits generated or used in the in-use credit 
program. Positive certification credits are generated when the FEL is 
set below the applicable standard. An in-use nonconformity occurs when 
the CL, which is the emission level determined by in-use testing for an 
engine family, is found to be above the FEL. Allowing a manufacturer to 
remedy an in-use nonconformity with positive certification credits 
generated by the same or another FEL setting would be a dubious policy. 
Such a policy does not appear to encourage manufacturers to make 
adequate effort to declare FELs during the certification process that 
predict in-use emission levels to the fullest extent possible. The 
Agency is concerned that if the in-use test results simply updated the 
certification FEL then manufacturers would attempt to set certification 
FELs that the engine would likely exceed in-use, because a manufacturer 
would have a chance after in-use testing to change the FEL if it had 
been set too low at certification. In this way, the manufacturer would 
generate more certification credits than the engine family actually 
should receive and would have already have used those credits to offset 
dirtier engines. This is referred to as ``gaming'' the ABT provisions 
by ``shaving'' the FELs. Therefore, to preserve the integrity of both 
the certification and in-use ABT programs and maintain accountability 
for manufacturers to meet their stated FELs in certification, 
production line, and in-use testing, EPA is proposing to restrict 
credit use by separating in-use credits from certification credits. The 
Agency requests comment on the necessity of separate certification and 
in-use ABT sets, especially with respect to providing the incentive for 
manufacturers to produce engines that meet designed emission levels in-
use and to choose a certification FEL which represents in-use emission 
levels.
    An engine family's in-use CL would be determined by averaging the 
results of testing in-use engines, as discussed in the NPRM. The test 
results would be rounded to the number of decimal places contained in 
the applicable emission standard or FEL, expressed to one additional 
significant figure. Rounding would be done in accordance with ASTM 29-
90, ``Standard Practice for Using Significant Digits in Test Data to 
Determine Conformance with Specifications.'' The CL would be compared 
to the applicable FEL to determine if the engine family will generate 
or require credits.
    In EPA's experience with the on-highway heavy-duty ABT 
program,19 manufacturers have changed FELs during the MY through 
running changes. The Agency would not want to restrict the ability of 
manufacturers to lower FELs when installing cleaner technology during 
the MY, or to raise FELs if emission data is obtained on an engine 
family indicating an inadequate safety margin. However, EPA also 
believes that an in-use testing program that tests only one declared 
FEL during a MY would not be representative of an engine family having 
multiple FELs throughout that MY. In a case where a manufacturer has 
changed the FEL of an engine family during the MY because of a design 
change, EPA may designate which FEL of the engine family is to be 
tested. If an in-use failure occurs for this FEL designation, EPA may 
request in-use emission results from other FEL designations within this 
engine family to ensure that the noncompliance is confined only to that 
portion of the engine family with the audited FEL. A manufacturer can 
also change their certification FEL to increase its safety margin or 
generate extra credits without a design change. In such cases, if a 
chosen FEL of an engine family were to fail in-use testing, the 
emission results would apply to the entire production of the engine 
family, encompassing all of the FEL changes. Since in this case only 
the FEL was changed, and not the design of the engine family, EPA 
believes it is reasonable to apply the emission results of in-use 
testing to all the FEL settings of the engine family.

    \19\ 40 CFR part 86 (subpart A).
---------------------------------------------------------------------------

    Separate calculations of credits would be required whenever an 
engine family contains multiple FELs. Therefore, it would be possible 
for an engine family (with multiple FELs) to both generate and require 
in-use credits as a result of in-use testing. The Agency requests 
comment on its proposed way of handling in-use testing of engine 
families with a number of declared FELs. In particular, EPA seeks 
comment whether it should require a minimum number of in-use tests for 
each FEL (or change in emissions characteristics) and if so, what 
number of tests would be appropriate.
    To provide a safeguard against potential environmental detriment, 
EPA believes that it should only grant in-use credits for amounts of 
emissions reductions in which EPA has a significant degree of 
confidence. Thus, EPA proposes to take into account the uncertainty in 
the in-use emissions tests when calculating credit generation by 
relating credit generation to the statistical accuracy of the tests. 
The ultimate purpose of testing a set of engines in-use is to estimate 
the average emissions rate of all of the engines in that family over 
their useful lives. For many reasons, the results of any one test of 
any one engine will generally constitute a fairly uncertain measure of 
fleet-wide average emission rates; various random factors in the way an 
individual engine is manufactured and used will cause its emission rate 
to deviate from the average of its engine family, and other random 
factors may cause the results of any one test of that engine to vary.
    According to the provisions in the NPRM for in-use testing, a 
manufacturer could pass an in-use audit after the first four tests of 
an audit. However, if another manufacturer failed the first four tests 
in an audit, it would continue testing to ten tests in order to make a 
compliance determination. If both manufacturers simply took the mean of 
the tests associated with each of these audits, the two manufacturers 
would not be generating and using credits for in-use emissions levels 
with the same degree of certainty.
    The Agency believes that the number of credits a manufacturer may 
generate should be related to the number of tests performed for that 
audit, because the more tests that are performed, the more certain EPA 
and the public are that the mean of those test results is near the true 
average for that engine family. In addition, an imbalance of certainty 
exists between credit generation and credit usage. This arises from the 
fact that manufacturers would be able to pass an audit and generate 
credits in four tests (or two for small volume manufacturers), but 
might not fail an audit or be required to use credits until ten tests 
were performed. The average of the ten tests will tend to be closer to 
the correct mean of the engine family. Thus, while EPA will award some 
credits for engine families that appear to be cleaner 

[[Page 4616]]
than their FEL on the basis of four tests, progressively more credits 
will be awarded if the compliance level is based on six, eight, or ten 
tests. (See Figure 1 below.) EPA requests comment on this proposal. In 
particular, EPA requests comment on the appropriateness of the 
magnitude of the relative sample size adjustments presented in Figure 
1, any preferred adjustments, as well as this approach of adjusting the 
credits based on the amount of testing done to determine the CL.

                                Figure 1                                
------------------------------------------------------------------------
                                                                        
------------------------------------------------------------------------
No. of engines tested................   2*,4       6         8        10
Credits earned (AF**)................       .5      .75       .9      1 
------------------------------------------------------------------------
* Small Volume Manufacturer.                                            
** Adjustment factor.                                                   

    If a manufacturer were to pass an in-use audit with the first four 
tests but wanted to participate in the voluntary in-use credit program, 
it would be allowed to test ten engines to maximize the credits it 
would generate.
    EPA is also considering a requirement that if in-use credits are 
generated and used from an engine family with a CL below the applicable 
emission standard or FEL, then the CL would become an enforceable limit 
for the engine family that generated the credits. This condition on the 
generation and use of credits would help ensure that the emissions 
reductions on which they are based would be real, permanent, and 
enforceable. EPA has consistently used such an approach before 
approving the use of credits for emissions reductions. Under such a 
system, an engine family would have to continue to meet the CL on which 
credits were based if it were tested again in-use. If it did not, then 
in-use credits would be required to avoid noncompliance. It is 
important to note that this would not affect or change in any way the 
generation or use of credits during certification.
    The Agency proposes to require manufacturers to use in-use credits 
at a higher rate than the credits were generated. While EPA believes 
that it is important for manufacturers to have the option of remedying 
in-use problems through in-use averaging rather than through much more 
costly and uncertain recall actions, it would be better to not have in-
use failures at all. EPA wants to maintain a strong incentive for 
manufacturers to produce engines that pass their in-use audits, and an 
incentive to achieve that is to require manufacturers who must use in-
use credits to do so at a greater rate than the credits are generated. 
Therefore, EPA proposes that manufacturers use credits at a rate of 1.2 
to 1. In other words, EPA is proposing that manufacturers offset 120 
percent of the negative credits identified by the in-use testing 
program with positive credits. In this way, the in-use credit program 
will achieve an additional environmental benefit when manufacturers 
underestimate FELs and will provide an added incentive to manufacturers 
to adequately identify expected full useful life emission levels when 
choosing the certification FEL. This would be a penalty for 
underestimating certification FELs. EPA requests comment on the 
appropriateness of this penalty, including whether the penalty should 
be larger, smaller, or not imposed at all.
    This penalty may result in a greater environmental benefit than 
accounted for in the cost-benefit calculation. However, EPA is not 
taking a benefit in that calculation, because it expects engines to 
comply in-use with the certification FEL. In other words, EPA expects 
there will be few engine families that need to use in-use credits.
    The credit calculation formula is as follows: Credits earned per 
engine family=

FEL=the pollutant specific family emission limit for the engine family 
in g/kW-hr.
CL=compliance level of the in-use testing in g/kW-hr.
SALES=the number of engines in the engine family sold in the U.S. 
calculated per the certification rules which are the ``first delivery'' 
concept.
Power=the average power of an engine family in kW. (sales weighted)
AF=adjustment factor for the number of tests conducted
U(t)=use in hours per year at age t, defined as

[GRAPHIC] [TIFF OMMITTED] TP07FE96.000

where

t=age of the engine in years
use=mean use in hours per year, usage rate specific to 
the application; for outboard engines, hours per year = 34.8; for 
personal watercraft, hours per year = 77.3; for sterndrive/inboard 
engines, hours per year = 47.6 life = the mean life in 
years of the engine; life=10 for personal watercraft and 
for outboards
[GRAPHIC] [TIFF OMMITTED] TP07FE96.001

S(t)=the cumulative fraction survived at time t
where life is the mean life in years of the engine; 
life= 10 for personal watercraft; and for outboards

[GRAPHIC] [TIFF OMMITTED] TP07FE96.002

    EPA requests comment on the use of the average power rating of an 
engine family. For certification provisions, EPA proposed to use the 
minimum power rating for engines below the applicable emission standard 
and the maximum power rating for engines above the applicable emission 
standard. EPA requests comments in light of the proposed certification 
requirements on power rating.
    The Agency proposes that results of in-use testing of an engine 
family may apply to similar engine families from other model years, 
provided the engine families had received carry-over certification 
because the emission characteristics of the engine family had not 
changed. Therefore, if a carry-over engine family was tested and the CL 
was below the FEL, the engine family could earn credits for a total of 
up to four model years (the MY of the engine family tested, plus the 
two model years prior and the MY after: ``minus two, plus one''). 
However, if the CL was above the FEL, then the engine family would owe 
credits for a total of up to four model years.
    For example, in the year 2002 the Agency may request testing of a 
manufacturer's MY 2001 engine family, which has received carry-over 
certification from 1998-2002. The manufacturer would conduct the audit. 
In this example, suppose the CL for the engine family were found to be 
below the FEL. Since the emission results of an audit of a carry-over 
engine family can apply to two previous years and one subsequent year 
of the MY of the engine family tested, this engine family would earn 
credits for the model years 1999, 2000, 2001, and 2002. Similarly, if 
the CL was greater than the FEL, it would require credits for those 
same years. Any generated credits would be identified as MY 2001 
credits for recordkeeping purposes.
    The Agency proposes to implement this carry-over by applying test 
results from a given MY engine family to the corresponding engine 
family from other model years that involve carry-over certification for 
a number of reasons. The Agency has limited itself to requiring a 
manufacturer to audit only 

[[Page 4617]]
25 percent of its engine families in any given MY. It would take at 
least four years of in-use auditing to cover all of a manufacturer's 
production. In fact, more than four years might be required, since 
manufacturers are allowed to drop and add engine families as their 
product line changes. Accordingly, the Agency believes it is reasonable 
to apply test results from an audit of an engine family that involves 
carry-over certification to other MY production. For example, a carry-
over engine family that has been produced for eight years may pass an 
in-use audit in year one and fail in year eight. The failure may have 
occurred in years two through seven. It appears reasonable to EPA that 
a manufacturer's liability be limited in such situations because some 
engine families may be produced for many years before they are tested 
in-use. The four year proposal in this SNPRM was chosen as a compromise 
between unlimited MY liability and no liability beyond the specific MY 
that was audited.
    In the administration of the Agency's in-use motor vehicle test 
program, the Agency has had occasion to be persuaded that an in-use 
remedy should not apply to a subclass of a given engine family or to a 
previous MY of a family that involved carry-over certification. The 
manufacturers have generally submitted test results and other 
information to support their cases. The Agency believes that a similar 
approach should apply to the marine in-use credit program. It would 
provide an opportunity for reductions in the amount of credits a 
manufacturer might owe for engine families that have been carried over 
for several years due to the automatic application of the ``minus two, 
plus one'' carry-over certification rule to credit calculations. The 
Agency anticipates using this approach infrequently, but believes it 
should be available due to EPA's experience in the motor vehicle in-use 
testing program.
    The Agency is proposing unlimited life for in-use credits. Because 
in-use credits are generated based on real in-use test results, the 
validity of the credits are not in question. With the concern about 
validity of credits removed, an economic rationale supports unlimited 
life. The banked positive credits represent emission reductions beyond 
the requirements of the regulations, or ``excess credits''. The present 
value concept applies to benefits (e.g., emission reductions) as well 
as cost. In other words, just as a dollar today is worth more than a 
dollar tomorrow, so too an environmental benefit today is worth more 
than a benefit tomorrow. However, EPA is not proposing to adjust upward 
the amount of credits banked to appropriate future value as would be 
required to properly account for present value with each year the 
credit is banked. Therefore, it is actually more beneficial to the 
environment for manufacturers to use the ``excess credits'' banked for 
exceedances in future years, because the banked credits inherently have 
a higher present value. Therefore, using the banked positive credits 
with a higher present value, although they are unadjusted, to offset 
negative credits in a future year yields a net environmental benefit 
because the banked credits have a real value higher than the value of 
the future year's negative credits. In this instance, the net 
environmental benefit is a ``shadow'' benefit insofar as it is 
philosophically valued yet unaccounted for.
    The Agency is proposing that the United States sales figures used 
in the marine certification program for each engine family would also 
apply to this in-use credit program. The Agency sees no need and little 
benefit to conducting two separate analyses of the engine sales in the 
United States.
    In order for EPA to determine manufacturer in-use compliance, the 
manufacturer would be required to submit an end of the MY in-use 
testing report. This report would have to be submitted within 90 days 
of the end of the in-use testing period for a given engine family for 
each MY, or at the same time as the final certification ABT report, 
whichever is later. The end of the MY in-use testing report would 
contain the calculated credits from all the in-use testing conducted by 
the manufacturer for a given MY. Also, within ten days after the end of 
an in-use audit for an engine family, the manufacturer would submit a 
report indicating the test results and the calculated CL for the engine 
family.
    To ensure that the environment would not be adversely affected, EPA 
proposes that manufacturers may not enter into a deficit situation as a 
result of credit trading with other manufacturers. For the same reason, 
manufacturers may not carry over deficits from one MY to another. A 
manufacturer must obtain sufficient credits to meet its needs each MY, 
whether those credits are generated by its own engine families or 
obtained through trading. Trading may occur only after the 
manufacturer's in-use testing for that MY has been completed, and a 
manufacturer may only trade to another manufacturer credits that are in 
the bank at that time.
    The integrity of the proposed marine in-use credit program depends 
on accurate recordkeeping and reporting by manufacturers and effective 
tracking and auditing by EPA. Failure of a manufacturer to maintain the 
required records would result in the certificates for the affected 
engine families being void ab initio. Violations of reporting 
requirements could result in a manufacturer being subject to penalties 
of up to $25,000 per day of violation as authorized by sections 205 and 
213 of the CAA.
    The Agency has prepared a supplementary document, available from 
the docket for this rulemaking, which discusses in-use credit issues in 
more detail. This document includes examples of calculations of credits 
in a variety of situations.
    The proposed regulations include hearing provisions which allow the 
manufacturer to challenge EPA's audit of in-use credit calculations and 
the manner in which those credits were used/generated.

G. Labeling Requirements

    As described in the NPRM, each certified engine must bear a label 
indicating the engine family name and the standard or FEL to which it 
is certified. Any engine imported into the United States in a vessel 
must have an engine which also complies with the labeling requirements.
    The Agency considered proposing in this SNPRM the idea of a system 
of labeling engines (or, possibly, watercraft in the case of SD/I 
applications) that would encourage purchase of the cleanest engines and 
discourage purchase of the highest-polluting engines. Such a system 
could be a marketing tool. For example, the cleanest engines could be 
designated as ``green engines'' or engines which are most 
environmentally friendly. The highest-polluting engines could also be 
designated in such a way as to let the consumer know that there are 
cleaner engines available for purchase. EPA proposed a ``green engine'' 
label in the NPRM. However, EPA did not propose to label engines that 
are dirtier. EPA seeks here to elicit comments on a system which would 
also identify which engines are the dirtier engines. One option would 
be to identify all engines that do not meet the MY 2006 average 
emission standard as a ``dirty engine.''
    EPA does not intend to go forward with such a proposal in this 
rulemaking. Nevertheless, EPA requests comment on the advisability of 
proposing labeling provisions of this type at some later date for use 
in conjunction with educational outreach to consumers. 

[[Page 4618]]


H. Addition of Competition Exclusion and Revised Criteria for National 
Security Exemption for Marine Rule and Other Nonroad Rules

    The Agency is proposing to amend or re-propose certain provisions 
of the existing land-based nonroad CI (>37 kW) rule,\20\ the existing 
nonroad SI (19 kW) rule,\21\ and this proposed marine engine 
rule, in order to make the exclusions and national security exemptions 
(NSEs) in these rules more closely follow EPA's long-standing treatment 
of exclusions and NSEs in the on-highway motor vehicle program.

    \20\ 59 FR 31306 (June 17, 1994); see also 40 CFR Part 89.
    \21\ 60 FR 34582 (July 3, 1995); to be codified at 40 CFR Part 
90.
---------------------------------------------------------------------------

    In the motor vehicle program, the regulations exclude from their 
scope any vehicle that exhibits features which render its use on a 
street or highway unsafe, impractical or highly unlikely, including 
features ordinarily associated with military combat or tactical 
vehicles such as armor and/or permanently affixed weaponry. 40 CFR 
85.1703. This exclusion criterion is grounded in the definition of 
``motor vehicle'' in the CAA, which restricts the term to vehicles that 
are designed for transporting persons or property on a street or 
highway. See Section 216(2) of the CAA.
    The statutory definition of ``nonroad engine'' provides no 
comparable basis for a combat exclusion. See Section 216(10). However, 
EPA believes that the national security exemption set forth in Section 
203(b)(1) of the Act allows EPA to grant a regulatory exemption to 
nonroad engines that exhibit ``combat features.'' There are many 
potential uses of nonroad engines in military and national defenses 
settings. Accordingly, the Agency proposes to include an automatic 
national security exemption for nonroad engines, nonroad vehicles, and 
nonroad equipment that exhibit combat features in the two existing 
nonroad rules (for CI engines greater than 37 kW and SI engines less 
than or equal to 19 kW), and in the marine engine rule. All nonroad 
engines vehicles, and equipment within the scope of the regulations 
which exhibit the combat features described in the regulations would 
automatically enjoy an NSE; manufacturers of such products would not be 
required to apply for this exemption.
    While the statutory basis for the automatic nonroad national 
security exemption differs from the statutory basis for the motor 
vehicle combat exclusion, the end result is substantially the same. EPA 
believes that establishing an automatic NSE for the nonroad programs 
accords with Congressional intent to offer a national security 
exemption and decreases significantly the burden for manufacturers and 
EPA that would exist if EPA limited the availability of an NSE to those 
manufacturers who apply to EPA and receive approval, as occurs in the 
motor vehicle program. See 40 CFR 85.1708.
    EPA also proposes that manufacturers who produce a nonroad engine, 
nonroad vehicle, or nonroad equipment which does not meet the ``combat 
features'' criterion, but may otherwise require an NSE, may apply to 
the Agency for an NSE in a manner similar to the national security 
exemption process offered in the motor vehicle program. See 40 CFR 
85.1703. (A slightly different version of the proposed regulatory text 
on this issue already appears in Parts 89 and 90.) Additionally, the 
Agency proposes to promulgate a requirement that EPA maintain a 
publicly available list of NSEs granted to nonroad engines, vehicles, 
and equipment by EPA after manufacturer application.
    Finally, EPA proposes to add a general competition exclusion to the 
marine rule; the NPRM had limited the competition exclusion to imported 
vessels. EPA believes this revised proposal accords with the CAA's 
definition of nonroad engine, which excludes nonroad engines used in a 
vehicle that is used solely for competition. See Section 216(2) of the 
Act.

I. Engine Family Definition

    The Agency proposed an engine family definition in the NPRM that 
allowed the manufacturers flexibility to further segregate engine 
families beyond the proposed criteria, but did not allow manufacturers 
the flexibility to consolidate engine families. Comments in response to 
the NPRM indicated that it would be appropriate to include flexibility 
allowing manufacturers to consolidate engine families.
    It is acceptable to consolidate engine families, particularly SD/I 
engine families, beyond the criteria proposed in the NPRM. For 
instance, SD/I engines may be marinized by different manufacturers yet 
have the same basic engine block produced by, for example, General 
Motors. The emission characteristics should be similar across most 
marinized engines with the same engine block, even if produced by other 
manufacturers. Generally, EPA would not expect the emission 
characteristics to be similar in the degree to which EPA expects on-
highway engine families to be similar. The degree of emission control 
that is necessary for on-highway applications requires that the concept 
of ``similar'' emission characteristics be more narrowly defined. For 
these reasons, EPA is proposing that engines differing in one or more 
of the characteristics proposed to define engine families (i.e., 
combustion cycle, cooling mechanism, cylinder configuration, number of 
cylinders, catalytic converter, thermal reactor characteristics) may be 
grouped in the same engine family if the manufacturer can show that the 
in-use emission characteristics are expected to be similar.

J. Harmonization With the International Maritime Organization

    As stated in the NPRM, EPA requests comment on harmonization with 
the International Maritime Organization (IMO) proposal to regulate 
emissions from new oceangoing vessels. A copy of this IMO proposal is 
located in the docket. EPA intends on harmonizing with the IMO emission 
standard levels for compression-ignition marine engines. EPA requests 
comment on specific ways to harmonize. EPA's NPRM proposed an average 
NOX emission standard of 9.2 g/kW-hr, while the IMO NOX 
emission standard varies from 9.8 g/kW-hr to 17.0 g/kW-hr, depending on 
engine speed. EPA's proposed NOX emission standard is an average 
in which the engine can be either below or above, so long as the 
emissions above the standard are compensated with emissions below the 
standard. On the other hand, the IMO NOX emission standard is a 
cap type standard that all engines must be less than.
    Although EPA is not prepared to re-propose a different NOX 
emission standard, there are several alternatives that seem to exist 
that would result in a harmonized NOX emission standard structure 
with IMO. One alternative would be to adopt the IMO NOX emission 
standard instead of the standard proposed in the NPRM. This would 
result in a cap type standard at the same NOX levels as the IMO 
NOX emission standard across the engine speed range. A second 
alternative would be to retain the proposed average NOX emission 
standard of 9.2 g/kW-hr and to also adopt the IMO emission standards 
across the engine speed range as a cap which no engine could exceed. In 
this way, clean engines would be encouraged through the market for 
emission credits. Third, it may be appropriate to determine an engine 
speed or engine power output cutoff point. Such a point could be used 
to apply the IMO cap emission standard to all engines of high 
horsepower and low 

[[Page 4619]]
and medium speeds. On the other hand, high speed engines with lower 
horsepower could meet the 9.2 g/kW-hr average standard proposed with 
the 9.8 g/kW-hr IMO level as a cap which no engine could exceed. This 
may be appropriate to encourage clean technology and because the high 
speed engines are used in other nonroad applications in addition to 
marine. Finally, EPA must determine whether and how to harmonize each 
of the emission standards for HC, CO, PM and smoke set forth in the 
NPRM with IMO's NOX-only emission control approach. With respect 
to each of these standards, EPA could retain the standard as proposed 
in the NPRM, drop it, or alter it in some way.
    EPA requests comment on ways to harmonize with the IMO emission 
standards, including the alternatives mentioned here and any 
alternatives that commenters can devise to integrate the standards. EPA 
thinks that harmonization is an important issue and intends on 
finalizing a harmonized NOX emission standard. EPA requests 
comment on the extent to which it is appropriate for EPA to harmonize 
the enforcement requirements in its final rule with the enforcement 
scheme proposed in the IMO regulation. For example, EPA may finalize 
its rule such that to the extent that ship owners are liable for engine 
emissions under the IMO's finalized Marpol Annex, EPA may exercise its 
discretion under the CAA to not hold engine manufacturers liable for 
the same emissions. Similarly, EPA would expect to revise its 
regulations to the extent necessary to harmonize the enforcement scheme 
with that of the IMO's finalized Marpol Annex However, EPA is concerned 
about the potential for a regulatory gap between the time EPA's 
regulation is implemented and the time when IMO's Marpol Annex would be 
implemented. EPA is considering applying harmonized or integrated 
emission standards until IMO's Marpol Annex is finalized so that EPA's 
regulation achieves emission reductions according to the schedule 
proposed in the NPRM (i.e., implementation of emission standards 
beginning in MY 1999).
    Finally, EPA is considering whether its test procedures proposed in 
the NPRM are appropriate for CI engines above 1500 kW. EPA's 
requirements are for test bed testing only, where as the IMO's Marpol 
Annex includes an option for testing such engines on-board vessels. EPA 
requests comment as to whether EPA test procedures are or should be 
harmonized with IMO test procedures, including details regarding any 
changes that are needed to bring EPA's procedures in harmony with the 
proposed IMO procedures.

V. Public Participation

A. Comments and the Public Docket

    The Agency welcomes comments on all aspects of this SNPRM. While 
EPA is not publishing the proposed regulatory language, EPA welcomes 
comment on it. The proposed regulatory language can be found in the 
docket, or can be requested from EPA on a floppy disk, or can be 
retrieved from the TTN (see information in section I. of this 
preamble). Commenters are especially encouraged to give suggestions for 
changing any aspects of the proposal that they find objectionable. 
Comments are also encouraged to identify those aspects of the proposal 
that they favor, since EPA may finalize some, but not all, of the 
proposals contained in this Notice. Also, commenters are encouraged to 
offer additional comments on the proposals contained in the NPRM should 
the proposals set forth in this SNPRM affect their views of the NPRM 
proposals. All comments, with the exception of proprietary information, 
should be directed to the EPA Air Docket Section, Docket No. A-92-28 
(see ADDRESSES).
    Commenters who wish to submit proprietary information for 
consideration should clearly separate such information from other 
comments by (1) labeling proprietary information ``Confidential 
Business Information'' and (2) sending proprietary information directly 
to the contact person listed (see FOR FURTHER INFORMATION CONTACT) and 
not to the public docket. This will help insure that proprietary 
information is not inadvertently placed in the docket. If a commenter 
wants EPA to use a submission labeled as confidential information as 
part of the basis for the final rule, then a nonconfidential version of 
the document that 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 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, it will be made available to 
the public without further notice to the commenter.

B. Public Hearing

    As noted above (see DATES), EPA will hold a public hearing on this 
SNPRM on February 22, 1996, if EPA receives from any party a request to 
testify at the hearing. Any person desiring to present testimony at the 
public hearing must notify the contact person listed above of such 
intent no later than February 20, 1996. The contact person should also 
be given an estimate of the time required for the presentation of the 
testimony and notification of any need for audio/visual equipment. 
Testimony will be scheduled on a first come, first served basis. A 
sign-up sheet also will be available at the registration table the 
morning of the hearing for scheduling testimony.
    The Agency suggests that approximately 50 copies of any 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 five days before the scheduled hearing date, in order 
to give EPA staff adequate time to review such material before the 
hearing. Advance copies should be submitted to the contact person 
listed.
    If a hearing is held, the official record of the hearing will be 
kept open for 30 days following the hearing to allow submission of 
rebuttal and supplementary testimony. All such submittals should be 
directed to the Air Docket, Docket No. A-92-28 (see ADDRESSES).
    The hearing will be conducted informally, and technical rules of 
evidence will not apply. A written transcript of the hearing will be 
placed in the above docket for review. Anyone desiring to purchase a 
copy of the transcript should make individual arrangements with the 
court reporter recording the proceeding.

VI. Administrative Requirements

A. Reporting and Recordkeeping Requirements

    The information collection requirements in the NPRM were submitted 
for approval to the OMB under the Paperwork Reduction Act, 44 U.S.C. 
3501 et seq. EPA prepared eight Information Collection Request (ICR) 
documents for the NPRM. Copies of the ICR documents may be obtained 
from Sandy Farmer, Information Policy Branch, EPA, 401 M St. SW. (mail 
code 2136), Washington, DC 20460 or by calling (202) 260-2740.
    The eight ICR documents that have been prepared are:

------------------------------------------------------------------------
       EPA ICR document No.                  Type of information        
------------------------------------------------------------------------
1722.01...........................  Certification/AB&T.                 
282.07............................  Emission Defect Information.        

[[Page 4620]]
                                                                        
1723.01...........................  Importation of Nonconforming        
                                     Engines.                           
1724.01...........................  Selective Enforcement Auditing.     
0012.08...........................  Engine Exclusion Determination.     
0095.07...........................  Precertification and Testing        
                                     Exemption.                         
1725.01...........................  Manufacturers' Assembly Line        
                                     Testing.                           
1726.01...........................  Manufacturers' In-use Testing.      
------------------------------------------------------------------------



    Each ICR document estimates the public reporting, recordkeeping, 
and testing burden for collecting the specified information, including 
time for reviewing instructions, searching existing data sources, 
gathering and maintaining the data needed, and completing the 
collection of information. In the NPRM, the Agency estimated that the 
public burden for the collection of information for all the ICRs would 
average approximately 6,050 hours annually for a typical engine 
manufacturer. The hours spent by a manufacturer for information 
collection activities in any given year would be highly dependent upon 
manufacturer specific variables, such as the number of engine families, 
production changes, emissions defects, etc.
    OMB originally denied all the ICRs that EPA submitted with the 
NPRM. OMB has subsequently approved two of these (1723.01, for 
Importation of Nonconforming Engines and 0012.08, for Engine Exclusion 
Determination), but the rest have not been approved as of the date of 
publication of this SNPRM. Without OMB approval of these information 
collection requests, EPA cannot implement the regulations once 
finalized. Therefore, EPA submitted new information collection requests 
in conjunction with this SNPRM that indicate that the reporting and 
recordkeeping requirements of the proposal as a whole are significantly 
less than estimated in the NPRM due to the small manufacturer criteria 
and provisions, the manufacturer production line testing program, the 
in-use credit program, the significantly reduced administrative 
programs for SD/I engines, and other proposals set forth in this SNPRM.
    The new estimates are also based on additional information 
indicating that the rule affects more manufacturers, and potentially a 
larger number of small manufacturers. This new information prompted EPA 
to reduce administrative program burdens as much as possible. EPA now 
estimates that the public burden for the collection of information for 
all ICRs under the proposed rule as a whole would average approximately 
4,200 hours annually for a typical engine manufacturer.
    Comments regarding the burden estimate or any other aspect of this 
collection of information, including suggestions for reducing this 
burden should be sent to Chief, Information Policy Branch, EPA, 401 M 
St., SW. (mail code 2136), 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 respond to any OMB or public comments on the 
information collection requirements contained in this SNPRM and the 
NPRM.

B. Impact on Small Entities

    The Regulatory Flexibility Act of 1980 requires federal agencies to 
identify potentially adverse impacts of federal regulations upon small 
entities. In instances where significant impacts are possible on a 
substantial number of these entities, agencies are required to perform 
a Regulatory Flexibility Analysis (RFA). The RFA explores options for 
minimizing those impacts.
    As mentioned in the NPRM, EPA considered, but rejected, the notion 
of exempting small manufacturers from enforcement programs or from the 
regulation entirely. A more proportionate sharing of cost burden was 
deemed appropriate. The pollution emitted by each of these engines not 
only contributes to ambient air quality problems but also has health 
impacts on the user of the engine who is in close proximity to the 
exhaust emissions.
    However, as stated in the NPRM, EPA has recently adopted a new 
approach to regulatory flexibility: 22 for purposes of EPA's 
implementation of the Act, any impact is a significant impact, and any 
number of small entities is a substantial number. Thus, EPA will 
consider regulatory options for every regulation subject to the Act 
that can reasonably be expected to have an impact on small entities. In 
light of this new approach, EPA has determined that, if no provisions 
were established to take economic effects into account, this rule would 
be likely to have a significant effect on a substantial number of small 
entities. As a result, in addition to the flexibility inherent in 
averaging, trading, and banking of emissions, EPA has tailored this 
rule to minimize the cost burdens imposed on smaller engine 
manufacturers.

    \22\ Habicht, F. Henry II, Deputy Administrator, Internal EPA 
Memorandum, ``Revised Guidelines for Implementing the Regulatory 
Flexibility Act,'' April 9, 1992.
---------------------------------------------------------------------------

    The Agency performed an RFA in conjunction with the NPRM.23 
Subsequent comments on the NPRM indicated that EPA's proposal would 
indeed adversely impact small manufacturers while providing little 
environmental benefit. Specifically, many small manufacturers of SD/I 
gasoline engines and marinized CI engines came forward to inform EPA of 
the severe impacts the proposed regulations would have on their 
businesses.

    \23\ 59 FR 55930 (November 9, 1994).
---------------------------------------------------------------------------

    In this SNPRM, EPA proposes small manufacturer exemptions and 
flexibility provisions, so as to ensure that this rulemaking does not 
unduly burden small manufacturers. The Agency is supplementing the RFA 
to reflect these proposals. EPA requests comment as to whether the 
proposed small manufacturer exemptions and provisions adequately 
address the needs of affected manufacturers.

C. Executive Order 12866

    Under Executive Order 12866,24 the Agency must determine 
whether the regulatory action is ``significant'' and therefore subject 
to review by the Office of Management and Budget (OMB) 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:

    \24\ 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;
    (3) Materially alter the budgetary impact of entitlements, 
grants, user fees, or loan programs or the rights and obligations of 
recipients thereof; or
    (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.

    Pursuant to the terms of Executive Order 12866, the Agency has 
determined that the NPRM, which this notice supplements, is a 
``significant regulatory action'' because it may adversely affect in a 
material way that sector of the economy involved with the production of 
marine engines. As such, this action was submitted to OMB for review. 
Changes made in response to OMB suggestions or recommendations will be 
documented in the public record. 

[[Page 4621]]


D. Unfunded Mandates Reform Act of 1995

    Under Section 202 of the Unfunded Mandates Reform Act of 1995 
(``UMRA''), Public Law 104-4, EPA must prepare a budgetary impact 
statement to accompany any general notice of proposed rulemaking or 
final rule that includes a Federal mandate which may result in 
estimated costs to State, local, or tribal governments in the 
aggregate, or to the private sector, of $100 million or more. Under 
Section 205, for any rule subject to Section 202 EPA generally must 
select the least costly, most cost-effective, or least burdensome 
alternative that achieves the objectives of the rule and is consistent 
with statutory requirements. Under Section 203, before establishing any 
regulatory requirements that may significantly or uniquely affect small 
governments, EPA must take steps to inform and advise small governments 
of the requirements and enable them to provide input.
    EPA has determined that today's supplemental proposal regarding 
marine engines and proposed revisions to Parts 89 and 90 of the CFR do 
not trigger the requirements of UMRA. EPA expects to prepare a 
budgetary impact statement in compliance with Section 202 of the UMRA, 
and to follow the requirements of Section 205 of the UMRA, at the time 
it issues a final rule on marine engines.

List of Subjects

40 CFR Part 89

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

40 CFR Part 90

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

40 CFR Part 91

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

    Dated: January 26, 1996.
Carol M. Browner,
Administrator.
[FR Doc. 96-2230 Filed 2-6-96; 8:45 am]
BILLING CODE 6560-50-P